Mercurial > hg > CbC > CbC_gcc
annotate NEWS @ 48:9907f3135723
update CbC on GCC from 4.4.2 to 4.4.3.
| author | kent <kent@cr.ie.u-ryukyu.ac.jp> |
|---|---|
| date | Sun, 07 Feb 2010 17:48:31 +0900 (2010-02-07) |
| parents | 3bfb6c00c1e0 |
| children |
| rev | line source |
|---|---|
| 0 | 1 This file contains information about GCC releases which has been generated |
| 2 automatically from the online release notes. It covers releases of GCC | |
| 3 (and the former EGCS project) since EGCS 1.0, on the line of development | |
| 4 that led to GCC 3. For information on GCC 2.8.1 and older releases of GCC 2, | |
| 5 see ONEWS. | |
| 6 | |
| 7 ====================================================================== | |
| 8 http://gcc.gnu.org/gcc-4.4/index.html | |
| 9 GCC 4.4 Release Series | |
| 10 | |
|
47
3bfb6c00c1e0
update it from 4.4.2 to 4.4.3.
kent <kent@cr.ie.u-ryukyu.ac.jp>
parents:
19
diff
changeset
|
11 October 15, 2009 |
| 0 | 12 |
| 13 The [1]GNU project and the GCC developers are pleased to announce the | |
|
47
3bfb6c00c1e0
update it from 4.4.2 to 4.4.3.
kent <kent@cr.ie.u-ryukyu.ac.jp>
parents:
19
diff
changeset
|
14 release of GCC 4.4.3. |
|
19
58ad6c70ea60
update gcc from 4.4.0 to 4.4.1.
kent@firefly.cr.ie.u-ryukyu.ac.jp
parents:
0
diff
changeset
|
15 |
|
58ad6c70ea60
update gcc from 4.4.0 to 4.4.1.
kent@firefly.cr.ie.u-ryukyu.ac.jp
parents:
0
diff
changeset
|
16 This release is a bug-fix release, containing fixes for regressions in |
|
47
3bfb6c00c1e0
update it from 4.4.2 to 4.4.3.
kent <kent@cr.ie.u-ryukyu.ac.jp>
parents:
19
diff
changeset
|
17 GCC 4.4.1 relative to previous releases of GCC. |
| 0 | 18 |
| 19 Release History | |
| 20 | |
|
47
3bfb6c00c1e0
update it from 4.4.2 to 4.4.3.
kent <kent@cr.ie.u-ryukyu.ac.jp>
parents:
19
diff
changeset
|
21 GCC 4.4.3 |
|
3bfb6c00c1e0
update it from 4.4.2 to 4.4.3.
kent <kent@cr.ie.u-ryukyu.ac.jp>
parents:
19
diff
changeset
|
22 January 21, 2010 ([2]changes) |
|
3bfb6c00c1e0
update it from 4.4.2 to 4.4.3.
kent <kent@cr.ie.u-ryukyu.ac.jp>
parents:
19
diff
changeset
|
23 |
|
3bfb6c00c1e0
update it from 4.4.2 to 4.4.3.
kent <kent@cr.ie.u-ryukyu.ac.jp>
parents:
19
diff
changeset
|
24 GCC 4.4.2 |
|
3bfb6c00c1e0
update it from 4.4.2 to 4.4.3.
kent <kent@cr.ie.u-ryukyu.ac.jp>
parents:
19
diff
changeset
|
25 October 15, 2009 ([3]changes) |
|
3bfb6c00c1e0
update it from 4.4.2 to 4.4.3.
kent <kent@cr.ie.u-ryukyu.ac.jp>
parents:
19
diff
changeset
|
26 |
|
19
58ad6c70ea60
update gcc from 4.4.0 to 4.4.1.
kent@firefly.cr.ie.u-ryukyu.ac.jp
parents:
0
diff
changeset
|
27 GCC 4.4.1 |
|
47
3bfb6c00c1e0
update it from 4.4.2 to 4.4.3.
kent <kent@cr.ie.u-ryukyu.ac.jp>
parents:
19
diff
changeset
|
28 July 22, 2009 ([4]changes) |
|
19
58ad6c70ea60
update gcc from 4.4.0 to 4.4.1.
kent@firefly.cr.ie.u-ryukyu.ac.jp
parents:
0
diff
changeset
|
29 |
| 0 | 30 GCC 4.4.0 |
|
47
3bfb6c00c1e0
update it from 4.4.2 to 4.4.3.
kent <kent@cr.ie.u-ryukyu.ac.jp>
parents:
19
diff
changeset
|
31 April 21, 2009 ([5]changes) |
| 0 | 32 |
| 33 References and Acknowledgements | |
| 34 | |
| 35 GCC used to stand for the GNU C Compiler, but since the compiler | |
| 36 supports several other languages aside from C, it now stands for the | |
| 37 GNU Compiler Collection. | |
| 38 | |
|
47
3bfb6c00c1e0
update it from 4.4.2 to 4.4.3.
kent <kent@cr.ie.u-ryukyu.ac.jp>
parents:
19
diff
changeset
|
39 A list of [6]successful builds is updated as new information becomes |
| 0 | 40 available. |
| 41 | |
| 42 The GCC developers would like to thank the numerous people that have | |
| 43 contributed new features, improvements, bug fixes, and other changes as | |
|
47
3bfb6c00c1e0
update it from 4.4.2 to 4.4.3.
kent <kent@cr.ie.u-ryukyu.ac.jp>
parents:
19
diff
changeset
|
44 well as test results to GCC. This [7]amazing group of volunteers is |
| 0 | 45 what makes GCC successful. |
| 46 | |
|
47
3bfb6c00c1e0
update it from 4.4.2 to 4.4.3.
kent <kent@cr.ie.u-ryukyu.ac.jp>
parents:
19
diff
changeset
|
47 For additional information about GCC please refer to the [8]GCC project |
|
3bfb6c00c1e0
update it from 4.4.2 to 4.4.3.
kent <kent@cr.ie.u-ryukyu.ac.jp>
parents:
19
diff
changeset
|
48 web site or contact the [9]GCC development mailing list. |
|
3bfb6c00c1e0
update it from 4.4.2 to 4.4.3.
kent <kent@cr.ie.u-ryukyu.ac.jp>
parents:
19
diff
changeset
|
49 |
|
3bfb6c00c1e0
update it from 4.4.2 to 4.4.3.
kent <kent@cr.ie.u-ryukyu.ac.jp>
parents:
19
diff
changeset
|
50 To obtain GCC please use [10]our mirror sites or [11]our SVN server. |
|
3bfb6c00c1e0
update it from 4.4.2 to 4.4.3.
kent <kent@cr.ie.u-ryukyu.ac.jp>
parents:
19
diff
changeset
|
51 |
|
3bfb6c00c1e0
update it from 4.4.2 to 4.4.3.
kent <kent@cr.ie.u-ryukyu.ac.jp>
parents:
19
diff
changeset
|
52 Please send FSF & GNU inquiries & questions to [12]gnu@gnu.org. There |
|
3bfb6c00c1e0
update it from 4.4.2 to 4.4.3.
kent <kent@cr.ie.u-ryukyu.ac.jp>
parents:
19
diff
changeset
|
53 are also [13]other ways to contact the FSF. |
|
3bfb6c00c1e0
update it from 4.4.2 to 4.4.3.
kent <kent@cr.ie.u-ryukyu.ac.jp>
parents:
19
diff
changeset
|
54 |
|
3bfb6c00c1e0
update it from 4.4.2 to 4.4.3.
kent <kent@cr.ie.u-ryukyu.ac.jp>
parents:
19
diff
changeset
|
55 These pages are maintained by [14]the GCC team. |
| 0 | 56 |
| 57 | |
| 58 For questions related to the use of GCC, please consult these web | |
|
47
3bfb6c00c1e0
update it from 4.4.2 to 4.4.3.
kent <kent@cr.ie.u-ryukyu.ac.jp>
parents:
19
diff
changeset
|
59 pages and the [15]GCC manuals. If that fails, the |
|
3bfb6c00c1e0
update it from 4.4.2 to 4.4.3.
kent <kent@cr.ie.u-ryukyu.ac.jp>
parents:
19
diff
changeset
|
60 [16]gcc-help@gcc.gnu.org mailing list might help. |
| 0 | 61 Please send comments on these web pages and the development of GCC to |
|
47
3bfb6c00c1e0
update it from 4.4.2 to 4.4.3.
kent <kent@cr.ie.u-ryukyu.ac.jp>
parents:
19
diff
changeset
|
62 our developer mailing list at [17]gcc@gnu.org or [18]gcc@gcc.gnu.org. |
|
3bfb6c00c1e0
update it from 4.4.2 to 4.4.3.
kent <kent@cr.ie.u-ryukyu.ac.jp>
parents:
19
diff
changeset
|
63 All of our lists have [19]public archives. |
| 0 | 64 |
| 65 Copyright (C) Free Software Foundation, Inc., 51 Franklin St, Fifth | |
| 66 Floor, Boston, MA 02110, USA. | |
| 67 | |
| 68 Verbatim copying and distribution of this entire article is permitted | |
| 69 in any medium, provided this notice is preserved. | |
|
47
3bfb6c00c1e0
update it from 4.4.2 to 4.4.3.
kent <kent@cr.ie.u-ryukyu.ac.jp>
parents:
19
diff
changeset
|
70 Last modified 2010-01-21 [20]Valid XHTML 1.0 |
| 0 | 71 |
| 72 References | |
| 73 | |
| 74 1. http://www.gnu.org/ | |
| 75 2. http://gcc.gnu.org/gcc-4.4/changes.html | |
|
19
58ad6c70ea60
update gcc from 4.4.0 to 4.4.1.
kent@firefly.cr.ie.u-ryukyu.ac.jp
parents:
0
diff
changeset
|
76 3. http://gcc.gnu.org/gcc-4.4/changes.html |
|
47
3bfb6c00c1e0
update it from 4.4.2 to 4.4.3.
kent <kent@cr.ie.u-ryukyu.ac.jp>
parents:
19
diff
changeset
|
77 4. http://gcc.gnu.org/gcc-4.4/changes.html |
|
3bfb6c00c1e0
update it from 4.4.2 to 4.4.3.
kent <kent@cr.ie.u-ryukyu.ac.jp>
parents:
19
diff
changeset
|
78 5. http://gcc.gnu.org/gcc-4.4/changes.html |
|
3bfb6c00c1e0
update it from 4.4.2 to 4.4.3.
kent <kent@cr.ie.u-ryukyu.ac.jp>
parents:
19
diff
changeset
|
79 6. http://gcc.gnu.org/gcc-4.4/buildstat.html |
|
3bfb6c00c1e0
update it from 4.4.2 to 4.4.3.
kent <kent@cr.ie.u-ryukyu.ac.jp>
parents:
19
diff
changeset
|
80 7. http://gcc.gnu.org/onlinedocs/gcc/Contributors.html |
|
3bfb6c00c1e0
update it from 4.4.2 to 4.4.3.
kent <kent@cr.ie.u-ryukyu.ac.jp>
parents:
19
diff
changeset
|
81 8. http://gcc.gnu.org/index.html |
|
3bfb6c00c1e0
update it from 4.4.2 to 4.4.3.
kent <kent@cr.ie.u-ryukyu.ac.jp>
parents:
19
diff
changeset
|
82 9. mailto:gcc@gcc.gnu.org |
|
3bfb6c00c1e0
update it from 4.4.2 to 4.4.3.
kent <kent@cr.ie.u-ryukyu.ac.jp>
parents:
19
diff
changeset
|
83 10. http://gcc.gnu.org/mirrors.html |
|
3bfb6c00c1e0
update it from 4.4.2 to 4.4.3.
kent <kent@cr.ie.u-ryukyu.ac.jp>
parents:
19
diff
changeset
|
84 11. http://gcc.gnu.org/svn.html |
|
3bfb6c00c1e0
update it from 4.4.2 to 4.4.3.
kent <kent@cr.ie.u-ryukyu.ac.jp>
parents:
19
diff
changeset
|
85 12. mailto:gnu@gnu.org |
|
3bfb6c00c1e0
update it from 4.4.2 to 4.4.3.
kent <kent@cr.ie.u-ryukyu.ac.jp>
parents:
19
diff
changeset
|
86 13. http://www.gnu.org/home.html#ContactInfo |
|
3bfb6c00c1e0
update it from 4.4.2 to 4.4.3.
kent <kent@cr.ie.u-ryukyu.ac.jp>
parents:
19
diff
changeset
|
87 14. http://gcc.gnu.org/about.html |
|
3bfb6c00c1e0
update it from 4.4.2 to 4.4.3.
kent <kent@cr.ie.u-ryukyu.ac.jp>
parents:
19
diff
changeset
|
88 15. http://gcc.gnu.org/onlinedocs/ |
|
3bfb6c00c1e0
update it from 4.4.2 to 4.4.3.
kent <kent@cr.ie.u-ryukyu.ac.jp>
parents:
19
diff
changeset
|
89 16. mailto:gcc-help@gcc.gnu.org |
|
3bfb6c00c1e0
update it from 4.4.2 to 4.4.3.
kent <kent@cr.ie.u-ryukyu.ac.jp>
parents:
19
diff
changeset
|
90 17. mailto:gcc@gnu.org |
|
3bfb6c00c1e0
update it from 4.4.2 to 4.4.3.
kent <kent@cr.ie.u-ryukyu.ac.jp>
parents:
19
diff
changeset
|
91 18. mailto:gcc@gcc.gnu.org |
|
3bfb6c00c1e0
update it from 4.4.2 to 4.4.3.
kent <kent@cr.ie.u-ryukyu.ac.jp>
parents:
19
diff
changeset
|
92 19. http://gcc.gnu.org/lists.html |
|
3bfb6c00c1e0
update it from 4.4.2 to 4.4.3.
kent <kent@cr.ie.u-ryukyu.ac.jp>
parents:
19
diff
changeset
|
93 20. http://validator.w3.org/check/referer |
| 0 | 94 ====================================================================== |
| 95 http://gcc.gnu.org/gcc-4.4/changes.html | |
| 96 GCC 4.4 Release Series | |
| 97 Changes, New Features, and Fixes | |
| 98 | |
|
47
3bfb6c00c1e0
update it from 4.4.2 to 4.4.3.
kent <kent@cr.ie.u-ryukyu.ac.jp>
parents:
19
diff
changeset
|
99 The latest release in the 4.4 release series is [1]GCC 4.4.3. |
|
19
58ad6c70ea60
update gcc from 4.4.0 to 4.4.1.
kent@firefly.cr.ie.u-ryukyu.ac.jp
parents:
0
diff
changeset
|
100 |
| 0 | 101 Caveats |
| 102 | |
| 103 * __builtin_stdarg_start has been completely removed from GCC. | |
| 104 Support for <varargs.h> had been deprecated since GCC 4.0. Use | |
| 105 __builtin_va_start as a replacement. | |
| 106 * Some of the errors issued by the C++ front end that could be | |
| 107 downgraded to warnings in previous releases by using -fpermissive | |
| 108 are now warnings by default. They can be converted into errors by | |
| 109 using -pedantic-errors. | |
| 110 * Use of the cpp assertion extension will now emit a warning when | |
| 111 -Wdeprecated or -pedantic is used. This extension has been | |
| 112 deprecated for many years, but never warned about. | |
| 113 * Packed bit-fields of type char were not properly bit-packed on many | |
| 114 targets prior to GCC 4.4. On these targets, the fix in GCC 4.4 | |
| 115 causes an ABI change. For example there is no longer a 4-bit | |
| 116 padding between field a and b in this structure: | |
| 117 struct foo | |
| 118 { | |
| 119 char a:4; | |
| 120 char b:8; | |
| 121 } __attribute__ ((packed)); | |
| 122 There is a new warning to help identify fields that are affected: | |
| 123 foo.c:5: note: Offset of packed bit-field 'b' has changed in GCC 4.4 | |
| 124 The warning can be disabled with -Wno-packed-bitfield-compat. | |
| 125 * On ARM EABI targets, the C++ mangling of the va_list type has been | |
| 126 changed to conform to the current revision of the EABI. This does | |
| 127 not affect the libstdc++ library included with GCC. | |
| 128 * The SCOUNT and POS bits of the MIPS DSP control register are now | |
| 129 treated as global. Previous versions of GCC treated these fields as | |
| 130 call-clobbered instead. | |
| 131 * The MIPS port no longer recognizes the h asm constraint. It was | |
| 132 necessary to remove this constraint in order to avoid generating | |
| 133 unpredictable code sequences. | |
| 134 One of the main uses of the h constraint was to extract the high | |
| 135 part of a multiplication on 64-bit targets. For example: | |
| 136 asm ("dmultu\t%1,%2" : "=h" (result) : "r" (x), "r" (y)); | |
| 137 You can now achieve the same effect using 128-bit types: | |
| 138 typedef unsigned int uint128_t __attribute__((mode(TI))); | |
| 139 result = ((uint128_t) x * y) >> 64; | |
| 140 The second sequence is better in many ways. For example, if x and y | |
| 141 are constants, the compiler can perform the multiplication at | |
| 142 compile time. If x and y are not constants, the compiler can | |
| 143 schedule the runtime multiplication better than it can schedule an | |
| 144 asm statement. | |
| 145 * Support for a number of older systems and recently unmaintained or | |
| 146 untested target ports of GCC has been declared obsolete in GCC 4.4. | |
| 147 Unless there is activity to revive them, the next release of GCC | |
| 148 will have their sources permanently removed. | |
| 149 The following ports for individual systems on particular | |
| 150 architectures have been obsoleted: | |
| 151 + Generic a.out on IA32 and m68k (i[34567]86-*-aout*, | |
| 152 m68k-*-aout*) | |
| 153 + Generic COFF on ARM, H8300, IA32, m68k and SH (arm-*-coff*, | |
| 154 armel-*-coff*, h8300-*-*, i[34567]86-*-coff*, m68k-*-coff*, | |
| 155 sh-*-*). This does not affect other more specific targets | |
| 156 using the COFF object format on those architectures, or the | |
| 157 more specific H8300 and SH targets (h8300-*-rtems*, | |
| 158 h8300-*-elf*, sh-*-elf*, sh-*-symbianelf*, sh-*-linux*, | |
| 159 sh-*-netbsdelf*, sh-*-rtems*, sh-wrs-vxworks). | |
| 160 + 2BSD on PDP-11 (pdp11-*-bsd) | |
| 161 + AIX 4.1 and 4.2 on PowerPC (rs6000-ibm-aix4.[12]*, | |
| 162 powerpc-ibm-aix4.[12]*) | |
| 163 + Tuning support for Itanium1 (Merced) variants. Note that code | |
| 164 tuned for Itanium2 should also run correctly on Itanium1. | |
| 165 * The protoize and unprotoize utilities have been obsoleted and will | |
| 166 be removed in GCC 4.5. These utilities have not been installed by | |
| 167 default since GCC 3.0. | |
|
19
58ad6c70ea60
update gcc from 4.4.0 to 4.4.1.
kent@firefly.cr.ie.u-ryukyu.ac.jp
parents:
0
diff
changeset
|
168 * Support has been removed for all the [2]configurations obsoleted in |
| 0 | 169 GCC 4.3. |
| 170 * Unknown -Wno-* options are now silently ignored by GCC if no other | |
| 171 diagnostics are issued. If other diagnostics are issued, then GCC | |
| 172 warns about the unknown options. | |
| 173 * More information on porting to GCC 4.4 from previous versions of | |
|
19
58ad6c70ea60
update gcc from 4.4.0 to 4.4.1.
kent@firefly.cr.ie.u-ryukyu.ac.jp
parents:
0
diff
changeset
|
174 GCC can be found in the [3]porting guide for this release. |
| 0 | 175 |
| 176 General Optimizer Improvements | |
| 177 | |
| 178 * A new command-line switch -findirect-inlining has been added. When | |
| 179 turned on it allows the inliner to also inline indirect calls that | |
| 180 are discovered to have known targets at compile time thanks to | |
| 181 previous inlining. | |
| 182 * A new command-line switch -ftree-switch-conversion has been added. | |
| 183 This new pass turns simple initializations of scalar variables in | |
| 184 switch statements into initializations from a static array, given | |
| 185 that all the values are known at compile time and the ratio between | |
| 186 the new array size and the original switch branches does not exceed | |
| 187 the parameter --param switch-conversion-max-branch-ratio (default | |
| 188 is eight). | |
| 189 * A new command-line switch -ftree-builtin-call-dce has been added. | |
| 190 This optimization eliminates unnecessary calls to certain builtin | |
| 191 functions when the return value is not used, in cases where the | |
| 192 calls can not be eliminated entirely because the function may set | |
| 193 errno. This optimization is on by default at -O2 and above. | |
| 194 * A new command-line switch -fconserve-stack directs the compiler to | |
| 195 minimize stack usage even if it makes the generated code slower. | |
| 196 This affects inlining decisions. | |
| 197 * When the assembler supports it, the compiler will now emit unwind | |
| 198 information using assembler .cfi directives. This makes it possible | |
| 199 to use such directives in inline assembler code. The new option | |
| 200 -fno-dwarf2-cfi-asm directs the compiler to not use .cfi | |
| 201 directives. | |
|
19
58ad6c70ea60
update gcc from 4.4.0 to 4.4.1.
kent@firefly.cr.ie.u-ryukyu.ac.jp
parents:
0
diff
changeset
|
202 * The [4]Graphite branch has been merged. This merge has brought in a |
| 0 | 203 new framework for loop optimizations based on a polyhedral |
| 204 intermediate representation. These optimizations apply to all the | |
| 205 languages supported by GCC. The following new code transformations | |
| 206 are available in GCC 4.4: | |
| 207 + -floop-interchange performs loop interchange transformations | |
| 208 on loops. Interchanging two nested loops switches the inner | |
| 209 and outer loops. For example, given a loop like: | |
| 210 DO J = 1, M | |
| 211 DO I = 1, N | |
| 212 A(J, I) = A(J, I) * C | |
| 213 ENDDO | |
| 214 ENDDO | |
| 215 | |
| 216 loop interchange will transform the loop as if the user had | |
| 217 written: | |
| 218 DO I = 1, N | |
| 219 DO J = 1, M | |
| 220 A(J, I) = A(J, I) * C | |
| 221 ENDDO | |
| 222 ENDDO | |
| 223 | |
| 224 which can be beneficial when N is larger than the caches, | |
| 225 because in Fortran, the elements of an array are stored in | |
| 226 memory contiguously by column, and the original loop iterates | |
| 227 over rows, potentially creating at each access a cache miss. | |
| 228 + -floop-strip-mine performs loop strip mining transformations | |
| 229 on loops. Strip mining splits a loop into two nested loops. | |
| 230 The outer loop has strides equal to the strip size and the | |
| 231 inner loop has strides of the original loop within a strip. | |
| 232 For example, given a loop like: | |
| 233 DO I = 1, N | |
| 234 A(I) = A(I) + C | |
| 235 ENDDO | |
| 236 | |
| 237 loop strip mining will transform the loop as if the user had | |
| 238 written: | |
| 239 DO II = 1, N, 4 | |
| 240 DO I = II, min (II + 3, N) | |
| 241 A(I) = A(I) + C | |
| 242 ENDDO | |
| 243 ENDDO | |
| 244 | |
| 245 + -floop-block performs loop blocking transformations on loops. | |
| 246 Blocking strip mines each loop in the loop nest such that the | |
| 247 memory accesses of the element loops fit inside caches. For | |
| 248 example, given a loop like: | |
| 249 DO I = 1, N | |
| 250 DO J = 1, M | |
| 251 A(J, I) = B(I) + C(J) | |
| 252 ENDDO | |
| 253 ENDDO | |
| 254 | |
| 255 loop blocking will transform the loop as if the user had | |
| 256 written: | |
| 257 DO II = 1, N, 64 | |
| 258 DO JJ = 1, M, 64 | |
| 259 DO I = II, min (II + 63, N) | |
| 260 DO J = JJ, min (JJ + 63, M) | |
| 261 A(J, I) = B(I) + C(J) | |
| 262 ENDDO | |
| 263 ENDDO | |
| 264 ENDDO | |
| 265 ENDDO | |
| 266 | |
| 267 which can be beneficial when M is larger than the caches, | |
| 268 because the innermost loop will iterate over a smaller amount | |
| 269 of data that can be kept in the caches. | |
| 270 * A new register allocator has replaced the old one. It is called | |
| 271 integrated register allocator (IRA) because coalescing, register | |
| 272 live range splitting, and hard register preferencing are done | |
| 273 on-the-fly during coloring. It also has better integration with the | |
| 274 reload pass. IRA is a regional register allocator which uses modern | |
| 275 Chaitin-Briggs coloring instead of Chow's priority coloring used in | |
| 276 the old register allocator. More info about IRA internals and | |
| 277 options can be found in the GCC manuals. | |
| 278 * A new instruction scheduler and software pipeliner, based on the | |
| 279 selective scheduling approach, has been added. The new pass | |
| 280 performs instruction unification, register renaming, substitution | |
| 281 through register copies, and speculation during scheduling. The | |
| 282 software pipeliner is able to pipeline non-countable loops. The new | |
| 283 pass is targeted at scheduling-eager in-order platforms. In GCC 4.4 | |
| 284 it is available for the Intel Itanium platform working by default | |
| 285 as the second scheduling pass (after register allocation) at the | |
| 286 -O3 optimization level. | |
| 287 * When using -fprofile-generate with a multi-threaded program, the | |
| 288 profile counts may be slightly wrong due to race conditions. The | |
| 289 new -fprofile-correction option directs the compiler to apply | |
| 290 heuristics to smooth out the inconsistencies. By default the | |
| 291 compiler will give an error message when it finds an inconsistent | |
| 292 profile. | |
| 293 * The new -fprofile-dir=PATH option permits setting the directory | |
| 294 where profile data files are stored when using -fprofile-generate | |
| 295 and friends, and the directory used when reading profile data files | |
| 296 using -fprofile-use and friends. | |
| 297 | |
| 298 New warning options | |
| 299 | |
| 300 * The new -Wframe-larger-than=NUMBER option directs GCC to emit a | |
| 301 warning if any stack frame is larger than NUMBER bytes. This may be | |
| 302 used to help ensure that code fits within a limited amount of stack | |
| 303 space. | |
| 304 * The new -Wno-mudflap option disables warnings about constructs | |
| 305 which can not be instrumented when using -fmudflap. | |
| 306 | |
| 307 New Languages and Language specific improvements | |
| 308 | |
|
19
58ad6c70ea60
update gcc from 4.4.0 to 4.4.1.
kent@firefly.cr.ie.u-ryukyu.ac.jp
parents:
0
diff
changeset
|
309 * Version 3.0 of the [5]OpenMP specification is now supported for the |
| 0 | 310 C, C++, and Fortran compilers. |
|
19
58ad6c70ea60
update gcc from 4.4.0 to 4.4.1.
kent@firefly.cr.ie.u-ryukyu.ac.jp
parents:
0
diff
changeset
|
311 * New character data types, per [6]TR 19769: New character types in |
|
58ad6c70ea60
update gcc from 4.4.0 to 4.4.1.
kent@firefly.cr.ie.u-ryukyu.ac.jp
parents:
0
diff
changeset
|
312 C, are now supported for the C compiler in -std=gnu99 mode, as |
|
58ad6c70ea60
update gcc from 4.4.0 to 4.4.1.
kent@firefly.cr.ie.u-ryukyu.ac.jp
parents:
0
diff
changeset
|
313 __CHAR16_TYPE__ and __CHAR32_TYPE__, and for the C++ compiler in |
|
58ad6c70ea60
update gcc from 4.4.0 to 4.4.1.
kent@firefly.cr.ie.u-ryukyu.ac.jp
parents:
0
diff
changeset
|
314 -std=c++0x and -std=gnu++0x modes, as char16_t and char32_t too. |
| 0 | 315 |
| 316 C family | |
| 317 | |
| 318 * A new optimize attribute was added to allow programmers to change | |
| 319 the optimization level and particular optimization options for an | |
| 320 individual function. You can also change the optimization options | |
| 321 via the GCC optimize pragma for functions defined after the pragma. | |
| 322 The GCC push_options pragma and the GCC pop_options pragma allow | |
| 323 you temporarily save and restore the options used. The GCC | |
| 324 reset_options pragma restores the options to what was specified on | |
| 325 the command line. | |
| 326 * Uninitialized warnings do not require enabling optimization | |
| 327 anymore, that is, -Wuninitialized can be used together with -O0. | |
| 328 Nonetheless, the warnings given by -Wuninitialized will probably be | |
| 329 more accurate if optimization is enabled. | |
| 330 * -Wparentheses now warns about expressions such as (!x | y) and (!x | |
| 331 & y). Using explicit parentheses, such as in ((!x) | y), silences | |
| 332 this warning. | |
|
19
58ad6c70ea60
update gcc from 4.4.0 to 4.4.1.
kent@firefly.cr.ie.u-ryukyu.ac.jp
parents:
0
diff
changeset
|
333 * -Wsequence-point now warns within if, while,do while and for |
| 0 | 334 conditions, and within for begin/end expressions. |
| 335 * A new option -dU is available to dump definitions of preprocessor | |
| 336 macros that are tested or expanded. | |
| 337 | |
| 338 C++ | |
| 339 | |
|
19
58ad6c70ea60
update gcc from 4.4.0 to 4.4.1.
kent@firefly.cr.ie.u-ryukyu.ac.jp
parents:
0
diff
changeset
|
340 * [7]Improved experimental support for the upcoming ISO C++ standard, |
| 0 | 341 C++0x. Including support for auto, inline namespaces, generalized |
| 342 initializer lists, defaulted and deleted functions, new character | |
| 343 types, and scoped enums. | |
| 344 * Those errors that may be downgraded to warnings to build legacy | |
| 345 code now mention -fpermissive when -fdiagnostics-show-option is | |
| 346 enabled. | |
| 347 * -Wconversion now warns if the result of a static_cast to enumeral | |
| 348 type is unspecified because the value is outside the range of the | |
| 349 enumeral type. | |
| 350 * -Wuninitialized now warns if a non-static reference or non-static | |
| 351 const member appears in a class without constructors. | |
| 352 * G++ now properly implements value-initialization, so objects with | |
| 353 an initializer of () and an implicitly defined default constructor | |
| 354 will be zero-initialized before the default constructor is called. | |
| 355 | |
| 356 Runtime Library (libstdc++) | |
| 357 | |
|
19
58ad6c70ea60
update gcc from 4.4.0 to 4.4.1.
kent@firefly.cr.ie.u-ryukyu.ac.jp
parents:
0
diff
changeset
|
358 * [8]Improved experimental support for the upcoming ISO C++ standard, |
|
58ad6c70ea60
update gcc from 4.4.0 to 4.4.1.
kent@firefly.cr.ie.u-ryukyu.ac.jp
parents:
0
diff
changeset
|
359 C++0x, including: |
|
58ad6c70ea60
update gcc from 4.4.0 to 4.4.1.
kent@firefly.cr.ie.u-ryukyu.ac.jp
parents:
0
diff
changeset
|
360 + Support for <chrono>, <condition_variable>, <cstdatomic>, |
|
58ad6c70ea60
update gcc from 4.4.0 to 4.4.1.
kent@firefly.cr.ie.u-ryukyu.ac.jp
parents:
0
diff
changeset
|
361 <forward_list>, <initializer_list>, <mutex>, <ratio>, |
|
58ad6c70ea60
update gcc from 4.4.0 to 4.4.1.
kent@firefly.cr.ie.u-ryukyu.ac.jp
parents:
0
diff
changeset
|
362 <system_error>, and <thread>. |
|
58ad6c70ea60
update gcc from 4.4.0 to 4.4.1.
kent@firefly.cr.ie.u-ryukyu.ac.jp
parents:
0
diff
changeset
|
363 + unique_ptr, <algorithm> additions, exception propagation, and |
|
58ad6c70ea60
update gcc from 4.4.0 to 4.4.1.
kent@firefly.cr.ie.u-ryukyu.ac.jp
parents:
0
diff
changeset
|
364 support for the new character types in <string> and <limits>. |
|
58ad6c70ea60
update gcc from 4.4.0 to 4.4.1.
kent@firefly.cr.ie.u-ryukyu.ac.jp
parents:
0
diff
changeset
|
365 + Existing facilities now exploit initializer lists, defaulted |
|
58ad6c70ea60
update gcc from 4.4.0 to 4.4.1.
kent@firefly.cr.ie.u-ryukyu.ac.jp
parents:
0
diff
changeset
|
366 and deleted functions, and the newly implemented core C++0x |
|
58ad6c70ea60
update gcc from 4.4.0 to 4.4.1.
kent@firefly.cr.ie.u-ryukyu.ac.jp
parents:
0
diff
changeset
|
367 features. |
|
58ad6c70ea60
update gcc from 4.4.0 to 4.4.1.
kent@firefly.cr.ie.u-ryukyu.ac.jp
parents:
0
diff
changeset
|
368 + Some standard containers are more efficient together with |
|
58ad6c70ea60
update gcc from 4.4.0 to 4.4.1.
kent@firefly.cr.ie.u-ryukyu.ac.jp
parents:
0
diff
changeset
|
369 stateful allocators, i.e., no allocator is constructed on the |
|
58ad6c70ea60
update gcc from 4.4.0 to 4.4.1.
kent@firefly.cr.ie.u-ryukyu.ac.jp
parents:
0
diff
changeset
|
370 fly at element construction time. |
| 0 | 371 * Experimental support for non-standard pointer types in containers. |
|
19
58ad6c70ea60
update gcc from 4.4.0 to 4.4.1.
kent@firefly.cr.ie.u-ryukyu.ac.jp
parents:
0
diff
changeset
|
372 * The long standing libstdc++/30928 has been fixed for targets |
|
58ad6c70ea60
update gcc from 4.4.0 to 4.4.1.
kent@firefly.cr.ie.u-ryukyu.ac.jp
parents:
0
diff
changeset
|
373 running glibc 2.10 or later. |
|
58ad6c70ea60
update gcc from 4.4.0 to 4.4.1.
kent@firefly.cr.ie.u-ryukyu.ac.jp
parents:
0
diff
changeset
|
374 * As usual, many small and larger bug fixes, in particular quite a |
|
58ad6c70ea60
update gcc from 4.4.0 to 4.4.1.
kent@firefly.cr.ie.u-ryukyu.ac.jp
parents:
0
diff
changeset
|
375 few corner cases in <locale>. |
| 0 | 376 |
| 377 Fortran | |
| 378 | |
| 379 * GNU Fortran now employs libcpp directly instead of using cc1 as an | |
|
19
58ad6c70ea60
update gcc from 4.4.0 to 4.4.1.
kent@firefly.cr.ie.u-ryukyu.ac.jp
parents:
0
diff
changeset
|
380 external preprocessor. The [9]-cpp option was added to allow manual |
| 0 | 381 invocation of the preprocessor without relying on filename |
| 382 extensions. | |
|
19
58ad6c70ea60
update gcc from 4.4.0 to 4.4.1.
kent@firefly.cr.ie.u-ryukyu.ac.jp
parents:
0
diff
changeset
|
383 * The [10]-Warray-temporaries option warns about array temporaries |
| 0 | 384 generated by the compiler, as an aid to optimization. |
|
19
58ad6c70ea60
update gcc from 4.4.0 to 4.4.1.
kent@firefly.cr.ie.u-ryukyu.ac.jp
parents:
0
diff
changeset
|
385 * The [11]-fcheck-array-temporaries option has been added, printing a |
| 0 | 386 notification at run time, when an array temporary had to be created |
| 387 for an function argument. Contrary to -Warray-temporaries the | |
| 388 warning is only printed if the array is noncontiguous. | |
| 389 * Improved generation of DWARF debugging symbols | |
| 390 * If using an intrinsic not part of the selected standard (via -std= | |
| 391 and -fall-intrinsics) gfortran will now treat it as if this | |
| 392 procedure were declared EXTERNAL and try to link to a user-supplied | |
| 393 procedure. -Wintrinsics-std will warn whenever this happens. The | |
| 394 now-useless option -Wnonstd-intrinsic was removed. | |
| 395 * The flag -falign-commons has been added to control the alignment of | |
| 396 variables in COMMON blocks, which is enabled by default in line | |
| 397 with previous GCC version. Using -fno-align-commons one can force | |
| 398 commons to be contiguous in memory as required by the Fortran | |
| 399 standard, however, this slows down the memory access. The option | |
| 400 -Walign-commons, which is enabled by default, warns when padding | |
| 401 bytes were added for alignment. The proper solution is to sort the | |
| 402 common objects by decreasing storage size, which avoids the | |
| 403 alignment problems. | |
| 404 * Fortran 2003 support has been extended: | |
| 405 + Wide characters (ISO 10646, UCS-4, kind=4) and UTF-8 I/O is | |
| 406 now supported (except internal reads from/writes to wide | |
|
19
58ad6c70ea60
update gcc from 4.4.0 to 4.4.1.
kent@firefly.cr.ie.u-ryukyu.ac.jp
parents:
0
diff
changeset
|
407 strings). [12]-fbackslash now supports also \unnnn and |
| 0 | 408 \Unnnnnnnn to enter Unicode characters. |
| 409 + Asynchronous I/O (implemented as synchronous I/O) and the | |
| 410 decimal=, size=, sign=, pad=, blank=, and delim= specifiers | |
| 411 are now supported in I/O statements. | |
| 412 + Support for Fortran 2003 structure constructors and for array | |
| 413 constructor with typespec has been added. | |
| 414 + Procedure Pointers (but not yet as component in derived types | |
| 415 and as function results) are now supported. | |
| 416 + Abstract types, type extension, and type-bound procedures | |
| 417 (both PROCEDURE and GENERIC but not as operators). Note: As | |
| 418 CLASS/polymorphyic types are not implemented, type-bound | |
| 419 procedures with PASS accept as non-standard extension TYPE | |
| 420 arguments. | |
| 421 * Fortran 2008 support has been added: | |
| 422 + The -std=f2008 option and support for the file extensions | |
| 423 .f2008 and .F2008 has been added. | |
| 424 + The g0 format descriptor is now supported. | |
| 425 + The Fortran 2008 mathematical intrinsics ASINH, ACOSH, ATANH, | |
| 426 ERF, ERFC, GAMMA, LOG_GAMMA, BESSEL_*, HYPOT, and ERFC_SCALED | |
| 427 are now available (some of them existed as GNU extension | |
| 428 before). Note: The hyperbolic functions are not yet supporting | |
| 429 complex arguments and the three- argument version of BESSEL_*N | |
| 430 is not available. | |
| 431 + The bit intrinsics LEADZ and TRAILZ have been added. | |
| 432 | |
| 433 Java (GCJ) | |
| 434 | |
| 435 Ada | |
| 436 | |
| 437 * The Ada runtime now supports multilibs on many platforms including | |
| 438 x86_64, SPARC and PowerPC. Their build is enabled by default. | |
| 439 | |
| 440 New Targets and Target Specific Improvements | |
| 441 | |
| 442 ARM | |
| 443 | |
| 444 * GCC now supports optimizing for the Cortex-A9, Cortex-R4 and | |
| 445 Cortex-R4F processors and has many other improvements to | |
| 446 optimization for ARM processors. | |
| 447 * GCC now supports the VFPv3 variant with 16 double-precision | |
| 448 registers with -mfpu=vfpv3-d16. The option -mfpu=vfp3 has been | |
| 449 renamed to -mfpu=vfpv3. | |
| 450 * GCC now supports the -mfix-cortex-m3-ldrd option to work around an | |
| 451 erratum on Cortex-M3 processors. | |
| 452 * GCC now supports the __sync_* atomic operations for ARM EABI | |
| 453 GNU/Linux. | |
| 454 * The section anchors optimization is now enabled by default when | |
| 455 optimizing for ARM. | |
| 456 * GCC now uses a new EABI-compatible profiling interface for EABI | |
| 457 targets. This requires a function __gnu_mcount_nc, which is | |
| 458 provided by GNU libc versions 2.8 and later. | |
| 459 | |
| 460 AVR | |
| 461 | |
| 462 * The -mno-tablejump option has been deprecated because it has the | |
| 463 same effect as the -fno-jump-tables option. | |
| 464 * Added support for these new AVR devices: | |
| 465 + ATA6289 | |
| 466 + ATtiny13A | |
| 467 + ATtiny87 | |
| 468 + ATtiny167 | |
| 469 + ATtiny327 | |
| 470 + ATmega8C1 | |
| 471 + ATmega16C1 | |
| 472 + ATmega32C1 | |
| 473 + ATmega8M1 | |
| 474 + ATmega16M1 | |
| 475 + ATmega32M1 | |
| 476 + ATmega32U4 | |
| 477 + ATmega16HVB | |
| 478 + ATmega4HVD | |
| 479 + ATmega8HVD | |
| 480 + ATmega64C1 | |
| 481 + ATmega64M1 | |
| 482 + ATmega16U4 | |
| 483 + ATmega32U6 | |
| 484 + ATmega128RFA1 | |
| 485 + AT90PWM81 | |
| 486 + AT90SCR100 | |
| 487 + M3000F | |
| 488 + M3000S | |
| 489 + M3001B | |
| 490 | |
| 491 IA-32/x86-64 | |
| 492 | |
| 493 * Support for Intel AES built-in functions and code generation is | |
| 494 available via -maes. | |
| 495 * Support for Intel PCLMUL built-in function and code generation is | |
| 496 available via -mpclmul. | |
| 497 * Support for Intel AVX built-in functions and code generation is | |
| 498 available via -mavx. | |
| 499 * Automatically align the stack for local variables with alignment | |
| 500 requirement. | |
| 501 * GCC can now utilize the SVML library for vectorizing calls to a set | |
| 502 of C99 functions if -mveclibabi=svml is specified and you link to | |
| 503 an SVML ABI compatible library. | |
|
19
58ad6c70ea60
update gcc from 4.4.0 to 4.4.1.
kent@firefly.cr.ie.u-ryukyu.ac.jp
parents:
0
diff
changeset
|
504 * On x86-64, the ABI has been changed in the following cases to |
|
58ad6c70ea60
update gcc from 4.4.0 to 4.4.1.
kent@firefly.cr.ie.u-ryukyu.ac.jp
parents:
0
diff
changeset
|
505 conform to the x86-64 ABI: |
|
58ad6c70ea60
update gcc from 4.4.0 to 4.4.1.
kent@firefly.cr.ie.u-ryukyu.ac.jp
parents:
0
diff
changeset
|
506 + Passing/returning structures with flexible array member: |
|
58ad6c70ea60
update gcc from 4.4.0 to 4.4.1.
kent@firefly.cr.ie.u-ryukyu.ac.jp
parents:
0
diff
changeset
|
507 struct foo |
|
58ad6c70ea60
update gcc from 4.4.0 to 4.4.1.
kent@firefly.cr.ie.u-ryukyu.ac.jp
parents:
0
diff
changeset
|
508 { |
|
58ad6c70ea60
update gcc from 4.4.0 to 4.4.1.
kent@firefly.cr.ie.u-ryukyu.ac.jp
parents:
0
diff
changeset
|
509 int i; |
|
58ad6c70ea60
update gcc from 4.4.0 to 4.4.1.
kent@firefly.cr.ie.u-ryukyu.ac.jp
parents:
0
diff
changeset
|
510 int flex[]; |
|
58ad6c70ea60
update gcc from 4.4.0 to 4.4.1.
kent@firefly.cr.ie.u-ryukyu.ac.jp
parents:
0
diff
changeset
|
511 }; |
|
58ad6c70ea60
update gcc from 4.4.0 to 4.4.1.
kent@firefly.cr.ie.u-ryukyu.ac.jp
parents:
0
diff
changeset
|
512 + Passing/returning structures with complex float member: |
|
58ad6c70ea60
update gcc from 4.4.0 to 4.4.1.
kent@firefly.cr.ie.u-ryukyu.ac.jp
parents:
0
diff
changeset
|
513 struct foo |
|
58ad6c70ea60
update gcc from 4.4.0 to 4.4.1.
kent@firefly.cr.ie.u-ryukyu.ac.jp
parents:
0
diff
changeset
|
514 { |
|
58ad6c70ea60
update gcc from 4.4.0 to 4.4.1.
kent@firefly.cr.ie.u-ryukyu.ac.jp
parents:
0
diff
changeset
|
515 int i; |
|
58ad6c70ea60
update gcc from 4.4.0 to 4.4.1.
kent@firefly.cr.ie.u-ryukyu.ac.jp
parents:
0
diff
changeset
|
516 __complex__ float f; |
|
58ad6c70ea60
update gcc from 4.4.0 to 4.4.1.
kent@firefly.cr.ie.u-ryukyu.ac.jp
parents:
0
diff
changeset
|
517 }; |
|
58ad6c70ea60
update gcc from 4.4.0 to 4.4.1.
kent@firefly.cr.ie.u-ryukyu.ac.jp
parents:
0
diff
changeset
|
518 + Passing/returning unions with long double member: |
|
58ad6c70ea60
update gcc from 4.4.0 to 4.4.1.
kent@firefly.cr.ie.u-ryukyu.ac.jp
parents:
0
diff
changeset
|
519 union foo |
|
58ad6c70ea60
update gcc from 4.4.0 to 4.4.1.
kent@firefly.cr.ie.u-ryukyu.ac.jp
parents:
0
diff
changeset
|
520 { |
|
58ad6c70ea60
update gcc from 4.4.0 to 4.4.1.
kent@firefly.cr.ie.u-ryukyu.ac.jp
parents:
0
diff
changeset
|
521 int x; |
|
58ad6c70ea60
update gcc from 4.4.0 to 4.4.1.
kent@firefly.cr.ie.u-ryukyu.ac.jp
parents:
0
diff
changeset
|
522 long double ld; |
|
58ad6c70ea60
update gcc from 4.4.0 to 4.4.1.
kent@firefly.cr.ie.u-ryukyu.ac.jp
parents:
0
diff
changeset
|
523 }; |
|
58ad6c70ea60
update gcc from 4.4.0 to 4.4.1.
kent@firefly.cr.ie.u-ryukyu.ac.jp
parents:
0
diff
changeset
|
524 Code built with previous versions of GCC that uses any of these is |
|
58ad6c70ea60
update gcc from 4.4.0 to 4.4.1.
kent@firefly.cr.ie.u-ryukyu.ac.jp
parents:
0
diff
changeset
|
525 not compatible with code built with GCC 4.4.0 or later. |
| 0 | 526 * A new target attribute was added to allow programmers to change the |
| 527 target options like -msse2 or -march=k8 for an individual function. | |
| 528 You can also change the target options via the GCC target pragma | |
| 529 for functions defined after the pragma. | |
| 530 * GCC can now be configured with options --with-arch-32, | |
| 531 --with-arch-64, --with-cpu-32, --with-cpu-64, --with-tune-32 and | |
| 532 --with-tune-64 to control the default optimization separately for | |
| 533 32-bit and 64-bit modes. | |
| 534 | |
| 535 IA-32/IA64 | |
| 536 | |
| 537 * Support for __float128 (TFmode) IEEE quad type and corresponding | |
| 538 TCmode IEEE complex quad type is available via the soft-fp library | |
| 539 on IA-32/IA64 targets. This includes basic arithmetic operations | |
| 540 (addition, subtraction, negation, multiplication and division) on | |
| 541 __float128 real and TCmode complex values, the full set of IEEE | |
| 542 comparisons between __float128 values, conversions to and from | |
| 543 float, double and long double floating point types, as well as | |
| 544 conversions to and from signed or unsigned integer, signed or | |
| 545 unsigned long integer and signed or unsigned quad (TImode, IA64 | |
| 546 only) integer types. Additionally, all operations generate the full | |
| 547 set of IEEE exceptions and support the full set of IEEE rounding | |
| 548 modes. | |
| 549 | |
| 550 M68K/ColdFire | |
| 551 | |
| 552 * GCC now supports instruction scheduling for ColdFire V1, V3 and V4 | |
| 553 processors. (Scheduling support for ColdFire V2 processors was | |
| 554 added in GCC 4.3.) | |
| 555 * GCC now supports the -mxgot option to support programs requiring | |
| 556 many GOT entries on ColdFire. | |
| 557 * The m68k-*-linux-gnu target now builds multilibs by default. | |
| 558 | |
| 559 MIPS | |
| 560 | |
| 561 * MIPS Technologies have extended the original MIPS SVR4 ABI to | |
| 562 include support for procedure linkage tables (PLTs) and copy | |
| 563 relocations. These extensions allow GNU/Linux executables to use a | |
| 564 significantly more efficient code model than the one defined by the | |
| 565 original ABI. | |
| 566 GCC support for this code model is available via a new command-line | |
| 567 option, -mplt. There is also a new configure-time option, | |
| 568 --with-mips-plt, to make -mplt the default. | |
| 569 The new code model requires support from the assembler, the linker, | |
| 570 and the runtime C library. This support is available in binutils | |
| 571 2.19 and GLIBC 2.9. | |
| 572 * GCC can now generate MIPS16 code for 32-bit GNU/Linux executables | |
| 573 and 32-bit GNU/Linux shared libraries. This feature requires GNU | |
| 574 binutils 2.19 or above. | |
| 575 * Support for RMI's XLR processor is now available through the | |
| 576 -march=xlr and -mtune=xlr options. | |
| 577 * 64-bit targets can now perform 128-bit multiplications inline, | |
| 578 instead of relying on a libgcc function. | |
| 579 * Native GNU/Linux toolchains now support -march=native and | |
| 580 -mtune=native, which select the host processor. | |
| 581 * GCC now supports the R10K, R12K, R14K and R16K processors. The | |
| 582 canonical -march= and -mtune= names for these processors are | |
| 583 r10000, r12000, r14000 and r16000 respectively. | |
| 584 * GCC can now work around the side effects of speculative execution | |
| 585 on R10K processors. Please see the documentation of the | |
| 586 -mr10k-cache-barrier option for details. | |
| 587 * Support for the MIPS64 Release 2 instruction set has been added. | |
| 588 The option -march=mips64r2 enables generation of these | |
| 589 instructions. | |
| 590 * GCC now supports Cavium Networks' Octeon processor. This support is | |
| 591 available through the -march=octeon and -mtune=octeon options. | |
| 592 * GCC now supports STMicroelectronics' Loongson 2E/2F processors. The | |
| 593 canonical -march= and -mtune= names for these processors are | |
| 594 loongson2e and loongson2f. | |
| 595 | |
| 596 picochip | |
| 597 | |
| 598 Picochip is a 16-bit processor. A typical picoChip contains over 250 | |
| 599 small cores, each with small amounts of memory. There are three | |
| 600 processor variants (STAN, MEM and CTRL) with different instruction sets | |
| 601 and memory configurations and they can be chosen using the -mae option. | |
| 602 | |
| 603 This port is intended to be a "C" only port. | |
| 604 | |
| 605 Power Architecture and PowerPC | |
| 606 | |
| 607 * GCC now supports the e300c2, e300c3 and e500mc processors. | |
| 608 * GCC now supports Xilinx processors with a single-precision FPU. | |
| 609 * Decimal floating point is now supported for e500 processors. | |
| 610 | |
| 611 S/390, zSeries and System z9/z10 | |
| 612 | |
| 613 * Support for the IBM System z10 EC/BC processor has been added. When | |
| 614 using the -march=z10 option, the compiler will generate code making | |
| 615 use of instructions provided by the General-Instruction-Extension | |
| 616 Facility and the Execute-Extension Facility. | |
| 617 | |
| 618 VxWorks | |
| 619 | |
| 620 * GCC now supports the thread-local storage mechanism used on | |
| 621 VxWorks. | |
| 622 | |
| 623 Xtensa | |
| 624 | |
| 625 * GCC now supports thread-local storage (TLS) for Xtensa processor | |
| 626 configurations that include the Thread Pointer option. TLS also | |
| 627 requires support from the assembler and linker; this support is | |
| 628 provided in the GNU binutils beginning with version 2.19. | |
| 629 | |
| 630 Documentation improvements | |
| 631 | |
| 632 Other significant improvements | |
| 633 | |
|
19
58ad6c70ea60
update gcc from 4.4.0 to 4.4.1.
kent@firefly.cr.ie.u-ryukyu.ac.jp
parents:
0
diff
changeset
|
634 GCC 4.4.1 |
|
58ad6c70ea60
update gcc from 4.4.0 to 4.4.1.
kent@firefly.cr.ie.u-ryukyu.ac.jp
parents:
0
diff
changeset
|
635 |
|
58ad6c70ea60
update gcc from 4.4.0 to 4.4.1.
kent@firefly.cr.ie.u-ryukyu.ac.jp
parents:
0
diff
changeset
|
636 This is the [13]list of problem reports (PRs) from GCC's bug tracking |
|
58ad6c70ea60
update gcc from 4.4.0 to 4.4.1.
kent@firefly.cr.ie.u-ryukyu.ac.jp
parents:
0
diff
changeset
|
637 system that are known to be fixed in the 4.4.1 release. This list might |
|
58ad6c70ea60
update gcc from 4.4.0 to 4.4.1.
kent@firefly.cr.ie.u-ryukyu.ac.jp
parents:
0
diff
changeset
|
638 not be complete (that is, it is possible that some PRs that have been |
|
58ad6c70ea60
update gcc from 4.4.0 to 4.4.1.
kent@firefly.cr.ie.u-ryukyu.ac.jp
parents:
0
diff
changeset
|
639 fixed are not listed here). |
|
58ad6c70ea60
update gcc from 4.4.0 to 4.4.1.
kent@firefly.cr.ie.u-ryukyu.ac.jp
parents:
0
diff
changeset
|
640 |
|
47
3bfb6c00c1e0
update it from 4.4.2 to 4.4.3.
kent <kent@cr.ie.u-ryukyu.ac.jp>
parents:
19
diff
changeset
|
641 GCC 4.4.2 |
|
3bfb6c00c1e0
update it from 4.4.2 to 4.4.3.
kent <kent@cr.ie.u-ryukyu.ac.jp>
parents:
19
diff
changeset
|
642 |
|
3bfb6c00c1e0
update it from 4.4.2 to 4.4.3.
kent <kent@cr.ie.u-ryukyu.ac.jp>
parents:
19
diff
changeset
|
643 This is the [14]list of problem reports (PRs) from GCC's bug tracking |
|
3bfb6c00c1e0
update it from 4.4.2 to 4.4.3.
kent <kent@cr.ie.u-ryukyu.ac.jp>
parents:
19
diff
changeset
|
644 system that are known to be fixed in the 4.4.2 release. This list might |
|
3bfb6c00c1e0
update it from 4.4.2 to 4.4.3.
kent <kent@cr.ie.u-ryukyu.ac.jp>
parents:
19
diff
changeset
|
645 not be complete (that is, it is possible that some PRs that have been |
|
3bfb6c00c1e0
update it from 4.4.2 to 4.4.3.
kent <kent@cr.ie.u-ryukyu.ac.jp>
parents:
19
diff
changeset
|
646 fixed are not listed here). |
|
3bfb6c00c1e0
update it from 4.4.2 to 4.4.3.
kent <kent@cr.ie.u-ryukyu.ac.jp>
parents:
19
diff
changeset
|
647 |
|
3bfb6c00c1e0
update it from 4.4.2 to 4.4.3.
kent <kent@cr.ie.u-ryukyu.ac.jp>
parents:
19
diff
changeset
|
648 GCC 4.4.3 |
|
3bfb6c00c1e0
update it from 4.4.2 to 4.4.3.
kent <kent@cr.ie.u-ryukyu.ac.jp>
parents:
19
diff
changeset
|
649 |
|
3bfb6c00c1e0
update it from 4.4.2 to 4.4.3.
kent <kent@cr.ie.u-ryukyu.ac.jp>
parents:
19
diff
changeset
|
650 This is the [15]list of problem reports (PRs) from GCC's bug tracking |
|
3bfb6c00c1e0
update it from 4.4.2 to 4.4.3.
kent <kent@cr.ie.u-ryukyu.ac.jp>
parents:
19
diff
changeset
|
651 system that are known to be fixed in the 4.4.3 release. This list might |
|
3bfb6c00c1e0
update it from 4.4.2 to 4.4.3.
kent <kent@cr.ie.u-ryukyu.ac.jp>
parents:
19
diff
changeset
|
652 not be complete (that is, it is possible that some PRs that have been |
|
3bfb6c00c1e0
update it from 4.4.2 to 4.4.3.
kent <kent@cr.ie.u-ryukyu.ac.jp>
parents:
19
diff
changeset
|
653 fixed are not listed here). |
|
3bfb6c00c1e0
update it from 4.4.2 to 4.4.3.
kent <kent@cr.ie.u-ryukyu.ac.jp>
parents:
19
diff
changeset
|
654 |
|
3bfb6c00c1e0
update it from 4.4.2 to 4.4.3.
kent <kent@cr.ie.u-ryukyu.ac.jp>
parents:
19
diff
changeset
|
655 Please send FSF & GNU inquiries & questions to [16]gnu@gnu.org. There |
|
3bfb6c00c1e0
update it from 4.4.2 to 4.4.3.
kent <kent@cr.ie.u-ryukyu.ac.jp>
parents:
19
diff
changeset
|
656 are also [17]other ways to contact the FSF. |
|
3bfb6c00c1e0
update it from 4.4.2 to 4.4.3.
kent <kent@cr.ie.u-ryukyu.ac.jp>
parents:
19
diff
changeset
|
657 |
|
3bfb6c00c1e0
update it from 4.4.2 to 4.4.3.
kent <kent@cr.ie.u-ryukyu.ac.jp>
parents:
19
diff
changeset
|
658 These pages are maintained by [18]the GCC team. |
| 0 | 659 |
| 660 | |
| 661 For questions related to the use of GCC, please consult these web | |
|
47
3bfb6c00c1e0
update it from 4.4.2 to 4.4.3.
kent <kent@cr.ie.u-ryukyu.ac.jp>
parents:
19
diff
changeset
|
662 pages and the [19]GCC manuals. If that fails, the |
|
3bfb6c00c1e0
update it from 4.4.2 to 4.4.3.
kent <kent@cr.ie.u-ryukyu.ac.jp>
parents:
19
diff
changeset
|
663 [20]gcc-help@gcc.gnu.org mailing list might help. |
| 0 | 664 Please send comments on these web pages and the development of GCC to |
|
47
3bfb6c00c1e0
update it from 4.4.2 to 4.4.3.
kent <kent@cr.ie.u-ryukyu.ac.jp>
parents:
19
diff
changeset
|
665 our developer mailing list at [21]gcc@gnu.org or [22]gcc@gcc.gnu.org. |
|
3bfb6c00c1e0
update it from 4.4.2 to 4.4.3.
kent <kent@cr.ie.u-ryukyu.ac.jp>
parents:
19
diff
changeset
|
666 All of our lists have [23]public archives. |
| 0 | 667 |
| 668 Copyright (C) Free Software Foundation, Inc., 51 Franklin St, Fifth | |
| 669 Floor, Boston, MA 02110, USA. | |
| 670 | |
| 671 Verbatim copying and distribution of this entire article is permitted | |
| 672 in any medium, provided this notice is preserved. | |
|
47
3bfb6c00c1e0
update it from 4.4.2 to 4.4.3.
kent <kent@cr.ie.u-ryukyu.ac.jp>
parents:
19
diff
changeset
|
673 Last modified 2010-01-21 [24]Valid XHTML 1.0 |
| 0 | 674 |
| 675 References | |
| 676 | |
|
47
3bfb6c00c1e0
update it from 4.4.2 to 4.4.3.
kent <kent@cr.ie.u-ryukyu.ac.jp>
parents:
19
diff
changeset
|
677 1. http://gcc.gnu.org/gcc-4.4/changes.html#4.4.3 |
|
19
58ad6c70ea60
update gcc from 4.4.0 to 4.4.1.
kent@firefly.cr.ie.u-ryukyu.ac.jp
parents:
0
diff
changeset
|
678 2. http://gcc.gnu.org/gcc-4.3/changes.html#obsoleted |
|
58ad6c70ea60
update gcc from 4.4.0 to 4.4.1.
kent@firefly.cr.ie.u-ryukyu.ac.jp
parents:
0
diff
changeset
|
679 3. http://gcc.gnu.org/gcc-4.4/porting_to.html |
|
58ad6c70ea60
update gcc from 4.4.0 to 4.4.1.
kent@firefly.cr.ie.u-ryukyu.ac.jp
parents:
0
diff
changeset
|
680 4. http://gcc.gnu.org/wiki/Graphite |
|
58ad6c70ea60
update gcc from 4.4.0 to 4.4.1.
kent@firefly.cr.ie.u-ryukyu.ac.jp
parents:
0
diff
changeset
|
681 5. http://openmp.org/wp/openmp-specifications/ |
|
58ad6c70ea60
update gcc from 4.4.0 to 4.4.1.
kent@firefly.cr.ie.u-ryukyu.ac.jp
parents:
0
diff
changeset
|
682 6. http://www.open-std.org/jtc1/sc22/wg14/www/docs/n1040.pdf |
|
58ad6c70ea60
update gcc from 4.4.0 to 4.4.1.
kent@firefly.cr.ie.u-ryukyu.ac.jp
parents:
0
diff
changeset
|
683 7. http://gcc.gnu.org/gcc-4.4/cxx0x_status.html |
|
58ad6c70ea60
update gcc from 4.4.0 to 4.4.1.
kent@firefly.cr.ie.u-ryukyu.ac.jp
parents:
0
diff
changeset
|
684 8. http://gcc.gnu.org/onlinedocs/libstdc++/manual/status.html#id476343 |
|
58ad6c70ea60
update gcc from 4.4.0 to 4.4.1.
kent@firefly.cr.ie.u-ryukyu.ac.jp
parents:
0
diff
changeset
|
685 9. http://gcc.gnu.org/onlinedocs/gfortran/Preprocessing-Options.html |
|
58ad6c70ea60
update gcc from 4.4.0 to 4.4.1.
kent@firefly.cr.ie.u-ryukyu.ac.jp
parents:
0
diff
changeset
|
686 10. http://gcc.gnu.org/onlinedocs/gfortran/Error-and-Warning-Options.html#index-g_t_0040code_007bWarray-temporaries_007d-125 |
|
58ad6c70ea60
update gcc from 4.4.0 to 4.4.1.
kent@firefly.cr.ie.u-ryukyu.ac.jp
parents:
0
diff
changeset
|
687 11. http://gcc.gnu.org/onlinedocs/gfortran/Code-Gen-Options.html#index-g_t_0040code_007bfcheck-array-temporaries_007d-221 |
|
58ad6c70ea60
update gcc from 4.4.0 to 4.4.1.
kent@firefly.cr.ie.u-ryukyu.ac.jp
parents:
0
diff
changeset
|
688 12. http://gcc.gnu.org/onlinedocs/gfortran/Fortran-Dialect-Options.html#index-g_t_0040code_007bbackslash_007d-34 |
|
58ad6c70ea60
update gcc from 4.4.0 to 4.4.1.
kent@firefly.cr.ie.u-ryukyu.ac.jp
parents:
0
diff
changeset
|
689 13. http://gcc.gnu.org/bugzilla/buglist.cgi?bug_status=RESOLVED&resolution=FIXED&target_milestone=4.4.1 |
|
47
3bfb6c00c1e0
update it from 4.4.2 to 4.4.3.
kent <kent@cr.ie.u-ryukyu.ac.jp>
parents:
19
diff
changeset
|
690 14. http://gcc.gnu.org/bugzilla/buglist.cgi?bug_status=RESOLVED&resolution=FIXED&target_milestone=4.4.2 |
|
3bfb6c00c1e0
update it from 4.4.2 to 4.4.3.
kent <kent@cr.ie.u-ryukyu.ac.jp>
parents:
19
diff
changeset
|
691 15. http://gcc.gnu.org/bugzilla/buglist.cgi?bug_status=RESOLVED&resolution=FIXED&target_milestone=4.4.3 |
|
3bfb6c00c1e0
update it from 4.4.2 to 4.4.3.
kent <kent@cr.ie.u-ryukyu.ac.jp>
parents:
19
diff
changeset
|
692 16. mailto:gnu@gnu.org |
|
3bfb6c00c1e0
update it from 4.4.2 to 4.4.3.
kent <kent@cr.ie.u-ryukyu.ac.jp>
parents:
19
diff
changeset
|
693 17. http://www.gnu.org/home.html#ContactInfo |
|
3bfb6c00c1e0
update it from 4.4.2 to 4.4.3.
kent <kent@cr.ie.u-ryukyu.ac.jp>
parents:
19
diff
changeset
|
694 18. http://gcc.gnu.org/about.html |
|
3bfb6c00c1e0
update it from 4.4.2 to 4.4.3.
kent <kent@cr.ie.u-ryukyu.ac.jp>
parents:
19
diff
changeset
|
695 19. http://gcc.gnu.org/onlinedocs/ |
|
3bfb6c00c1e0
update it from 4.4.2 to 4.4.3.
kent <kent@cr.ie.u-ryukyu.ac.jp>
parents:
19
diff
changeset
|
696 20. mailto:gcc-help@gcc.gnu.org |
|
3bfb6c00c1e0
update it from 4.4.2 to 4.4.3.
kent <kent@cr.ie.u-ryukyu.ac.jp>
parents:
19
diff
changeset
|
697 21. mailto:gcc@gnu.org |
|
3bfb6c00c1e0
update it from 4.4.2 to 4.4.3.
kent <kent@cr.ie.u-ryukyu.ac.jp>
parents:
19
diff
changeset
|
698 22. mailto:gcc@gcc.gnu.org |
|
3bfb6c00c1e0
update it from 4.4.2 to 4.4.3.
kent <kent@cr.ie.u-ryukyu.ac.jp>
parents:
19
diff
changeset
|
699 23. http://gcc.gnu.org/lists.html |
|
3bfb6c00c1e0
update it from 4.4.2 to 4.4.3.
kent <kent@cr.ie.u-ryukyu.ac.jp>
parents:
19
diff
changeset
|
700 24. http://validator.w3.org/check/referer |
| 0 | 701 ====================================================================== |
| 702 http://gcc.gnu.org/gcc-4.3/index.html | |
| 703 GCC 4.3 Release Series | |
| 704 | |
| 705 January 24, 2009 | |
| 706 | |
| 707 The [1]GNU project and the GCC developers are pleased to announce the | |
|
47
3bfb6c00c1e0
update it from 4.4.2 to 4.4.3.
kent <kent@cr.ie.u-ryukyu.ac.jp>
parents:
19
diff
changeset
|
708 release of GCC 4.3.4. |
| 0 | 709 |
| 710 This release is a bug-fix release, containing fixes for regressions in | |
|
47
3bfb6c00c1e0
update it from 4.4.2 to 4.4.3.
kent <kent@cr.ie.u-ryukyu.ac.jp>
parents:
19
diff
changeset
|
711 GCC 4.3.3 relative to previous releases of GCC. |
| 0 | 712 |
| 713 Release History | |
| 714 | |
|
47
3bfb6c00c1e0
update it from 4.4.2 to 4.4.3.
kent <kent@cr.ie.u-ryukyu.ac.jp>
parents:
19
diff
changeset
|
715 GCC 4.3.4 |
|
3bfb6c00c1e0
update it from 4.4.2 to 4.4.3.
kent <kent@cr.ie.u-ryukyu.ac.jp>
parents:
19
diff
changeset
|
716 August 4, 2009 ([2]changes) |
|
3bfb6c00c1e0
update it from 4.4.2 to 4.4.3.
kent <kent@cr.ie.u-ryukyu.ac.jp>
parents:
19
diff
changeset
|
717 |
| 0 | 718 GCC 4.3.3 |
|
47
3bfb6c00c1e0
update it from 4.4.2 to 4.4.3.
kent <kent@cr.ie.u-ryukyu.ac.jp>
parents:
19
diff
changeset
|
719 January 24, 2009 ([3]changes) |
| 0 | 720 |
| 721 GCC 4.3.2 | |
|
47
3bfb6c00c1e0
update it from 4.4.2 to 4.4.3.
kent <kent@cr.ie.u-ryukyu.ac.jp>
parents:
19
diff
changeset
|
722 August 27, 2008 ([4]changes) |
| 0 | 723 |
| 724 GCC 4.3.1 | |
|
47
3bfb6c00c1e0
update it from 4.4.2 to 4.4.3.
kent <kent@cr.ie.u-ryukyu.ac.jp>
parents:
19
diff
changeset
|
725 June 6, 2008 ([5]changes) |
| 0 | 726 |
| 727 GCC 4.3.0 | |
|
47
3bfb6c00c1e0
update it from 4.4.2 to 4.4.3.
kent <kent@cr.ie.u-ryukyu.ac.jp>
parents:
19
diff
changeset
|
728 March 5, 2008 ([6]changes) |
| 0 | 729 |
| 730 References and Acknowledgements | |
| 731 | |
| 732 GCC used to stand for the GNU C Compiler, but since the compiler | |
| 733 supports several other languages aside from C, it now stands for the | |
| 734 GNU Compiler Collection. | |
| 735 | |
|
47
3bfb6c00c1e0
update it from 4.4.2 to 4.4.3.
kent <kent@cr.ie.u-ryukyu.ac.jp>
parents:
19
diff
changeset
|
736 A list of [7]successful builds is updated as new information becomes |
| 0 | 737 available. |
| 738 | |
| 739 The GCC developers would like to thank the numerous people that have | |
| 740 contributed new features, improvements, bug fixes, and other changes as | |
|
47
3bfb6c00c1e0
update it from 4.4.2 to 4.4.3.
kent <kent@cr.ie.u-ryukyu.ac.jp>
parents:
19
diff
changeset
|
741 well as test results to GCC. This [8]amazing group of volunteers is |
| 0 | 742 what makes GCC successful. |
| 743 | |
|
47
3bfb6c00c1e0
update it from 4.4.2 to 4.4.3.
kent <kent@cr.ie.u-ryukyu.ac.jp>
parents:
19
diff
changeset
|
744 For additional information about GCC please refer to the [9]GCC project |
|
3bfb6c00c1e0
update it from 4.4.2 to 4.4.3.
kent <kent@cr.ie.u-ryukyu.ac.jp>
parents:
19
diff
changeset
|
745 web site or contact the [10]GCC development mailing list. |
|
3bfb6c00c1e0
update it from 4.4.2 to 4.4.3.
kent <kent@cr.ie.u-ryukyu.ac.jp>
parents:
19
diff
changeset
|
746 |
|
3bfb6c00c1e0
update it from 4.4.2 to 4.4.3.
kent <kent@cr.ie.u-ryukyu.ac.jp>
parents:
19
diff
changeset
|
747 To obtain GCC please use [11]our mirror sites or [12]our SVN server. |
|
3bfb6c00c1e0
update it from 4.4.2 to 4.4.3.
kent <kent@cr.ie.u-ryukyu.ac.jp>
parents:
19
diff
changeset
|
748 |
|
3bfb6c00c1e0
update it from 4.4.2 to 4.4.3.
kent <kent@cr.ie.u-ryukyu.ac.jp>
parents:
19
diff
changeset
|
749 Please send FSF & GNU inquiries & questions to [13]gnu@gnu.org. There |
|
3bfb6c00c1e0
update it from 4.4.2 to 4.4.3.
kent <kent@cr.ie.u-ryukyu.ac.jp>
parents:
19
diff
changeset
|
750 are also [14]other ways to contact the FSF. |
|
3bfb6c00c1e0
update it from 4.4.2 to 4.4.3.
kent <kent@cr.ie.u-ryukyu.ac.jp>
parents:
19
diff
changeset
|
751 |
|
3bfb6c00c1e0
update it from 4.4.2 to 4.4.3.
kent <kent@cr.ie.u-ryukyu.ac.jp>
parents:
19
diff
changeset
|
752 These pages are maintained by [15]the GCC team. |
| 0 | 753 |
| 754 | |
| 755 For questions related to the use of GCC, please consult these web | |
|
47
3bfb6c00c1e0
update it from 4.4.2 to 4.4.3.
kent <kent@cr.ie.u-ryukyu.ac.jp>
parents:
19
diff
changeset
|
756 pages and the [16]GCC manuals. If that fails, the |
|
3bfb6c00c1e0
update it from 4.4.2 to 4.4.3.
kent <kent@cr.ie.u-ryukyu.ac.jp>
parents:
19
diff
changeset
|
757 [17]gcc-help@gcc.gnu.org mailing list might help. |
| 0 | 758 Please send comments on these web pages and the development of GCC to |
|
47
3bfb6c00c1e0
update it from 4.4.2 to 4.4.3.
kent <kent@cr.ie.u-ryukyu.ac.jp>
parents:
19
diff
changeset
|
759 our developer mailing list at [18]gcc@gnu.org or [19]gcc@gcc.gnu.org. |
|
3bfb6c00c1e0
update it from 4.4.2 to 4.4.3.
kent <kent@cr.ie.u-ryukyu.ac.jp>
parents:
19
diff
changeset
|
760 All of our lists have [20]public archives. |
| 0 | 761 |
| 762 Copyright (C) Free Software Foundation, Inc., 51 Franklin St, Fifth | |
| 763 Floor, Boston, MA 02110, USA. | |
| 764 | |
| 765 Verbatim copying and distribution of this entire article is permitted | |
| 766 in any medium, provided this notice is preserved. | |
|
47
3bfb6c00c1e0
update it from 4.4.2 to 4.4.3.
kent <kent@cr.ie.u-ryukyu.ac.jp>
parents:
19
diff
changeset
|
767 Last modified 2009-10-15 [21]Valid XHTML 1.0 |
| 0 | 768 |
| 769 References | |
| 770 | |
| 771 1. http://www.gnu.org/ | |
| 772 2. http://gcc.gnu.org/gcc-4.3/changes.html | |
| 773 3. http://gcc.gnu.org/gcc-4.3/changes.html | |
| 774 4. http://gcc.gnu.org/gcc-4.3/changes.html | |
| 775 5. http://gcc.gnu.org/gcc-4.3/changes.html | |
|
47
3bfb6c00c1e0
update it from 4.4.2 to 4.4.3.
kent <kent@cr.ie.u-ryukyu.ac.jp>
parents:
19
diff
changeset
|
776 6. http://gcc.gnu.org/gcc-4.3/changes.html |
|
3bfb6c00c1e0
update it from 4.4.2 to 4.4.3.
kent <kent@cr.ie.u-ryukyu.ac.jp>
parents:
19
diff
changeset
|
777 7. http://gcc.gnu.org/gcc-4.3/buildstat.html |
|
3bfb6c00c1e0
update it from 4.4.2 to 4.4.3.
kent <kent@cr.ie.u-ryukyu.ac.jp>
parents:
19
diff
changeset
|
778 8. http://gcc.gnu.org/onlinedocs/gcc/Contributors.html |
|
3bfb6c00c1e0
update it from 4.4.2 to 4.4.3.
kent <kent@cr.ie.u-ryukyu.ac.jp>
parents:
19
diff
changeset
|
779 9. http://gcc.gnu.org/index.html |
|
3bfb6c00c1e0
update it from 4.4.2 to 4.4.3.
kent <kent@cr.ie.u-ryukyu.ac.jp>
parents:
19
diff
changeset
|
780 10. mailto:gcc@gcc.gnu.org |
|
3bfb6c00c1e0
update it from 4.4.2 to 4.4.3.
kent <kent@cr.ie.u-ryukyu.ac.jp>
parents:
19
diff
changeset
|
781 11. http://gcc.gnu.org/mirrors.html |
|
3bfb6c00c1e0
update it from 4.4.2 to 4.4.3.
kent <kent@cr.ie.u-ryukyu.ac.jp>
parents:
19
diff
changeset
|
782 12. http://gcc.gnu.org/svn.html |
|
3bfb6c00c1e0
update it from 4.4.2 to 4.4.3.
kent <kent@cr.ie.u-ryukyu.ac.jp>
parents:
19
diff
changeset
|
783 13. mailto:gnu@gnu.org |
|
3bfb6c00c1e0
update it from 4.4.2 to 4.4.3.
kent <kent@cr.ie.u-ryukyu.ac.jp>
parents:
19
diff
changeset
|
784 14. http://www.gnu.org/home.html#ContactInfo |
|
3bfb6c00c1e0
update it from 4.4.2 to 4.4.3.
kent <kent@cr.ie.u-ryukyu.ac.jp>
parents:
19
diff
changeset
|
785 15. http://gcc.gnu.org/about.html |
|
3bfb6c00c1e0
update it from 4.4.2 to 4.4.3.
kent <kent@cr.ie.u-ryukyu.ac.jp>
parents:
19
diff
changeset
|
786 16. http://gcc.gnu.org/onlinedocs/ |
|
3bfb6c00c1e0
update it from 4.4.2 to 4.4.3.
kent <kent@cr.ie.u-ryukyu.ac.jp>
parents:
19
diff
changeset
|
787 17. mailto:gcc-help@gcc.gnu.org |
|
3bfb6c00c1e0
update it from 4.4.2 to 4.4.3.
kent <kent@cr.ie.u-ryukyu.ac.jp>
parents:
19
diff
changeset
|
788 18. mailto:gcc@gnu.org |
|
3bfb6c00c1e0
update it from 4.4.2 to 4.4.3.
kent <kent@cr.ie.u-ryukyu.ac.jp>
parents:
19
diff
changeset
|
789 19. mailto:gcc@gcc.gnu.org |
|
3bfb6c00c1e0
update it from 4.4.2 to 4.4.3.
kent <kent@cr.ie.u-ryukyu.ac.jp>
parents:
19
diff
changeset
|
790 20. http://gcc.gnu.org/lists.html |
|
3bfb6c00c1e0
update it from 4.4.2 to 4.4.3.
kent <kent@cr.ie.u-ryukyu.ac.jp>
parents:
19
diff
changeset
|
791 21. http://validator.w3.org/check/referer |
| 0 | 792 ====================================================================== |
| 793 http://gcc.gnu.org/gcc-4.3/changes.html | |
| 794 GCC 4.3 Release Series | |
| 795 Changes, New Features, and Fixes | |
| 796 | |
|
47
3bfb6c00c1e0
update it from 4.4.2 to 4.4.3.
kent <kent@cr.ie.u-ryukyu.ac.jp>
parents:
19
diff
changeset
|
797 The latest release in the 4.3 release series is [1]GCC 4.3.4. |
| 0 | 798 |
| 799 Caveats | |
| 800 | |
| 801 * GCC requires the [2]GMP and [3]MPFR libraries for building all the | |
| 802 various front-end languages it supports. See the [4]prerequisites | |
| 803 page for version requirements. | |
| 804 * ColdFire targets now treat long double as having the same format as | |
| 805 double. In earlier versions of GCC, they used the 68881 long double | |
| 806 format instead. | |
| 807 * The m68k-uclinux target now uses the same calling conventions as | |
| 808 m68k-linux-gnu. You can select the original calling conventions by | |
| 809 configuring for m68k-uclinuxoldabi instead. Note that | |
| 810 m68k-uclinuxoldabi also retains the original 80-bit long double on | |
| 811 ColdFire targets. | |
| 812 * The -fforce-mem option has been removed because it has had no | |
| 813 effect in the last few GCC releases. | |
| 814 * The i386 -msvr3-shlib option has been removed since it is no longer | |
| 815 used. | |
| 816 * Fastcall for i386 has been changed not to pass aggregate arguments | |
| 817 in registers, following Microsoft compilers. | |
| 818 * Support for the AOF assembler has been removed from the ARM back | |
| 819 end; this affects only the targets arm-semi-aof and armel-semi-aof, | |
| 820 which are no longer recognized. We removed these targets without a | |
| 821 deprecation period because we discovered that they have been | |
| 822 unusable since GCC 4.0.0. | |
| 823 * Support for the TMS320C3x/C4x processor (targets c4x-* and tic4x-*) | |
| 824 has been removed. This support had been deprecated since GCC 4.0.0. | |
| 825 * Support for a number of older systems and recently unmaintained or | |
| 826 untested target ports of GCC has been declared obsolete in GCC 4.3. | |
| 827 Unless there is activity to revive them, the next release of GCC | |
| 828 will have their sources permanently removed. | |
| 829 All GCC ports for the following processor architectures have been | |
| 830 declared obsolete: | |
| 831 + Morpho MT (mt-*) | |
| 832 The following aliases for processor architectures have been | |
| 833 declared obsolete. Users should use the indicated generic target | |
| 834 names instead, with compile-time options such as -mcpu or | |
| 835 configure-time options such as --with-cpu to control the | |
| 836 configuration more precisely. | |
| 837 + strongarm*-*-*, ep9312*-*-*, xscale*-*-* (use arm*-*-* | |
| 838 instead). | |
| 839 + parisc*-*-* (use hppa*-*-* instead). | |
| 840 + m680[012]0-*-* (use m68k-*-* instead). | |
| 841 All GCC ports for the following operating systems have been | |
| 842 declared obsolete: | |
| 843 + BeOS (*-*-beos*) | |
| 844 + kaOS (*-*-kaos*) | |
| 845 + GNU/Linux using the a.out object format (*-*-linux*aout*) | |
| 846 + GNU/Linux using version 1 of the GNU C Library | |
| 847 (*-*-linux*libc1*) | |
| 848 + Solaris versions before Solaris 7 (*-*-solaris2.[0-6], | |
| 849 *-*-solaris2.[0-6].*) | |
| 850 + Miscellaneous System V (*-*-sysv*) | |
| 851 + WindISS (*-*-windiss*) | |
| 852 Also, those for some individual systems on particular architectures | |
| 853 have been obsoleted: | |
| 854 + UNICOS/mk on DEC Alpha (alpha*-*-unicosmk*) | |
| 855 + CRIS with a.out object format (cris-*-aout) | |
| 856 + BSD 4.3 on PA-RISC (hppa1.1-*-bsd*) | |
| 857 + OSF/1 on PA-RISC (hppa1.1-*-osf*) | |
| 858 + PRO on PA-RISC (hppa1.1-*-pro*) | |
| 859 + Sequent PTX on IA32 (i[34567]86-sequent-ptx4*, | |
| 860 i[34567]86-sequent-sysv4*) | |
| 861 + SCO Open Server 5 on IA32 (i[34567]86-*-sco3.2v5*) | |
| 862 + UWIN on IA32 (i[34567]86-*-uwin*) (support for UWIN as a host | |
| 863 was previously [5]removed in 2001, leaving only the support | |
| 864 for UWIN as a target now being deprecated) | |
| 865 + ChorusOS on PowerPC (powerpc-*-chorusos*) | |
| 866 + All VAX configurations apart from NetBSD and OpenBSD | |
| 867 (vax-*-bsd*, vax-*-sysv*, vax-*-ultrix*) | |
| 868 * The [6]-Wconversion option has been modified. Its purpose now is to | |
| 869 warn for implicit conversions that may alter a value. This new | |
| 870 behavior is available for both C and C++. Warnings about | |
| 871 conversions between signed and unsigned integers can be disabled by | |
| 872 using -Wno-sign-conversion. In C++, they are disabled by default | |
| 873 unless -Wsign-conversion is explicitly requested. The old behavior | |
| 874 of -Wconversion, that is, warn for prototypes causing a type | |
| 875 conversion that is different from what would happen to the same | |
| 876 argument in the absence of a prototype, has been moved to a new | |
| 877 option -Wtraditional-conversion, which is only available for C. | |
| 878 * The -m386, -m486, -mpentium and -mpentiumpro tuning options have | |
| 879 been removed because they were deprecated for more than 3 GCC major | |
| 880 releases. Use -mtune=i386, -mtune=i486, -mtune=pentium or | |
| 881 -mtune=pentiumpro as a replacement. | |
| 882 * The -funsafe-math-optimizations option now automatically turns on | |
| 883 -fno-trapping-math in addition to -fno-signed-zeros, as it enables | |
| 884 reassociation and thus may introduce or remove traps. | |
| 885 * The -ftree-vectorize option is now on by default under -O3. In | |
| 886 order to generate code for a SIMD extension, it has to be enabled | |
| 887 as well: use -maltivec for PowerPC platforms and -msse/-msse2 for | |
| 888 i?86 and x86_64. | |
| 889 * More information on porting to GCC 4.3 from previous versions of | |
| 890 GCC can be found in the [7]porting guide for this release. | |
| 891 | |
| 892 General Optimizer Improvements | |
| 893 | |
| 894 * The GCC middle-end has been integrated with the [8]MPFR library. | |
| 895 This allows GCC to evaluate and replace at compile-time calls to | |
| 896 built-in math functions having constant arguments with their | |
| 897 mathematically equivalent results. In making use of [9]MPFR, GCC | |
| 898 can generate correct results regardless of the math library | |
| 899 implementation or floating point precision of the host platform. | |
| 900 This also allows GCC to generate identical results regardless of | |
| 901 whether one compiles in native or cross-compile configurations to a | |
| 902 particular target. The following built-in functions take advantage | |
| 903 of this new capability: acos, acosh, asin, asinh, atan2, atan, | |
| 904 atanh, cbrt, cos, cosh, drem, erf, erfc, exp10, exp2, exp, expm1, | |
| 905 fdim, fma, fmax, fmin, gamma_r, hypot, j0, j1, jn, lgamma_r, log10, | |
| 906 log1p, log2, log, pow10, pow, remainder, remquo, sin, sincos, sinh, | |
| 907 tan, tanh, tgamma, y0, y1 and yn. The float and long double | |
| 908 variants of these functions (e.g. sinf and sinl) are also handled. | |
| 909 The sqrt and cabs functions with constant arguments were already | |
| 910 optimized in prior GCC releases. Now they also use [10]MPFR. | |
| 911 * A new forward propagation pass on RTL was added. The new pass | |
| 912 replaces several slower transformations, resulting in compile-time | |
| 913 improvements as well as better code generation in some cases. | |
| 914 * A new command-line switch -frecord-gcc-switches has been added to | |
| 915 GCC, although it is only enabled for some targets. The switch | |
| 916 causes the command line that was used to invoke the compiler to be | |
| 917 recorded into the object file that is being created. The exact | |
| 918 format of this recording is target and binary file format | |
| 919 dependent, but it usually takes the form of a note section | |
| 920 containing ASCII text. The switch is related to the -fverbose-asm | |
| 921 switch, but that one only records the information in the assembler | |
| 922 output file as comments, so the information never reaches the | |
| 923 object file. | |
| 924 * The inliner heuristic is now aware of stack frame consumption. New | |
| 925 command-line parameters --param large-stack-frame and --param | |
| 926 large-stack-frame-growth can be used to limit stack frame size | |
| 927 growth caused by inlining. | |
| 928 * During feedback directed optimizations, the expected block size the | |
| 929 memcpy, memset and bzero functions operate on is discovered and for | |
| 930 cases of commonly used small sizes, specialized inline code is | |
| 931 generated. | |
| 932 * __builtin_expect no longer requires its argument to be a compile | |
| 933 time constant. | |
| 934 * Interprocedural optimization was reorganized to work on functions | |
| 935 in SSA form. This enables more precise and cheaper dataflow | |
| 936 analysis and makes writing interprocedural optimizations easier. | |
| 937 The following improvements have been implemented on top of this | |
| 938 framework: | |
| 939 + Pre-inline optimization: Selected local optimization passes | |
| 940 are run before the inliner (and other interprocedural passes) | |
| 941 are executed. This significantly improves the accuracy of code | |
| 942 growth estimates used by the inliner and reduces the overall | |
| 943 memory footprint for large compilation units. | |
| 944 + Early inlining (a simple bottom-up inliner pass inlining only | |
| 945 functions whose body is smaller than the expected call | |
| 946 overhead) is now executed with the early optimization passes, | |
| 947 thus inlining already optimized function bodies into an | |
| 948 unoptimized function that is subsequently optimized by early | |
| 949 optimizers. This enables the compiler to quickly eliminate | |
| 950 abstraction penalty in C++ programs. | |
| 951 + Interprocedural constant propagation now operate on SSA form | |
| 952 increasing accuracy of the analysis. | |
| 953 * A new internal representation for GIMPLE statements has been | |
| 954 contributed, resulting in compile-time memory savings. | |
| 955 * The vectorizer was enhanced to support vectorization of outer | |
| 956 loops, intra-iteration parallelism (loop-aware SLP), vectorization | |
| 957 of strided accesses and loops with multiple data-types. Run-time | |
| 958 dependency testing using loop versioning was added. The cost model, | |
| 959 turned on by -fvect-cost-model, was developed. | |
| 960 | |
| 961 New Languages and Language specific improvements | |
| 962 | |
| 963 * We have added new command-line options | |
| 964 -finstrument-functions-exclude-function-list and | |
| 965 -finstrument-functions-exclude-file-list. They provide more control | |
| 966 over which functions are annotated by the -finstrument-functions | |
| 967 option. | |
| 968 | |
| 969 C family | |
| 970 | |
| 971 * Implicit conversions between generic vector types are now only | |
| 972 permitted when the two vectors in question have the same number of | |
| 973 elements and compatible element types. (Note that the restriction | |
| 974 involves compatible element types, not implicitly-convertible | |
| 975 element types: thus, a vector type with element type int may not be | |
| 976 implicitly converted to a vector type with element type unsigned | |
| 977 int.) This restriction, which is in line with specifications for | |
| 978 SIMD architectures such as AltiVec, may be relaxed using the flag | |
| 979 -flax-vector-conversions. This flag is intended only as a | |
| 980 compatibility measure and should not be used for new code. | |
| 981 * -Warray-bounds has been added and is now enabled by default for | |
| 982 -Wall . It produces warnings for array subscripts that can be | |
| 983 determined at compile time to be always out of bounds. | |
| 984 -Wno-array-bounds will disable the warning. | |
| 985 * The constructor and destructor function attributes now accept | |
| 986 optional priority arguments which control the order in which the | |
| 987 constructor and destructor functions are run. | |
| 988 * New [11]command-line options -Wtype-limits, | |
| 989 -Wold-style-declaration, -Wmissing-parameter-type, -Wempty-body, | |
| 990 -Wclobbered and -Wignored-qualifiers have been added for finer | |
| 991 control of the diverse warnings enabled by -Wextra. | |
| 992 * A new function attribute alloc_size has been added to mark up | |
| 993 malloc style functions. For constant sized allocations this can be | |
| 994 used to find out the size of the returned pointer using the | |
| 995 __builtin_object_size() function for buffer overflow checking and | |
| 996 similar. This supplements the already built-in malloc and calloc | |
| 997 constant size handling. | |
| 998 * Integer constants written in binary are now supported as a GCC | |
| 999 extension. They consist of a prefix 0b or 0B, followed by a | |
| 1000 sequence of 0 and 1 digits. | |
| 1001 * A new predefined macro __COUNTER__ has been added. It expands to | |
| 1002 sequential integral values starting from 0. In conjunction with the | |
| 1003 ## operator, this provides a convenient means to generate unique | |
| 1004 identifiers. | |
| 1005 * A new command-line option -fdirectives-only has been added. It | |
| 1006 enables a special preprocessing mode which improves the performance | |
| 1007 of applications like distcc and ccache. | |
| 1008 * Fixed-point data types and operators have been added. They are | |
| 1009 based on Chapter 4 of the Embedded-C specification (n1169.pdf). | |
| 1010 Currently, only MIPS targets are supported. | |
| 1011 * Decimal floating-point arithmetic based on draft ISO/IEC TR 24732, | |
| 1012 N1241, is now supported as a GCC extension to C for targets | |
| 1013 i[34567]86-*-linux-gnu, powerpc*-*-linux-gnu, s390*-ibm-linux-gnu, | |
| 1014 and x86_64-*-linux-gnu. The feature introduces new data types | |
| 1015 _Decimal32, _Decimal64, and _Decimal128 with constant suffixes DF, | |
| 1016 DD, and DL. | |
| 1017 | |
| 1018 C++ | |
| 1019 | |
| 1020 * [12]Experimental support for the upcoming ISO C++ standard, C++0x. | |
| 1021 * -Wc++0x-compat has been added and is now enabled by default for | |
| 1022 -Wall. It produces warnings for constructs whose meaning differs | |
| 1023 between ISO C++ 1998 and C++0x. | |
| 1024 * The -Wparentheses option now works for C++ as it does for C. It | |
| 1025 warns if parentheses are omitted when operators with confusing | |
| 1026 precedence are nested. It also warns about ambiguous else | |
| 1027 statements. Since -Wparentheses is enabled by -Wall, this may cause | |
| 1028 additional warnings with existing C++ code which uses -Wall. These | |
| 1029 new warnings may be disabled by using -Wall -Wno-parentheses. | |
| 1030 * The -Wmissing-declarations now works for C++ as it does for C. | |
| 1031 * The -fvisibility-ms-compat flag was added, to make it easier to | |
| 1032 port larger projects using shared libraries from Microsoft's Visual | |
| 1033 Studio to ELF and Mach-O systems. | |
| 1034 * C++ attribute handling has been overhauled for template arguments | |
| 1035 (ie dependent types). In particular, __attribute__((aligned(T))); | |
| 1036 works for C++ types. | |
| 1037 | |
| 1038 Runtime Library (libstdc++) | |
| 1039 | |
| 1040 * [13]Experimental support for the upcoming ISO C++ standard, C++0x. | |
| 1041 * Support for TR1 mathematical special functions and regular | |
| 1042 expressions. ([14]Implementation status of TR1) | |
| 1043 * Default what implementations give more elaborate exception strings | |
| 1044 for bad_cast, bad_typeid, bad_exception, and bad_alloc. | |
| 1045 * Header dependencies have been streamlined, reducing unnecessary | |
| 1046 includes and pre-processed bloat. | |
| 1047 * Variadic template implementations of items in <tuple> and | |
| 1048 <functional>. | |
| 1049 * An experimental [15]parallel mode has been added. This is a | |
| 1050 parallel implementation of many C++ Standard library algorithms, | |
| 1051 like std::accumulate, std::for_each, std::transform, or std::sort, | |
| 1052 to give but four examples. These algorithms can be substituted for | |
| 1053 the normal (sequential) libstdc++ algorithms on a piecemeal basis, | |
| 1054 or all existing algorithms can be transformed via the | |
| 1055 -D_GLIBCXX_PARALLEL macro. | |
| 1056 * Debug mode versions of classes in <unordered_set> and | |
| 1057 <unordered_map>. | |
| 1058 * Formal deprecation of <ext/hash_set> and <ext/hash_map>, which are | |
| 1059 now <backward/hash_set> and <backward/hash_map>. This code: | |
| 1060 #include <ext/hash_set> | |
| 1061 __gnu_cxx::hash_set<int> s; | |
| 1062 | |
| 1063 Can be transformed (in order of preference) to: | |
| 1064 #include <tr1/unordered_set> | |
| 1065 std::tr1::unordered_set<int> s; | |
| 1066 | |
| 1067 or | |
| 1068 #include <backward/hash_set> | |
| 1069 __gnu_cxx::hash_set<int> s; | |
| 1070 | |
| 1071 Similar transformations apply to __gnu_cxx::hash_map, | |
| 1072 __gnu_cxx::hash_multimap, __gnu_cxx::hash_set, | |
| 1073 __gnu_cxx::hash_multiset. | |
| 1074 | |
| 1075 Fortran | |
| 1076 | |
| 1077 * Due to the fact that the [16]GMP and [17]MPFR libraries are | |
| 1078 required for all languages, Fortran is no longer special in this | |
| 1079 regard and is available by default. | |
| 1080 * The [18]-fexternal-blas option has been added, which generates | |
| 1081 calls to BLAS routines for intrinsic matrix operations such as | |
| 1082 matmul rather than using the built-in algorithms. | |
| 1083 * Support to give a backtrace (compiler flag -fbacktrace or | |
| 1084 environment variable GFORTRAN_ERROR_BACKTRACE; on glibc systems | |
| 1085 only) or a core dump (-fdump-core, GFORTRAN_ERROR_DUMPCORE) when a | |
| 1086 run-time error occured. | |
| 1087 * GNU Fortran now defines __GFORTRAN__ when it runs the C | |
| 1088 preprocessor (CPP). | |
| 1089 * The [19]-finit-local-zero, -finit-real, -finit-integer, | |
| 1090 -finit-character, and -finit-logical options have been added, which | |
| 1091 can be used to initialize local variables. | |
| 1092 * The intrinsic procedures [20]GAMMA and [21]LGAMMA have been added, | |
| 1093 which calculate the Gamma function and its logarithm. Use EXTERNAL | |
| 1094 gamma if you want to use your own gamma function. | |
| 1095 * GNU Fortran now regards the backslash character as literal (as | |
| 1096 required by the Fortran 2003 standard); using [22]-fbackslash GNU | |
| 1097 Fortran interprets backslashes as C-style escape characters. | |
| 1098 * The [23]interpretation of binary, octal and hexadecimal (BOZ) | |
| 1099 literal constants has been changed. Before they were always | |
| 1100 interpreted as integer; now they are bit-wise transferred as | |
| 1101 argument of INT, REAL, DBLE and CMPLX as required by the Fortran | |
| 1102 2003 standard, and for real and complex variables in DATA | |
| 1103 statements or when directly assigned to real and complex variables. | |
| 1104 Everywhere else and especially in expressions they are still | |
| 1105 regarded as integer constants. | |
| 1106 * Fortran 2003 support has been extended: | |
| 1107 + Intrinsic statements IMPORT, PROTECTED, VALUE and VOLATILE | |
| 1108 + Pointer intent | |
| 1109 + Intrinsic module ISO_ENV_FORTRAN | |
| 1110 + Interoperability with C (ISO C Bindings) | |
| 1111 + ABSTRACT INTERFACES and PROCEDURE statements (without POINTER | |
| 1112 attribute) | |
| 1113 + Fortran 2003 BOZ | |
| 1114 | |
| 1115 Java (GCJ) | |
| 1116 | |
| 1117 * gcj now uses the Eclipse Java compiler for its Java parsing needs. | |
| 1118 This enables the use of all 1.5 language features, and fixes most | |
| 1119 existing front end bugs. | |
| 1120 * libgcj now supports all 1.5 language features which require runtime | |
| 1121 support: foreach, enum, annotations, generics, and auto-boxing. | |
| 1122 * We've made many changes to the tools shipped with gcj. | |
| 1123 + The old jv-scan tool has been removed. This tool never really | |
| 1124 worked properly. There is no replacement. | |
| 1125 + gcjh has been rewritten. Some of its more obscure options no | |
| 1126 longer work, but are still recognized in an attempt at | |
| 1127 compatibility. gjavah is a new program with similar | |
| 1128 functionality but different command-line options. | |
| 1129 + grmic and grmiregistry have been rewritten. grmid has been | |
| 1130 added. | |
| 1131 + gjar replaces the old fastjar. | |
| 1132 + gjarsigner (used for signing jars), gkeytool (used for key | |
| 1133 management), gorbd (for CORBA), gserialver (computes | |
| 1134 serialization UIDs), and gtnameserv (also for CORBA) are now | |
| 1135 installed. | |
| 1136 * The ability to dump the contents of the java run time heap to a | |
| 1137 file for off-line analysis has been added. The heap dumps may be | |
| 1138 analyzed with the new gc-analyze tool. They may be generated on | |
| 1139 out-of-memory conditions or on demand and are controlled by the new | |
| 1140 run time class gnu.gcj.util.GCInfo. | |
| 1141 * java.util.TimeZone can now read files from /usr/share/zoneinfo to | |
| 1142 provide correct, updated, timezone information. This means that | |
| 1143 packagers no longer have to update libgcj when a time zone change | |
| 1144 is published. | |
| 1145 | |
| 1146 New Targets and Target Specific Improvements | |
| 1147 | |
| 1148 IA-32/x86-64 | |
| 1149 | |
| 1150 * Tuning for Intel Core 2 processors is available via -mtune=core2 | |
| 1151 and -march=core2. | |
| 1152 * Tuning for AMD Geode processors is available via -mtune=geode and | |
| 1153 -march=geode. | |
| 1154 * Code generation of block move (memcpy) and block set (memset) was | |
| 1155 rewritten. GCC can now pick the best algorithm (loop, unrolled | |
| 1156 loop, instruction with rep prefix or a library call) based on the | |
| 1157 size of the block being copied and the CPU being optimized for. A | |
| 1158 new option -minline-stringops-dynamically has been added. With this | |
| 1159 option string operations of unknown size are expanded such that | |
| 1160 small blocks are copied by in-line code, while for large blocks a | |
| 1161 library call is used. This results in faster code than | |
| 1162 -minline-all-stringops when the library implementation is capable | |
| 1163 of using cache hierarchy hints. The heuristic choosing the | |
| 1164 particular algorithm can be overwritten via -mstringop-strategy. | |
| 1165 Newly also memset of values different from 0 is inlined. | |
| 1166 * GCC no longer places the cld instruction before string operations. | |
| 1167 Both i386 and x86-64 ABI documents mandate the direction flag to be | |
| 1168 clear at the entry of a function. It is now invalid to set the flag | |
| 1169 in asm statement without reseting it afterward. | |
| 1170 * Support for SSSE3 built-in functions and code generation are | |
| 1171 available via -mssse3. | |
| 1172 * Support for SSE4.1 built-in functions and code generation are | |
| 1173 available via -msse4.1. | |
| 1174 * Support for SSE4.2 built-in functions and code generation are | |
| 1175 available via -msse4.2. | |
| 1176 * Both SSE4.1 and SSE4.2 support can be enabled via -msse4. | |
| 1177 * A new set of options -mpc32, -mpc64 and -mpc80 have been added to | |
| 1178 allow explicit control of x87 floating point precision. | |
| 1179 * Support for __float128 (TFmode) IEEE quad type and corresponding | |
| 1180 TCmode IEEE complex quad type is available via the soft-fp library | |
| 1181 on x86_64 targets. This includes basic arithmetic operations | |
| 1182 (addition, subtraction, negation, multiplication and division) on | |
| 1183 __float128 real and TCmode complex values, the full set of IEEE | |
| 1184 comparisons between __float128 values, conversions to and from | |
| 1185 float, double and long double floating point types, as well as | |
| 1186 conversions to and from signed or unsigned integer, signed or | |
| 1187 unsigned long integer and signed or unsigned quad (TImode) integer | |
| 1188 types. Additionally, all operations generate the full set of IEEE | |
| 1189 exceptions and support the full set of IEEE rounding modes. | |
| 1190 * GCC can now utilize the ACML library for vectorizing calls to a set | |
| 1191 of C99 functions on x86_64 if -mveclibabi=acml is specified and you | |
| 1192 link to an ACML ABI compatible library. | |
| 1193 | |
| 1194 ARM | |
| 1195 | |
| 1196 * Compiler and Library support for Thumb-2 and the ARMv7 architecture | |
| 1197 has been added. | |
| 1198 | |
| 1199 CRIS | |
| 1200 | |
| 1201 New features | |
| 1202 | |
| 1203 * Compiler and Library support for the CRIS v32 architecture, as | |
| 1204 found in Axis Communications ETRAX FS and ARTPEC-3 chips, has been | |
| 1205 added. | |
| 1206 | |
| 1207 Configuration changes | |
| 1208 | |
| 1209 * The cris-*-elf target now includes support for CRIS v32, including | |
| 1210 libraries, through the -march=v32 option. | |
| 1211 * A new crisv32-*-elf target defaults to generate code for CRIS v32. | |
| 1212 * A new crisv32-*-linux* target defaults to generate code for CRIS | |
| 1213 v32. | |
| 1214 * The cris-*-aout target has been obsoleted. | |
| 1215 | |
| 1216 Improved support for built-in functions | |
| 1217 | |
| 1218 * GCC can now use the lz and swapwbr instructions to implement the | |
| 1219 __builtin_clz, __builtin_ctz and __builtin_ffs family of functions. | |
| 1220 * __builtin_bswap32 is now implemented using the swapwb instruction, | |
| 1221 when available. | |
| 1222 | |
| 1223 m68k and ColdFire | |
| 1224 | |
| 1225 New features | |
| 1226 | |
| 1227 * Support for several new ColdFire processors has been added. You can | |
| 1228 generate code for them using the new -mcpu option. | |
| 1229 * All targets now support ColdFire processors. | |
| 1230 * m68k-uclinux targets have improved support for C++ constructors and | |
| 1231 destructors, and for shared libraries. | |
| 1232 * It is now possible to set breakpoints on the first or last line of | |
| 1233 a function, even if there are no statements on that line. | |
| 1234 | |
| 1235 Optimizations | |
| 1236 | |
| 1237 * Support for sibling calls has been added. | |
| 1238 * More use is now made of the ColdFire mov3q instruction. | |
| 1239 * __builtin_clz is now implemented using the ff1 ColdFire | |
| 1240 instruction, when available. | |
| 1241 * GCC now honors the -m68010 option. 68010 code now uses clr rather | |
| 1242 than move to zero volatile memory. | |
| 1243 * 68020 targets and above can now use symbol(index.size*scale) | |
| 1244 addresses for indexed array accesses. Earlier compilers would | |
| 1245 always load the symbol into a base register first. | |
| 1246 | |
| 1247 Configuration changes | |
| 1248 | |
| 1249 * All m68k and ColdFire targets now allow the default processor to be | |
| 1250 set at configure time using --with-cpu. | |
| 1251 * A --with-arch configuration option has been added. This option | |
| 1252 allows you to restrict a target to ColdFire or non-ColdFire | |
| 1253 processors. | |
| 1254 | |
| 1255 Preprocessor macros | |
| 1256 | |
| 1257 * An __mcfv*__ macro is now defined for all ColdFire targets. | |
| 1258 (Earlier versions of GCC only defined __mcfv4e__.) | |
| 1259 * __mcf_cpu_*, __mcf_family_* and __mcffpu__ macros have been added. | |
| 1260 * All targets now define __mc68010 and __mc68010__ when generating | |
| 1261 68010 code. | |
| 1262 | |
| 1263 Command-line changes | |
| 1264 | |
| 1265 * New command-line options -march, -mcpu, -mtune and -mhard-float | |
| 1266 have been added. These options apply to both m68k and ColdFire | |
| 1267 targets. | |
| 1268 * -mno-short, -mno-bitfield and -mno-rtd are now accepted as negative | |
| 1269 versions of -mshort, etc. | |
| 1270 * -fforce-addr has been removed. It is now ignored by the compiler. | |
| 1271 | |
| 1272 Other improvements | |
| 1273 | |
| 1274 * ColdFire targets now try to maintain a 4-byte-aligned stack where | |
| 1275 possible. | |
| 1276 * m68k-uclinux targets now try to avoid situations that lead to the | |
| 1277 load-time error: BINFMT_FLAT: reloc outside program. | |
| 1278 | |
| 1279 MIPS | |
| 1280 | |
| 1281 Changes to existing configurations | |
| 1282 | |
| 1283 * libffi and libjava now support all three GNU/Linux ABIs: o32, n32 | |
| 1284 and n64. Every GNU/Linux configuration now builds these libraries | |
| 1285 by default. | |
| 1286 * GNU/Linux configurations now generate -mno-shared code unless | |
| 1287 overridden by -fpic, -fPIC, -fpie or -fPIE. | |
| 1288 * mipsisa32*-linux-gnu configurations now generate hard-float code by | |
| 1289 default, just like other mipsisa32* and mips*-linux-gnu | |
| 1290 configurations. You can build a soft-float version of any | |
| 1291 mips*-linux-gnu configuration by passing --with-float=soft to | |
| 1292 configure. | |
| 1293 * mips-wrs-vxworks now supports run-time processes (RTPs). | |
| 1294 | |
| 1295 Changes to existing command-line options | |
| 1296 | |
| 1297 * The -march and -mtune options no longer accept 24k as a processor | |
| 1298 name. Please use 24kc, 24kf2_1 or 24kf1_1 instead. | |
| 1299 * The -march and -mtune options now accept 24kf2_1, 24kef2_1 and | |
| 1300 34kf2_1 as synonyms for 24kf, 24kef and 34kf respectively. The | |
| 1301 options also accept 24kf1_1, 24kef1_1 and 34kf1_1 as synonyms for | |
| 1302 24kx, 24kex and 34kx. | |
| 1303 | |
| 1304 New configurations | |
| 1305 | |
| 1306 GCC now supports the following configurations: | |
| 1307 * mipsisa32r2*-linux-gnu*, which generates MIPS32 revision 2 code by | |
| 1308 default. Earlier releases also recognized this configuration, but | |
| 1309 they treated it in the same way as mipsisa32*-linux-gnu*. Note that | |
| 1310 you can customize any mips*-linux-gnu* configuration to a | |
| 1311 particular ISA or processor by passing an appropriate --with-arch | |
| 1312 option to configure. | |
| 1313 * mipsisa*-sde-elf*, which provides compatibility with MIPS | |
| 1314 Technologies' SDE toolchains. The configuration uses the SDE | |
| 1315 libraries by default, but you can use it like other newlib-based | |
| 1316 ELF configurations by passing --with-newlib to configure. It is the | |
| 1317 only configuration besides mips64vr*-elf* to build MIPS16 as well | |
| 1318 as non-MIPS16 libraries. | |
| 1319 * mipsisa*-elfoabi*, which is similar to the general mipsisa*-elf* | |
| 1320 configuration, but uses the o32 and o64 ABIs instead of the 32-bit | |
| 1321 and 64-bit forms of the EABI. | |
| 1322 | |
| 1323 New processors and application-specific extensions | |
| 1324 | |
| 1325 * Support for the SmartMIPS ASE is available through the new | |
| 1326 -msmartmips option. | |
| 1327 * Support for revision 2 of the DSP ASE is available through the new | |
| 1328 -mdspr2 option. A new preprocessor macro called __mips_dsp_rev | |
| 1329 indicates the revision of the ASE in use. | |
| 1330 * Support for the 4KS and 74K families of processors is available | |
| 1331 through the -march and -mtune options. | |
| 1332 | |
| 1333 Improved support for built-in functions | |
| 1334 | |
| 1335 * GCC can now use load-linked, store-conditional and sync | |
| 1336 instructions to implement atomic built-in functions such as | |
| 1337 __sync_fetch_and_add. The memory reference must be 4 bytes wide for | |
| 1338 32-bit targets and either 4 or 8 bytes wide for 64-bit targets. | |
| 1339 * GCC can now use the clz and dclz instructions to implement the | |
| 1340 __builtin_ctz and __builtin_ffs families of functions. | |
| 1341 * There is a new __builtin___clear_cache function for flushing the | |
| 1342 instruction cache. GCC expands this function inline on MIPS32 | |
| 1343 revision 2 targets, otherwise it calls the function specified by | |
| 1344 -mcache-flush-func. | |
| 1345 | |
| 1346 MIPS16 improvements | |
| 1347 | |
| 1348 * GCC can now compile objects that contain a mixture of MIPS16 and | |
| 1349 non-MIPS16 code. There are two new attributes, mips16 and nomips16, | |
| 1350 for specifying which mode a function should use. | |
| 1351 * A new option called -minterlink-mips16 makes non-MIPS16 code | |
| 1352 link-compatible with MIPS16 code. | |
| 1353 * After many bug fixes, the long-standing MIPS16 -mhard-float support | |
| 1354 should now work fairly reliably. | |
| 1355 * GCC can now use the MIPS16e save and restore instructions. | |
| 1356 * -fsection-anchors now works in MIPS16 mode. MIPS16 code compiled | |
| 1357 with -G0 -fsection-anchors is often smaller than code compiled with | |
| 1358 -G8. However, please note that you must usually compile all objects | |
| 1359 in your application with the same -G option; see the documentation | |
| 1360 of -G for details. | |
| 1361 * A new option called-mcode-readable specifies which instructions are | |
| 1362 allowed to load from the code segment. -mcode-readable=yes is the | |
| 1363 default and says that any instruction may load from the code | |
| 1364 segment. The other alternatives are -mcode-readable=pcrel, which | |
| 1365 says that only PC-relative MIPS16 instructions may load from the | |
| 1366 code segment, and -mcode-readable=no, which says that no | |
| 1367 instruction may do so. Please see the documentation for more | |
| 1368 details, including example uses. | |
| 1369 | |
| 1370 Small-data improvements | |
| 1371 | |
| 1372 There are three new options for controlling small data: | |
| 1373 * -mno-extern-sdata, which disables small-data accesses for | |
| 1374 externally-defined variables. Code compiled with -Gn | |
| 1375 -mno-extern-sdata will be link-compatible with any -G setting | |
| 1376 between -G0 and -Gn inclusive. | |
| 1377 * -mno-local-sdata, which disables the use of small-data sections for | |
| 1378 data that is not externally visible. This option can be a useful | |
| 1379 way of reducing small-data usage in less performance-critical parts | |
| 1380 of an application. | |
| 1381 * -mno-gpopt, which disables the use of the $gp register while still | |
| 1382 honoring the -G limit when placing externally-visible data. This | |
| 1383 option implies -mno-extern-sdata and -mno-local-sdata and it can be | |
| 1384 useful in situations where $gp does not necessarily hold the | |
| 1385 expected value. | |
| 1386 | |
| 1387 Miscellaneous improvements | |
| 1388 | |
| 1389 * There is a new option called -mbranch-cost for tweaking the | |
| 1390 perceived cost of branches. | |
| 1391 * If GCC is configured to use a version of GAS that supports the | |
| 1392 .gnu_attribute directive, it will use that directive to record | |
| 1393 certain properties of the output code. .gnu_attribute is new to GAS | |
| 1394 2.18. | |
| 1395 * There are two new function attributes, near and far, for overriding | |
| 1396 the command-line setting of -mlong-calls on a function-by-function | |
| 1397 basis. | |
| 1398 * -mfp64, which previously required a 64-bit target, now works with | |
| 1399 MIPS32 revision 2 targets as well. The mipsisa*-elfoabi* and | |
| 1400 mipsisa*-sde-elf* configurations provide suitable library support. | |
| 1401 * GCC now recognizes the -mdmx and -mmt options and passes them down | |
| 1402 to the assembler. It does nothing else with the options at present. | |
| 1403 | |
| 1404 SPU (Synergistic Processor Unit) of the Cell Broadband Engine Architecture | |
| 1405 (BEA) | |
| 1406 | |
| 1407 * Support has been added for this new architecture. | |
| 1408 | |
| 1409 RS6000 (POWER/PowerPC) | |
| 1410 | |
| 1411 * Support for the PowerPC 750CL paired-single instructions has been | |
| 1412 added with a new powerpc-*-linux*paired* target configuration. It | |
| 1413 is enabled by an associated -mpaired option and can be accessed | |
| 1414 using new built-in functions. | |
| 1415 * Support for auto-detecting architecture and system configuration to | |
| 1416 auto-select processor optimization tuning. | |
| 1417 * Support for VMX on AIX 5.3 has been added. | |
| 1418 * Support for AIX Version 6.1 has been added. | |
| 1419 | |
| 1420 S/390, zSeries and System z9 | |
| 1421 | |
| 1422 * Support for the IBM System z9 EC/BC processor (z9 GA3) has been | |
| 1423 added. When using the -march=z9-ec option, the compiler will | |
| 1424 generate code making use of instructions provided by the decimal | |
| 1425 floating point facility and the floating point conversion facility | |
| 1426 (pfpo). Besides the instructions used to implement decimal floating | |
| 1427 point operations these facilities also contain instructions to move | |
| 1428 between general purpose and floating point registers and to modify | |
| 1429 and copy the sign-bit of floating point values. | |
| 1430 * When the -march=z9-ec option is used the new | |
| 1431 -mhard-dfp/-mno-hard-dfp options can be used to specify whether the | |
| 1432 decimal floating point hardware instructions will be used or not. | |
| 1433 If none of them is given the hardware support is enabled by | |
| 1434 default. | |
| 1435 * The -mstack-guard option can now be omitted when using stack | |
| 1436 checking via -mstack-size in order to let GCC choose a sensible | |
| 1437 stack guard value according to the frame size of each function. | |
| 1438 * Various changes to improve performance of generated code have been | |
| 1439 implemented, including: | |
| 1440 + The condition code set by an add logical with carry | |
| 1441 instruction is now available for overflow checks like: a + b + | |
| 1442 carry < b. | |
| 1443 + The test data class instruction is now used to implement | |
| 1444 sign-bit and infinity checks of binary and decimal floating | |
| 1445 point numbers. | |
| 1446 | |
| 1447 Xtensa | |
| 1448 | |
| 1449 * Stack unwinding for exception handling now uses by default a | |
| 1450 specialized version of DWARF unwinding. This is not | |
| 1451 binary-compatible with the setjmp/longjmp (sjlj) unwinding used for | |
| 1452 Xtensa with previous versions of GCC. | |
| 1453 * For Xtensa processors that include the Conditional Store option, | |
| 1454 the built-in functions for atomic memory access are now implemented | |
| 1455 using S32C1I instructions. | |
| 1456 * If the Xtensa NSA option is available, GCC will use it to implement | |
| 1457 the __builtin_ctz and __builtin_clz functions. | |
| 1458 | |
| 1459 Documentation improvements | |
| 1460 | |
| 1461 * Existing libstdc++ documentation has been edited and restructured | |
| 1462 into a single DocBook XML manual. The results can be viewed online | |
| 1463 [24]here. | |
| 1464 | |
| 1465 Other significant improvements | |
| 1466 | |
| 1467 * The compiler's --help command-line option has been extended so that | |
| 1468 it now takes an optional set of arguments. These arguments restrict | |
| 1469 the information displayed to specific classes of command-line | |
| 1470 options, and possibly only a subset of those options. It is also | |
| 1471 now possible to replace the descriptive text associated with each | |
| 1472 displayed option with an indication of its current value, or for | |
| 1473 binary options, whether it has been enabled or disabled. | |
| 1474 Here are some examples. The following will display all the options | |
| 1475 controlling warning messages: | |
| 1476 --help=warnings | |
| 1477 | |
| 1478 Whereas this will display all the undocumented, target specific | |
| 1479 options: | |
| 1480 --help=target,undocumented | |
| 1481 | |
| 1482 This sequence of commands will display the binary optimizations | |
| 1483 that are enabled by -O3: | |
| 1484 gcc -c -Q -O3 --help=optimizers > /tmp/O3-opts | |
| 1485 gcc -c -Q -O2 --help=optimizers > /tmp/O2-opts | |
| 1486 diff /tmp/O2-opts /tmp/O3-opts | grep enabled | |
| 1487 | |
| 1488 * The configure options --with-pkgversion and --with-bugurl have been | |
| 1489 added. These allow distributors of GCC to include a | |
| 1490 distributor-specific string in manuals and --version output and to | |
| 1491 specify the URL for reporting bugs in their versions of GCC. | |
| 1492 | |
| 1493 GCC 4.3.1 | |
| 1494 | |
| 1495 This is the [25]list of problem reports (PRs) from GCC's bug tracking | |
| 1496 system that are known to be fixed in the 4.3.1 release. This list might | |
| 1497 not be complete (that is, it is possible that some PRs that have been | |
| 1498 fixed are not listed here). | |
| 1499 | |
| 1500 Target Specific Changes | |
| 1501 | |
| 1502 IA-32/x86-64 | |
| 1503 | |
| 1504 ABI changes | |
| 1505 | |
| 1506 * Starting with GCC 4.3.1, decimal floating point variables are | |
| 1507 aligned to their natural boundaries when they are passed on the | |
| 1508 stack for i386. | |
| 1509 | |
| 1510 Command-line changes | |
| 1511 | |
| 1512 * Starting with GCC 4.3.1, the -mcld option has been added to | |
| 1513 automatically generate a cld instruction in the prologue of | |
| 1514 functions that use string instructions. This option is used for | |
| 1515 backward compatibility on some operating systems and can be enabled | |
| 1516 by default for 32-bit x86 targets by configuring GCC with the | |
| 1517 --enable-cld configure option. | |
| 1518 | |
| 1519 GCC 4.3.2 | |
| 1520 | |
| 1521 This is the [26]list of problem reports (PRs) from GCC's bug tracking | |
| 1522 system that are known to be fixed in the 4.3.2 release. This list might | |
| 1523 not be complete (that is, it is possible that some PRs that have been | |
| 1524 fixed are not listed here). | |
| 1525 | |
| 1526 GCC 4.3.3 | |
| 1527 | |
| 1528 This is the [27]list of problem reports (PRs) from GCC's bug tracking | |
| 1529 system that are known to be fixed in the 4.3.3 release. This list might | |
| 1530 not be complete (that is, it is possible that some PRs that have been | |
| 1531 fixed are not listed here). | |
| 1532 | |
|
47
3bfb6c00c1e0
update it from 4.4.2 to 4.4.3.
kent <kent@cr.ie.u-ryukyu.ac.jp>
parents:
19
diff
changeset
|
1533 GCC 4.3.4 |
|
3bfb6c00c1e0
update it from 4.4.2 to 4.4.3.
kent <kent@cr.ie.u-ryukyu.ac.jp>
parents:
19
diff
changeset
|
1534 |
|
3bfb6c00c1e0
update it from 4.4.2 to 4.4.3.
kent <kent@cr.ie.u-ryukyu.ac.jp>
parents:
19
diff
changeset
|
1535 This is the [28]list of problem reports (PRs) from GCC's bug tracking |
|
3bfb6c00c1e0
update it from 4.4.2 to 4.4.3.
kent <kent@cr.ie.u-ryukyu.ac.jp>
parents:
19
diff
changeset
|
1536 system that are known to be fixed in the 4.3.4 release. This list might |
|
3bfb6c00c1e0
update it from 4.4.2 to 4.4.3.
kent <kent@cr.ie.u-ryukyu.ac.jp>
parents:
19
diff
changeset
|
1537 not be complete (that is, it is possible that some PRs that have been |
|
3bfb6c00c1e0
update it from 4.4.2 to 4.4.3.
kent <kent@cr.ie.u-ryukyu.ac.jp>
parents:
19
diff
changeset
|
1538 fixed are not listed here). |
|
3bfb6c00c1e0
update it from 4.4.2 to 4.4.3.
kent <kent@cr.ie.u-ryukyu.ac.jp>
parents:
19
diff
changeset
|
1539 |
|
3bfb6c00c1e0
update it from 4.4.2 to 4.4.3.
kent <kent@cr.ie.u-ryukyu.ac.jp>
parents:
19
diff
changeset
|
1540 Please send FSF & GNU inquiries & questions to [29]gnu@gnu.org. There |
|
3bfb6c00c1e0
update it from 4.4.2 to 4.4.3.
kent <kent@cr.ie.u-ryukyu.ac.jp>
parents:
19
diff
changeset
|
1541 are also [30]other ways to contact the FSF. |
|
3bfb6c00c1e0
update it from 4.4.2 to 4.4.3.
kent <kent@cr.ie.u-ryukyu.ac.jp>
parents:
19
diff
changeset
|
1542 |
|
3bfb6c00c1e0
update it from 4.4.2 to 4.4.3.
kent <kent@cr.ie.u-ryukyu.ac.jp>
parents:
19
diff
changeset
|
1543 These pages are maintained by [31]the GCC team. |
| 0 | 1544 |
| 1545 | |
| 1546 For questions related to the use of GCC, please consult these web | |
|
47
3bfb6c00c1e0
update it from 4.4.2 to 4.4.3.
kent <kent@cr.ie.u-ryukyu.ac.jp>
parents:
19
diff
changeset
|
1547 pages and the [32]GCC manuals. If that fails, the |
|
3bfb6c00c1e0
update it from 4.4.2 to 4.4.3.
kent <kent@cr.ie.u-ryukyu.ac.jp>
parents:
19
diff
changeset
|
1548 [33]gcc-help@gcc.gnu.org mailing list might help. |
| 0 | 1549 Please send comments on these web pages and the development of GCC to |
|
47
3bfb6c00c1e0
update it from 4.4.2 to 4.4.3.
kent <kent@cr.ie.u-ryukyu.ac.jp>
parents:
19
diff
changeset
|
1550 our developer mailing list at [34]gcc@gnu.org or [35]gcc@gcc.gnu.org. |
|
3bfb6c00c1e0
update it from 4.4.2 to 4.4.3.
kent <kent@cr.ie.u-ryukyu.ac.jp>
parents:
19
diff
changeset
|
1551 All of our lists have [36]public archives. |
| 0 | 1552 |
| 1553 Copyright (C) Free Software Foundation, Inc., 51 Franklin St, Fifth | |
| 1554 Floor, Boston, MA 02110, USA. | |
| 1555 | |
| 1556 Verbatim copying and distribution of this entire article is permitted | |
| 1557 in any medium, provided this notice is preserved. | |
|
47
3bfb6c00c1e0
update it from 4.4.2 to 4.4.3.
kent <kent@cr.ie.u-ryukyu.ac.jp>
parents:
19
diff
changeset
|
1558 Last modified 2009-08-13 [37]Valid XHTML 1.0 |
| 0 | 1559 |
| 1560 References | |
| 1561 | |
|
47
3bfb6c00c1e0
update it from 4.4.2 to 4.4.3.
kent <kent@cr.ie.u-ryukyu.ac.jp>
parents:
19
diff
changeset
|
1562 1. http://gcc.gnu.org/gcc-4.3/changes.html#4.3.4 |
| 0 | 1563 2. http://gmplib.org/ |
| 1564 3. http://www.mpfr.org/ | |
| 1565 4. http://gcc.gnu.org/install/prerequisites.html | |
| 1566 5. http://gcc.gnu.org/ml/gcc-announce/2001/msg00000.html | |
| 1567 6. http://gcc.gnu.org/onlinedocs/gcc/Warning-Options.html#Warning-Options | |
| 1568 7. http://gcc.gnu.org/gcc-4.3/porting_to.html | |
| 1569 8. http://www.mpfr.org/ | |
| 1570 9. http://www.mpfr.org/ | |
| 1571 10. http://www.mpfr.org/ | |
| 1572 11. http://gcc.gnu.org/onlinedocs/gcc/Warning-Options.html | |
| 1573 12. http://gcc.gnu.org/gcc-4.3/cxx0x_status.html | |
| 1574 13. http://gcc.gnu.org/gcc-4.3/cxx0x_status.html | |
| 1575 14. http://gcc.gnu.org/onlinedocs/libstdc++/manual/bk01pt01ch01.html#m anual.intro.status.standard.tr1 | |
| 1576 15. http://gcc.gnu.org/onlinedocs/libstdc++/manual/parallel_mode.html | |
| 1577 16. http://gmplib.org/ | |
| 1578 17. http://www.mpfr.org/ | |
| 1579 18. http://gcc.gnu.org/onlinedocs/gfortran/Code-Gen-Options.html#Code-Gen-Options | |
| 1580 19. http://gcc.gnu.org/onlinedocs/gfortran/Code-Gen-Options.html#index-g_t_0040code_007bfinit-local-zero_007d-167 | |
| 1581 20. http://gcc.gnu.org/onlinedocs/gcc-4.3.0/gfortran/GAMMA.html | |
| 1582 21. http://gcc.gnu.org/onlinedocs/gcc-4.3.0/gfortran/LGAMMA.html | |
| 1583 22. http://gcc.gnu.org/onlinedocs/gfortran/Fortran-Dialect-Options.html | |
| 1584 23. http://gcc.gnu.org/onlinedocs/gfortran/BOZ-literal-constants.html | |
| 1585 24. http://gcc.gnu.org/onlinedocs/libstdc++/ | |
| 1586 25. http://gcc.gnu.org/bugzilla/buglist.cgi?bug_status=RESOLVED&resolution=FIXED&target_milestone=4.3.1 | |
| 1587 26. http://gcc.gnu.org/bugzilla/buglist.cgi?bug_status=RESOLVED&resolution=FIXED&target_milestone=4.3.2 | |
| 1588 27. http://gcc.gnu.org/bugzilla/buglist.cgi?bug_status=RESOLVED&resolution=FIXED&target_milestone=4.3.3 | |
|
47
3bfb6c00c1e0
update it from 4.4.2 to 4.4.3.
kent <kent@cr.ie.u-ryukyu.ac.jp>
parents:
19
diff
changeset
|
1589 28. http://gcc.gnu.org/bugzilla/buglist.cgi?bug_status=RESOLVED&resolution=FIXED&target_milestone=4.3.4 |
|
3bfb6c00c1e0
update it from 4.4.2 to 4.4.3.
kent <kent@cr.ie.u-ryukyu.ac.jp>
parents:
19
diff
changeset
|
1590 29. mailto:gnu@gnu.org |
|
3bfb6c00c1e0
update it from 4.4.2 to 4.4.3.
kent <kent@cr.ie.u-ryukyu.ac.jp>
parents:
19
diff
changeset
|
1591 30. http://www.gnu.org/home.html#ContactInfo |
|
3bfb6c00c1e0
update it from 4.4.2 to 4.4.3.
kent <kent@cr.ie.u-ryukyu.ac.jp>
parents:
19
diff
changeset
|
1592 31. http://gcc.gnu.org/about.html |
|
3bfb6c00c1e0
update it from 4.4.2 to 4.4.3.
kent <kent@cr.ie.u-ryukyu.ac.jp>
parents:
19
diff
changeset
|
1593 32. http://gcc.gnu.org/onlinedocs/ |
|
3bfb6c00c1e0
update it from 4.4.2 to 4.4.3.
kent <kent@cr.ie.u-ryukyu.ac.jp>
parents:
19
diff
changeset
|
1594 33. mailto:gcc-help@gcc.gnu.org |
|
3bfb6c00c1e0
update it from 4.4.2 to 4.4.3.
kent <kent@cr.ie.u-ryukyu.ac.jp>
parents:
19
diff
changeset
|
1595 34. mailto:gcc@gnu.org |
|
3bfb6c00c1e0
update it from 4.4.2 to 4.4.3.
kent <kent@cr.ie.u-ryukyu.ac.jp>
parents:
19
diff
changeset
|
1596 35. mailto:gcc@gcc.gnu.org |
|
3bfb6c00c1e0
update it from 4.4.2 to 4.4.3.
kent <kent@cr.ie.u-ryukyu.ac.jp>
parents:
19
diff
changeset
|
1597 36. http://gcc.gnu.org/lists.html |
|
3bfb6c00c1e0
update it from 4.4.2 to 4.4.3.
kent <kent@cr.ie.u-ryukyu.ac.jp>
parents:
19
diff
changeset
|
1598 37. http://validator.w3.org/check/referer |
| 0 | 1599 ====================================================================== |
| 1600 http://gcc.gnu.org/gcc-4.2/index.html | |
| 1601 GCC 4.2 Release Series | |
| 1602 | |
| 1603 May 19, 2008 | |
| 1604 | |
| 1605 The [1]GNU project and the GCC developers are pleased to announce the | |
| 1606 release of GCC 4.2.4. | |
| 1607 | |
| 1608 This release is a bug-fix release, containing fixes for regressions in | |
| 1609 GCC 4.2.3 relative to previous releases of GCC. | |
| 1610 | |
| 1611 Release History | |
| 1612 | |
| 1613 GCC 4.2.4 | |
| 1614 May 19, 2008 ([2]changes) | |
| 1615 | |
| 1616 GCC 4.2.3 | |
| 1617 February 1, 2008 ([3]changes) | |
| 1618 | |
| 1619 GCC 4.2.2 | |
| 1620 October 7, 2007 ([4]changes) | |
| 1621 | |
| 1622 GCC 4.2.1 | |
| 1623 July 18, 2007 ([5]changes) | |
| 1624 | |
| 1625 GCC 4.2.0 | |
| 1626 May 13, 2007 ([6]changes) | |
| 1627 | |
| 1628 References and Acknowledgements | |
| 1629 | |
| 1630 GCC used to stand for the GNU C Compiler, but since the compiler | |
| 1631 supports several other languages aside from C, it now stands for the | |
| 1632 GNU Compiler Collection. | |
| 1633 | |
| 1634 A list of [7]successful builds is updated as new information becomes | |
| 1635 available. | |
| 1636 | |
| 1637 The GCC developers would like to thank the numerous people that have | |
| 1638 contributed new features, improvements, bug fixes, and other changes as | |
| 1639 well as test results to GCC. This [8]amazing group of volunteers is | |
| 1640 what makes GCC successful. | |
| 1641 | |
| 1642 For additional information about GCC please refer to the [9]GCC project | |
| 1643 web site or contact the [10]GCC development mailing list. | |
| 1644 | |
| 1645 To obtain GCC please use [11]our mirror sites or [12]our SVN server. | |
| 1646 | |
| 1647 Please send FSF & GNU inquiries & questions to [13]gnu@gnu.org. There | |
| 1648 are also [14]other ways to contact the FSF. | |
| 1649 | |
| 1650 These pages are maintained by [15]the GCC team. | |
| 1651 | |
| 1652 | |
| 1653 For questions related to the use of GCC, please consult these web | |
| 1654 pages and the [16]GCC manuals. If that fails, the | |
| 1655 [17]gcc-help@gcc.gnu.org mailing list might help. | |
| 1656 Please send comments on these web pages and the development of GCC to | |
| 1657 our developer mailing list at [18]gcc@gnu.org or [19]gcc@gcc.gnu.org. | |
| 1658 All of our lists have [20]public archives. | |
| 1659 | |
| 1660 Copyright (C) Free Software Foundation, Inc., 51 Franklin St, Fifth | |
| 1661 Floor, Boston, MA 02110, USA. | |
| 1662 | |
| 1663 Verbatim copying and distribution of this entire article is permitted | |
| 1664 in any medium, provided this notice is preserved. | |
| 1665 Last modified 2008-10-04 [21]Valid XHTML 1.0 | |
| 1666 | |
| 1667 References | |
| 1668 | |
| 1669 1. http://www.gnu.org/ | |
| 1670 2. http://gcc.gnu.org/gcc-4.2/changes.html | |
| 1671 3. http://gcc.gnu.org/gcc-4.2/changes.html | |
| 1672 4. http://gcc.gnu.org/gcc-4.2/changes.html | |
| 1673 5. http://gcc.gnu.org/gcc-4.2/changes.html | |
| 1674 6. http://gcc.gnu.org/gcc-4.2/changes.html | |
| 1675 7. http://gcc.gnu.org/gcc-4.2/buildstat.html | |
| 1676 8. http://gcc.gnu.org/onlinedocs/gcc/Contributors.html | |
| 1677 9. http://gcc.gnu.org/index.html | |
| 1678 10. mailto:gcc@gcc.gnu.org | |
| 1679 11. http://gcc.gnu.org/mirrors.html | |
| 1680 12. http://gcc.gnu.org/svn.html | |
| 1681 13. mailto:gnu@gnu.org | |
| 1682 14. http://www.gnu.org/home.html#ContactInfo | |
| 1683 15. http://gcc.gnu.org/about.html | |
| 1684 16. http://gcc.gnu.org/onlinedocs/ | |
| 1685 17. mailto:gcc-help@gcc.gnu.org | |
| 1686 18. mailto:gcc@gnu.org | |
| 1687 19. mailto:gcc@gcc.gnu.org | |
| 1688 20. http://gcc.gnu.org/lists.html | |
| 1689 21. http://validator.w3.org/check/referer | |
| 1690 ====================================================================== | |
| 1691 http://gcc.gnu.org/gcc-4.2/changes.html | |
| 1692 GCC 4.2 Release Series | |
| 1693 Changes, New Features, and Fixes | |
| 1694 | |
| 1695 Caveats | |
| 1696 | |
| 1697 * GCC no longer accepts the -fshared-data option. This option has had | |
| 1698 no effect in any GCC 4 release; the targets to which the option | |
| 1699 used to apply had been removed before GCC 4.0. | |
| 1700 | |
| 1701 General Optimizer Improvements | |
| 1702 | |
| 1703 * New command-line options specify the possible relationships among | |
| 1704 parameters and between parameters and global data. For example, | |
| 1705 -fargument-noalias-anything specifies that arguments do not alias | |
| 1706 any other storage. | |
| 1707 Each language will automatically use whatever option is required by | |
| 1708 the language standard. You should not need to use these options | |
| 1709 yourself. | |
| 1710 | |
| 1711 New Languages and Language specific improvements | |
| 1712 | |
| 1713 * [1]OpenMP is now supported for the C, C++ and Fortran compilers. | |
| 1714 * New command-line options -fstrict-overflow and -Wstrict-overflow | |
| 1715 have been added. -fstrict-overflow tells the compiler that it may | |
| 1716 assume that the program follows the strict signed overflow | |
| 1717 semantics permitted for the language: for C and C++ this means that | |
| 1718 the compiler may assume that signed overflow does not occur. For | |
| 1719 example, a loop like | |
| 1720 for (i = 1; i > 0; i *= 2) | |
| 1721 | |
| 1722 is presumably intended to continue looping until i overflows. With | |
| 1723 -fstrict-overflow, the compiler may assume that signed overflow | |
| 1724 will not occur, and transform this into an infinite loop. | |
| 1725 -fstrict-overflow is turned on by default at -O2, and may be | |
| 1726 disabled via -fno-strict-overflow. The -Wstrict-overflow option may | |
| 1727 be used to warn about cases where the compiler assumes that signed | |
| 1728 overflow will not occur. It takes five different levels: | |
| 1729 -Wstrict-overflow=1 to 5. See the [2]documentation for details. | |
| 1730 -Wstrict-overflow=1 is enabled by -Wall. | |
| 1731 * The new command-line option -fno-toplevel-reorder directs GCC to | |
| 1732 emit top-level functions, variables, and asm statements in the same | |
| 1733 order that they appear in the input file. This is intended to | |
| 1734 support existing code which relies on a particular ordering (for | |
| 1735 example, code which uses top-level asm statements to switch | |
| 1736 sections). For new code, it is generally better to use function and | |
| 1737 variable attributes. The -fno-toplevel-reorder option may be used | |
| 1738 for most cases which currently use -fno-unit-at-a-time. The | |
| 1739 -fno-unit-at-a-time option will be removed in some future version | |
| 1740 of GCC. If you know of a case which requires -fno-unit-at-a-time | |
|
47
3bfb6c00c1e0
update it from 4.4.2 to 4.4.3.
kent <kent@cr.ie.u-ryukyu.ac.jp>
parents:
19
diff
changeset
|
1741 which is not fixed by -fno-toplevel-reorder, please open a bug |
| 0 | 1742 report. |
| 1743 | |
| 1744 C family | |
| 1745 | |
| 1746 * The pragma redefine_extname will now macro expand its tokens for | |
| 1747 compatibility with SunPRO. | |
| 1748 * In the next release of GCC, 4.3, -std=c99 or -std=gnu99 will direct | |
| 1749 GCC to handle inline functions as specified in the C99 standard. In | |
| 1750 preparation for this, GCC 4.2 will warn about any use of non-static | |
| 1751 inline functions in gnu99 or c99 mode. This new warning may be | |
| 1752 disabled with the new gnu_inline function attribute or the new | |
| 1753 -fgnu89-inline command-line option. Also, GCC 4.2 and later will | |
| 1754 define one of the preprocessor macros __GNUC_GNU_INLINE__ or | |
| 1755 __GNUC_STDC_INLINE__ to indicate the semantics of inline functions | |
| 1756 in the current compilation. | |
| 1757 * A new command-line option -Waddress has been added to warn about | |
| 1758 suspicious uses of memory addresses as, for example, using the | |
| 1759 address of a function in a conditional expression, and comparisons | |
| 1760 against the memory address of a string literal. This warning is | |
| 1761 enabled by -Wall. | |
| 1762 | |
| 1763 C++ | |
| 1764 | |
| 1765 * C++ visibility handling has been overhauled. | |
| 1766 Restricted visiblity is propagated from classes to members, from | |
| 1767 functions to local statics, and from templates and template | |
| 1768 arguments to instantiations, unless the latter has explicitly | |
| 1769 declared visibility. | |
| 1770 The visibility attribute for a class must come between the | |
| 1771 class-key and the name, not after the closing brace. | |
| 1772 Attributes are now allowed for enums and elaborated-type-specifiers | |
| 1773 that only declare a type. | |
| 1774 Members of the anonymous namespace are now local to a particular | |
| 1775 translation unit, along with any other declarations which use them, | |
| 1776 though they are still treated as having external linkage for | |
| 1777 language semantics. | |
| 1778 * The (undocumented) extension which permitted templates with default | |
| 1779 arguments to be bound to template template parameters with fewer | |
| 1780 parameters has been removed. For example: | |
| 1781 template <template <typename> class C> | |
| 1782 void f(C<double>) {} | |
| 1783 | |
| 1784 template <typename T, typename U = int> | |
| 1785 struct S {}; | |
| 1786 | |
| 1787 template void f(S<double>); | |
| 1788 | |
| 1789 is no longer accepted by G++. The reason this code is not accepted | |
| 1790 is that S is a template with two parameters; therefore, it cannot | |
| 1791 be bound to C which has only one parameter. | |
| 1792 * The <?, >?, <?=, and >?= operators, deprecated in previous GCC | |
| 1793 releases, have been removed. | |
| 1794 * The command-line option -fconst-strings, deprecated in previous GCC | |
| 1795 releases, has been removed. | |
| 1796 * The configure variable enable-__cxa_atexit is now enabled by | |
| 1797 default for more targets. Enabling this variable is necessary in | |
| 1798 order for static destructors to be executed in the correct order, | |
| 1799 but it depends upon the presence of a non-standard C library in the | |
| 1800 target library in order to work. The variable is now enabled for | |
| 1801 more targets which are known to have suitable C libraries. | |
| 1802 * -Wextra will produce warnings for if statements with a semicolon as | |
| 1803 the only body, to catch code like: | |
| 1804 if (a); | |
| 1805 return 1; | |
| 1806 return 0; | |
| 1807 | |
| 1808 To suppress the warning in valid cases, use { } instead. | |
| 1809 * The C++ frontend now also produces strict aliasing warnings when | |
| 1810 -fstrict-aliasing -Wstrict-aliasing is in effect. | |
| 1811 | |
| 1812 Runtime Library (libstdc++) | |
| 1813 | |
| 1814 * Added support for TR1 <random>, <complex>, and C compatibility | |
| 1815 headers. In addition, a lock-free version of shared_ptr was | |
| 1816 contributed as part of Phillip Jordan's Google Summer of Code | |
|
47
3bfb6c00c1e0
update it from 4.4.2 to 4.4.3.
kent <kent@cr.ie.u-ryukyu.ac.jp>
parents:
19
diff
changeset
|
1817 project on lock-free containers. ([3]Implementation status of TR1) |
| 0 | 1818 * In association with the Summer of Code work on lock-free |
| 1819 containers, the interface for atomic builtins was adjusted, | |
| 1820 creating simpler alternatives for non-threaded code paths. Also, | |
| 1821 usage was consolidated and all elements were moved from namespace | |
| 1822 std to namespace__gnu_cxx. Affected interfaces are the functions | |
| 1823 __exchange_and_add, __atomic_add, and the objects __mutex, | |
| 1824 __recursive_mutex, and __scoped_lock. | |
| 1825 * Support for versioning weak symbol names via namespace association | |
| 1826 was added. However, as this changes the names of exported symbols, | |
| 1827 this is turned off by default in the current ABI. Intrepid users | |
| 1828 can enable this feature by using | |
| 1829 --enable-symvers=gnu-versioned-namespace during configuration. | |
| 1830 * Revised, simplified, and expanded policy-based associative | |
| 1831 containers, including data types for tree and trie forms | |
| 1832 (basic_tree, tree, trie), lists (list_update), and both | |
| 1833 collision-chaining and probing hash-based containers | |
| 1834 (basic_hash_table, cc_hash_table, gp_hash_table). More details per | |
|
47
3bfb6c00c1e0
update it from 4.4.2 to 4.4.3.
kent <kent@cr.ie.u-ryukyu.ac.jp>
parents:
19
diff
changeset
|
1835 the [4]documentation. |
| 0 | 1836 * The implementation of the debug mode was modified, whereby the |
| 1837 debug namespaces were nested inside of namespace std and namespace | |
| 1838 __gnu_cxx in order to resolve some long standing corner cases | |
| 1839 involving name lookup. Debug functionality from the policy-based | |
| 1840 data structures was consolidated and enabled with the single macro, | |
| 1841 _GLIBCXX_DEBUG. See PR 26142 for more information. | |
| 1842 * Added extensions for type traits: __conditional_type, | |
| 1843 __numeric_traits, __add_unsigned, __removed_unsigned, __enable_if. | |
| 1844 * Added a typelist implementation for compile-time meta-programming. | |
| 1845 Elements for typelist construction and operation can be found | |
| 1846 within namespace __gnu_cxx::typelist. | |
| 1847 * Added a new allocator, __gnu_cxx::throw_allocator, for testing | |
| 1848 exception-safety. | |
| 1849 * Enabled library-wide visibility control, allowing -fvisibility to | |
| 1850 be used. | |
| 1851 * Consolidated all nested namespaces and the conversion of | |
| 1852 __gnu_internal implementation-private details to anonymous | |
| 1853 namespaces whenever possible. | |
| 1854 * Implemented LWG resolutions DR 431 and DR 538. | |
| 1855 | |
| 1856 Fortran | |
| 1857 | |
| 1858 * Support for allocatable components has been added (TR 15581 and | |
| 1859 Fortran 2003). | |
| 1860 * Support for the Fortran 2003 streaming IO extension has been added. | |
| 1861 * The GNU Fortran compiler now uses 4-byte record markers by default | |
| 1862 for unformatted files to be compatible with g77 and most other | |
| 1863 compilers. The implementation allows for records greater than 2 GB | |
| 1864 and is compatible with several other compilers. Older versions of | |
| 1865 gfortran used 8-byte record markers by default (on most systems). | |
| 1866 In order to change the length of the record markers, e.g. to read | |
| 1867 unformatted files created by older gfortran versions, the | |
|
47
3bfb6c00c1e0
update it from 4.4.2 to 4.4.3.
kent <kent@cr.ie.u-ryukyu.ac.jp>
parents:
19
diff
changeset
|
1868 [5]-frecord-marker=8 option can be used. |
| 0 | 1869 |
| 1870 Java (GCJ) | |
| 1871 | |
| 1872 * A new command-line option -static-libgcj has been added for targets | |
| 1873 that use a linker compatible with GNU Binutils. As its name | |
| 1874 implies, this causes libgcj to be linked statically. In some cases | |
| 1875 this causes the resulting executable to start faster and use less | |
| 1876 memory than if the shared version of libgcj were used. However | |
| 1877 caution should be used as it can also cause essential parts of the | |
| 1878 library to be omitted. Some of these issues are discussed in: | |
|
47
3bfb6c00c1e0
update it from 4.4.2 to 4.4.3.
kent <kent@cr.ie.u-ryukyu.ac.jp>
parents:
19
diff
changeset
|
1879 [6]http://gcc.gnu.org/wiki/Statically_linking_libgcj |
| 0 | 1880 * fastjar is no longer bundled with GCC. To build libgcj, you will |
| 1881 need either InfoZIP (both zip and unzip) or an external jar | |
| 1882 program. In the former case, the GCC build will install a jar shell | |
| 1883 script that is based on InfoZIP and provides the same functionality | |
| 1884 as fastjar. | |
| 1885 | |
| 1886 New Targets and Target Specific Improvements | |
| 1887 | |
| 1888 IA-32/x86-64 | |
| 1889 | |
| 1890 * -mtune=generic can now be used to generate code running well on | |
| 1891 common x86 chips. This includes AMD Athlon, AMD Opteron, Intel | |
| 1892 Pentium-M, Intel Pentium 4 and Intel Core 2. | |
| 1893 * -mtune=native and -march=native will produce code optimized for the | |
| 1894 host architecture as detected using the cpuid instruction. | |
| 1895 * Added a new command-line option -fstackrealign and and | |
| 1896 __attribute__ ((force_align_arg_pointer)) to realign the stack at | |
| 1897 runtime. This allows functions compiled with a vector-aligned stack | |
| 1898 to be invoked from legacy objects that keep only word-alignment. | |
| 1899 | |
| 1900 SPARC | |
| 1901 | |
| 1902 * The default CPU setting has been changed from V7 to V9 in 32-bit | |
| 1903 mode on Solaris 7 and above. This is already the case in 64-bit | |
| 1904 mode. It can be overridden by specifying --with-cpu at configure | |
| 1905 time. | |
| 1906 * Back-end support of built-in functions for atomic memory access has | |
| 1907 been implemented. | |
| 1908 * Support for the Sun UltraSPARC T1 (Niagara) processor has been | |
| 1909 added. | |
| 1910 | |
| 1911 M32C | |
| 1912 | |
| 1913 * Various bug fixes have made some functions (notably, functions | |
| 1914 returning structures) incompatible with previous releases. | |
| 1915 Recompiling all libraries is recommended. Note that code quality | |
| 1916 has considerably improved since 4.1, making a recompile even more | |
| 1917 beneficial. | |
| 1918 | |
| 1919 MIPS | |
| 1920 | |
| 1921 * Added support for the Broadcom SB-1A core. | |
| 1922 | |
| 1923 IA-64 | |
| 1924 | |
| 1925 * Added support for IA-64 data and control speculation. By default | |
| 1926 speculation is enabled only during second scheduler pass. A number | |
| 1927 of machine flags was introduced to control the usage of speculation | |
| 1928 for both scheduler passes. | |
| 1929 | |
| 1930 HPPA | |
| 1931 | |
| 1932 * Added Java language support (libffi and libjava) for 32-bit HP-UX | |
| 1933 11 target. | |
| 1934 | |
| 1935 Obsolete Systems | |
| 1936 | |
| 1937 Documentation improvements | |
| 1938 | |
| 1939 PDF Documentation | |
| 1940 | |
| 1941 * A make pdf target has been added to the top-level makefile, | |
| 1942 enabling automated production of PDF documentation files. | |
| 1943 (Front-ends external to GCC should modify their Make-lang.in file | |
| 1944 to add a lang.pdf: target.) | |
| 1945 | |
| 1946 Other significant improvements | |
| 1947 | |
| 1948 Build system improvements | |
| 1949 | |
| 1950 * All the components of the compiler are now bootstrapped by default. | |
| 1951 This improves the resilience to bugs in the system compiler or | |
| 1952 binary compatibility problems, as well as providing better testing | |
| 1953 of GCC 4.2 itself. In addition, if you build the compiler from a | |
| 1954 combined tree, the assembler, linker, etc. will also be | |
| 1955 bootstrapped (i.e. built with themselves). | |
| 1956 You can disable this behavior, and go back to the pre-GCC 4.2 set | |
| 1957 up, by configuring GCC with --disable-bootstrap. | |
| 1958 * The rules that configure follows to find target tools resemble more | |
| 1959 closely the locations that the built compiler will search. In | |
| 1960 addition, you can use the new configure option --with-target-tools | |
| 1961 to specify where to find the target tools used during the build, | |
| 1962 without affecting what the built compiler will use. | |
| 1963 This can be especially useful when building packages of GCC. For | |
| 1964 example, you may want to build GCC with GNU as or ld, even if the | |
| 1965 resulting compiler to work with the native assembler and linker. To | |
| 1966 do so, you can use --with-target-tools to point to the native | |
| 1967 tools. | |
| 1968 | |
| 1969 Incompatible changes to the build system | |
| 1970 | |
| 1971 * Front-ends external to GCC should modify their Make-lang.in file to | |
| 1972 replace double-colon rules (e.g. dvi::) with normal rules (like | |
| 1973 lang.dvi:). Front-end makefile hooks do not use double-colon rules | |
| 1974 anymore. | |
| 1975 * Up to GCC 4.1, a popular way to specify the target tools used | |
| 1976 during the build was to create directories named gas, binutils, | |
| 1977 etc. in the build tree, and create links to the tools from there. | |
| 1978 This does not work any more when the compiler is bootstrapped. The | |
| 1979 new configure option --with-target-tools provides a better way to | |
| 1980 achieve the same effect, and works for all native and cross | |
| 1981 settings. | |
| 1982 | |
|
47
3bfb6c00c1e0
update it from 4.4.2 to 4.4.3.
kent <kent@cr.ie.u-ryukyu.ac.jp>
parents:
19
diff
changeset
|
1983 Please send FSF & GNU inquiries & questions to [7]gnu@gnu.org. There |
|
3bfb6c00c1e0
update it from 4.4.2 to 4.4.3.
kent <kent@cr.ie.u-ryukyu.ac.jp>
parents:
19
diff
changeset
|
1984 are also [8]other ways to contact the FSF. |
|
3bfb6c00c1e0
update it from 4.4.2 to 4.4.3.
kent <kent@cr.ie.u-ryukyu.ac.jp>
parents:
19
diff
changeset
|
1985 |
|
3bfb6c00c1e0
update it from 4.4.2 to 4.4.3.
kent <kent@cr.ie.u-ryukyu.ac.jp>
parents:
19
diff
changeset
|
1986 These pages are maintained by [9]the GCC team. |
| 0 | 1987 |
| 1988 | |
| 1989 For questions related to the use of GCC, please consult these web | |
|
47
3bfb6c00c1e0
update it from 4.4.2 to 4.4.3.
kent <kent@cr.ie.u-ryukyu.ac.jp>
parents:
19
diff
changeset
|
1990 pages and the [10]GCC manuals. If that fails, the |
|
3bfb6c00c1e0
update it from 4.4.2 to 4.4.3.
kent <kent@cr.ie.u-ryukyu.ac.jp>
parents:
19
diff
changeset
|
1991 [11]gcc-help@gcc.gnu.org mailing list might help. |
| 0 | 1992 Please send comments on these web pages and the development of GCC to |
|
47
3bfb6c00c1e0
update it from 4.4.2 to 4.4.3.
kent <kent@cr.ie.u-ryukyu.ac.jp>
parents:
19
diff
changeset
|
1993 our developer mailing list at [12]gcc@gnu.org or [13]gcc@gcc.gnu.org. |
|
3bfb6c00c1e0
update it from 4.4.2 to 4.4.3.
kent <kent@cr.ie.u-ryukyu.ac.jp>
parents:
19
diff
changeset
|
1994 All of our lists have [14]public archives. |
| 0 | 1995 |
| 1996 Copyright (C) Free Software Foundation, Inc., 51 Franklin St, Fifth | |
| 1997 Floor, Boston, MA 02110, USA. | |
| 1998 | |
| 1999 Verbatim copying and distribution of this entire article is permitted | |
| 2000 in any medium, provided this notice is preserved. | |
|
47
3bfb6c00c1e0
update it from 4.4.2 to 4.4.3.
kent <kent@cr.ie.u-ryukyu.ac.jp>
parents:
19
diff
changeset
|
2001 Last modified 2009-11-08 [15]Valid XHTML 1.0 |
| 0 | 2002 |
| 2003 References | |
| 2004 | |
| 2005 1. http://gcc.gnu.org/projects/gomp/ | |
| 2006 2. http://gcc.gnu.org/onlinedocs/gcc/Warning-Options.html | |
|
47
3bfb6c00c1e0
update it from 4.4.2 to 4.4.3.
kent <kent@cr.ie.u-ryukyu.ac.jp>
parents:
19
diff
changeset
|
2007 3. http://gcc.gnu.org/onlinedocs/libstdc++/manual/bk01pt01ch01.html#manual.intro.status.standard.tr1 |
|
3bfb6c00c1e0
update it from 4.4.2 to 4.4.3.
kent <kent@cr.ie.u-ryukyu.ac.jp>
parents:
19
diff
changeset
|
2008 4. http://gcc.gnu.org/onlinedocs/libstdc++/ext/pb_ds/index.html |
|
3bfb6c00c1e0
update it from 4.4.2 to 4.4.3.
kent <kent@cr.ie.u-ryukyu.ac.jp>
parents:
19
diff
changeset
|
2009 5. http://gcc.gnu.org/onlinedocs/gfortran/Runtime-Options.html |
|
3bfb6c00c1e0
update it from 4.4.2 to 4.4.3.
kent <kent@cr.ie.u-ryukyu.ac.jp>
parents:
19
diff
changeset
|
2010 6. http://gcc.gnu.org/wiki/Statically_linking_libgcj |
|
3bfb6c00c1e0
update it from 4.4.2 to 4.4.3.
kent <kent@cr.ie.u-ryukyu.ac.jp>
parents:
19
diff
changeset
|
2011 7. mailto:gnu@gnu.org |
|
3bfb6c00c1e0
update it from 4.4.2 to 4.4.3.
kent <kent@cr.ie.u-ryukyu.ac.jp>
parents:
19
diff
changeset
|
2012 8. http://www.gnu.org/home.html#ContactInfo |
|
3bfb6c00c1e0
update it from 4.4.2 to 4.4.3.
kent <kent@cr.ie.u-ryukyu.ac.jp>
parents:
19
diff
changeset
|
2013 9. http://gcc.gnu.org/about.html |
|
3bfb6c00c1e0
update it from 4.4.2 to 4.4.3.
kent <kent@cr.ie.u-ryukyu.ac.jp>
parents:
19
diff
changeset
|
2014 10. http://gcc.gnu.org/onlinedocs/ |
|
3bfb6c00c1e0
update it from 4.4.2 to 4.4.3.
kent <kent@cr.ie.u-ryukyu.ac.jp>
parents:
19
diff
changeset
|
2015 11. mailto:gcc-help@gcc.gnu.org |
|
3bfb6c00c1e0
update it from 4.4.2 to 4.4.3.
kent <kent@cr.ie.u-ryukyu.ac.jp>
parents:
19
diff
changeset
|
2016 12. mailto:gcc@gnu.org |
|
3bfb6c00c1e0
update it from 4.4.2 to 4.4.3.
kent <kent@cr.ie.u-ryukyu.ac.jp>
parents:
19
diff
changeset
|
2017 13. mailto:gcc@gcc.gnu.org |
|
3bfb6c00c1e0
update it from 4.4.2 to 4.4.3.
kent <kent@cr.ie.u-ryukyu.ac.jp>
parents:
19
diff
changeset
|
2018 14. http://gcc.gnu.org/lists.html |
|
3bfb6c00c1e0
update it from 4.4.2 to 4.4.3.
kent <kent@cr.ie.u-ryukyu.ac.jp>
parents:
19
diff
changeset
|
2019 15. http://validator.w3.org/check/referer |
| 0 | 2020 ====================================================================== |
| 2021 http://gcc.gnu.org/gcc-4.1/index.html | |
| 2022 GCC 4.1 Release Series | |
| 2023 | |
| 2024 February 13, 2007 | |
| 2025 | |
| 2026 The [1]GNU project and the GCC developers are pleased to announce the | |
| 2027 release of GCC 4.1.2. | |
| 2028 | |
| 2029 This release is a bug-fix release, containing fixes for regressions in | |
| 2030 GCC 4.1.1 relative to previous releases of GCC. | |
| 2031 | |
| 2032 Release History | |
| 2033 | |
| 2034 GCC 4.1.2 | |
| 2035 February 13, 2007 ([2]changes) | |
| 2036 | |
| 2037 GCC 4.1.1 | |
| 2038 May 24, 2006 ([3]changes) | |
| 2039 | |
| 2040 GCC 4.1.0 | |
| 2041 February 28, 2006 ([4]changes) | |
| 2042 | |
| 2043 References and Acknowledgements | |
| 2044 | |
| 2045 GCC used to stand for the GNU C Compiler, but since the compiler | |
| 2046 supports several other languages aside from C, it now stands for the | |
| 2047 GNU Compiler Collection. | |
| 2048 | |
| 2049 A list of [5]successful builds is updated as new information becomes | |
| 2050 available. | |
| 2051 | |
| 2052 The GCC developers would like to thank the numerous people that have | |
| 2053 contributed new features, improvements, bug fixes, and other changes as | |
| 2054 well as test results to GCC. This [6]amazing group of volunteers is | |
| 2055 what makes GCC successful. | |
| 2056 | |
| 2057 For additional information about GCC please refer to the [7]GCC project | |
| 2058 web site or contact the [8]GCC development mailing list. | |
| 2059 | |
| 2060 To obtain GCC please use [9]our mirror sites or [10]our SVN server. | |
| 2061 | |
| 2062 Please send FSF & GNU inquiries & questions to [11]gnu@gnu.org. There | |
| 2063 are also [12]other ways to contact the FSF. | |
| 2064 | |
| 2065 These pages are maintained by [13]the GCC team. | |
| 2066 | |
| 2067 | |
| 2068 For questions related to the use of GCC, please consult these web | |
| 2069 pages and the [14]GCC manuals. If that fails, the | |
| 2070 [15]gcc-help@gcc.gnu.org mailing list might help. | |
| 2071 Please send comments on these web pages and the development of GCC to | |
| 2072 our developer mailing list at [16]gcc@gnu.org or [17]gcc@gcc.gnu.org. | |
| 2073 All of our lists have [18]public archives. | |
| 2074 | |
| 2075 Copyright (C) Free Software Foundation, Inc., 51 Franklin St, Fifth | |
| 2076 Floor, Boston, MA 02110, USA. | |
| 2077 | |
| 2078 Verbatim copying and distribution of this entire article is permitted | |
| 2079 in any medium, provided this notice is preserved. | |
| 2080 Last modified 2008-10-04 [19]Valid XHTML 1.0 | |
| 2081 | |
| 2082 References | |
| 2083 | |
| 2084 1. http://www.gnu.org/ | |
| 2085 2. http://gcc.gnu.org/gcc-4.1/changes.html#4.1.2 | |
| 2086 3. http://gcc.gnu.org/gcc-4.1/changes.html | |
| 2087 4. http://gcc.gnu.org/gcc-4.1/changes.html | |
| 2088 5. http://gcc.gnu.org/gcc-4.1/buildstat.html | |
| 2089 6. http://gcc.gnu.org/onlinedocs/gcc/Contributors.html | |
| 2090 7. http://gcc.gnu.org/index.html | |
| 2091 8. mailto:gcc@gcc.gnu.org | |
| 2092 9. http://gcc.gnu.org/mirrors.html | |
| 2093 10. http://gcc.gnu.org/svn.html | |
| 2094 11. mailto:gnu@gnu.org | |
| 2095 12. http://www.gnu.org/home.html#ContactInfo | |
| 2096 13. http://gcc.gnu.org/about.html | |
| 2097 14. http://gcc.gnu.org/onlinedocs/ | |
| 2098 15. mailto:gcc-help@gcc.gnu.org | |
| 2099 16. mailto:gcc@gnu.org | |
| 2100 17. mailto:gcc@gcc.gnu.org | |
| 2101 18. http://gcc.gnu.org/lists.html | |
| 2102 19. http://validator.w3.org/check/referer | |
| 2103 ====================================================================== | |
| 2104 http://gcc.gnu.org/gcc-4.1/changes.html | |
| 2105 GCC 4.1 Release Series | |
| 2106 Changes, New Features, and Fixes | |
| 2107 | |
| 2108 The latest release in the 4.1 release series is [1]GCC 4.1.2. | |
| 2109 | |
| 2110 Caveats | |
| 2111 | |
| 2112 General Optimizer Improvements | |
| 2113 | |
| 2114 * GCC now has infrastructure for inter-procedural optimizations and | |
| 2115 the following inter-procedural optimizations are implemented: | |
| 2116 + Profile guided inlining. When doing profile feedback guided | |
| 2117 optimization, GCC can now use the profile to make better | |
| 2118 informed decisions on whether inlining of a function is | |
| 2119 profitable or not. This means that GCC will no longer inline | |
| 2120 functions at call sites that are not executed very often, and | |
| 2121 that functions at hot call sites are more likely to be | |
| 2122 inlined. | |
| 2123 A new parameter min-inline-recursive-probability is also now | |
| 2124 available to throttle recursive inlining of functions with | |
| 2125 small average recursive depths. | |
| 2126 + Discovery of pure and const functions, a form of side-effects | |
| 2127 analysis. While older GCC releases could also discover such | |
| 2128 special functions, the new IPA-based pass runs earlier so that | |
| 2129 the results are available to more optimizers. The pass is also | |
| 2130 simply more powerful than the old one. | |
| 2131 + Analysis of references to static variables and type escape | |
| 2132 analysis, also forms of side-effects analysis. The results of | |
| 2133 these passes allow the compiler to be less conservative about | |
| 2134 call-clobbered variables and references. This results in more | |
| 2135 redundant loads being eliminated and in making static | |
| 2136 variables candidates for register promotion. | |
| 2137 + Improvement of RTL-based alias analysis. The results of type | |
| 2138 escape analysis are fed to the RTL type-based alias analyzer, | |
| 2139 allowing it to disambiguate more memory references. | |
| 2140 + Interprocedural constant propagation and function versioning. | |
| 2141 This pass looks for functions that are always called with the | |
| 2142 same constant value for one or more of the function arguments, | |
| 2143 and propagates those constants into those functions. | |
| 2144 + GCC will now eliminate static variables whose usage was | |
| 2145 optimized out. | |
| 2146 + -fwhole-program --combine can now be used to make all | |
| 2147 functions in program static allowing whole program | |
| 2148 optimization. As an exception, the main function and all | |
| 2149 functions marked with the new externally_visible attribute are | |
| 2150 kept global so that programs can link with runtime libraries. | |
| 2151 * GCC can now do a form of partial dead code elimination (PDCE) that | |
| 2152 allows code motion of expressions to the paths where the result of | |
| 2153 the expression is actually needed. This is not always a win, so the | |
| 2154 pass has been limited to only consider profitable cases. Here is an | |
| 2155 example: | |
| 2156 int foo (int *, int *); | |
| 2157 int | |
| 2158 bar (int d) | |
| 2159 { | |
| 2160 int a, b, c; | |
| 2161 b = d + 1; | |
| 2162 c = d + 2; | |
| 2163 a = b + c; | |
| 2164 if (d) | |
| 2165 { | |
| 2166 foo (&b, &c); | |
| 2167 a = b + c; | |
| 2168 } | |
| 2169 printf ("%d\n", a); | |
| 2170 } | |
| 2171 | |
| 2172 The a = b + c can be sunk to right before the printf. Normal code | |
| 2173 sinking will not do this, it will sink the first one above into the | |
| 2174 else-branch of the conditional jump, which still gives you two | |
| 2175 copies of the code. | |
| 2176 * GCC now has a value range propagation pass. This allows the | |
| 2177 compiler to eliminate bounds checks and branches. The results of | |
| 2178 the pass can also be used to accurately compute branch | |
| 2179 probabilities. | |
| 2180 * The pass to convert PHI nodes to straight-line code (a form of | |
| 2181 if-conversion for GIMPLE) has been improved significantly. The two | |
| 2182 most significant improvements are an improved algorithm to | |
| 2183 determine the order in which the PHI nodes are considered, and an | |
| 2184 improvement that allow the pass to consider if-conversions of basic | |
| 2185 blocks with more than two predecessors. | |
| 2186 * Alias analysis improvements. GCC can now differentiate between | |
| 2187 different fields of structures in Tree-SSA's virtual operands form. | |
| 2188 This lets stores/loads from non-overlapping structure fields not | |
| 2189 conflict. A new algorithm to compute points-to sets was contributed | |
| 2190 that can allows GCC to see now that p->a and p->b, where p is a | |
| 2191 pointer to a structure, can never point to the same field. | |
| 2192 * Various enhancements to auto-vectorization: | |
| 2193 + Incrementally preserve SSA form when vectorizing. | |
| 2194 + Incrementally preserve loop-closed form when vectorizing. | |
| 2195 + Improvements to peeling for alignment: generate better code | |
| 2196 when the misalignment of an access is known at compile time, | |
| 2197 or when different accesses are known to have the same | |
| 2198 misalignment, even if the misalignment amount itself is | |
| 2199 unknown. | |
| 2200 + Consider dependence distance in the vectorizer. | |
| 2201 + Externalize generic parts of data reference analysis to make | |
| 2202 this analysis available to other passes. | |
| 2203 + Vectorization of conditional code. | |
| 2204 + Reduction support. | |
| 2205 * GCC can now partition functions in sections of hot and cold code. | |
| 2206 This can significantly improve performance due to better | |
| 2207 instruction cache locality. This feature works best together with | |
| 2208 profile feedback driven optimization. | |
| 2209 * A new pass to avoid saving of unneeded arguments to the stack in | |
| 2210 vararg functions if the compiler can prove that they will not be | |
| 2211 needed. | |
| 2212 * Transition of basic block profiling to tree level implementation | |
| 2213 has been completed. The new implementation should be considerably | |
| 2214 more reliable (hopefully avoiding profile mismatch errors when | |
| 2215 using -fprofile-use or -fbranch-probabilities) and can be used to | |
| 2216 drive higher level optimizations, such as inlining. | |
| 2217 The -ftree-based-profiling command-line option was removed and | |
| 2218 -fprofile-use now implies disabling old RTL level loop optimizer | |
| 2219 (-fno-loop-optimize). Speculative prefetching optimization | |
| 2220 (originally enabled by -fspeculative-prefetching) was removed. | |
| 2221 | |
| 2222 New Languages and Language specific improvements | |
| 2223 | |
| 2224 C and Objective-C | |
| 2225 | |
| 2226 * The old Bison-based C and Objective-C parser has been replaced by a | |
| 2227 new, faster hand-written recursive-descent parser. | |
| 2228 | |
| 2229 Ada | |
| 2230 | |
| 2231 * The build infrastructure for the Ada runtime library and tools has | |
| 2232 been changed to be better integrated with the rest of the build | |
| 2233 infrastructure of GCC. This should make doing cross builds of Ada a | |
| 2234 bit easier. | |
| 2235 | |
| 2236 C++ | |
| 2237 | |
| 2238 * ARM-style name-injection of friend declarations is no longer the | |
| 2239 default. For example: | |
| 2240 struct S { | |
| 2241 friend void f(); | |
| 2242 }; | |
| 2243 | |
| 2244 void g() { f(); } | |
| 2245 will not be accepted; instead a declaration of f will need to be | |
| 2246 present outside of the scope of S. The new -ffriend-injection | |
| 2247 option will enable the old behavior. | |
| 2248 * The (undocumented) extension which permitted templates with default | |
| 2249 arguments to be bound to template template parameters with fewer | |
| 2250 parameters has been deprecated, and will be removed in the next | |
| 2251 major release of G++. For example: | |
| 2252 template <template <typename> class C> | |
| 2253 void f(C<double>) {} | |
| 2254 | |
| 2255 template <typename T, typename U = int> | |
| 2256 struct S {}; | |
| 2257 | |
| 2258 template void f(S<double>); | |
| 2259 | |
| 2260 makes use of the deprecated extension. The reason this code is not | |
| 2261 valid ISO C++ is that S is a template with two parameters; | |
| 2262 therefore, it cannot be bound to C which has only one parameter. | |
| 2263 | |
| 2264 Runtime Library (libstdc++) | |
| 2265 | |
| 2266 * Optimization work: | |
| 2267 + A new implementation of std::search_n is provided, better | |
| 2268 performing in case of random access iterators. | |
| 2269 + Added further efficient specializations of istream functions, | |
| 2270 i.e., character array and string extractors. | |
| 2271 + Other smaller improvements throughout. | |
| 2272 * Policy-based associative containers, designed for high-performance, | |
| 2273 flexibility and semantic safety are delivered in ext/pb_assoc. | |
| 2274 * A versatile string class, __gnu_cxx::__versa_string, providing | |
| 2275 facilities conforming to the standard requirements for | |
| 2276 basic_string, is delivered in <ext/vstring.h>. In particular: | |
| 2277 + Two base classes are provided: the default one avoids | |
| 2278 reference counting and is optimized for short strings; the | |
| 2279 alternate one, still uses it while improving in a few low | |
| 2280 level areas (e.g., alignment). See vstring_fwd.h for some | |
| 2281 useful typedefs. | |
| 2282 + Various algorithms have been rewritten (e.g., replace), the | |
| 2283 code streamlined and simple optimizations added. | |
| 2284 + Option 3 of DR 431 is implemented for both available bases, | |
| 2285 thus improving the support for stateful allocators. | |
| 2286 * As usual, many bugs have been fixed (e.g., libstdc++/13583, | |
| 2287 libstdc++/23953) and LWG resolutions put into effect for the first | |
| 2288 time (e.g., DR 280, DR 464, N1780 recommendations for DR 233, TR1 | |
| 2289 Issue 6.19). The implementation status of TR1 is now tracked in the | |
| 2290 docs in tr1.html. | |
| 2291 | |
| 2292 Objective-C++ | |
| 2293 | |
| 2294 * A new language front end for Objective-C++ has been added. This | |
| 2295 language allows users to mix the object oriented features of | |
| 2296 Objective-C with those of C++. | |
| 2297 | |
| 2298 Java (GCJ) | |
| 2299 | |
| 2300 * Core library (libgcj) updates based on GNU Classpath 0.15 - 0.19 | |
| 2301 features (plus some 0.20 bug-fixes) | |
| 2302 + Networking | |
| 2303 o The java.net.HttpURLConnection implementation no longer | |
| 2304 buffers the entire response body in memory. This means | |
| 2305 that response bodies larger than available memory can now | |
| 2306 be handled. | |
| 2307 + (N)IO | |
| 2308 o NIO FileChannel.map implementation, fast bulk put | |
| 2309 implementation for DirectByteBuffer (speeds up this | |
| 2310 method 10x). | |
| 2311 o FileChannel.lock() and FileChannel.force() implemented. | |
| 2312 + XML | |
| 2313 o gnu.xml fix for nodes created outside a namespace | |
| 2314 context. | |
| 2315 o Add support for output indenting and | |
| 2316 cdata-section-elements output instruction in | |
| 2317 xml.transform. | |
| 2318 o xml.xpath corrections for cases where elements/attributes | |
| 2319 might have been created in non-namespace-aware mode. | |
| 2320 Corrections to handling of XSL variables and minor | |
| 2321 conformance updates. | |
| 2322 + AWT | |
| 2323 o GNU JAWT implementation, the AWT Native Interface, which | |
| 2324 allows direct access to native screen resources from | |
| 2325 within a Canvas's paint method. GNU Classpath Examples | |
| 2326 comes with a Demo, see libjava/classpath/examples/README. | |
| 2327 o awt.datatransfer updated to 1.5 with support for | |
| 2328 FlavorEvents. The gtk+ awt peers now allow copy/paste of | |
| 2329 text, images, URIs/files and serialized objects with | |
| 2330 other applications and tracking clipboard change events | |
| 2331 with gtk+ 2.6 (for gtk+ 2.4 only text and serialized | |
| 2332 objects are supported). A GNU Classpath Examples | |
| 2333 datatransfer Demo was added to show the new | |
| 2334 functionality. | |
| 2335 o Split gtk+ awt peers event handling in two threads and | |
| 2336 improve gdk lock handling (solves several awt lock ups). | |
| 2337 o Speed up awt Image loading. | |
| 2338 o Better gtk+ scrollbar peer implementation when using gtk+ | |
| 2339 >= 2.6. | |
| 2340 o Handle image loading errors correctly for gdkpixbuf and | |
| 2341 MediaTracker. | |
| 2342 o Better handle GDK lock. Properly prefix gtkpeer native | |
| 2343 functions (cp_gtk). | |
| 2344 o GdkGraphics2D has been updated to use Cairo 0.5.x or | |
| 2345 higher. | |
| 2346 o BufferedImage and GtkImage rewrites. All image drawing | |
| 2347 operations should now work correctly (flipping requires | |
| 2348 gtk+ >= 2.6) | |
| 2349 o Future Graphics2D, image and text work is documented at: | |
| 2350 [2]http://developer.classpath.org/mediation/ClasspathGrap | |
| 2351 hicsImagesText | |
| 2352 o When gtk+ 2.6 or higher is installed the default log | |
| 2353 handler will produce stack traces whenever a WARNING, | |
| 2354 CRITICAL or ERROR message is produced. | |
| 2355 + Free Swing | |
| 2356 o The RepaintManager has been reworked for more efficient | |
| 2357 painting, especially for large GUIs. | |
| 2358 o The layout manager OverlayLayout has been implemented, | |
| 2359 the BoxLayout has been rewritten to make use of the | |
| 2360 SizeRequirements utility class and caching for more | |
| 2361 efficient layout. | |
| 2362 o Improved accessibility support. | |
| 2363 o Significant progress has been made in the implementation | |
| 2364 of the javax.swing.plaf.metal package, with most UI | |
| 2365 delegates in a working state now. Please test this with | |
| 2366 your own applications and provide feedback that will help | |
| 2367 us to improve this package. | |
| 2368 o The GUI demo (gnu.classpath.examples.swing.Demo) has been | |
| 2369 extended to highlight various features in our Free Swing | |
| 2370 implementation. And it includes a look and feel switcher | |
| 2371 for Metal (default), Ocean and GNU themes. | |
| 2372 o The javax.swing.plaf.multi package is now implemented. | |
| 2373 o Editing and several key actions for JTree and JTable were | |
| 2374 implemented. | |
| 2375 o Lots of icons and look and feel improvements for Free | |
| 2376 Swing basic and metal themes were added. Try running the | |
| 2377 GNU Classpath Swing Demo in examples | |
| 2378 (gnu.classpath.examples.swing.Demo) with: | |
| 2379 -Dswing.defaultlaf=javax.swing.plaf.basic.BasicLookAndFee | |
| 2380 l or | |
| 2381 -Dswing.defaultlaf=javax.swing.plaf.metal.MetalLookAndFee | |
| 2382 l | |
| 2383 o Start of styled text capabilites for java.swing.text. | |
| 2384 o DefaultMutableTreeNode pre-order, post-order, depth-first | |
| 2385 and breadth-first traversal enumerations implemented. | |
| 2386 o JInternalFrame colors and titlebar draw properly. | |
| 2387 o JTree is working up to par (icons, selection and keyboard | |
| 2388 traversal). | |
| 2389 o JMenus were made more compatible in visual and | |
| 2390 programmatic behavior. | |
| 2391 o JTable changeSelection and multiple selections | |
| 2392 implemented. | |
| 2393 o JButton and JToggleButton change states work properly | |
| 2394 now. | |
| 2395 o JFileChooser fixes. | |
| 2396 o revalidate() and repaint() fixes which make Free Swing | |
| 2397 much more responsive. | |
| 2398 o MetalIconFactory implemented. | |
| 2399 o Free Swing Top-Level Compatibility. JFrame, JDialog, | |
| 2400 JApplet, JInternalFrame, and JWindow are now 1.5 | |
| 2401 compatible in the sense that you can call add() and | |
| 2402 setLayout() directly on them, which will have the same | |
| 2403 effect as calling getContentPane().add() and | |
| 2404 getContentPane().setLayout(). | |
| 2405 o The JTree interface has been completed. JTrees now | |
| 2406 recognizes mouse clicks and selections work. | |
| 2407 o BoxLayout works properly now. | |
| 2408 o Fixed GrayFilter to actually work. | |
| 2409 o Metal SplitPane implemented. | |
| 2410 o Lots of Free Swing text and editor stuff work now. | |
| 2411 + Free RMI and Corba | |
| 2412 o Andrew Watson, Vice President and Technical Director of | |
| 2413 the Object Management Group, has officially assigned us | |
| 2414 20 bit Vendor Minor Code Id: 0x47430 ("GC") that will | |
| 2415 mark remote classpath-specific system exceptions. | |
| 2416 Obtaining the VMCID means that GNU Classpath now is a | |
| 2417 recogniseable type of node in a highly interoperable | |
| 2418 CORBA world. | |
| 2419 o GNU Classpath now includes the first working draft to | |
| 2420 support the RMI over IIOP protocol. The current | |
| 2421 implementation is capable of remote invocations, | |
| 2422 transferring various Serializables and Externalizables | |
| 2423 via RMI-IIOP protocol. It can flatten graphs and, at | |
| 2424 least for the simple cases, is interoperable with 1.5 | |
| 2425 JDKs. | |
| 2426 o org.omg.PortableInterceptor and related functionality in | |
| 2427 other packages is now implemented: | |
| 2428 # The sever and client interceptors work as required | |
| 2429 since 1.4. | |
| 2430 # The IOR interceptor works as needed for 1.5. | |
| 2431 o The org.omg.DynamicAny package is completed and passes | |
| 2432 the prepared tests. | |
| 2433 o The Portable Object Adapter should now support the output | |
| 2434 of the recent IDL to java compilers. These compilers now | |
| 2435 generate servants and not CORBA objects as before, making | |
| 2436 the output depend on the existing POA implementation. | |
| 2437 Completing POA means that such code can already be tried | |
| 2438 to run on Classpath. Our POA is tested for the following | |
| 2439 usager scenarios: | |
| 2440 # POA converts servant to the CORBA object. | |
| 2441 # Servant provides to the CORBA object. | |
| 2442 # POA activates new CORBA object with the given Object | |
| 2443 Id (byte array) that is later accessible for the | |
| 2444 servant. | |
| 2445 # During the first call, the ServantActivator provides | |
| 2446 servant for this and all subsequent calls on the | |
| 2447 current object. | |
| 2448 # During each call, the ServantLocator provides | |
| 2449 servant for this call only. | |
| 2450 # ServantLocator or ServantActivator forwards call to | |
| 2451 another server. | |
| 2452 # POA has a single servant, responsible for all | |
| 2453 objects. | |
| 2454 # POA has a default servant, but some objects are | |
| 2455 explicitly connected to they specific servants. | |
| 2456 The POA is verified using tests from the former | |
| 2457 cost.omg.org. | |
| 2458 o The CORBA implementation is now a working prototype that | |
| 2459 should support features up to 1.3 inclusive. We invite | |
| 2460 groups writing CORBA dependent applications to try | |
| 2461 Classpath implementation, reporting any possible bugs. | |
| 2462 The CORBA prototype is interoperable with Sun's | |
| 2463 implementation v 1.4, transferring object references, | |
| 2464 primitive types, narrow and wide strings, arrays, | |
| 2465 structures, trees, abstract interfaces and value types | |
| 2466 (feature of CORBA 2.3) between these two platforms. | |
| 2467 Remote exceptions are transferred and handled correctly. | |
| 2468 The stringified object references (IORs) from various | |
| 2469 sources are parsed as required. The transient (for | |
| 2470 current session) and permanent (till jre restart) | |
| 2471 redirections work. Both Little and Big Endian encoded | |
| 2472 messages are accepted. The implementation is verified | |
| 2473 using tests from the former cost.omg.org. The current | |
| 2474 release includes working examples (see the examples | |
| 2475 directory), demonstrating the client-server | |
| 2476 communication, using either CORBA Request or IDL-based | |
| 2477 stub (usually generated by a IDL to java compiler). These | |
| 2478 examples also show how to use the Classpath CORBA naming | |
| 2479 service. The IDL to java compiler is not yet written, but | |
| 2480 as our library must be compatible, it naturally accepts | |
| 2481 the output of other idlj implementations. | |
| 2482 + Misc | |
| 2483 o Updated TimeZone data against Olson tzdata2005l. | |
| 2484 o Make zip and jar packages UTF-8 clean. | |
| 2485 o "native" code builds and compiles (warning free) on | |
| 2486 Darwin and Solaris. | |
| 2487 o java.util.logging.FileHandler now rotates files. | |
| 2488 o Start of a generic JDWP framework in gnu/classpath/jdwp. | |
| 2489 This is unfinished, but feedback (at classpath@gnu.org) | |
| 2490 from runtime hackers is greatly appreciated. Although | |
| 2491 most of the work is currently being done around gcj/gij | |
| 2492 we want this framework to be as VM neutral as possible. | |
| 2493 Early design is described in: | |
| 2494 [3]http://gcc.gnu.org/ml/java/2005-05/msg00260.html | |
| 2495 o QT4 AWT peers, enable by giving configure | |
| 2496 --enable-qt-peer. Included, but not ready for production | |
| 2497 yet. They are explicitly disabled and not supported. But | |
| 2498 if you want to help with the development of these new | |
| 2499 features we are interested in feedback. You will have to | |
| 2500 explicitly enable them to try them out (and they will | |
| 2501 most likely contain bugs). | |
| 2502 o Documentation fixes all over the place. See | |
| 2503 [4]http://developer.classpath.org/doc/ | |
| 2504 | |
| 2505 New Targets and Target Specific Improvements | |
| 2506 | |
| 2507 IA-32/x86-64 | |
| 2508 | |
| 2509 * The x86-64 medium model (that allows building applications whose | |
| 2510 data segment exceeds 4GB) was redesigned to match latest ABI draft. | |
| 2511 New implementation split large datastructures into separate segment | |
| 2512 improving performance of accesses to small datastructures and also | |
| 2513 allows linking of small model libraries into medium model programs | |
| 2514 as long as the libraries are not accessing the large datastructures | |
| 2515 directly. Medium model is also supported in position independent | |
| 2516 code now. | |
| 2517 The ABI change results in partial incompatibility among medium | |
| 2518 model objects. Linking medium model libraries (or objects) compiled | |
| 2519 with new compiler into medium model program compiled with older | |
| 2520 will likely result in exceeding ranges of relocations. | |
| 2521 Binutils 2.16.91 or newer are required for compiling medium model | |
| 2522 now. | |
| 2523 | |
| 2524 RS6000 (POWER/PowerPC) | |
| 2525 | |
| 2526 * The AltiVec vector primitives in <altivec.h> are now implemented in | |
| 2527 a way that puts a smaller burden on the preprocessor, instead | |
| 2528 processing the "overloading" in the front ends. This should benefit | |
| 2529 compilation speed on AltiVec vector code. | |
| 2530 * AltiVec initializers now are generated more efficiently. | |
| 2531 * The popcountb instruction available on POWER5 now is generated. | |
| 2532 * The floating point round to integer instructions available on | |
| 2533 POWER5+ now is generated. | |
| 2534 * Floating point divides can be synthesized using the floating point | |
| 2535 reciprocal estimate instructions. | |
| 2536 * Double precision floating point constants are initialized as single | |
| 2537 precision values if they can be represented exactly. | |
| 2538 | |
| 2539 S/390, zSeries and System z9 | |
| 2540 | |
| 2541 * Support for the IBM System z9 109 processor has been added. When | |
| 2542 using the -march=z9-109 option, the compiler will generate code | |
| 2543 making use of instructions provided by the extended immediate | |
| 2544 facility. | |
| 2545 * Support for 128-bit IEEE floating point has been added. When using | |
| 2546 the -mlong-double-128 option, the compiler will map the long double | |
| 2547 data type to 128-bit IEEE floating point. Using this option | |
| 2548 constitutes an ABI change, and requires glibc support. | |
| 2549 * Various changes to improve performance of generated code have been | |
| 2550 implemented, including: | |
| 2551 + In functions that do not require a literal pool, register %r13 | |
| 2552 (which is traditionally reserved as literal pool pointer), can | |
| 2553 now be freely used for other purposes by the compiler. | |
| 2554 + More precise tracking of register use allows the compiler to | |
| 2555 generate more efficient function prolog and epilog code in | |
| 2556 certain cases. | |
| 2557 + The SEARCH STRING, COMPARE LOGICAL STRING, and MOVE STRING | |
| 2558 instructions are now used to implement C string functions. | |
| 2559 + The MOVE CHARACTER instruction with single byte overlap is now | |
| 2560 used to implement the memset function with non-zero fill byte. | |
| 2561 + The LOAD ZERO instructions are now used where appropriate. | |
| 2562 + The INSERT CHARACTERS UNDER MASK, STORE CHARACTERS UNDER MASK, | |
| 2563 and INSERT IMMEDIATE instructions are now used more frequently | |
| 2564 to optimize bitfield operations. | |
| 2565 + The BRANCH ON COUNT instruction is now used more frequently. | |
| 2566 In particular, the fact that a loop contains a subroutine call | |
| 2567 no longer prevents the compiler from using this instruction. | |
| 2568 + The compiler is now aware that all shift and rotate | |
| 2569 instructions implicitly truncate the shift count to six bits. | |
| 2570 * Back-end support for the following generic features has been | |
| 2571 implemented: | |
| 2572 + The full set of [5]built-in functions for atomic memory | |
| 2573 access. | |
| 2574 + The -fstack-protector feature. | |
| 2575 + The optimization pass avoiding unnecessary stores of incoming | |
| 2576 argument registers in functions with variable argument list. | |
| 2577 | |
| 2578 SPARC | |
| 2579 | |
| 2580 * The default code model in 64-bit mode has been changed from | |
| 2581 Medium/Anywhere to Medium/Middle on Solaris. | |
| 2582 * TLS support is disabled by default on Solaris prior to release 10. | |
| 2583 It can be enabled on TLS-capable Solaris 9 versions (4/04 release | |
| 2584 and later) by specifying --enable-tls at configure time. | |
| 2585 | |
| 2586 MorphoSys | |
| 2587 | |
| 2588 * Support has been added for this new architecture. | |
| 2589 | |
| 2590 Obsolete Systems | |
| 2591 | |
| 2592 Documentation improvements | |
| 2593 | |
| 2594 Other significant improvements | |
| 2595 | |
| 2596 * GCC can now emit code for protecting applications from | |
| 2597 stack-smashing attacks. The protection is realized by buffer | |
| 2598 overflow detection and reordering of stack variables to avoid | |
| 2599 pointer corruption. | |
| 2600 * Some built-in functions have been fortified to protect them against | |
| 2601 various buffer overflow (and format string) vulnerabilities. | |
| 2602 Compared to the mudflap bounds checking feature, the safe builtins | |
| 2603 have far smaller overhead. This means that programs built using | |
| 2604 safe builtins should not experience any measurable slowdown. | |
| 2605 | |
| 2606 GCC 4.1.2 | |
| 2607 | |
| 2608 This is the [6]list of problem reports (PRs) from GCC's bug tracking | |
| 2609 system that are known to be fixed in the 4.1.2 release. This list might | |
| 2610 not be complete (that is, it is possible that some PRs that have been | |
| 2611 fixed are not listed here). | |
| 2612 | |
| 2613 When generating code for a shared library, GCC now recognizes that | |
| 2614 global functions may be replaced when the program runs. Therefore, it | |
| 2615 is now more conservative in deducing information from the bodies of | |
| 2616 functions. For example, in this example: | |
| 2617 void f() {} | |
| 2618 void g() { | |
| 2619 try { f(); } | |
| 2620 catch (...) { | |
| 2621 cout << "Exception"; | |
| 2622 } | |
| 2623 } | |
| 2624 | |
| 2625 G++ would previously have optimized away the catch clause, since it | |
| 2626 would have concluded that f cannot throw exceptions. Because users may | |
| 2627 replace f with another function in the main body of the program, this | |
| 2628 optimization is unsafe, and is no longer performed. If you wish G++ to | |
| 2629 continue to optimize as before, you must add a throw() clause to the | |
| 2630 declaration of f to make clear that it does not throw exceptions. | |
| 2631 | |
| 2632 Please send FSF & GNU inquiries & questions to [7]gnu@gnu.org. There | |
| 2633 are also [8]other ways to contact the FSF. | |
| 2634 | |
| 2635 These pages are maintained by [9]the GCC team. | |
| 2636 | |
| 2637 | |
| 2638 For questions related to the use of GCC, please consult these web | |
| 2639 pages and the [10]GCC manuals. If that fails, the | |
| 2640 [11]gcc-help@gcc.gnu.org mailing list might help. | |
| 2641 Please send comments on these web pages and the development of GCC to | |
| 2642 our developer mailing list at [12]gcc@gnu.org or [13]gcc@gcc.gnu.org. | |
| 2643 All of our lists have [14]public archives. | |
| 2644 | |
| 2645 Copyright (C) Free Software Foundation, Inc., 51 Franklin St, Fifth | |
| 2646 Floor, Boston, MA 02110, USA. | |
| 2647 | |
| 2648 Verbatim copying and distribution of this entire article is permitted | |
| 2649 in any medium, provided this notice is preserved. | |
| 2650 Last modified 2009-01-25 [15]Valid XHTML 1.0 | |
| 2651 | |
| 2652 References | |
| 2653 | |
| 2654 1. http://gcc.gnu.org/gcc-4.1/changes.html#4.1.2 | |
| 2655 2. http://developer.classpath.org/mediation/ClasspathGraphicsImagesText | |
| 2656 3. http://gcc.gnu.org/ml/java/2005-05/msg00260.html | |
| 2657 4. http://developer.classpath.org/doc/ | |
| 2658 5. http://gcc.gnu.org/onlinedocs/gcc-4.1.0/gcc/Atomic-Builtins.html | |
| 2659 6. http://gcc.gnu.org/bugzilla/buglist.cgi?bug_status=RESOLVED&resolution=FIXED&target_milestone=4.1.2 | |
| 2660 7. mailto:gnu@gnu.org | |
| 2661 8. http://www.gnu.org/home.html#ContactInfo | |
| 2662 9. http://gcc.gnu.org/about.html | |
| 2663 10. http://gcc.gnu.org/onlinedocs/ | |
| 2664 11. mailto:gcc-help@gcc.gnu.org | |
| 2665 12. mailto:gcc@gnu.org | |
| 2666 13. mailto:gcc@gcc.gnu.org | |
| 2667 14. http://gcc.gnu.org/lists.html | |
| 2668 15. http://validator.w3.org/check/referer | |
| 2669 ====================================================================== | |
| 2670 http://gcc.gnu.org/gcc-4.0/index.html | |
| 2671 GCC 4.0 Release Series | |
| 2672 | |
| 2673 January 31, 2007 | |
| 2674 | |
| 2675 The [1]GNU project and the GCC developers are pleased to announce the | |
| 2676 release of GCC 4.0.4. | |
| 2677 | |
| 2678 This release is a bug-fix release, containing fixes for regressions in | |
| 2679 GCC 4.0.3 relative to previous releases of GCC. | |
| 2680 | |
| 2681 Release History | |
| 2682 | |
| 2683 GCC 4.0.4 | |
| 2684 January 31, 2007 ([2]changes) | |
| 2685 | |
| 2686 GCC 4.0.3 | |
| 2687 March 10, 2006 ([3]changes) | |
| 2688 | |
| 2689 GCC 4.0.2 | |
| 2690 September 28, 2005 ([4]changes) | |
| 2691 | |
| 2692 GCC 4.0.1 | |
| 2693 July 7, 2005 ([5]changes) | |
| 2694 | |
| 2695 GCC 4.0.0 | |
| 2696 April 20, 2005 ([6]changes) | |
| 2697 | |
| 2698 References and Acknowledgements | |
| 2699 | |
| 2700 GCC used to stand for the GNU C Compiler, but since the compiler | |
| 2701 supports several other languages aside from C, it now stands for the | |
| 2702 GNU Compiler Collection. | |
| 2703 | |
| 2704 A list of [7]successful builds is updated as new information becomes | |
| 2705 available. | |
| 2706 | |
| 2707 The GCC developers would like to thank the numerous people that have | |
| 2708 contributed new features, improvements, bug fixes, and other changes as | |
| 2709 well as test results to GCC. This [8]amazing group of volunteers is | |
| 2710 what makes GCC successful. | |
| 2711 | |
| 2712 For additional information about GCC please refer to the [9]GCC project | |
| 2713 web site or contact the [10]GCC development mailing list. | |
| 2714 | |
| 2715 To obtain GCC please use [11]our mirror sites, or [12]our SVN server. | |
| 2716 | |
| 2717 Please send FSF & GNU inquiries & questions to [13]gnu@gnu.org. There | |
| 2718 are also [14]other ways to contact the FSF. | |
| 2719 | |
| 2720 These pages are maintained by [15]the GCC team. | |
| 2721 | |
| 2722 | |
| 2723 For questions related to the use of GCC, please consult these web | |
| 2724 pages and the [16]GCC manuals. If that fails, the | |
| 2725 [17]gcc-help@gcc.gnu.org mailing list might help. | |
| 2726 Please send comments on these web pages and the development of GCC to | |
| 2727 our developer mailing list at [18]gcc@gnu.org or [19]gcc@gcc.gnu.org. | |
| 2728 All of our lists have [20]public archives. | |
| 2729 | |
| 2730 Copyright (C) Free Software Foundation, Inc., 51 Franklin St, Fifth | |
| 2731 Floor, Boston, MA 02110, USA. | |
| 2732 | |
| 2733 Verbatim copying and distribution of this entire article is permitted | |
| 2734 in any medium, provided this notice is preserved. | |
| 2735 Last modified 2008-07-26 [21]Valid XHTML 1.0 | |
| 2736 | |
| 2737 References | |
| 2738 | |
| 2739 1. http://www.gnu.org/ | |
| 2740 2. http://gcc.gnu.org/gcc-4.0/changes.html#4.0.4 | |
| 2741 3. http://gcc.gnu.org/gcc-4.0/changes.html#4.0.3 | |
| 2742 4. http://gcc.gnu.org/gcc-4.0/changes.html#4.0.2 | |
| 2743 5. http://gcc.gnu.org/gcc-4.0/changes.html#4.0.1 | |
| 2744 6. http://gcc.gnu.org/gcc-4.0/changes.html | |
| 2745 7. http://gcc.gnu.org/gcc-4.0/buildstat.html | |
| 2746 8. http://gcc.gnu.org/onlinedocs/gcc/Contributors.html | |
| 2747 9. http://gcc.gnu.org/index.html | |
| 2748 10. mailto:gcc@gcc.gnu.org | |
| 2749 11. http://gcc.gnu.org/mirrors.html | |
| 2750 12. http://gcc.gnu.org/svn.html | |
| 2751 13. mailto:gnu@gnu.org | |
| 2752 14. http://www.gnu.org/home.html#ContactInfo | |
| 2753 15. http://gcc.gnu.org/about.html | |
| 2754 16. http://gcc.gnu.org/onlinedocs/ | |
| 2755 17. mailto:gcc-help@gcc.gnu.org | |
| 2756 18. mailto:gcc@gnu.org | |
| 2757 19. mailto:gcc@gcc.gnu.org | |
| 2758 20. http://gcc.gnu.org/lists.html | |
| 2759 21. http://validator.w3.org/check/referer | |
| 2760 ====================================================================== | |
| 2761 http://gcc.gnu.org/gcc-4.0/changes.html | |
| 2762 GCC 4.0 Release Series | |
| 2763 Changes, New Features, and Fixes | |
| 2764 | |
| 2765 The latest release in the 4.0 release series is [1]GCC 4.0.4. | |
| 2766 | |
| 2767 Caveats | |
| 2768 | |
| 2769 * GCC now generates location lists by default when compiling with | |
| 2770 debug info and optimization. | |
| 2771 + GDB 6.0 and older crashes when it sees location lists. GDB 6.1 | |
| 2772 or later is needed to debug binaries containing location | |
| 2773 lists. | |
| 2774 + When you are trying to view a value of a variable in a part of | |
| 2775 a function where it has no location (for example when the | |
| 2776 variable is no longer used and thus its location was used for | |
| 2777 something else) GDB will say that it is not available. | |
| 2778 You can disable generating location lists by -fno-var-tracking. | |
| 2779 * GCC no longer accepts the -fwritable-strings option. Use named | |
| 2780 character arrays when you need a writable string. | |
| 2781 * The options -freduce-all-givs and -fmove-all-movables have been | |
| 2782 discontinued. They were used to circumvent a shortcoming in the | |
| 2783 heuristics of the old loop optimization code with respect to common | |
| 2784 Fortran constructs. The new (tree) loop optimizer works differently | |
| 2785 and doesn't need those work-arounds. | |
| 2786 * The graph-coloring register allocator, formerly enabled by the | |
| 2787 option -fnew-ra, has been discontinued. | |
| 2788 * -I- has been deprecated. -iquote is meant to replace the need for | |
| 2789 this option. | |
| 2790 * The MIPS -membedded-pic and -mrnames options have been removed. | |
| 2791 * All MIPS targets now require the GNU assembler. In particular, IRIX | |
| 2792 configurations can no longer use the MIPSpro assemblers, although | |
| 2793 they do still support the MIPSpro linkers. | |
| 2794 * The SPARC option -mflat has been removed. | |
| 2795 * English-language diagnostic messages will now use Unicode quotation | |
| 2796 marks in UTF-8 locales. (Non-English messages already used the | |
| 2797 quotes appropriate for the language in previous releases.) If your | |
| 2798 terminal does not support UTF-8 but you are using a UTF-8 locale | |
| 2799 (such locales are the default on many GNU/Linux systems) then you | |
| 2800 should set LC_CTYPE=C in the environment to disable that locale. | |
| 2801 Programs that parse diagnostics and expect plain ASCII | |
| 2802 English-language messages should set LC_ALL=C. See [2]Markus Kuhn's | |
| 2803 explanation of Unicode quotation marks for more information. | |
| 2804 * The specs file is no longer installed on most platforms. Most users | |
| 2805 will be totally unaffected. However, if you are accustomed to | |
| 2806 editing the specs file yourself, you will now have to use the | |
| 2807 -dumpspecs option to generate the specs file, and then edit the | |
| 2808 resulting file. | |
| 2809 | |
| 2810 General Optimizer Improvements | |
| 2811 | |
| 2812 * The [3]tree ssa branch has been merged. This merge has brought in a | |
| 2813 completely new optimization framework based on a higher level | |
| 2814 intermediate representation than the existing RTL representation. | |
| 2815 Numerous new code transformations based on the new framework are | |
| 2816 available in GCC 4.0, including: | |
| 2817 + Scalar replacement of aggregates | |
| 2818 + Constant propagation | |
| 2819 + Value range propagation | |
| 2820 + Partial redundancy elimination | |
| 2821 + Load and store motion | |
| 2822 + Strength reduction | |
| 2823 + Dead store elimination | |
| 2824 + Dead and unreachable code elimination | |
| 2825 + [4]Autovectorization | |
| 2826 + Loop interchange | |
| 2827 + Tail recursion by accumulation | |
| 2828 Many of these passes outperform their counterparts from previous | |
| 2829 GCC releases. | |
| 2830 * [5]Swing Modulo Scheduling (SMS). An RTL level instruction | |
| 2831 scheduling optimization intended for loops that perform heavy | |
| 2832 computations. | |
| 2833 | |
| 2834 New Languages and Language specific improvements | |
| 2835 | |
| 2836 C family | |
| 2837 | |
| 2838 * The sentinel attribute has been added to GCC. This function | |
| 2839 attribute allows GCC to warn when variadic functions such as execl | |
| 2840 are not NULL terminated. See the GCC manual for a complete | |
| 2841 description of its behavior. | |
| 2842 * Given __attribute__((alias("target"))) it is now an error if target | |
| 2843 is not a symbol, defined in the same translation unit. This also | |
| 2844 applies to aliases created by #pragma weak alias=target. This is | |
| 2845 because it's meaningless to define an alias to an undefined symbol. | |
| 2846 On Solaris, the native assembler would have caught this error, but | |
| 2847 GNU as does not. | |
| 2848 | |
| 2849 C and Objective-C | |
| 2850 | |
| 2851 * The -Wstrict-aliasing=2 option has been added. This warning catches | |
| 2852 all unsafe cases, but it may also give a warning for some cases | |
| 2853 that are safe. | |
| 2854 * The cast-as-lvalue, conditional-expression-as-lvalue and | |
| 2855 compound-expression-as-lvalue extensions, which were deprecated in | |
| 2856 3.3.4 and 3.4, have been removed. | |
| 2857 * The -fwritable-strings option, which was deprecated in 3.4, has | |
| 2858 been removed. | |
| 2859 * #pragma pack() semantics have been brought closer to those used by | |
| 2860 other compilers. This also applies to C++. | |
| 2861 * Taking the address of a variable with register storage is invalid | |
| 2862 in C. GCC now issues an error instead of a warning. | |
| 2863 * Arrays of incomplete element type are invalid in C. GCC now issues | |
| 2864 an error for such arrays. Declarations such as extern struct s x[]; | |
| 2865 (where struct s has not been defined) can be moved after the | |
| 2866 definition of struct s. Function parameters declared as arrays of | |
| 2867 incomplete type can instead be declared as pointers. | |
| 2868 | |
| 2869 C++ | |
| 2870 | |
| 2871 * When compiling without optimizations (-O0), the C++ frontend is | |
| 2872 much faster than in any previous versions of GCC. Independent | |
| 2873 testers have measured speed-ups up to 25% in real-world production | |
| 2874 code, compared to the 3.4 family (which was already the fastest | |
| 2875 version to date). Upgrading from older versions might show even | |
| 2876 bigger improvements. | |
| 2877 * ELF visibility attributes can now be applied to a class type, so | |
| 2878 that it affects every member function of a class at once, without | |
| 2879 having to specify each individually: | |
| 2880 class __attribute__ ((visibility("hidden"))) Foo | |
| 2881 { | |
| 2882 int foo1(); | |
| 2883 void foo2(); | |
| 2884 }; | |
| 2885 The syntax is deliberately similar to the __declspec() system used | |
| 2886 by Microsoft Windows based compilers, allowing cross-platform | |
| 2887 projects to easily reuse their existing macro system for denoting | |
| 2888 exports and imports. By explicitly marking internal classes never | |
| 2889 used outside a binary as hidden, one can completely avoid PLT | |
| 2890 indirection overheads during their usage by the compiler. You can | |
| 2891 find out more about the advantages of this at | |
| 2892 [6]http://people.redhat.com/drepper/dsohowto.pdf | |
| 2893 * The -fvisibility-inlines-hidden option has been added which marks | |
| 2894 all inlineable functions as having hidden ELF visibility, thus | |
| 2895 removing their symbol and typeinfo from the exported symbol table | |
| 2896 of the output ELF binary. Using this option can reduce the exported | |
| 2897 symbol count of template-heavy code by up to 40% with no code | |
| 2898 change at all, thus notably improving link and load times for the | |
| 2899 binary as well as a reduction in size of up to 10%. Also, check the | |
| 2900 new [7]-fvisibility option. | |
| 2901 * The compiler now uses the library interface specified by the [8]C++ | |
| 2902 ABI for thread-safe initialization of function-scope static | |
| 2903 variables. Most users should leave this alone, but embedded | |
| 2904 programmers may want to disable this by specifying | |
| 2905 -fno-threadsafe-statics for a small savings in code size. | |
| 2906 * Taking the address of an explicit register variable is no longer | |
| 2907 supported. Note that C++ allows taking the address of variables | |
| 2908 with register storage so this will continue to compile with a | |
| 2909 warning. For example, assuming that r0 is a machine register: | |
| 2910 register int foo asm ("r0"); | |
| 2911 register int bar; | |
| 2912 &foo; // error, no longer accepted | |
| 2913 &bar; // OK, with a warning | |
| 2914 * G++ has an undocumented extension to virtual function covariancy | |
| 2915 rules that allowed the overrider to return a type that was | |
| 2916 implicitly convertable to the overridden function's return type. | |
| 2917 For instance a function returning void * could be overridden by a | |
| 2918 function returning T *. This is now deprecated and will be removed | |
| 2919 in a future release. | |
| 2920 * The G++ minimum and maximum operators (<? and >?) and their | |
| 2921 compound forms (<?=) and >?=) have been deprecated and will be | |
| 2922 removed in a future version. Code using these operators should be | |
| 2923 modified to use std::min and std::max instead. | |
| 2924 * Declaration of nested classes of class templates as friends are | |
| 2925 supported: | |
| 2926 template <typename T> struct A { | |
| 2927 class B {}; | |
| 2928 }; | |
| 2929 class C { | |
| 2930 template <typename T> friend class A<T>::B; | |
| 2931 }; | |
| 2932 This complements the feature member functions of class templates as | |
| 2933 friends introduced in GCC 3.4.0. | |
| 2934 * When declaring a friend class using an unqualified name, classes | |
| 2935 outside the innermost non-class scope are not searched: | |
| 2936 class A; | |
| 2937 namespace N { | |
| 2938 class B { | |
| 2939 friend class A; // Refer to N::A which has not been declared yet | |
| 2940 // because name outside namespace N are not searched | |
| 2941 friend class ::A; // Refer to ::A | |
| 2942 }; | |
| 2943 } | |
| 2944 Hiding the friend name until declaration is still not implemented. | |
| 2945 * Friends of classes defined outside their namespace are correctly | |
| 2946 handled: | |
| 2947 namespace N { | |
| 2948 class A; | |
| 2949 } | |
| 2950 class N::A { | |
| 2951 friend class B; // Refer to N::B in GCC 4.0.0 | |
| 2952 // but ::B in earlier versions of GCC | |
| 2953 }; | |
| 2954 | |
| 2955 Runtime Library (libstdc++) | |
| 2956 | |
| 2957 * Optimization work: | |
| 2958 + Added efficient specializations of istream functions for char | |
| 2959 and wchar_t. | |
| 2960 + Further performance tuning of strings, in particular wrt | |
| 2961 single-char append and getline. | |
| 2962 + iter_swap - and therefore most of the mutating algorithms - | |
| 2963 now makes an unqualified call to swap when the value_type of | |
| 2964 the two iterators is the same. | |
| 2965 * A large subset of the features in Technical Report 1 (TR1 for | |
| 2966 short) is experimentally delivered (i.e., no guarantees about the | |
| 2967 implementation are provided. In particular it is not promised that | |
| 2968 the library will remain link-compatible when code using TR1 is | |
| 2969 used): | |
| 2970 + General utilities such as reference_wrapper and shared_ptr. | |
| 2971 + Function objects, i.e., result_of, mem_fn, bind, function. | |
| 2972 + Support for metaprogramming. | |
| 2973 + New containers such as tuple, array, unordered_set, | |
| 2974 unordered_map, unordered_multiset, unordered_multimap. | |
| 2975 * As usual, many bugs have been fixed and LWG resolutions implemented | |
| 2976 for the first time (e.g., DR 409). | |
| 2977 | |
| 2978 Java | |
| 2979 | |
| 2980 * In order to prevent naming conflicts with other implementations of | |
| 2981 these tools, some GCJ binaries have been renamed: | |
| 2982 + rmic is now grmic, | |
| 2983 + rmiregistry is now grmiregistry, and | |
| 2984 + jar is now fastjar. | |
| 2985 In particular, these names were problematic for the jpackage.org | |
| 2986 packaging conventions which install symlinks in /usr/bin that point | |
| 2987 to the preferred versions of these tools. | |
| 2988 * The -findirect-dispatch argument to the compiler now works and | |
| 2989 generates code following a new "binary compatibility" ABI. Code | |
| 2990 compiled this way follows the binary compatibility rules of the | |
| 2991 Java Language Specification. | |
| 2992 * libgcj now has support for using GCJ as a JIT, using the | |
| 2993 gnu.gcj.jit family of system properties. | |
| 2994 * libgcj can now find a shared library corresponding to the bytecode | |
| 2995 representation of a class. See the documentation for the new | |
| 2996 gcj-dbtool program, and the new gnu.gcj.precompiled.db.path system | |
| 2997 property. | |
| 2998 * There have been many improvements to the class library. Here are | |
| 2999 some highlights: | |
| 3000 + Much more of AWT and Swing exist. | |
| 3001 + Many new packages and classes were added, including | |
| 3002 java.util.regex, java.net.URI, javax.crypto, | |
| 3003 javax.crypto.interfaces, javax.crypto.spec, javax.net, | |
| 3004 javax.net.ssl, javax.security.auth, | |
| 3005 javax.security.auth.callback, javax.security.auth.login, | |
| 3006 javax.security.auth.x500, javax.security.sasl, org.ietf.jgss, | |
| 3007 javax.imageio, javax.imageio.event, javax.imageio.spi, | |
| 3008 javax.print, javax.print.attribute, | |
| 3009 javax.print.attribute.standard, javax.print.event, and | |
| 3010 javax.xml | |
| 3011 + Updated SAX and DOM, and imported GNU JAXP | |
| 3012 | |
| 3013 Fortran | |
| 3014 | |
| 3015 * A new [9]Fortran front end has replaced the aging GNU Fortran 77 | |
| 3016 front end. The new front end supports Fortran 90 and Fortran 95. It | |
| 3017 may not yet be as stable as the old Fortran front end. | |
| 3018 | |
| 3019 Ada | |
| 3020 | |
| 3021 * Ada (with tasking and Zero Cost Exceptions) is now available on | |
| 3022 many more targets, including but not limited to: alpha-linux, | |
| 3023 hppa-hpux, hppa-linux, powerpc-darwin, powerpc-linux, s390-linux, | |
| 3024 s390x-linux, sparc-linux. | |
| 3025 * Some of the new Ada 2005 features are now implemented like | |
| 3026 Wide_Wide_Character and Ada.Containers. | |
| 3027 * Many bugs have been fixed, tools and documentation improved. | |
| 3028 * To compile Ada from the sources, install an older working Ada | |
| 3029 compiler and then use --enable-languages=ada at configuration time, | |
| 3030 since the Ada frontend is not currently activated by default. See | |
| 3031 the [10]Installing GCC for details. | |
| 3032 | |
| 3033 New Targets and Target Specific Improvements | |
| 3034 | |
| 3035 H8/300 | |
| 3036 | |
| 3037 * The frame layout has changed. In the new layout, the prologue of a | |
| 3038 function first saves registers and then allocate space for locals, | |
| 3039 resulting in an 1% improvement on code size. | |
| 3040 | |
| 3041 IA-32/x86-64 (AMD64) | |
| 3042 | |
| 3043 * The acos, asin, drem, exp10, exp2, expm1, fmod, ilogb, log10, | |
| 3044 log1p, log2, logb and tan mathematical builtins (and their float | |
| 3045 and long double variants) are now implemented as inline x87 | |
| 3046 intrinsics when using -ffast-math. | |
| 3047 * The ceil, floor, nearbyint, rint and trunc mathematical builtins | |
| 3048 (and their float and long double variants) are now implemented as | |
| 3049 inline x87 intrinsics when using -ffast-math. | |
| 3050 * The x87's fsincos instruction is now used automatically with | |
| 3051 -ffast-math when calculating both the sin and cos of the same | |
| 3052 argument. | |
| 3053 * Instruction selection for multiplication and division by constants | |
| 3054 has been improved. | |
| 3055 | |
| 3056 IA-64 | |
| 3057 | |
| 3058 * Floating point division, integer division and sqrt are now inlined, | |
| 3059 resulting in significant performance improvements on some codes. | |
| 3060 | |
| 3061 MIPS | |
| 3062 | |
| 3063 * Division by zero checks now use conditional traps if the target | |
| 3064 processor supports them. This decreases code size by one word per | |
| 3065 division operation. The old behavior (branch and break) can be | |
| 3066 obtained either at configure time by passing --with-divide=breaks | |
| 3067 to configure or at runtime by passing -mdivide-breaks to GCC. | |
| 3068 * Support for MIPS64 paired-single instructions has been added. It is | |
| 3069 enabled by -mpaired-single and can be accessed using both the | |
| 3070 target-independent vector extensions and new MIPS-specific built-in | |
| 3071 functions. | |
| 3072 * Support for the MIPS-3D ASE has been added. It is enabled by | |
| 3073 -mips3d and provides new MIPS-3D-specific built-in functions. | |
| 3074 * The -mexplicit-relocs option now supports static n64 code (as is | |
| 3075 used, for example, in 64-bit linux kernels). -mexplicit-relocs | |
| 3076 should now be feature-complete and is enabled by default when GCC | |
| 3077 is configured to use a compatible assembler. | |
| 3078 * Support for the NEC VR4130 series has been added. This support | |
| 3079 includes the use of VR-specific instructions and a new VR4130 | |
| 3080 scheduler. Full VR4130 support can be selected with -march=vr4130 | |
| 3081 while code for any ISA can be tuned for the VR4130 using | |
| 3082 -mtune=vr4130. There is also a new -mvr4130-align option that | |
| 3083 produces better schedules at the cost of increased code size. | |
| 3084 * Support for the Broadcom SB-1 has been extended. There is now an | |
| 3085 SB-1 scheduler as well as support for the SB-1-specific | |
| 3086 paired-single instructions. Full SB-1 support can be selected with | |
| 3087 -march=sb1 while code for any ISA can be optimized for the SB-1 | |
| 3088 using -mtune=sb1. | |
| 3089 * The compiler can now work around errata in R4000, R4400, VR4120 and | |
| 3090 VR4130 processors. These workarounds are enabled by -mfix-r4000, | |
| 3091 -mfix-r4400, -mfix-vr4120 and -mfix-vr4130 respectively. The VR4120 | |
| 3092 and VR4130 workarounds need binutils 2.16 or above. | |
| 3093 * IRIX shared libraries are now installed into the standard library | |
| 3094 directories: o32 libraries go into lib/, n32 libraries go into | |
| 3095 lib32/ and n64 libraries go into lib64/. | |
| 3096 * The compiler supports a new -msym32 option. It can be used to | |
| 3097 optimize n64 code in which all symbols are known to have 32-bit | |
| 3098 values. | |
| 3099 | |
| 3100 S/390 and zSeries | |
| 3101 | |
| 3102 * New command-line options help to generate code intended to run in | |
| 3103 an environment where stack space is restricted, e.g. Linux kernel | |
| 3104 code: | |
| 3105 + -mwarn-framesize and -mwarn-dynamicstack trigger compile-time | |
| 3106 warnings for single functions that require large or dynamic | |
| 3107 stack frames. | |
| 3108 + -mstack-size and -mstack-guard generate code that checks for | |
| 3109 stack overflow at run time. | |
| 3110 + -mpacked-stack generates code that reduces the stack frame | |
| 3111 size of many functions by reusing unneeded parts of the stack | |
| 3112 bias area. | |
| 3113 * The -msoft-float option now ensures that generated code never | |
| 3114 accesses floating point registers. | |
| 3115 * The s390x-ibm-tpf target now fully supports C++, including | |
| 3116 exceptions and threads. | |
| 3117 * Various changes to improve performance of the generated code have | |
| 3118 been implemented, including: | |
| 3119 + GCC now uses sibling calls where possible. | |
| 3120 + Condition code handling has been optimized, allowing GCC to | |
| 3121 omit redundant comparisons in certain cases. | |
| 3122 + The cost function guiding many optimizations has been refined | |
| 3123 to more accurately represent the z900 and z990 processors. | |
| 3124 + The ADD LOGICAL WITH CARRY and SUBTRACT LOGICAL WITH BORROW | |
| 3125 instructions are now used to avoid conditional branches in | |
| 3126 certain cases. | |
| 3127 + The back end now uses the LEGITIMIZE_RELOAD_ADDRESS feature to | |
| 3128 optimize address arithmetic required to access large stack | |
| 3129 frames. | |
| 3130 + GCC now makes more efficient use of memory-to-memory type | |
| 3131 instructions (MVC, CLC, ...). | |
| 3132 + More precise tracking of special register use allows better | |
| 3133 instruction scheduling, in particular of the function prologue | |
| 3134 and epilogue sequences. | |
| 3135 + The Java front end now generates inline code to implement | |
| 3136 integer division, instead of calling library routines. | |
| 3137 | |
| 3138 SPARC | |
| 3139 | |
| 3140 * The options -mv8, -msparclite, -mcypress, -msupersparc, -mf930 and | |
| 3141 -mf934 have been removed. They have been replaced with -mcpu=xxx. | |
| 3142 * The internal model used to estimate the relative cost of each | |
| 3143 instruction has been updated. It is expected to give better results | |
| 3144 on recent UltraSPARC processors. | |
| 3145 * Code generation for function prologues and epilogues has been | |
| 3146 improved, resulting in better scheduling and allowing multiple exit | |
| 3147 points in functions. | |
| 3148 * Support for Sun's Visual Instruction Set (VIS) has been enhanced. | |
| 3149 It is enabled by -mvis and provides new built-in functions for VIS | |
| 3150 instructions on UltraSPARC processors. | |
| 3151 * The option -mapp-regs has been turned on by default on Solaris too. | |
| 3152 | |
| 3153 NetWare | |
| 3154 | |
| 3155 * Novell NetWare (on ix86, no other hardware platform was ever really | |
| 3156 supported by this OS) has been re-enabled and the ABI supported by | |
| 3157 GCC has been brought into sync with that of MetroWerks CodeWarrior | |
| 3158 (the ABI previously supported was that of some Unix systems, which | |
| 3159 NetWare never tried to support). | |
| 3160 | |
| 3161 Obsolete Systems | |
| 3162 | |
| 3163 Support for a number of older systems has been declared obsolete in GCC | |
| 3164 4.0. Unless there is activity to revive them, the next release of GCC | |
| 3165 will have their sources permanently removed. | |
| 3166 | |
| 3167 All GCC ports for the following processor architectures have been | |
| 3168 declared obsolete: | |
| 3169 * Intel i860 | |
| 3170 * Ubicom IP2022 | |
| 3171 * National Semiconductor NS32K | |
| 3172 * Texas Instruments TMS320C[34]x | |
| 3173 | |
| 3174 Also, those for some individual systems have been obsoleted: | |
| 3175 * SPARC family | |
| 3176 + SPARClite-based systems (sparclite-*-coff, sparclite-*-elf, | |
| 3177 sparc86x-*-elf) | |
| 3178 + OpenBSD 32-bit (sparc-*-openbsd*) | |
| 3179 | |
| 3180 Documentation improvements | |
| 3181 | |
| 3182 Other significant improvements | |
| 3183 | |
| 3184 * Location lists are now generated by default when compiling with | |
| 3185 debug info and optimization. Location lists provide more accurate | |
| 3186 debug info about locations of variables and they allow debugging | |
| 3187 code compiled with -fomit-frame-pointer. | |
| 3188 * The -fvisibility option has been added which allows the default ELF | |
| 3189 visibility of all symbols to be set per compilation and the new | |
| 3190 #pragma GCC visibility preprocessor command allows the setting of | |
| 3191 default ELF visibility for a region of code. Using | |
| 3192 -fvisibility=hidden especially in combination with the new | |
| 3193 -fvisibility-inlines-hidden can yield substantial improvements in | |
| 3194 output binary quality including avoiding PLT indirection overheads, | |
| 3195 reduction of the exported symbol count by up to 60% (with resultant | |
| 3196 improvements to link and load times), better scope for the | |
| 3197 optimizer to improve code and up to a 20% reduction in binary size. | |
| 3198 Using these options correctly yields a binary with a similar symbol | |
| 3199 count to a Windows DLL. | |
| 3200 Perhaps more importantly, this new feature finally allows (with | |
| 3201 careful planning) complete avoidance of symbol clashes when | |
| 3202 manually loading shared objects with RTLD_GLOBAL, thus finally | |
| 3203 solving problems many projects such as python were forced to use | |
| 3204 RTLD_LOCAL for (with its resulting issues for C++ correctness). You | |
| 3205 can find more information about using these options at | |
| 3206 [11]http://gcc.gnu.org/wiki/Visibility. | |
| 3207 __________________________________________________________________ | |
| 3208 | |
| 3209 GCC 4.0.1 | |
| 3210 | |
| 3211 This is the [12]list of problem reports (PRs) from GCC's bug tracking | |
| 3212 system that are known to be fixed in the 4.0.1 release. This list might | |
| 3213 not be complete (that is, it is possible that some PRs that have been | |
| 3214 fixed are not listed here). | |
| 3215 | |
| 3216 GCC 4.0.2 | |
| 3217 | |
| 3218 This is the [13]list of problem reports (PRs) from GCC's bug tracking | |
| 3219 system that are known to be fixed in the 4.0.2 release. This list might | |
| 3220 not be complete (that is, it is possible that some PRs that have been | |
| 3221 fixed are not listed here). | |
| 3222 | |
| 3223 Unfortunately, due to a release engineering failure, this release has a | |
| 3224 regression on Solaris that will affect some C++ programs. We suggest | |
| 3225 that Solaris users apply a [14]patch that corrects the problem. Users | |
| 3226 who do not wish to apply the patch should explicitly link C++ programs | |
| 3227 with the -pthreads option, even if they do not use threads. This | |
| 3228 problem has been corrected in the current 4.0 branch sources and will | |
| 3229 not be present in GCC 4.0.3. | |
| 3230 | |
| 3231 GCC 4.0.3 | |
| 3232 | |
| 3233 Starting with this release, the function getcontext is recognized by | |
| 3234 the compiler as having the same semantics as the setjmp function. In | |
| 3235 particular, the compiler will ensure that all registers are dead before | |
| 3236 calling such a function and will emit a warning about the variables | |
| 3237 that may be clobbered after the second return from the function. | |
| 3238 | |
| 3239 GCC 4.0.4 | |
| 3240 | |
| 3241 This is the [15]list of problem reports (PRs) from GCC's bug tracking | |
| 3242 system that are known to be fixed in the 4.0.4 release. This list might | |
| 3243 not be complete (that is, it is possible that some PRs that have been | |
| 3244 fixed are not listed here). | |
| 3245 | |
| 3246 The 4.0.4 release is provided for those that require a high degree of | |
| 3247 binary compatibility with previous 4.0.x releases. For most users, the | |
| 3248 GCC team recommends that version 4.1.1 or later be used instead." | |
| 3249 | |
| 3250 Please send FSF & GNU inquiries & questions to [16]gnu@gnu.org. There | |
| 3251 are also [17]other ways to contact the FSF. | |
| 3252 | |
| 3253 These pages are maintained by [18]the GCC team. | |
| 3254 | |
| 3255 | |
| 3256 For questions related to the use of GCC, please consult these web | |
| 3257 pages and the [19]GCC manuals. If that fails, the | |
| 3258 [20]gcc-help@gcc.gnu.org mailing list might help. | |
| 3259 Please send comments on these web pages and the development of GCC to | |
| 3260 our developer mailing list at [21]gcc@gnu.org or [22]gcc@gcc.gnu.org. | |
| 3261 All of our lists have [23]public archives. | |
| 3262 | |
| 3263 Copyright (C) Free Software Foundation, Inc., 51 Franklin St, Fifth | |
| 3264 Floor, Boston, MA 02110, USA. | |
| 3265 | |
| 3266 Verbatim copying and distribution of this entire article is permitted | |
| 3267 in any medium, provided this notice is preserved. | |
| 3268 Last modified 2009-01-25 [24]Valid XHTML 1.0 | |
| 3269 | |
| 3270 References | |
| 3271 | |
| 3272 1. http://gcc.gnu.org/gcc-4.0/changes.html#4.0.4 | |
| 3273 2. http://www.cl.cam.ac.uk/~mgk25/ucs/quotes.html | |
| 3274 3. http://gcc.gnu.org/projects/tree-ssa/ | |
| 3275 4. http://gcc.gnu.org/projects/tree-ssa/vectorization.html | |
| 3276 5. http://gcc.gnu.org/news/sms.html | |
| 3277 6. http://people.redhat.com/drepper/dsohowto.pdf | |
| 3278 7. http://gcc.gnu.org/gcc-4.0/changes.html#visibility | |
| 3279 8. http://www.codesourcery.com/public/cxx-abi/ | |
| 3280 9. http://gcc.gnu.org/fortran/ | |
| 3281 10. http://gcc.gnu.org/install/ | |
| 3282 11. http://gcc.gnu.org/wiki/Visibility | |
| 3283 12. http://gcc.gnu.org/bugzilla/buglist.cgi?bug_status=RESOLVED&resolution=FIXED&target_milestone=4.0.1 | |
| 3284 13. http://gcc.gnu.org/bugzilla/buglist.cgi?bug_status=RESOLVED&resolution=FIXED&target_milestone=4.0.2 | |
| 3285 14. http://gcc.gnu.org/ml/gcc-cvs/2005-09/msg00984.html | |
| 3286 15. http://gcc.gnu.org/bugzilla/buglist.cgi?bug_status=RESOLVED&resolution=FIXED&target_milestone=4.0.4 | |
| 3287 16. mailto:gnu@gnu.org | |
| 3288 17. http://www.gnu.org/home.html#ContactInfo | |
| 3289 18. http://gcc.gnu.org/about.html | |
| 3290 19. http://gcc.gnu.org/onlinedocs/ | |
| 3291 20. mailto:gcc-help@gcc.gnu.org | |
| 3292 21. mailto:gcc@gnu.org | |
| 3293 22. mailto:gcc@gcc.gnu.org | |
| 3294 23. http://gcc.gnu.org/lists.html | |
| 3295 24. http://validator.w3.org/check/referer | |
| 3296 ====================================================================== | |
| 3297 http://gcc.gnu.org/gcc-3.4/index.html | |
| 3298 GCC 3.4 Release Series | |
| 3299 | |
| 3300 May 26, 2006 | |
| 3301 | |
| 3302 The [1]GNU project and the GCC developers are pleased to announce the | |
| 3303 release of GCC 3.4.6. | |
| 3304 | |
| 3305 This release is a bug-fix release, containing fixes for regressions in | |
| 3306 GCC 3.4.4 relative to previous releases of GCC. This is the last of the | |
| 3307 3.4.x series. | |
| 3308 | |
| 3309 The GCC 3.4 release series includes numerous [2]new features, | |
| 3310 improvements, bug fixes, and other changes, thanks to an [3]amazing | |
| 3311 group of volunteers. | |
| 3312 | |
| 3313 Release History | |
| 3314 | |
| 3315 GCC 3.4.6 | |
| 3316 March 6, 2006 ([4]changes) | |
| 3317 | |
| 3318 GCC 3.4.5 | |
| 3319 November 30, 2005 ([5]changes) | |
| 3320 | |
| 3321 GCC 3.4.4 | |
| 3322 May 18, 2005 ([6]changes) | |
| 3323 | |
| 3324 GCC 3.4.3 | |
| 3325 November 4, 2004 ([7]changes) | |
| 3326 | |
| 3327 GCC 3.4.2 | |
| 3328 September 6, 2004 ([8]changes) | |
| 3329 | |
| 3330 GCC 3.4.1 | |
| 3331 July 1, 2004 ([9]changes) | |
| 3332 | |
| 3333 GCC 3.4.0 | |
| 3334 April 18, 2004 ([10]changes) | |
| 3335 | |
| 3336 References and Acknowledgements | |
| 3337 | |
| 3338 GCC used to stand for the GNU C Compiler, but since the compiler | |
| 3339 supports several other languages aside from C, it now stands for the | |
| 3340 GNU Compiler Collection. | |
| 3341 | |
| 3342 A list of [11]successful builds is updated as new information becomes | |
| 3343 available. | |
| 3344 | |
| 3345 The GCC developers would like to thank the numerous people that have | |
| 3346 contributed new features, improvements, bug fixes, and other changes as | |
| 3347 well as test results to GCC. This [12]amazing group of volunteers is | |
| 3348 what makes GCC successful. | |
| 3349 | |
| 3350 For additional information about GCC please refer to the [13]GCC | |
| 3351 project web site or contact the [14]GCC development mailing list. | |
| 3352 | |
| 3353 To obtain GCC please use [15]our mirror sites, or [16]our SVN server. | |
| 3354 | |
| 3355 Please send FSF & GNU inquiries & questions to [17]gnu@gnu.org. There | |
| 3356 are also [18]other ways to contact the FSF. | |
| 3357 | |
| 3358 These pages are maintained by [19]the GCC team. | |
| 3359 | |
| 3360 | |
| 3361 For questions related to the use of GCC, please consult these web | |
| 3362 pages and the [20]GCC manuals. If that fails, the | |
| 3363 [21]gcc-help@gcc.gnu.org mailing list might help. | |
| 3364 Please send comments on these web pages and the development of GCC to | |
| 3365 our developer mailing list at [22]gcc@gnu.org or [23]gcc@gcc.gnu.org. | |
| 3366 All of our lists have [24]public archives. | |
| 3367 | |
| 3368 Copyright (C) Free Software Foundation, Inc., 51 Franklin St, Fifth | |
| 3369 Floor, Boston, MA 02110, USA. | |
| 3370 | |
| 3371 Verbatim copying and distribution of this entire article is permitted | |
| 3372 in any medium, provided this notice is preserved. | |
| 3373 Last modified 2008-07-26 [25]Valid XHTML 1.0 | |
| 3374 | |
| 3375 References | |
| 3376 | |
| 3377 1. http://www.gnu.org/ | |
| 3378 2. http://gcc.gnu.org/gcc-3.4/changes.html | |
| 3379 3. http://gcc.gnu.org/onlinedocs/gcc/Contributors.html | |
| 3380 4. http://gcc.gnu.org/gcc-3.4/changes.html#3.4.6 | |
| 3381 5. http://gcc.gnu.org/gcc-3.4/changes.html#3.4.5 | |
| 3382 6. http://gcc.gnu.org/gcc-3.4/changes.html#3.4.4 | |
| 3383 7. http://gcc.gnu.org/gcc-3.4/changes.html#3.4.3 | |
| 3384 8. http://gcc.gnu.org/gcc-3.4/changes.html#3.4.2 | |
| 3385 9. http://gcc.gnu.org/gcc-3.4/changes.html#3.4.1 | |
| 3386 10. http://gcc.gnu.org/gcc-3.4/changes.html | |
| 3387 11. http://gcc.gnu.org/gcc-3.4/buildstat.html | |
| 3388 12. http://gcc.gnu.org/onlinedocs/gcc/Contributors.html | |
| 3389 13. http://gcc.gnu.org/index.html | |
| 3390 14. mailto:gcc@gcc.gnu.org | |
| 3391 15. http://gcc.gnu.org/mirrors.html | |
| 3392 16. http://gcc.gnu.org/svn.html | |
| 3393 17. mailto:gnu@gnu.org | |
| 3394 18. http://www.gnu.org/home.html#ContactInfo | |
| 3395 19. http://gcc.gnu.org/about.html | |
| 3396 20. http://gcc.gnu.org/onlinedocs/ | |
| 3397 21. mailto:gcc-help@gcc.gnu.org | |
| 3398 22. mailto:gcc@gnu.org | |
| 3399 23. mailto:gcc@gcc.gnu.org | |
| 3400 24. http://gcc.gnu.org/lists.html | |
| 3401 25. http://validator.w3.org/check/referer | |
| 3402 ====================================================================== | |
| 3403 http://gcc.gnu.org/gcc-3.4/changes.html | |
| 3404 GCC 3.4 Release Series | |
| 3405 Changes, New Features, and Fixes | |
| 3406 | |
| 3407 The final release in the 3.4 release series is [1]GCC 3.4.6. The series | |
| 3408 is now closed. | |
| 3409 | |
| 3410 GCC 3.4 has [2]many improvements in the C++ frontend. Before reporting | |
| 3411 a bug, please make sure it's really GCC, and not your code, that is | |
| 3412 broken. | |
| 3413 | |
| 3414 Caveats | |
| 3415 | |
| 3416 * GNU Make is now required to build GCC. | |
| 3417 * With -nostdinc the preprocessor used to ignore both standard | |
| 3418 include paths and include paths contained in environment variables. | |
| 3419 It was neither documented nor intended that environment variable | |
| 3420 paths be ignored, so this has been corrected. | |
| 3421 * GCC no longer accepts the options -fvolatile, -fvolatile-global and | |
| 3422 -fvolatile-static. It is unlikely that they worked correctly in any | |
| 3423 3.x release. | |
| 3424 * GCC no longer ships <varargs.h>. Use <stdarg.h> instead. | |
| 3425 * Support for all the systems [3]obsoleted in GCC 3.3 has been | |
| 3426 removed from GCC 3.4. See below for a [4]list of systems which are | |
| 3427 obsoleted in this release. | |
| 3428 * GCC now requires an ISO C90 (ANSI C89) C compiler to build. K&R C | |
| 3429 compilers will not work. | |
| 3430 * The implementation of the [5]MIPS ABIs has changed. As a result, | |
| 3431 the code generated for certain MIPS targets will not be binary | |
| 3432 compatible with earlier releases. | |
| 3433 * In previous releases, the MIPS port had a fake "hilo" register with | |
| 3434 the user-visible name accum. This register has been removed. | |
| 3435 * The implementation of the [6]SPARC ABIs has changed. As a result, | |
| 3436 the code generated will not be binary compatible with earlier | |
| 3437 releases in certain cases. | |
| 3438 * The configure option --enable-threads=pthreads has been removed; | |
| 3439 use --enable-threads=posix instead, which should have the same | |
| 3440 effect. | |
| 3441 * Code size estimates used by inlining heuristics for C, Objective-C, | |
| 3442 C++ and Java have been redesigned significantly. As a result the | |
| 3443 parameters of -finline-insns, --param max-inline-insns-single and | |
| 3444 --param max-inline-insns-auto need to be reconsidered. | |
| 3445 * --param max-inline-slope and --param min-inline-insns have been | |
| 3446 removed; they are not needed for the new bottom-up inlining | |
| 3447 heuristics. | |
| 3448 * The new unit-at-a-time compilation scheme has several compatibility | |
| 3449 issues: | |
| 3450 + The order in which functions, variables, and top-level asm | |
| 3451 statements are emitted may have changed. Code relying on some | |
| 3452 particular ordering needs to be updated. The majority of such | |
| 3453 top-level asm statements can be replaced by section | |
| 3454 attributes. | |
| 3455 + Unreferenced static variables and functions are removed. This | |
| 3456 may result in undefined references when an asm statement | |
| 3457 refers to the variable/function directly. In that case either | |
| 3458 the variable/function shall be listed in asm statement operand | |
| 3459 or in the case of top-level asm statements the attribute used | |
| 3460 shall be used to force function/variable to be always output | |
| 3461 and considered as a possibly used by unknown code. | |
| 3462 For variables the attribute is accepted only by GCC 3.4 and | |
| 3463 newer, while for earlier versions it is sufficient to use | |
| 3464 unused to silence warnings about the variables not being | |
| 3465 referenced. To keep code portable across different GCC | |
| 3466 versions, you can use appropriate preprocessor conditionals. | |
| 3467 + Static functions now can use non-standard passing conventions | |
| 3468 that may break asm statements calling functions directly. | |
| 3469 Again the attribute used shall be used to prevent this | |
| 3470 behavior. | |
| 3471 As a temporary workaround, -fno-unit-at-a-time can be used, but | |
| 3472 this scheme may not be supported by future releases of GCC. | |
| 3473 * GCC 3.4 automatically places zero-initialized variables in the .bss | |
| 3474 section on some operating systems. Versions of GNU Emacs up to (and | |
| 3475 including) 21.3 will not work correctly when using this | |
| 3476 optimization; you can use -fno-zero-initialized-in-bss to disable | |
| 3477 it. | |
| 3478 * If GCC 3.4 is configured with --enable-threads=posix (the default | |
| 3479 on most targets that support pthreads) then _REENTRANT will be | |
| 3480 defined unconditionally by some libstdc++ headers. C++ code which | |
| 3481 relies on that macro to detect whether multi-threaded code is being | |
| 3482 compiled might change in meaning, possibly resulting in linker | |
| 3483 errors for single-threaded programs. Affected users of [7]Boost | |
| 3484 should compile single-threaded code with -DBOOST_DISABLE_THREADS. | |
| 3485 See Bugzilla for [8]more information. | |
| 3486 | |
| 3487 General Optimizer Improvements | |
| 3488 | |
| 3489 * Usability of the profile feedback and coverage testing has been | |
| 3490 improved. | |
| 3491 + Performance of profiled programs has been improved by faster | |
| 3492 profile merging code. | |
| 3493 + Better use of the profile feedback for optimization (loop | |
| 3494 unrolling and loop peeling). | |
| 3495 + File locking support allowing fork() calls and parallel runs | |
| 3496 of profiled programs. | |
| 3497 + Coverage file format has been redesigned. | |
| 3498 + gcov coverage tool has been improved. | |
| 3499 + make profiledbootstrap available to build a faster compiler. | |
| 3500 Experiments made on i386 hardware showed an 11% speedup on -O0 | |
| 3501 and a 7.5% speedup on -O2 compilation of a [9]large C++ | |
| 3502 testcase. | |
| 3503 + New value profiling pass enabled via -fprofile-values | |
| 3504 + New value profile transformations pass enabled via -fvpt aims | |
| 3505 to optimize some code sequences by exploiting knowledge about | |
| 3506 value ranges or other properties of the operands. At the | |
| 3507 moment a conversion of expensive divisions into cheaper | |
| 3508 operations has been implemented. | |
| 3509 + New -fprofile-generate and -fprofile-use command-line options | |
| 3510 to simplify the use of profile feedback. | |
| 3511 * A new unit-at-a-time compilation scheme for C, Objective-C, C++ and | |
| 3512 Java which is enabled via -funit-at-a-time (and implied by -O2). In | |
| 3513 this scheme a whole file is parsed first and optimized later. The | |
| 3514 following basic inter-procedural optimizations are implemented: | |
| 3515 + Removal of unreachable functions and variables | |
| 3516 + Discovery of local functions (functions with static linkage | |
| 3517 whose address is never taken) | |
| 3518 + On i386, these local functions use register parameter passing | |
| 3519 conventions. | |
| 3520 + Reordering of functions in topological order of the call graph | |
| 3521 to enable better propagation of optimizing hints (such as the | |
| 3522 stack alignments needed by functions) in the back end. | |
| 3523 + Call graph based out-of-order inlining heuristics which allows | |
| 3524 to limit overall compilation unit growth (--param | |
| 3525 inline-unit-growth). | |
| 3526 Overall, the unit-at-a-time scheme produces a 1.3% improvement for | |
| 3527 the SPECint2000 benchmark on the i386 architecture (AMD Athlon | |
| 3528 CPU). | |
| 3529 * More realistic code size estimates used by inlining for C, | |
| 3530 Objective-C, C++ and Java. The growth of large functions can now be | |
| 3531 limited via --param large-function-insns and --param | |
| 3532 large-function-growth. | |
| 3533 * A new cfg-level loop optimizer pass replaces the old loop unrolling | |
| 3534 pass and adds two other loop transformations -- loop peeling and | |
| 3535 loop unswitching -- and also uses the profile feedback to limit | |
| 3536 code growth. (The three optimizations are enabled by | |
| 3537 -funroll-loops, -fpeel-loops and -funswitch-loops flags, | |
| 3538 respectively). | |
| 3539 The old loop unroller still can be enabled by -fold-unroll-loops | |
| 3540 and may produce better code in some cases, especially when the | |
| 3541 webizer optimization pass is not run. | |
| 3542 * A new web construction pass enabled via -fweb (and implied by -O3) | |
| 3543 improves the quality of register allocation, CSE, first scheduling | |
| 3544 pass and some other optimization passes by avoiding re-use of | |
| 3545 pseudo registers with non-overlapping live ranges. The pass almost | |
| 3546 always improves code quality but does make debugging difficult and | |
| 3547 thus is not enabled by default by -O2 | |
| 3548 The pass is especially effective as cleanup after code duplication | |
| 3549 passes, such as the loop unroller or the tracer. | |
| 3550 * Experimental implementations of superblock or trace scheduling in | |
| 3551 the second scheduling pass can be enabled via | |
| 3552 -fsched2-use-superblocks and -fsched2-use-traces, respectively. | |
| 3553 | |
| 3554 New Languages and Language specific improvements | |
| 3555 | |
| 3556 Ada | |
| 3557 | |
| 3558 * The Ada front end has been updated to include numerous bug fixes | |
| 3559 and enhancements. These include: | |
| 3560 + Improved project file support | |
| 3561 + Additional set of warnings about potential wrong code | |
| 3562 + Improved error messages | |
| 3563 + Improved code generation | |
| 3564 + Improved cross reference information | |
| 3565 + Improved inlining | |
| 3566 + Better run-time check elimination | |
| 3567 + Better error recovery | |
| 3568 + More efficient implementation of unbounded strings | |
| 3569 + Added features in GNAT.Sockets, GNAT.OS_Lib, GNAT.Debug_Pools, | |
| 3570 ... | |
| 3571 + New GNAT.xxxx packages (e.g. GNAT.Strings, | |
| 3572 GNAT.Exception_Action) | |
| 3573 + New pragmas | |
| 3574 + New -gnatS switch replacing gnatpsta | |
| 3575 + Implementation of new Ada features (in particular limited | |
| 3576 with, limited aggregates) | |
| 3577 | |
| 3578 C/Objective-C/C++ | |
| 3579 | |
| 3580 * Precompiled headers are now supported. Precompiled headers can | |
| 3581 dramatically speed up compilation of some projects. There are some | |
| 3582 known defects in the current precompiled header implementation that | |
| 3583 will result in compiler crashes in relatively rare situations. | |
| 3584 Therefore, precompiled headers should be considered a "technology | |
| 3585 preview" in this release. Read the manual for details about how to | |
| 3586 use precompiled headers. | |
| 3587 * File handling in the preprocessor has been rewritten. GCC no longer | |
| 3588 gets confused by symlinks and hardlinks, and now has a correct | |
| 3589 implementation of #import and #pragma once. These two directives | |
| 3590 have therefore been un-deprecated. | |
| 3591 * The undocumented extension that allowed C programs to have a label | |
| 3592 at the end of a compound statement, which has been deprecated since | |
| 3593 GCC 3.0, has been removed. | |
| 3594 * The cast-as-lvalue extension has been removed for C++ and | |
| 3595 deprecated for C and Objective-C. In particular, code like this: | |
| 3596 int i; | |
| 3597 (char) i = 5; | |
| 3598 | |
| 3599 or this: | |
| 3600 char *p; | |
| 3601 ((int *) p)++; | |
| 3602 | |
| 3603 is no longer accepted for C++ and will not be accepted for C and | |
| 3604 Objective-C in a future version. | |
| 3605 * The conditional-expression-as-lvalue extension has been deprecated | |
| 3606 for C and Objective-C. In particular, code like this: | |
| 3607 int a, b, c; | |
| 3608 (a ? b : c) = 2; | |
| 3609 | |
| 3610 will not be accepted for C and Objective-C in a future version. | |
| 3611 * The compound-expression-as-lvalue extension has been deprecated for | |
| 3612 C and Objective-C. In particular, code like this: | |
| 3613 int a, b; | |
| 3614 (a, b) = 2; | |
| 3615 | |
| 3616 will not be accepted for C and Objective-C in a future version. A | |
| 3617 possible non-intrusive workaround is the following: | |
| 3618 (*(a, &b)) = 2; | |
| 3619 | |
| 3620 * Several [10]built-in functions such as __builtin_popcount for | |
| 3621 counting bits, finding the highest and lowest bit in a word, and | |
| 3622 parity have been added. | |
| 3623 * The -fwritable-strings option has been deprecated and will be | |
| 3624 removed. | |
| 3625 * Many C math library functions are now recognized as built-ins and | |
| 3626 optimized. | |
| 3627 * The C, C++, and Objective-C compilers can now handle source files | |
| 3628 written in any character encoding supported by the host C library. | |
| 3629 The default input character set is taken from the current locale, | |
| 3630 and may be overridden with the -finput-charset command line option. | |
| 3631 In the future we will add support for inline encoding markers. | |
| 3632 | |
| 3633 C++ | |
| 3634 | |
| 3635 * G++ is now much closer to full conformance to the ISO/ANSI C++ | |
| 3636 standard. This means, among other things, that a lot of invalid | |
| 3637 constructs which used to be accepted in previous versions will now | |
| 3638 be rejected. It is very likely that existing C++ code will need to | |
| 3639 be fixed. This document lists some of the most common issues. | |
| 3640 * A hand-written recursive-descent C++ parser has replaced the | |
| 3641 YACC-derived C++ parser from previous GCC releases. The new parser | |
| 3642 contains much improved infrastructure needed for better parsing of | |
| 3643 C++ source codes, handling of extensions, and clean separation | |
| 3644 (where possible) between proper semantics analysis and parsing. The | |
| 3645 new parser fixes many bugs that were found in the old parser. | |
| 3646 * You must now use the typename and template keywords to disambiguate | |
| 3647 dependent names, as required by the C++ standard. | |
| 3648 struct K { | |
| 3649 typedef int mytype_t; | |
| 3650 }; | |
| 3651 | |
| 3652 template <class T1> struct A { | |
| 3653 template <class T2> struct B { | |
| 3654 void callme(void); | |
| 3655 }; | |
| 3656 | |
| 3657 template <int N> void bar(void) | |
| 3658 { | |
| 3659 // Use 'typename' to tell the parser that T1::mytype_t names | |
| 3660 // a type. This is needed because the name is dependent (in | |
| 3661 // this case, on template parameter T1). | |
| 3662 typename T1::mytype_t x; | |
| 3663 x = 0; | |
| 3664 } | |
| 3665 }; | |
| 3666 | |
| 3667 template <class T> void template_func(void) | |
| 3668 { | |
| 3669 // Use 'template' to prefix member templates within | |
| 3670 // dependent types (a has type A<T>, which depends on | |
| 3671 // the template parameter T). | |
| 3672 A<T> a; | |
| 3673 a.template bar<0>(); | |
| 3674 | |
| 3675 // Use 'template' to tell the parser that B is a nested | |
| 3676 // template class (dependent on template parameter T), and | |
| 3677 // 'typename' because the whole A<T>::B<int> is | |
| 3678 // the name of a type (again, dependent). | |
| 3679 typename A<T>::template B<int> b; | |
| 3680 b.callme(); | |
| 3681 } | |
| 3682 | |
| 3683 void non_template_func(void) | |
| 3684 { | |
| 3685 // Outside of any template class or function, no names can be | |
| 3686 // dependent, so the use of the keyword 'typename' and 'template' | |
| 3687 // is not needed (and actually forbidden). | |
| 3688 A<K> a; | |
| 3689 a.bar<0>(); | |
| 3690 A<K>::B<float> b; | |
| 3691 b.callme(); | |
| 3692 } | |
| 3693 * In a template definition, unqualified names will no longer find | |
| 3694 members of a dependent base (as specified by [temp.dep]/3 in the | |
| 3695 C++ standard). For example, | |
| 3696 template <typename T> struct B { | |
| 3697 int m; | |
| 3698 int n; | |
| 3699 int f (); | |
| 3700 int g (); | |
| 3701 }; | |
| 3702 int n; | |
| 3703 int g (); | |
| 3704 template <typename T> struct C : B<T> { | |
| 3705 void h () | |
| 3706 { | |
| 3707 m = 0; // error | |
| 3708 f (); // error | |
| 3709 n = 0; // ::n is modified | |
| 3710 g (); // ::g is called | |
| 3711 } | |
| 3712 }; | |
| 3713 You must make the names dependent, e.g. by prefixing them with | |
| 3714 this->. Here is the corrected definition of C<T>::h, | |
| 3715 template <typename T> void C<T>::h () | |
| 3716 { | |
| 3717 this->m = 0; | |
| 3718 this->f (); | |
| 3719 this->n = 0 | |
| 3720 this->g (); | |
| 3721 } | |
| 3722 As an alternative solution (unfortunately not backwards compatible | |
| 3723 with GCC 3.3), you may use using declarations instead of this->: | |
| 3724 template <typename T> struct C : B<T> { | |
| 3725 using B<T>::m; | |
| 3726 using B<T>::f; | |
| 3727 using B<T>::n; | |
| 3728 using B<T>::g; | |
| 3729 void h () | |
| 3730 { | |
| 3731 m = 0; | |
| 3732 f (); | |
| 3733 n = 0; | |
| 3734 g (); | |
| 3735 } | |
| 3736 }; | |
| 3737 * In templates, all non-dependent names are now looked up and bound | |
| 3738 at definition time (while parsing the code), instead of later when | |
| 3739 the template is instantiated. For instance: | |
| 3740 void foo(int); | |
| 3741 | |
| 3742 template <int> struct A { | |
| 3743 static void bar(void){ | |
| 3744 foo('a'); | |
| 3745 } | |
| 3746 }; | |
| 3747 | |
| 3748 void foo(char); | |
| 3749 | |
| 3750 int main() | |
| 3751 { | |
| 3752 A<0>::bar(); // Calls foo(int), used to call foo(char). | |
| 3753 } | |
| 3754 | |
| 3755 * In an explicit instantiation of a class template, you must use | |
| 3756 class or struct before the template-id: | |
| 3757 template <int N> | |
| 3758 class A {}; | |
| 3759 | |
| 3760 template A<0>; // error, not accepted anymore | |
| 3761 template class A<0>; // OK | |
| 3762 * The "named return value" and "implicit typename" extensions have | |
| 3763 been removed. | |
| 3764 * Default arguments in function types have been deprecated and will | |
| 3765 be removed. | |
| 3766 * ARM-style name-injection of friend declarations has been deprecated | |
| 3767 and will be removed. For example: struct S { friend void f(); }; | |
| 3768 void g() { f(); } will not be accepted by future versions of G++; | |
| 3769 instead a declaration of "f" will need to be present outside of the | |
| 3770 scope of "S". | |
| 3771 * Covariant returns are implemented for all but varadic functions | |
| 3772 that require an adjustment. | |
| 3773 * When -pedantic is used, G++ now issues errors about spurious | |
| 3774 semicolons. For example, | |
| 3775 namespace N {}; // Invalid semicolon. | |
| 3776 void f() {}; // Invalid semicolon. | |
| 3777 * G++ no longer accepts attributes for a declarator after the | |
| 3778 initializer associated with that declarator. For example, | |
| 3779 X x(1) __attribute__((...)); | |
| 3780 is no longer accepted. Instead, use: | |
| 3781 X x __attribute__((...)) (1); | |
| 3782 * Inside the scope of a template class, the name of the class itself | |
| 3783 can be treated as either a class or a template. So GCC used to | |
| 3784 accept the class name as argument of type template, and template | |
| 3785 template parameter. However this is not C++ standard compliant. Now | |
| 3786 the name is not treated as a valid template template argument | |
| 3787 unless you qualify the name by its scope. For example, the code | |
| 3788 below no longer compiles. | |
| 3789 template <template <class> class TT> class X {}; | |
| 3790 template <class T> class Y { | |
| 3791 X<Y> x; // Invalid, Y is always a type template parameter. | |
| 3792 }; | |
| 3793 The valid code for the above example is | |
| 3794 X< ::Y> x; // Valid. | |
| 3795 (Notice the space between < and : to prevent GCC to interpret this | |
| 3796 as a digraph for [.) | |
| 3797 * Friend declarations that refer to template specializations are | |
| 3798 rejected if the template has not already been declared. For | |
| 3799 example, | |
| 3800 template <typename T> | |
| 3801 class C { | |
| 3802 friend void f<> (C&); | |
| 3803 }; | |
| 3804 is rejected. You must first declare f as a template, | |
| 3805 template <typename T> | |
| 3806 void f(T); | |
| 3807 * In case of friend declarations, every name used in the friend | |
| 3808 declaration must be accessible at the point of that declaration. | |
| 3809 Previous versions of G++ used to be less strict about this and | |
| 3810 allowed friend declarations for private class members, for example. | |
| 3811 See the ISO C++ Standard Committee's [11]defect report #209 for | |
| 3812 details. | |
| 3813 * Declaration of member functions of class templates as friends are | |
| 3814 supported. For example, | |
| 3815 template <typename T> struct A { | |
| 3816 void f(); | |
| 3817 }; | |
| 3818 class C { | |
| 3819 template <typename T> friend void A<T>::f(); | |
| 3820 }; | |
| 3821 * You must use template <> to introduce template specializations, as | |
| 3822 required by the standard. For example, | |
| 3823 template <typename T> | |
| 3824 struct S; | |
| 3825 | |
| 3826 struct S<int> { }; | |
| 3827 is rejected. You must write, | |
| 3828 template <> struct S<int> {}; | |
| 3829 * G++ used to accept code like this, | |
| 3830 struct S { | |
| 3831 int h(); | |
| 3832 void f(int i = g()); | |
| 3833 int g(int i = h()); | |
| 3834 }; | |
| 3835 This behavior is not mandated by the standard. Now G++ issues an | |
| 3836 error about this code. To avoid the error, you must move the | |
| 3837 declaration of g before the declaration of f. The default arguments | |
| 3838 for g must be visible at the point where it is called. | |
| 3839 * The C++ ABI Section 3.3.3 specifications for the array construction | |
| 3840 routines __cxa_vec_new2 and __cxa_vec_new3 were changed to return | |
| 3841 NULL when the allocator argument returns NULL. These changes are | |
| 3842 incorporated into the libstdc++ runtime library. | |
| 3843 * Using a name introduced by a typedef in a friend declaration or in | |
| 3844 an explicit instantiation is now rejected, as specified by the ISO | |
| 3845 C++ standard. | |
| 3846 class A; | |
| 3847 typedef A B; | |
| 3848 class C { | |
| 3849 friend class B; // error, no typedef name here | |
| 3850 friend B; // error, friend always needs class/struct/enum | |
| 3851 friend class A; // OK | |
| 3852 }; | |
| 3853 | |
| 3854 template <int> class Q {}; | |
| 3855 typedef Q<0> R; | |
| 3856 template class R; // error, no typedef name here | |
| 3857 template class Q<0>; // OK | |
| 3858 * When allocating an array with a new expression, GCC used to allow | |
| 3859 parentheses around the type name. This is actually ill-formed and | |
| 3860 it is now rejected: | |
| 3861 int* a = new (int)[10]; // error, not accepted anymore | |
| 3862 int* a = new int[10]; // OK | |
| 3863 * When binding an rvalue of class type to a reference, the copy | |
| 3864 constructor of the class must be accessible. For instance, consider | |
| 3865 the following code: | |
| 3866 class A | |
| 3867 { | |
| 3868 public: | |
| 3869 A(); | |
| 3870 | |
| 3871 private: | |
| 3872 A(const A&); // private copy ctor | |
| 3873 }; | |
| 3874 | |
| 3875 A makeA(void); | |
| 3876 void foo(const A&); | |
| 3877 | |
| 3878 void bar(void) | |
| 3879 { | |
| 3880 foo(A()); // error, copy ctor is not accessible | |
| 3881 foo(makeA()); // error, copy ctor is not accessible | |
| 3882 | |
| 3883 A a1; | |
| 3884 foo(a1); // OK, a1 is a lvalue | |
| 3885 } | |
| 3886 This might be surprising at first sight, especially since most | |
| 3887 popular compilers do not correctly implement this rule ([12]further | |
| 3888 details). | |
| 3889 * When forming a pointer to member or a pointer to member function, | |
| 3890 access checks for class visibility (public, protected, private) are | |
| 3891 now performed using the qualifying scope of the name itself. This | |
| 3892 is better explained with an example: | |
| 3893 class A | |
| 3894 { | |
| 3895 public: | |
| 3896 void pub_func(); | |
| 3897 protected: | |
| 3898 void prot_func(); | |
| 3899 private: | |
| 3900 void priv_func(); | |
| 3901 }; | |
| 3902 | |
| 3903 class B : public A | |
| 3904 { | |
| 3905 public: | |
| 3906 void foo() | |
| 3907 { | |
| 3908 &A::pub_func; // OK, pub_func is accessible through A | |
| 3909 &A::prot_func; // error, cannot access prot_func through A | |
| 3910 &A::priv_func; // error, cannot access priv_func through A | |
| 3911 | |
| 3912 &B::pub_func; // OK, pub_func is accessible through B | |
| 3913 &B::prot_func; // OK, can access prot_func through B (within B) | |
| 3914 &B::priv_func; // error, cannot access priv_func through B | |
| 3915 } | |
| 3916 }; | |
| 3917 | |
| 3918 Runtime Library (libstdc++) | |
| 3919 | |
| 3920 * Optimization work: | |
| 3921 + Streamlined streambuf, filebuf, separate synched with C | |
| 3922 Standard I/O streambuf. | |
| 3923 + All formatted I/O now uses cached locale information. | |
| 3924 + STL optimizations (memory/speed for list, red-black trees as | |
| 3925 used by sets and maps). | |
| 3926 + More use of GCC builtins. | |
| 3927 + String optimizations (avoid contention on | |
| 3928 increment/decrement-and-test of the reference count in the | |
| 3929 empty-string object, constructor from input_iterators | |
| 3930 speedup). | |
| 3931 * Static linkage size reductions. | |
| 3932 * Large File Support (files larger than 2 GB on 32-bit systems). | |
| 3933 * Wide character and variable encoding filebuf work (UTF-8, Unicode). | |
| 3934 * Generic character traits. | |
| 3935 * Also support wchar_t specializations on Mac OS 10.3.x, FreeBSD 5.x, | |
| 3936 Solaris 2.7 and above, AIX 5.x, Irix 6.5. | |
| 3937 * The allocator class is now standard-conformant, and two additional | |
| 3938 extension allocators have been added, mt_alloc and | |
| 3939 bitmap_allocator. | |
| 3940 * PCH support: -include bits/stdc++.h (2x compile speedup). | |
| 3941 * Rewrote __cxa_demangle with support for C++ style allocators. | |
| 3942 * New debug modes for STL containers and iterators. | |
| 3943 * Testsuite rewrite: five times as many tests, plus increasingly | |
| 3944 sophisticated tests, including I/O, MT, multi-locale, wide and | |
| 3945 narrow characters. | |
| 3946 * Use current versions of GNU "autotools" for build/configuration. | |
| 3947 | |
| 3948 Objective-C | |
| 3949 | |
| 3950 * The Objective-C front end has been updated to include the numerous | |
| 3951 bug fixes and enhancements previously available only in Apple's | |
| 3952 version of GCC. These include: | |
| 3953 + Structured exception (@try... @catch... @finally, @throw) and | |
| 3954 synchronization (@synchronized) support. These are accessible | |
| 3955 via the -fobjc-exceptions switch; as of this writing, they may | |
| 3956 only be used in conjunction with -fnext-runtime on Mac OS X | |
| 3957 10.3 and later. See [13]Options Controlling Objective-C | |
| 3958 Dialect for more information. | |
| 3959 + An overhaul of @encode logic. The C99 _Bool and C++ bool type | |
| 3960 may now be encoded as 'B'. In addition, the back-end/codegen | |
| 3961 dependencies have been removed. | |
| 3962 + An overhaul of message dispatch construction, ensuring that | |
| 3963 the various receiver types (and casts thereof) are handled | |
| 3964 properly, and that correct diagnostics are issued. | |
| 3965 + Support for "Zero-Link" (-fzero-link) and "Fix-and-Continue" | |
| 3966 (-freplace-objc-classes) debugging modes, currently available | |
| 3967 on Mac OS X 10.3 and later. See [14]Options Controlling | |
| 3968 Objective-C Dialect for more information. | |
| 3969 + Access to optimized runtime entry points (-fno-nil-receivers ) | |
| 3970 on the assumption that message receivers are never nil. This | |
| 3971 is currently available on Mac OS X 10.3 and later. See | |
| 3972 [15]Options Controlling Objective-C Dialect for more | |
| 3973 information. | |
| 3974 | |
| 3975 Java | |
| 3976 | |
| 3977 * Compiling a .jar file will now cause non-.class entries to be | |
| 3978 automatically compiled as resources. | |
| 3979 * libgcj has been ported to Darwin. | |
| 3980 * Jeff Sturm has adapted Jan Hubicka's call graph optimization code | |
| 3981 to gcj. | |
| 3982 * libgcj has a new gcjlib URL type; this lets URLClassLoader load | |
| 3983 code from shared libraries. | |
| 3984 * libgcj has been much more completely merged with [16]GNU Classpath. | |
| 3985 * Class loading is now much more correct; in particular the caller's | |
| 3986 class loader is now used when that is required. | |
| 3987 * [17]Eclipse 2.x will run out of the box using gij. | |
| 3988 * Parts of java.nio have been implemented. Direct and indirect | |
| 3989 buffers work, as do fundamental file and socket operations. | |
| 3990 * java.awt has been improved, though it is still not ready for | |
| 3991 general use. | |
| 3992 * The HTTP protocol handler now uses HTTP/1.1 and can handle the POST | |
| 3993 method. | |
| 3994 * The MinGW port has matured. Enhancements include socket timeout | |
| 3995 support, thread interruption, improved Runtime.exec() handling and | |
| 3996 support for accented characters in filenames. | |
| 3997 | |
| 3998 Fortran | |
| 3999 | |
| 4000 * Fortran improvements are listed in the [18]Fortran documentation. | |
| 4001 | |
| 4002 New Targets and Target Specific Improvements | |
| 4003 | |
| 4004 Alpha | |
| 4005 | |
| 4006 * Several [19]built-in functions have been added such as | |
| 4007 __builtin_alpha_zap to allow utilizing the more obscure | |
| 4008 instructions of the CPU. | |
| 4009 * Parameter passing of complex arguments has changed to match the | |
| 4010 [20]ABI. This change is incompatible with previous GCC versions, | |
| 4011 but does fix compatibility with the Tru64 compiler and several | |
| 4012 corner cases where GCC was incompatible with itself. | |
| 4013 | |
| 4014 ARM | |
| 4015 | |
| 4016 * Nicolas Pitre has contributed his hand-coded floating-point support | |
| 4017 code for ARM. It is both significantly smaller and faster than the | |
| 4018 existing C-based implementation, even when building applications | |
| 4019 for Thumb. The arm-elf configuration has been converted to use the | |
| 4020 new code. | |
| 4021 * Support for the Intel's iWMMXt architecture, a second generation | |
| 4022 XScale processor, has been added. Enabled at run time with the | |
| 4023 -mcpu=iwmmxt command line switch. | |
| 4024 * A new ARM target has been added: arm-wince-pe. This is similar to | |
| 4025 the arm-pe target, but it defaults to using the APCS32 ABI. | |
| 4026 * The existing ARM pipeline description has been converted to the use | |
| 4027 the [21]DFA processor pipeline model. There is not much change in | |
| 4028 code performance, but the description is now [22]easier to | |
| 4029 understand. | |
| 4030 * Support for the Cirrus EP9312 Maverick floating point co-processor | |
| 4031 added. Enabled at run time with the -mcpu=ep9312 command line | |
| 4032 switch. Note however that the multilibs to support this chip are | |
| 4033 currently disabled in gcc/config/arm/t-arm-elf, so if you want to | |
| 4034 enable their production you will have to uncomment the entries in | |
| 4035 that file. | |
| 4036 | |
| 4037 H8/300 | |
| 4038 | |
| 4039 * Support for long long has been added. | |
| 4040 * Support for saveall attribute has been added. | |
| 4041 * Pavel Pisa contributed hand-written 32-bit-by-32-bit division code | |
| 4042 for H8/300H and H8S, which is much faster than the previous | |
| 4043 implementation. | |
| 4044 * A lot of small performance improvements. | |
| 4045 | |
| 4046 IA-32/AMD64 (x86-64) | |
| 4047 | |
| 4048 * Tuning for K8 (AMD Opteron/Athlon64) core is available via | |
| 4049 -march=k8 and -mcpu=k8. | |
| 4050 * Scalar SSE code generation carefully avoids reformatting penalties, | |
| 4051 hidden dependencies and minimizes the number of uops generated on | |
| 4052 both Intel and AMD CPUs. | |
| 4053 * Vector MMX and SSE operands are now passed in registers to improve | |
| 4054 performance and match the argument passing convention used by the | |
| 4055 Intel C++ Compiler. As a result it is not possible to call | |
| 4056 functions accepting vector arguments compiled by older GCC version. | |
| 4057 * Conditional jump elimination is now more aggressive on modern CPUs. | |
| 4058 * The Athlon ports has been converted to use the DFA processor | |
| 4059 pipeline description. | |
| 4060 * Optimization of indirect tail calls is now possible in a similar | |
| 4061 fashion as direct sibcall optimization. | |
| 4062 * Further small performance improvements. | |
| 4063 * -m128bit-long-double is now less buggy. | |
| 4064 * __float128 support in 64-bit compilation. | |
| 4065 * Support for data structures exceeding 2GB in 64-bit mode. | |
| 4066 * -mcpu has been renamed to -mtune. | |
| 4067 | |
| 4068 IA-64 | |
| 4069 | |
| 4070 * Tuning code for the Itanium 2 processor has been added. The | |
| 4071 generation of code tuned for Itanium 2 (option -mtune=itanium2) is | |
| 4072 enabled by default now. To generate code tuned for Itanium 1 the | |
| 4073 option -mtune=itanium1 should be used. | |
| 4074 * [23]DFA processor pipeline descriptions for the IA-64 processors | |
| 4075 have been added. This resulted in about 3% improvement on the | |
| 4076 SPECInt2000 benchmark for Itanium 2. | |
| 4077 * Instruction bundling for the IA-64 processors has been rewritten | |
| 4078 using the DFA pipeline hazard recognizer. It resulted in about 60% | |
| 4079 compiler speedup on the SPECInt2000 C programs. | |
| 4080 | |
| 4081 M32R | |
| 4082 | |
| 4083 * Support for the M32R/2 processor has been added by Renesas. | |
| 4084 * Support for an M32R Linux target and PIC code generation has been | |
| 4085 added by Renesas. | |
| 4086 | |
| 4087 M68000 | |
| 4088 | |
| 4089 * Bernardo Innocenti (Develer S.r.l.) has contributed the | |
| 4090 m68k-uclinux target, based on former work done by Paul Dale | |
| 4091 (SnapGear Inc.). Code generation for the ColdFire processors family | |
| 4092 has been enhanced and extended to support the MCF 53xx and MCF 54xx | |
| 4093 cores, integrating former work done by Peter Barada (Motorola). | |
| 4094 | |
| 4095 MIPS | |
| 4096 | |
| 4097 Processor-specific changes | |
| 4098 | |
| 4099 * Support for the RM7000 and RM9000 processors has been added. It can | |
| 4100 be selected using the -march compiler option and should work with | |
| 4101 any MIPS I (mips-*) or MIPS III (mips64-*) configuration. | |
| 4102 * Support for revision 2 of the MIPS32 ISA has been added. It can be | |
| 4103 selected with the command-line option -march=mips32r2. | |
| 4104 * There is a new option, -mfix-sb1, to work around certain SB-1 | |
| 4105 errata. | |
| 4106 | |
| 4107 Configuration | |
| 4108 | |
| 4109 * It is possible to customize GCC using the following configure-time | |
| 4110 options: | |
| 4111 + --with-arch, which specifies the default value of the -march | |
| 4112 option. | |
| 4113 + --with-tune, which specifies the default value of the -mtune | |
| 4114 option. | |
| 4115 + --with-abi, which specifies the default ABI. | |
| 4116 + --with-float=soft, which tells GCC to use software floating | |
| 4117 point by default. | |
| 4118 + --with-float=hard, which tells GCC to use hardware floating | |
| 4119 point by default. | |
| 4120 * A 64-bit GNU/Linux port has been added. The associated | |
| 4121 configurations are mips64-linux-gnu and mips64el-linux-gnu. | |
| 4122 * The 32-bit GNU/Linux port now supports Java. | |
| 4123 * The IRIX 6 configuration now supports the o32 ABI and will build | |
| 4124 o32 multilibs by default. This support is compatible with both | |
| 4125 binutils and the SGI tools, but note that several features, | |
| 4126 including debugging information and DWARF2 exception handling, are | |
| 4127 only available when using the GNU assembler. Use of the GNU | |
| 4128 assembler and linker (version 2.15 or above) is strongly | |
| 4129 recommended. | |
| 4130 * The IRIX 6 configuration now supports 128-bit long doubles. | |
| 4131 * There are two new RTEMS-specific configurations, mips-rtems and | |
| 4132 mipsel-rtems. | |
| 4133 * There are two new *-elf configurations, mipsisa32r2-elf and | |
| 4134 mipsisa32r2el-elf. | |
| 4135 | |
| 4136 General | |
| 4137 | |
| 4138 * Several [24]ABI bugs have been fixed. Unfortunately, these changes | |
| 4139 will break binary compatibility with earlier releases. | |
| 4140 * GCC can now use explicit relocation operators when generating | |
| 4141 -mabicalls code. This behavior is controlled by -mexplicit-relocs | |
| 4142 and can have several performance benefits. For example: | |
| 4143 + It allows for more optimization of GOT accesses, including | |
| 4144 better scheduling and redundancy elimination. | |
| 4145 + It allows sibling calls to be implemented as jumps. | |
| 4146 + n32 and n64 leaf functions can use a call-clobbered global | |
| 4147 pointer instead of $28. | |
| 4148 + The code to set up $gp can be removed from functions that | |
| 4149 don't need it. | |
| 4150 * A new option, -mxgot, allows the GOT to be bigger than 64k. This | |
| 4151 option is equivalent to the assembler's -xgot option and should be | |
| 4152 used instead of -Wa,-xgot. | |
| 4153 * Frame pointer elimination is now supported when generating 64-bit | |
| 4154 MIPS16 code. | |
| 4155 * Inline block moves have been optimized to take more account of | |
| 4156 alignment information. | |
| 4157 * Many internal changes have been made to the MIPS port, mostly aimed | |
| 4158 at reducing the reliance on assembler macros. | |
| 4159 | |
| 4160 PowerPC | |
| 4161 | |
| 4162 * GCC 3.4 releases have a number of fixes for PowerPC and PowerPC64 | |
| 4163 [25]ABI incompatibilities regarding the way parameters are passed | |
| 4164 during functions calls. These changes may result in incompatibility | |
| 4165 between code compiled with GCC 3.3 and GCC 3.4. | |
| 4166 | |
| 4167 PowerPC Darwin | |
| 4168 | |
| 4169 * Support for shared/dylib gcc libraries has been added. It is | |
| 4170 enabled by default on powerpc-apple-darwin7.0.0 and up. | |
| 4171 * Libgcj is enabled by default. On systems older than | |
| 4172 powerpc-apple-darwin7.0.0 you need to install dlcompat. | |
| 4173 * 128-bit IBM extended precision format support added for long | |
| 4174 double. | |
| 4175 | |
| 4176 PowerPC64 GNU/Linux | |
| 4177 | |
| 4178 * By default, PowerPC64 GNU/Linux now uses natural alignment of | |
| 4179 structure elements. The old four byte alignment for double, with | |
| 4180 special rules for a struct starting with a double, can be chosen | |
| 4181 with -malign-power. This change may result in incompatibility | |
| 4182 between code compiled with GCC 3.3 and GCC 3.4. | |
| 4183 * -mabi=altivec is now the default rather than -mabi=no-altivec. | |
| 4184 * 128-bit IBM extended precision format support added for long | |
| 4185 double. | |
| 4186 | |
| 4187 S/390 and zSeries | |
| 4188 | |
| 4189 * New command-line options allow to specify the intended execution | |
| 4190 environment for generated code: | |
| 4191 + -mesa/-mzarch allows to specify whether to generate code | |
| 4192 running in ESA/390 mode or in z/Architecture mode (this is | |
| 4193 applicable to 31-bit code only). | |
| 4194 + -march allows to specify a minimum processor architecture | |
| 4195 level (g5, g6, z900, or z990). | |
| 4196 + -mtune allows to specify which processor to tune for. | |
| 4197 * It is possible to customize GCC using the following configure-time | |
| 4198 options: | |
| 4199 + --with-mode, which specifies whether to default to assuming | |
| 4200 ESA/390 or z/Architecture mode. | |
| 4201 + --with-arch, which specifies the default value of the -march | |
| 4202 option. | |
| 4203 + --with-tune, which specifies the default value of the -mtune | |
| 4204 option. | |
| 4205 * Support for the z990 processor has been added, and can be selected | |
| 4206 using -march=z990 or -mtune=z990. This includes instruction | |
| 4207 scheduling tuned for the superscalar instruction pipeline of the | |
| 4208 z990 processor as well as support for all new instructions provided | |
| 4209 by the long-displacement facility. | |
| 4210 * Support to generate 31-bit code optimized for zSeries processors | |
| 4211 (running in ESA/390 or in z/Architecture mode) has been added. This | |
| 4212 can be selected using -march=z900 and -mzarch respectively. | |
| 4213 * Instruction scheduling for the z900 and z990 processors now uses | |
| 4214 the DFA pipeline hazard recognizer. | |
| 4215 * GCC no longer generates code to maintain a stack backchain, | |
| 4216 previously used to generate stack backtraces for debugging | |
| 4217 purposes. As replacement that does not incur runtime overhead, | |
| 4218 DWARF-2 call frame information is provided by GCC; this is | |
| 4219 supported by GDB 6.1. The old behavior can be restored using the | |
| 4220 -mbackchain option. | |
| 4221 * The stack frame size of functions may now exceed 2 GB in 64-bit | |
| 4222 code. | |
| 4223 * A port for the 64-bit IBM TPF operating system has been added; the | |
| 4224 configuration is s390x-ibm-tpf. This configuration is supported as | |
| 4225 cross-compilation target only. | |
| 4226 * Various changes to improve the generated code have been | |
| 4227 implemented, including: | |
| 4228 + GCC now uses the MULTIPLY AND ADD and MULTIPLY AND SUBTRACT | |
| 4229 instructions to significantly speed up many floating-point | |
| 4230 applications. | |
| 4231 + GCC now uses the ADD LOGICAL WITH CARRY and SUBTRACT LOGICAL | |
| 4232 WITH BORROW instructions to speed up long long arithmetic. | |
| 4233 + GCC now uses the SEARCH STRING instruction to implement | |
| 4234 strlen(). | |
| 4235 + In many cases, function call overhead for 31-bit code has been | |
| 4236 reduced by placing the literal pool after the function code | |
| 4237 instead of after the function prolog. | |
| 4238 + Register 14 is no longer reserved in 64-bit code. | |
| 4239 + Handling of global register variables has been improved. | |
| 4240 | |
| 4241 SPARC | |
| 4242 | |
| 4243 * The option -mflat is deprecated. | |
| 4244 * Support for large (> 2GB) frames has been added to the 64-bit port. | |
| 4245 * Several [26]ABI bugs have been fixed. Unfortunately, these changes | |
| 4246 will break binary compatibility with earlier releases. | |
| 4247 * The default debugging format has been switched from STABS to | |
| 4248 DWARF-2 for 32-bit code on Solaris 7 and later. DWARF-2 is already | |
| 4249 the default debugging format for 64-bit code on Solaris. | |
| 4250 | |
| 4251 SuperH | |
| 4252 | |
| 4253 * Support for the SH2E processor has been added. Enabled at run time | |
| 4254 with the -m2e command line switch, or at configure time by | |
| 4255 specifying sh2e as the machine part of the target triple. | |
| 4256 | |
| 4257 V850 | |
| 4258 | |
| 4259 * Support for the Mitsubishi V850E1 processor has been added. This is | |
| 4260 a variant of the V850E processor with some additional debugging | |
| 4261 instructions. | |
| 4262 | |
| 4263 Xtensa | |
| 4264 | |
| 4265 * Several ABI bugs have been fixed. Unfortunately, these changes | |
| 4266 break binary compatibility with earlier releases. | |
| 4267 + For big-endian processors, the padding of aggregate return | |
| 4268 values larger than a word has changed. If the size of an | |
| 4269 aggregate return value is not a multiple of 32 bits, previous | |
| 4270 versions of GCC inserted padding in the most-significant bytes | |
| 4271 of the first return value register. Aggregates larger than a | |
| 4272 word are now padded in the least-significant bytes of the last | |
| 4273 return value register used. Aggregates smaller than a word are | |
| 4274 still padded in the most-significant bytes. The return value | |
| 4275 padding has not changed for little-endian processors. | |
| 4276 + Function arguments with 16-byte alignment are now properly | |
| 4277 aligned. | |
| 4278 + The implementation of the va_list type has changed. A va_list | |
| 4279 value created by va_start from a previous release cannot be | |
| 4280 used with va_arg from this release, or vice versa. | |
| 4281 * More processor configuration options for Xtensa processors are | |
| 4282 supported: | |
| 4283 + the ABS instruction is now optional; | |
| 4284 + the ADDX* and SUBX* instructions are now optional; | |
| 4285 + an experimental CONST16 instruction can be used to synthesize | |
| 4286 constants instead of loading them from constant pools. | |
| 4287 These and other Xtensa processor configuration options can no | |
| 4288 longer be enabled or disabled by command-line options; the | |
| 4289 processor configuration must be specified by the xtensa-config.h | |
| 4290 header file when building GCC. Additionally, the | |
| 4291 -mno-serialize-volatile option is no longer supported. | |
| 4292 | |
| 4293 Obsolete Systems | |
| 4294 | |
| 4295 Support for a number of older systems has been declared obsolete in GCC | |
| 4296 3.4. Unless there is activity to revive them, the next release of GCC | |
| 4297 will have their sources permanently removed. | |
| 4298 | |
| 4299 All configurations of the following processor architectures have been | |
| 4300 declared obsolete: | |
| 4301 * Mitsubishi D30V, d30v-* | |
| 4302 * AT&T DSP1600 and DSP1610, dsp16xx-* | |
| 4303 * Intel 80960, i960 | |
| 4304 | |
| 4305 Also, some individual systems have been obsoleted: | |
| 4306 * ARM Family | |
| 4307 + Support for generating code for operation in APCS/26 mode | |
| 4308 (-mapcs-26). | |
| 4309 * IBM ESA/390 | |
| 4310 + "Bigfoot" port, i370-*. (The other port, s390-*, is actively | |
| 4311 maintained and supported.) | |
| 4312 * Intel 386 family | |
| 4313 + MOSS, i?86-moss-msdos and i?86-*-moss* | |
| 4314 + NCR 3000 running System V r.4, i?86-ncr-sysv4* | |
| 4315 + FreeBSD with a.out object format, i?86-*-freebsd*aout* and | |
| 4316 i?86-*-freebsd2* | |
| 4317 + Linux with a.out object format, i?86-linux*aout* | |
| 4318 + Linux with libc5, a.k.a. glibc1, i?86-linux*libc1* | |
| 4319 + Interix versions before Interix 3, i?86-*-interix | |
| 4320 + Mach microkernel, i?86-mach* | |
| 4321 + SCO UnixWare with UDK, i?86-*-udk* | |
| 4322 + Generic System V releases 1, 2, and 3, i?86-*-sysv[123]* | |
| 4323 + VSTa microkernel, i386-*-vsta | |
| 4324 * Motorola M68000 family | |
| 4325 + HPUX, m68k-hp-hpux* and m68000-hp-hpux* | |
| 4326 + NetBSD with a.out object format (before NetBSD 1.4), | |
| 4327 m68k-*-*-netbsd* except m68k-*-*-netbsdelf* | |
| 4328 + Generic System V r.4, m68k-*-sysv4* | |
| 4329 * VAX | |
| 4330 + Generic VAX, vax-*-* (This is generic VAX only; we have not | |
| 4331 obsoleted any VAX triples for specific operating systems.) | |
| 4332 | |
| 4333 Documentation improvements | |
| 4334 | |
| 4335 Other significant improvements | |
| 4336 | |
| 4337 * The build system has undergone several significant cleanups. | |
| 4338 Subdirectories will only be configured if they are being built, and | |
| 4339 all subdirectory configures are run from the make command. The top | |
| 4340 level has been autoconfiscated. | |
| 4341 * Building GCC no longer writes to its source directory. This should | |
| 4342 help those wishing to share a read-only source directory over NFS | |
| 4343 or build from a CD. The exceptions to this feature are if you | |
| 4344 configure with either --enable-maintainer-mode or | |
| 4345 --enable-generated-files-in-srcdir. | |
| 4346 * The -W warning option has been renamed to -Wextra, which is more | |
| 4347 easily understood. The older spelling will be retained for | |
| 4348 backwards compatibility. | |
| 4349 * Substantial improvements in compile time have been made, | |
| 4350 particularly for non-optimizing compilations. | |
| 4351 __________________________________________________________________ | |
| 4352 | |
| 4353 GCC 3.4.0 | |
| 4354 | |
| 4355 Bug Fixes | |
| 4356 | |
| 4357 A vast number of bugs have been fixed in 3.4.0, too many to publish a | |
| 4358 complete list here. [27]Follow this link to query the Bugzilla database | |
| 4359 for the list of over 900 bugs fixed in 3.4.0. This is the list of all | |
| 4360 bugs marked as resolved and fixed in 3.4.0 that are not flagged as 3.4 | |
| 4361 regressions. | |
| 4362 __________________________________________________________________ | |
| 4363 | |
| 4364 GCC 3.4.1 | |
| 4365 | |
| 4366 Bug Fixes | |
| 4367 | |
| 4368 This section lists the problem reports (PRs) from GCC's bug tracking | |
| 4369 system that are known to be fixed in the 3.4.1 release. This list might | |
| 4370 not be complete (that is, it is possible that some PRs that have been | |
| 4371 fixed are not listed here). | |
| 4372 | |
| 4373 Bootstrap failures | |
| 4374 | |
| 4375 * [28]10129 Ada bootstrap fails on PPC-Darwin - invalid assembler | |
| 4376 emitted - PIC related | |
| 4377 * [29]14576 [ARM] ICE in libiberty when building gcc-3.4 for arm-elf | |
| 4378 * [30]14760 A bug in configure.in prevents using both | |
| 4379 --program-suffix and --program-prefix | |
| 4380 * [31]14671 [hppa64] bootstrap fails: ICE in | |
| 4381 save_call_clobbered_regs, in caller_save.c | |
| 4382 * [32]15093 [alpha][Java] make bootstrap fails to configure libffi on | |
| 4383 Alpha | |
| 4384 * [33]15178 Solaris 9/x86 fails linking after stage 3 | |
| 4385 | |
| 4386 Multi-platform internal compiler errors (ICEs) | |
| 4387 | |
| 4388 * [34]12753 (preprocessor) Memory corruption in preprocessor on bad | |
| 4389 input | |
| 4390 * [35]13985 ICE in gcc.c-torture/compile/930621-1.c | |
| 4391 * [36]14810 (c++) tree check failures with invalid code involving | |
| 4392 templates | |
| 4393 * [37]14883 (c++) ICE on invalid code, in cp_parser_lookup_name, in | |
| 4394 cp/parser.c | |
| 4395 * [38]15044 (c++) ICE on syntax error, template header | |
| 4396 * [39]15057 (c++) Compiling of conditional value throw constructs | |
| 4397 cause a segmentation violation | |
| 4398 * [40]15064 (c++) typeid of template parameter gives ICE | |
| 4399 * [41]15142 (c++) ICE when passing a string where a char* is expected | |
| 4400 in a throw statement | |
| 4401 * [42]15159 ICE in rtl_verify_flow_info_1 | |
| 4402 * [43]15165 (c++) ICE in instantiate_template | |
| 4403 * [44]15193 Unary minus using pointer to V4SF vector causes | |
| 4404 -fforce-mem to exhaust all memory | |
| 4405 * [45]15209 (c++) Runs out of memory with packed structs | |
| 4406 * [46]15227 (c++) Trouble with invalid function definition | |
| 4407 * [47]15285 (c++) instantiate_type ICE when forming pointer to | |
| 4408 template function | |
| 4409 * [48]15299 (c++) ICE in resolve_overloaded_unification | |
| 4410 * [49]15329 (c++) ICE on constructor of member template | |
| 4411 * [50]15550 ICE in extract_insn, in recog.c | |
| 4412 * [51]15554 (c++) ICE in tsubst_copy, in cp/pt.c | |
| 4413 * [52]15640 (c++) ICE on invalid code in arg_assoc, in | |
| 4414 cp/name-lookup.c | |
| 4415 * [53]15666 [unit-at-a-time] Gcc abort on valid code | |
| 4416 * [54]15696 (c++) ICE with bad pointer-to-member code | |
| 4417 * [55]15701 (c++) ICE with friends and template template parameter | |
| 4418 * [56]15761 ICE in do_SUBST, in combine.c | |
| 4419 * [57]15829 (c++) ICE on Botan-1.3.13 due to -funroll-loops | |
| 4420 | |
| 4421 Ada | |
| 4422 | |
| 4423 * [58]14538 All RTEMS targets broken for gnat | |
| 4424 | |
| 4425 C front end | |
| 4426 | |
| 4427 * [59]12391 missing warning about assigning to an incomplete type | |
| 4428 * [60]14649 atan(1.0) should not be a constant expression | |
| 4429 * [61]15004 [unit-at-a-time] no warning for unused paramater in | |
| 4430 static function | |
| 4431 * [62]15749 --pedantic-errors behaves differently from --pedantic | |
| 4432 with C-compiler on GNU/Linux | |
| 4433 | |
| 4434 C++ compiler and library | |
| 4435 | |
| 4436 * [63]10646 non-const reference is incorrectly matched in a "const T" | |
| 4437 partial specialization | |
| 4438 * [64]12077 wcin.rdbuf()->in_avail() return value too high | |
| 4439 * [65]13598 enc_filebuf doesn't work | |
| 4440 * [66]14211 const_cast returns lvalue but should be rvalue | |
| 4441 * [67]14220 num_put::do_put() undesired float/double behavior | |
| 4442 * [68]14245 problem with user-defined allocators in std::basic_string | |
| 4443 * [69]14340 libstdc++ Debug mode: failure to convert iterator to | |
| 4444 const_iterator | |
| 4445 * [70]14600 __gnu_cxx::stdio_sync_filebuf should expose internal | |
| 4446 FILE* | |
| 4447 * [71]14668 no warning anymore for reevaluation of declaration | |
| 4448 * [72]14775 LFS (large file support) tests missing | |
| 4449 * [73]14821 Duplicate namespace alias declaration should not conflict | |
| 4450 * [74]14930 Friend declaration ignored | |
| 4451 * [75]14932 cannot use offsetof to get offsets of array elements in | |
| 4452 g++ 3.4.0 | |
| 4453 * [76]14950 [non unit-at-a-time] always_inline does not mix with | |
| 4454 templates and -O0 | |
| 4455 * [77]14962 g++ ignores #pragma redefine_extname | |
| 4456 * [78]14975 Segfault on low-level write error during imbue | |
| 4457 * [79]15002 Linewise stream input is unusably slow (std::string slow) | |
| 4458 * [80]15025 compiler accepts redeclaration of template as | |
| 4459 non-template | |
| 4460 * [81]15046 [arm] Math functions misdetected by cross configuration | |
| 4461 * [82]15069 a bit test on a variable of enum type is miscompiled | |
| 4462 * [83]15074 g++ -lsupc++ still links against libstdc++ | |
| 4463 * [84]15083 spurious "statement has no effect" warning | |
| 4464 * [85]15096 parse error with templates and pointer to const member | |
| 4465 * [86]15287 combination of operator[] and operator .* fails in | |
| 4466 templates | |
| 4467 * [87]15317 __attribute__ unused in first parameter of constructor | |
| 4468 gives error | |
| 4469 * [88]15337 sizeof on incomplete type diagnostic | |
| 4470 * [89]15361 bitset<>::_Find_next fails | |
| 4471 * [90]15412 _GLIBCXX_ symbols symbols defined and used in different | |
| 4472 namespaces | |
| 4473 * [91]15427 valid code results in incomplete type error | |
| 4474 * [92]15471 Incorrect member pointer offsets in anonymous | |
| 4475 structs/unions | |
| 4476 * [93]15503 nested template problem | |
| 4477 * [94]15507 compiler hangs while laying out union | |
| 4478 * [95]15542 operator & and template definitions | |
| 4479 * [96]15565 SLES9: leading + sign for unsigned int with showpos | |
| 4480 * [97]15625 friend defined inside a template fails to find static | |
| 4481 function | |
| 4482 * [98]15629 Function templates, overloads, and friend name injection | |
| 4483 * [99]15742 'noreturn' attribute ignored in method of template | |
| 4484 functions. | |
| 4485 * [100]15775 Allocator::pointer consistently ignored | |
| 4486 * [101]15821 Duplicate namespace alias within namespace rejected | |
| 4487 * [102]15862 'enum yn' fails (confict with undeclared builtin) | |
| 4488 * [103]15875 rejects pointer to member in template | |
| 4489 * [104]15877 valid code using templates and anonymous enums is | |
| 4490 rejected | |
| 4491 * [105]15947 Puzzling error message for wrong destructor declaration | |
| 4492 in template class | |
| 4493 * [106]16020 cannot copy __gnu_debug::bitset | |
| 4494 * [107]16154 input iterator concept too restrictive | |
| 4495 * [108]16174 deducing top-level consts | |
| 4496 | |
| 4497 Java | |
| 4498 | |
| 4499 * [109]14315 Java compiler is not parallel make safe | |
| 4500 | |
| 4501 Fortran | |
| 4502 | |
| 4503 * [110]15151 [g77] incorrect logical i/o in 64-bit mode | |
| 4504 | |
| 4505 Objective-C | |
| 4506 | |
| 4507 * [111]7993 private variables cannot be shadowed in subclasses | |
| 4508 | |
| 4509 Optimization bugs | |
| 4510 | |
| 4511 * [112]15228 useless copies of floating point operands | |
| 4512 * [113]15345 [non-unit-at-a-time] unreferenced nested inline | |
| 4513 functions not optimized away | |
| 4514 * [114]15945 Incorrect floating point optimization | |
| 4515 * [115]15526 ftrapv aborts on 0 * (-1) | |
| 4516 * [116]14690 Miscompiled POOMA tests | |
| 4517 * [117]15112 GCC generates code to write to unchanging memory | |
| 4518 | |
| 4519 Preprocessor | |
| 4520 | |
| 4521 * [118]15067 Minor glitch in the source of cpp | |
| 4522 | |
| 4523 Main driver program bugs | |
| 4524 | |
| 4525 * [119]1963 collect2 interprets -oldstyle_liblookup as -o | |
| 4526 ldstyle_liblookup | |
| 4527 | |
| 4528 x86-specific (Intel/AMD) | |
| 4529 | |
| 4530 * [120]15717 Error: can't resolve `L0' {*ABS* section} - `xx' {*UND* | |
| 4531 section} | |
| 4532 | |
| 4533 HPPA-specific | |
| 4534 | |
| 4535 * [121]14782 GCC produces an unaligned data access at -O2 | |
| 4536 * [122]14828 FAIL: gcc.c-torture/execute/20030408-1.c execution, -O2 | |
| 4537 * [123]15202 ICE in reload_cse_simplify_operands, in postreload.c | |
| 4538 | |
| 4539 IA64-specific | |
| 4540 | |
| 4541 * [124]14610 __float80 constants incorrectly emitted | |
| 4542 * [125]14813 init_array sections are initialized in the wrong order | |
| 4543 * [126]14857 GCC segfault on duplicated asm statement | |
| 4544 * [127]15598 Gcc 3.4 ICE on valid code | |
| 4545 * [128]15653 Gcc 3.4 ICE on valid code | |
| 4546 | |
| 4547 MIPS-specific | |
| 4548 | |
| 4549 * [129]15189 wrong filling of delay slot with -march=mips1 -G0 | |
| 4550 -mno-split-addresses -mno-explicit-relocs | |
| 4551 * [130]15331 Assembler error building gnatlib on IRIX 6.5 with GNU as | |
| 4552 2.14.91 | |
| 4553 * [131]16144 Bogus reference to __divdf3 when -O1 | |
| 4554 * [132]16176 Miscompilation of unaligned data in MIPS backend | |
| 4555 | |
| 4556 PowerPC-specific | |
| 4557 | |
| 4558 * [133]11591 ICE in gcc.dg/altivec-5.c | |
| 4559 * [134]12028 powerpc-eabispe produces bad sCOND operation | |
| 4560 * [135]14478 rs6000 geu/ltu patterns generate incorrect code | |
| 4561 * [136]14567 long double and va_arg complex args | |
| 4562 * [137]14715 Altivec stack layout may overlap gpr save with stack | |
| 4563 temps | |
| 4564 * [138]14902 (libstdc++) Stream checking functions fail when -pthread | |
| 4565 option is used. | |
| 4566 * [139]14924 Compiler ICE on valid code | |
| 4567 * [140]14960 -maltivec affects vector return with -mabi=no-altivec | |
| 4568 * [141]15106 vector varargs failure passing from altivec to | |
| 4569 non-altivec code for -m32 | |
| 4570 * [142]16026 ICE in function.c:4804, assign_parms, when -mpowerpc64 & | |
| 4571 half-word operation | |
| 4572 * [143]15191 -maltivec -mabi=no-altivec results in mis-aligned lvx | |
| 4573 and stvx | |
| 4574 * [144]15662 Segmentation fault when an exception is thrown - even if | |
| 4575 try and catch are specified | |
| 4576 | |
| 4577 s390-specific | |
| 4578 | |
| 4579 * [145]15054 Bad code due to overlapping stack temporaries | |
| 4580 | |
| 4581 SPARC-specific | |
| 4582 | |
| 4583 * [146]15783 ICE with union assignment in 64-bit mode | |
| 4584 * [147]15626 GCC 3.4 emits "ld: warning: relocation error: | |
| 4585 R_SPARC_UA32" | |
| 4586 | |
| 4587 x86-64-specific | |
| 4588 | |
| 4589 * [148]14326 boehm-gc hardcodes to 3DNow! prefetch for x86_64 | |
| 4590 * [149]14723 Backported -march=nocona from mainline | |
| 4591 * [150]15290 __float128 failed to pass to function properly | |
| 4592 | |
| 4593 Cygwin/Mingw32-specific | |
| 4594 | |
| 4595 * [151]15250 Option -mms-bitfields support on GCC 3.4 is not | |
| 4596 conformant to MS layout | |
| 4597 * [152]15551 -mtune=pentium4 -O2 with sjlj EH breaks stack probe | |
| 4598 worker on windows32 targets | |
| 4599 | |
| 4600 Bugs specific to embedded processors | |
| 4601 | |
| 4602 * [153]8309 [m68k] -m5200 produces erroneous SImode set of short | |
| 4603 varaible on stack | |
| 4604 * [154]13250 [SH] Gcc code for rotation clobbers the register, but | |
| 4605 gcc continues to use the register as if it was not clobbered | |
| 4606 * [155]13803 [coldfire] movqi operand constraints too restrictivefor | |
| 4607 TARGET_COLDFIRE | |
| 4608 * [156]14093 [SH] ICE for code when using -mhitachi option in SH | |
| 4609 * [157]14457 [m6811hc] ICE with simple c++ source | |
| 4610 * [158]14542 [m6811hc] ICE on simple source | |
| 4611 * [159]15100 [SH] cc1plus got hang-up on | |
| 4612 libstdc++-v3/testsuite/abi_check.cc | |
| 4613 * [160]15296 [CRIS] Delayed branch scheduling causing invalid code on | |
| 4614 cris-* | |
| 4615 * [161]15396 [SH] ICE with -O2 -fPIC | |
| 4616 * [162]15782 [coldfire] m68k_output_mi_thunk emits wrong code for | |
| 4617 ColdFire | |
| 4618 | |
| 4619 Testsuite problems (compiler not affected) | |
| 4620 | |
| 4621 * [163]11610 libstdc++ testcases 27_io/* don't work properly remotely | |
| 4622 * [164]15488 (libstdc++) possibly insufficient file permissions for | |
| 4623 executing test suite | |
| 4624 * [165]15489 (libstdc++) testsuite_files determined incorrectly | |
| 4625 | |
| 4626 Documentation bugs | |
| 4627 | |
| 4628 * [166]13928 (libstdc++) no whatis info in some man pages generated | |
| 4629 by doxygen | |
| 4630 * [167]14150 Ada documentation out of date | |
| 4631 * [168]14949 (c++) Need to document method visibility changes | |
| 4632 * [169]15123 libstdc++-doc: Allocators.3 manpage is empty | |
| 4633 __________________________________________________________________ | |
| 4634 | |
| 4635 GCC 3.4.2 | |
| 4636 | |
| 4637 Bug Fixes | |
| 4638 | |
| 4639 This section lists the problem reports (PRs) from GCC's bug tracking | |
| 4640 system that are known to be fixed in the 3.4.2 release. This list might | |
| 4641 not be complete (that is, it is possible that some PRs that have been | |
| 4642 fixed are not listed here). | |
| 4643 | |
| 4644 Bootstrap failures and issues | |
| 4645 | |
| 4646 * [170]16469 [mips-sgi-irix5.3] bootstrap fails in | |
| 4647 libstdc++-v3/testsuite | |
| 4648 * [171]16344 [hppa-linux-gnu] libstdc++'s PCH built by | |
| 4649 profiledbootstrap does not work with the built compiler | |
| 4650 * [172]16842 [Solaris/x86] mkheaders can not find mkheaders.conf | |
| 4651 | |
| 4652 Multi-platform internal compiler errors (ICEs) | |
| 4653 | |
| 4654 * [173]12608 (c++) ICE: expected class 't', have 'x' (error_mark) in | |
| 4655 cp_parser_class_specifier, in cp/parser.c | |
| 4656 * [174]14492 ICE in loc_descriptor_from_tree, in dwarf2out.c | |
| 4657 * [175]15461 (c++) ICE due to NRV and inlining | |
| 4658 * [176]15890 (c++) ICE in c_expand_expr, in c-common.c | |
| 4659 * [177]16180 ICE: segmentation fault in RTL optimization | |
| 4660 * [178]16224 (c++) ICE in write_unscoped_name (template/namespace) | |
| 4661 * [179]16408 ICE: in delete_insn, in cfgrtl.c | |
| 4662 * [180]16529 (c++) ICE for: namespace-alias shall not be declared as | |
| 4663 the name of any other entity | |
| 4664 * [181]16698 (c++) ICE with exceptions and declaration of __cxa_throw | |
| 4665 * [182]16706 (c++) ICE in finish_member_declaration, in | |
| 4666 cp/semantics.c | |
| 4667 * [183]16810 (c++) Legal C++ program with cast gives ICE in | |
| 4668 build_ptrmemfunc | |
| 4669 * [184]16851 (c++) ICE when throwing a comma expression | |
| 4670 * [185]16870 (c++) Boost.Spirit causes ICE in tsubst, in cp/pt.c | |
| 4671 * [186]16904 (c++) ICE in finish_class_member_access_expr, in | |
| 4672 cp/typeck.c | |
| 4673 * [187]16905 (c++) ICE (segfault) with exceptions | |
| 4674 * [188]16964 (c++) ICE in cp_parser_class_specifier due to | |
| 4675 redefinition | |
| 4676 * [189]17068 (c++) ICE: tree check: expected class 'd', have 'x' | |
| 4677 (identifier_node) in dependent_template_p, in cp/pt.c | |
| 4678 | |
| 4679 Preprocessor bugs | |
| 4680 | |
| 4681 * [190]16366 Preprocessor option -remap causes memory corruption | |
| 4682 | |
| 4683 Optimization | |
| 4684 | |
| 4685 * [191]15345 unreferenced nested inline functions not optimized away | |
| 4686 * [192]16590 Incorrect execution when compiling with -O2 | |
| 4687 * [193]16693 Bitwise AND is lost when used within a cast to an enum | |
| 4688 of the same precision | |
| 4689 * [194]17078 Jump into if(0) substatement fails | |
| 4690 | |
| 4691 Problems in generated debug information | |
| 4692 | |
| 4693 * [195]13956 incorrect stabs for nested local variables | |
| 4694 | |
| 4695 C front end bugs | |
| 4696 | |
| 4697 * [196]16684 GCC should not warn about redundant redeclarations of | |
| 4698 built-ins | |
| 4699 | |
| 4700 C++ compiler and library | |
| 4701 | |
| 4702 * [197]12658 Thread safety problems in locale::global() and | |
| 4703 locale::locale() | |
| 4704 * [198]13092 g++ accepts invalid pointer-to-member conversion | |
| 4705 * [199]15320 Excessive memory consumption | |
| 4706 * [200]16246 Incorrect template argument deduction | |
| 4707 * [201]16273 Memory exhausted when using nested classes and virtual | |
| 4708 functions | |
| 4709 * [202]16401 ostringstream in gcc 3.4.x very slow for big data | |
| 4710 * [203]16411 undefined reference to | |
| 4711 __gnu_cxx::stdio_sync_filebuf<char, std::char_traits<char> | |
| 4712 >::file() | |
| 4713 * [204]16489 G++ incorrectly rejects use of a null constant integral | |
| 4714 expression as a null constant pointer | |
| 4715 * [205]16618 offsetof fails with constant member | |
| 4716 * [206]16637 syntax error reported for valid input code | |
| 4717 * [207]16717 __attribute__((constructor)) broken in C++ | |
| 4718 * [208]16813 compiler error in DEBUG version of range insertion | |
| 4719 std::map::insert | |
| 4720 * [209]16853 pointer-to-member initialization from incompatible one | |
| 4721 accepted | |
| 4722 * [210]16889 ambiguity is not detected | |
| 4723 * [211]16959 Segmentation fault in ios_base::sync_with_stdio | |
| 4724 | |
| 4725 Java compiler and library | |
| 4726 | |
| 4727 * [212]7587 direct threaded interpreter not thread-safe | |
| 4728 * [213]16473 ServerSocket accept() leaks file descriptors | |
| 4729 * [214]16478 Hash synchronization deadlock with finalizers | |
| 4730 | |
| 4731 Alpha-specific | |
| 4732 | |
| 4733 * [215]10695 ICE in dwarf2out_frame_debug_expr, in dwarf2out.c | |
| 4734 * [216]16974 could not split insn (ice in final_scan_insn, in | |
| 4735 final.c) | |
| 4736 | |
| 4737 x86-specific | |
| 4738 | |
| 4739 * [217]16298 ICE in output_operand | |
| 4740 * [218]17113 ICE with SSE2 intrinsics | |
| 4741 | |
| 4742 x86-64 specific | |
| 4743 | |
| 4744 * [219]14697 libstdc++ couldn't find 32bit libgcc_s | |
| 4745 | |
| 4746 MIPS-specific | |
| 4747 | |
| 4748 * [220]15869 [mips64] No NOP after LW (with -mips1 -O0) | |
| 4749 * [221]16325 [mips64] value profiling clobbers gp on mips | |
| 4750 * [222]16357 [mipsisa64-elf] ICE copying 7 bytes between extern | |
| 4751 char[]s | |
| 4752 * [223]16380 [mips64] Use of uninitialised register after dbra | |
| 4753 conversion | |
| 4754 * [224]16407 [mips64] Unaligned access to local variables | |
| 4755 * [225]16643 [mips64] verify_local_live_at_start ICE after | |
| 4756 crossjumping & cfgcleanup | |
| 4757 | |
| 4758 ARM-specific | |
| 4759 | |
| 4760 * [226]15927 THUMB -O2: strength-reduced iteration variable ends up | |
| 4761 off by 1 | |
| 4762 * [227]15948 THUMB: ICE with non-commutative cbranch | |
| 4763 * [228]17019 THUMB: bad switch statement in md code for | |
| 4764 addsi3_cbranch_scratch | |
| 4765 | |
| 4766 IA64-specific | |
| 4767 | |
| 4768 * [229]16130 ICE on valid code: in bundling, in config/ia64/ia64.c | |
| 4769 (-mtune=merced) | |
| 4770 * [230]16142 ICE on valid code: in bundling, in config/ia64/ia64.c | |
| 4771 (-mtune=itanium) | |
| 4772 * [231]16278 Gcc failed to build Linux kernel with -mtune=merced | |
| 4773 * [232]16414 ICE on valid code: typo in comparison of asm_noperands | |
| 4774 result | |
| 4775 * [233]16445 ICE on valid code: don't count ignored insns | |
| 4776 * [234]16490 ICE (segfault) while compiling with -fprofile-use | |
| 4777 * [235]16683 ia64 does not honor SUBTARGET_EXTRA_SPECS | |
| 4778 | |
| 4779 PowerPC-specific | |
| 4780 | |
| 4781 * [236]16195 (ppc64): Miscompilation of GCC 3.3.x by 3.4.x | |
| 4782 * [237]16239 ICE on ppc64 (mozilla 1.7 compile, -O1 -fno-exceptions | |
| 4783 issue) | |
| 4784 | |
| 4785 SPARC-specific | |
| 4786 | |
| 4787 * [238]16199 ICE while compiling apache 2.0.49 | |
| 4788 * [239]16416 -m64 doesn't imply -mcpu=v9 anymore | |
| 4789 * [240]16430 ICE when returning non-C aggregates larger than 16 bytes | |
| 4790 | |
| 4791 Bugs specific to embedded processors | |
| 4792 | |
| 4793 * [241]16379 [m32r] can't output large model function call of memcpy | |
| 4794 * [242]17093 [m32r] ICE with -msdata=use -O0 | |
| 4795 * [243]17119 [m32r] ICE at switch case 0x8000 | |
| 4796 | |
| 4797 DJGPP-specific | |
| 4798 | |
| 4799 * [244]15928 libstdc++ in 3.4.x doesn't cross-compile for djgpp | |
| 4800 | |
| 4801 Alpha Tru64-specific | |
| 4802 | |
| 4803 * [245]16210 libstdc++ gratuitously omits "long long" I/O | |
| 4804 | |
| 4805 Testsuite, documentation issues (compiler is not affected): | |
| 4806 | |
| 4807 * [246]15488 (libstdc++) possibly insufficient file permissions for | |
| 4808 executing test suite | |
| 4809 * [247]16250 ada/doctools runs makeinfo even in release tarball | |
| 4810 __________________________________________________________________ | |
| 4811 | |
| 4812 GCC 3.4.3 | |
| 4813 | |
| 4814 This is the [248]list of problem reports (PRs) from GCC's bug tracking | |
| 4815 system that are known to be fixed in the 3.4.3 release. This list might | |
| 4816 not be complete (that is, it is possible that some PRs that have been | |
| 4817 fixed are not listed here). | |
| 4818 | |
| 4819 Bootstrap failures | |
| 4820 | |
| 4821 * [249]17369 [ia64] Bootstrap failure with binutils-2.15.90.0.1.1 | |
| 4822 * [250]17850 [arm-elf] bootstrap failure - libstdc++ uses strtold | |
| 4823 when undeclared | |
| 4824 | |
| 4825 Internal compiler errors (ICEs) affecting multiple platforms | |
| 4826 | |
| 4827 * [251]13948 (java) GCJ segmentation fault while compiling GL4Java | |
| 4828 .class files | |
| 4829 * [252]14492 ICE in loc_descriptor_from_tree, in dwarf2out.c | |
| 4830 * [253]16301 (c++) ICE when "strong" attribute is attached to a using | |
| 4831 directive | |
| 4832 * [254]16566 ICE with flexible arrays | |
| 4833 * [255]17023 ICE with nested functions in parameter declaration | |
| 4834 * [256]17027 ICE with noreturn function in loop at -O2 | |
| 4835 * [257]17524 ICE in grokdeclarator, in cp/decl.c | |
| 4836 * [258]17826 (c++) ICE in cp_tree_equal | |
| 4837 | |
| 4838 C and optimization bugs | |
| 4839 | |
| 4840 * [259]15526 -ftrapv aborts on 0 * (-1) | |
| 4841 * [260]16999 #ident stopped working | |
| 4842 * [261]17503 quadratic behaviour in invalid_mode_change_p | |
| 4843 * [262]17581 Long long arithmetic fails inside a switch/case | |
| 4844 statement when compiled with -O2 | |
| 4845 * [263]18129 -fwritable-strings doesn't work | |
| 4846 | |
| 4847 C++ compiler and library bugs | |
| 4848 | |
| 4849 * [264]10975 incorrect initial ostringstream::tellp() | |
| 4850 * [265]11722 Unbuffered filebuf::sgetn is slow | |
| 4851 * [266]14534 Unrecognizing static function as a template parameter | |
| 4852 when its return value is also templated | |
| 4853 * [267]15172 Copy constructor optimization in aggregate | |
| 4854 initialization | |
| 4855 * [268]15786 Bad error message for frequently occuring error. | |
| 4856 * [269]16162 Rejects valid member-template-definition | |
| 4857 * [270]16612 empty basic_strings can't live in shared memory | |
| 4858 * [271]16715 std::basic_iostream is instantiated when used, even | |
| 4859 though instantiations are already contained in libstdc++ | |
| 4860 * [272]16848 code in /ext/demangle.h appears broken | |
| 4861 * [273]17132 GCC fails to eliminate function template specialization | |
| 4862 when argument deduction fails | |
| 4863 * [274]17259 One more _S_leaf incorrectly qualified with _RopeRep:: | |
| 4864 in ropeimpl.h | |
| 4865 * [275]17327 use of `enumeral_type' in template type unification | |
| 4866 * [276]17393 "unused variable '._0'" warning with -Wall | |
| 4867 * [277]17501 Confusion with member templates | |
| 4868 * [278]17537 g++ not passing -lstdc++ to linker when all command line | |
| 4869 arguments are libraries | |
| 4870 * [279]17585 usage of unqualified name of static member from within | |
| 4871 class not allowed | |
| 4872 * [280]17821 Poor diagnostic for using "." instead of "->" | |
| 4873 * [281]17829 wrong error: call of overloaded function is ambiguous | |
| 4874 * [282]17851 Misleading diagnostic for invalid function declarations | |
| 4875 with undeclared types | |
| 4876 * [283]17976 Destructor is called twice | |
| 4877 * [284]18020 rejects valid definition of enum value in template | |
| 4878 * [285]18093 bogus conflict in namespace aliasing | |
| 4879 * [286]18140 C++ parser bug when using >> in templates | |
| 4880 | |
| 4881 Fortran | |
| 4882 | |
| 4883 * [287]17541 data statements with double precision constants fail | |
| 4884 | |
| 4885 x86-specific | |
| 4886 | |
| 4887 * [288]17853 -O2 ICE for MMX testcase | |
| 4888 | |
| 4889 SPARC-specific | |
| 4890 | |
| 4891 * [289]17245 ICE compiling gsl-1.5 statistics/lag1.c | |
| 4892 | |
| 4893 Darwin-specific | |
| 4894 | |
| 4895 * [290]17167 FATAL:Symbol L_foo$stub already defined. | |
| 4896 | |
| 4897 AIX-specific | |
| 4898 | |
| 4899 * [291]17277 could not catch an exception when specified -maix64 | |
| 4900 | |
| 4901 Solaris-specific | |
| 4902 | |
| 4903 * [292]17505 <cmath> calls acosf(), ceilf(), and other functions | |
| 4904 missing from system libraries | |
| 4905 | |
| 4906 HP/UX specific: | |
| 4907 | |
| 4908 * [293]17684 /usr/ccs/bin/ld: Can't create libgcc_s.sl | |
| 4909 | |
| 4910 ARM-specific | |
| 4911 | |
| 4912 * [294]17384 ICE with mode attribute on structures | |
| 4913 | |
| 4914 MIPS-specific | |
| 4915 | |
| 4916 * [295]17770 No NOP after LWL with -mips1 | |
| 4917 | |
| 4918 Other embedded target specific | |
| 4919 | |
| 4920 * [296]11476 [arc-elf] gcc ICE on newlib's vfprintf.c | |
| 4921 * [297]14064 [avr-elf] -fdata-sections triggers ICE | |
| 4922 * [298]14678 [m68hc11-elf] gcc ICE | |
| 4923 * [299]15583 [powerpc-rtems] powerpc-rtems lacks __USE_INIT_FINI__ | |
| 4924 * [300]15790 [i686-coff] Alignment error building gcc with i686-coff | |
| 4925 target | |
| 4926 * [301]15886 [SH] Miscompilation with -O2 -fPIC | |
| 4927 * [302]16884 [avr-elf] [fweb related] bug while initializing | |
| 4928 variables | |
| 4929 | |
| 4930 Bugs relating to debugger support | |
| 4931 | |
| 4932 * [303]13841 missing debug info for _Complex function arguments | |
| 4933 * [304]15860 [big-endian targets] No DW_AT_location debug info is | |
| 4934 emitted for formal arguments to a function that uses "register" | |
| 4935 qualifiers | |
| 4936 | |
| 4937 Testsuite issues (compiler not affected) | |
| 4938 | |
| 4939 * [305]17465 Testsuite in libffi overrides LD_LIBRARY_PATH | |
| 4940 * [306]17469 Testsuite in libstdc++ overrides LD_LIBRARY_PATH | |
| 4941 * [307]18138 [mips-sgi-irix6.5] libgcc_s.so.1 not found by 64-bit | |
| 4942 testsuite | |
| 4943 | |
| 4944 Documentation | |
| 4945 | |
| 4946 * [308]15498 typo in gcc manual: non-existing locale example en_UK, | |
| 4947 should be en_GB | |
| 4948 * [309]15747 [mips-sgi-irix5.3] /bin/sh hangs during bootstrap: | |
| 4949 document broken shell | |
| 4950 * [310]16406 USE_LD_AS_NEEDED undocumented | |
| 4951 __________________________________________________________________ | |
| 4952 | |
| 4953 GCC 3.4.4 | |
| 4954 | |
| 4955 This is the [311]list of problem reports (PRs) from GCC's bug tracking | |
| 4956 system that are known to be fixed in the 3.4.4 release. This list might | |
| 4957 not be complete (that is, it is possible that some PRs that have been | |
| 4958 fixed are not listed here). | |
| 4959 __________________________________________________________________ | |
| 4960 | |
| 4961 GCC 3.4.5 | |
| 4962 | |
| 4963 This is the [312]list of problem reports (PRs) from GCC's bug tracking | |
| 4964 system that are known to be fixed in the 3.4.5 release. This list might | |
| 4965 not be complete (that is, it is possible that some PRs that have been | |
| 4966 fixed are not listed here). | |
| 4967 | |
| 4968 Bootstrap issues | |
| 4969 | |
| 4970 * [313]24688 sco_math fixincl breaks math.h | |
| 4971 | |
| 4972 C compiler bugs | |
| 4973 | |
| 4974 * [314]17188 struct Foo { } redefinition | |
| 4975 * [315]20187 wrong code for ((unsigned char)(unsigned long | |
| 4976 long)((a?a:1)&(a*b)))?0:1) | |
| 4977 * [316]21873 infinite warning loop on bad array initializer | |
| 4978 * [317]21899 enum definition accepts values to be overriden | |
| 4979 * [318]22061 ICE in find_function_data, in function.c | |
| 4980 * [319]22308 Failure to diagnose violation of constraint 6.516p2 | |
| 4981 * [320]22458 ICE on missing brace | |
| 4982 * [321]22589 ICE casting to long long | |
| 4983 * [322]24101 Segfault with preprocessed source | |
| 4984 | |
| 4985 C++ compiler and library bugs | |
| 4986 | |
| 4987 * [323]10611 operations on vector mode not recognized in C++ | |
| 4988 * [324]13377 unexpected behavior of namespace usage directive | |
| 4989 * [325]16002 Strange error message with new parser | |
| 4990 * [326]17413 local classes as template argument | |
| 4991 * [327]17609 spurious error message after using keyword | |
| 4992 * [328]17618 ICE in cp_convert_to_pointer, in cp/cvt.c | |
| 4993 * [329]18124 ICE with invalid template template parameter | |
| 4994 * [330]18155 typedef in template declaration not rejected | |
| 4995 * [331]18177 ICE with const_cast for undeclared variable | |
| 4996 * [332]18368 C++ error message regression | |
| 4997 * [333]16378 ICE when returning a copy of a packed member | |
| 4998 * [334]18466 int ::i; accepted | |
| 4999 * [335]18512 ICE on invalid usage of template base class | |
| 5000 * [336]18454 ICE when returning undefined type | |
| 5001 * [337]18738 typename not allowed with non-dependent qualified name | |
| 5002 * [338]18803 rejects access to operator() in template | |
| 5003 * [339]19004 ICE in uses_template_parms, in cp/pt.c | |
| 5004 * [340]19208 Spurious error about variably modified type | |
| 5005 * [341]18253 bad error message / ICE for invalid template parameter | |
| 5006 * [342]19608 ICE after friend function definition in local class | |
| 5007 * [343]19884 ICE on explicit instantiation of a non-template | |
| 5008 constructor | |
| 5009 * [344]20153 ICE when C++ template function contains anonymous union | |
| 5010 * [345]20563 Infinite loop in diagnostic (and ice after error | |
| 5011 message) | |
| 5012 * [346]20789 ICE with incomplete type in template | |
| 5013 * [347]21336 Internal compiler error when using custom new operators | |
| 5014 * [348]21768 ICE in error message due to violation of coding | |
| 5015 conventions | |
| 5016 * [349]21853 constness of pointer to data member ignored | |
| 5017 * [350]21903 Default argument of template function causes a | |
| 5018 compile-time error | |
| 5019 * [351]21983 multiple diagnostics | |
| 5020 * [352]21987 New testsuite failure | |
| 5021 g++.dg/warn/conversion-function-1.C | |
| 5022 * [353]22153 ICE on invalid template specialization | |
| 5023 * [354]22172 Internal compiler error, seg fault. | |
| 5024 * [355]21286 filebuf::xsgetn vs pipes | |
| 5025 * [356]22233 ICE with wrong number of template parameters | |
| 5026 * [357]22508 ICE after invalid operator new | |
| 5027 * [358]22545 ICE with pointer to class member & user defined | |
| 5028 conversion operator | |
| 5029 * [359]23528 Wrong default allocator in ext/hash_map | |
| 5030 * [360]23550 char_traits requirements/1.cc test bad math | |
| 5031 * [361]23586 Bad diagnostic for invalid namespace-name | |
| 5032 * [362]23624 ICE in invert_truthvalue, in fold-const.c | |
| 5033 * [363]23639 Bad error message: not a member of '<declaration error>' | |
| 5034 * [364]23797 ICE on typename outside template | |
| 5035 * [365]23965 Bogus error message: no matching function for call to | |
| 5036 'foo(<type error>)' | |
| 5037 * [366]24052 &#`label_decl' not supported by dump_expr#<expression | |
| 5038 error> | |
| 5039 * [367]24580 virtual base class cause exception not to be caught | |
| 5040 | |
| 5041 Problems in generated debug information | |
| 5042 | |
| 5043 * [368]24267 Bad DWARF for altivec vectors | |
| 5044 | |
| 5045 Optimizations issues | |
| 5046 | |
| 5047 * [369]17810 ICE in verify_local_live_at_start | |
| 5048 * [370]17860 Wrong generated code for loop with varying bound | |
| 5049 * [371]21709 ICE on compile-time complex NaN | |
| 5050 * [372]21964 broken tail call at -O2 or more | |
| 5051 * [373]22167 Strange optimization bug when using -Os | |
| 5052 * [374]22619 Compilation failure for real_const_1.f and | |
| 5053 real_const_2.f90 | |
| 5054 * [375]23241 Invalid code generated for comparison of uchar to 255 | |
| 5055 * [376]23478 Miscompilation due to reloading of a var that is also | |
| 5056 used in EH pad | |
| 5057 * [377]24470 segmentation fault in cc1plus when compiling with -O | |
| 5058 * [378]24950 ICE in operand_subword_force | |
| 5059 | |
| 5060 Precompiled headers problems | |
| 5061 | |
| 5062 * [379]14400 Cannot compile qt-x11-free-3.3.0 | |
| 5063 * [380]14940 PCH largefile test fails on various platforms | |
| 5064 | |
| 5065 Preprocessor bugs | |
| 5066 | |
| 5067 * [381]20239 ICE on empty preprocessed input | |
| 5068 * [382]15220 "gcc -E -MM -MG" reports missing system headers in | |
| 5069 source directory | |
| 5070 | |
| 5071 Testsuite issues | |
| 5072 | |
| 5073 * [383]19275 gcc.dg/20020919-1.c fails with -fpic/-fPIC on | |
| 5074 i686-pc-linux-gnu | |
| 5075 | |
| 5076 Alpha specific | |
| 5077 | |
| 5078 * [384]21888 bootstrap failure with linker relaxation enabled | |
| 5079 | |
| 5080 ARM specific | |
| 5081 | |
| 5082 * [385]15342 [arm-linux]: ICE in verify_local_live_at_start | |
| 5083 * [386]23985 Memory aliasing information incorrect in inlined memcpy | |
| 5084 | |
| 5085 ColdFile specific | |
| 5086 | |
| 5087 * [387]16719 Illegal move of byte into address register causes | |
| 5088 compiler to ICE | |
| 5089 | |
| 5090 HPPA specific | |
| 5091 | |
| 5092 * [388]21723 ICE while building libgfortran | |
| 5093 * [389]21841 -mhp-ld/-mgnu-ld documentation | |
| 5094 | |
| 5095 IA-64 specific | |
| 5096 | |
| 5097 * [390]23644 IA-64 hardware models and configuration options | |
| 5098 documentation error | |
| 5099 * [391]24718 Shared libgcc not used for linking by default | |
| 5100 | |
| 5101 M68000 specific | |
| 5102 | |
| 5103 * [392]18421 ICE in reload_cse_simplify_operands, in postreload.c | |
| 5104 | |
| 5105 MIPS specific | |
| 5106 | |
| 5107 * [393]20621 ICE in change_address_1, in emit-rtl.c | |
| 5108 | |
| 5109 PowerPC and PowerPC64 specific | |
| 5110 | |
| 5111 * [394]18583 error on valid code: const | |
| 5112 __attribute__((altivec(vector__))) doesn't work in arrays | |
| 5113 * [395]20191 ICE in reload_cse_simplify_operands | |
| 5114 * [396]22083 AIX: TARGET_C99_FUNCTIONS is wrongly defined | |
| 5115 * [397]23070 CALL_V4_CLEAR_FP_ARGS flag not properly set | |
| 5116 * [398]23404 gij trashes args of functions with more than 8 fp args | |
| 5117 * [399]23539 C & C++ compiler generating misaligned references | |
| 5118 regardless of compiler flags | |
| 5119 * [400]24102 floatdisf2_internal2 broken | |
| 5120 * [401]24465 -mminimal-toc miscompilation of __thread vars | |
| 5121 | |
| 5122 Solaris specific | |
| 5123 | |
| 5124 * [402]19933 Problem with define of HUGE_VAL in math_c99 | |
| 5125 * [403]21889 Native Solaris assembler cannot grok DTP-relative debug | |
| 5126 symbols | |
| 5127 | |
| 5128 SPARC specific | |
| 5129 | |
| 5130 * [404]19300 PCH failures on sparc-linux | |
| 5131 * [405]20301 Assembler labels have a leading "-" | |
| 5132 * [406]20673 C PCH testsuite assembly comparison failure | |
| 5133 | |
| 5134 x86 and x86_64 specific | |
| 5135 | |
| 5136 * [407]18582 ICE with arrays of type V2DF | |
| 5137 * [408]19340 Compilation SEGFAULTs with -O1 -fschedule-insns2 | |
| 5138 -fsched2-use-traces | |
| 5139 * [409]21716 ICE in reg-stack.c's swap_rtx_condition | |
| 5140 * [410]24315 amd64 fails -fpeephole2 | |
| 5141 __________________________________________________________________ | |
| 5142 | |
| 5143 GCC 3.4.6 | |
| 5144 | |
| 5145 This is the [411]list of problem reports (PRs) from GCC's bug tracking | |
| 5146 system that are known to be fixed in the 3.4.6 release. This list might | |
| 5147 not be complete (that is, it is possible that some PRs that have been | |
| 5148 fixed are not listed here). | |
| 5149 | |
| 5150 Please send FSF & GNU inquiries & questions to [412]gnu@gnu.org. There | |
| 5151 are also [413]other ways to contact the FSF. | |
| 5152 | |
| 5153 These pages are maintained by [414]the GCC team. | |
| 5154 | |
| 5155 | |
| 5156 For questions related to the use of GCC, please consult these web | |
| 5157 pages and the [415]GCC manuals. If that fails, the | |
| 5158 [416]gcc-help@gcc.gnu.org mailing list might help. | |
| 5159 Please send comments on these web pages and the development of GCC to | |
| 5160 our developer mailing list at [417]gcc@gnu.org or | |
| 5161 [418]gcc@gcc.gnu.org. All of our lists have [419]public archives. | |
| 5162 | |
| 5163 Copyright (C) Free Software Foundation, Inc., 51 Franklin St, Fifth | |
| 5164 Floor, Boston, MA 02110, USA. | |
| 5165 | |
| 5166 Verbatim copying and distribution of this entire article is permitted | |
| 5167 in any medium, provided this notice is preserved. | |
|
47
3bfb6c00c1e0
update it from 4.4.2 to 4.4.3.
kent <kent@cr.ie.u-ryukyu.ac.jp>
parents:
19
diff
changeset
|
5168 Last modified 2009-11-08 [420]Valid XHTML 1.0 |
| 0 | 5169 |
| 5170 References | |
| 5171 | |
| 5172 1. http://gcc.gnu.org/gcc-3.4/changes.html#3.4.6 | |
| 5173 2. http://gcc.gnu.org/gcc-3.4/changes.html#cplusplus | |
| 5174 3. http://gcc.gnu.org/gcc-3.3/changes.html#obsolete_systems | |
| 5175 4. http://gcc.gnu.org/gcc-3.4/changes.html#obsolete_systems | |
| 5176 5. http://gcc.gnu.org/gcc-3.4/mips-abi.html | |
| 5177 6. http://gcc.gnu.org/gcc-3.4/sparc-abi.html | |
| 5178 7. http://www.boost.org/ | |
| 5179 8. http://gcc.gnu.org/bugzilla/show_bug.cgi?id=11953 | |
| 5180 9. http://gcc.gnu.org/bugzilla/show_bug.cgi?id=8361 | |
| 5181 10. http://gcc.gnu.org/onlinedocs/gcc-3.4.3/gcc/Other-Builtins.html#Other%20Builtins | |
| 5182 11. http://www.open-std.org/jtc1/sc22/wg21/docs/cwg_closed.html#209 | |
|
47
3bfb6c00c1e0
update it from 4.4.2 to 4.4.3.
kent <kent@cr.ie.u-ryukyu.ac.jp>
parents:
19
diff
changeset
|
5183 12. http://gcc.gnu.org/bugs/#cxx_rvalbind |
| 0 | 5184 13. http://gcc.gnu.org/onlinedocs/gcc-3.4.3/gcc/Objective-C-Dialect-Options.html |
| 5185 14. http://gcc.gnu.org/onlinedocs/gcc-3.4.3/gcc/Objective-C-Dialect-Options.html | |
| 5186 15. http://gcc.gnu.org/onlinedocs/gcc-3.4.3/gcc/Objective-C-Dialect-Options.html | |
| 5187 16. http://www.gnu.org/software/classpath/ | |
| 5188 17. http://www.eclipse.org/ | |
| 5189 18. http://gcc.gnu.org/onlinedocs/gcc-3.4.3/g77/News.html | |
| 5190 19. http://gcc.gnu.org/onlinedocs/gcc-3.4.3/gcc/Alpha-Built-in-Functions.html | |
| 5191 20. http://h30097.www3.hp.com/docs/base_doc/DOCUMENTATION/V51A_HTML/ARH9MBTE/DTMNPLTN.HTM#normal-argument-list-structure | |
| 5192 21. http://gcc.gnu.org/onlinedocs/gcc-3.4.3/gccint/Processor-pipeline-description.html | |
| 5193 22. http://gcc.gnu.org/onlinedocs/gcc-3.4.3/gccint/Comparison-of-the-two-descriptions.html | |
| 5194 23. http://gcc.gnu.org/onlinedocs/gcc-3.4.3/gccint/Processor-pipeline-description.html | |
| 5195 24. http://gcc.gnu.org/gcc-3.4/mips-abi.html | |
| 5196 25. http://gcc.gnu.org/gcc-3.4/powerpc-abi.html | |
| 5197 26. http://gcc.gnu.org/gcc-3.4/sparc-abi.html | |
| 5198 27. http://gcc.gnu.org/bugzilla/buglist.cgi?short_desc_type=notregexp&short_desc=%5C%5B3%5C.4.*%5BRr%5Degression&target_milestone=3.4.0&bug_status=RESOLVED&resolution=FIXED | |
| 5199 28. http://gcc.gnu.org/PR10129 | |
| 5200 29. http://gcc.gnu.org/PR14576 | |
| 5201 30. http://gcc.gnu.org/PR14760 | |
| 5202 31. http://gcc.gnu.org/PR14671 | |
| 5203 32. http://gcc.gnu.org/PR15093 | |
| 5204 33. http://gcc.gnu.org/PR15178 | |
| 5205 34. http://gcc.gnu.org/PR12753 | |
| 5206 35. http://gcc.gnu.org/PR13985 | |
| 5207 36. http://gcc.gnu.org/PR14810 | |
| 5208 37. http://gcc.gnu.org/PR14883 | |
| 5209 38. http://gcc.gnu.org/PR15044 | |
| 5210 39. http://gcc.gnu.org/PR15057 | |
| 5211 40. http://gcc.gnu.org/PR15064 | |
| 5212 41. http://gcc.gnu.org/PR15142 | |
| 5213 42. http://gcc.gnu.org/PR15159 | |
| 5214 43. http://gcc.gnu.org/PR15165 | |
| 5215 44. http://gcc.gnu.org/PR15193 | |
| 5216 45. http://gcc.gnu.org/PR15209 | |
| 5217 46. http://gcc.gnu.org/PR15227 | |
| 5218 47. http://gcc.gnu.org/PR15285 | |
| 5219 48. http://gcc.gnu.org/PR15299 | |
| 5220 49. http://gcc.gnu.org/PR15329 | |
| 5221 50. http://gcc.gnu.org/PR15550 | |
| 5222 51. http://gcc.gnu.org/PR15554 | |
| 5223 52. http://gcc.gnu.org/PR15640 | |
| 5224 53. http://gcc.gnu.org/PR15666 | |
| 5225 54. http://gcc.gnu.org/PR15696 | |
| 5226 55. http://gcc.gnu.org/PR15701 | |
| 5227 56. http://gcc.gnu.org/PR15761 | |
| 5228 57. http://gcc.gnu.org/PR15829 | |
| 5229 58. http://gcc.gnu.org/PR14538 | |
| 5230 59. http://gcc.gnu.org/PR12391 | |
| 5231 60. http://gcc.gnu.org/PR14649 | |
| 5232 61. http://gcc.gnu.org/PR15004 | |
| 5233 62. http://gcc.gnu.org/PR15749 | |
| 5234 63. http://gcc.gnu.org/PR10646 | |
| 5235 64. http://gcc.gnu.org/PR12077 | |
| 5236 65. http://gcc.gnu.org/PR13598 | |
| 5237 66. http://gcc.gnu.org/PR14211 | |
| 5238 67. http://gcc.gnu.org/PR14220 | |
| 5239 68. http://gcc.gnu.org/PR14245 | |
| 5240 69. http://gcc.gnu.org/PR14340 | |
| 5241 70. http://gcc.gnu.org/PR14600 | |
| 5242 71. http://gcc.gnu.org/PR14668 | |
| 5243 72. http://gcc.gnu.org/PR14775 | |
| 5244 73. http://gcc.gnu.org/PR14821 | |
| 5245 74. http://gcc.gnu.org/PR14930 | |
| 5246 75. http://gcc.gnu.org/PR14932 | |
| 5247 76. http://gcc.gnu.org/PR14950 | |
| 5248 77. http://gcc.gnu.org/PR14962 | |
| 5249 78. http://gcc.gnu.org/PR14975 | |
| 5250 79. http://gcc.gnu.org/PR15002 | |
| 5251 80. http://gcc.gnu.org/PR15025 | |
| 5252 81. http://gcc.gnu.org/PR15046 | |
| 5253 82. http://gcc.gnu.org/PR15069 | |
| 5254 83. http://gcc.gnu.org/PR15074 | |
| 5255 84. http://gcc.gnu.org/PR15083 | |
| 5256 85. http://gcc.gnu.org/PR15096 | |
| 5257 86. http://gcc.gnu.org/PR15287 | |
| 5258 87. http://gcc.gnu.org/PR15317 | |
| 5259 88. http://gcc.gnu.org/PR15337 | |
| 5260 89. http://gcc.gnu.org/PR15361 | |
| 5261 90. http://gcc.gnu.org/PR15412 | |
| 5262 91. http://gcc.gnu.org/PR15427 | |
| 5263 92. http://gcc.gnu.org/PR15471 | |
| 5264 93. http://gcc.gnu.org/PR15503 | |
| 5265 94. http://gcc.gnu.org/PR15507 | |
| 5266 95. http://gcc.gnu.org/PR15542 | |
| 5267 96. http://gcc.gnu.org/PR15565 | |
| 5268 97. http://gcc.gnu.org/PR15625 | |
| 5269 98. http://gcc.gnu.org/PR15629 | |
| 5270 99. http://gcc.gnu.org/PR15742 | |
| 5271 100. http://gcc.gnu.org/PR15775 | |
| 5272 101. http://gcc.gnu.org/PR15821 | |
| 5273 102. http://gcc.gnu.org/PR15862 | |
| 5274 103. http://gcc.gnu.org/PR15875 | |
| 5275 104. http://gcc.gnu.org/PR15877 | |
| 5276 105. http://gcc.gnu.org/PR15947 | |
| 5277 106. http://gcc.gnu.org/PR16020 | |
| 5278 107. http://gcc.gnu.org/PR16154 | |
| 5279 108. http://gcc.gnu.org/PR16174 | |
| 5280 109. http://gcc.gnu.org/PR14315 | |
| 5281 110. http://gcc.gnu.org/PR15151 | |
| 5282 111. http://gcc.gnu.org/PR7993 | |
| 5283 112. http://gcc.gnu.org/PR15228 | |
| 5284 113. http://gcc.gnu.org/PR15345 | |
| 5285 114. http://gcc.gnu.org/PR15945 | |
| 5286 115. http://gcc.gnu.org/PR15526 | |
| 5287 116. http://gcc.gnu.org/PR14690 | |
| 5288 117. http://gcc.gnu.org/PR15112 | |
| 5289 118. http://gcc.gnu.org/PR15067 | |
| 5290 119. http://gcc.gnu.org/PR1963 | |
| 5291 120. http://gcc.gnu.org/PR15717 | |
| 5292 121. http://gcc.gnu.org/PR14782 | |
| 5293 122. http://gcc.gnu.org/PR14828 | |
| 5294 123. http://gcc.gnu.org/PR15202 | |
| 5295 124. http://gcc.gnu.org/PR14610 | |
| 5296 125. http://gcc.gnu.org/PR14813 | |
| 5297 126. http://gcc.gnu.org/PR14857 | |
| 5298 127. http://gcc.gnu.org/PR15598 | |
| 5299 128. http://gcc.gnu.org/PR15653 | |
| 5300 129. http://gcc.gnu.org/PR15189 | |
| 5301 130. http://gcc.gnu.org/PR15331 | |
| 5302 131. http://gcc.gnu.org/PR16144 | |
| 5303 132. http://gcc.gnu.org/PR16176 | |
| 5304 133. http://gcc.gnu.org/PR11591 | |
| 5305 134. http://gcc.gnu.org/PR12028 | |
| 5306 135. http://gcc.gnu.org/PR14478 | |
| 5307 136. http://gcc.gnu.org/PR14567 | |
| 5308 137. http://gcc.gnu.org/PR14715 | |
| 5309 138. http://gcc.gnu.org/PR14902 | |
| 5310 139. http://gcc.gnu.org/PR14924 | |
| 5311 140. http://gcc.gnu.org/PR14960 | |
| 5312 141. http://gcc.gnu.org/PR15106 | |
| 5313 142. http://gcc.gnu.org/PR16026 | |
| 5314 143. http://gcc.gnu.org/PR15191 | |
| 5315 144. http://gcc.gnu.org/PR15662 | |
| 5316 145. http://gcc.gnu.org/PR15054 | |
| 5317 146. http://gcc.gnu.org/PR15783 | |
| 5318 147. http://gcc.gnu.org/PR15626 | |
| 5319 148. http://gcc.gnu.org/PR14326 | |
| 5320 149. http://gcc.gnu.org/PR14723 | |
| 5321 150. http://gcc.gnu.org/PR15290 | |
| 5322 151. http://gcc.gnu.org/PR15250 | |
| 5323 152. http://gcc.gnu.org/PR15551 | |
| 5324 153. http://gcc.gnu.org/PR8309 | |
| 5325 154. http://gcc.gnu.org/PR13250 | |
| 5326 155. http://gcc.gnu.org/PR13803 | |
| 5327 156. http://gcc.gnu.org/PR14093 | |
| 5328 157. http://gcc.gnu.org/PR14457 | |
| 5329 158. http://gcc.gnu.org/PR14542 | |
| 5330 159. http://gcc.gnu.org/PR15100 | |
| 5331 160. http://gcc.gnu.org/PR15296 | |
| 5332 161. http://gcc.gnu.org/PR15396 | |
| 5333 162. http://gcc.gnu.org/PR15782 | |
| 5334 163. http://gcc.gnu.org/PR11610 | |
| 5335 164. http://gcc.gnu.org/PR15488 | |
| 5336 165. http://gcc.gnu.org/PR15489 | |
| 5337 166. http://gcc.gnu.org/PR13928 | |
| 5338 167. http://gcc.gnu.org/PR14150 | |
| 5339 168. http://gcc.gnu.org/PR14949 | |
| 5340 169. http://gcc.gnu.org/PR15123 | |
| 5341 170. http://gcc.gnu.org/PR16469 | |
| 5342 171. http://gcc.gnu.org/PR16344 | |
| 5343 172. http://gcc.gnu.org/PR16842 | |
| 5344 173. http://gcc.gnu.org/PR12608 | |
| 5345 174. http://gcc.gnu.org/PR14492 | |
| 5346 175. http://gcc.gnu.org/PR15461 | |
| 5347 176. http://gcc.gnu.org/PR15890 | |
| 5348 177. http://gcc.gnu.org/PR16180 | |
| 5349 178. http://gcc.gnu.org/PR16224 | |
| 5350 179. http://gcc.gnu.org/PR16408 | |
| 5351 180. http://gcc.gnu.org/PR16529 | |
| 5352 181. http://gcc.gnu.org/PR16698 | |
| 5353 182. http://gcc.gnu.org/PR16706 | |
| 5354 183. http://gcc.gnu.org/PR16810 | |
| 5355 184. http://gcc.gnu.org/PR16851 | |
| 5356 185. http://gcc.gnu.org/PR16870 | |
| 5357 186. http://gcc.gnu.org/PR16904 | |
| 5358 187. http://gcc.gnu.org/PR16905 | |
| 5359 188. http://gcc.gnu.org/PR16964 | |
| 5360 189. http://gcc.gnu.org/PR17068 | |
| 5361 190. http://gcc.gnu.org/PR16366 | |
| 5362 191. http://gcc.gnu.org/PR15345 | |
| 5363 192. http://gcc.gnu.org/PR16590 | |
| 5364 193. http://gcc.gnu.org/PR16693 | |
| 5365 194. http://gcc.gnu.org/PR17078 | |
| 5366 195. http://gcc.gnu.org/PR13956 | |
| 5367 196. http://gcc.gnu.org/PR16684 | |
| 5368 197. http://gcc.gnu.org/PR12658 | |
| 5369 198. http://gcc.gnu.org/PR13092 | |
| 5370 199. http://gcc.gnu.org/PR15320 | |
| 5371 200. http://gcc.gnu.org/PR16246 | |
| 5372 201. http://gcc.gnu.org/PR16273 | |
| 5373 202. http://gcc.gnu.org/PR16401 | |
| 5374 203. http://gcc.gnu.org/PR16411 | |
| 5375 204. http://gcc.gnu.org/PR16489 | |
| 5376 205. http://gcc.gnu.org/PR16618 | |
| 5377 206. http://gcc.gnu.org/PR16637 | |
| 5378 207. http://gcc.gnu.org/PR16717 | |
| 5379 208. http://gcc.gnu.org/PR16813 | |
| 5380 209. http://gcc.gnu.org/PR16853 | |
| 5381 210. http://gcc.gnu.org/PR16889 | |
| 5382 211. http://gcc.gnu.org/PR16959 | |
| 5383 212. http://gcc.gnu.org/PR7587 | |
| 5384 213. http://gcc.gnu.org/PR16473 | |
| 5385 214. http://gcc.gnu.org/PR16478 | |
| 5386 215. http://gcc.gnu.org/PR10695 | |
| 5387 216. http://gcc.gnu.org/PR16974 | |
| 5388 217. http://gcc.gnu.org/PR16298 | |
| 5389 218. http://gcc.gnu.org/PR17113 | |
| 5390 219. http://gcc.gnu.org/PR14697 | |
| 5391 220. http://gcc.gnu.org/PR15869 | |
| 5392 221. http://gcc.gnu.org/PR16325 | |
| 5393 222. http://gcc.gnu.org/PR16357 | |
| 5394 223. http://gcc.gnu.org/PR16380 | |
| 5395 224. http://gcc.gnu.org/PR16407 | |
| 5396 225. http://gcc.gnu.org/PR16643 | |
| 5397 226. http://gcc.gnu.org/PR15927 | |
| 5398 227. http://gcc.gnu.org/PR15948 | |
| 5399 228. http://gcc.gnu.org/PR17019 | |
| 5400 229. http://gcc.gnu.org/PR16130 | |
| 5401 230. http://gcc.gnu.org/PR16142 | |
| 5402 231. http://gcc.gnu.org/PR16278 | |
| 5403 232. http://gcc.gnu.org/PR16414 | |
| 5404 233. http://gcc.gnu.org/PR16445 | |
| 5405 234. http://gcc.gnu.org/PR16490 | |
| 5406 235. http://gcc.gnu.org/PR16683 | |
| 5407 236. http://gcc.gnu.org/PR16195 | |
| 5408 237. http://gcc.gnu.org/PR16239 | |
| 5409 238. http://gcc.gnu.org/PR16199 | |
| 5410 239. http://gcc.gnu.org/PR16416 | |
| 5411 240. http://gcc.gnu.org/PR16430 | |
| 5412 241. http://gcc.gnu.org/PR16379 | |
| 5413 242. http://gcc.gnu.org/PR17093 | |
| 5414 243. http://gcc.gnu.org/PR17119 | |
| 5415 244. http://gcc.gnu.org/PR15928 | |
| 5416 245. http://gcc.gnu.org/PR16210 | |
| 5417 246. http://gcc.gnu.org/PR15488 | |
| 5418 247. http://gcc.gnu.org/PR16250 | |
| 5419 248. http://gcc.gnu.org/bugzilla/buglist.cgi?bug_status=RESOLVED&resolution=FIXED&target_milestone=3.4.3 | |
| 5420 249. http://gcc.gnu.org/PR17369 | |
| 5421 250. http://gcc.gnu.org/PR17850 | |
| 5422 251. http://gcc.gnu.org/PR13948 | |
| 5423 252. http://gcc.gnu.org/PR14492 | |
| 5424 253. http://gcc.gnu.org/PR16301 | |
| 5425 254. http://gcc.gnu.org/PR16566 | |
| 5426 255. http://gcc.gnu.org/PR17023 | |
| 5427 256. http://gcc.gnu.org/PR17027 | |
| 5428 257. http://gcc.gnu.org/PR17524 | |
| 5429 258. http://gcc.gnu.org/PR17826 | |
| 5430 259. http://gcc.gnu.org/PR15526 | |
| 5431 260. http://gcc.gnu.org/PR16999 | |
| 5432 261. http://gcc.gnu.org/PR17503 | |
| 5433 262. http://gcc.gnu.org/PR17581 | |
| 5434 263. http://gcc.gnu.org/PR18129 | |
| 5435 264. http://gcc.gnu.org/PR10975 | |
| 5436 265. http://gcc.gnu.org/PR11722 | |
| 5437 266. http://gcc.gnu.org/PR14534 | |
| 5438 267. http://gcc.gnu.org/PR15172 | |
| 5439 268. http://gcc.gnu.org/PR15786 | |
| 5440 269. http://gcc.gnu.org/PR16162 | |
| 5441 270. http://gcc.gnu.org/PR16612 | |
| 5442 271. http://gcc.gnu.org/PR16715 | |
| 5443 272. http://gcc.gnu.org/PR16848 | |
| 5444 273. http://gcc.gnu.org/PR17132 | |
| 5445 274. http://gcc.gnu.org/PR17259 | |
| 5446 275. http://gcc.gnu.org/PR17327 | |
| 5447 276. http://gcc.gnu.org/PR17393 | |
| 5448 277. http://gcc.gnu.org/PR17501 | |
| 5449 278. http://gcc.gnu.org/PR17537 | |
| 5450 279. http://gcc.gnu.org/PR17585 | |
| 5451 280. http://gcc.gnu.org/PR17821 | |
| 5452 281. http://gcc.gnu.org/PR17829 | |
| 5453 282. http://gcc.gnu.org/PR17851 | |
| 5454 283. http://gcc.gnu.org/PR17976 | |
| 5455 284. http://gcc.gnu.org/PR18020 | |
| 5456 285. http://gcc.gnu.org/PR18093 | |
| 5457 286. http://gcc.gnu.org/PR18140 | |
| 5458 287. http://gcc.gnu.org/PR17541 | |
| 5459 288. http://gcc.gnu.org/PR17853 | |
| 5460 289. http://gcc.gnu.org/PR17245 | |
| 5461 290. http://gcc.gnu.org/PR17167 | |
| 5462 291. http://gcc.gnu.org/PR17277 | |
| 5463 292. http://gcc.gnu.org/PR17505 | |
| 5464 293. http://gcc.gnu.org/PR17684 | |
| 5465 294. http://gcc.gnu.org/PR17384 | |
| 5466 295. http://gcc.gnu.org/PR17770 | |
| 5467 296. http://gcc.gnu.org/PR11476 | |
| 5468 297. http://gcc.gnu.org/PR14064 | |
| 5469 298. http://gcc.gnu.org/PR14678 | |
| 5470 299. http://gcc.gnu.org/PR15583 | |
| 5471 300. http://gcc.gnu.org/PR15790 | |
| 5472 301. http://gcc.gnu.org/PR15886 | |
| 5473 302. http://gcc.gnu.org/PR16884 | |
| 5474 303. http://gcc.gnu.org/PR13841 | |
| 5475 304. http://gcc.gnu.org/PR15860 | |
| 5476 305. http://gcc.gnu.org/PR17465 | |
| 5477 306. http://gcc.gnu.org/PR17469 | |
| 5478 307. http://gcc.gnu.org/PR18138 | |
| 5479 308. http://gcc.gnu.org/PR15498 | |
| 5480 309. http://gcc.gnu.org/PR15747 | |
| 5481 310. http://gcc.gnu.org/PR16406 | |
| 5482 311. http://gcc.gnu.org/bugzilla/buglist.cgi?bug_status=RESOLVED&resolution=FIXED&target_milestone=3.4.4 | |
| 5483 312. http://gcc.gnu.org/bugzilla/buglist.cgi?bug_status=RESOLVED&resolution=FIXED&target_milestone=3.4.5 | |
| 5484 313. http://gcc.gnu.org/PR24688 | |
| 5485 314. http://gcc.gnu.org/PR17188 | |
| 5486 315. http://gcc.gnu.org/PR20187 | |
| 5487 316. http://gcc.gnu.org/PR21873 | |
| 5488 317. http://gcc.gnu.org/PR21899 | |
| 5489 318. http://gcc.gnu.org/PR22061 | |
| 5490 319. http://gcc.gnu.org/PR22208 | |
| 5491 320. http://gcc.gnu.org/PR22458 | |
| 5492 321. http://gcc.gnu.org/PR22589 | |
| 5493 322. http://gcc.gnu.org/PR24101 | |
| 5494 323. http://gcc.gnu.org/PR10611 | |
| 5495 324. http://gcc.gnu.org/PR13377 | |
| 5496 325. http://gcc.gnu.org/PR16002 | |
| 5497 326. http://gcc.gnu.org/PR17413 | |
| 5498 327. http://gcc.gnu.org/PR17609 | |
| 5499 328. http://gcc.gnu.org/PR17618 | |
| 5500 329. http://gcc.gnu.org/PR18124 | |
| 5501 330. http://gcc.gnu.org/PR18155 | |
| 5502 331. http://gcc.gnu.org/PR18177 | |
| 5503 332. http://gcc.gnu.org/PR18368 | |
| 5504 333. http://gcc.gnu.org/PR18378 | |
| 5505 334. http://gcc.gnu.org/PR18466 | |
| 5506 335. http://gcc.gnu.org/PR18512 | |
| 5507 336. http://gcc.gnu.org/PR18545 | |
| 5508 337. http://gcc.gnu.org/PR18738 | |
| 5509 338. http://gcc.gnu.org/PR18803 | |
| 5510 339. http://gcc.gnu.org/PR19004 | |
| 5511 340. http://gcc.gnu.org/PR19208 | |
| 5512 341. http://gcc.gnu.org/PR19253 | |
| 5513 342. http://gcc.gnu.org/PR19608 | |
| 5514 343. http://gcc.gnu.org/PR19884 | |
| 5515 344. http://gcc.gnu.org/PR20153 | |
| 5516 345. http://gcc.gnu.org/PR20563 | |
| 5517 346. http://gcc.gnu.org/PR20789 | |
| 5518 347. http://gcc.gnu.org/PR21336 | |
| 5519 348. http://gcc.gnu.org/PR21768 | |
| 5520 349. http://gcc.gnu.org/PR21853 | |
| 5521 350. http://gcc.gnu.org/PR21903 | |
| 5522 351. http://gcc.gnu.org/PR21983 | |
| 5523 352. http://gcc.gnu.org/PR21987 | |
| 5524 353. http://gcc.gnu.org/PR22153 | |
| 5525 354. http://gcc.gnu.org/PR22172 | |
| 5526 355. http://gcc.gnu.org/PR21286 | |
| 5527 356. http://gcc.gnu.org/PR22233 | |
| 5528 357. http://gcc.gnu.org/PR22508 | |
| 5529 358. http://gcc.gnu.org/PR22545 | |
| 5530 359. http://gcc.gnu.org/PR23528 | |
| 5531 360. http://gcc.gnu.org/PR23550 | |
| 5532 361. http://gcc.gnu.org/PR23586 | |
| 5533 362. http://gcc.gnu.org/PR23624 | |
| 5534 363. http://gcc.gnu.org/PR23639 | |
| 5535 364. http://gcc.gnu.org/PR23797 | |
| 5536 365. http://gcc.gnu.org/PR23965 | |
| 5537 366. http://gcc.gnu.org/PR24052 | |
| 5538 367. http://gcc.gnu.org/PR24580 | |
| 5539 368. http://gcc.gnu.org/PR24267 | |
| 5540 369. http://gcc.gnu.org/PR17810 | |
| 5541 370. http://gcc.gnu.org/PR17860 | |
| 5542 371. http://gcc/gnu.org/PR21709 | |
| 5543 372. http://gcc.gnu.org/PR21964 | |
| 5544 373. http://gcc.gnu.org/PR22167 | |
| 5545 374. http://gcc.gnu.org/PR22619 | |
| 5546 375. http://gcc.gnu.org/PR23241 | |
| 5547 376. http://gcc.gnu.org/PR23478 | |
| 5548 377. http://gcc.gnu.org/PR24470 | |
| 5549 378. http://gcc.gnu.org/PR24950 | |
| 5550 379. http://gcc.gnu.org/PR14400 | |
| 5551 380. http://gcc.gnu.org/PR14940 | |
| 5552 381. http://gcc.gnu.org/PR20239 | |
| 5553 382. http://gcc.gnu.org/PR15220 | |
| 5554 383. http://gcc.gnu.org/PR19275 | |
| 5555 384. http://gcc.gnu.org/PR21888 | |
| 5556 385. http://gcc.gnu.org/PR15342 | |
| 5557 386. http://gcc.gnu.org/PR23985 | |
| 5558 387. http://gcc.gnu.org/PR16719 | |
| 5559 388. http://gcc.gnu.org/PR21723 | |
| 5560 389. http://gcc.gnu.org/PR21841 | |
| 5561 390. http://gcc.gnu.org/PR23644 | |
| 5562 391. http://gcc.gnu.org/PR24718 | |
| 5563 392. http://gcc.gnu.org/PR18421 | |
| 5564 393. http://gcc.gnu.org/PR20621 | |
| 5565 394. http://gcc.gnu.org/PR18583 | |
| 5566 395. http://gcc.gnu.org/PR20191 | |
| 5567 396. http://gcc.gnu.org/PR22083 | |
| 5568 397. http://gcc.gnu.org/PR23070 | |
| 5569 398. http://gcc.gnu.org/PR23404 | |
| 5570 399. http://gcc.gnu.org/PR23539 | |
| 5571 400. http://gcc.gnu.org/PR24102 | |
| 5572 401. http://gcc.gnu.org/PR24465 | |
| 5573 402. http://gcc.gnu.org/PR19933 | |
| 5574 403. http://gcc.gnu.org/PR21889 | |
| 5575 404. http://gcc.gnu.org/PR19300 | |
| 5576 405. http://gcc.gnu.org/PR20301 | |
| 5577 406. http://gcc.gnu.org/PR20673 | |
| 5578 407. http://gcc.gnu.org/PR18582 | |
| 5579 408. http://gcc.gnu.org/PR19340 | |
| 5580 409. http://gcc.gnu.org/PR21716 | |
| 5581 410. http://gcc.gnu.org/PR24315 | |
| 5582 411. http://gcc.gnu.org/bugzilla/buglist.cgi?bug_status=RESOLVED&resolution=FIXED&target_milestone=3.4.6 | |
| 5583 412. mailto:gnu@gnu.org | |
| 5584 413. http://www.gnu.org/home.html#ContactInfo | |
| 5585 414. http://gcc.gnu.org/about.html | |
| 5586 415. http://gcc.gnu.org/onlinedocs/ | |
| 5587 416. mailto:gcc-help@gcc.gnu.org | |
| 5588 417. mailto:gcc@gnu.org | |
| 5589 418. mailto:gcc@gcc.gnu.org | |
| 5590 419. http://gcc.gnu.org/lists.html | |
| 5591 420. http://validator.w3.org/check/referer | |
| 5592 ====================================================================== | |
| 5593 http://gcc.gnu.org/gcc-3.3/index.html | |
| 5594 GCC 3.3 Release Series | |
| 5595 | |
| 5596 May 03, 2005 | |
| 5597 | |
| 5598 The [1]GNU project and the GCC developers are pleased to announce the | |
| 5599 release of GCC 3.3.6. | |
| 5600 | |
| 5601 This release is a bug-fix release, containing fixes for regressions in | |
| 5602 GCC 3.3.5 relative to previous releases of GCC. | |
| 5603 | |
| 5604 This release is the last of the series 3.3.x. | |
| 5605 | |
| 5606 The GCC 3.3 release series includes numerous [2]new features, | |
| 5607 improvements, bug fixes, and other changes, thanks to an [3]amazing | |
| 5608 group of volunteers. | |
| 5609 | |
| 5610 Release History | |
| 5611 | |
| 5612 GCC 3.3.6 | |
| 5613 May 3, 2005 ([4]changes) | |
| 5614 | |
| 5615 GCC 3.3.5 | |
| 5616 September 30, 2004 ([5]changes) | |
| 5617 | |
| 5618 GCC 3.3.4 | |
| 5619 May 31, 2004 ([6]changes) | |
| 5620 | |
| 5621 GCC 3.3.3 | |
| 5622 February 14, 2004 ([7]changes) | |
| 5623 | |
| 5624 GCC 3.3.2 | |
| 5625 October 16, 2003 ([8]changes) | |
| 5626 | |
| 5627 GCC 3.3.1 | |
| 5628 August 8, 2003 ([9]changes) | |
| 5629 | |
| 5630 GCC 3.3 | |
| 5631 May 14, 2003 ([10]changes) | |
| 5632 | |
| 5633 References and Acknowledgements | |
| 5634 | |
| 5635 GCC used to stand for the GNU C Compiler, but since the compiler | |
| 5636 supports several other languages aside from C, it now stands for the | |
| 5637 GNU Compiler Collection. | |
| 5638 | |
| 5639 A list of [11]successful builds is updated as new information becomes | |
| 5640 available. | |
| 5641 | |
| 5642 The GCC developers would like to thank the numerous people that have | |
| 5643 contributed new features, improvements, bug fixes, and other changes as | |
| 5644 well as test results to GCC. This [12]amazing group of volunteers is | |
| 5645 what makes GCC successful. | |
| 5646 | |
| 5647 For additional information about GCC please refer to the [13]GCC | |
| 5648 project web site or contact the [14]GCC development mailing list. | |
| 5649 | |
| 5650 To obtain GCC please use [15]our mirror sites, or our CVS server. | |
| 5651 | |
| 5652 Please send FSF & GNU inquiries & questions to [16]gnu@gnu.org. There | |
| 5653 are also [17]other ways to contact the FSF. | |
| 5654 | |
| 5655 These pages are maintained by [18]the GCC team. | |
| 5656 | |
| 5657 | |
| 5658 For questions related to the use of GCC, please consult these web | |
| 5659 pages and the [19]GCC manuals. If that fails, the | |
| 5660 [20]gcc-help@gcc.gnu.org mailing list might help. | |
| 5661 Please send comments on these web pages and the development of GCC to | |
| 5662 our developer mailing list at [21]gcc@gnu.org or [22]gcc@gcc.gnu.org. | |
| 5663 All of our lists have [23]public archives. | |
| 5664 | |
| 5665 Copyright (C) Free Software Foundation, Inc., 51 Franklin St, Fifth | |
| 5666 Floor, Boston, MA 02110, USA. | |
| 5667 | |
| 5668 Verbatim copying and distribution of this entire article is permitted | |
| 5669 in any medium, provided this notice is preserved. | |
| 5670 Last modified 2008-07-26 [24]Valid XHTML 1.0 | |
| 5671 | |
| 5672 References | |
| 5673 | |
| 5674 1. http://www.gnu.org/ | |
| 5675 2. http://gcc.gnu.org/gcc-3.3/changes.html | |
| 5676 3. http://gcc.gnu.org/onlinedocs/gcc/Contributors.html | |
| 5677 4. http://gcc.gnu.org/gcc-3.3/changes.html#3.3.6 | |
| 5678 5. http://gcc.gnu.org/gcc-3.3/changes.html#3.3.5 | |
| 5679 6. http://gcc.gnu.org/gcc-3.3/changes.html#3.3.4 | |
| 5680 7. http://gcc.gnu.org/gcc-3.3/changes.html#3.3.3 | |
| 5681 8. http://gcc.gnu.org/gcc-3.3/changes.html#3.3.2 | |
| 5682 9. http://gcc.gnu.org/gcc-3.3/changes.html#3.3.1 | |
| 5683 10. http://gcc.gnu.org/gcc-3.3/changes.html | |
| 5684 11. http://gcc.gnu.org/gcc-3.3/buildstat.html | |
| 5685 12. http://gcc.gnu.org/onlinedocs/gcc/Contributors.html | |
| 5686 13. http://gcc.gnu.org/index.html | |
| 5687 14. mailto:gcc@gcc.gnu.org | |
| 5688 15. http://gcc.gnu.org/mirrors.html | |
| 5689 16. mailto:gnu@gnu.org | |
| 5690 17. http://www.gnu.org/home.html#ContactInfo | |
| 5691 18. http://gcc.gnu.org/about.html | |
| 5692 19. http://gcc.gnu.org/onlinedocs/ | |
| 5693 20. mailto:gcc-help@gcc.gnu.org | |
| 5694 21. mailto:gcc@gnu.org | |
| 5695 22. mailto:gcc@gcc.gnu.org | |
| 5696 23. http://gcc.gnu.org/lists.html | |
| 5697 24. http://validator.w3.org/check/referer | |
| 5698 ====================================================================== | |
| 5699 http://gcc.gnu.org/gcc-3.3/changes.html | |
| 5700 GCC 3.3 Release Series | |
| 5701 Changes, New Features, and Fixes | |
| 5702 | |
| 5703 The latest release in the 3.3 release series is [1]GCC 3.3.6. | |
| 5704 | |
| 5705 Caveats | |
| 5706 | |
| 5707 * The preprocessor no longer accepts multi-line string literals. They | |
| 5708 were deprecated in 3.0, 3.1, and 3.2. | |
| 5709 * The preprocessor no longer supports the -A- switch when appearing | |
| 5710 alone. -A- followed by an assertion is still supported. | |
| 5711 * Support for all the systems [2]obsoleted in GCC 3.1 has been | |
| 5712 removed from GCC 3.3. See below for a [3]list of systems which are | |
| 5713 obsoleted in this release. | |
| 5714 * Checking for null format arguments has been decoupled from the rest | |
| 5715 of the format checking mechanism. Programs which use the format | |
| 5716 attribute may regain this functionality by using the new [4]nonnull | |
| 5717 function attribute. Note that all functions for which GCC has a | |
| 5718 built-in format attribute, an appropriate built-in nonnull | |
| 5719 attribute is also applied. | |
| 5720 * The DWARF (version 1) debugging format has been deprecated and will | |
| 5721 be removed in a future version of GCC. Version 2 of the DWARF | |
| 5722 debugging format will continue to be supported for the foreseeable | |
| 5723 future. | |
| 5724 * The C and Objective-C compilers no longer accept the "Naming Types" | |
| 5725 extension (typedef foo = bar); it was already unavailable in C++. | |
| 5726 Code which uses it will need to be changed to use the "typeof" | |
| 5727 extension instead: typedef typeof(bar) foo. (We have removed this | |
| 5728 extension without a period of deprecation because it has caused the | |
| 5729 compiler to crash since version 3.0 and no one noticed until very | |
| 5730 recently. Thus we conclude it is not in widespread use.) | |
| 5731 * The -traditional C compiler option has been removed. It was | |
| 5732 deprecated in 3.1 and 3.2. (Traditional preprocessing remains | |
| 5733 available.) The <varargs.h> header, used for writing variadic | |
| 5734 functions in traditional C, still exists but will produce an error | |
| 5735 message if used. | |
| 5736 * GCC 3.3.1 automatically places zero-initialized variables in the | |
| 5737 .bss section on some operating systems. Versions of GNU Emacs up to | |
| 5738 (and including) 21.3 will not work correctly when using this | |
| 5739 optimization; you can use -fno-zero-initialized-in-bss to disable | |
| 5740 it. | |
| 5741 | |
| 5742 General Optimizer Improvements | |
| 5743 | |
| 5744 * A new scheme for accurately describing processor pipelines, the | |
| 5745 [5]DFA scheduler, has been added. | |
| 5746 * Pavel Nejedly, Charles University Prague, has contributed new file | |
| 5747 format used by the edge coverage profiler (-fprofile-arcs). | |
| 5748 The new format is robust and diagnoses common mistakes where | |
| 5749 profiles from different versions (or compilations) of the program | |
| 5750 are combined resulting in nonsensical profiles and slow code to | |
| 5751 produced with profile feedback. Additionally this format allows | |
| 5752 extra data to be gathered. Currently, overall statistics are | |
| 5753 produced helping optimizers to identify hot spots of a program | |
| 5754 globally replacing the old intra-procedural scheme and resulting in | |
| 5755 better code. Note that the gcov tool from older GCC versions will | |
| 5756 not be able to parse the profiles generated by GCC 3.3 and vice | |
| 5757 versa. | |
| 5758 * Jan Hubicka, SuSE Labs, has contributed a new superblock formation | |
| 5759 pass enabled using -ftracer. This pass simplifies the control flow | |
| 5760 of functions allowing other optimizations to do better job. | |
| 5761 He also contributed the function reordering pass | |
| 5762 (-freorder-functions) to optimize function placement using profile | |
| 5763 feedback. | |
| 5764 | |
| 5765 New Languages and Language specific improvements | |
| 5766 | |
| 5767 C/ObjC/C++ | |
| 5768 | |
| 5769 * The preprocessor now accepts directives within macro arguments. It | |
| 5770 processes them just as if they had not been within macro arguments. | |
| 5771 * The separate ISO and traditional preprocessors have been completely | |
| 5772 removed. The front end handles either type of preprocessed output | |
| 5773 if necessary. | |
| 5774 * In C99 mode preprocessor arithmetic is done in the precision of the | |
| 5775 target's intmax_t, as required by that standard. | |
| 5776 * The preprocessor can now copy comments inside macros to the output | |
| 5777 file when the macro is expanded. This feature, enabled using the | |
| 5778 -CC option, is intended for use by applications which place | |
| 5779 metadata or directives inside comments, such as lint. | |
| 5780 * The method of constructing the list of directories to be searched | |
| 5781 for header files has been revised. If a directory named by a -I | |
| 5782 option is a standard system include directory, the option is | |
| 5783 ignored to ensure that the default search order for system | |
| 5784 directories and the special treatment of system header files are | |
| 5785 not defeated. | |
| 5786 * A few more [6]ISO C99 features now work correctly. | |
| 5787 * A new function attribute, nonnull, has been added which allows | |
| 5788 pointer arguments to functions to be specified as requiring a | |
| 5789 non-null value. The compiler currently uses this information to | |
| 5790 issue a warning when it detects a null value passed in such an | |
| 5791 argument slot. | |
| 5792 * A new type attribute, may_alias, has been added. Accesses to | |
| 5793 objects with types with this attribute are not subjected to | |
| 5794 type-based alias analysis, but are instead assumed to be able to | |
| 5795 alias any other type of objects, just like the char type. | |
| 5796 | |
| 5797 C++ | |
| 5798 | |
| 5799 * Type based alias analysis has been implemented for C++ aggregate | |
| 5800 types. | |
| 5801 | |
| 5802 Objective-C | |
| 5803 | |
| 5804 * Generate an error if Objective-C objects are passed by value in | |
| 5805 function and method calls. | |
| 5806 * When -Wselector is used, check the whole list of selectors at the | |
| 5807 end of compilation, and emit a warning if a @selector() is not | |
| 5808 known. | |
| 5809 * Define __NEXT_RUNTIME__ when compiling for the NeXT runtime. | |
| 5810 * No longer need to include objc/objc-class.h to compile self calls | |
| 5811 in class methods (NeXT runtime only). | |
| 5812 * New -Wundeclared-selector option. | |
| 5813 * Removed selector bloating which was causing object files to be 10% | |
| 5814 bigger on average (GNU runtime only). | |
| 5815 * Using at run time @protocol() objects has been fixed in certain | |
| 5816 situations (GNU runtime only). | |
| 5817 * Type checking has been fixed and improved in many situations | |
| 5818 involving protocols. | |
| 5819 | |
| 5820 Java | |
| 5821 | |
| 5822 * The java.sql and javax.sql packages now implement the JDBC 3.0 (JDK | |
| 5823 1.4) API. | |
| 5824 * The JDK 1.4 assert facility has been implemented. | |
| 5825 * The bytecode interpreter is now direct threaded and thus faster. | |
| 5826 | |
| 5827 Fortran | |
| 5828 | |
| 5829 * Fortran improvements are listed in [7]the Fortran documentation. | |
| 5830 | |
| 5831 Ada | |
| 5832 | |
| 5833 * Ada tasking now works with glibc 2.3.x threading libraries. | |
| 5834 | |
| 5835 New Targets and Target Specific Improvements | |
| 5836 | |
| 5837 * The following changes have been made to the HP-PA port: | |
| 5838 + The port now defaults to scheduling for the PA8000 series of | |
| 5839 processors. | |
| 5840 + Scheduling support for the PA7300 processor has been added. | |
| 5841 + The 32-bit port now supports weak symbols under HP-UX 11. | |
| 5842 + The handling of initializers and finalizers has been improved | |
| 5843 under HP-UX 11. The 64-bit port no longer uses collect2. | |
| 5844 + Dwarf2 EH support has been added to the 32-bit linux port. | |
| 5845 + ABI fixes to correct the passing of small structures by value. | |
| 5846 * The SPARC, HP-PA, SH4, and x86/pentium ports have been converted to | |
| 5847 use the DFA processor pipeline description. | |
| 5848 * The following NetBSD configurations for the SuperH processor family | |
| 5849 have been added: | |
| 5850 + SH3, big-endian, sh-*-netbsdelf* | |
| 5851 + SH3, little-endian, shle-*-netbsdelf* | |
| 5852 + SH5, SHmedia, big-endian, 32-bit default, sh5-*-netbsd* | |
| 5853 + SH5, SHmedia, little-endian, 32-bit default, sh5le-*-netbsd* | |
| 5854 + SH5, SHmedia, big-endian, 64-bit default, sh64-*-netbsd* | |
| 5855 + SH5, SHmedia, little-endian, 64-bit default, sh64le-*-netbsd* | |
| 5856 * The following changes have been made to the IA-32/x86-64 port: | |
| 5857 + SSE2 and 3dNOW! intrinsics are now supported. | |
| 5858 + Support for thread local storage has been added to the IA-32 | |
| 5859 and x86-64 ports. | |
| 5860 + The x86-64 port has been significantly improved. | |
| 5861 * The following changes have been made to the MIPS port: | |
| 5862 + All configurations now accept the -mabi switch. Note that you | |
| 5863 will need appropriate multilibs for this option to work | |
| 5864 properly. | |
| 5865 + ELF configurations will always pass an ABI flag to the | |
| 5866 assembler, except when the MIPS EABI is selected. | |
| 5867 + -mabi=64 no longer selects MIPS IV code. | |
| 5868 + The -mcpu option, which was deprecated in 3.1 and 3.2, has | |
| 5869 been removed from this release. | |
| 5870 + -march now changes the core ISA level. In previous releases, | |
| 5871 it would change the use of processor-specific extensions, but | |
| 5872 would leave the core ISA unchanged. For example, mips64-elf | |
| 5873 -march=r8000 will now generate MIPS IV code. | |
| 5874 + Under most configurations, -mipsN now acts as a synonym for | |
| 5875 -march. | |
| 5876 + There are some new preprocessor macros to describe the -march | |
| 5877 and -mtune settings. See the documentation of those options | |
| 5878 for details. | |
| 5879 + Support for the NEC VR-Series processors has been added. This | |
| 5880 includes the 54xx, 5500, and 41xx series. | |
| 5881 + Support for the Sandcraft sr71k processor has been added. | |
| 5882 * The following changes have been made to the S/390 port: | |
| 5883 + Support to build the Java runtime libraries has been added. | |
| 5884 Java is now enabled by default on s390-*-linux* and | |
| 5885 s390x-*-linux* targets. | |
| 5886 + Multilib support for the s390x-*-linux* target has been added; | |
| 5887 this allows to build 31-bit binaries using the -m31 option. | |
| 5888 + Support for thread local storage has been added. | |
| 5889 + Inline assembler code may now use the 'Q' constraint to | |
| 5890 specify memory operands without index register. | |
| 5891 + Various platform-specific performance improvements have been | |
| 5892 implemented; in particular, the compiler now uses the BRANCH | |
| 5893 ON COUNT family of instructions and makes more frequent use of | |
| 5894 the TEST UNDER MASK family of instructions. | |
| 5895 * The following changes have been made to the PowerPC port: | |
| 5896 + Support for IBM Power4 processor added. | |
| 5897 + Support for Motorola e500 SPE added. | |
| 5898 + Support for AIX 5.2 added. | |
| 5899 + Function and Data sections now supported on AIX. | |
| 5900 + Sibcall optimizations added. | |
| 5901 * The support for H8 Tiny is added to the H8/300 port with -mn. | |
| 5902 | |
| 5903 Obsolete Systems | |
| 5904 | |
| 5905 Support for a number of older systems has been declared obsolete in GCC | |
| 5906 3.3. Unless there is activity to revive them, the next release of GCC | |
| 5907 will have their sources permanently removed. | |
| 5908 | |
| 5909 All configurations of the following processor architectures have been | |
| 5910 declared obsolete: | |
| 5911 * Matsushita MN10200, mn10200-*-* | |
| 5912 * Motorola 88000, m88k-*-* | |
| 5913 * IBM ROMP, romp-*-* | |
| 5914 | |
| 5915 Also, some individual systems have been obsoleted: | |
| 5916 * Alpha | |
| 5917 + Interix, alpha*-*-interix* | |
| 5918 + Linux libc1, alpha*-*-linux*libc1* | |
| 5919 + Linux ECOFF, alpha*-*-linux*ecoff* | |
| 5920 * ARM | |
| 5921 + Generic a.out, arm*-*-aout* | |
| 5922 + Conix, arm*-*-conix* | |
| 5923 + "Old ABI," arm*-*-oabi | |
| 5924 + StrongARM/COFF, strongarm-*-coff* | |
| 5925 * HPPA (PA-RISC) | |
| 5926 + Generic OSF, hppa1.0-*-osf* | |
| 5927 + Generic BSD, hppa1.0-*-bsd* | |
| 5928 + HP/UX versions 7, 8, and 9, hppa1.[01]-*-hpux[789]* | |
| 5929 + HiUX, hppa*-*-hiux* | |
| 5930 + Mach Lites, hppa*-*-lites* | |
| 5931 * Intel 386 family | |
| 5932 + Windows NT 3.x, i?86-*-win32 | |
| 5933 * MC68000 family | |
| 5934 + HP systems, m68000-hp-bsd* and m68k-hp-bsd* | |
| 5935 + Sun systems, m68000-sun-sunos*, m68k-sun-sunos*, and | |
| 5936 m68k-sun-mach* | |
| 5937 + AT&T systems, m68000-att-sysv* | |
| 5938 + Atari systems, m68k-atari-sysv* | |
| 5939 + Motorola systems, m68k-motorola-sysv* | |
| 5940 + NCR systems, m68k-ncr-sysv* | |
| 5941 + Plexus systems, m68k-plexus-sysv* | |
| 5942 + Commodore systems, m68k-cbm-sysv* | |
| 5943 + Citicorp TTI, m68k-tti-* | |
| 5944 + Unos, m68k-crds-unos* | |
| 5945 + Concurrent RTU, m68k-ccur-rtu* | |
| 5946 + Linux a.out, m68k-*-linux*aout* | |
| 5947 + Linux libc1, m68k-*-linux*libc1* | |
| 5948 + pSOS, m68k-*-psos* | |
| 5949 * MIPS | |
| 5950 + Generic ECOFF, mips*-*-ecoff* | |
| 5951 + SINIX, mips-sni-sysv4 | |
| 5952 + Orion RTEMS, mips64orion-*-rtems* | |
| 5953 * National Semiconductor 32000 | |
| 5954 + OpenBSD, ns32k-*-openbsd* | |
| 5955 * POWER (aka RS/6000) and PowerPC | |
| 5956 + AIX versions 1, 2, and 3, rs6000-ibm-aix[123]* | |
| 5957 + Bull BOSX, rs6000-bull-bosx | |
| 5958 + Generic Mach, rs6000-*-mach* | |
| 5959 + Generic SysV, powerpc*-*-sysv* | |
| 5960 + Linux libc1, powerpc*-*-linux*libc1* | |
| 5961 * Sun SPARC | |
| 5962 + Generic a.out, sparc-*-aout*, sparclet-*-aout*, | |
| 5963 sparclite-*-aout*, and sparc86x-*-aout* | |
| 5964 + NetBSD a.out, sparc-*-netbsd*aout* | |
| 5965 + Generic BSD, sparc-*-bsd* | |
| 5966 + ChorusOS, sparc-*-chorusos* | |
| 5967 + Linux a.out, sparc-*-linux*aout* | |
| 5968 + Linux libc1, sparc-*-linux*libc1* | |
| 5969 + LynxOS, sparc-*-lynxos* | |
| 5970 + Solaris on HAL hardware, sparc-hal-solaris2* | |
| 5971 + SunOS versions 3 and 4, sparc-*-sunos[34]* | |
| 5972 * NEC V850 | |
| 5973 + RTEMS, v850-*-rtems* | |
| 5974 * VAX | |
| 5975 + VMS, vax-*-vms* | |
| 5976 | |
| 5977 Documentation improvements | |
| 5978 | |
| 5979 Other significant improvements | |
| 5980 | |
| 5981 * Almost all front-end dependencies in the compiler have been | |
| 5982 separated out into a set of language hooks. This should make adding | |
| 5983 a new front end clearer and easier. | |
| 5984 * One effect of removing the separate preprocessor is a small | |
| 5985 increase in the robustness of the compiler in general, and the | |
| 5986 maintainability of target descriptions. Previously target-specific | |
| 5987 built-in macros and others, such as __FAST_MATH__, had to be | |
| 5988 handled with so-called specs that were hard to maintain. Often they | |
| 5989 would fail to behave properly when conflicting options were | |
| 5990 supplied on the command line, and define macros in the user's | |
| 5991 namespace even when strict ISO compliance was requested. | |
| 5992 Integrating the preprocessor has cleanly solved these issues. | |
| 5993 * The Makefile suite now supports redirection of make install by | |
| 5994 means of the variable DESTDIR. | |
| 5995 __________________________________________________________________ | |
| 5996 | |
| 5997 GCC 3.3 | |
| 5998 | |
| 5999 Detailed release notes for the GCC 3.3 release follow. | |
| 6000 | |
| 6001 Bug Fixes | |
| 6002 | |
| 6003 bootstrap failures | |
| 6004 | |
| 6005 * [8]10140 cross compiler build failures: missing __mempcpy (DUP: | |
| 6006 [9]10198,[10]10338) | |
| 6007 | |
| 6008 Internal compiler errors (multi-platform) | |
| 6009 | |
| 6010 * [11]3581 large string causes segmentation fault in cc1 | |
| 6011 * [12]4382 __builtin_{set,long}jmp with -O3 can crash the compiler | |
| 6012 * [13]5533 (c++) ICE when processing std::accumulate(begin, end, | |
| 6013 init, invalid_op) | |
| 6014 * [14]6387 -fpic -gdwarf-2 -g1 combination gives ICE in dwarf2out | |
| 6015 * [15]6412 (c++) ICE in retrieve_specialization | |
| 6016 * [16]6620 (c++) partial template specialization causes an ICE | |
| 6017 (segmentation fault) | |
| 6018 * [17]6663 (c++) ICE with attribute aligned | |
| 6019 * [18]7068 ICE with incomplete types | |
| 6020 * [19]7083 (c++) ICE using -gstabs with dodgy class derivation | |
| 6021 * [20]7647 (c++) ICE when data member has the name of the enclosing | |
| 6022 class | |
| 6023 * [21]7675 ICE in fixup_var_refs_1 | |
| 6024 * [22]7718 'complex' template instantiation causes ICE | |
| 6025 * [23]8116 (c++) ICE in member template function | |
| 6026 * [24]8358 (ada) Ada compiler accesses freed memory, crashes | |
| 6027 * [25]8511 (c++) ICE: (hopefully) reproducible cc1plus segmentation | |
| 6028 fault | |
| 6029 * [26]8564 (c++) ICE in find_function_data, in function.c | |
| 6030 * [27]8660 (c++) template overloading ICE in tsubst_expr, in cp/pt.c | |
| 6031 * [28]8766 (c++) ICE after failed initialization of static template | |
| 6032 variable | |
| 6033 * [29]8803 ICE in instantiate_virtual_regs_1, in function.c | |
| 6034 * [30]8846 (c++) ICE after diagnostic if fr_FR@euro locale is set | |
| 6035 * [31]8906 (c++) ICE (Segmentation fault) when parsing nested-class | |
| 6036 definition | |
| 6037 * [32]9216 (c++) ICE on missing template parameter | |
| 6038 * [33]9261 (c++) ICE in arg_assoc, in cp/decl2.c | |
| 6039 * [34]9263 (fortran) ICE caused by invalid PARAMETER in implied DO | |
| 6040 loop | |
| 6041 * [35]9429 (c++) ICE in template instantiation with a pointered new | |
| 6042 operator | |
| 6043 * [36]9516 Internal error when using a big array | |
| 6044 * [37]9600 (c++) ICE with typedefs in template class | |
| 6045 * [38]9629 (c++) virtual inheritance segfault | |
| 6046 * [39]9672 (c++) ICE: Error reporting routines re-entered | |
| 6047 * [40]9749 (c++) ICE in write_expression on invalid function | |
| 6048 prototype | |
| 6049 * [41]9794 (fortran) ICE: floating point exception during constant | |
| 6050 folding | |
| 6051 * [42]9829 (c++) Missing colon in nested namespace usage causes ICE | |
| 6052 * [43]9916 (c++) ICE with noreturn function in ?: statement | |
| 6053 * [44]9936 ICE with local function and variable-length 2d array | |
| 6054 * [45]10262 (c++) cc1plus crashes with large generated code | |
| 6055 * [46]10278 (c++) ICE in parser for invalid code | |
| 6056 * [47]10446 (c++) ICE on definition of nonexistent member function of | |
| 6057 nested class in a class template | |
| 6058 * [48]10451 (c++) ICE in grokdeclarator on spurious mutable | |
| 6059 declaration | |
| 6060 * [49]10506 (c++) ICE in build_new at cp/init.c with | |
| 6061 -fkeep-inline-functions and multiple inheritance | |
| 6062 * [50]10549 (c++) ICE in store_bit_field on bitfields that exceed the | |
| 6063 precision of the declared type | |
| 6064 | |
| 6065 Optimization bugs | |
| 6066 | |
| 6067 * [51]2001 Inordinately long compile times in reload CSE regs | |
| 6068 * [52]2391 Exponential compilation time explosion in combine | |
| 6069 * [53]2960 Duplicate loop conditions even with -Os | |
| 6070 * [54]4046 redundant conditional branch | |
| 6071 * [55]6405 Loop-unrolling related performance regressions | |
| 6072 * [56]6798 very long compile time with large case-statement | |
| 6073 * [57]6871 const objects shouldn't be moved to .bss | |
| 6074 * [58]6909 problem w/ -Os on modified loop-2c.c test case | |
| 6075 * [59]7189 gcc -O2 -Wall does not print ``control reaches end of | |
| 6076 non-void function'' warning | |
| 6077 * [60]7642 optimization problem with signbit() | |
| 6078 * [61]8634 incorrect code for inlining of memcpy under -O2 | |
| 6079 * [62]8750 Cygwin prolog generation erroneously emitting __alloca as | |
| 6080 regular function call | |
| 6081 | |
| 6082 C front end | |
| 6083 | |
| 6084 * [63]2161 long if-else cascade overflows parser stack | |
| 6085 * [64]4319 short accepted on typedef'd char | |
| 6086 * [65]8602 incorrect line numbers in warning messages when using | |
| 6087 inline functions | |
| 6088 * [66]9177 -fdump-translation-unit: C front end deletes function_decl | |
| 6089 AST nodes and breaks debugging dumps | |
| 6090 * [67]9853 miscompilation of non-constant structure initializer | |
| 6091 | |
| 6092 c++ compiler and library | |
| 6093 | |
| 6094 * [68]45 legal template specialization code is rejected (DUP: | |
| 6095 [69]3784) | |
| 6096 * [70]764 lookup failure: friend operator and dereferencing a pointer | |
| 6097 and templates (DUP: [71]5116) | |
| 6098 * [72]2862 gcc accepts invalid explicit instantiation syntax (DUP: | |
| 6099 2863) | |
| 6100 * [73]3663 G++ doesn't check access control during template | |
| 6101 instantiation | |
| 6102 * [74]3797 gcc fails to emit explicit specialization of a template | |
| 6103 member | |
| 6104 * [75]3948 Two destructors are called when no copy destructor is | |
| 6105 defined (ABI change) | |
| 6106 * [76]4137 Conversion operator within template is not accepted | |
| 6107 * [77]4361 bogus ambiguity taking the address of a member template | |
| 6108 * [78]4802 g++ accepts illegal template code (access to private | |
| 6109 member; DUP: [79]5837) | |
| 6110 * [80]4803 inline function is used but never defined, and g++ does | |
| 6111 not object | |
| 6112 * [81]5094 Partial specialization cannot be friend? | |
| 6113 * [82]5730 complex<double>::norm() -- huge slowdown from egcs-2.91.66 | |
| 6114 * [83]6713 Regression wrt 3.0.4: g++ -O2 leads to seg fault at run | |
| 6115 time | |
| 6116 * [84]7015 certain __asm__ constructs rejected | |
| 6117 * [85]7086 compile time regression (quadratic behavior in | |
| 6118 fixup_var_refs) | |
| 6119 * [86]7099 G++ doesn't set the noreturn attribute on std::exit and | |
| 6120 std::abort | |
| 6121 * [87]7247 copy constructor missing when inlining enabled (invalid | |
| 6122 optimization?) | |
| 6123 * [88]7441 string array initialization compilation time regression | |
| 6124 from seconds to minutes | |
| 6125 * [89]7768 __PRETTY_FUNCTION__ for template destructor is wrong | |
| 6126 * [90]7804 bad printing of floating point constant in warning message | |
| 6127 * [91]8099 Friend classes and template specializations | |
| 6128 * [92]8117 member function pointers and multiple inheritance | |
| 6129 * [93]8205 using declaration and multiple inheritance | |
| 6130 * [94]8645 unnecessary non-zero checks in stl_tree.h | |
| 6131 * [95]8724 explicit destructor call for incomplete class allowed | |
| 6132 * [96]8805 compile time regression with many member variables | |
| 6133 * [97]8691 -O3 and -fno-implicit-templates are incompatible | |
| 6134 * [98]8700 unhelpful error message for binding temp to reference | |
| 6135 * [99]8724 explicit destructor call for incomplete class allowed | |
| 6136 * [100]8949 numeric_limits<>::denorm_min() and is_iec559 problems | |
| 6137 * [101]9016 Failure to consistently constant fold "constant" C++ | |
| 6138 objects | |
| 6139 * [102]9053 g++ confused about ambiguity of overloaded function | |
| 6140 templates | |
| 6141 * [103]9152 undefined virtual thunks | |
| 6142 * [104]9182 basic_filebuf<> does not report errors in codecvt<>::out | |
| 6143 * [105]9297 data corruption due to codegen bug (when copying.) | |
| 6144 * [106]9318 i/ostream::operator>>/<<(streambuf*) broken | |
| 6145 * [107]9320 Incorrect usage of traits_type::int_type in stdio_filebuf | |
| 6146 * [108]9400 bogus -Wshadow warning: shadowed declaration of this in | |
| 6147 local classes | |
| 6148 * [109]9424 i/ostream::operator>>/<<(streambuf*) drops characters | |
| 6149 * [110]9425 filebuf::pbackfail broken (DUP: [111]9439) | |
| 6150 * [112]9474 GCC freezes in compiling a weird code mixing <iostream> | |
| 6151 and <iostream.h> | |
| 6152 * [113]9548 Incorrect results from setf(ios::fixed) and precision(-1) | |
| 6153 [114][DR 231] | |
| 6154 * [115]9555 ostream inserters fail to set badbit on exception | |
| 6155 * [116]9561 ostream inserters rethrow exception of wrong type | |
| 6156 * [117]9563 ostream::sentry returns true after a failed preparation | |
| 6157 * [118]9582 one-definition rule violation in std::allocator | |
| 6158 * [119]9622 __PRETTY_FUNCTION__ incorrect in template destructors | |
| 6159 * [120]9683 bug in initialization chains for static const variables | |
| 6160 from template classes | |
| 6161 * [121]9791 -Woverloaded-virtual reports hiding of destructor | |
| 6162 * [122]9817 collate::compare doesn't handle nul characters | |
| 6163 * [123]9825 filebuf::sputbackc breaks sbumpc | |
| 6164 * [124]9826 operator>>(basic_istream, basic_string) fails to compile | |
| 6165 with custom traits | |
| 6166 * [125]9924 Multiple using statements for builtin functions not | |
| 6167 allowed | |
| 6168 * [126]9946 destructor is not called for temporary object | |
| 6169 * [127]9964 filebuf::close() sometimes fails to close file | |
| 6170 * [128]9988 filebuf::overflow writes EOF to file | |
| 6171 * [129]10033 optimization breaks polymorphic references w/ typeid | |
| 6172 operator | |
| 6173 * [130]10097 filebuf::underflow drops characters | |
| 6174 * [131]10132 filebuf destructor can throw exceptions | |
| 6175 * [132]10180 gcc fails to warn about non-inlined function | |
| 6176 * [133]10199 method parametrized by template does not work everywhere | |
| 6177 * [134]10300 use of array-new (nothrow) in segfaults on NULL return | |
| 6178 * [135]10427 Stack corruption with variable-length automatic arrays | |
| 6179 and virtual destructors | |
| 6180 * [136]10503 Compilation never stops in fixed_type_or_null | |
| 6181 | |
| 6182 Objective-C | |
| 6183 | |
| 6184 * [137]5956 selectors aren't matched properly when added to the | |
| 6185 selector table | |
| 6186 | |
| 6187 Fortran compiler and library | |
| 6188 | |
| 6189 * [138]1832 list directed i/o overflow hangs, -fbounds-check doesn't | |
| 6190 detect | |
| 6191 * [139]3924 g77 generates code that is rejected by GAS if COFF debug | |
| 6192 info requested | |
| 6193 * [140]5634 doc: explain that configure --prefix=~/... does not work | |
| 6194 * [141]6367 multiple repeat counts confuse namelist read into array | |
| 6195 * [142]6491 Logical operations error on logicals when using | |
| 6196 -fugly-logint | |
| 6197 * [143]6742 Generation of C++ Prototype for FORTRAN and extern "C" | |
| 6198 * [144]7113 Failure of g77.f-torture/execute/f90-intrinsic-bit.f -Os | |
| 6199 on irix6.5 | |
| 6200 * [145]7236 OPEN(...,RECL=nnn,...) without ACCESS='DIRECT' should | |
| 6201 assume a direct access file | |
| 6202 * [146]7278 g77 "bug"; the executable misbehaves (with -O2 | |
| 6203 -fno-automatic) | |
| 6204 * [147]7384 DATE_AND_TIME milliseconds field inactive on Windows | |
| 6205 * [148]7388 Incorrect output with 0-based array of characters | |
| 6206 * [149]8587 Double complex zero ** double precision number -> NaN | |
| 6207 instead of zero | |
| 6208 * [150]9038 -ffixed-line-length-none -x f77-cpp-input gives: Warning: | |
| 6209 unknown register name line-length-none | |
| 6210 * [151]10197 Direct access files not unformatted by default | |
| 6211 | |
| 6212 Java compiler and library | |
| 6213 | |
| 6214 * [152]6005 gcj fails to build rhug on alpha | |
| 6215 * [153]6389 System.getProperty("") should always throw an | |
| 6216 IllegalArgumentException | |
| 6217 * [154]6576 java.util.ResourceBundle.getResource ignores locale | |
| 6218 * [155]6652 new java.io.File("").getCanonicalFile() throws exception | |
| 6219 * [156]7060 getMethod() doesn't search super interface | |
| 6220 * [157]7073 bytecode interpreter gives wrong answer for interface | |
| 6221 getSuperclass() | |
| 6222 * [158]7180 possible bug in | |
| 6223 javax.naming.spi.NamingManager.getPlusPath() | |
| 6224 * [159]7416 java.security startup refs "GNU libgcj.security" | |
| 6225 * [160]7570 Runtime.exec with null envp: child doesn't inherit parent | |
| 6226 env (DUP: [161]7578) | |
| 6227 * [162]7611 Internal error while compiling libjava with -O | |
| 6228 * [163]7709 NullPointerException in _Jv_ResolvePoolEntry | |
| 6229 * [164]7766 ZipInputStream.available returns 0 immediately after | |
| 6230 construction | |
| 6231 * [165]7785 Calendar.getTimeInMillis/setTimeInMillis should be public | |
| 6232 * [166]7786 TimeZone.getDSTSavings() from JDK1.4 not implemented | |
| 6233 * [167]8142 '$' in class names vs. dlopen 'dynamic string tokens' | |
| 6234 * [168]8234 ZipInputStream chokes when InputStream.read() returns | |
| 6235 small chunks | |
| 6236 * [169]8415 reflection bug: exception info for Method | |
| 6237 * [170]8481 java.Random.nextInt(int) may return negative | |
| 6238 * [171]8593 Error reading GZIPped files with BufferedReader | |
| 6239 * [172]8759 java.beans.Introspector has no flushCaches() or | |
| 6240 flushFromCaches() methods | |
| 6241 * [173]8997 spin() calls Thread.sleep | |
| 6242 * [174]9253 on win32, java.io.File.listFiles("C:\\") returns pwd | |
| 6243 instead of the root content of C: | |
| 6244 * [175]9254 java::lang::Object::wait(), threads-win32.cc returns | |
| 6245 wrong return codes | |
| 6246 * [176]9271 Severe bias in java.security.SecureRandom | |
| 6247 | |
| 6248 Ada compiler and library | |
| 6249 | |
| 6250 * [177]6767 make gnatlib-shared fails on -laddr2line | |
| 6251 * [178]9911 gnatmake fails to link when GCC configured with | |
| 6252 --with-sjlj-exceptions=yes | |
| 6253 * [179]10020 Can't bootstrap gcc on AIX with Ada enabled | |
| 6254 * [180]10546 Ada tasking not working on Red Hat 9 | |
| 6255 | |
| 6256 preprocessor | |
| 6257 | |
| 6258 * [181]7029 preprocessor should ignore #warning with -M | |
| 6259 | |
| 6260 ARM-specific | |
| 6261 | |
| 6262 * [182]2903 [arm] Optimization bug with long long arithmetic | |
| 6263 * [183]7873 arm-linux-gcc fails when assigning address to a bit field | |
| 6264 | |
| 6265 FreeBSD-specific | |
| 6266 | |
| 6267 * [184]7680 float functions undefined in math.h/cmath with #define | |
| 6268 _XOPEN_SOURCE | |
| 6269 | |
| 6270 HP-UX or HP-PA-specific | |
| 6271 | |
| 6272 * [185]8705 [HP-PA] ICE in emit_move_insn_1, in expr.c | |
| 6273 * [186]9986 [HP-UX] Incorrect transformation of fputs_unlocked to | |
| 6274 fputc_unlocked | |
| 6275 * [187]10056 [HP-PA] ICE at -O2 when building c++ code from doxygen | |
| 6276 | |
| 6277 m68hc11-specific | |
| 6278 | |
| 6279 * [188]6744 Bad assembler code generated: reference to pseudo | |
| 6280 register z | |
| 6281 * [189]7361 Internal compiler error in reload_cse_simplify_operands, | |
| 6282 in reload1.c | |
| 6283 | |
| 6284 MIPS-specific | |
| 6285 | |
| 6286 * [190]9496 [mips-linux] bug in optimizer? | |
| 6287 | |
| 6288 PowerPC-specific | |
| 6289 | |
| 6290 * [191]7067 -Os with -mcpu=powerpc optimizes for speed (?) instead of | |
| 6291 space | |
| 6292 * [192]8480 reload ICEs for LAPACK code on powerpc64-linux | |
| 6293 * [193]8784 [AIX] Internal compiler error in simplify_gen_subreg | |
| 6294 * [194]10315 [powerpc] ICE: in extract_insn, in recog.c | |
| 6295 | |
| 6296 SPARC-specific | |
| 6297 | |
| 6298 * [195]10267 (documentation) Wrong build instructions for | |
| 6299 *-*-solaris2* | |
| 6300 | |
| 6301 x86-specific (Intel/AMD) | |
| 6302 | |
| 6303 * [196]7916 ICE in instantiate_virtual_register_1 | |
| 6304 * [197]7926 (c++) i486 instructions in header files make c++ programs | |
| 6305 crash on i386 | |
| 6306 * [198]8555 ICE in gen_split_1231 | |
| 6307 * [199]8994 ICE with -O -march=pentium4 | |
| 6308 * [200]9426 ICE with -fssa -funroll-loops -fprofile-arcs | |
| 6309 * [201]9806 ICE in inline assembly with -fPIC flag | |
| 6310 * [202]10077 gcc -msse2 generates movd to move dwords between xmm | |
| 6311 regs | |
| 6312 * [203]10233 64-bit comparison only comparing bottom 32-bits | |
| 6313 * [204]10286 type-punning doesn't work with __m64 and -O | |
| 6314 * [205]10308 [x86] ICE with -O -fgcse or -O2 | |
| 6315 __________________________________________________________________ | |
| 6316 | |
| 6317 GCC 3.3.1 | |
| 6318 | |
| 6319 Bug Fixes | |
| 6320 | |
| 6321 This section lists the problem reports (PRs) from GCC's bug tracking | |
| 6322 system that are known to be fixed in the 3.3.1 release. This list might | |
| 6323 not be complete (that is, it is possible that some PRs that have been | |
| 6324 fixed are not listed here). | |
| 6325 | |
| 6326 Bootstrap failures | |
| 6327 | |
| 6328 * [206]11272 [Solaris] make bootstrap fails while building libstdc++ | |
| 6329 | |
| 6330 Internal compiler errors (multi-platform) | |
| 6331 | |
| 6332 * [207]5754 ICE on invalid nested template class | |
| 6333 * [208]6597 ICE in set_mem_alias_set compiling Qt with -O2 on ia64 | |
| 6334 and --enable-checking | |
| 6335 * [209]6949 (c++) ICE in tsubst_decl, in cp/pt.c | |
| 6336 * [210]7053 (c++) ICE when declaring a function already defined as a | |
| 6337 friend method of a template class | |
| 6338 * [211]8164 (c++) ICE when using different const expressions as | |
| 6339 template parameter | |
| 6340 * [212]8384 (c++) ICE in is_base_type, in dwarf2out.c | |
| 6341 * [213]9559 (c++) ICE with invalid initialization of a static const | |
| 6342 * [214]9649 (c++) ICE in finish_member_declaration, in cp/semantics.c | |
| 6343 when redeclaring a static member variable | |
| 6344 * [215]9864 (fortran) ICE in add_abstract_origin_attribute, in | |
| 6345 dwarfout.c with -g -O -finline-functions | |
| 6346 * [216]10432 (c++) ICE in poplevel, in cp/decl.c | |
| 6347 * [217]10475 ICE in subreg_highpart_offset for code with long long | |
| 6348 * [218]10635 (c++) ICE when dereferencing an incomplete type casted | |
| 6349 from a void pointer | |
| 6350 * [219]10661 (c++) ICE in instantiate_decl, in cp/pt.c while | |
| 6351 instantiating static member variables | |
| 6352 * [220]10700 ICE in copy_to_mode_reg on 64-bit targets | |
| 6353 * [221]10712 (c++) ICE in constructor_name_full, in cp/decl2.c | |
| 6354 * [222]10796 (c++) ICE when defining an enum with two values: -1 and | |
| 6355 MAX_INT_64BIT | |
| 6356 * [223]10890 ICE in merge_assigned_reloads building Linux 2.4.2x | |
| 6357 sched.c | |
| 6358 * [224]10939 (c++) ICE with template code | |
| 6359 * [225]10956 (c++) ICE when specializing a template member function | |
| 6360 of a template class, in tsubst, in cp/pt.c | |
| 6361 * [226]11041 (c++) ICE: const myclass &x = *x; (when operator*() | |
| 6362 defined) | |
| 6363 * [227]11059 (c++) ICE with empty union | |
| 6364 * [228]11083 (c++) ICE in commit_one_edge_insertion, in cfgrtl.c with | |
| 6365 -O2 -fnon-call-exceptions | |
| 6366 * [229]11105 (c++) ICE in mangle_conv_op_name_for_type | |
| 6367 * [230]11149 (c++) ICE on error when instantiation with call function | |
| 6368 of a base type | |
| 6369 * [231]11228 (c++) ICE on new-expression using array operator new and | |
| 6370 default-initialization | |
| 6371 * [232]11282 (c++) Infinite memory usage after syntax error | |
| 6372 * [233]11301 (fortran) ICE with -fno-globals | |
| 6373 * [234]11308 (c++) ICE when using an enum type name as if it were a | |
| 6374 class or namespace | |
| 6375 * [235]11473 (c++) ICE with -gstabs when empty struct inherits from | |
| 6376 an empty struct | |
| 6377 * [236]11503 (c++) ICE when instantiating template with ADDR_EXPR | |
| 6378 * [237]11513 (c++) ICE in push_template_decl_real, in cp/pt.c: | |
| 6379 template member functions | |
| 6380 | |
| 6381 Optimization bugs | |
| 6382 | |
| 6383 * [238]11198 -O2 -frename-registers generates wrong code (aliasing | |
| 6384 problem) | |
| 6385 * [239]11304 Wrong code production with -fomit-frame-pointer | |
| 6386 * [240]11381 volatile memory access optimized away | |
| 6387 * [241]11536 [strength-reduce] -O2 optimization produces wrong code | |
| 6388 * [242]11557 constant folding bug generates wrong code | |
| 6389 | |
| 6390 C front end | |
| 6391 | |
| 6392 * [243]5897 No warning for statement after return | |
| 6393 * [244]11279 DWARF-2 output mishandles large enums | |
| 6394 | |
| 6395 Preprocessor bugs | |
| 6396 | |
| 6397 * [245]11022 no warning for non-compatible macro redefinition | |
| 6398 | |
| 6399 C++ compiler and library | |
| 6400 | |
| 6401 * [246]2330 static_cast<>() to a private base is allowed | |
| 6402 * [247]5388 Incorrect message "operands to ?: have different types" | |
| 6403 * [248]5390 Libiberty fails to demangle multi-digit template | |
| 6404 parameters | |
| 6405 * [249]7877 Incorrect parameter passing to specializations of member | |
| 6406 function templates | |
| 6407 * [250]9393 Anonymous namespaces and compiling the same file twice | |
| 6408 * [251]10032 -pedantic converts some errors to warnings | |
| 6409 * [252]10468 const typeof(x) is non-const, but only in templates | |
| 6410 * [253]10527 confused error message with "new int()" parameter | |
| 6411 initializer | |
| 6412 * [254]10679 parameter MIN_INLINE_INSNS is not honored | |
| 6413 * [255]10682 gcc chokes on a typedef for an enum inside a class | |
| 6414 template | |
| 6415 * [256]10689 pow(std::complex(0),1/3) returns (nan, nan) instead of | |
| 6416 0. | |
| 6417 * [257]10845 template member function (with nested template as | |
| 6418 parameter) cannot be called anymore if another unrelated template | |
| 6419 member function is defined | |
| 6420 * [258]10849 Cannot define an out-of-class specialization of a | |
| 6421 private nested template class | |
| 6422 * [259]10888 Suppress -Winline warnings for system headers | |
| 6423 * [260]10929 -Winline warns about functions for which no definition | |
| 6424 is visible | |
| 6425 * [261]10931 valid conversion static_cast<const unsigned | |
| 6426 int&>(lvalue-of-type-int) is rejected | |
| 6427 * [262]10940 Bad code with explicit specialization | |
| 6428 * [263]10968 If member function implicitly instantiated, explicit | |
| 6429 instantiation of class fails to instantiate it | |
| 6430 * [264]10990 Cannot convert with dynamic_cast<> to a private base | |
| 6431 class from within a member function | |
| 6432 * [265]11039 Bad interaction between implicit typename deprecation | |
| 6433 and friendship | |
| 6434 * [266]11062 (libstdc++) avoid __attribute__ ((unused)); say | |
| 6435 "__unused__" instead | |
| 6436 * [267]11095 C++ iostream manipulator causes segfault when called | |
| 6437 with negative argument | |
| 6438 * [268]11098 g++ doesn't emit complete debugging information for | |
| 6439 local variables in destructors | |
| 6440 * [269]11137 Linux shared library constructors not called unless | |
| 6441 there's one global object | |
| 6442 * [270]11154 spurious ambiguity report for template class | |
| 6443 specialization | |
| 6444 * [271]11329 Compiler cannot find user defined implicit typecast | |
| 6445 * [272]11332 Spurious error with casts in ?: expression | |
| 6446 * [273]11431 static_cast behavior with subclasses when default | |
| 6447 constructor available | |
| 6448 * [274]11528 money_get facet does not accept "$.00" as valid | |
| 6449 * [275]11546 Type lookup problems in out-of-line definition of a | |
| 6450 class doubly nested from a template class | |
| 6451 * [276]11567 C++ code containing templated member function with same | |
| 6452 name as pure virtual member function results in linking failure | |
| 6453 * [277]11645 Failure to deal with using and private inheritance | |
| 6454 | |
| 6455 Java compiler and library | |
| 6456 | |
| 6457 * [278]5179 Qualified static field access doesn't initialize its | |
| 6458 class | |
| 6459 * [279]8204 gcj -O2 to native reorders certain instructions | |
| 6460 improperly | |
| 6461 * [280]10838 java.io.ObjectInputStream syntax error | |
| 6462 * [281]10886 The RMI registry that comes with GCJ does not work | |
| 6463 correctly | |
| 6464 * [282]11349 JNDI URL context factories not located correctly | |
| 6465 | |
| 6466 x86-specific (Intel/AMD) | |
| 6467 | |
| 6468 * [283]4823 ICE on inline assembly code | |
| 6469 * [284]8878 miscompilation with -O and SSE | |
| 6470 * [285]9815 (c++ library) atomicity.h - fails to compile with -O3 | |
| 6471 -masm=intel | |
| 6472 * [286]10402 (inline assembly) [x86] ICE in merge_assigned_reloads, | |
| 6473 in reload1.c | |
| 6474 * [287]10504 ICE with SSE2 code and -O3 -mcpu=pentium4 -msse2 | |
| 6475 * [288]10673 ICE for x86-64 on freebsd libc vfprintf.c source | |
| 6476 * [289]11044 [x86] out of range loop instructions for FP code on K6 | |
| 6477 * [290]11089 ICE: instantiate_virtual_regs_lossage while using SSE | |
| 6478 built-ins | |
| 6479 * [291]11420 [x86_64] gcc generates invalid asm code when "-O -fPIC" | |
| 6480 is used | |
| 6481 | |
| 6482 SPARC- or Solaris- specific | |
| 6483 | |
| 6484 * [292]9362 solaris 'as' dies when fed .s and "-gstabs" | |
| 6485 * [293]10142 [SPARC64] gcc produces wrong code when passing | |
| 6486 structures by value | |
| 6487 * [294]10663 New configure check aborts with Sun tools. | |
| 6488 * [295]10835 combinatorial explosion in scheduler on HyperSPARC | |
| 6489 * [296]10876 ICE in calculate_giv_inc when building KDE | |
| 6490 * [297]10955 wrong code at -O3 for structure argument in context of | |
| 6491 structure return | |
| 6492 * [298]11018 -mcpu=ultrasparc busts tar-1.13.25 | |
| 6493 * [299]11556 [sparc64] ICE in gen_reg_rtx() while compiling 2.6.x | |
| 6494 Linux kernel | |
| 6495 | |
| 6496 ia64 specific | |
| 6497 | |
| 6498 * [300]10907 gcc violates the ia64 ABI (GP must be preserved) | |
| 6499 * [301]11320 scheduler bug (in machine depended reorganization pass) | |
| 6500 * [302]11599 bug with conditional and __builtin_prefetch | |
| 6501 | |
| 6502 PowerPC specific | |
| 6503 | |
| 6504 * [303]9745 [powerpc] gcc mis-compiles libmcrypt (alias problem | |
| 6505 during loop) | |
| 6506 * [304]10871 error in rs6000_stack_info save_size computation | |
| 6507 * [305]11440 gcc mis-compiles c++ code (libkhtml) with -O2, -fno-gcse | |
| 6508 cures it | |
| 6509 | |
| 6510 m68k-specific | |
| 6511 | |
| 6512 * [306]7594 [m68k] ICE on legal code associated with simplify-rtx | |
| 6513 * [307]10557 [m68k] ICE in subreg_offset_representable_p | |
| 6514 * [308]11054 [m68k] ICE in reg_overlap_mentioned_p | |
| 6515 | |
| 6516 ARM-specific | |
| 6517 | |
| 6518 * [309]10834 [arm] GCC 3.3 still generates incorrect instructions for | |
| 6519 functions with __attribute__ ((interrupt ("IRQ"))) | |
| 6520 * [310]10842 [arm] Clobbered link register is copied to pc under | |
| 6521 certain circumstances | |
| 6522 * [311]11052 [arm] noce_process_if_block() can lose REG_INC notes | |
| 6523 * [312]11183 [arm] ICE in change_address_1 (3.3) / subreg_hard_regno | |
| 6524 (3.4) | |
| 6525 | |
| 6526 MIPS-specific | |
| 6527 | |
| 6528 * [313]11084 ICE in propagate_one_insn, in flow.c | |
| 6529 | |
| 6530 SH-specific | |
| 6531 | |
| 6532 * [314]10331 can't compile c++ part of gcc cross compiler for sh-elf | |
| 6533 * [315]10413 [SH] ICE in reload_cse_simplify_operands, in reload1.c | |
| 6534 * [316]11096 i686-linux to sh-linux cross compiler fails to compile | |
| 6535 C++ files | |
| 6536 | |
| 6537 GNU/Linux (or Hurd?) specific | |
| 6538 | |
| 6539 * [317]2873 Bogus fixinclude of stdio.h from glibc 2.2.3 | |
| 6540 | |
| 6541 UnixWare specific | |
| 6542 | |
| 6543 * [318]3163 configure bug: gcc/aclocal.m4 mmap test fails on UnixWare | |
| 6544 7.1.1 | |
| 6545 | |
| 6546 Cygwin (or mingw) specific | |
| 6547 | |
| 6548 * [319]5287 ICE with dllimport attribute | |
| 6549 * [320]10148 [MingW/CygWin] Compiler dumps core | |
| 6550 | |
| 6551 DJGPP specific | |
| 6552 | |
| 6553 * [321]8787 GCC fails to emit .intel_syntax when invoked with | |
| 6554 -masm=intel on DJGPP | |
| 6555 | |
| 6556 Darwin (and MacOS X) specific | |
| 6557 | |
| 6558 * [322]10900 trampolines crash | |
| 6559 | |
| 6560 Documentation | |
| 6561 | |
| 6562 * [323]1607 (c++) Format attributes on methods undocumented | |
| 6563 * [324]4252 Invalid option `-fdump-translation-unit' | |
| 6564 * [325]4490 Clarify restrictions on -m96bit-long-double, | |
| 6565 -m128bit-long-double | |
| 6566 * [326]10355 document an issue with regparm attribute on some systems | |
| 6567 (e.g. Solaris) | |
| 6568 * [327]10726 (fortran) Documentation for function "IDate Intrinsic | |
| 6569 (Unix)" is wrong | |
| 6570 * [328]10805 document bug in old version of Sun assembler | |
| 6571 * [329]10815 warn against GNU binutils on AIX | |
| 6572 * [330]10877 document need for newer binutils on i?86-*-linux-gnu | |
| 6573 * [331]11280 Manual incorrect with respect to -freorder-blocks | |
| 6574 * [332]11466 Document -mlittle-endian and its restrictions for the | |
| 6575 sparc64 port | |
| 6576 | |
| 6577 Testsuite bugs (compiler itself is not affected) | |
| 6578 | |
| 6579 * [333]10737 newer bison causes g++.dg/parse/crash2.C to incorrectly | |
| 6580 report failure | |
| 6581 * [334]10810 gcc-3.3 fails make check: buffer overrun in | |
| 6582 test_demangle.c | |
| 6583 __________________________________________________________________ | |
| 6584 | |
| 6585 GCC 3.3.2 | |
| 6586 | |
| 6587 Bug Fixes | |
| 6588 | |
| 6589 This section lists the problem reports (PRs) from [335]GCC's bug | |
| 6590 tracking system that are known to be fixed in the 3.3.2 release. This | |
| 6591 list might not be complete (that is, it is possible that some PRs that | |
| 6592 have been fixed are not listed here). | |
| 6593 | |
| 6594 Bootstrap failures and problems | |
| 6595 | |
| 6596 * [336]8336 [SCO5] bootstrap config still tries to use COFF options | |
| 6597 * [337]9330 [alpha-osf] Bootstrap failure on Compaq Tru64 with | |
| 6598 --enable-threads=posix | |
| 6599 * [338]9631 [hppa64-linux] gcc-3.3 fails to bootstrap | |
| 6600 * [339]9877 fixincludes makes a bad sys/byteorder.h on svr5 (UnixWare | |
| 6601 7.1.1) | |
| 6602 * [340]11687 xstormy16-elf build fails in libf2c | |
| 6603 * [341]12263 [SGI IRIX] bootstrap fails during compile of | |
| 6604 libf2c/libI77/backspace.c | |
| 6605 * [342]12490 buffer overflow in scan-decls.c (during Solaris 9 | |
| 6606 fix-header processing) | |
| 6607 | |
| 6608 Internal compiler errors (multi-platform) | |
| 6609 | |
| 6610 * [343]7277 Casting integers to vector types causes ICE | |
| 6611 * [344]7939 (c++) ICE on invalid function template specialization | |
| 6612 * [345]11063 (c++) ICE on parsing initialization list of const array | |
| 6613 member | |
| 6614 * [346]11207 ICE with negative index in array element designator | |
| 6615 * [347]11522 (fortran) g77 dwarf-2 ICE in | |
| 6616 add_abstract_origin_attribute | |
| 6617 * [348]11595 (c++) ICE on duplicate label definition | |
| 6618 * [349]11646 (c++) ICE in commit_one_edge_insertion with | |
| 6619 -fnon-call-exceptions -fgcse -O | |
| 6620 * [350]11665 ICE in struct initializer when taking address | |
| 6621 * [351]11852 (c++) ICE with bad struct initializer. | |
| 6622 * [352]11878 (c++) ICE in cp_expr_size | |
| 6623 * [353]11883 ICE with any -O on mercury-generated C code | |
| 6624 * [354]11991 (c++) ICE in cxx_incomplete_type_diagnostic, in | |
| 6625 cp/typeck2.c when applying typeid operator to template template | |
| 6626 parameter | |
| 6627 * [355]12146 ICE in lookup_template_function, in cp/pt.c | |
| 6628 * [356]12215 ICE in make_label_edge with -fnon-call-exceptions | |
| 6629 -fno-gcse -O2 | |
| 6630 * [357]12369 (c++) ICE with templates and friends | |
| 6631 * [358]12446 ICE in emit_move_insn on complicated array reference | |
| 6632 * [359]12510 ICE in final_scan_insn | |
| 6633 * [360]12544 ICE with large parameters used in nested functions | |
| 6634 | |
| 6635 C and optimization bugs | |
| 6636 | |
| 6637 * [361]9862 spurious warnings with -W -finline-functions | |
| 6638 * [362]10962 lookup_field is a linear search on a linked list (can be | |
| 6639 slow if large struct) | |
| 6640 * [363]11370 -Wunreachable-code gives false complaints | |
| 6641 * [364]11637 invalid assembly with -fnon-call-exceptions | |
| 6642 * [365]11885 Problem with bitfields in packed structs | |
| 6643 * [366]12082 Inappropriate unreachable code warnings | |
| 6644 * [367]12180 Inline optimization fails for variadic function | |
| 6645 * [368]12340 loop unroller + gcse produces wrong code | |
| 6646 | |
| 6647 C++ compiler and library | |
| 6648 | |
| 6649 * [369]3907 nested template parameter collides with member name | |
| 6650 * [370]5293 confusing message when binding a temporary to a reference | |
| 6651 * [371]5296 [DR115] Pointers to functions and to template functions | |
| 6652 behave differently in deduction | |
| 6653 * [372]7939 ICE on function template specialization | |
| 6654 * [373]8656 Unable to assign function with __attribute__ and pointer | |
| 6655 return type to an appropriate variable | |
| 6656 * [374]10147 Confusing error message for invalid template function | |
| 6657 argument | |
| 6658 * [375]11400 std::search_n() makes assumptions about Size parameter | |
| 6659 * [376]11409 issues with using declarations, overloading, and | |
| 6660 built-in functions | |
| 6661 * [377]11740 ctype<wchar_t>::do_is(mask, wchar_t) doesn't handle | |
| 6662 multiple bits in mask | |
| 6663 * [378]11786 operator() call on variable in other namespace not | |
| 6664 recognized | |
| 6665 * [379]11867 static_cast ignores ambiguity | |
| 6666 * [380]11928 bug with conversion operators that are typedefs | |
| 6667 * [381]12114 Uninitialized memory accessed in dtor | |
| 6668 * [382]12163 static_cast + explicit constructor regression | |
| 6669 * [383]12181 Wrong code with comma operator and c++ | |
| 6670 * [384]12236 regparm and fastcall messes up parameters | |
| 6671 * [385]12266 incorrect instantiation of unneeded template during | |
| 6672 overload resolution | |
| 6673 * [386]12296 istream::peek() doesn't set eofbit | |
| 6674 * [387]12298 [sjlj exceptions] Stack unwind destroys | |
| 6675 not-yet-constructed object | |
| 6676 * [388]12369 ICE with templates and friends | |
| 6677 * [389]12337 apparently infinite loop in g++ | |
| 6678 * [390]12344 stdcall attribute ignored if function returns a pointer | |
| 6679 * [391]12451 missing(late) class forward declaration in cxxabi.h | |
| 6680 * [392]12486 g++ accepts invalid use of a qualified name | |
| 6681 | |
| 6682 x86 specific (Intel/AMD) | |
| 6683 | |
| 6684 * [393]8869 [x86 MMX] ICE with const variable optimization and MMX | |
| 6685 builtins | |
| 6686 * [394]9786 ICE in fixup_abnormal_edges with -fnon-call-exceptions | |
| 6687 -O2 | |
| 6688 * [395]11689 g++3.3 emits un-assembleable code for k6 architecture | |
| 6689 * [396]12116 [k6] Invalid assembly output values with X-MAME code | |
| 6690 * [397]12070 ICE converting between double and long double with | |
| 6691 -msoft-float | |
| 6692 | |
| 6693 ia64-specific | |
| 6694 | |
| 6695 * [398]11184 [ia64 hpux] ICE on __builtin_apply building libobjc | |
| 6696 * [399]11535 __builtin_return_address may not work on ia64 | |
| 6697 * [400]11693 [ia64] ICE in gen_nop_type | |
| 6698 * [401]12224 [ia64] Thread-local storage doesn't work | |
| 6699 | |
| 6700 PowerPC-specific | |
| 6701 | |
| 6702 * [402]11087 [powerpc64-linux] GCC miscompiles raid1.c from linux | |
| 6703 kernel | |
| 6704 * [403]11319 loop miscompiled on ppc32 | |
| 6705 * [404]11949 ICE Compiler segfault with ffmpeg -maltivec code | |
| 6706 | |
| 6707 SPARC-specific | |
| 6708 | |
| 6709 * [405]11662 wrong code for expr. with cast to long long and | |
| 6710 exclusive or | |
| 6711 * [406]11965 invalid assembler code for a shift < 32 operation | |
| 6712 * [407]12301 (c++) stack corruption when a returned expression throws | |
| 6713 an exception | |
| 6714 | |
| 6715 Alpha-specific | |
| 6716 | |
| 6717 * [408]11717 [alpha-linux] unrecognizable insn compiling for.c of | |
| 6718 kernel 2.4.22-pre8 | |
| 6719 | |
| 6720 HPUX-specific | |
| 6721 | |
| 6722 * [409]11313 problem with #pragma weak and static inline functions | |
| 6723 * [410]11712 __STDC_EXT__ not defined for C++ by default anymore? | |
| 6724 | |
| 6725 Solaris specific | |
| 6726 | |
| 6727 * [411]12166 Profiled programs crash if PROFDIR is set | |
| 6728 | |
| 6729 Solaris-x86 specific | |
| 6730 | |
| 6731 * [412]12101 i386 Solaris no longer works with GNU as? | |
| 6732 | |
| 6733 Miscellaneous embedded target-specific bugs | |
| 6734 | |
| 6735 * [413]10988 [m32r-elf] wrong blockmove code with -O3 | |
| 6736 * [414]11805 [h8300-unknown-coff] [H8300] ICE for simple code with | |
| 6737 -O2 | |
| 6738 * [415]11902 [sh4] spec file improperly inserts rpath even when none | |
| 6739 needed | |
| 6740 * [416]11903 [sh4] -pthread fails to link due to error in spec file | |
| 6741 on sh4 | |
| 6742 __________________________________________________________________ | |
| 6743 | |
| 6744 GCC 3.3.3 | |
| 6745 | |
| 6746 Minor features | |
| 6747 | |
| 6748 In addition to the bug fixes documented below, this release contains | |
| 6749 few minor features such as: | |
| 6750 * Support for --with-sysroot | |
| 6751 * Support for automatic detection of executable stacks | |
| 6752 * Support for SSE3 instructions | |
| 6753 * Support for thread local storage debugging under GDB on S390 | |
| 6754 | |
| 6755 Bug Fixes | |
| 6756 | |
| 6757 This section lists the problem reports (PRs) from [417]GCC's bug | |
| 6758 tracking system that are known to be fixed in the 3.3.3 release. This | |
| 6759 list might not be complete (that is, it is possible that some PRs that | |
| 6760 have been fixed are not listed here). | |
| 6761 | |
| 6762 Bootstrap failures and issues | |
| 6763 | |
| 6764 * [418]11890 Building cross gcc-3.3.1 for sparc-sun-solaris2.6 fails | |
| 6765 * [419]12399 boehm-gc fails (when building a cross compiler): libtool | |
| 6766 unable to infer tagged configuration | |
| 6767 * [420]13068 mklibgcc.in doesn't handle multi-level multilib | |
| 6768 subdirectories properly | |
| 6769 | |
| 6770 Internal compiler errors (multi-platform) | |
| 6771 | |
| 6772 * [421]10060 ICE (stack overflow) on huge file (300k lines) due to | |
| 6773 recursive behaviour of copy_rtx_if_shared, in emit_rtl.c | |
| 6774 * [422]10555 (c++) ICE on undefined template argument | |
| 6775 * [423]10706 (c++) ICE in mangle_class_name_for_template | |
| 6776 * [424]11496 (fortran) error in flow_loops_find when -funroll-loops | |
| 6777 active | |
| 6778 * [425]11741 ICE in pre_insert_copy_insn, in gcse.c | |
| 6779 * [426]12440 GCC crashes during compilation of quicktime4linux 2.0.0 | |
| 6780 * [427]12632 (fortran) -fbounds-check ICE | |
| 6781 * [428]12712 (c++) ICE on short legit C++ code fragment with gcc | |
| 6782 3.3.2 | |
| 6783 * [429]12726 (c++) ICE (segfault) on trivial code | |
| 6784 * [430]12890 (c++) ICE on compilation of class with throwing method | |
| 6785 * [431]12900 (c++) ICE in rtl_verify_flow_info_1 | |
| 6786 * [432]13060 (fortran) ICE in fixup_var_refs_1, in function.c on | |
| 6787 correct code with -O2 -fno-force-mem | |
| 6788 * [433]13289 (c++) ICE in regenerate_decl_from_template on recursive | |
| 6789 template | |
| 6790 * [434]13318 ICE: floating point exception in the loop optimizer | |
| 6791 * [435]13392 (c++) ICE in convert_from_eh_region_ranges_1, in | |
| 6792 except.c | |
| 6793 * [436]13574 (c++) invalid array default initializer in class lets | |
| 6794 gcc consume all memory and die | |
| 6795 * [437]13475 ICE on SIMD variables with partial value initialization | |
| 6796 * [438]13797 (c++) ICE on invalid template parameter | |
| 6797 * [439]13824 (java) gcj SEGV with simple .java program | |
| 6798 | |
| 6799 C and optimization bugs | |
| 6800 | |
| 6801 * [440]8776 loop invariants are not removed (most likely) | |
| 6802 * [441]10339 [sparc,ppc,ppc64] Invalid optimization: replacing | |
| 6803 strncmp by memcmp | |
| 6804 * [442]11350 undefined labels with -Os -fPIC | |
| 6805 * [443]12826 Optimizer removes reference through volatile pointer | |
| 6806 * [444]12500 stabs debug info: void no longer a predefined / builtin | |
| 6807 type | |
| 6808 * [445]12941 builtin-bitops-1.c miscompilation (latent bug) | |
| 6809 * [446]12953 tree inliner bug (in inline_forbidden_p) and fix | |
| 6810 * [447]13041 linux-2.6/sound/core/oss/rate.c miscompiled | |
| 6811 * [448]13507 spurious printf format warning | |
| 6812 * [449]13382 Type information for const pointer disappears during | |
| 6813 optimization. | |
| 6814 * [450]13394 noreturn attribute ignored on recursive invokation | |
| 6815 * [451]13400 Compiled code crashes storing to read-only location | |
| 6816 * [452]13521 Endless loop in calculate_global_regs_live | |
| 6817 | |
| 6818 C++ compiler and library | |
| 6819 | |
| 6820 Some of the bug fixes in this list were made to implement decisions | |
| 6821 that the ISO C++ standards committee has made concerning several defect | |
| 6822 reports (DRs). Links in the list below point to detailed discussion of | |
| 6823 the relevant defect report. | |
| 6824 * [453]2094 unimplemented: use of `ptrmem_cst' in template type | |
| 6825 unification | |
| 6826 * [454]2294 using declaration confusion | |
| 6827 * [455]5050 template instantiation depth exceeds limit: recursion | |
| 6828 problem? | |
| 6829 * [456]9371 Bad exception handling in | |
| 6830 i/ostream::operator>>/<<(streambuf*) | |
| 6831 * [457]9546 bad exception handling in ostream members | |
| 6832 * [458]10081 basic_ios::_M_cache_locale leaves NULL members in the | |
| 6833 face of unknown locales | |
| 6834 * [459]10093 [460][DR 61] Setting failbit in exceptions doesn't work | |
| 6835 * [461]10095 istream::operator>>(int&) sets ios::badbit when | |
| 6836 ios::failbit is set. | |
| 6837 * [462]11554 Warning about reordering of initializers doesn't mention | |
| 6838 location of constructor | |
| 6839 * [463]12297 istream::sentry::sentry() handles eof() incorrectly. | |
| 6840 * [464]12352 Exception safety problems in src/localename.cc | |
| 6841 * [465]12438 Memory leak in locale::combine() | |
| 6842 * [466]12540 Memory leak in locale::locale(const char*) | |
| 6843 * [467]12594 DRs [468]60 [TC] and [469]63 [TC] not implemented | |
| 6844 * [470]12657 Resolution of [471]DR 292 (WP) still unimplemented | |
| 6845 * [472]12696 memory eating infinite loop in diagnostics (error | |
| 6846 recovery problem) | |
| 6847 * [473]12815 Code compiled with optimization behaves unexpectedly | |
| 6848 * [474]12862 Conflicts between typedefs/enums and namespace member | |
| 6849 declarations | |
| 6850 * [475]12926 Wrong value after assignment in initialize list using | |
| 6851 bit-fields | |
| 6852 * [476]12967 Resolution of [477]DR 300 [WP] still unimplemented | |
| 6853 * [478]12971 Resolution of [479]DR 328 [WP] still unimplemented | |
| 6854 * [480]13007 basic_streambuf::pubimbue, imbue wrong | |
| 6855 * [481]13009 Implicitly-defined assignment operator writes to wrong | |
| 6856 memory | |
| 6857 * [482]13057 regparm attribute not applied to destructor | |
| 6858 * [483]13070 -Wformat option ignored in g++ | |
| 6859 * [484]13081 forward template declarations in <complex> let inlining | |
| 6860 fail | |
| 6861 * [485]13239 Assertion does not seem to work correctly anymore | |
| 6862 * [486]13262 "xxx is private within this context" when initializing a | |
| 6863 self-contained template class | |
| 6864 * [487]13290 simple typo in concept checking for std::generate_n | |
| 6865 * [488]13323 Template code does not compile in presence of typedef | |
| 6866 * [489]13369 __verify_grouping (and __add_grouping?) not correct | |
| 6867 * [490]13371 infinite loop with packed struct and inlining | |
| 6868 * [491]13445 Template argument replacement "dereferences" a typedef | |
| 6869 * [492]13461 Fails to access protected-ctor from public constant | |
| 6870 * [493]13462 Non-standard-conforming type set::pointer | |
| 6871 * [494]13478 gcc uses wrong constructor to initialize a const | |
| 6872 reference | |
| 6873 * [495]13544 "conflicting types" for enums in different scopes | |
| 6874 * [496]13650 string::compare should not (always) use | |
| 6875 traits_type::length() | |
| 6876 * [497]13683 bogus warning about passing non-PODs through ellipsis | |
| 6877 * [498]13688 Derived class is denied access to protected base class | |
| 6878 member class | |
| 6879 * [499]13774 Member variable cleared in virtual multiple inheritance | |
| 6880 class | |
| 6881 * [500]13884 Protect sstream.tcc from extern template use | |
| 6882 | |
| 6883 Java compiler and library | |
| 6884 | |
| 6885 * [501]10746 [win32] garbage collection crash in GCJ | |
| 6886 | |
| 6887 Objective-C compiler and library | |
| 6888 | |
| 6889 * [502]11433 Crash due to dereferencing null pointer when querying | |
| 6890 protocol | |
| 6891 | |
| 6892 Fortran compiler and library | |
| 6893 | |
| 6894 * [503]12633 logical expression gives incorrect result with | |
| 6895 -fugly-logint option | |
| 6896 * [504]13037 [gcse-lm] g77 generates incorrect code | |
| 6897 * [505]13213 Hex constant problem when compiling with -fugly-logint | |
| 6898 and -ftypeless-boz | |
| 6899 | |
| 6900 x86-specific (Intel/AMD) | |
| 6901 | |
| 6902 * [506]4490 ICE with -m128bit-long-double | |
| 6903 * [507]12292 [x86_64] ICE: RTL check: expected code `const_int', have | |
| 6904 `reg' in make_field_assignment, in combine.c | |
| 6905 * [508]12441 ICE: can't find a register to spill | |
| 6906 * [509]12943 array static-init failure under -fpic, -fPIC | |
| 6907 * [510]13608 Incorrect code with -O3 -ffast-math | |
| 6908 | |
| 6909 PowerPC-specific | |
| 6910 | |
| 6911 * [511]11598 testcase gcc.dg/20020118-1.c fails runtime check of | |
| 6912 __attribute__((aligned(16))) | |
| 6913 * [512]11793 ICE in extract_insn, in recog.c (const_vector's) | |
| 6914 * [513]12467 vmsumubm emitted when vmsummbm appropriate (typo in | |
| 6915 altivec.md) | |
| 6916 * [514]12537 g++ generates writeable text sections | |
| 6917 | |
| 6918 SPARC-specific | |
| 6919 | |
| 6920 * [515]12496 wrong result for __atomic_add(&value, -1) when using -O0 | |
| 6921 -m64 | |
| 6922 * [516]12865 mprotect call to make trampoline executable may fail | |
| 6923 * [517]13354 ICE in sparc_emit_set_const32 | |
| 6924 | |
| 6925 ARM-specific | |
| 6926 | |
| 6927 * [518]10467 [arm] ICE in pre_insert_copy_insn, | |
| 6928 | |
| 6929 ia64-specific | |
| 6930 | |
| 6931 * [519]11226 ICE passing struct arg with two floats | |
| 6932 * [520]11227 ICE for _Complex float, _Complex long double args | |
| 6933 * [521]12644 GCC 3.3.2 fails to compile glibc on ia64 | |
| 6934 * [522]13149 build gcc-3.3.2 1305 error:unrecognizable insn | |
| 6935 * Various fixes for libunwind | |
| 6936 | |
| 6937 Alpha-specific | |
| 6938 | |
| 6939 * [523]12654 Incorrect comparison code generated for Alpha | |
| 6940 * [524]12965 SEGV+ICE in cc1plus on alpha-linux with -O2 | |
| 6941 * [525]13031 ICE (unrecognizable insn) when building gnome-libs-1.4.2 | |
| 6942 | |
| 6943 HPPA-specific | |
| 6944 | |
| 6945 * [526]11634 [hppa] ICE in verify_local_live_at_start, in flow.c | |
| 6946 * [527]12158 [hppa] compilation does not terminate at -O1 | |
| 6947 | |
| 6948 S390-specific | |
| 6949 | |
| 6950 * [528]11992 Wrong built-in code for memcmp with length 1<<24: only | |
| 6951 (1<<24)-1 possible for CLCL-Instruction | |
| 6952 | |
| 6953 SH-specific | |
| 6954 | |
| 6955 * [529]9365 segfault in gen_far_branch (config/sh/sh.c) | |
| 6956 * [530]10392 optimizer generates faulty array indexing | |
| 6957 * [531]11322 SH profiler outputs multiple definitions of symbol | |
| 6958 * [532]13069 gcc/config/sh/rtems.h broken | |
| 6959 * [533]13302 Putting a va_list in a struct causes seg fault | |
| 6960 * [534]13585 Incorrect optimization of call to sfunc | |
| 6961 * Fix inappropriately exported libgcc functions from the shared | |
| 6962 library | |
| 6963 | |
| 6964 Other embedded target specific | |
| 6965 | |
| 6966 * [535]8916 [mcore] unsigned char assign gets hosed. | |
| 6967 * [536]11576 [h8300] ICE in change_address_1, in emit-rtl.c | |
| 6968 * [537]13122 [h8300] local variable gets corrupted by function call | |
| 6969 when -fomit-frame-pointer is given | |
| 6970 * [538]13256 [cris] strict_low_part mistreated in delay slots | |
| 6971 * [539]13373 [mcore] optimization with -frerun-cse-after-loop | |
| 6972 -fexpensive-optimizations produces wrong code on mcore | |
| 6973 | |
| 6974 GNU HURD-specific | |
| 6975 | |
| 6976 * [540]12561 gcc/config/t-gnu needs updating to work with | |
| 6977 --with-sysroot | |
| 6978 | |
| 6979 Tru64 Unix specific | |
| 6980 | |
| 6981 * [541]6243 testsuite fails almost all tests due to no libintl in | |
| 6982 LD_LIBRARY_PATH during test. | |
| 6983 * [542]11397 weak aliases broken on Tru64 UNIX | |
| 6984 | |
| 6985 AIX-specific | |
| 6986 | |
| 6987 * [543]12505 build failure due to defines of uchar in cpphash.h and | |
| 6988 sys/types.h | |
| 6989 * [544]13150 WEAK symbols not exported by collect2 | |
| 6990 | |
| 6991 IRIX-specific | |
| 6992 | |
| 6993 * [545]12666 fixincludes problem on IRIX 6.5.19m | |
| 6994 | |
| 6995 Solaris-specific | |
| 6996 | |
| 6997 * [546]12969 Including sys/byteorder.h breaks configure checks | |
| 6998 | |
| 6999 Testsuite problems (compiler is not affected) | |
| 7000 | |
| 7001 * [547]10819 testsuite creates CR+LF on compiler version lines in | |
| 7002 test summary files | |
| 7003 * [548]11612 abi_check not finding correct libgcc_s.so.1 | |
| 7004 | |
| 7005 Miscellaneous | |
| 7006 | |
| 7007 * [549]13211 using -###, incorrect warnings about unused linker file | |
| 7008 are produced | |
| 7009 __________________________________________________________________ | |
| 7010 | |
| 7011 GCC 3.3.4 | |
| 7012 | |
| 7013 This is the [550]list of problem reports (PRs) from GCC's bug tracking | |
| 7014 system that are known to be fixed in the 3.3.4 release. This list might | |
| 7015 not be complete (that is, it is possible that some PRs that have been | |
| 7016 fixed are not listed here). | |
| 7017 __________________________________________________________________ | |
| 7018 | |
| 7019 GCC 3.3.5 | |
| 7020 | |
| 7021 This is the [551]list of problem reports (PRs) from GCC's bug tracking | |
| 7022 system that are known to be fixed in the 3.3.5 release. This list might | |
| 7023 not be complete (that is, it is possible that some PRs that have been | |
| 7024 fixed are not listed here). | |
| 7025 __________________________________________________________________ | |
| 7026 | |
| 7027 GCC 3.3.6 | |
| 7028 | |
| 7029 This is the [552]list of problem reports (PRs) from GCC's bug tracking | |
| 7030 system that are known to be fixed in the 3.3.6 release. This list might | |
| 7031 not be complete (that is, it is possible that some PRs that have been | |
| 7032 fixed are not listed here). | |
| 7033 | |
| 7034 Please send FSF & GNU inquiries & questions to [553]gnu@gnu.org. There | |
| 7035 are also [554]other ways to contact the FSF. | |
| 7036 | |
| 7037 These pages are maintained by [555]the GCC team. | |
| 7038 | |
| 7039 | |
| 7040 For questions related to the use of GCC, please consult these web | |
| 7041 pages and the [556]GCC manuals. If that fails, the | |
| 7042 [557]gcc-help@gcc.gnu.org mailing list might help. | |
| 7043 Please send comments on these web pages and the development of GCC to | |
| 7044 our developer mailing list at [558]gcc@gnu.org or | |
| 7045 [559]gcc@gcc.gnu.org. All of our lists have [560]public archives. | |
| 7046 | |
| 7047 Copyright (C) Free Software Foundation, Inc., 51 Franklin St, Fifth | |
| 7048 Floor, Boston, MA 02110, USA. | |
| 7049 | |
| 7050 Verbatim copying and distribution of this entire article is permitted | |
| 7051 in any medium, provided this notice is preserved. | |
| 7052 Last modified 2007-08-20 [561]Valid XHTML 1.0 | |
| 7053 | |
| 7054 References | |
| 7055 | |
| 7056 1. http://gcc.gnu.org/gcc-3.3/changes.html#3.3.6 | |
| 7057 2. http://gcc.gnu.org/gcc-3.1/changes.html#obsolete_systems | |
| 7058 3. http://gcc.gnu.org/gcc-3.3/changes.html#obsolete_systems | |
| 7059 4. http://gcc.gnu.org/gcc-3.3/changes.html#nonnull_attribute | |
| 7060 5. http://gcc.gnu.org/news/dfa.html | |
| 7061 6. http://gcc.gnu.org/gcc-3.3/c99status.html | |
| 7062 7. http://gcc.gnu.org/onlinedocs/gcc-3.3.6/g77/News.html | |
| 7063 8. http://gcc.gnu.org/PR10140 | |
| 7064 9. http://gcc.gnu.org/PR10198 | |
| 7065 10. http://gcc.gnu.org/PR10338 | |
| 7066 11. http://gcc.gnu.org/PR3581 | |
| 7067 12. http://gcc.gnu.org/PR4382 | |
| 7068 13. http://gcc.gnu.org/PR5533 | |
| 7069 14. http://gcc.gnu.org/PR6387 | |
| 7070 15. http://gcc.gnu.org/PR6412 | |
| 7071 16. http://gcc.gnu.org/PR6620 | |
| 7072 17. http://gcc.gnu.org/PR6663 | |
| 7073 18. http://gcc.gnu.org/PR7068 | |
| 7074 19. http://gcc.gnu.org/PR7083 | |
| 7075 20. http://gcc.gnu.org/PR7647 | |
| 7076 21. http://gcc.gnu.org/PR7675 | |
| 7077 22. http://gcc.gnu.org/PR7718 | |
| 7078 23. http://gcc.gnu.org/PR8116 | |
| 7079 24. http://gcc.gnu.org/PR8358 | |
| 7080 25. http://gcc.gnu.org/PR8511 | |
| 7081 26. http://gcc.gnu.org/PR8564 | |
| 7082 27. http://gcc.gnu.org/PR8660 | |
| 7083 28. http://gcc.gnu.org/PR8766 | |
| 7084 29. http://gcc.gnu.org/PR8803 | |
| 7085 30. http://gcc.gnu.org/PR8846 | |
| 7086 31. http://gcc.gnu.org/PR8906 | |
| 7087 32. http://gcc.gnu.org/PR9216 | |
| 7088 33. http://gcc.gnu.org/PR9261 | |
| 7089 34. http://gcc.gnu.org/PR9263 | |
| 7090 35. http://gcc.gnu.org/PR9429 | |
| 7091 36. http://gcc.gnu.org/PR9516 | |
| 7092 37. http://gcc.gnu.org/PR9600 | |
| 7093 38. http://gcc.gnu.org/PR9629 | |
| 7094 39. http://gcc.gnu.org/PR9672 | |
| 7095 40. http://gcc.gnu.org/PR9749 | |
| 7096 41. http://gcc.gnu.org/PR9794 | |
| 7097 42. http://gcc.gnu.org/PR9829 | |
| 7098 43. http://gcc.gnu.org/PR9916 | |
| 7099 44. http://gcc.gnu.org/PR9936 | |
| 7100 45. http://gcc.gnu.org/PR10262 | |
| 7101 46. http://gcc.gnu.org/PR10278 | |
| 7102 47. http://gcc.gnu.org/PR10446 | |
| 7103 48. http://gcc.gnu.org/PR10451 | |
| 7104 49. http://gcc.gnu.org/PR10506 | |
| 7105 50. http://gcc.gnu.org/PR10549 | |
| 7106 51. http://gcc.gnu.org/PR2001 | |
| 7107 52. http://gcc.gnu.org/PR2391 | |
| 7108 53. http://gcc.gnu.org/PR2960 | |
| 7109 54. http://gcc.gnu.org/PR4046 | |
| 7110 55. http://gcc.gnu.org/PR6405 | |
| 7111 56. http://gcc.gnu.org/PR6798 | |
| 7112 57. http://gcc.gnu.org/PR6871 | |
| 7113 58. http://gcc.gnu.org/PR6909 | |
| 7114 59. http://gcc.gnu.org/PR7189 | |
| 7115 60. http://gcc.gnu.org/PR7642 | |
| 7116 61. http://gcc.gnu.org/PR8634 | |
| 7117 62. http://gcc.gnu.org/PR8750 | |
| 7118 63. http://gcc.gnu.org/PR2161 | |
| 7119 64. http://gcc.gnu.org/PR4319 | |
| 7120 65. http://gcc.gnu.org/PR8602 | |
| 7121 66. http://gcc.gnu.org/PR9177 | |
| 7122 67. http://gcc.gnu.org/PR9853 | |
| 7123 68. http://gcc.gnu.org/PR45 | |
| 7124 69. http://gcc.gnu.org/PR3784 | |
| 7125 70. http://gcc.gnu.org/PR764 | |
| 7126 71. http://gcc.gnu.org/PR5116 | |
| 7127 72. http://gcc.gnu.org/PR2862 | |
| 7128 73. http://gcc.gnu.org/PR3663 | |
| 7129 74. http://gcc.gnu.org/PR3797 | |
| 7130 75. http://gcc.gnu.org/PR3948 | |
| 7131 76. http://gcc.gnu.org/PR4137 | |
| 7132 77. http://gcc.gnu.org/PR4361 | |
| 7133 78. http://gcc.gnu.org/PR4802 | |
| 7134 79. http://gcc.gnu.org/PR5837 | |
| 7135 80. http://gcc.gnu.org/PR4803 | |
| 7136 81. http://gcc.gnu.org/PR5094 | |
| 7137 82. http://gcc.gnu.org/PR5730 | |
| 7138 83. http://gcc.gnu.org/PR6713 | |
| 7139 84. http://gcc.gnu.org/PR7015 | |
| 7140 85. http://gcc.gnu.org/PR7086 | |
| 7141 86. http://gcc.gnu.org/PR7099 | |
| 7142 87. http://gcc.gnu.org/PR7247 | |
| 7143 88. http://gcc.gnu.org/PR7441 | |
| 7144 89. http://gcc.gnu.org/PR7768 | |
| 7145 90. http://gcc.gnu.org/PR7804 | |
| 7146 91. http://gcc.gnu.org/PR8099 | |
| 7147 92. http://gcc.gnu.org/PR8117 | |
| 7148 93. http://gcc.gnu.org/PR8205 | |
| 7149 94. http://gcc.gnu.org/PR8645 | |
| 7150 95. http://gcc.gnu.org/PR8724 | |
| 7151 96. http://gcc.gnu.org/PR8805 | |
| 7152 97. http://gcc.gnu.org/PR8691 | |
| 7153 98. http://gcc.gnu.org/PR8700 | |
| 7154 99. http://gcc.gnu.org/PR8724 | |
| 7155 100. http://gcc.gnu.org/PR8949 | |
| 7156 101. http://gcc.gnu.org/PR9016 | |
| 7157 102. http://gcc.gnu.org/PR9053 | |
| 7158 103. http://gcc.gnu.org/PR9152 | |
| 7159 104. http://gcc.gnu.org/PR9182 | |
| 7160 105. http://gcc.gnu.org/PR9297 | |
| 7161 106. http://gcc.gnu.org/PR9318 | |
| 7162 107. http://gcc.gnu.org/PR9320 | |
| 7163 108. http://gcc.gnu.org/PR9400 | |
| 7164 109. http://gcc.gnu.org/PR9424 | |
| 7165 110. http://gcc.gnu.org/PR9425 | |
| 7166 111. http://gcc.gnu.org/PR9439 | |
| 7167 112. http://gcc.gnu.org/PR9474 | |
| 7168 113. http://gcc.gnu.org/PR9548 | |
| 7169 114. http://www.open-std.org/jtc1/sc22/wg21/docs/cwg_defects.html#231 | |
| 7170 115. http://gcc.gnu.org/PR9555 | |
| 7171 116. http://gcc.gnu.org/PR9561 | |
| 7172 117. http://gcc.gnu.org/PR9563 | |
| 7173 118. http://gcc.gnu.org/PR9582 | |
| 7174 119. http://gcc.gnu.org/PR9622 | |
| 7175 120. http://gcc.gnu.org/PR9683 | |
| 7176 121. http://gcc.gnu.org/PR9791 | |
| 7177 122. http://gcc.gnu.org/PR9817 | |
| 7178 123. http://gcc.gnu.org/PR9825 | |
| 7179 124. http://gcc.gnu.org/PR9826 | |
| 7180 125. http://gcc.gnu.org/PR9924 | |
| 7181 126. http://gcc.gnu.org/PR9946 | |
| 7182 127. http://gcc.gnu.org/PR9964 | |
| 7183 128. http://gcc.gnu.org/PR9988 | |
| 7184 129. http://gcc.gnu.org/PR10033 | |
| 7185 130. http://gcc.gnu.org/PR10097 | |
| 7186 131. http://gcc.gnu.org/PR10132 | |
| 7187 132. http://gcc.gnu.org/PR10180 | |
| 7188 133. http://gcc.gnu.org/PR10199 | |
| 7189 134. http://gcc.gnu.org/PR10300 | |
| 7190 135. http://gcc.gnu.org/PR10427 | |
| 7191 136. http://gcc.gnu.org/PR10503 | |
| 7192 137. http://gcc.gnu.org/PR5956 | |
| 7193 138. http://gcc.gnu.org/PR1832 | |
| 7194 139. http://gcc.gnu.org/PR3924 | |
| 7195 140. http://gcc.gnu.org/PR5634 | |
| 7196 141. http://gcc.gnu.org/PR6367 | |
| 7197 142. http://gcc.gnu.org/PR6491 | |
| 7198 143. http://gcc.gnu.org/PR6742 | |
| 7199 144. http://gcc.gnu.org/PR7113 | |
| 7200 145. http://gcc.gnu.org/PR7236 | |
| 7201 146. http://gcc.gnu.org/PR7278 | |
| 7202 147. http://gcc.gnu.org/PR7384 | |
| 7203 148. http://gcc.gnu.org/PR7388 | |
| 7204 149. http://gcc.gnu.org/PR8587 | |
| 7205 150. http://gcc.gnu.org/PR9038 | |
| 7206 151. http://gcc.gnu.org/PR10197 | |
| 7207 152. http://gcc.gnu.org/PR6005 | |
| 7208 153. http://gcc.gnu.org/PR6389 | |
| 7209 154. http://gcc.gnu.org/PR6576 | |
| 7210 155. http://gcc.gnu.org/PR6652 | |
| 7211 156. http://gcc.gnu.org/PR7060 | |
| 7212 157. http://gcc.gnu.org/PR7073 | |
| 7213 158. http://gcc.gnu.org/PR7180 | |
| 7214 159. http://gcc.gnu.org/PR7416 | |
| 7215 160. http://gcc.gnu.org/PR7570 | |
| 7216 161. http://gcc.gnu.org/PR7578 | |
| 7217 162. http://gcc.gnu.org/PR7611 | |
| 7218 163. http://gcc.gnu.org/PR7709 | |
| 7219 164. http://gcc.gnu.org/PR7766 | |
| 7220 165. http://gcc.gnu.org/PR7785 | |
| 7221 166. http://gcc.gnu.org/PR7786 | |
| 7222 167. http://gcc.gnu.org/PR8142 | |
| 7223 168. http://gcc.gnu.org/PR8234 | |
| 7224 169. http://gcc.gnu.org/PR8415 | |
| 7225 170. http://gcc.gnu.org/PR8481 | |
| 7226 171. http://gcc.gnu.org/PR8593 | |
| 7227 172. http://gcc.gnu.org/PR8759 | |
| 7228 173. http://gcc.gnu.org/PR8997 | |
| 7229 174. http://gcc.gnu.org/PR9253 | |
| 7230 175. http://gcc.gnu.org/PR9254 | |
| 7231 176. http://gcc.gnu.org/PR9271 | |
| 7232 177. http://gcc.gnu.org/PR6767 | |
| 7233 178. http://gcc.gnu.org/PR9911 | |
| 7234 179. http://gcc.gnu.org/PR10020 | |
| 7235 180. http://gcc.gnu.org/PR10546 | |
| 7236 181. http://gcc.gnu.org/PR7029 | |
| 7237 182. http://gcc.gnu.org/PR2903 | |
| 7238 183. http://gcc.gnu.org/PR7873 | |
| 7239 184. http://gcc.gnu.org/PR7680 | |
| 7240 185. http://gcc.gnu.org/PR8705 | |
| 7241 186. http://gcc.gnu.org/PR9986 | |
| 7242 187. http://gcc.gnu.org/PR10056 | |
| 7243 188. http://gcc.gnu.org/PR6744 | |
| 7244 189. http://gcc.gnu.org/PR7361 | |
| 7245 190. http://gcc.gnu.org/PR9496 | |
| 7246 191. http://gcc.gnu.org/PR7067 | |
| 7247 192. http://gcc.gnu.org/PR8480 | |
| 7248 193. http://gcc.gnu.org/PR8784 | |
| 7249 194. http://gcc.gnu.org/PR10315 | |
| 7250 195. http://gcc.gnu.org/PR10267 | |
| 7251 196. http://gcc.gnu.org/PR7916 | |
| 7252 197. http://gcc.gnu.org/PR7926 | |
| 7253 198. http://gcc.gnu.org/PR8555 | |
| 7254 199. http://gcc.gnu.org/PR8994 | |
| 7255 200. http://gcc.gnu.org/PR9426 | |
| 7256 201. http://gcc.gnu.org/PR9806 | |
| 7257 202. http://gcc.gnu.org/PR10077 | |
| 7258 203. http://gcc.gnu.org/PR10233 | |
| 7259 204. http://gcc.gnu.org/PR10286 | |
| 7260 205. http://gcc.gnu.org/PR10308 | |
| 7261 206. http://gcc.gnu.org/PR11272 | |
| 7262 207. http://gcc.gnu.org/PR5754 | |
| 7263 208. http://gcc.gnu.org/PR6597 | |
| 7264 209. http://gcc.gnu.org/PR6949 | |
| 7265 210. http://gcc.gnu.org/PR7053 | |
| 7266 211. http://gcc.gnu.org/PR8164 | |
| 7267 212. http://gcc.gnu.org/PR8384 | |
| 7268 213. http://gcc.gnu.org/PR9559 | |
| 7269 214. http://gcc.gnu.org/PR9649 | |
| 7270 215. http://gcc.gnu.org/PR9864 | |
| 7271 216. http://gcc.gnu.org/PR10432 | |
| 7272 217. http://gcc.gnu.org/PR10475 | |
| 7273 218. http://gcc.gnu.org/PR10635 | |
| 7274 219. http://gcc.gnu.org/PR10661 | |
| 7275 220. http://gcc.gnu.org/PR10700 | |
| 7276 221. http://gcc.gnu.org/PR10712 | |
| 7277 222. http://gcc.gnu.org/PR10796 | |
| 7278 223. http://gcc.gnu.org/PR10890 | |
| 7279 224. http://gcc.gnu.org/PR10939 | |
| 7280 225. http://gcc.gnu.org/PR10956 | |
| 7281 226. http://gcc.gnu.org/PR11041 | |
| 7282 227. http://gcc.gnu.org/PR11059 | |
| 7283 228. http://gcc.gnu.org/PR11083 | |
| 7284 229. http://gcc.gnu.org/PR11105 | |
| 7285 230. http://gcc.gnu.org/PR11149 | |
| 7286 231. http://gcc.gnu.org/PR11228 | |
| 7287 232. http://gcc.gnu.org/PR11282 | |
| 7288 233. http://gcc.gnu.org/PR11301 | |
| 7289 234. http://gcc.gnu.org/PR11308 | |
| 7290 235. http://gcc.gnu.org/PR11473 | |
| 7291 236. http://gcc.gnu.org/PR11503 | |
| 7292 237. http://gcc.gnu.org/PR11513 | |
| 7293 238. http://gcc.gnu.org/PR11198 | |
| 7294 239. http://gcc.gnu.org/PR11304 | |
| 7295 240. http://gcc.gnu.org/PR11381 | |
| 7296 241. http://gcc.gnu.org/PR11536 | |
| 7297 242. http://gcc.gnu.org/PR11557 | |
| 7298 243. http://gcc.gnu.org/PR5897 | |
| 7299 244. http://gcc.gnu.org/PR11279 | |
| 7300 245. http://gcc.gnu.org/PR11022 | |
| 7301 246. http://gcc.gnu.org/PR2330 | |
| 7302 247. http://gcc.gnu.org/PR5388 | |
| 7303 248. http://gcc.gnu.org/PR5390 | |
| 7304 249. http://gcc.gnu.org/PR7877 | |
| 7305 250. http://gcc.gnu.org/PR9393 | |
| 7306 251. http://gcc.gnu.org/PR10032 | |
| 7307 252. http://gcc.gnu.org/PR10468 | |
| 7308 253. http://gcc.gnu.org/PR10527 | |
| 7309 254. http://gcc.gnu.org/PR10679 | |
| 7310 255. http://gcc.gnu.org/PR10682 | |
| 7311 256. http://gcc.gnu.org/PR10689 | |
| 7312 257. http://gcc.gnu.org/PR10845 | |
| 7313 258. http://gcc.gnu.org/PR10849 | |
| 7314 259. http://gcc.gnu.org/PR10888 | |
| 7315 260. http://gcc.gnu.org/PR10929 | |
| 7316 261. http://gcc.gnu.org/PR10931 | |
| 7317 262. http://gcc.gnu.org/PR10940 | |
| 7318 263. http://gcc.gnu.org/PR10968 | |
| 7319 264. http://gcc.gnu.org/PR10990 | |
| 7320 265. http://gcc.gnu.org/PR11039 | |
| 7321 266. http://gcc.gnu.org/PR11062 | |
| 7322 267. http://gcc.gnu.org/PR11095 | |
| 7323 268. http://gcc.gnu.org/PR11098 | |
| 7324 269. http://gcc.gnu.org/PR11137 | |
| 7325 270. http://gcc.gnu.org/PR11154 | |
| 7326 271. http://gcc.gnu.org/PR11329 | |
| 7327 272. http://gcc.gnu.org/PR11332 | |
| 7328 273. http://gcc.gnu.org/PR11431 | |
| 7329 274. http://gcc.gnu.org/PR11528 | |
| 7330 275. http://gcc.gnu.org/PR11546 | |
| 7331 276. http://gcc.gnu.org/PR11567 | |
| 7332 277. http://gcc.gnu.org/PR11645 | |
| 7333 278. http://gcc.gnu.org/PR5179 | |
| 7334 279. http://gcc.gnu.org/PR8204 | |
| 7335 280. http://gcc.gnu.org/PR10838 | |
| 7336 281. http://gcc.gnu.org/PR10886 | |
| 7337 282. http://gcc.gnu.org/PR11349 | |
| 7338 283. http://gcc.gnu.org/PR4823 | |
| 7339 284. http://gcc.gnu.org/PR8878 | |
| 7340 285. http://gcc.gnu.org/PR9815 | |
| 7341 286. http://gcc.gnu.org/PR10402 | |
| 7342 287. http://gcc.gnu.org/PR10504 | |
| 7343 288. http://gcc.gnu.org/PR10673 | |
| 7344 289. http://gcc.gnu.org/PR11044 | |
| 7345 290. http://gcc.gnu.org/PR11089 | |
| 7346 291. http://gcc.gnu.org/PR11420 | |
| 7347 292. http://gcc.gnu.org/PR9362 | |
| 7348 293. http://gcc.gnu.org/PR10142 | |
| 7349 294. http://gcc.gnu.org/PR10663 | |
| 7350 295. http://gcc.gnu.org/PR10835 | |
| 7351 296. http://gcc.gnu.org/PR10876 | |
| 7352 297. http://gcc.gnu.org/PR10955 | |
| 7353 298. http://gcc.gnu.org/PR11018 | |
| 7354 299. http://gcc.gnu.org/PR11556 | |
| 7355 300. http://gcc.gnu.org/PR10907 | |
| 7356 301. http://gcc.gnu.org/PR11320 | |
| 7357 302. http://gcc.gnu.org/PR11599 | |
| 7358 303. http://gcc.gnu.org/PR9745 | |
| 7359 304. http://gcc.gnu.org/PR10871 | |
| 7360 305. http://gcc.gnu.org/PR11440 | |
| 7361 306. http://gcc.gnu.org/PR7594 | |
| 7362 307. http://gcc.gnu.org/PR10557 | |
| 7363 308. http://gcc.gnu.org/PR11054 | |
| 7364 309. http://gcc.gnu.org/PR10834 | |
| 7365 310. http://gcc.gnu.org/PR10842 | |
| 7366 311. http://gcc.gnu.org/PR11052 | |
| 7367 312. http://gcc.gnu.org/PR11183 | |
| 7368 313. http://gcc.gnu.org/PR11084 | |
| 7369 314. http://gcc.gnu.org/PR10331 | |
| 7370 315. http://gcc.gnu.org/PR10413 | |
| 7371 316. http://gcc.gnu.org/PR11096 | |
| 7372 317. http://gcc.gnu.org/PR2873 | |
| 7373 318. http://gcc.gnu.org/PR3163 | |
| 7374 319. http://gcc.gnu.org/PR5287 | |
| 7375 320. http://gcc.gnu.org/PR10148 | |
| 7376 321. http://gcc.gnu.org/PR8787 | |
| 7377 322. http://gcc.gnu.org/PR10900 | |
| 7378 323. http://gcc.gnu.org/PR1607 | |
| 7379 324. http://gcc.gnu.org/PR4252 | |
| 7380 325. http://gcc.gnu.org/PR4490 | |
| 7381 326. http://gcc.gnu.org/PR10355 | |
| 7382 327. http://gcc.gnu.org/PR10726 | |
| 7383 328. http://gcc.gnu.org/PR10805 | |
| 7384 329. http://gcc.gnu.org/PR10815 | |
| 7385 330. http://gcc.gnu.org/PR10877 | |
| 7386 331. http://gcc.gnu.org/PR11280 | |
| 7387 332. http://gcc.gnu.org/PR11466 | |
| 7388 333. http://gcc.gnu.org/PR10737 | |
| 7389 334. http://gcc.gnu.org/PR10810 | |
| 7390 335. http://gcc.gnu.org/bugzilla/ | |
| 7391 336. http://gcc.gnu.org/PR8336 | |
| 7392 337. http://gcc.gnu.org/PR9330 | |
| 7393 338. http://gcc.gnu.org/PR9631 | |
| 7394 339. http://gcc.gnu.org/PR9877 | |
| 7395 340. http://gcc.gnu.org/PR11687 | |
| 7396 341. http://gcc.gnu.org/PR12263 | |
| 7397 342. http://gcc.gnu.org/PR12490 | |
| 7398 343. http://gcc.gnu.org/PR7277 | |
| 7399 344. http://gcc.gnu.org/PR7939 | |
| 7400 345. http://gcc.gnu.org/PR11063 | |
| 7401 346. http://gcc.gnu.org/PR11207 | |
| 7402 347. http://gcc.gnu.org/PR11522 | |
| 7403 348. http://gcc.gnu.org/PR11595 | |
| 7404 349. http://gcc.gnu.org/PR11646 | |
| 7405 350. http://gcc.gnu.org/PR11665 | |
| 7406 351. http://gcc.gnu.org/PR11852 | |
| 7407 352. http://gcc.gnu.org/PR11878 | |
| 7408 353. http://gcc.gnu.org/PR11883 | |
| 7409 354. http://gcc.gnu.org/PR11991 | |
| 7410 355. http://gcc.gnu.org/PR12146 | |
| 7411 356. http://gcc.gnu.org/PR12215 | |
| 7412 357. http://gcc.gnu.org/PR12369 | |
| 7413 358. http://gcc.gnu.org/PR12446 | |
| 7414 359. http://gcc.gnu.org/PR12510 | |
| 7415 360. http://gcc.gnu.org/PR12544 | |
| 7416 361. http://gcc.gnu.org/PR9862 | |
| 7417 362. http://gcc.gnu.org/PR10962 | |
| 7418 363. http://gcc.gnu.org/PR11370 | |
| 7419 364. http://gcc.gnu.org/PR11637 | |
| 7420 365. http://gcc.gnu.org/PR11885 | |
| 7421 366. http://gcc.gnu.org/PR12082 | |
| 7422 367. http://gcc.gnu.org/PR12180 | |
| 7423 368. http://gcc.gnu.org/PR12340 | |
| 7424 369. http://gcc.gnu.org/PR3907 | |
| 7425 370. http://gcc.gnu.org/PR5293 | |
| 7426 371. http://gcc.gnu.org/PR5296 | |
| 7427 372. http://gcc.gnu.org/PR7939 | |
| 7428 373. http://gcc.gnu.org/PR8656 | |
| 7429 374. http://gcc.gnu.org/PR10147 | |
| 7430 375. http://gcc.gnu.org/PR11400 | |
| 7431 376. http://gcc.gnu.org/PR11409 | |
| 7432 377. http://gcc.gnu.org/PR11740 | |
| 7433 378. http://gcc.gnu.org/PR11786 | |
| 7434 379. http://gcc.gnu.org/PR11867 | |
| 7435 380. http://gcc.gnu.org/PR11928 | |
| 7436 381. http://gcc.gnu.org/PR12114 | |
| 7437 382. http://gcc.gnu.org/PR12163 | |
| 7438 383. http://gcc.gnu.org/PR12181 | |
| 7439 384. http://gcc.gnu.org/PR12236 | |
| 7440 385. http://gcc.gnu.org/PR12266 | |
| 7441 386. http://gcc.gnu.org/PR12296 | |
| 7442 387. http://gcc.gnu.org/PR12298 | |
| 7443 388. http://gcc.gnu.org/PR12369 | |
| 7444 389. http://gcc.gnu.org/PR12337 | |
| 7445 390. http://gcc.gnu.org/PR12344 | |
| 7446 391. http://gcc.gnu.org/PR12451 | |
| 7447 392. http://gcc.gnu.org/PR12486 | |
| 7448 393. http://gcc.gnu.org/PR8869 | |
| 7449 394. http://gcc.gnu.org/PR9786 | |
| 7450 395. http://gcc.gnu.org/PR11689 | |
| 7451 396. http://gcc.gnu.org/PR12116 | |
| 7452 397. http://gcc.gnu.org/PR12070 | |
| 7453 398. http://gcc.gnu.org/PR11184 | |
| 7454 399. http://gcc.gnu.org/PR11535 | |
| 7455 400. http://gcc.gnu.org/PR11693 | |
| 7456 401. http://gcc.gnu.org/PR12224 | |
| 7457 402. http://gcc.gnu.org/PR11087 | |
| 7458 403. http://gcc.gnu.org/PR11319 | |
| 7459 404. http://gcc.gnu.org/PR11949 | |
| 7460 405. http://gcc.gnu.org/PR11662 | |
| 7461 406. http://gcc.gnu.org/PR11965 | |
| 7462 407. http://gcc.gnu.org/PR12301 | |
| 7463 408. http://gcc.gnu.org/PR11717 | |
| 7464 409. http://gcc.gnu.org/PR11313 | |
| 7465 410. http://gcc.gnu.org/PR11712 | |
| 7466 411. http://gcc.gnu.org/PR12166 | |
| 7467 412. http://gcc.gnu.org/PR12101 | |
| 7468 413. http://gcc.gnu.org/PR10988 | |
| 7469 414. http://gcc.gnu.org/PR11805 | |
| 7470 415. http://gcc.gnu.org/PR11902 | |
| 7471 416. http://gcc.gnu.org/PR11903 | |
| 7472 417. http://gcc.gnu.org/bugzilla/ | |
| 7473 418. http://gcc.gnu.org/PR11890 | |
| 7474 419. http://gcc.gnu.org/PR12399 | |
| 7475 420. http://gcc.gnu.org/PR13068 | |
| 7476 421. http://gcc.gnu.org/PR10060 | |
| 7477 422. http://gcc.gnu.org/PR10555 | |
| 7478 423. http://gcc.gnu.org/PR10706 | |
| 7479 424. http://gcc.gnu.org/PR11496 | |
| 7480 425. http://gcc.gnu.org/PR11741 | |
| 7481 426. http://gcc.gnu.org/PR12440 | |
| 7482 427. http://gcc.gnu.org/PR12632 | |
| 7483 428. http://gcc.gnu.org/PR12712 | |
| 7484 429. http://gcc.gnu.org/PR12726 | |
| 7485 430. http://gcc.gnu.org/PR12890 | |
| 7486 431. http://gcc.gnu.org/PR12900 | |
| 7487 432. http://gcc.gnu.org/PR13060 | |
| 7488 433. http://gcc.gnu.org/PR13289 | |
| 7489 434. http://gcc.gnu.org/PR13318 | |
| 7490 435. http://gcc.gnu.org/PR13392 | |
| 7491 436. http://gcc.gnu.org/PR13574 | |
| 7492 437. http://gcc.gnu.org/PR13475 | |
| 7493 438. http://gcc.gnu.org/PR13797 | |
| 7494 439. http://gcc.gnu.org/PR13824 | |
| 7495 440. http://gcc.gnu.org/PR8776 | |
| 7496 441. http://gcc.gnu.org/PR10339 | |
| 7497 442. http://gcc.gnu.org/PR11350 | |
| 7498 443. http://gcc.gnu.org/PR12826 | |
| 7499 444. http://gcc.gnu.org/PR12500 | |
| 7500 445. http://gcc.gnu.org/PR12941 | |
| 7501 446. http://gcc.gnu.org/PR12953 | |
| 7502 447. http://gcc.gnu.org/PR13041 | |
| 7503 448. http://gcc.gnu.org/PR13507 | |
| 7504 449. http://gcc.gnu.org/PR13382 | |
| 7505 450. http://gcc.gnu.org/PR13394 | |
| 7506 451. http://gcc.gnu.org/PR13400 | |
| 7507 452. http://gcc.gnu.org/PR13521 | |
| 7508 453. http://gcc.gnu.org/PR2094 | |
| 7509 454. http://gcc.gnu.org/PR2294 | |
| 7510 455. http://gcc.gnu.org/PR5050 | |
| 7511 456. http://gcc.gnu.org/PR9371 | |
| 7512 457. http://gcc.gnu.org/PR9546 | |
| 7513 458. http://gcc.gnu.org/PR10081 | |
| 7514 459. http://gcc.gnu.org/PR10093 | |
| 7515 460. http://www.open-std.org/jtc1/sc22/wg21/docs/cwg_defects.html#61 | |
| 7516 461. http://gcc.gnu.org/PR10095 | |
| 7517 462. http://gcc.gnu.org/PR11554 | |
| 7518 463. http://gcc.gnu.org/PR12297 | |
| 7519 464. http://gcc.gnu.org/PR12352 | |
| 7520 465. http://gcc.gnu.org/PR12438 | |
| 7521 466. http://gcc.gnu.org/PR12540 | |
| 7522 467. http://gcc.gnu.org/PR12594 | |
| 7523 468. http://www.open-std.org/jtc1/sc22/wg21/docs/cwg_defects.html#60 | |
| 7524 469. http://www.open-std.org/jtc1/sc22/wg21/docs/cwg_defects.html#63 | |
| 7525 470. http://gcc.gnu.org/PR12657 | |
| 7526 471. http://www.open-std.org/jtc1/sc22/wg21/docs/cwg_defects.html#292 | |
| 7527 472. http://gcc.gnu.org/PR12696 | |
| 7528 473. http://gcc.gnu.org/PR12815 | |
| 7529 474. http://gcc.gnu.org/PR12862 | |
| 7530 475. http://gcc.gnu.org/PR12926 | |
| 7531 476. http://gcc.gnu.org/PR12967 | |
| 7532 477. http://www.open-std.org/jtc1/sc22/wg21/docs/cwg_defects.html | |
| 7533 478. http://gcc.gnu.org/PR12971 | |
| 7534 479. http://www.open-std.org/jtc1/sc22/wg21/docs/cwg_defects.html#328 | |
| 7535 480. http://gcc.gnu.org/PR13007 | |
| 7536 481. http://gcc.gnu.org/PR13009 | |
| 7537 482. http://gcc.gnu.org/PR13057 | |
| 7538 483. http://gcc.gnu.org/PR13070 | |
| 7539 484. http://gcc.gnu.org/PR13081 | |
| 7540 485. http://gcc.gnu.org/PR13239 | |
| 7541 486. http://gcc.gnu.org/PR13262 | |
| 7542 487. http://gcc.gnu.org/PR13290 | |
| 7543 488. http://gcc.gnu.org/PR13323 | |
| 7544 489. http://gcc.gnu.org/PR13369 | |
| 7545 490. http://gcc.gnu.org/PR13371 | |
| 7546 491. http://gcc.gnu.org/PR13445 | |
| 7547 492. http://gcc.gnu.org/PR13461 | |
| 7548 493. http://gcc.gnu.org/PR13462 | |
| 7549 494. http://gcc.gnu.org/PR13478 | |
| 7550 495. http://gcc.gnu.org/PR13544 | |
| 7551 496. http://gcc.gnu.org/PR13650 | |
| 7552 497. http://gcc.gnu.org/PR13683 | |
| 7553 498. http://gcc.gnu.org/PR13688 | |
| 7554 499. http://gcc.gnu.org/PR13774 | |
| 7555 500. http://gcc.gnu.org/PR13884 | |
| 7556 501. http://gcc.gnu.org/PR10746 | |
| 7557 502. http://gcc.gnu.org/PR11433 | |
| 7558 503. http://gcc.gnu.org/PR12633 | |
| 7559 504. http://gcc.gnu.org/PR13037 | |
| 7560 505. http://gcc.gnu.org/PR13213 | |
| 7561 506. http://gcc.gnu.org/PR4490 | |
| 7562 507. http://gcc.gnu.org/PR12292 | |
| 7563 508. http://gcc.gnu.org/PR12441 | |
| 7564 509. http://gcc.gnu.org/PR12943 | |
| 7565 510. http://gcc.gnu.org/PR13608 | |
| 7566 511. http://gcc.gnu.org/PR11598 | |
| 7567 512. http://gcc.gnu.org/PR11793 | |
| 7568 513. http://gcc.gnu.org/PR12467 | |
| 7569 514. http://gcc.gnu.org/PR12537 | |
| 7570 515. http://gcc.gnu.org/PR12496 | |
| 7571 516. http://gcc.gnu.org/PR12865 | |
| 7572 517. http://gcc.gnu.org/PR13354 | |
| 7573 518. http://gcc.gnu.org/PR10467 | |
| 7574 519. http://gcc.gnu.org/PR11226 | |
| 7575 520. http://gcc.gnu.org/PR11227 | |
| 7576 521. http://gcc.gnu.org/PR12644 | |
| 7577 522. http://gcc.gnu.org/PR13149 | |
| 7578 523. http://gcc.gnu.org/PR12654 | |
| 7579 524. http://gcc.gnu.org/PR12965 | |
| 7580 525. http://gcc.gnu.org/PR13031 | |
| 7581 526. http://gcc.gnu.org/PR11634 | |
| 7582 527. http://gcc.gnu.org/PR12158 | |
| 7583 528. http://gcc.gnu.org/PR11992 | |
| 7584 529. http://gcc.gnu.org/PR9365 | |
| 7585 530. http://gcc.gnu.org/PR10392 | |
| 7586 531. http://gcc.gnu.org/PR11322 | |
| 7587 532. http://gcc.gnu.org/PR13069 | |
| 7588 533. http://gcc.gnu.org/PR13302 | |
| 7589 534. http://gcc.gnu.org/PR13585 | |
| 7590 535. http://gcc.gnu.org/PR8916 | |
| 7591 536. http://gcc.gnu.org/PR11576 | |
| 7592 537. http://gcc.gnu.org/PR13122 | |
| 7593 538. http://gcc.gnu.org/PR13256 | |
| 7594 539. http://gcc.gnu.org/PR13373 | |
| 7595 540. http://gcc.gnu.org/PR12561 | |
| 7596 541. http://gcc.gnu.org/PR6243 | |
| 7597 542. http://gcc.gnu.org/PR11397 | |
| 7598 543. http://gcc.gnu.org/PR12505 | |
| 7599 544. http://gcc.gnu.org/PR13150 | |
| 7600 545. http://gcc.gnu.org/PR12666 | |
| 7601 546. http://gcc.gnu.org/PR12969 | |
| 7602 547. http://gcc.gnu.org/PR10819 | |
| 7603 548. http://gcc.gnu.org/PR11612 | |
| 7604 549. http://gcc.gnu.org/PR13211 | |
| 7605 550. http://gcc.gnu.org/bugzilla/buglist.cgi?bug_status=RESOLVED&resolution=FIXED&target_milestone=3.3.4 | |
| 7606 551. http://gcc.gnu.org/bugzilla/buglist.cgi?bug_status=RESOLVED&resolution=FIXED&target_milestone=3.3.5 | |
| 7607 552. http://gcc.gnu.org/bugzilla/buglist.cgi?bug_status=RESOLVED&resolution=FIXED&target_milestone=3.3.6 | |
| 7608 553. mailto:gnu@gnu.org | |
| 7609 554. http://www.gnu.org/home.html#ContactInfo | |
| 7610 555. http://gcc.gnu.org/about.html | |
| 7611 556. http://gcc.gnu.org/onlinedocs/ | |
| 7612 557. mailto:gcc-help@gcc.gnu.org | |
| 7613 558. mailto:gcc@gnu.org | |
| 7614 559. mailto:gcc@gcc.gnu.org | |
| 7615 560. http://gcc.gnu.org/lists.html | |
| 7616 561. http://validator.w3.org/check/referer | |
| 7617 ====================================================================== | |
| 7618 http://gcc.gnu.org/gcc-3.2/index.html | |
| 7619 GCC 3.2 Release Series | |
| 7620 | |
| 7621 April 25, 2003 | |
| 7622 | |
| 7623 The [1]GNU project and the GCC developers are pleased to announce the | |
| 7624 release of GCC 3.2.3. | |
| 7625 | |
| 7626 The purpose of the GCC 3.2 release series is to provide a stable | |
| 7627 platform for OS distributors to use building their next releases. A | |
| 7628 primary objective was to stabilize the C++ ABI; we believe that the | |
| 7629 interface to the compiler and the C++ standard library are now | |
| 7630 relatively stable. | |
| 7631 | |
| 7632 Be aware that C++ code compiled by GCC 3.2.x will (in general) not | |
| 7633 interoperate with code compiled by GCC 3.1.1 or earlier. | |
| 7634 | |
| 7635 Please refer to our [2]detailed list of news, caveats, and bug-fixes | |
| 7636 for further information. | |
| 7637 | |
| 7638 Release History | |
| 7639 | |
| 7640 GCC 3.2.3 | |
| 7641 April 25, 2003 ([3]changes) | |
| 7642 | |
| 7643 GCC 3.2.2 | |
| 7644 February 5, 2003 ([4]changes) | |
| 7645 | |
| 7646 GCC 3.2.1 | |
| 7647 November 19, 2002 ([5]changes) | |
| 7648 | |
| 7649 GCC 3.2 | |
| 7650 August 14, 2002 ([6]changes) | |
| 7651 | |
| 7652 References and Acknowledgements | |
| 7653 | |
| 7654 GCC used to stand for the GNU C Compiler, but since the compiler | |
| 7655 supports several other languages aside from C, it now stands for the | |
| 7656 GNU Compiler Collection. | |
| 7657 | |
| 7658 A list of [7]successful builds is updated as new information becomes | |
| 7659 available. | |
| 7660 | |
| 7661 The GCC developers would like to thank the numerous people that have | |
| 7662 contributed new features, improvements, bug fixes, and other changes as | |
| 7663 well as test results to GCC. This [8]amazing group of volunteers is | |
| 7664 what makes GCC successful. | |
| 7665 | |
| 7666 For additional information about GCC please refer to the [9]GCC project | |
| 7667 web site or contact the [10]GCC development mailing list. | |
| 7668 | |
| 7669 To obtain GCC please use [11]our mirror sites, or our CVS server. | |
| 7670 | |
| 7671 Please send FSF & GNU inquiries & questions to [12]gnu@gnu.org. There | |
| 7672 are also [13]other ways to contact the FSF. | |
| 7673 | |
| 7674 These pages are maintained by [14]the GCC team. | |
| 7675 | |
| 7676 | |
| 7677 For questions related to the use of GCC, please consult these web | |
| 7678 pages and the [15]GCC manuals. If that fails, the | |
| 7679 [16]gcc-help@gcc.gnu.org mailing list might help. | |
| 7680 Please send comments on these web pages and the development of GCC to | |
| 7681 our developer mailing list at [17]gcc@gnu.org or [18]gcc@gcc.gnu.org. | |
| 7682 All of our lists have [19]public archives. | |
| 7683 | |
| 7684 Copyright (C) Free Software Foundation, Inc., 51 Franklin St, Fifth | |
| 7685 Floor, Boston, MA 02110, USA. | |
| 7686 | |
| 7687 Verbatim copying and distribution of this entire article is permitted | |
| 7688 in any medium, provided this notice is preserved. | |
| 7689 Last modified 2008-07-26 [20]Valid XHTML 1.0 | |
| 7690 | |
| 7691 References | |
| 7692 | |
| 7693 1. http://www.gnu.org/ | |
| 7694 2. http://gcc.gnu.org/gcc-3.2/changes.html | |
| 7695 3. http://gcc.gnu.org/gcc-3.2/changes.html#3.2.3 | |
| 7696 4. http://gcc.gnu.org/gcc-3.2/changes.html#3.2.2 | |
| 7697 5. http://gcc.gnu.org/gcc-3.2/changes.html#3.2.1 | |
| 7698 6. http://gcc.gnu.org/gcc-3.2/changes.html#3.2 | |
| 7699 7. http://gcc.gnu.org/gcc-3.2/buildstat.html | |
| 7700 8. http://gcc.gnu.org/onlinedocs/gcc/Contributors.html | |
| 7701 9. http://gcc.gnu.org/index.html | |
| 7702 10. mailto:gcc@gcc.gnu.org | |
| 7703 11. http://gcc.gnu.org/mirrors.html | |
| 7704 12. mailto:gnu@gnu.org | |
| 7705 13. http://www.gnu.org/home.html#ContactInfo | |
| 7706 14. http://gcc.gnu.org/about.html | |
| 7707 15. http://gcc.gnu.org/onlinedocs/ | |
| 7708 16. mailto:gcc-help@gcc.gnu.org | |
| 7709 17. mailto:gcc@gnu.org | |
| 7710 18. mailto:gcc@gcc.gnu.org | |
| 7711 19. http://gcc.gnu.org/lists.html | |
| 7712 20. http://validator.w3.org/check/referer | |
| 7713 ====================================================================== | |
| 7714 http://gcc.gnu.org/gcc-3.2/changes.html | |
| 7715 GCC 3.2 Release Series | |
| 7716 Changes, New Features, and Fixes | |
| 7717 | |
| 7718 The latest release in the 3.2 release series is [1]GCC 3.2.3. | |
| 7719 | |
| 7720 Caveats and New Features | |
| 7721 | |
| 7722 Caveats | |
| 7723 | |
| 7724 * The C++ compiler does not correctly zero-initialize | |
| 7725 pointers-to-data members. You must explicitly initialize them. For | |
| 7726 example: int S::*m(0); will work, but depending on | |
| 7727 default-initialization to zero will not work. This bug cannot be | |
| 7728 fixed in GCC 3.2 without inducing unacceptable risks. It will be | |
| 7729 fixed in GCC 3.3. | |
| 7730 * This GCC release is based on the GCC 3.1 sourcebase, and thus has | |
| 7731 all the [2]changes in the GCC 3.1 series. In addition, GCC 3.2 has | |
| 7732 a number of C++ ABI fixes which make its C++ compiler generate | |
| 7733 binary code which is incompatible with the C++ compilers found in | |
| 7734 earlier GCC releases, including GCC 3.1 and GCC 3.1.1. | |
| 7735 | |
| 7736 Frontend Enhancements | |
| 7737 | |
| 7738 C/C++/Objective-C | |
| 7739 | |
| 7740 * The method of constructing the list of directories to be searched | |
| 7741 for header files has been revised. If a directory named by a -I | |
| 7742 option is a standard system include directory, the option is | |
| 7743 ignored to ensure that the default search order for system | |
| 7744 directories and the special treatment of system header files are | |
| 7745 not defeated. | |
| 7746 * The C and Objective-C compilers no longer accept the "Naming Types" | |
| 7747 extension (typedef foo = bar); it was already unavailable in C++. | |
| 7748 Code which uses it will need to be changed to use the "typeof" | |
| 7749 extension instead: typedef typeof(bar) foo. (We have removed this | |
| 7750 extension without a period of deprecation because it has caused the | |
| 7751 compiler to crash since version 3.0 and no one noticed until very | |
| 7752 recently. Thus we conclude it is not in widespread use.) | |
| 7753 | |
| 7754 C++ | |
| 7755 | |
| 7756 * GCC 3.2 fixed serveral differences between the C++ ABI implemented | |
| 7757 in GCC and the multi-vendor standard, but more have been found | |
| 7758 since the release. 3.2.1 adds a new warning, -Wabi, to warn about | |
| 7759 code which is affected by these bugs. We will fix these bugs in | |
| 7760 some future release, once we are confident that all have been | |
| 7761 found; until then, it is our intention to make changes to the ABI | |
| 7762 only if they are necessary for correct compilation of C++, as | |
| 7763 opposed to conformance to the ABI documents. | |
| 7764 * For details on how to build an ABI compliant compiler for GNU/Linux | |
| 7765 systems, check the [3]common C++ ABI page. | |
| 7766 | |
| 7767 New Targets and Target Specific Improvements | |
| 7768 | |
| 7769 IA-32 | |
| 7770 | |
| 7771 * Fixed a number of bugs in SSE and MMX intrinsics. | |
| 7772 * Fixed common compiler crashes with SSE instruction set enabled | |
| 7773 (implied by -march=pentium3, pentium4, athlon-xp) | |
| 7774 * __m128 and __m128i is not 128bit aligned when used in structures. | |
| 7775 | |
| 7776 x86-64 | |
| 7777 | |
| 7778 * A bug whereby the compiler could generate bad code for bzero has | |
| 7779 been fixed. | |
| 7780 * ABI fixes (implying ABI incompatibilities with previous version in | |
| 7781 some corner cases) | |
| 7782 * Fixed prefetch code generation | |
| 7783 __________________________________________________________________ | |
| 7784 | |
| 7785 GCC 3.2.3 | |
| 7786 | |
| 7787 3.2.3 is a bug fix release only; there are no new features that were | |
| 7788 not present in GCC 3.2.2. | |
| 7789 | |
| 7790 Bug Fixes | |
| 7791 | |
| 7792 This section lists the problem reports (PRs) from GCC's bug tracking | |
| 7793 system that are known to be fixed in the 3.2.3 release. This list might | |
| 7794 not be complete (that is, it is possible that some PRs that have been | |
| 7795 fixed are not listed here), and some of the titles have been changed to | |
| 7796 make them more clear. | |
| 7797 | |
| 7798 Internal Compiler Errors (multi-platform) | |
| 7799 | |
| 7800 * [4]3782: (c++) -quiet -fstats produces a segmentation fault in | |
| 7801 cc1plus | |
| 7802 * [5]6440: (c++) template specializations cause ICE | |
| 7803 * [6]7050: (c++) ICE on: (i ? get_string() : throw) | |
| 7804 * [7]7741: ICE on conflicting types (make_decl_rtl in varasm.c) | |
| 7805 * [8]7982: (c++) ICE due to infinite recursion (using STL set) | |
| 7806 * [9]8068: exceedingly high (infinite) memory usage | |
| 7807 * [10]8178: ICE with __builtin_ffs | |
| 7808 * [11]8396: ICE in copy_to_mode_reg, in explow.c | |
| 7809 * [12]8674: (c++) ICE in cp_expr_size, in cp/cp-lang.c | |
| 7810 * [13]9768: ICE when optimizing inline code at -O2 | |
| 7811 * [14]9798: (c++) Infinite recursion (segfault) in | |
| 7812 cp/decl.c:push_using_directive with recursive using directives | |
| 7813 * [15]9799: mismatching structure initializer with nested flexible | |
| 7814 array member: ICE | |
| 7815 * [16]9928: ICE on duplicate enum declaration | |
| 7816 * [17]10114: ICE in mem_loc_descriptor, in dwarf2out.c (affects | |
| 7817 sparc, alpha) | |
| 7818 * [18]10352: ICE in find_reloads_toplev | |
| 7819 * [19]10336: ICE with -Wunreachable-code | |
| 7820 | |
| 7821 C/optimizer bugs: | |
| 7822 | |
| 7823 * [20]8224: Incorrect joining of signed and unsigned division | |
| 7824 * [21]8613: -O2 produces wrong code with builtin strlen and | |
| 7825 postincrements | |
| 7826 * [22]8828: gcc reports some code is unreachable when it is not | |
| 7827 * [23]9226: GCSE breaking argument passing | |
| 7828 * [24]9853: miscompilation of non-constant structure initializer | |
| 7829 * [25]9797: C99-style struct initializers are miscompiled | |
| 7830 * [26]9967: Some standard C function calls should not be replaced | |
| 7831 when optimizing for size | |
| 7832 * [27]10116: ce2: invalid merge of join_bb in the context of switch | |
| 7833 statements | |
| 7834 * [28]10171: wrong code for inlined function | |
| 7835 * [29]10175: -Wunreachable-code doesn't work for single lines | |
| 7836 | |
| 7837 C++ compiler and library: | |
| 7838 | |
| 7839 * [30]8316: Confusing diagnostic for code that misuses conversion | |
| 7840 operators | |
| 7841 * [31]9169: filebuf output fails if codecvt<>::out returns noconv | |
| 7842 * [32]9420: incomplete type incorrectly reported | |
| 7843 * [33]9459: typeof in return type specification of template not | |
| 7844 supported | |
| 7845 * [34]9507: filebuf::open handles ios_base::ate incorrectly | |
| 7846 * [35]9538: Out-of-bounds memory access in streambuf::sputbackc | |
| 7847 * [36]9602: Total confusion about template/friend/virtual/abstract | |
| 7848 * [37]9993: destructor not called for local object created within and | |
| 7849 returned from infinite loop | |
| 7850 * [38]10167: ieee_1003.1-2001 locale specialisations on a glibc-2.3.2 | |
| 7851 system | |
| 7852 | |
| 7853 Java compiler and library: | |
| 7854 | |
| 7855 * [39]9652: libgcj build fails on irix6.5.1[78] | |
| 7856 * [40]10144: gas on solaris complains about bad .stabs lines for | |
| 7857 java, native as unaffected | |
| 7858 | |
| 7859 x86-specific (Intel/AMD): | |
| 7860 | |
| 7861 * [41]8746: gcc miscompiles Linux kernel ppa driver on x86 | |
| 7862 * [42]9888: -mcpu=k6 -Os produces out of range loop instructions | |
| 7863 * [43]9638: Cross-build for target i386-elf and i586-pc-linux-gnu | |
| 7864 failed | |
| 7865 * [44]9954: Cross-build for target i586-pc-linux-gnu (--with-newlib) | |
| 7866 failed | |
| 7867 | |
| 7868 SPARC-specific: | |
| 7869 | |
| 7870 * [45]7784: [Sparc] ICE in extract_insn, in recog.c | |
| 7871 * [46]7796: sparc extra failure with -m64 on execute/930921-1.c in | |
| 7872 unroll.c | |
| 7873 * [47]8281: ICE when compiling with -O2 -fPIC for Ultrasparc | |
| 7874 * [48]8366: [Sparc] C testsuite failure with -m64 -fpic -O in | |
| 7875 execute/loop-2d.c | |
| 7876 * [49]8726: gcc -O2 miscompiles Samba 2.2.7 on 32-bit sparc | |
| 7877 * [50]9414: Scheduling bug on Ultrasparc | |
| 7878 * [51]10067: GCC-3.2.2 outputs invalid asm on sparc64 | |
| 7879 | |
| 7880 m68k-specific: | |
| 7881 | |
| 7882 * [52]7248: broken "inclusive or" code | |
| 7883 * [53]8343: m68k-elf/rtems ICE at instantiate_virtual_regs_1 | |
| 7884 | |
| 7885 PowerPC-specific: | |
| 7886 | |
| 7887 * [54]9732: Wrong code with -O2 -fPIC | |
| 7888 * [55]10073: ICE: powerpc cannot split insn | |
| 7889 | |
| 7890 Alpha-specific: | |
| 7891 | |
| 7892 * [56]7702: optimization problem on a DEC alpha under OSF1 | |
| 7893 * [57]9671: gcc.3.2.2 does not build on a HP Tru64 Unix v5.1B system | |
| 7894 | |
| 7895 HP-specific: | |
| 7896 | |
| 7897 * [58]8694: <string> breaks <ctype.h> on HP-UX 10.20 (DUP: 9275) | |
| 7898 * [59]9953: (ada) gcc 3.2.x can't build 3.3-branch ada on HP-UX 10 | |
| 7899 (missing symbol) | |
| 7900 * [60]10271: Floating point args don't get reloaded across function | |
| 7901 calls with -O2 | |
| 7902 | |
| 7903 MIPS specific: | |
| 7904 | |
| 7905 * [61]6362: mips-irix6 gcc-3.1 C testsuite failure with -mips4 in | |
| 7906 compile/920501-4.c | |
| 7907 | |
| 7908 CRIS specific: | |
| 7909 | |
| 7910 * [62]10377: gcc-3.2.2 creates bad assembler code for cris | |
| 7911 | |
| 7912 Miscellaneous and minor bugs: | |
| 7913 | |
| 7914 * [63]6955: collect2 says "core dumped" when there is no core | |
| 7915 __________________________________________________________________ | |
| 7916 | |
| 7917 GCC 3.2.2 | |
| 7918 | |
| 7919 Beginning with 3.2.2, GCC's Makefile suite supports redirection of make | |
| 7920 install by means of the DESTDIR variable. Parts of the GCC tree have | |
| 7921 featured that support long before, but now it is available even from | |
| 7922 the top level. | |
| 7923 | |
| 7924 Other than that, GCC 3.2.2 is a bug fix release only; there are no new | |
| 7925 features that were not present in GCC 3.2.1. | |
| 7926 | |
| 7927 Bug Fixes | |
| 7928 | |
| 7929 On the following i386-based systems GCC 3.2.1 broke the C ABI wrt. | |
| 7930 functions returning structures: Cygwin, FreeBSD (GCC 3.2.1 as shipped | |
| 7931 with FreeBSD 5.0 does not have this problem), Interix, a.out-based | |
| 7932 Linux and NetBSD, OpenBSD, and Darwin. GCC 3.2.2 reverts this ABI | |
| 7933 change, and thus restores ABI-compatibility with previous releases | |
| 7934 (except GCC 3.2.1) on these platforms. | |
| 7935 | |
| 7936 This section lists the problem reports (PRs) from GCC's bug tracking | |
| 7937 system that are known to be fixed in the 3.2.2 release. This list might | |
| 7938 not be complete (that is, it is possible that some PRs that have been | |
| 7939 fixed are not listed here) and some of the titles have been changed to | |
| 7940 make them more clear. | |
| 7941 | |
| 7942 Internal Compiler Errors (multi-platform) | |
| 7943 | |
| 7944 * [64]5919: (c++) ICE when passing variable array to template | |
| 7945 function | |
| 7946 * [65]7129: (c++) ICE with min/max assignment operators (<?= and >?=) | |
| 7947 * [66]7507: ICE with -O2 when address of called function is a | |
| 7948 complicated expression | |
| 7949 * [67]7622: ICE with nested inline functions if function's address is | |
| 7950 taken | |
| 7951 * [68]7681: (fortran) ICE in compensate_edge, in reg-stack.c (also PR | |
| 7952 [69]9258) | |
| 7953 * [70]8031: (c++) ICE in code comparing typeids and casting from | |
| 7954 virtual base | |
| 7955 * [71]8275: ICE in simplify_subreg | |
| 7956 * [72]8332: (c++) builtin strlen/template interaction causes ICE | |
| 7957 * [73]8372: (c++) ICE on explicit call of destructor | |
| 7958 * [74]8439: (c, not c++) empty struct causes ICE | |
| 7959 * [75]8442: (c++) ICE with nested template classes | |
| 7960 * [76]8518: ICE when compiling mplayer ("extern inline" issue) | |
| 7961 * [77]8615: (c++) ICE with out-of-range character constant template | |
| 7962 argument | |
| 7963 * [78]8663: (c++) ICE in cp_expr_size, at cp-lang.c:307 | |
| 7964 * [79]8799: (c++) ICE: error reporting routines re-entered | |
| 7965 * [80]9328: (c++) ICE with typeof(X) for overloaded X | |
| 7966 * [81]9465: (preprocessor) cpp -traditional ICE on null bytes | |
| 7967 | |
| 7968 C++ (compiler and library) bugs | |
| 7969 | |
| 7970 * [82]47: scoping in nested classes is broken | |
| 7971 * [83]6745: problems with iostream rdbuf() member function | |
| 7972 * [84]8214: conversion from const char* const to char* sometimes | |
| 7973 accepted illegally | |
| 7974 * [85]8493: builtin strlen and overload resolution (same bug as | |
| 7975 [86]8332) | |
| 7976 * [87]8503: strange behaviour of function types | |
| 7977 * [88]8727: compiler confused by inheritance from an anonymous struct | |
| 7978 * [89]7445: poor performance of std::locale::classic() in | |
| 7979 multi-threaded applications | |
| 7980 * [90]8230: mishandling of overflow in vector<T>::resize | |
| 7981 * [91]8399: sync_with_stdio(false) breaks unformatted input | |
| 7982 * [92]8662: illegal access of private member of unnamed class is | |
| 7983 accepted | |
| 7984 * [93]8707: "make distclean" fails in libstdc++-v3 directory | |
| 7985 * [94]8708: __USE_MALLOC doesn't work | |
| 7986 * [95]8790: Use of non-thread-safe strtok in src/localename.cc | |
| 7987 * [96]8887: Bug in date formats with --enable-clocale=generic | |
| 7988 * [97]9076: Call Frame Instructions are not handled correctly during | |
| 7989 unwind operation | |
| 7990 * [98]9151: std::setprecision limited to 16 digits when outputting a | |
| 7991 double to a stream | |
| 7992 * [99]9168: codecvt<char, char, mbstate_t> overwrites output buffers | |
| 7993 * [100]9269: libstdc++ headers: explicit specialization of function | |
| 7994 must precede its first use | |
| 7995 * [101]9322: return value of basic_streambuf<>::getloc affected by | |
| 7996 locale::global | |
| 7997 * [102]9433: segfault in runtime support for dynamic_cast | |
| 7998 | |
| 7999 C and optimizer bugs | |
| 8000 | |
| 8001 * [103]8032: GCC incorrectly initializes static structs that have | |
| 8002 flexible arrays | |
| 8003 * [104]8639: simple arithmetic expression broken | |
| 8004 * [105]8794: optimization improperly eliminates certain expressions | |
| 8005 * [106]8832: traditional "asm volatile" code is illegally optimized | |
| 8006 * [107]8988: loop optimizer bug: with -O2, code is generated that | |
| 8007 segfaults (found on i386, bug present for all platforms) | |
| 8008 * [108]9492: structure copy clobbers subsequent stores to structure | |
| 8009 | |
| 8010 Objective-C bugs | |
| 8011 | |
| 8012 * [109]9267: Objective-C parser won't build with newer bison versions | |
| 8013 (e.g. 1.875) | |
| 8014 | |
| 8015 Ada bugs | |
| 8016 | |
| 8017 * [110]8344: Ada build problem due to conflict between gcc/final.o, | |
| 8018 gcc/ada/final.o | |
| 8019 | |
| 8020 Preprocessor bugs | |
| 8021 | |
| 8022 * [111]8524: _Pragma within macros is improperly expanded | |
| 8023 * [112]8880: __WCHAR_TYPE__ macro incorrectly set to "long int" with | |
| 8024 -fshort-wchar | |
| 8025 | |
| 8026 ARM-specific | |
| 8027 | |
| 8028 * [113]9090: arm ICE with >= -O2; regression from gcc-2.95 | |
| 8029 | |
| 8030 x86-specific (Intel/AMD) | |
| 8031 | |
| 8032 * [114]8588: ICE in extract_insn, at recog.c:NNNN (shift instruction) | |
| 8033 * [115]8599: loop unroll bug with -march=k6-3 | |
| 8034 * [116]9506: ABI breakage in structure return (affects BSD and | |
| 8035 Cygwin, but not GNU/Linux) | |
| 8036 | |
| 8037 FreeBSD 5.0 specific | |
| 8038 | |
| 8039 * [117]9484: GCC 3.2.1 Bootstrap failure on FreeBSD 5.0 | |
| 8040 | |
| 8041 RTEMS-specific | |
| 8042 | |
| 8043 * [118]9292: hppa1.1-rtems configurery problems | |
| 8044 * [119]9293: [m68k-elf/rtems] config/m68k/t-crtstuff bug | |
| 8045 * [120]9295: [mips-rtems] config/mips/rtems.h init/fini issue | |
| 8046 * [121]9296: gthr-rtems regression | |
| 8047 * [122]9316: powerpc-rtems: extending multilibs | |
| 8048 | |
| 8049 HP-PA specific | |
| 8050 | |
| 8051 * [123]9493: ICE with -O2 when building a simple function | |
| 8052 | |
| 8053 Documentation | |
| 8054 | |
| 8055 * [124]7341: hyperlink to gcov in GCC documentation doesn't work | |
| 8056 * [125]8947: Please add a warning about "-malign-double" in docs | |
| 8057 * [126]7448, [127]8882: typo cleanups | |
| 8058 __________________________________________________________________ | |
| 8059 | |
| 8060 GCC 3.2.1 | |
| 8061 | |
| 8062 3.2.1 adds a new warning, -Wabi. This option warns when GNU C++ | |
| 8063 generates code that is known not to be binary-compatible with the | |
| 8064 vendor-neutral ia32/ia64 ABI. Please consult the GCC manual, included | |
| 8065 in the distribution, for details. | |
| 8066 | |
| 8067 This release also removes an old GCC extension, "naming types", and the | |
| 8068 documentation now directs users to use a different GCC extension, | |
| 8069 __typeof__, instead. The feature had evidently been broken for a while. | |
| 8070 | |
| 8071 Otherwise, 3.2.1 is a bug fix release only; other than bug fixes and | |
| 8072 the new warning there are no new features that were not present in GCC | |
| 8073 3.2. | |
| 8074 | |
| 8075 In addition, the previous fix for [128]PR 7445 (poor performance of | |
| 8076 std::locale::classic() in multi-threaded applications) was reverted | |
| 8077 ("unfixed"), because the "fix" was not thread-safe. | |
| 8078 | |
| 8079 Bug Fixes | |
| 8080 | |
| 8081 This section lists the problem reports (PRs) from GCC's bug tracking | |
| 8082 system that are known to be fixed in the 3.2.1 release. This list might | |
| 8083 not be complete (that is, it is possible that some PRs that have been | |
| 8084 fixed are not listed here). As you can see, the number of bug fixes is | |
| 8085 quite large, so it is strongly recommended that users of earlier GCC | |
| 8086 3.x releases upgrade to GCC 3.2.1. | |
| 8087 | |
| 8088 Internal Compiler Errors (multi-platform) | |
| 8089 | |
| 8090 * [129]2521: (c++) ICE in build_ptrmemfunc, in cp/typeck.c | |
| 8091 * [130]5661: (c++) ICE instantiating template on array of unknown | |
| 8092 size (bad code) | |
| 8093 * [131]6419: (c++) ICE in make_decl_rtl for "longest" attribute on | |
| 8094 64-bit platforms | |
| 8095 * [132]6994: (c++) ICE in find_function_data | |
| 8096 * [133]7150: preprocessor: GCC -dM -E gives an ICE | |
| 8097 * [134]7160: ICE when optimizing branches without a return value | |
| 8098 * [135]7228: (c++) ICE when using member template and template | |
| 8099 function | |
| 8100 * [136]7266: (c++) ICE with -pedantic on missing typename | |
| 8101 * [137]7353: ICE from use of "Naming Types" extension, see above | |
| 8102 * [138]7411: ICE in instantiate_virtual_regs_1, in function.c | |
| 8103 * [139]7478: (c++) ICE on static_cast inside template | |
| 8104 * [140]7526: preprocessor core dump when _Pragma implies #pragma | |
| 8105 dependency | |
| 8106 * [141]7721: (c++) ICE on simple (but incorrect) template ([142]7803 | |
| 8107 is a duplicate) | |
| 8108 * [143]7754: (c++) ICE on union with template parameter | |
| 8109 * [144]7788: (c++) redeclaring a definition as an incomplete class | |
| 8110 causes ICE | |
| 8111 * [145]8031: (c++) ICE in comptypes, in cp/typeck.c | |
| 8112 * [146]8055: preprocessor dies with SIG11 when building FreeBSD | |
| 8113 kernel | |
| 8114 * [147]8067: (c++) ICE due to mishandling of __FUNCTION__ and related | |
| 8115 variables | |
| 8116 * [148]8134: (c++) ICE in force_store_init_value on legal code | |
| 8117 * [149]8149: (c++) ICE on incomplete type | |
| 8118 * [150]8160: (c++) ICE in build_modify_expr, in cp/typeck.c: array | |
| 8119 initialization | |
| 8120 | |
| 8121 C++ (compiler and library) bugs | |
| 8122 | |
| 8123 * [151]5607: No pointer adjustment in covariant return types | |
| 8124 * [152]6579: Infinite loop with statement expressions in member | |
| 8125 initialization | |
| 8126 * [153]6803: Default copy constructor bug in GCC 3.1 | |
| 8127 * [154]7176: g++ confused by friend and static member with same name | |
| 8128 * [155]7188: Segfault with template class and recursive (incorrect) | |
| 8129 initializer list | |
| 8130 * [156]7306: Regression: GCC 3.x fails to compile code with virtual | |
| 8131 inheritance if a method has a variable number of arguments | |
| 8132 * [157]7461: ctype<char>::classic_table() returns offset array on | |
| 8133 Cygwin | |
| 8134 * [158]7524: f(const float arg[3]) fails | |
| 8135 * [159]7584: Erroneous ambiguous base error on using declaration | |
| 8136 * [160]7676: Member template overloading problem | |
| 8137 * [161]7679: infinite loop when a right parenthesis is missing | |
| 8138 * [162]7811: default locale not taken from environment | |
| 8139 * [163]7961: compare( char *) implemented incorrectly in | |
| 8140 basic_string<> | |
| 8141 * [164]8071: basic_ostream::operator<<(streambuf*) loops forever if | |
| 8142 streambuf::underflow() leaves gptr() NULL (dups: [165]8127, | |
| 8143 [166]6745) | |
| 8144 * [167]8096: deque::at() throws std::range_error instead of | |
| 8145 std::out_of_range | |
| 8146 * [168]8127: cout << cin.rdbuf() infinite loop | |
| 8147 * [169]8218: Excessively large memory consumed for classes with large | |
| 8148 array members | |
| 8149 * [170]8287: GCC 3.2: Destructor called for non-constructed local | |
| 8150 object | |
| 8151 * [171]8347: empty vector range used in string construction causes | |
| 8152 core dump | |
| 8153 * [172]8348: fail() flag is set in istringstream when eof() flag is | |
| 8154 set | |
| 8155 * [173]8391: regression: infinite loop in cp/decl2.c(finish_file) | |
| 8156 | |
| 8157 C and optimizer bugs | |
| 8158 | |
| 8159 * [174]6627: -fno-align-functions doesn't seem to disable function | |
| 8160 alignment | |
| 8161 * [175]6631: life_analysis misoptimizes code to initialize fields of | |
| 8162 a structure | |
| 8163 * [176]7102: unsigned char division results in floating exception | |
| 8164 * [177]7120: Run once loop should *always* be unrolled | |
| 8165 (pessimization) | |
| 8166 * [178]7209: Bug involving array referencing and ?: operator | |
| 8167 * [179]7515: invalid inlining of global function with -O3 | |
| 8168 * [180]7814: incorrect scheduling for glibc-2.2.92 strcpy test | |
| 8169 * [181]8467: bug in sibling call optimization | |
| 8170 | |
| 8171 Preprocessor bugs | |
| 8172 | |
| 8173 * [182]4890: incorrect line markers from the traditional preprocessor | |
| 8174 * [183]7357: -M option omits system headers files (making it the same | |
| 8175 as -MM) | |
| 8176 * [184]7358: Changes to Sun's make Dependencies | |
| 8177 * [185]7602: C++ header files found in CPLUS_INCLUDE_PATH treated as | |
| 8178 C headers | |
| 8179 * [186]7862: Interrupting GCC -MD removes .d file but not .o | |
| 8180 * [187]8190: Failed compilation deletes -MD dependency file | |
| 8181 * [188]8524: _Pragma within macro is improperly expanded | |
| 8182 | |
| 8183 x86 specific (Intel/AMD) | |
| 8184 | |
| 8185 * [189]5351: (i686-only) function pass-by-value structure copy | |
| 8186 corrupts stack ([190]7591 is a duplicate) | |
| 8187 * [191]6845, [192]7034, [193]7124, [194]7174: ICE's with | |
| 8188 -march=pentium3/pentium2/athlon (these are all the same underlying | |
| 8189 bug, in MMX register use) | |
| 8190 * [195]7134, [196]7375, [197]7390: ICE with -march=athlon (maybe same | |
| 8191 as above?) | |
| 8192 * [198]6890: xmmintrin.h, _MM_TRANSPOSE4_PS is broken | |
| 8193 * [199]6981: wrong code in 64-bit manipulation on x86 | |
| 8194 * [200]7242: GCC -mcpu=pentium[23] doesn't define __tune_pentiumpro__ | |
| 8195 macro | |
| 8196 * [201]7396: ix86: cmpgt_ss, cmpge_ss, cmpngt_ss, and cmpnge_ss SSE | |
| 8197 intrinsics are broken | |
| 8198 * [202]7630: GCC 3.2 breaks on Mozilla 1.0's JS sources with | |
| 8199 -march=pentium4 | |
| 8200 * [203]7693: Typo in i386 mmintrin.h header | |
| 8201 * [204]7723: ICE - Pentium3 sse - GCC 3.2 | |
| 8202 * [205]7951: ICE on -march=pentium4 -O2 -mfpmath=sse | |
| 8203 * [206]8146: (i686 only) gcc 3.2 miscompiles gcc 2.95.3 | |
| 8204 | |
| 8205 PowerPC specific | |
| 8206 | |
| 8207 * [207]5967: GCC bug when profiling nested functions on powerpc | |
| 8208 * [208]6984: wrong code generated with -O2, -O3, -Os for do-while | |
| 8209 loop on PowerPC | |
| 8210 * [209]7114: PowerPC: ICE building strcoll.op from glibc-2.2.5 | |
| 8211 * [210]7130: miscompiled code for GCC-3.1 in powerpc linux with | |
| 8212 -funroll-all-loops | |
| 8213 * [211]7133: PowerPC ICE: unrecognizable insn | |
| 8214 * [212]7380: ICE in extract_insn, at recog.c:2148 | |
| 8215 * [213]8252: ICE on Altivec code with optimization turned on | |
| 8216 * [214]8451: Altivec ICE in GCC 3.2 | |
| 8217 | |
| 8218 HP/PA specific | |
| 8219 | |
| 8220 * [215]7250: __ashrdi3 returns wrong value on 32 bit hppa | |
| 8221 | |
| 8222 SPARC specific | |
| 8223 | |
| 8224 * [216]6668: when using --disable-multilib, libgcc_s.so is installed | |
| 8225 in the wrong place on sparc-solaris | |
| 8226 * [217]7151: ICE when compiling for UltraSPARC | |
| 8227 * [218]7335: SPARC: ICE in verify_wide_reg (flow.c:557) with long | |
| 8228 double and -O1 | |
| 8229 * [219]7842: [REGRESSION] SPARC code gen bug | |
| 8230 | |
| 8231 ARM specific | |
| 8232 | |
| 8233 * [220]7856: [arm] invalid offset in constant pool reference | |
| 8234 * [221]7967: optimization produces wrong code (ARM) | |
| 8235 | |
| 8236 Alpha specific | |
| 8237 | |
| 8238 * [222]7374: __builtin_fabsl broken on alpha | |
| 8239 | |
| 8240 IBM s390 specific | |
| 8241 | |
| 8242 * [223]7370: ICE in fixup_var_refs_1 on s390x | |
| 8243 * [224]7409: loop optimization bug on s390x-linux-gnu | |
| 8244 * [225]8232: s390x: ICE when using bcmp with int length argument | |
| 8245 | |
| 8246 SCO specific | |
| 8247 | |
| 8248 * [226]7623: SCO OpenServer build fails with machmode.def: undefined | |
| 8249 symbol: BITS_PER_UNIT | |
| 8250 | |
| 8251 m68k/Coldfire specific | |
| 8252 | |
| 8253 * [227]8314: crtbegin, crtend need to be multilib'ed for this | |
| 8254 platform | |
| 8255 | |
| 8256 Documentation | |
| 8257 | |
| 8258 * [228]761: Document some undocumented options | |
| 8259 * [229]5610: Fix documentation about invoking SSE instructions | |
| 8260 (-mfpmath=sse) | |
| 8261 * [230]7484: List -Wmissing-declarations as C-only option | |
| 8262 * [231]7531: -mcmodel not documented for x86-64 | |
| 8263 * [232]8120: Update documentation of bad use of ## | |
| 8264 __________________________________________________________________ | |
| 8265 | |
| 8266 GCC 3.2 | |
| 8267 | |
| 8268 3.2 is a small bug fix release, but there is a change to the | |
| 8269 application binary interface (ABI), hence the change to the second part | |
| 8270 of the version number. | |
| 8271 | |
| 8272 The main purpose of the 3.2 release is to correct a couple of problems | |
| 8273 in the C++ ABI, with the intention of providing a stable interface | |
| 8274 going forward. Accordingly, 3.2 is only a small change to 3.1.1. | |
| 8275 | |
| 8276 Bug Fixes | |
| 8277 | |
| 8278 C++ | |
| 8279 | |
| 8280 * [233]7320: g++ 3.2 relocation problem | |
| 8281 * [234]7470: vtable: virtual function pointers not in declaration | |
| 8282 order | |
| 8283 | |
| 8284 libstdc++ | |
| 8285 | |
| 8286 * [235]6410: Trouble with non-ASCII monetary symbols and wchar_t | |
| 8287 * [236]6503, [237]6642, [238]7186: Problems with comparing or | |
| 8288 subtracting various types of const and non-const iterators | |
| 8289 * [239]7216: ambiguity with basic_iostream::traits_type | |
| 8290 * [240]7220: problem with basic_istream::ignore(0,delimiter) | |
| 8291 * [241]7222: locale::operator==() doesn't work on std::locale("") | |
| 8292 * [242]7286: placement operator delete issue | |
| 8293 * [243]7442: cxxabi.h does not match the C++ ABI | |
| 8294 * [244]7445: poor performance of std::locale::classic() in | |
| 8295 multi-threaded applications | |
| 8296 | |
| 8297 x86-64 specific | |
| 8298 | |
| 8299 * [245]7291: off-by-one in generated inline bzero code for x86-64 | |
| 8300 | |
| 8301 Please send FSF & GNU inquiries & questions to [246]gnu@gnu.org. There | |
| 8302 are also [247]other ways to contact the FSF. | |
| 8303 | |
| 8304 These pages are maintained by [248]the GCC team. | |
| 8305 | |
| 8306 | |
| 8307 For questions related to the use of GCC, please consult these web | |
| 8308 pages and the [249]GCC manuals. If that fails, the | |
| 8309 [250]gcc-help@gcc.gnu.org mailing list might help. | |
| 8310 Please send comments on these web pages and the development of GCC to | |
| 8311 our developer mailing list at [251]gcc@gnu.org or | |
| 8312 [252]gcc@gcc.gnu.org. All of our lists have [253]public archives. | |
| 8313 | |
| 8314 Copyright (C) Free Software Foundation, Inc., 51 Franklin St, Fifth | |
| 8315 Floor, Boston, MA 02110, USA. | |
| 8316 | |
| 8317 Verbatim copying and distribution of this entire article is permitted | |
| 8318 in any medium, provided this notice is preserved. | |
| 8319 Last modified 2007-07-25 [254]Valid XHTML 1.0 | |
| 8320 | |
| 8321 References | |
| 8322 | |
| 8323 1. http://gcc.gnu.org/gcc-3.2/changes.html#3.2.3 | |
| 8324 2. http://gcc.gnu.org/gcc-3.1/changes.html | |
| 8325 3. http://gcc.gnu.org/gcc-3.2/c++-abi.html | |
| 8326 4. http://gcc.gnu.org/PR3782 | |
| 8327 5. http://gcc.gnu.org/PR6440 | |
| 8328 6. http://gcc.gnu.org/PR7050 | |
| 8329 7. http://gcc.gnu.org/PR7741 | |
| 8330 8. http://gcc.gnu.org/PR7982 | |
| 8331 9. http://gcc.gnu.org/PR8068 | |
| 8332 10. http://gcc.gnu.org/PR8178 | |
| 8333 11. http://gcc.gnu.org/PR8396 | |
| 8334 12. http://gcc.gnu.org/PR8674 | |
| 8335 13. http://gcc.gnu.org/PR9768 | |
| 8336 14. http://gcc.gnu.org/PR9798 | |
| 8337 15. http://gcc.gnu.org/PR9799 | |
| 8338 16. http://gcc.gnu.org/PR9928 | |
| 8339 17. http://gcc.gnu.org/PR10114 | |
| 8340 18. http://gcc.gnu.org/PR10352 | |
| 8341 19. http://gcc.gnu.org/PR10336 | |
| 8342 20. http://gcc.gnu.org/PR8224 | |
| 8343 21. http://gcc.gnu.org/PR8613 | |
| 8344 22. http://gcc.gnu.org/PR8828 | |
| 8345 23. http://gcc.gnu.org/PR9226 | |
| 8346 24. http://gcc.gnu.org/PR9853 | |
| 8347 25. http://gcc.gnu.org/PR9797 | |
| 8348 26. http://gcc.gnu.org/PR9967 | |
| 8349 27. http://gcc.gnu.org/PR10116 | |
| 8350 28. http://gcc.gnu.org/PR10171 | |
| 8351 29. http://gcc.gnu.org/PR10175 | |
| 8352 30. http://gcc.gnu.org/PR8316 | |
| 8353 31. http://gcc.gnu.org/PR9169 | |
| 8354 32. http://gcc.gnu.org/PR9420 | |
| 8355 33. http://gcc.gnu.org/PR9459 | |
| 8356 34. http://gcc.gnu.org/PR9507 | |
| 8357 35. http://gcc.gnu.org/PR9538 | |
| 8358 36. http://gcc.gnu.org/PR9602 | |
| 8359 37. http://gcc.gnu.org/PR9993 | |
| 8360 38. http://gcc.gnu.org/PR10167 | |
| 8361 39. http://gcc.gnu.org/PR9652 | |
| 8362 40. http://gcc.gnu.org/PR10144 | |
| 8363 41. http://gcc.gnu.org/PR8746 | |
| 8364 42. http://gcc.gnu.org/PR9888 | |
| 8365 43. http://gcc.gnu.org/PR9638 | |
| 8366 44. http://gcc.gnu.org/PR9954 | |
| 8367 45. http://gcc.gnu.org/PR7784 | |
| 8368 46. http://gcc.gnu.org/PR7796 | |
| 8369 47. http://gcc.gnu.org/PR8281 | |
| 8370 48. http://gcc.gnu.org/PR8366 | |
| 8371 49. http://gcc.gnu.org/PR8726 | |
| 8372 50. http://gcc.gnu.org/PR9414 | |
| 8373 51. http://gcc.gnu.org/PR10067 | |
| 8374 52. http://gcc.gnu.org/PR7248 | |
| 8375 53. http://gcc.gnu.org/PR8343 | |
| 8376 54. http://gcc.gnu.org/PR9732 | |
| 8377 55. http://gcc.gnu.org/PR10073 | |
| 8378 56. http://gcc.gnu.org/PR7702 | |
| 8379 57. http://gcc.gnu.org/PR9671 | |
| 8380 58. http://gcc.gnu.org/PR8694 | |
| 8381 59. http://gcc.gnu.org/PR9953 | |
| 8382 60. http://gcc.gnu.org/PR10271 | |
| 8383 61. http://gcc.gnu.org/PR6362 | |
| 8384 62. http://gcc.gnu.org/PR10377 | |
| 8385 63. http://gcc.gnu.org/PR6955 | |
| 8386 64. http://gcc.gnu.org/PR5919 | |
| 8387 65. http://gcc.gnu.org/PR7129 | |
| 8388 66. http://gcc.gnu.org/PR7507 | |
| 8389 67. http://gcc.gnu.org/PR7622 | |
| 8390 68. http://gcc.gnu.org/PR7681 | |
| 8391 69. http://gcc.gnu.org/PR9528 | |
| 8392 70. http://gcc.gnu.org/PR8031 | |
| 8393 71. http://gcc.gnu.org/PR8275 | |
| 8394 72. http://gcc.gnu.org/PR8332 | |
| 8395 73. http://gcc.gnu.org/PR8372 | |
| 8396 74. http://gcc.gnu.org/PR8439 | |
| 8397 75. http://gcc.gnu.org/PR8442 | |
| 8398 76. http://gcc.gnu.org/PR8518 | |
| 8399 77. http://gcc.gnu.org/PR8615 | |
| 8400 78. http://gcc.gnu.org/PR8663 | |
| 8401 79. http://gcc.gnu.org/PR8799 | |
| 8402 80. http://gcc.gnu.org/PR9328 | |
| 8403 81. http://gcc.gnu.org/PR9465 | |
| 8404 82. http://gcc.gnu.org/PR47 | |
| 8405 83. http://gcc.gnu.org/PR6745 | |
| 8406 84. http://gcc.gnu.org/PR8214 | |
| 8407 85. http://gcc.gnu.org/PR8493 | |
| 8408 86. http://gcc.gnu.org/PR8332 | |
| 8409 87. http://gcc.gnu.org/PR8503 | |
| 8410 88. http://gcc.gnu.org/PR8727 | |
| 8411 89. http://gcc.gnu.org/PR7445 | |
| 8412 90. http://gcc.gnu.org/PR8230 | |
| 8413 91. http://gcc.gnu.org/PR8399 | |
| 8414 92. http://gcc.gnu.org/PR8662 | |
| 8415 93. http://gcc.gnu.org/PR8707 | |
| 8416 94. http://gcc.gnu.org/PR8708 | |
| 8417 95. http://gcc.gnu.org/PR8790 | |
| 8418 96. http://gcc.gnu.org/PR8887 | |
| 8419 97. http://gcc.gnu.org/PR9076 | |
| 8420 98. http://gcc.gnu.org/PR9151 | |
| 8421 99. http://gcc.gnu.org/PR9168 | |
| 8422 100. http://gcc.gnu.org/PR9269 | |
| 8423 101. http://gcc.gnu.org/PR9322 | |
| 8424 102. http://gcc.gnu.org/PR9433 | |
| 8425 103. http://gcc.gnu.org/PR8032 | |
| 8426 104. http://gcc.gnu.org/PR8639 | |
| 8427 105. http://gcc.gnu.org/PR8794 | |
| 8428 106. http://gcc.gnu.org/PR8832 | |
| 8429 107. http://gcc.gnu.org/PR8988 | |
| 8430 108. http://gcc.gnu.org/PR9492 | |
| 8431 109. http://gcc.gnu.org/PR9267 | |
| 8432 110. http://gcc.gnu.org/PR8344 | |
| 8433 111. http://gcc.gnu.org/PR8524 | |
| 8434 112. http://gcc.gnu.org/PR8880 | |
| 8435 113. http://gcc.gnu.org/PR9090 | |
| 8436 114. http://gcc.gnu.org/PR8588 | |
| 8437 115. http://gcc.gnu.org/PR8599 | |
| 8438 116. http://gcc.gnu.org/PR9506 | |
| 8439 117. http://gcc.gnu.org/PR9484 | |
| 8440 118. http://gcc.gnu.org/PR9292 | |
| 8441 119. http://gcc.gnu.org/PR9293 | |
| 8442 120. http://gcc.gnu.org/PR9295 | |
| 8443 121. http://gcc.gnu.org/PR9296 | |
| 8444 122. http://gcc.gnu.org/PR9316 | |
| 8445 123. http://gcc.gnu.org/PR9493 | |
| 8446 124. http://gcc.gnu.org/PR7341 | |
| 8447 125. http://gcc.gnu.org/PR8947 | |
| 8448 126. http://gcc.gnu.org/PR7448 | |
| 8449 127. http://gcc.gnu.org/PR8882 | |
| 8450 128. http://gcc.gnu.org/PR7445 | |
| 8451 129. http://gcc.gnu.org/PR2521 | |
| 8452 130. http://gcc.gnu.org/PR5661 | |
| 8453 131. http://gcc.gnu.org/PR6419 | |
| 8454 132. http://gcc.gnu.org/PR6994 | |
| 8455 133. http://gcc.gnu.org/PR7150 | |
| 8456 134. http://gcc.gnu.org/PR7160 | |
| 8457 135. http://gcc.gnu.org/PR7228 | |
| 8458 136. http://gcc.gnu.org/PR7266 | |
| 8459 137. http://gcc.gnu.org/PR7353 | |
| 8460 138. http://gcc.gnu.org/PR7411 | |
| 8461 139. http://gcc.gnu.org/PR7478 | |
| 8462 140. http://gcc.gnu.org/PR7526 | |
| 8463 141. http://gcc.gnu.org/PR7721 | |
| 8464 142. http://gcc.gnu.org/PR7803 | |
| 8465 143. http://gcc.gnu.org/PR7754 | |
| 8466 144. http://gcc.gnu.org/PR7788 | |
| 8467 145. http://gcc.gnu.org/PR8031 | |
| 8468 146. http://gcc.gnu.org/PR8055 | |
| 8469 147. http://gcc.gnu.org/PR8067 | |
| 8470 148. http://gcc.gnu.org/PR8134 | |
| 8471 149. http://gcc.gnu.org/PR8149 | |
| 8472 150. http://gcc.gnu.org/PR8160 | |
| 8473 151. http://gcc.gnu.org/PR5607 | |
| 8474 152. http://gcc.gnu.org/PR6579 | |
| 8475 153. http://gcc.gnu.org/PR6803 | |
| 8476 154. http://gcc.gnu.org/PR7176 | |
| 8477 155. http://gcc.gnu.org/PR7188 | |
| 8478 156. http://gcc.gnu.org/PR7306 | |
| 8479 157. http://gcc.gnu.org/PR7461 | |
| 8480 158. http://gcc.gnu.org/PR7524 | |
| 8481 159. http://gcc.gnu.org/PR7584 | |
| 8482 160. http://gcc.gnu.org/PR7676 | |
| 8483 161. http://gcc.gnu.org/PR7679 | |
| 8484 162. http://gcc.gnu.org/PR7811 | |
| 8485 163. http://gcc.gnu.org/PR7961 | |
| 8486 164. http://gcc.gnu.org/PR8071 | |
| 8487 165. http://gcc.gnu.org/PR8127 | |
| 8488 166. http://gcc.gnu.org/PR6745 | |
| 8489 167. http://gcc.gnu.org/PR8096 | |
| 8490 168. http://gcc.gnu.org/PR8127 | |
| 8491 169. http://gcc.gnu.org/PR8218 | |
| 8492 170. http://gcc.gnu.org/PR8287 | |
| 8493 171. http://gcc.gnu.org/PR8347 | |
| 8494 172. http://gcc.gnu.org/PR8348 | |
| 8495 173. http://gcc.gnu.org/PR8391 | |
| 8496 174. http://gcc.gnu.org/PR6627 | |
| 8497 175. http://gcc.gnu.org/PR6631 | |
| 8498 176. http://gcc.gnu.org/PR7102 | |
| 8499 177. http://gcc.gnu.org/PR7120 | |
| 8500 178. http://gcc.gnu.org/PR7209 | |
| 8501 179. http://gcc.gnu.org/PR7515 | |
| 8502 180. http://gcc.gnu.org/PR7814 | |
| 8503 181. http://gcc.gnu.org/PR8467 | |
| 8504 182. http://gcc.gnu.org/PR4890 | |
| 8505 183. http://gcc.gnu.org/PR7357 | |
| 8506 184. http://gcc.gnu.org/PR7358 | |
| 8507 185. http://gcc.gnu.org/PR7602 | |
| 8508 186. http://gcc.gnu.org/PR7862 | |
| 8509 187. http://gcc.gnu.org/PR8190 | |
| 8510 188. http://gcc.gnu.org/PR8524 | |
| 8511 189. http://gcc.gnu.org/PR5351 | |
| 8512 190. http://gcc.gnu.org/PR7591 | |
| 8513 191. http://gcc.gnu.org/PR6845 | |
| 8514 192. http://gcc.gnu.org/PR7034 | |
| 8515 193. http://gcc.gnu.org/PR7124 | |
| 8516 194. http://gcc.gnu.org/PR7174 | |
| 8517 195. http://gcc.gnu.org/PR7134 | |
| 8518 196. http://gcc.gnu.org/PR7375 | |
| 8519 197. http://gcc.gnu.org/PR7390 | |
| 8520 198. http://gcc.gnu.org/PR6890 | |
| 8521 199. http://gcc.gnu.org/PR6981 | |
| 8522 200. http://gcc.gnu.org/PR7242 | |
| 8523 201. http://gcc.gnu.org/PR7396 | |
| 8524 202. http://gcc.gnu.org/PR7630 | |
| 8525 203. http://gcc.gnu.org/PR7693 | |
| 8526 204. http://gcc.gnu.org/PR7723 | |
| 8527 205. http://gcc.gnu.org/PR7951 | |
| 8528 206. http://gcc.gnu.org/PR8146 | |
| 8529 207. http://gcc.gnu.org/PR5967 | |
| 8530 208. http://gcc.gnu.org/PR6984 | |
| 8531 209. http://gcc.gnu.org/PR7114 | |
| 8532 210. http://gcc.gnu.org/PR7130 | |
| 8533 211. http://gcc.gnu.org/PR7133 | |
| 8534 212. http://gcc.gnu.org/PR7380 | |
| 8535 213. http://gcc.gnu.org/PR8252 | |
| 8536 214. http://gcc.gnu.org/PR8451 | |
| 8537 215. http://gcc.gnu.org/PR7250 | |
| 8538 216. http://gcc.gnu.org/PR6668 | |
| 8539 217. http://gcc.gnu.org/PR7151 | |
| 8540 218. http://gcc.gnu.org/PR7335 | |
| 8541 219. http://gcc.gnu.org/PR7842 | |
| 8542 220. http://gcc.gnu.org/PR7856 | |
| 8543 221. http://gcc.gnu.org/PR7967 | |
| 8544 222. http://gcc.gnu.org/PR7374 | |
| 8545 223. http://gcc.gnu.org/PR7370 | |
| 8546 224. http://gcc.gnu.org/PR7409 | |
| 8547 225. http://gcc.gnu.org/PR8232 | |
| 8548 226. http://gcc.gnu.org/PR7623 | |
| 8549 227. http://gcc.gnu.org/PR8314 | |
| 8550 228. http://gcc.gnu.org/PR761 | |
| 8551 229. http://gcc.gnu.org/PR5610 | |
| 8552 230. http://gcc.gnu.org/PR7484 | |
| 8553 231. http://gcc.gnu.org/PR7531 | |
| 8554 232. http://gcc.gnu.org/PR8120 | |
| 8555 233. http://gcc.gnu.org/PR7320 | |
| 8556 234. http://gcc.gnu.org/PR7470 | |
| 8557 235. http://gcc.gnu.org/PR6410 | |
| 8558 236. http://gcc.gnu.org/PR6503 | |
| 8559 237. http://gcc.gnu.org/PR6642 | |
| 8560 238. http://gcc.gnu.org/PR7186 | |
| 8561 239. http://gcc.gnu.org/PR7216 | |
| 8562 240. http://gcc.gnu.org/PR7220 | |
| 8563 241. http://gcc.gnu.org/PR7222 | |
| 8564 242. http://gcc.gnu.org/PR7286 | |
| 8565 243. http://gcc.gnu.org/PR7442 | |
| 8566 244. http://gcc.gnu.org/PR7445 | |
| 8567 245. http://gcc.gnu.org/PR7291 | |
| 8568 246. mailto:gnu@gnu.org | |
| 8569 247. http://www.gnu.org/home.html#ContactInfo | |
| 8570 248. http://gcc.gnu.org/about.html | |
| 8571 249. http://gcc.gnu.org/onlinedocs/ | |
| 8572 250. mailto:gcc-help@gcc.gnu.org | |
| 8573 251. mailto:gcc@gnu.org | |
| 8574 252. mailto:gcc@gcc.gnu.org | |
| 8575 253. http://gcc.gnu.org/lists.html | |
| 8576 254. http://validator.w3.org/check/referer | |
| 8577 ====================================================================== | |
| 8578 http://gcc.gnu.org/gcc-3.1/index.html | |
| 8579 GCC 3.1 | |
| 8580 | |
| 8581 July 27, 2002 | |
| 8582 | |
| 8583 The [1]GNU project and the GCC developers are pleased to announce the | |
| 8584 release of GCC 3.1.1. | |
| 8585 | |
| 8586 The links below still apply to GCC 3.1.1. | |
| 8587 | |
| 8588 May 15, 2002 | |
| 8589 | |
| 8590 The [2]GNU project and the GCC developers are pleased to announce the | |
| 8591 release of GCC 3.1. | |
| 8592 | |
| 8593 GCC used to stand for the GNU C Compiler, but since the compiler | |
| 8594 supports several other languages aside from C, it now stands for the | |
| 8595 GNU Compiler Collection. | |
| 8596 | |
| 8597 A list of [3]successful builds is updated as new information becomes | |
| 8598 available. | |
| 8599 | |
| 8600 The GCC developers would like to thank the numerous people that have | |
| 8601 contributed [4]new features, improvements, bug fixes, and other changes | |
| 8602 as well as test results to GCC. This [5]amazing group of volunteers is | |
| 8603 what makes GCC successful. | |
| 8604 | |
| 8605 For additional information about GCC please refer to the [6]GCC project | |
| 8606 web site or contact the [7]GCC development mailing list. | |
| 8607 | |
| 8608 To obtain GCC please use [8]our mirror sites, or our CVS server. | |
| 8609 __________________________________________________________________ | |
| 8610 | |
| 8611 Please send FSF & GNU inquiries & questions to [9]gnu@gnu.org. There | |
| 8612 are also [10]other ways to contact the FSF. | |
| 8613 | |
| 8614 These pages are maintained by [11]the GCC team. | |
| 8615 | |
| 8616 | |
| 8617 For questions related to the use of GCC, please consult these web | |
| 8618 pages and the [12]GCC manuals. If that fails, the | |
| 8619 [13]gcc-help@gcc.gnu.org mailing list might help. | |
| 8620 Please send comments on these web pages and the development of GCC to | |
| 8621 our developer mailing list at [14]gcc@gnu.org or [15]gcc@gcc.gnu.org. | |
| 8622 All of our lists have [16]public archives. | |
| 8623 | |
| 8624 Copyright (C) Free Software Foundation, Inc., 51 Franklin St, Fifth | |
| 8625 Floor, Boston, MA 02110, USA. | |
| 8626 | |
| 8627 Verbatim copying and distribution of this entire article is permitted | |
| 8628 in any medium, provided this notice is preserved. | |
| 8629 Last modified 2008-07-26 [17]Valid XHTML 1.0 | |
| 8630 | |
| 8631 References | |
| 8632 | |
| 8633 1. http://www.gnu.org/ | |
| 8634 2. http://www.gnu.org/ | |
| 8635 3. http://gcc.gnu.org/gcc-3.1/buildstat.html | |
| 8636 4. http://gcc.gnu.org/gcc-3.1/changes.html | |
| 8637 5. http://gcc.gnu.org/onlinedocs/gcc/Contributors.html | |
| 8638 6. http://gcc.gnu.org/index.html | |
| 8639 7. mailto:gcc@gcc.gnu.org | |
| 8640 8. http://gcc.gnu.org/mirrors.html | |
| 8641 9. mailto:gnu@gnu.org | |
| 8642 10. http://www.gnu.org/home.html#ContactInfo | |
| 8643 11. http://gcc.gnu.org/about.html | |
| 8644 12. http://gcc.gnu.org/onlinedocs/ | |
| 8645 13. mailto:gcc-help@gcc.gnu.org | |
| 8646 14. mailto:gcc@gnu.org | |
| 8647 15. mailto:gcc@gcc.gnu.org | |
| 8648 16. http://gcc.gnu.org/lists.html | |
| 8649 17. http://validator.w3.org/check/referer | |
| 8650 ====================================================================== | |
| 8651 http://gcc.gnu.org/gcc-3.1/changes.html | |
| 8652 GCC 3.1 Release Series | |
| 8653 Changes, New Features, and Fixes | |
| 8654 | |
| 8655 Additional changes in GCC 3.1.1 | |
| 8656 | |
| 8657 * A bug related to how structures and unions are returned has been | |
| 8658 fixed for powerpc-*-netbsd*. | |
| 8659 * An important bug in the implementation of -fprefetch-loop-arrays | |
| 8660 has been fixed. Previously the optimization prefetched random | |
| 8661 blocks of memory for most targets except for i386. | |
| 8662 * The Java compiler now compiles Java programs much faster and also | |
| 8663 works with parallel make. | |
| 8664 * Nested functions have been fixed for mips*-*-netbsd*. | |
| 8665 * Some missing floating point support routines have beed added for | |
| 8666 mips*-*-netbsd*. | |
| 8667 * This [1]message gives additional information about the bugs fixed | |
| 8668 in this release. | |
| 8669 | |
| 8670 Caveats | |
| 8671 | |
| 8672 * The -traditional C compiler option has been deprecated and will be | |
| 8673 removed in GCC 3.3. (It remains possible to preprocess non-C code | |
| 8674 with the traditional preprocessor.) | |
| 8675 * The default debugging format for most ELF platforms (including | |
| 8676 GNU/Linux and FreeBSD; notable exception is Solaris) has changed | |
| 8677 from stabs to DWARF2. This requires GDB 5.1.1 or later. | |
| 8678 | |
| 8679 General Optimizer Improvements | |
| 8680 | |
| 8681 * Jan Hubicka, SuSE Labs, together with Richard Henderson, Red Hat, | |
| 8682 and Andreas Jaeger, SuSE Labs, has contributed [2]infrastructure | |
| 8683 for profile driven optimizations. | |
| 8684 Options -fprofile-arcs and -fbranch-probabilities can now be used | |
| 8685 to improve speed of the generated code by profiling the actual | |
| 8686 program behaviour on typical runs. In the absence of profile info | |
| 8687 the compiler attempts to guess the profile statically. | |
| 8688 * [3]SPEC2000 and SPEC95 benchmark suites are now used daily to | |
| 8689 monitor performance of the generated code. | |
| 8690 According to the SPECInt2000 results on an AMD Athlon CPU, the code | |
| 8691 generated by GCC 3.1 is 6% faster on the average (8.2% faster with | |
| 8692 profile feedback) compared to GCC 3.0. The code produced by GCC 3.0 | |
| 8693 is about 2.1% faster compared to 2.95.3. Tests were done using the | |
| 8694 -O2 -march=athlon command-line options. | |
| 8695 * Alexandre Oliva, of Red Hat, has generalized the tree inlining | |
| 8696 infrastructure developed by CodeSourcery, LLC for the C++ front | |
| 8697 end, so that it is now used in the C front end too. Inlining | |
| 8698 functions as trees exposes them earlier to the compiler, giving it | |
| 8699 more opportunities for optimization. | |
| 8700 * Support for data prefetching instructions has been added to the GCC | |
| 8701 back end and several targets. A new __builtin_prefetch intrinsic is | |
| 8702 available to explicitly insert prefetch instructions and | |
| 8703 experimental support for loop array prefetching has been added (see | |
| 8704 -fprefetch-loop-array documentation). | |
| 8705 * Support for emitting debugging information for macros has been | |
| 8706 added for DWARF2. It is activated using -g3. | |
| 8707 | |
| 8708 New Languages and Language specific improvements | |
| 8709 | |
| 8710 C/C++ | |
| 8711 | |
| 8712 * A few more [4]ISO C99 features. | |
| 8713 * The preprocessor is 10-50% faster than the preprocessor in GCC 3.0. | |
| 8714 * The preprocessor's symbol table has been merged with the symbol | |
| 8715 table of the C, C++ and Objective-C front ends. | |
| 8716 * The preprocessor consumes less memory than the preprocessor in GCC | |
| 8717 3.0, often significantly so. On normal input files, it typically | |
| 8718 consumes less memory than pre-3.0 cccp-based GCC, too. | |
| 8719 | |
| 8720 C++ | |
| 8721 | |
| 8722 * -fhonor-std and -fno-honor-std have been removed. -fno-honor-std | |
| 8723 was a workaround to allow std compliant code to work with the | |
| 8724 non-std compliant libstdc++-v2. libstdc++-v3 is std compliant. | |
| 8725 * The C++ ABI has been fixed so that void (A::*)() const is mangled | |
| 8726 as "M1AKFvvE", rather than "MK1AFvvE" as before. This change only | |
| 8727 affects pointer to cv-qualified member function types. | |
| 8728 * The C++ ABI has been changed to correctly handle this code: | |
| 8729 struct A { | |
| 8730 void operator delete[] (void *, size_t); | |
| 8731 }; | |
| 8732 | |
| 8733 struct B : public A { | |
| 8734 }; | |
| 8735 | |
| 8736 new B[10]; | |
| 8737 | |
| 8738 The amount of storage allocated for the array will be greater than | |
| 8739 it was in 3.0, in order to store the number of elements in the | |
| 8740 array, so that the correct size can be passed to operator delete[] | |
| 8741 when the array is deleted. Previously, the value passed to operator | |
| 8742 delete[] was unpredictable. | |
| 8743 This change will only affect code that declares a two-argument | |
| 8744 operator delete[] with a second parameter of type size_t in a base | |
| 8745 class, and does not override that definition in a derived class. | |
| 8746 * The C++ ABI has been changed so that: | |
| 8747 struct A { | |
| 8748 void operator delete[] (void *, size_t); | |
| 8749 void operator delete[] (void *); | |
| 8750 }; | |
| 8751 | |
| 8752 does not cause unnecessary storage to be allocated when an array of | |
| 8753 A objects is allocated. | |
| 8754 This change will only affect code that declares both of these forms | |
| 8755 of operator delete[], and declared the two-argument form before the | |
| 8756 one-argument form. | |
| 8757 * The C++ ABI has been changed so that when a parameter is passed by | |
| 8758 value, any cleanup for that parameter is performed in the caller, | |
| 8759 as specified by the ia64 C++ ABI, rather than the called function | |
| 8760 as before. As a result, classes with a non-trivial destructor but a | |
| 8761 trivial copy constructor will be passed and returned by invisible | |
| 8762 reference, rather than by bitwise copy as before. | |
| 8763 * G++ now supports the "named return value optimization": for code | |
| 8764 like | |
| 8765 A f () { | |
| 8766 A a; | |
| 8767 ... | |
| 8768 return a; | |
| 8769 } | |
| 8770 | |
| 8771 G++ will allocate a in the return value slot, so that the return | |
| 8772 becomes a no-op. For this to work, all return statements in the | |
| 8773 function must return the same variable. | |
| 8774 * Improvements to the C++ library are listed in [5]the libstdc++-v3 | |
| 8775 FAQ. | |
| 8776 | |
| 8777 Objective-C | |
| 8778 | |
| 8779 * Annoying linker warnings (due to incorrect code being generated) | |
| 8780 have been fixed. | |
| 8781 * If a class method cannot be found, the compiler no longer issues a | |
| 8782 warning if a corresponding instance method exists in the root | |
| 8783 class. | |
| 8784 * Forward @protocol declarations have been fixed. | |
| 8785 * Loading of categories has been fixed in certain situations (GNU run | |
| 8786 time only). | |
| 8787 * The class lookup in the run-time library has been rewritten so that | |
| 8788 class method dispatch is more than twice as fast as it used to be | |
| 8789 (GNU run time only). | |
| 8790 | |
| 8791 Java | |
| 8792 | |
| 8793 * libgcj now includes RMI, java.lang.ref.*, javax.naming, and | |
| 8794 javax.transaction. | |
| 8795 * Property files and other system resources can be compiled into | |
| 8796 executables which use libgcj using the new gcj --resource feature. | |
| 8797 * libgcj has been ported to more platforms. In particular there is | |
| 8798 now a mostly-functional mingw32 (Windows) target port. | |
| 8799 * JNI and CNI invocation interfaces were implemented, so gcj-compiled | |
| 8800 Java code can now be called from a C/C++ application. | |
| 8801 * gcj can now use builtin functions for certain known methods, for | |
| 8802 instance Math.cos. | |
| 8803 * gcj can now automatically remove redundant array-store checks in | |
| 8804 some common cases. | |
| 8805 * The --no-store-checks optimization option was added. This can be | |
| 8806 used to omit runtime store checks for code which is known not to | |
| 8807 throw ArrayStoreException | |
| 8808 * The following third party interface standards were added to libgcj: | |
| 8809 org.w3c.dom and org.xml.sax. | |
| 8810 * java.security has been merged with GNU Classpath. The new package | |
| 8811 is now JDK 1.2 compliant, and much more complete. | |
| 8812 * A bytecode verifier was added to the libgcj interpreter. | |
| 8813 * java.lang.Character was rewritten to comply with the Unicode 3.0 | |
| 8814 standard, and improve performance. | |
| 8815 * Partial support for many more locales was added to libgcj. | |
| 8816 * Socket timeouts have been implemented. | |
| 8817 * libgcj has been merged into a single shared library. There are no | |
| 8818 longer separate shared libraries for the garbage collector and | |
| 8819 zlib. | |
| 8820 * Several performance improvements were made to gcj and libgcj: | |
| 8821 + Hash synchronization (thin locks) | |
| 8822 + A special allocation path for finalizer-free objects | |
| 8823 + Thread-local allocation | |
| 8824 + Parallel GC, and other GC tweaks | |
| 8825 | |
| 8826 Fortran | |
| 8827 | |
| 8828 Fortran improvements are listed in [6]the Fortran documentation. | |
| 8829 | |
| 8830 Ada | |
| 8831 | |
| 8832 [7]Ada Core Technologies, Inc, has contributed its GNAT Ada 95 front | |
| 8833 end and associated tools. The GNAT compiler fully implements the Ada | |
| 8834 language as defined by the ISO/IEC 8652 standard. | |
| 8835 | |
| 8836 Please note that the integration of the Ada front end is still work in | |
| 8837 progress. | |
| 8838 | |
| 8839 New Targets and Target Specific Improvements | |
| 8840 | |
| 8841 * Hans-Peter Nilsson has contributed a port to [8]MMIX, the CPU | |
| 8842 architecture used in new editions of Donald E. Knuth's The Art of | |
| 8843 Computer Programming. | |
| 8844 * [9]Axis Communications has contributed its port to the CRIS CPU | |
| 8845 architecture, used in the ETRAX system-on-a-chip series. See | |
| 8846 [10]Axis' developer site for technical information. | |
| 8847 * Alexandre Oliva, of Red Hat, has contributed a port to the | |
| 8848 [11]SuperH SH5 64-bit RISC microprocessor architecture, extending | |
| 8849 the existing SH port. | |
| 8850 * UltraSPARC is fully supported in 64-bit mode. The option -m64 | |
| 8851 enables it. | |
| 8852 * For compatibility with the Sun compiler #pragma redefine_extname | |
| 8853 has been implemented on Solaris. | |
| 8854 * The x86 back end has had some noticeable work done to it. | |
| 8855 + SuSE Labs developers Jan Hubicka, Bo Thorsen and Andreas | |
| 8856 Jaeger have contributed a port to the AMD x86-64 architecture. | |
| 8857 For more information on x86-64 see [12]http://www.x86-64.org. | |
| 8858 + The compiler now supports MMX, 3DNow!, SSE, and SSE2 | |
| 8859 instructions. Options -mmmx, -m3dnow, -msse, and -msse2 will | |
| 8860 enable the respective instruction sets. Intel C++ compatible | |
| 8861 MMX/3DNow!/SSE intrinsics are implemented. SSE2 intrinsics | |
| 8862 will be added in next major release. | |
| 8863 + Following those improvements, targets for Pentium MMX, K6-2, | |
| 8864 K6-3, Pentium III, Pentium 4, and Athlon 4 Mobile/XP/MP were | |
| 8865 added. Refer to the documentation on -march= and -mcpu= | |
| 8866 options for details. | |
| 8867 + For those targets that support it, -mfpmath=sse will cause the | |
| 8868 compiler to generate SSE/SSE2 instructions for floating point | |
| 8869 math instead of x87 instructions. Usually, this will lead to | |
| 8870 quicker code -- especially on the Pentium 4. Note that only | |
| 8871 scalar floating point instructions are used and GCC does not | |
| 8872 exploit SIMD features yet. | |
| 8873 + Prefetch support has been added to the Pentium III, Pentium 4, | |
| 8874 K6-2, K6-3, and Athlon series. | |
| 8875 + Code generated for floating point to integer conversions has | |
| 8876 been improved leading to better performance of many 3D | |
| 8877 applications. | |
| 8878 * The PowerPC back end has added 64-bit PowerPC GNU/Linux support. | |
| 8879 * C++ support for AIX has been improved. | |
| 8880 * Aldy Hernandez, of Red Hat, Inc, has contributed extensions to the | |
| 8881 PowerPC port supporting the AltiVec programming model (SIMD). The | |
| 8882 support, though presently useful, is experimental and is expected | |
| 8883 to stabilize for 3.2. The support is written to conform to | |
| 8884 Motorola's AltiVec specs. See -maltivec. | |
| 8885 | |
| 8886 Obsolete Systems | |
| 8887 | |
| 8888 Support for a number of older systems has been declared obsolete in GCC | |
| 8889 3.1. Unless there is activity to revive them, the next release of GCC | |
| 8890 will have their sources permanently removed. | |
| 8891 | |
| 8892 All configurations of the following processor architectures have been | |
| 8893 declared obsolete: | |
| 8894 * MIL-STD-1750A, 1750a-*-* | |
| 8895 * AMD A29k, a29k-*-* | |
| 8896 * Convex, c*-convex-* | |
| 8897 * Clipper, clipper-*-* | |
| 8898 * Elxsi, elxsi-*-* | |
| 8899 * Intel i860, i860-*-* | |
| 8900 * Sun picoJava, pj-*-* and pjl-*-* | |
| 8901 * Western Electric 32000, we32k-*-* | |
| 8902 | |
| 8903 Most configurations of the following processor architectures have been | |
| 8904 declared obsolete, but we are preserving a few systems which may have | |
| 8905 active developers. It is unlikely that the remaining systems will | |
| 8906 survive much longer unless we see definite signs of port activity. | |
| 8907 * Motorola 88000 except | |
| 8908 + Generic a.out, m88k-*-aout* | |
| 8909 + Generic SVR4, m88k-*-sysv4 | |
| 8910 + OpenBSD, m88k-*-openbsd* | |
| 8911 * NS32k except | |
| 8912 + NetBSD, ns32k-*-netbsd* | |
| 8913 + OpenBSD, ns32k-*-openbsd*. | |
| 8914 * ROMP except | |
| 8915 + OpenBSD, romp-*-openbsd*. | |
| 8916 | |
| 8917 Finally, only some configurations of these processor architectures are | |
| 8918 being obsoleted. | |
| 8919 * Alpha: | |
| 8920 + OSF/1, alpha*-*-osf[123]*. (Digital Unix and Tru64 Unix, aka | |
| 8921 alpha*-*-osf[45], are still supported.) | |
| 8922 * ARM: | |
| 8923 + RISCiX, arm-*-riscix*. | |
| 8924 * i386: | |
| 8925 + 386BSD, i?86-*-bsd* | |
| 8926 + Chorus, i?86-*-chorusos* | |
| 8927 + DG/UX, i?86-*-dgux* | |
| 8928 + FreeBSD 1.x, i?86-*-freebsd1.* | |
| 8929 + IBM AIX, i?86-*-aix* | |
| 8930 + ISC UNIX, i?86-*-isc* | |
| 8931 + Linux with pre-BFD linker, i?86-*-linux*oldld* | |
| 8932 + NEXTstep, i?86-next-* | |
| 8933 + OSF UNIX, i?86-*-osf1* and i?86-*-osfrose* | |
| 8934 + RTEMS/coff, i?86-*-rtemscoff* | |
| 8935 + RTEMS/go32, i?86-go32-rtems* | |
| 8936 + Sequent/BSD, i?86-sequent-bsd* | |
| 8937 + Sequent/ptx before version 3, i?86-sequent-ptx[12]* and | |
| 8938 i?86-sequent-sysv3* | |
| 8939 + SunOS, i?86-*-sunos* | |
| 8940 * Motorola 68000: | |
| 8941 + Altos, m68[k0]*-altos-* | |
| 8942 + Apollo, m68[k0]*-apollo-* | |
| 8943 + Apple A/UX, m68[k0]*-apple-* | |
| 8944 + Bull, m68[k0]*-bull-* | |
| 8945 + Convergent, m68[k0]*-convergent-* | |
| 8946 + Generic SVR3, m68[k0]*-*-sysv3* | |
| 8947 + ISI, m68[k0]*-isi-* | |
| 8948 + LynxOS, m68[k0]*-*-lynxos* | |
| 8949 + NEXT, m68[k0]*-next-* | |
| 8950 + RTEMS/coff, m68[k0]*-*-rtemscoff* | |
| 8951 + Sony, m68[k0]*-sony-* | |
| 8952 * MIPS: | |
| 8953 + DEC Ultrix, mips-*-ultrix* and mips-dec-* | |
| 8954 + Generic BSD, mips-*-bsd* | |
| 8955 + Generic System V, mips-*-sysv* | |
| 8956 + IRIX before version 5, mips-sgi-irix[1234]* | |
| 8957 + RiscOS, mips-*-riscos* | |
| 8958 + Sony, mips-sony-* | |
| 8959 + Tandem, mips-tandem-* | |
| 8960 * SPARC: | |
| 8961 + RTEMS/a.out, sparc-*-rtemsaout*. | |
| 8962 | |
| 8963 Documentation improvements | |
| 8964 | |
| 8965 * The old manual ("Using and Porting the GNU Compiler Collection") | |
| 8966 has been replaced by a users manual ("Using the GNU Compiler | |
| 8967 Collection") and a separate internals reference manual ("GNU | |
| 8968 Compiler Collection Internals"). | |
| 8969 * More complete and much improved documentation about GCC's internal | |
| 8970 representation used by the C and C++ front ends. | |
| 8971 * Many cleanups and improvements in general. | |
| 8972 | |
| 8973 Please send FSF & GNU inquiries & questions to [13]gnu@gnu.org. There | |
| 8974 are also [14]other ways to contact the FSF. | |
| 8975 | |
| 8976 These pages are maintained by [15]the GCC team. | |
| 8977 | |
| 8978 | |
| 8979 For questions related to the use of GCC, please consult these web | |
| 8980 pages and the [16]GCC manuals. If that fails, the | |
| 8981 [17]gcc-help@gcc.gnu.org mailing list might help. | |
| 8982 Please send comments on these web pages and the development of GCC to | |
| 8983 our developer mailing list at [18]gcc@gnu.org or [19]gcc@gcc.gnu.org. | |
| 8984 All of our lists have [20]public archives. | |
| 8985 | |
| 8986 Copyright (C) Free Software Foundation, Inc., 51 Franklin St, Fifth | |
| 8987 Floor, Boston, MA 02110, USA. | |
| 8988 | |
| 8989 Verbatim copying and distribution of this entire article is permitted | |
| 8990 in any medium, provided this notice is preserved. | |
| 8991 Last modified 2008-05-03 [21]Valid XHTML 1.0 | |
| 8992 | |
| 8993 References | |
| 8994 | |
| 8995 1. http://gcc.gnu.org/ml/gcc/2002-07/msg01208.html | |
| 8996 2. http://gcc.gnu.org/news/profiledriven.html | |
| 8997 3. http://gcc.gnu.org/benchmarks/ | |
| 8998 4. http://gcc.gnu.org/gcc-3.1/c99status.html | |
| 8999 5. http://gcc.gnu.org/onlinedocs/libstdc++/faq.html | |
| 9000 6. http://gcc.gnu.org/onlinedocs/gcc-3.1.1/g77/News.html | |
| 9001 7. http://www.adacore.com/home/ | |
| 9002 8. http://www-cs-faculty.stanford.edu/~knuth/mmix.html | |
| 9003 9. http://www.axis.com/ | |
| 9004 10. http://developer.axis.com/ | |
| 9005 11. http://www.superh.com/ | |
| 9006 12. http://www.x86-64.org/ | |
| 9007 13. mailto:gnu@gnu.org | |
| 9008 14. http://www.gnu.org/home.html#ContactInfo | |
| 9009 15. http://gcc.gnu.org/about.html | |
| 9010 16. http://gcc.gnu.org/onlinedocs/ | |
| 9011 17. mailto:gcc-help@gcc.gnu.org | |
| 9012 18. mailto:gcc@gnu.org | |
| 9013 19. mailto:gcc@gcc.gnu.org | |
| 9014 20. http://gcc.gnu.org/lists.html | |
| 9015 21. http://validator.w3.org/check/referer | |
| 9016 ====================================================================== | |
| 9017 http://gcc.gnu.org/gcc-3.0/gcc-3.0.html | |
| 9018 GCC 3.0.4 | |
| 9019 | |
| 9020 February 20, 2002 | |
| 9021 | |
| 9022 The [1]GNU project and the GCC developers are pleased to announce the | |
| 9023 release of GCC 3.0.4, which is a bug-fix release for the GCC 3.0 | |
| 9024 series. | |
| 9025 | |
| 9026 GCC used to stand for the GNU C Compiler, but since the compiler | |
| 9027 supports several other languages aside from C, it now stands for the | |
| 9028 GNU Compiler Collection. | |
| 9029 | |
| 9030 GCC 3.0.x has several new optimizations, new targets, new languages and | |
| 9031 many other new features, relative to GCC 2.95.x. See the [2]new | |
| 9032 features page for a more complete list. | |
| 9033 | |
| 9034 A list of [3]successful builds is updated as new information becomes | |
| 9035 available. | |
| 9036 | |
| 9037 The GCC developers would like to thank the numerous people that have | |
| 9038 contributed new features, test results, bug fixes, etc to GCC. This | |
| 9039 [4]amazing group of volunteers is what makes GCC successful. | |
| 9040 | |
| 9041 And finally, we can't in good conscience fail to mention some | |
| 9042 [5]caveats to using GCC 3.0.x. | |
| 9043 | |
| 9044 For additional information about GCC please refer to the [6]GCC project | |
| 9045 web site or contact the [7]GCC development mailing list. | |
| 9046 | |
| 9047 To obtain GCC please use [8]our mirror sites, or our CVS server. | |
| 9048 __________________________________________________________________ | |
| 9049 | |
| 9050 Previous 3.0.x Releases | |
| 9051 | |
| 9052 December 20, 2001: GCC 3.0.3 has been released. | |
| 9053 October 25, 2001: GCC 3.0.2 has been released. | |
| 9054 August 20, 2001: GCC 3.0.1 has been released. | |
| 9055 June 18, 2001: GCC 3.0 has been released. | |
| 9056 | |
| 9057 Please send FSF & GNU inquiries & questions to [9]gnu@gnu.org. There | |
| 9058 are also [10]other ways to contact the FSF. | |
| 9059 | |
| 9060 These pages are maintained by [11]the GCC team. | |
| 9061 | |
| 9062 | |
| 9063 For questions related to the use of GCC, please consult these web | |
| 9064 pages and the [12]GCC manuals. If that fails, the | |
| 9065 [13]gcc-help@gcc.gnu.org mailing list might help. | |
| 9066 Please send comments on these web pages and the development of GCC to | |
| 9067 our developer mailing list at [14]gcc@gnu.org or [15]gcc@gcc.gnu.org. | |
| 9068 All of our lists have [16]public archives. | |
| 9069 | |
| 9070 Copyright (C) Free Software Foundation, Inc., 51 Franklin St, Fifth | |
| 9071 Floor, Boston, MA 02110, USA. | |
| 9072 | |
| 9073 Verbatim copying and distribution of this entire article is permitted | |
| 9074 in any medium, provided this notice is preserved. | |
| 9075 Last modified 2008-07-26 [17]Valid XHTML 1.0 | |
| 9076 | |
| 9077 References | |
| 9078 | |
| 9079 1. http://www.gnu.org/ | |
| 9080 2. http://gcc.gnu.org/gcc-3.0/features.html | |
| 9081 3. http://gcc.gnu.org/gcc-3.0/buildstat.html | |
| 9082 4. http://gcc.gnu.org/onlinedocs/gcc/Contributors.html | |
| 9083 5. http://gcc.gnu.org/gcc-3.0/caveats.html | |
| 9084 6. http://gcc.gnu.org/index.html | |
| 9085 7. mailto:gcc@gcc.gnu.org | |
| 9086 8. http://gcc.gnu.org/mirrors.html | |
| 9087 9. mailto:gnu@gnu.org | |
| 9088 10. http://www.gnu.org/home.html#ContactInfo | |
| 9089 11. http://gcc.gnu.org/about.html | |
| 9090 12. http://gcc.gnu.org/onlinedocs/ | |
| 9091 13. mailto:gcc-help@gcc.gnu.org | |
| 9092 14. mailto:gcc@gnu.org | |
| 9093 15. mailto:gcc@gcc.gnu.org | |
| 9094 16. http://gcc.gnu.org/lists.html | |
| 9095 17. http://validator.w3.org/check/referer | |
| 9096 ====================================================================== | |
| 9097 http://gcc.gnu.org/gcc-3.0/features.html | |
| 9098 GCC 3.0 New Features | |
| 9099 | |
| 9100 Additional changes in GCC 3.0.4 | |
| 9101 | |
| 9102 * GCC 3.0 now supports newer versions of the [1]NetBSD operating | |
| 9103 system, which use the ELF object file format, on x86 processors. | |
| 9104 * Correct debugging information is generated from functions that have | |
| 9105 lines from multiple files (e.g. yacc output). | |
| 9106 * A fix for whitespace handling in the -traditional preprocessor, | |
| 9107 which can affect Fortran. | |
| 9108 * Fixes to the exception handling runtime. | |
| 9109 * More fixes for bad code generation in C++. | |
| 9110 * A fix for shared library generation under AIX 4.3. | |
| 9111 * Documentation updates. | |
| 9112 * Port of GCC to Tensilica's Xtensa processor contributed. | |
| 9113 * A fix for compiling the PPC Linux kernel (FAT fs wouldn't link). | |
| 9114 | |
| 9115 Additional changes in GCC 3.0.3 | |
| 9116 | |
| 9117 * A fix to correct an accidental change to the PowerPC ABI. | |
| 9118 * Fixes for bad code generation on a variety of architectures. | |
| 9119 * Improvements to the debugging information generated for C++ | |
| 9120 classes. | |
| 9121 * Fixes for bad code generation in C++. | |
| 9122 * A fix to avoid crashes in the C++ demangler. | |
| 9123 * A fix to the C++ standard library to avoid buffer overflows. | |
| 9124 * Miscellaneous improvements for a variety of architectures. | |
| 9125 | |
| 9126 Additional changes in GCC 3.0.2 | |
| 9127 | |
| 9128 * Fixes for bad code generation during loop unrolling. | |
| 9129 * Fixes for bad code generation by the sibling call optimization. | |
| 9130 * Minor improvements to x86 code generation. | |
| 9131 * Implementation of function descriptors in C++ vtables for IA64. | |
| 9132 * Numerous minor bug-fixes. | |
| 9133 | |
| 9134 Additional changes in GCC 3.0.1 | |
| 9135 | |
| 9136 * C++ fixes for incorrect code-generation. | |
| 9137 * Improved cross-compiling support for the C++ standard library. | |
| 9138 * Fixes for some embedded targets that worked in GCC 2.95.3, but not | |
| 9139 in GCC 3.0. | |
| 9140 * Fixes for various exception-handling bugs. | |
| 9141 * A port to the S/390 architecture. | |
| 9142 | |
| 9143 General Optimizer Improvements | |
| 9144 | |
| 9145 * [2]Basic block reordering pass. | |
| 9146 * New if-conversion pass with support for conditional (predicated) | |
| 9147 execution. | |
| 9148 * New tail call and sibling call elimination optimizations. | |
| 9149 * New register renaming pass. | |
| 9150 * New (experimental) [3]static single assignment (SSA) representation | |
| 9151 support. | |
| 9152 * New dead-code elimination pass implemented using the SSA | |
| 9153 representation. | |
| 9154 * [4]Global null pointer test elimination. | |
| 9155 * [5]Global code hoisting/unification. | |
| 9156 * More builtins and optimizations for stdio.h, string.h and old BSD | |
| 9157 functions, as well as for ISO C99 functions. | |
| 9158 * New builtin __builtin_expect for giving hints to the branch | |
| 9159 predictor. | |
| 9160 | |
| 9161 New Languages and Language specific improvements | |
| 9162 | |
| 9163 * The GNU Compiler for the Java(TM) language (GCJ) is now integrated | |
| 9164 and supported, including the run-time library containing most | |
| 9165 common non-GUI Java classes, a bytecode interpreter, and the Boehm | |
| 9166 conservative garbage collector. Many bugs have been fixed. GCJ can | |
| 9167 compile Java source or Java bytecodes to either native code or Java | |
| 9168 class files, and supports native methods written in either the | |
| 9169 standard JNI or the more efficient and convenient CNI. | |
| 9170 * Here is a [6]partial list of C++ improvements, both new features | |
| 9171 and those no longer supported. | |
| 9172 * New C++ ABI. On the IA-64 platform GCC is capable of | |
| 9173 inter-operating with other IA-64 compilers. | |
| 9174 * The new ABI also significantly reduces the size of symbol and debug | |
| 9175 information. | |
| 9176 * New [7]C++ support library and many C++ bug fixes, vastly improving | |
| 9177 our conformance to the ISO C++ standard. | |
| 9178 * New [8]inliner for C++. | |
| 9179 * Rewritten C preprocessor, integrated into the C, C++ and Objective | |
| 9180 C compilers, with very many improvements including ISO C99 support | |
| 9181 and [9]improvements to dependency generation. | |
| 9182 * Support for more [10]ISO C99 features. | |
| 9183 * Many improvements to support for checking calls to format functions | |
| 9184 such as printf and scanf, including support for ISO C99 format | |
| 9185 features, extensions from the Single Unix Specification and GNU | |
| 9186 libc 2.2, checking of strfmon formats and features to assist in | |
| 9187 auditing for format string security bugs. | |
| 9188 * New warnings for C code that may have undefined semantics because | |
| 9189 of violations of sequence point rules in the C standard (such as a | |
| 9190 = a++;, a[n] = b[n++]; and a[i++] = i;), included in -Wall. | |
| 9191 * Additional warning option -Wfloat-equal. | |
| 9192 * Improvements to -Wtraditional. | |
| 9193 * Fortran improvements are listed in [11]the Fortran documentation. | |
| 9194 | |
| 9195 New Targets and Target Specific Improvements | |
| 9196 | |
| 9197 * New x86 back-end, generating much improved code. | |
| 9198 * Support for a generic i386-elf target contributed. | |
| 9199 * New option to emit x86 assembly code using Intel style syntax | |
| 9200 (-mintel-syntax). | |
| 9201 * HPUX 11 support contributed. | |
| 9202 * Improved PowerPC code generation, including scheduled prologue and | |
| 9203 epilogue. | |
| 9204 * Port of GCC to Intel's IA-64 processor contributed. | |
| 9205 * Port of GCC to Motorola's MCore 210 and 340 contributed. | |
| 9206 * New unified back-end for Arm, Thumb and StrongArm contributed. | |
| 9207 * Port of GCC to Intel's XScale processor contributed. | |
| 9208 * Port of GCC to Atmel's AVR microcontrollers contributed. | |
| 9209 * Port of GCC to Mitsubishi's D30V processor contributed. | |
| 9210 * Port of GCC to Matsushita's AM33 processor (a member of the MN10300 | |
| 9211 processor family) contributed. | |
| 9212 * Port of GCC to Fujitsu's FR30 processor contributed. | |
| 9213 * Port of GCC to Motorola's 68HC11 and 68HC12 processors contributed. | |
| 9214 * Port of GCC to Sun's picoJava processor core contributed. | |
| 9215 | |
| 9216 Documentation improvements | |
| 9217 | |
| 9218 * Substantially rewritten and improved C preprocessor manual. | |
| 9219 * Many improvements to other documentation. | |
| 9220 * Manpages for gcc, cpp and gcov are now generated automatically from | |
| 9221 the master Texinfo manual, eliminating the problem of manpages | |
| 9222 being out of date. (The generated manpages are only extracts from | |
| 9223 the full manual, which is provided in Texinfo form, from which | |
| 9224 info, HTML, other formats and a printed manual can be generated.) | |
| 9225 * Generated info files are included in the release tarballs alongside | |
| 9226 their Texinfo sources, avoiding problems on some platforms with | |
| 9227 building makeinfo as part of the GCC distribution. | |
| 9228 | |
| 9229 Other significant improvements | |
| 9230 | |
| 9231 * Garbage collection used internally by the compiler for most memory | |
| 9232 allocation instead of obstacks. | |
| 9233 * Lengauer and Tarjan algorithm used for computing dominators in the | |
| 9234 CFG. This algorithm can be significantly faster and more space | |
| 9235 efficient than our older algorithm. | |
| 9236 * gccbug script provided to assist in submitting bug reports to our | |
| 9237 bug tracking system. (Bug reports previously submitted directly to | |
| 9238 our mailing lists, for which you received no bug tracking number, | |
| 9239 should be submitted again using gccbug if you can reproduce the | |
| 9240 problem with GCC 3.0.) | |
| 9241 * The internal libgcc library is [12]built as a shared library on | |
| 9242 systems that support it. | |
| 9243 * Extensive testsuite included with GCC, with many new tests. In | |
| 9244 addition to tests for GCC bugs that have been fixed, many tests | |
| 9245 have been added for language features, compiler warnings and | |
| 9246 builtin functions. | |
| 9247 * Additional language-independent warning options -Wpacked, -Wpadded, | |
| 9248 -Wunreachable-code and -Wdisabled-optimization. | |
| 9249 * Target-independent options -falign-functions, -falign-loops and | |
| 9250 -falign-jumps. | |
| 9251 | |
| 9252 Plus a great many bug fixes and almost all the [13]features found in | |
| 9253 GCC 2.95. | |
| 9254 | |
| 9255 Please send FSF & GNU inquiries & questions to [14]gnu@gnu.org. There | |
| 9256 are also [15]other ways to contact the FSF. | |
| 9257 | |
| 9258 These pages are maintained by [16]the GCC team. | |
| 9259 | |
| 9260 | |
| 9261 For questions related to the use of GCC, please consult these web | |
| 9262 pages and the [17]GCC manuals. If that fails, the | |
| 9263 [18]gcc-help@gcc.gnu.org mailing list might help. | |
| 9264 Please send comments on these web pages and the development of GCC to | |
| 9265 our developer mailing list at [19]gcc@gnu.org or [20]gcc@gcc.gnu.org. | |
| 9266 All of our lists have [21]public archives. | |
| 9267 | |
| 9268 Copyright (C) Free Software Foundation, Inc., 51 Franklin St, Fifth | |
| 9269 Floor, Boston, MA 02110, USA. | |
| 9270 | |
| 9271 Verbatim copying and distribution of this entire article is permitted | |
| 9272 in any medium, provided this notice is preserved. | |
| 9273 Last modified 2007-07-25 [22]Valid XHTML 1.0 | |
| 9274 | |
| 9275 References | |
| 9276 | |
| 9277 1. http://www.netbsd.org/ | |
| 9278 2. http://gcc.gnu.org/news/reorder.html | |
| 9279 3. http://gcc.gnu.org/news/ssa.html | |
| 9280 4. http://gcc.gnu.org/news/null.html | |
| 9281 5. http://gcc.gnu.org/news/unify.html | |
| 9282 6. http://gcc.gnu.org/gcc-3.0/c++features.html | |
| 9283 7. http://gcc.gnu.org/libstdc++/ | |
| 9284 8. http://gcc.gnu.org/news/inlining.html | |
| 9285 9. http://gcc.gnu.org/news/dependencies.html | |
| 9286 10. http://gcc.gnu.org/gcc-3.0/c99status.html | |
| 9287 11. http://gcc.gnu.org/onlinedocs/gcc-3.4.6/g77/News.html | |
| 9288 12. http://gcc.gnu.org/gcc-3.0/libgcc.html | |
| 9289 13. http://gcc.gnu.org/gcc-2.95/features.html | |
| 9290 14. mailto:gnu@gnu.org | |
| 9291 15. http://www.gnu.org/home.html#ContactInfo | |
| 9292 16. http://gcc.gnu.org/about.html | |
| 9293 17. http://gcc.gnu.org/onlinedocs/ | |
| 9294 18. mailto:gcc-help@gcc.gnu.org | |
| 9295 19. mailto:gcc@gnu.org | |
| 9296 20. mailto:gcc@gcc.gnu.org | |
| 9297 21. http://gcc.gnu.org/lists.html | |
| 9298 22. http://validator.w3.org/check/referer | |
| 9299 ====================================================================== | |
| 9300 http://gcc.gnu.org/gcc-3.0/caveats.html | |
| 9301 GCC 3.0 Caveats | |
| 9302 | |
| 9303 * -fstrict-aliasing is now part of -O2 and higher optimization | |
| 9304 levels. This allows the compiler to assume the strictest aliasing | |
| 9305 rules applicable to the language being compiled. For C and C++, | |
| 9306 this activates optimizations based on the type of expressions. This | |
| 9307 optimization may thus break old, non-compliant code. | |
| 9308 * Enumerations are now properly promoted to int in function | |
| 9309 parameters and function returns. Normally this change is not | |
| 9310 visible, but when using -fshort-enums this is an ABI change. | |
| 9311 * The undocumented extension that allowed C programs to have a label | |
| 9312 at the end of a compound statement has been deprecated and may be | |
| 9313 removed in a future version. Programs that now generate a warning | |
| 9314 about this may be fixed by adding a null statement (a single | |
| 9315 semicolon) after the label. | |
| 9316 * The poorly documented extension that allowed string constants in C, | |
| 9317 C++ and Objective C to contain unescaped newlines has been | |
| 9318 deprecated and may be removed in a future version. Programs using | |
| 9319 this extension may be fixed in several ways: the bare newline may | |
| 9320 be replaced by \n, or preceded by \n\, or string concatenation may | |
| 9321 be used with the bare newline preceded by \n" and " placed at the | |
| 9322 start of the next line. | |
| 9323 * The Chill compiler is not included in GCC 3.0, because of the lack | |
| 9324 of a volunteer to convert it to use garbage collection. | |
| 9325 * Certain non-standard iostream methods from earlier versions of | |
| 9326 libstdc++ are not included in libstdc++ v3, i.e. filebuf::attach, | |
| 9327 ostream::form, and istream::gets. | |
| 9328 * The new C++ ABI is not yet fully supported by current (as of | |
| 9329 2001-07-01) releases and development versions of GDB, or any | |
| 9330 earlier versions. There is a problem setting breakpoints by line | |
| 9331 number, and other related issues that have been fixed in GCC 3.0 | |
| 9332 but not yet handled in GDB: | |
| 9333 [1]http://gcc.gnu.org/ml/gcc-bugs/2001-06/msg00421.html | |
| 9334 | |
| 9335 Please send FSF & GNU inquiries & questions to [2]gnu@gnu.org. There | |
| 9336 are also [3]other ways to contact the FSF. | |
| 9337 | |
| 9338 These pages are maintained by [4]the GCC team. | |
| 9339 | |
| 9340 | |
| 9341 For questions related to the use of GCC, please consult these web | |
| 9342 pages and the [5]GCC manuals. If that fails, the | |
| 9343 [6]gcc-help@gcc.gnu.org mailing list might help. | |
| 9344 Please send comments on these web pages and the development of GCC to | |
| 9345 our developer mailing list at [7]gcc@gnu.org or [8]gcc@gcc.gnu.org. | |
| 9346 All of our lists have [9]public archives. | |
| 9347 | |
| 9348 Copyright (C) Free Software Foundation, Inc., 51 Franklin St, Fifth | |
| 9349 Floor, Boston, MA 02110, USA. | |
| 9350 | |
| 9351 Verbatim copying and distribution of this entire article is permitted | |
| 9352 in any medium, provided this notice is preserved. | |
| 9353 Last modified 2008-05-07 [10]Valid XHTML 1.0 | |
| 9354 | |
| 9355 References | |
| 9356 | |
| 9357 1. http://gcc.gnu.org/ml/gcc-bugs/2001-06/msg00421.html | |
| 9358 2. mailto:gnu@gnu.org | |
| 9359 3. http://www.gnu.org/home.html#ContactInfo | |
| 9360 4. http://gcc.gnu.org/about.html | |
| 9361 5. http://gcc.gnu.org/onlinedocs/ | |
| 9362 6. mailto:gcc-help@gcc.gnu.org | |
| 9363 7. mailto:gcc@gnu.org | |
| 9364 8. mailto:gcc@gcc.gnu.org | |
| 9365 9. http://gcc.gnu.org/lists.html | |
| 9366 10. http://validator.w3.org/check/referer | |
| 9367 ====================================================================== | |
| 9368 http://gcc.gnu.org/gcc-2.95/index.html | |
| 9369 GCC 2.95 | |
| 9370 | |
| 9371 March 16, 2001: The GNU project and the GCC developers are pleased to | |
| 9372 announce the release of GCC version 2.95.3. | |
| 9373 | |
| 9374 Release History | |
| 9375 | |
| 9376 GCC 2.95.3 | |
| 9377 March 16, 2001 | |
| 9378 | |
| 9379 GCC 2.95.2 | |
| 9380 October 27, 1999 | |
| 9381 | |
| 9382 GCC 2.95.1 | |
| 9383 August 19, 1999 | |
| 9384 | |
| 9385 GCC 2.95 | |
| 9386 July 31, 1999. This is the first release of GCC since the April | |
| 9387 1999 GCC/EGCS reunification and includes nearly a year's worth | |
| 9388 of new development and bugfixes. | |
| 9389 | |
| 9390 References and Acknowledgements | |
| 9391 | |
| 9392 GCC used to stand for the GNU C Compiler, but since the compiler | |
| 9393 supports several other languages aside from C, it now stands for the | |
| 9394 GNU Compiler Collection. | |
| 9395 | |
| 9396 The whole suite has been extensively [1]regression tested and | |
| 9397 [2]package tested. It should be reliable and suitable for widespread | |
| 9398 use. | |
| 9399 | |
| 9400 The compiler has several new optimizations, new targets, new languages | |
| 9401 and other new features. See the [3]new features page for a more | |
| 9402 complete list of new features found in the GCC 2.95 releases. | |
| 9403 | |
| 9404 The sources include installation instructions in both HTML and | |
| 9405 plaintext forms in the install directory in the distribution. However, | |
| 9406 the most up to date [4]installation instructions and [5]build/test | |
| 9407 status are on the web pages. We will update those pages as new | |
| 9408 information becomes available. | |
| 9409 | |
| 9410 The GCC developers would like to thank the numerous people that have | |
| 9411 contributed new features, test results, bugfixes, etc to GCC. This | |
| 9412 [6]amazing group of volunteers is what makes GCC successful. | |
| 9413 | |
| 9414 And finally, we can't in good conscience fail to mention some | |
| 9415 [7]caveats to using GCC 2.95. | |
| 9416 | |
| 9417 Download GCC 2.95 from one of our many [8]mirror sites. | |
| 9418 | |
| 9419 For additional information about GCC please see the [9]GCC project web | |
| 9420 server or contact the [10]GCC development mailing list. | |
| 9421 | |
| 9422 Please send FSF & GNU inquiries & questions to [11]gnu@gnu.org. There | |
| 9423 are also [12]other ways to contact the FSF. | |
| 9424 | |
| 9425 These pages are maintained by [13]the GCC team. | |
| 9426 | |
| 9427 | |
| 9428 For questions related to the use of GCC, please consult these web | |
| 9429 pages and the [14]GCC manuals. If that fails, the | |
| 9430 [15]gcc-help@gcc.gnu.org mailing list might help. | |
| 9431 Please send comments on these web pages and the development of GCC to | |
| 9432 our developer mailing list at [16]gcc@gnu.org or [17]gcc@gcc.gnu.org. | |
| 9433 All of our lists have [18]public archives. | |
| 9434 | |
| 9435 Copyright (C) Free Software Foundation, Inc., 51 Franklin St, Fifth | |
| 9436 Floor, Boston, MA 02110, USA. | |
| 9437 | |
| 9438 Verbatim copying and distribution of this entire article is permitted | |
| 9439 in any medium, provided this notice is preserved. | |
| 9440 Last modified 2008-07-26 [19]Valid XHTML 1.0 | |
| 9441 | |
| 9442 References | |
| 9443 | |
| 9444 1. http://gcc.gnu.org/gcc-2.95/regress.html | |
| 9445 2. http://gcc.gnu.org/gcc-2.95/othertest.html | |
| 9446 3. http://gcc.gnu.org/gcc-2.95/features.html | |
| 9447 4. http://gcc.gnu.org/install/ | |
| 9448 5. http://gcc.gnu.org/gcc-2.95/buildstat.html | |
| 9449 6. http://gcc.gnu.org/onlinedocs/gcc/Contributors.html | |
| 9450 7. http://gcc.gnu.org/gcc-2.95/caveats.html | |
| 9451 8. http://gcc.gnu.org/mirrors.html | |
| 9452 9. http://gcc.gnu.org/index.html | |
| 9453 10. mailto:gcc@gcc.gnu.org | |
| 9454 11. mailto:gnu@gnu.org | |
| 9455 12. http://www.gnu.org/home.html#ContactInfo | |
| 9456 13. http://gcc.gnu.org/about.html | |
| 9457 14. http://gcc.gnu.org/onlinedocs/ | |
| 9458 15. mailto:gcc-help@gcc.gnu.org | |
| 9459 16. mailto:gcc@gnu.org | |
| 9460 17. mailto:gcc@gcc.gnu.org | |
| 9461 18. http://gcc.gnu.org/lists.html | |
| 9462 19. http://validator.w3.org/check/referer | |
| 9463 ====================================================================== | |
| 9464 http://gcc.gnu.org/gcc-2.95/features.html | |
| 9465 GCC 2.95 New Features | |
| 9466 | |
| 9467 * General Optimizer Improvements: | |
| 9468 + [1]Localized register spilling to improve speed and code | |
| 9469 density especially on small register class machines. | |
| 9470 + [2]Global CSE using lazy code motion algorithms. | |
| 9471 + [3]Improved global constant/copy propagation. | |
| 9472 + [4]Improved control flow graph analysis and manipulation. | |
| 9473 + [5]Local dead store elimination. | |
| 9474 + [6]Memory Load hoisting/store sinking in loops. | |
| 9475 + [7]Type based alias analysis is enabled by default. Note this | |
| 9476 feature will expose bugs in the Linux kernel. Please refer to | |
| 9477 the FAQ (as shipped with GCC 2.95) for additional information | |
| 9478 on this issue. | |
| 9479 + Major revamp of GIV detection, combination and simplification | |
| 9480 to improve loop performance. | |
| 9481 + Major improvements to register allocation and reloading. | |
| 9482 * New Languages and Language specific improvements | |
| 9483 + [8]Many C++ improvements. | |
| 9484 + [9]Many Fortran improvements. | |
| 9485 + [10]Java front-end has been integrated. [11]runtime library is | |
| 9486 available separately. | |
| 9487 + [12]ISO C99 support | |
| 9488 + [13]Chill front-end and runtime has been integrated. | |
| 9489 + Boehm garbage collector support in libobjc. | |
| 9490 + More support for various pragmas which appear in vendor | |
| 9491 include files | |
| 9492 * New Targets and Target Specific Improvements | |
| 9493 + [14]SPARC backend rewrite. | |
| 9494 + -mschedule=8000 will optimize code for PA8000 class | |
| 9495 processors; -mpa-risc-2-0 will generate code for PA2.0 | |
| 9496 processors | |
| 9497 + Various micro-optimizations for the ia32 port. K6 | |
| 9498 optimizations | |
| 9499 + Compiler will attempt to align doubles in the stack on the | |
| 9500 ia32 port | |
| 9501 + Alpha EV6 support | |
| 9502 + PowerPC 750 | |
| 9503 + RS6000/PowerPC: -mcpu=401 was added as an alias for -mcpu=403. | |
| 9504 -mcpu=e603e was added to do -mcpu=603e and -msoft-float. | |
| 9505 + c3x, c4x | |
| 9506 + HyperSPARC | |
| 9507 + SparcLite86x | |
| 9508 + sh4 | |
| 9509 + Support for new systems (OpenBSD, FreeBSD, UWIN, Interix, | |
| 9510 arm-linux) | |
| 9511 + vxWorks targets include support for vxWorks threads | |
| 9512 + StrongARM 110 and ARM9 support added. ARM Scheduling | |
| 9513 parameters rewritten. | |
| 9514 + Various changes to the MIPS port to avoid assembler macros, | |
| 9515 which in turn improves performance | |
| 9516 + Various performance improvements to the i960 port. | |
| 9517 + Major rewrite of ns32k port | |
| 9518 * Other significant improvements | |
| 9519 + [15]Ability to dump cfg information and display it using vcg. | |
| 9520 + The new faster scheme for fixing vendor header files is | |
| 9521 enabled by default. | |
| 9522 + Experimental internationalization support. | |
| 9523 + multibyte character support | |
| 9524 + Some compile-time speedups for pathological problems | |
| 9525 + Better support for complex types | |
| 9526 * Plus the usual mountain of bugfixes | |
| 9527 * Core compiler is based on the gcc2 development tree from Sept 30, | |
| 9528 1998, so we have all of the [16]features found in GCC 2.8. | |
| 9529 | |
| 9530 Additional Changes in GCC 2.95.1 | |
| 9531 | |
| 9532 * Generic bugfixes and improvements | |
| 9533 + Various documentation fixes related to the GCC/EGCS merger. | |
| 9534 + Fix memory management bug which could lead to spurious aborts, | |
| 9535 core dumps or random parsing errors in the compiler. | |
| 9536 + Fix a couple bugs in the dwarf1 and dwarf2 debug record | |
| 9537 support. | |
| 9538 + Fix infinite loop in the CSE optimizer. | |
| 9539 + Avoid undefined behavior in compiler FP emulation code | |
| 9540 + Fix install problem when prefix is overridden on the make | |
| 9541 install command. | |
| 9542 + Fix problem with unwanted installation of assert.h on some | |
| 9543 systems. | |
| 9544 + Fix problem with finding the wrong assembler in a single tree | |
| 9545 build. | |
| 9546 + Avoid increasing the known alignment of a register that is | |
| 9547 already known to be a pointer. | |
| 9548 * Platform specific bugfixes and improvements | |
| 9549 + Codegen bugfix for prologue/epilogue for cpu32 target. | |
| 9550 + Fix long long code generation bug for the Coldfire target. | |
| 9551 + Fix various aborts in the SH compiler. | |
| 9552 + Fix bugs in libgcc support library for the SH. | |
| 9553 + Fix alpha ev6 code generation bug. | |
| 9554 + Fix problems with EXIT_SUCCESS/EXIT_FAILURE redefinitions on | |
| 9555 AIX platforms. | |
| 9556 + Fix -fpic code generation bug for rs6000/ppc svr4 targets. | |
| 9557 + Fix varargs/stdarg code generation bug for rs6000/ppc svr4 | |
| 9558 targets. | |
| 9559 + Fix weak symbol handling for rs6000/ppc svr4 targets. | |
| 9560 + Fix various problems with 64bit code generation for the | |
| 9561 rs6000/ppc port. | |
| 9562 + Fix codegen bug which caused tetex to be mis-compiled on the | |
| 9563 x86. | |
| 9564 + Fix compiler abort in new cfg code exposed by x86 port. | |
| 9565 + Fix out of range array reference in code convert flat | |
| 9566 registers to the x87 stacked FP register file. | |
| 9567 + Fix minor vxworks configuration bug. | |
| 9568 + Fix return type of bsearch for SunOS 4.x. | |
| 9569 * Language & Runtime specific fixes. | |
| 9570 + The G++ signature extension has been deprecated. It will be | |
| 9571 removed in the next major release of G++. Use of signatures | |
| 9572 will result in a warning from the compiler. | |
| 9573 + Several bugs relating to templates and namespaces were fixed. | |
| 9574 + A bug that caused crashes when combining templates with -g on | |
| 9575 DWARF1 platforms was fixed. | |
| 9576 + Pointers-to-members, virtual functions, and multiple | |
| 9577 inheritance should now work together correctly. | |
| 9578 + Some code-generation bugs relating to function try blocks were | |
| 9579 fixed. | |
| 9580 + G++ is a little bit more lenient with certain archaic | |
| 9581 constructs than in GCC 2.95. | |
| 9582 + Fix to prevent shared library version #s from bring truncated | |
| 9583 to 1 digit | |
| 9584 + Fix missing std:: in the libstdc++ library. | |
| 9585 + Fix stream locking problems in libio. | |
| 9586 + Fix problem in java compiler driver. | |
| 9587 | |
| 9588 Additional Changes in GCC 2.95.2 | |
| 9589 | |
| 9590 The -fstrict-aliasing is not enabled by default for GCC 2.95.2. While | |
| 9591 the optimizations performed by -fstrict-aliasing are valid according to | |
| 9592 the C and C++ standards, the optimization have caused some problems, | |
| 9593 particularly with old non-conforming code. | |
| 9594 | |
| 9595 The GCC developers are experimenting with ways to warn users about code | |
| 9596 which violates the C/C++ standards, but those warnings are not ready | |
| 9597 for widespread use at this time. Rather than wait for those warnings | |
| 9598 the GCC developers have chosen to disable -fstrict-aliasing by default | |
| 9599 for the GCC 2.95.2 release. | |
| 9600 | |
| 9601 We strongly encourage developers to find and fix code which violates | |
| 9602 the C/C++ standards as -fstrict-aliasing may be enabled by default in | |
| 9603 future releases. Use the option -fstrict-aliasing to re-enable these | |
| 9604 optimizations. | |
| 9605 * Generic bugfixes and improvements | |
| 9606 + Fix incorrectly optimized memory reference in global common | |
| 9607 subexpression elimination (GCSE) optimization pass. | |
| 9608 + Fix code generation bug in regmove.c in which it could | |
| 9609 incorrectly change a "const" value. | |
| 9610 + Fix bug in optimization of conditionals involving volatile | |
| 9611 memory references. | |
| 9612 + Avoid over-allocation of stack space for some procedures. | |
| 9613 + Fixed bug in the compiler which caused incorrect optimization | |
| 9614 of an obscure series of bit manipulations, shifts and | |
| 9615 arithmetic. | |
| 9616 + Fixed register allocator bug which caused teTeX to be | |
| 9617 mis-compiled on SPARC targets. | |
| 9618 + Avoid incorrect optimization of degenerate case statements for | |
| 9619 certain targets such as the ARM. | |
| 9620 + Fix out of range memory reference in the jump optimizer. | |
| 9621 + Avoid dereferencing null pointer in fix-header. | |
| 9622 + Fix test for GCC specific features so that it is possible to | |
| 9623 bootstrap with gcc-2.6.2 and older versions of GCC. | |
| 9624 + Fix typo in scheduler which could potentially cause out of | |
| 9625 range memory accesses. | |
| 9626 + Avoid incorrect loop reversal which caused incorrect code for | |
| 9627 certain loops on PowerPC targets. | |
| 9628 + Avoid incorrect optimization of switch statements on certain | |
| 9629 targets (for example the ARM). | |
| 9630 * Platform specific bugfixes and improvements | |
| 9631 + Work around bug in Sun V5.0 compilers which caused bootstrap | |
| 9632 comparison failures on SPARC targets. | |
| 9633 + Fix SPARC backend bug which caused aborts in final.c. | |
| 9634 + Fix sparc-hal-solaris2* configuration fragments. | |
| 9635 + Fix bug in sparc block profiling. | |
| 9636 + Fix obscure code generation bug for the PARISC targets. | |
| 9637 + Define __STDC_EXT__ for HPUX configurations. | |
| 9638 + Various POWERPC64 code generation bugfixes. | |
| 9639 + Fix abort for PPC targets using ELF (ex GNU/Linux). | |
| 9640 + Fix collect2 problems for AIX targets. | |
| 9641 + Correct handling of .file directive for PPC targets. | |
| 9642 + Fix bug in fix_trunc x86 patterns. | |
| 9643 + Fix x86 port to correctly pop the FP stack for functions that | |
| 9644 return structures in memory. | |
| 9645 + Fix minor bug in strlen x86 pattern. | |
| 9646 + Use stabs debugging instead of dwarf1 for x86-solaris targets. | |
| 9647 + Fix template repository code to handle leading underscore in | |
| 9648 mangled names. | |
| 9649 + Fix weak/weak alias support for OpenBSD. | |
| 9650 + GNU/Linux for the ARM has C++ compatible include files. | |
| 9651 * Language & Runtime specific fixes. | |
| 9652 + Fix handling of constructor attribute in the C front-end which | |
| 9653 caused problems building the Chill runtime library on some | |
| 9654 targets. | |
| 9655 + Fix minor problem merging type qualifiers in the C front-end. | |
| 9656 + Fix aliasing bug for pointers and references (C/C++). | |
| 9657 + Fix incorrect "non-constant initializer bug" when -traditional | |
| 9658 or -fwritable-strings is enabled. | |
| 9659 + Fix build error for Chill front-end on SunOS. | |
| 9660 + Do not complain about duplicate instantiations when using | |
| 9661 -frepo (C++). | |
| 9662 + Fix array bounds handling in C++ front-end which caused | |
| 9663 problems with dwarf debugging information in some | |
| 9664 circumstances. | |
| 9665 + Fix minor namespace problem. | |
| 9666 + Fix problem linking java programs. | |
| 9667 | |
| 9668 Additional Changes in GCC 2.95.3 | |
| 9669 | |
| 9670 * Generic bugfixes and improvements | |
| 9671 + Fix numerous problems that caused incorrect optimization in | |
| 9672 the register reloading code. | |
| 9673 + Fix numerous problems that caused incorrect optimization in | |
| 9674 the loop optimizer. | |
| 9675 + Fix aborts in the functions build_insn_chain and scan_loops | |
| 9676 under some circumstances. | |
| 9677 + Fix an alias analysis bug. | |
| 9678 + Fix an infinite compilation bug in the combiner. | |
| 9679 + A few problems with complex number support have been fixed. | |
| 9680 + It is no longer possible for gcc to act as a fork bomb when | |
| 9681 installed incorrectly. | |
| 9682 + The -fpack-struct option should be recognized now. | |
| 9683 + Fixed a bug that caused incorrect code to be generated due to | |
| 9684 a lost stack adjustment. | |
| 9685 * Platform specific bugfixes and improvements | |
| 9686 + Support building ARM toolchains hosted on Windows. | |
| 9687 + Fix attribute calculations in ARM toolchains. | |
| 9688 + arm-linux support has been improved. | |
| 9689 + Fix a PIC failure on sparc targets. | |
| 9690 + On ix86 targets, the regparm attribute should now work | |
| 9691 reliably. | |
| 9692 + Several updates for the h8300 port. | |
| 9693 + Fix problem building libio with glibc 2.2. | |
| 9694 | |
| 9695 Please send FSF & GNU inquiries & questions to [17]gnu@gnu.org. There | |
| 9696 are also [18]other ways to contact the FSF. | |
| 9697 | |
| 9698 These pages are maintained by [19]the GCC team. | |
| 9699 | |
| 9700 | |
| 9701 For questions related to the use of GCC, please consult these web | |
| 9702 pages and the [20]GCC manuals. If that fails, the | |
| 9703 [21]gcc-help@gcc.gnu.org mailing list might help. | |
| 9704 Please send comments on these web pages and the development of GCC to | |
| 9705 our developer mailing list at [22]gcc@gnu.org or [23]gcc@gcc.gnu.org. | |
| 9706 All of our lists have [24]public archives. | |
| 9707 | |
| 9708 Copyright (C) Free Software Foundation, Inc., 51 Franklin St, Fifth | |
| 9709 Floor, Boston, MA 02110, USA. | |
| 9710 | |
| 9711 Verbatim copying and distribution of this entire article is permitted | |
| 9712 in any medium, provided this notice is preserved. | |
| 9713 Last modified 2008-06-15 [25]Valid XHTML 1.0 | |
| 9714 | |
| 9715 References | |
| 9716 | |
| 9717 1. http://gcc.gnu.org/news/spill.html | |
| 9718 2. http://gcc.gnu.org/news/lcm.html | |
| 9719 3. http://gcc.gnu.org/news/cprop.html | |
| 9720 4. http://gcc.gnu.org/news/cfg.html | |
| 9721 5. http://gcc.gnu.org/news/dse.html | |
| 9722 6. http://gcc.gnu.org/news/hoist.html | |
| 9723 7. http://gcc.gnu.org/news/alias.html | |
| 9724 8. http://gcc.gnu.org/gcc-2.95/c++features.html | |
| 9725 9. http://gcc.gnu.org/onlinedocs/gcc-3.4.6/g77/News.html | |
| 9726 10. http://gcc.gnu.org/java/gcj-announce.txt | |
| 9727 11. http://gcc.gnu.org/news/javaannounce.html | |
| 9728 12. http://gcc.gnu.org/c99status.html | |
| 9729 13. http://gcc.gnu.org/news/chill.html | |
| 9730 14. http://gcc.gnu.org/news/sparc.html | |
| 9731 15. http://gcc.gnu.org/news/egcs-vcg.html | |
| 9732 16. http://gcc.gnu.org/egcs-1.0/features-2.8.html | |
| 9733 17. mailto:gnu@gnu.org | |
| 9734 18. http://www.gnu.org/home.html#ContactInfo | |
| 9735 19. http://gcc.gnu.org/about.html | |
| 9736 20. http://gcc.gnu.org/onlinedocs/ | |
| 9737 21. mailto:gcc-help@gcc.gnu.org | |
| 9738 22. mailto:gcc@gnu.org | |
| 9739 23. mailto:gcc@gcc.gnu.org | |
| 9740 24. http://gcc.gnu.org/lists.html | |
| 9741 25. http://validator.w3.org/check/referer | |
| 9742 ====================================================================== | |
| 9743 http://gcc.gnu.org/gcc-2.95/caveats.html | |
| 9744 GCC 2.95 Caveats | |
| 9745 | |
| 9746 * GCC 2.95 will issue an error for invalid asm statements that had | |
| 9747 been silently accepted by earlier versions of the compiler. This is | |
| 9748 particularly noticeable when compiling older versions of the Linux | |
| 9749 kernel (2.0.xx). Please refer to the FAQ (as shipped with GCC 2.95) | |
| 9750 for more information on this issue. | |
| 9751 * GCC 2.95 implements type based alias analysis to disambiguate | |
| 9752 memory references. Some programs, particularly the Linux kernel | |
| 9753 violate ANSI/ISO aliasing rules and therefore may not operate | |
| 9754 correctly when compiled with GCC 2.95. Please refer to the FAQ (as | |
| 9755 shipped with GCC 2.95) for more information on this issue. | |
| 9756 * GCC 2.95 has a known bug in its handling of complex variables for | |
| 9757 64bit targets. Instead of silently generating incorrect code, GCC | |
| 9758 2.95 will issue a fatal error for situations it can not handle. | |
| 9759 This primarily affects the Fortran community as Fortran makes more | |
| 9760 use of complex variables than C or C++. | |
| 9761 * GCC 2.95 has an integrated libstdc++, but does not have an | |
| 9762 integrated libg++. Furthermore old libg++ releases will not work | |
| 9763 with GCC 2.95. You can retrieve a recent copy of libg++ from the | |
| 9764 [1]GCC ftp server. | |
| 9765 Note most C++ programs only need libstdc++. | |
| 9766 * Exception handling may not work with shared libraries, particularly | |
| 9767 on alphas, hppas, rs6000/powerpc and mips based platforms. | |
| 9768 Exception handling is known to work on x86 GNU/Linux platforms with | |
| 9769 shared libraries. | |
| 9770 * In general, GCC 2.95 is more rigorous about rejecting invalid C++ | |
| 9771 code or deprecated C++ constructs than G++ 2.7, G++ 2.8, EGCS 1.0, | |
| 9772 or EGCS 1.1. As a result it may be necessary to fix C++ code before | |
| 9773 it will compile with GCC 2.95. | |
| 9774 * G++ is also converting toward the ISO C++ standard; as a result | |
| 9775 code which was previously valid (and thus accepted by other | |
| 9776 compilers and older versions of g++) may no longer be accepted. The | |
| 9777 flag -fpermissive may allow some non-conforming code to compile | |
| 9778 with GCC 2.95. | |
| 9779 * GCC 2.95 compiled C++ code is not binary compatible with EGCS | |
| 9780 1.1.x, EGCS 1.0.x or GCC 2.8.x. | |
| 9781 * GCC 2.95 does not have changes from the GCC 2.8 tree that were made | |
| 9782 between Sept 30, 1998 and April 30, 1999 (the official end of the | |
| 9783 GCC 2.8 project). Future GCC releases will include all the changes | |
| 9784 from the defunct GCC 2.8 sources. | |
| 9785 | |
| 9786 Please send FSF & GNU inquiries & questions to [2]gnu@gnu.org. There | |
| 9787 are also [3]other ways to contact the FSF. | |
| 9788 | |
| 9789 These pages are maintained by [4]the GCC team. | |
| 9790 | |
| 9791 | |
| 9792 For questions related to the use of GCC, please consult these web | |
| 9793 pages and the [5]GCC manuals. If that fails, the | |
| 9794 [6]gcc-help@gcc.gnu.org mailing list might help. | |
| 9795 Please send comments on these web pages and the development of GCC to | |
| 9796 our developer mailing list at [7]gcc@gnu.org or [8]gcc@gcc.gnu.org. | |
| 9797 All of our lists have [9]public archives. | |
| 9798 | |
| 9799 Copyright (C) Free Software Foundation, Inc., 51 Franklin St, Fifth | |
| 9800 Floor, Boston, MA 02110, USA. | |
| 9801 | |
| 9802 Verbatim copying and distribution of this entire article is permitted | |
| 9803 in any medium, provided this notice is preserved. | |
| 9804 Last modified 2007-07-25 [10]Valid XHTML 1.0 | |
| 9805 | |
| 9806 References | |
| 9807 | |
| 9808 1. ftp://gcc.gnu.org/pub/gcc/infrastructure/libg++-2.8.1.3.tar.gz | |
| 9809 2. mailto:gnu@gnu.org | |
| 9810 3. http://www.gnu.org/home.html#ContactInfo | |
| 9811 4. http://gcc.gnu.org/about.html | |
| 9812 5. http://gcc.gnu.org/onlinedocs/ | |
| 9813 6. mailto:gcc-help@gcc.gnu.org | |
| 9814 7. mailto:gcc@gnu.org | |
| 9815 8. mailto:gcc@gcc.gnu.org | |
| 9816 9. http://gcc.gnu.org/lists.html | |
| 9817 10. http://validator.w3.org/check/referer | |
| 9818 ====================================================================== | |
| 9819 http://gcc.gnu.org/egcs-1.1/index.html | |
| 9820 EGCS 1.1 | |
| 9821 | |
| 9822 September 3, 1998: We are pleased to announce the release of EGCS 1.1. | |
| 9823 December 1, 1998: We are pleased to announce the release of EGCS 1.1.1. | |
| 9824 March 15, 1999: We are pleased to announce the release of EGCS 1.1.2. | |
| 9825 | |
| 9826 EGCS is a free software project to further the development of the GNU | |
| 9827 compilers using an open development environment. | |
| 9828 | |
| 9829 EGCS 1.1 is a major new release of the EGCS compiler system. It has | |
| 9830 been [1]extensively tested and is believed to be stable and suitable | |
| 9831 for widespread use. | |
| 9832 | |
| 9833 EGCS 1.1 is based on an June 6, 1998 snapshot of the GCC 2.8 | |
| 9834 development sources; it contains all of the new features found in GCC | |
| 9835 2.8.1 as well as all new development from GCC up to June 6, 1998. | |
| 9836 | |
| 9837 EGCS 1.1 also contains many improvements and features not found in GCC | |
| 9838 or in older versions of EGCS: | |
| 9839 * Global common subexpression elimination and global constant/copy | |
| 9840 propagation (aka [2]gcse) | |
| 9841 * Ongoing improvements to the [3]alias analysis support to allow for | |
| 9842 better optimizations throughout the compiler. | |
| 9843 * Vastly improved [4]C++ compiler and integrated C++ runtime | |
| 9844 libraries. | |
| 9845 * Fixes for the /tmp symlink race security problems. | |
| 9846 * New targets including mips16, arm-thumb and 64 bit PowerPC. | |
| 9847 * Improvements to GNU Fortran (g77) compiler and runtime library made | |
| 9848 since g77 version 0.5.23. | |
| 9849 | |
| 9850 See the [5]new features page for a more complete list of new features | |
| 9851 found in EGCS 1.1 releases. | |
| 9852 | |
| 9853 EGCS 1.1.1 is a minor update to fix several serious problems in EGCS | |
| 9854 1.1: | |
| 9855 * General improvements and fixes | |
| 9856 + Avoid some stack overflows when compiling large functions. | |
| 9857 + Avoid incorrect loop invariant code motions. | |
| 9858 + Fix some core dumps on Linux kernel code. | |
| 9859 + Bring back the imake -Di386 and friends fix from EGCS 1.0.2. | |
| 9860 + Fix code generation problem in gcse. | |
| 9861 + Various documentation related fixes. | |
| 9862 * g++/libstdc++ improvements and fixes | |
| 9863 + MT safe EH fix for setjmp/longjmp based exception handling. | |
| 9864 + Fix a few bad interactions between optimization and exception | |
| 9865 handling. | |
| 9866 + Fixes for demangling of template names starting with "__". | |
| 9867 + Fix a bug that would fail to run destructors in some cases | |
| 9868 with -O2. | |
| 9869 + Fix 'new' of classes with virtual bases. | |
| 9870 + Fix crash building Qt on the Alpha. | |
| 9871 + Fix failure compiling WIFEXITED macro on GNU/Linux. | |
| 9872 + Fix some -frepo failures. | |
| 9873 * g77 and libf2c improvements and fixes | |
| 9874 + Various documentation fixes. | |
| 9875 + Avoid compiler crash on RAND intrinsic. | |
| 9876 + Fix minor bugs in makefiles exposed by BSD make programs. | |
| 9877 + Define _XOPEN_SOURCE for libI77 build to avoid potential | |
| 9878 problems on some 64-bit systems. | |
| 9879 + Fix problem with implicit endfile on rewind. | |
| 9880 + Fix spurious recursive I/O errors. | |
| 9881 * platform specific improvements and fixes | |
| 9882 + Match all versions of UnixWare7. | |
| 9883 + Do not assume x86 SVR4 or UnixWare targets can handle stabs. | |
| 9884 + Fix PPC/RS6000 LEGITIMIZE_ADDRESS macro and bug in conversion | |
| 9885 from unsigned ints to double precision floats. | |
| 9886 + Fix ARM ABI issue with NetBSD. | |
| 9887 + Fix a few arm code generation bugs. | |
| 9888 + Fixincludes will fix additional broken SCO OpenServer header | |
| 9889 files. | |
| 9890 + Fix a m68k backend bug which caused invalid offsets in reg+d | |
| 9891 addresses. | |
| 9892 + Fix problems with 64bit AIX 4.3 support. | |
| 9893 + Fix handling of long longs for varargs/stdarg functions on the | |
| 9894 ppc. | |
| 9895 + Minor fixes to CPP predefines for Windows. | |
| 9896 + Fix code generation problems with gpr<->fpr copies for 64bit | |
| 9897 ppc. | |
| 9898 + Fix a few coldfire code generation bugs. | |
| 9899 + Fix some more header file problems on SunOS 4.x. | |
| 9900 + Fix assert.h handling for RTEMS. | |
| 9901 + Fix Windows handling of TREE_SYMBOL_REFERENCED. | |
| 9902 + Fix x86 compiler abort in reg-stack pass. | |
| 9903 + Fix cygwin/windows problem with section attributes. | |
| 9904 + Fix Alpha code generation problem exposed by SMP Linux | |
| 9905 kernels. | |
| 9906 + Fix typo in m68k 32->64bit integer conversion. | |
| 9907 + Make sure target libraries build with -fPIC for PPC & Alpha | |
| 9908 targets. | |
| 9909 | |
| 9910 EGCS 1.1.2 is a minor update to fix several serious problems in EGCS | |
| 9911 1.1.1: | |
| 9912 * General improvements and fixes | |
| 9913 + Fix bug in loop optimizer which caused the SPARC (and | |
| 9914 potentially other) ports to segfault. | |
| 9915 + Fix infinite recursion in alias analysis and combiner code. | |
| 9916 + Fix bug in regclass preferencing. | |
| 9917 + Fix incorrect loop reversal which caused incorrect code to be | |
| 9918 generated for several targets. | |
| 9919 + Fix return value for builtin memcpy. | |
| 9920 + Reduce compile time for certain loops which exposed quadratic | |
| 9921 behavior in the loop optimizer. | |
| 9922 + Fix bug which caused volatile memory to be written multiple | |
| 9923 times when only one write was needed/desired. | |
| 9924 + Fix compiler abort in caller-save.c | |
| 9925 + Fix combiner bug which caused incorrect code generation for | |
| 9926 certain division by constant operations. | |
| 9927 + Fix incorrect code generation due to a bug in range check | |
| 9928 optimizations. | |
| 9929 + Fix incorrect code generation due to mis-handling of clobbered | |
| 9930 values in CSE. | |
| 9931 + Fix compiler abort/segfault due to incorrect register | |
| 9932 splitting when unrolling loops. | |
| 9933 + Fix code generation involving autoincremented addresses with | |
| 9934 ternary operators. | |
| 9935 + Work around bug in the scheduler which caused qt to be | |
| 9936 mis-compiled on some platforms. | |
| 9937 + Fix code generation problems with -fshort-enums. | |
| 9938 + Tighten security for temporary files. | |
| 9939 + Improve compile time for codes which make heavy use of | |
| 9940 overloaded functions. | |
| 9941 + Fix multiply defined constructor/destructor symbol problems. | |
| 9942 + Avoid setting bogus RPATH environment variable during | |
| 9943 bootstrap. | |
| 9944 + Avoid GNU-make dependencies in the texinfo subdir. | |
| 9945 + Install CPP wrapper script in $(prefix)/bin if --enable-cpp. | |
| 9946 --enable-cpp=<dirname> can be used to specify an additional | |
| 9947 install directory for the cpp wrapper script. | |
| 9948 + Fix CSE bug which caused incorrect label-label refs to appear | |
| 9949 on some platforms. | |
| 9950 + Avoid linking in EH routines from libgcc if they are not | |
| 9951 needed. | |
| 9952 + Avoid obscure bug in aliasing code. | |
| 9953 + Fix bug in weak symbol handling. | |
| 9954 * Platform-specific improvements and fixes | |
| 9955 + Fix detection of PPro/PII on Unixware 7. | |
| 9956 + Fix compiler segfault when building spec99 and other programs | |
| 9957 for SPARC targets. | |
| 9958 + Fix code-generation bugs for integer and floating point | |
| 9959 conditional move instructions on the PPro/PII. | |
| 9960 + Use fixincludes to fix byteorder problems on i?86-*-sysv. | |
| 9961 + Fix build failure for the arc port. | |
| 9962 + Fix floating point format configuration for i?86-gnu port. | |
| 9963 + Fix problems with hppa1.0-hp-hpux10.20 configuration when | |
| 9964 threads are enabled. | |
| 9965 + Fix coldfire code generation bugs. | |
| 9966 + Fix "unrecognized insn" problems for Alpha and PPC ports. | |
| 9967 + Fix h8/300 code generation problem with floating point values | |
| 9968 in memory. | |
| 9969 + Fix unrecognized insn problems for the m68k port. | |
| 9970 + Fix namespace-pollution problem for the x86 port. | |
| 9971 + Fix problems with old assembler on x86 NeXT systems. | |
| 9972 + Fix PIC code-generation problems for the SPARC port. | |
| 9973 + Fix minor bug with LONG_CALLS in PowerPC SVR4 support. | |
| 9974 + Fix minor ISO namespace violation in Alpha varargs/stdarg | |
| 9975 support. | |
| 9976 + Fix incorrect "braf" instruction usage for the SH port. | |
| 9977 + Fix minor bug in va-sh which prevented its use with -ansi. | |
| 9978 + Fix problems recognizing and supporting FreeBSD. | |
| 9979 + Handle OpenBSD systems correctly. | |
| 9980 + Minor fixincludes fix for Digital UNIX 4.0B. | |
| 9981 + Fix problems with ctors/dtors in SCO shared libraries. | |
| 9982 + Abort instead of generating incorrect code for PPro/PII | |
| 9983 floating point conditional moves. | |
| 9984 + Avoid multiply defined symbols on Linux/GNU systems using | |
| 9985 libc-5.4.xx. | |
| 9986 + Fix abort in alpha compiler. | |
| 9987 * Fortran-specific fixes | |
| 9988 + Fix the IDate intrinsic (VXT) (in libg2c) so the returned year | |
| 9989 is in the documented, non-Y2K-compliant range of 0-99, instead | |
| 9990 of being returned as 100 in the year 2000. | |
| 9991 + Fix the `Date_and_Time' intrinsic (in libg2c) to return the | |
| 9992 milliseconds value properly in Values(8). | |
| 9993 + Fix the `LStat' intrinsic (in libg2c) to return device-ID | |
| 9994 information properly in SArray(7). | |
| 9995 | |
| 9996 Each release includes installation instructions in both HTML and | |
| 9997 plaintext forms (see the INSTALL directory in the toplevel directory of | |
| 9998 the distribution). However, we also keep the most up to date | |
| 9999 [6]installation instructions and [7]build/test status on our web page. | |
| 10000 We will update those pages as new information becomes available. | |
| 10001 | |
| 10002 The EGCS project would like to thank the numerous people that have | |
| 10003 contributed new features, test results, bugfixes, etc. This [8]amazing | |
| 10004 group of volunteers is what makes EGCS successful. | |
| 10005 | |
| 10006 And finally, we can't in good conscience fail to mention some | |
| 10007 [9]caveats to using EGCS 1.1. | |
| 10008 | |
| 10009 Download EGCS from egcs.cygnus.com (USA California). | |
| 10010 | |
| 10011 The EGCS 1.1 release is also available on many mirror sites. | |
| 10012 [10]Goto mirror list to find a closer site. | |
| 10013 | |
| 10014 Please send FSF & GNU inquiries & questions to [11]gnu@gnu.org. There | |
| 10015 are also [12]other ways to contact the FSF. | |
| 10016 | |
| 10017 These pages are maintained by [13]the GCC team. | |
| 10018 | |
| 10019 | |
| 10020 For questions related to the use of GCC, please consult these web | |
| 10021 pages and the [14]GCC manuals. If that fails, the | |
| 10022 [15]gcc-help@gcc.gnu.org mailing list might help. | |
| 10023 Please send comments on these web pages and the development of GCC to | |
| 10024 our developer mailing list at [16]gcc@gnu.org or [17]gcc@gcc.gnu.org. | |
| 10025 All of our lists have [18]public archives. | |
| 10026 | |
| 10027 Copyright (C) Free Software Foundation, Inc., 51 Franklin St, Fifth | |
| 10028 Floor, Boston, MA 02110, USA. | |
| 10029 | |
| 10030 Verbatim copying and distribution of this entire article is permitted | |
| 10031 in any medium, provided this notice is preserved. | |
| 10032 Last modified 2007-07-25 [19]Valid XHTML 1.0 | |
| 10033 | |
| 10034 References | |
| 10035 | |
| 10036 1. http://gcc.gnu.org/egcs-1.1/egcs-1.1-test.html | |
| 10037 2. http://gcc.gnu.org/news/gcse.html | |
| 10038 3. http://gcc.gnu.org/news/alias.html | |
| 10039 4. http://gcc.gnu.org/egcs-1.1/c++features.html | |
| 10040 5. http://gcc.gnu.org/egcs-1.1/features.html | |
| 10041 6. http://gcc.gnu.org/install/ | |
| 10042 7. http://gcc.gnu.org/egcs-1.1/buildstat.html | |
| 10043 8. http://gcc.gnu.org/onlinedocs/gcc/Contributors.html | |
| 10044 9. http://gcc.gnu.org/egcs-1.1/caveats.html | |
| 10045 10. http://gcc.gnu.org/mirrors.html | |
| 10046 11. mailto:gnu@gnu.org | |
| 10047 12. http://www.gnu.org/home.html#ContactInfo | |
| 10048 13. http://gcc.gnu.org/about.html | |
| 10049 14. http://gcc.gnu.org/onlinedocs/ | |
| 10050 15. mailto:gcc-help@gcc.gnu.org | |
| 10051 16. mailto:gcc@gnu.org | |
| 10052 17. mailto:gcc@gcc.gnu.org | |
| 10053 18. http://gcc.gnu.org/lists.html | |
| 10054 19. http://validator.w3.org/check/referer | |
| 10055 ====================================================================== | |
| 10056 http://gcc.gnu.org/egcs-1.1/features.html | |
| 10057 EGCS 1.1 new features | |
| 10058 | |
| 10059 * Integrated GNU Fortran (g77) compiler and runtime library with | |
| 10060 improvements, based on g77 version 0.5.23. | |
| 10061 * Vast improvements in the C++ compiler; so many they have [1]page of | |
| 10062 their own! | |
| 10063 * Compiler implements [2]global common subexpression elimination and | |
| 10064 global copy/constant propagation. | |
| 10065 * More major improvements in the [3]alias analysis code. | |
| 10066 * More major improvements in the exception handling code to improve | |
| 10067 performance, lower static overhead and provide the infrastructure | |
| 10068 for future improvements. | |
| 10069 * The infamous /tmp symlink race security problems have been fixed. | |
| 10070 * The regmove optimization pass has been nearly completely rewritten | |
| 10071 to improve performance of generated code. | |
| 10072 * The compiler now recomputes register usage information before local | |
| 10073 register allocation. By providing more accurate information to the | |
| 10074 priority based allocator, we get better register allocation. | |
| 10075 * The register reloading phase of the compiler optimizes spill code | |
| 10076 much better than in previous releases. | |
| 10077 * Some bad interactions between the register allocator and | |
| 10078 instruction scheduler have been fixed, resulting in much better | |
| 10079 code for certain programs. Additionally, we have tuned the | |
| 10080 scheduler in various ways to improve performance of generated code | |
| 10081 for some architectures. | |
| 10082 * The compiler's branch shortening algorithms have been significantly | |
| 10083 improved to work better on targets which align jump targets. | |
| 10084 * The compiler now supports -Os to prefer optimizing for code space | |
| 10085 over optimizing for code speed. | |
| 10086 * The compiler will now totally eliminate library calls which compute | |
| 10087 constant values. This primarily helps targets with no integer | |
| 10088 div/mul support and targets without floating point support. | |
| 10089 * The compiler now supports an extensive "--help" option. | |
| 10090 * cpplib has been greatly improved and may be suitable for limited | |
| 10091 use. | |
| 10092 * Memory footprint for the compiler has been significantly reduced | |
| 10093 for some pathological cases. | |
| 10094 * The time to build EGCS has been improved for certain targets | |
| 10095 (particularly the alpha and mips platforms). | |
| 10096 * Many infrastructure improvements throughout the compiler, plus the | |
| 10097 usual mountain of bugfixes and minor improvements. | |
| 10098 * Target dependent improvements: | |
| 10099 + SPARC port now includes V8 plus and V9 support as well as | |
| 10100 performance tuning for Ultra class machines. The SPARC port | |
| 10101 now uses the Haifa scheduler. | |
| 10102 + Alpha port has been tuned for the EV6 processor and has an | |
| 10103 optimized expansion of memcpy/bzero. The Alpha port now uses | |
| 10104 the Haifa scheduler. | |
| 10105 + RS6000/PowerPC: support for the Power64 architecture and AIX | |
| 10106 4.3. The RS6000/PowerPC port now uses the Haifa scheduler. | |
| 10107 + x86: Alignment of static store data and jump targets is per | |
| 10108 Intel recommendations now. Various improvements throughout the | |
| 10109 x86 port to improve performance on Pentium processors | |
| 10110 (including improved epilogue sequences for Pentium chips and | |
| 10111 backend improvements which should help register allocation on | |
| 10112 all x86 variants. Conditional move support has been fixed and | |
| 10113 enabled for PPro processors. The x86 port also better supports | |
| 10114 64bit operations now. Unixware 7, a System V Release 5 target, | |
| 10115 is now supported and SCO OpenServer targets can support GAS. | |
| 10116 + MIPS has improved multiply/multiply-add support and now | |
| 10117 includes mips16 ISA support. | |
| 10118 + M68k has many micro-optimizations and Coldfire fixes. | |
| 10119 * Core compiler is based on the GCC development tree from June 9, | |
| 10120 1998, so we have all of the [4]features found in GCC 2.8. | |
| 10121 | |
| 10122 Please send FSF & GNU inquiries & questions to [5]gnu@gnu.org. There | |
| 10123 are also [6]other ways to contact the FSF. | |
| 10124 | |
| 10125 These pages are maintained by [7]the GCC team. | |
| 10126 | |
| 10127 | |
| 10128 For questions related to the use of GCC, please consult these web | |
| 10129 pages and the [8]GCC manuals. If that fails, the | |
| 10130 [9]gcc-help@gcc.gnu.org mailing list might help. | |
| 10131 Please send comments on these web pages and the development of GCC to | |
| 10132 our developer mailing list at [10]gcc@gnu.org or [11]gcc@gcc.gnu.org. | |
| 10133 All of our lists have [12]public archives. | |
| 10134 | |
| 10135 Copyright (C) Free Software Foundation, Inc., 51 Franklin St, Fifth | |
| 10136 Floor, Boston, MA 02110, USA. | |
| 10137 | |
| 10138 Verbatim copying and distribution of this entire article is permitted | |
| 10139 in any medium, provided this notice is preserved. | |
| 10140 Last modified 2008-07-26 [13]Valid XHTML 1.0 | |
| 10141 | |
| 10142 References | |
| 10143 | |
| 10144 1. http://gcc.gnu.org/egcs-1.1/c++features.html | |
| 10145 2. http://gcc.gnu.org/news/gcse.html | |
| 10146 3. http://gcc.gnu.org/news/alias.html | |
| 10147 4. http://gcc.gnu.org/egcs-1.0/features-2.8.html | |
| 10148 5. mailto:gnu@gnu.org | |
| 10149 6. http://www.gnu.org/home.html#ContactInfo | |
| 10150 7. http://gcc.gnu.org/about.html | |
| 10151 8. http://gcc.gnu.org/onlinedocs/ | |
| 10152 9. mailto:gcc-help@gcc.gnu.org | |
| 10153 10. mailto:gcc@gnu.org | |
| 10154 11. mailto:gcc@gcc.gnu.org | |
| 10155 12. http://gcc.gnu.org/lists.html | |
| 10156 13. http://validator.w3.org/check/referer | |
| 10157 ====================================================================== | |
| 10158 http://gcc.gnu.org/egcs-1.1/caveats.html | |
| 10159 EGCS 1.1 Caveats | |
| 10160 | |
| 10161 * EGCS has an integrated libstdc++, but does not have an integrated | |
| 10162 libg++. Furthermore old libg++ releases will not work with EGCS; HJ | |
| 10163 Lu has made a libg++-2.8.1.2 snapshot available which may work with | |
| 10164 EGCS. | |
| 10165 Note most C++ programs only need libstdc++. | |
| 10166 * Exception handling may not work with shared libraries, particularly | |
| 10167 on alphas, hppas, rs6000/powerpc and mips based platforms. | |
| 10168 Exception handling is known to work on x86-linux platforms with | |
| 10169 shared libraries. | |
| 10170 * Some versions of the Linux kernel have bugs which prevent them from | |
| 10171 being compiled or from running when compiled by EGCS. See the FAQ | |
| 10172 (as shipped with EGCS 1.1) for additional information. | |
| 10173 * In general, EGCS is more rigorous about rejecting invalid C++ code | |
| 10174 or deprecated C++ constructs than g++-2.7, g++-2.8 or EGCS 1.0. As | |
| 10175 a result it may be necessary to fix C++ code before it will compile | |
| 10176 with EGCS. | |
| 10177 * G++ is also converting toward the ISO C++ standard; as a result | |
| 10178 code which was previously valid (and thus accepted by other | |
| 10179 compilers and older versions of g++) may no longer be accepted. | |
| 10180 * EGCS 1.1 compiled C++ code is not binary compatible with EGCS 1.0.x | |
| 10181 or GCC 2.8.x due to changes necessary to support thread safe | |
| 10182 exception handling. | |
| 10183 | |
| 10184 Please send FSF & GNU inquiries & questions to [1]gnu@gnu.org. There | |
| 10185 are also [2]other ways to contact the FSF. | |
| 10186 | |
| 10187 These pages are maintained by [3]the GCC team. | |
| 10188 | |
| 10189 | |
| 10190 For questions related to the use of GCC, please consult these web | |
| 10191 pages and the [4]GCC manuals. If that fails, the | |
| 10192 [5]gcc-help@gcc.gnu.org mailing list might help. | |
| 10193 Please send comments on these web pages and the development of GCC to | |
| 10194 our developer mailing list at [6]gcc@gnu.org or [7]gcc@gcc.gnu.org. | |
| 10195 All of our lists have [8]public archives. | |
| 10196 | |
| 10197 Copyright (C) Free Software Foundation, Inc., 51 Franklin St, Fifth | |
| 10198 Floor, Boston, MA 02110, USA. | |
| 10199 | |
| 10200 Verbatim copying and distribution of this entire article is permitted | |
| 10201 in any medium, provided this notice is preserved. | |
| 10202 Last modified 2007-07-25 [9]Valid XHTML 1.0 | |
| 10203 | |
| 10204 References | |
| 10205 | |
| 10206 1. mailto:gnu@gnu.org | |
| 10207 2. http://www.gnu.org/home.html#ContactInfo | |
| 10208 3. http://gcc.gnu.org/about.html | |
| 10209 4. http://gcc.gnu.org/onlinedocs/ | |
| 10210 5. mailto:gcc-help@gcc.gnu.org | |
| 10211 6. mailto:gcc@gnu.org | |
| 10212 7. mailto:gcc@gcc.gnu.org | |
| 10213 8. http://gcc.gnu.org/lists.html | |
| 10214 9. http://validator.w3.org/check/referer | |
| 10215 ====================================================================== | |
| 10216 http://gcc.gnu.org/egcs-1.0/index.html | |
| 10217 EGCS 1.0 | |
| 10218 | |
| 10219 December 3, 1997: We are pleased to announce the release of EGCS 1.0. | |
| 10220 January 6, 1998: We are pleased to announce the release of EGCS 1.0.1. | |
| 10221 March 16, 1998: We are pleased to announce the release of EGCS 1.0.2. | |
| 10222 May 15, 1998 We are pleased to announce the release of EGCS 1.0.3. | |
| 10223 | |
| 10224 EGCS is a collaborative effort involving several groups of hackers | |
| 10225 using an open development model to accelerate development and testing | |
| 10226 of GNU compilers and runtime libraries. | |
| 10227 | |
| 10228 An important goal of EGCS is to allow wide scale testing of | |
| 10229 experimental features and optimizations; therefore, EGCS contains some | |
| 10230 features and optimizations which are still under development. However, | |
| 10231 EGCS has been carefully tested and should be comparable in quality to | |
| 10232 most GCC releases. | |
| 10233 | |
| 10234 EGCS 1.0 is based on an August 2, 1997 snapshot of the GCC 2.8 | |
| 10235 development sources; it contains nearly all of the new features found | |
| 10236 in GCC 2.8. | |
| 10237 | |
| 10238 EGCS 1.0 also contains many improvements and features not found in GCC | |
| 10239 2.7 and even the GCC 2.8 series (which was released after the original | |
| 10240 EGCS 1.0 release). | |
| 10241 * Integrated C++ runtime libraries, including support for most major | |
| 10242 GNU/Linux systems! | |
| 10243 * The integrated libstdc++ library includes a verbatim copy of SGI's | |
| 10244 STL release. | |
| 10245 * Integrated GNU Fortran compiler. | |
| 10246 * New instruction scheduler. | |
| 10247 * New alias analysis code. | |
| 10248 | |
| 10249 See the [1]new features page for a more complete list of new features. | |
| 10250 | |
| 10251 EGCS 1.0.1 is a minor update to the EGCS 1.0 compiler to fix a few | |
| 10252 critical bugs and add support for Red Hat 5.0 Linux. Changes since the | |
| 10253 EGCS 1.0 release: | |
| 10254 * Add support for Red Hat 5.0 Linux and better support for Linux | |
| 10255 systems using glibc2. | |
| 10256 Many programs failed to link when compiled with EGCS 1.0 on Red Hat | |
| 10257 5.0 or on systems with newer versions of glibc2. EGCS 1.0.1 should | |
| 10258 fix these problems. | |
| 10259 * Compatibility with both EGCS 1.0 and GCC 2.8 libgcc exception | |
| 10260 handling interfaces. | |
| 10261 To avoid future compatibility problems, we strongly urge anyone who | |
| 10262 is planning on distributing shared libraries that contain C++ code | |
| 10263 to upgrade to EGCS 1.0.1 first. | |
| 10264 Soon after EGCS 1.0 was released, the GCC developers made some | |
| 10265 incompatible changes in libgcc's exception handling interfaces. | |
| 10266 These changes were needed to solve problems on some platforms. This | |
| 10267 means that GCC 2.8.0, when released, will not be seamlessly | |
| 10268 compatible with shared libraries built by EGCS 1.0. The reason is | |
| 10269 that the libgcc.a in GCC 2.8.0 will not contain a function needed | |
| 10270 by the old interface. | |
| 10271 The result of this is that there may be compatibility problems with | |
| 10272 shared libraries built by EGCS 1.0 when used with GCC 2.8.0. | |
| 10273 With EGCS 1.0.1, generated code uses the new (GCC 2.8.0) interface, | |
| 10274 and libgcc.a has the support routines for both the old and the new | |
| 10275 interfaces (so EGCS 1.0.1 and EGCS 1.0 code can be freely mixed, | |
| 10276 and EGCS 1.0.1 and GCC 2.8.0 code can be freely mixed). | |
| 10277 The maintainers of GCC 2.x have decided against including seamless | |
| 10278 support for the old interface in 2.8.0, since it was never | |
| 10279 "official", so to avoid future compatibility problems we recommend | |
| 10280 against distributing any shared libraries built by EGCS 1.0 that | |
| 10281 contain C++ code (upgrade to 1.0.1 and use that). | |
| 10282 * Various bugfixes in the x86, hppa, mips, and rs6000/ppc backends. | |
| 10283 The x86 changes fix code generation errors exposed when building | |
| 10284 glibc2 and the Linux dynamic linker (ld.so). | |
| 10285 The hppa change fixes a compiler abort when configured for use with | |
| 10286 RTEMS. | |
| 10287 The MIPS changes fix problems with the definition of LONG_MAX on | |
| 10288 newer systems, allow for command line selection of the target ABI, | |
| 10289 and fix one code generation problem. | |
| 10290 The rs6000/ppc change fixes some problems with passing structures | |
| 10291 to varargs/stdarg functions. | |
| 10292 * A few machine independent bugfixes, mostly to fix code generation | |
| 10293 errors when building Linux kernels or glibc. | |
| 10294 * Fix a few critical exception handling and template bugs in the C++ | |
| 10295 compiler. | |
| 10296 * Fix Fortran namelist bug on alphas. | |
| 10297 * Fix build problems on x86-solaris systems. | |
| 10298 | |
| 10299 EGCS 1.0.2 is a minor update to the EGCS 1.0.1 compiler to fix several | |
| 10300 serious problems in EGCS 1.0.1. | |
| 10301 * General improvements and fixes | |
| 10302 + Memory consumption significantly reduced, especially for | |
| 10303 templates and inline functions. | |
| 10304 + Fix various problems with glibc2.1. | |
| 10305 + Fix loop optimization bug exposed by rs6000/ppc port. | |
| 10306 + Fix to avoid potential code generation problems in jump.c. | |
| 10307 + Fix some undefined symbol problems in dwarf1 debug support. | |
| 10308 * g++/libstdc++ improvements and fixes | |
| 10309 + libstdc++ in the EGCS release has been updated and should be | |
| 10310 link compatible with libstdc++-2.8. | |
| 10311 + Various fixes in libio/libstdc++ to work better on Linux | |
| 10312 systems. | |
| 10313 + Fix problems with duplicate symbols on systems that do not | |
| 10314 support weak symbols. | |
| 10315 + Memory corruption bug and undefined symbols in bastring have | |
| 10316 been fixed. | |
| 10317 + Various exception handling fixes. | |
| 10318 + Fix compiler abort for very long thunk names. | |
| 10319 * g77 improvements and fixes | |
| 10320 + Fix compiler crash for omitted bound in Fortran CASE | |
| 10321 statement. | |
| 10322 + Add missing entries to g77 lang-options. | |
| 10323 + Fix problem with -fpedantic in the g77 compiler. | |
| 10324 + Fix "backspace" problem with g77 on alphas. | |
| 10325 + Fix x86 backend problem with Fortran literals and -fpic. | |
| 10326 + Fix some of the problems with negative subscripts for g77 on | |
| 10327 alphas. | |
| 10328 + Fixes for Fortran builds on cygwin32/mingw32. | |
| 10329 * platform specific improvements and fixes | |
| 10330 + Fix long double problems on x86 (exposed by glibc). | |
| 10331 + x86 ports define i386 again to keep imake happy. | |
| 10332 + Fix exception handling support on NetBSD ports. | |
| 10333 + Several changes to collect2 to fix many problems with AIX. | |
| 10334 + Define __ELF__ for rs6000/linux. | |
| 10335 + Fix -mcall-linux problem on rs6000/linux. | |
| 10336 + Fix stdarg/vararg problem for rs6000/linux. | |
| 10337 + Allow autoconf to select a proper install problem on AIX 3.1. | |
| 10338 + m68k port support includes -mcpu32 option as well as cpu32 | |
| 10339 multilibs. | |
| 10340 + Fix stdarg bug for irix6. | |
| 10341 + Allow EGCS to build on irix5 without the gnu assembler. | |
| 10342 + Fix problem with static linking on sco5. | |
| 10343 + Fix bootstrap on sco5 with native compiler. | |
| 10344 + Fix for abort building newlib on H8 target. | |
| 10345 + Fix fixincludes handling of math.h on SunOS. | |
| 10346 + Minor fix for Motorola 3300 m68k systems. | |
| 10347 | |
| 10348 EGCS 1.0.3 is a minor update to the EGCS 1.0.2 compiler to fix a few | |
| 10349 problems reported by Red Hat for builds of Red Hat 5.1. | |
| 10350 * Generic bugfixes: | |
| 10351 + Fix a typo in the libio library which resulted in incorrect | |
| 10352 behavior of istream::get. | |
| 10353 + Fix the Fortran negative array index problem. | |
| 10354 + Fix a major problem with the ObjC runtime thread support | |
| 10355 exposed by glibc2. | |
| 10356 + Reduce memory consumption of the Haifa scheduler. | |
| 10357 * Target specific bugfixes: | |
| 10358 + Fix one x86 floating point code generation bug exposed by | |
| 10359 glibc2 builds. | |
| 10360 + Fix one x86 internal compiler error exposed by glibc2 builds. | |
| 10361 + Fix profiling bugs on the Alpha. | |
| 10362 + Fix ImageMagick & emacs 20.2 build problems on the Alpha. | |
| 10363 + Fix rs6000/ppc bug when converting values from integer types | |
| 10364 to floating point types. | |
| 10365 | |
| 10366 The EGCS 1.0 releases include installation instructions in both HTML | |
| 10367 and plaintext forms (see the INSTALL directory in the toplevel | |
| 10368 directory of the distribution). However, we also keep the most up to | |
| 10369 date [2]installation instructions and [3]build/test status on our web | |
| 10370 page. We will update those pages as new information becomes available. | |
| 10371 | |
| 10372 And, we can't in good conscience fail to mention some [4]caveats to | |
| 10373 using EGCS. | |
| 10374 | |
| 10375 Update: Big thanks to Stanford for providing a high speed link for | |
| 10376 downloading EGCS (go.cygnus.com)! | |
| 10377 | |
| 10378 Download EGCS from ftp.cygnus.com (USA California) or go.cygnus.com | |
| 10379 (USA California -- High speed link provided by Stanford). | |
| 10380 | |
| 10381 The EGCS 1.0 release is also available many mirror sites. | |
| 10382 [5]Goto mirror list to find a closer site | |
| 10383 | |
| 10384 We'd like to thank the numerous people that have contributed new | |
| 10385 features, test results, bugfixes, etc. Unfortunately, they're far too | |
| 10386 numerous to mention by name. | |
| 10387 | |
| 10388 Please send FSF & GNU inquiries & questions to [6]gnu@gnu.org. There | |
| 10389 are also [7]other ways to contact the FSF. | |
| 10390 | |
| 10391 These pages are maintained by [8]the GCC team. | |
| 10392 | |
| 10393 | |
| 10394 For questions related to the use of GCC, please consult these web | |
| 10395 pages and the [9]GCC manuals. If that fails, the | |
| 10396 [10]gcc-help@gcc.gnu.org mailing list might help. | |
| 10397 Please send comments on these web pages and the development of GCC to | |
| 10398 our developer mailing list at [11]gcc@gnu.org or [12]gcc@gcc.gnu.org. | |
| 10399 All of our lists have [13]public archives. | |
| 10400 | |
| 10401 Copyright (C) Free Software Foundation, Inc., 51 Franklin St, Fifth | |
| 10402 Floor, Boston, MA 02110, USA. | |
| 10403 | |
| 10404 Verbatim copying and distribution of this entire article is permitted | |
| 10405 in any medium, provided this notice is preserved. | |
| 10406 Last modified 2007-07-25 [14]Valid XHTML 1.0 | |
| 10407 | |
| 10408 References | |
| 10409 | |
| 10410 1. http://gcc.gnu.org/egcs-1.0/features.html | |
| 10411 2. http://gcc.gnu.org/install/ | |
| 10412 3. http://gcc.gnu.org/egcs-1.0/buildstat.html | |
| 10413 4. http://gcc.gnu.org/egcs-1.0/caveats.html | |
| 10414 5. http://gcc.gnu.org/mirrors.html | |
| 10415 6. mailto:gnu@gnu.org | |
| 10416 7. http://www.gnu.org/home.html#ContactInfo | |
| 10417 8. http://gcc.gnu.org/about.html | |
| 10418 9. http://gcc.gnu.org/onlinedocs/ | |
| 10419 10. mailto:gcc-help@gcc.gnu.org | |
| 10420 11. mailto:gcc@gnu.org | |
| 10421 12. mailto:gcc@gcc.gnu.org | |
| 10422 13. http://gcc.gnu.org/lists.html | |
| 10423 14. http://validator.w3.org/check/referer | |
| 10424 ====================================================================== | |
| 10425 http://gcc.gnu.org/egcs-1.0/features.html | |
| 10426 EGCS 1.0 features | |
| 10427 | |
| 10428 * Core compiler is based on the gcc2 development tree from Aug 2, | |
| 10429 1997, so we have most of the [1]features found in GCC 2.8. | |
| 10430 * Integrated GNU Fortran compiler based on g77-0.5.22-19970929. | |
| 10431 * Vast improvements in the C++ compiler; so many they have [2]page of | |
| 10432 their own! | |
| 10433 * Integrated C++ runtime libraries, including support for most major | |
| 10434 linux systems! | |
| 10435 * New instruction scheduler from IBM Haifa which includes support for | |
| 10436 function wide instruction scheduling as well as superscalar | |
| 10437 scheduling. | |
| 10438 * Significantly improved alias analysis code. | |
| 10439 * Improved register allocation for two address machines. | |
| 10440 * Significant code generation improvements for Fortran code on | |
| 10441 Alphas. | |
| 10442 * Various optimizations from the g77 project as well as improved loop | |
| 10443 optimizations. | |
| 10444 * Dwarf2 debug format support for some targets. | |
| 10445 * egcs libstdc++ includes the SGI STL implementation without changes. | |
| 10446 * As a result of these and other changes, egcs libstc++ is not binary | |
| 10447 compatible with previous releases of libstdc++. | |
| 10448 * Various new ports -- UltraSPARC, Irix6.2 & Irix6.3 support, The SCO | |
| 10449 Openserver 5 family (5.0.{0,2,4} and Internet FastStart 1.0 and | |
| 10450 1.1), Support for RTEMS on several embedded targets, Support for | |
| 10451 arm-linux, Mitsubishi M32R, Hitachi H8/S, Matsushita MN102 and | |
| 10452 MN103, NEC V850, Sparclet, Solaris & Linux on PowerPCs, etc. | |
| 10453 * Integrated testsuites for gcc, g++, g77, libstdc++ and libio. | |
| 10454 * RS6000/PowerPC ports generate code which can run on all | |
| 10455 RS6000/PowerPC variants by default. | |
| 10456 * -mcpu= and -march= switches for the x86 port to allow better | |
| 10457 control over how the x86 port generates code. | |
| 10458 * Includes the template repository patch (aka repo patch); note the | |
| 10459 new template code makes repo obsolete for ELF systems using gnu-ld | |
| 10460 such as Linux. | |
| 10461 * Plus the usual assortment of bugfixes and improvements. | |
| 10462 | |
| 10463 Please send FSF & GNU inquiries & questions to [3]gnu@gnu.org. There | |
| 10464 are also [4]other ways to contact the FSF. | |
| 10465 | |
| 10466 These pages are maintained by [5]the GCC team. | |
| 10467 | |
| 10468 | |
| 10469 For questions related to the use of GCC, please consult these web | |
| 10470 pages and the [6]GCC manuals. If that fails, the | |
| 10471 [7]gcc-help@gcc.gnu.org mailing list might help. | |
| 10472 Please send comments on these web pages and the development of GCC to | |
| 10473 our developer mailing list at [8]gcc@gnu.org or [9]gcc@gcc.gnu.org. | |
| 10474 All of our lists have [10]public archives. | |
| 10475 | |
| 10476 Copyright (C) Free Software Foundation, Inc., 51 Franklin St, Fifth | |
| 10477 Floor, Boston, MA 02110, USA. | |
| 10478 | |
| 10479 Verbatim copying and distribution of this entire article is permitted | |
| 10480 in any medium, provided this notice is preserved. | |
| 10481 Last modified 2007-07-25 [11]Valid XHTML 1.0 | |
| 10482 | |
| 10483 References | |
| 10484 | |
| 10485 1. http://gcc.gnu.org/egcs-1.0/features-2.8.html | |
| 10486 2. http://gcc.gnu.org/egcs-1.0/c++features.html | |
| 10487 3. mailto:gnu@gnu.org | |
| 10488 4. http://www.gnu.org/home.html#ContactInfo | |
| 10489 5. http://gcc.gnu.org/about.html | |
| 10490 6. http://gcc.gnu.org/onlinedocs/ | |
| 10491 7. mailto:gcc-help@gcc.gnu.org | |
| 10492 8. mailto:gcc@gnu.org | |
| 10493 9. mailto:gcc@gcc.gnu.org | |
| 10494 10. http://gcc.gnu.org/lists.html | |
| 10495 11. http://validator.w3.org/check/referer | |
| 10496 ====================================================================== | |
| 10497 http://gcc.gnu.org/egcs-1.0/caveats.html | |
| 10498 EGCS 1.0 Caveats | |
| 10499 | |
| 10500 * EGCS has an integrated libstdc++, but does not have an integrated | |
| 10501 libg++. Furthermore old libg++ releases will not work with egc; HJ | |
| 10502 Lu has made a libg++-2.8.1.2 available which may work with EGCS. | |
| 10503 Note most C++ programs only need libstdc++. | |
| 10504 * Note that using -pedantic or -Wreturn-type can cause an explosion | |
| 10505 in the amount of memory needed for template-heavy C++ code, such as | |
| 10506 code that uses STL. Also note that -Wall includes -Wreturn-type, so | |
| 10507 if you use -Wall you will need to specify -Wno-return-type to turn | |
| 10508 it off. | |
| 10509 * Exception handling may not work with shared libraries, particularly | |
| 10510 on alphas, hppas, and mips based platforms. Exception handling is | |
| 10511 known to work on x86-linux platforms with shared libraries. | |
| 10512 * Some versions of the Linux kernel have bugs which prevent them from | |
| 10513 being compiled or from running when compiled by EGCS. See the FAQ | |
| 10514 (as shipped with EGCS 1.0) for additional information. | |
| 10515 * In general, EGCS is more rigorous about rejecting invalid C++ code | |
| 10516 or deprecated C++ constructs than G++ 2.7. As a result it may be | |
| 10517 necessary to fix C++ code before it will compile with EGCS. | |
| 10518 * G++ is also aggressively tracking the C++ standard; as a result | |
| 10519 code which was previously valid (and thus accepted by other | |
| 10520 compilers and older versions of G++) may no longer be accepted. | |
| 10521 * EGCS 1.0 may not work with Red Hat Linux 5.0 on all targets. EGCS | |
| 10522 1.0.x and later releases should work with Red Hat Linux 5.0. | |
| 10523 | |
| 10524 Please send FSF & GNU inquiries & questions to [1]gnu@gnu.org. There | |
| 10525 are also [2]other ways to contact the FSF. | |
| 10526 | |
| 10527 These pages are maintained by [3]the GCC team. | |
| 10528 | |
| 10529 | |
| 10530 For questions related to the use of GCC, please consult these web | |
| 10531 pages and the [4]GCC manuals. If that fails, the | |
| 10532 [5]gcc-help@gcc.gnu.org mailing list might help. | |
| 10533 Please send comments on these web pages and the development of GCC to | |
| 10534 our developer mailing list at [6]gcc@gnu.org or [7]gcc@gcc.gnu.org. | |
| 10535 All of our lists have [8]public archives. | |
| 10536 | |
| 10537 Copyright (C) Free Software Foundation, Inc., 51 Franklin St, Fifth | |
| 10538 Floor, Boston, MA 02110, USA. | |
| 10539 | |
| 10540 Verbatim copying and distribution of this entire article is permitted | |
| 10541 in any medium, provided this notice is preserved. | |
| 10542 Last modified 2007-07-25 [9]Valid XHTML 1.0 | |
| 10543 | |
| 10544 References | |
| 10545 | |
| 10546 1. mailto:gnu@gnu.org | |
| 10547 2. http://www.gnu.org/home.html#ContactInfo | |
| 10548 3. http://gcc.gnu.org/about.html | |
| 10549 4. http://gcc.gnu.org/onlinedocs/ | |
| 10550 5. mailto:gcc-help@gcc.gnu.org | |
| 10551 6. mailto:gcc@gnu.org | |
| 10552 7. mailto:gcc@gcc.gnu.org | |
| 10553 8. http://gcc.gnu.org/lists.html | |
| 10554 9. http://validator.w3.org/check/referer | |
| 10555 ====================================================================== |