annotate TaskManager/ChangeLog @ 304:e679d401eab3

fix Makefile.cell
author kono@localhost.localdomain
date Mon, 08 Jun 2009 18:28:18 +0900
parents 99dccf47e684
children 7075842aa155 81b25e5d5379
Ignore whitespace changes - Everywhere: Within whitespace: At end of lines:
rev   line source
279
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents: 274
diff changeset
1 2009-06-4 Shinji KONO <kono@ie.u-ryukyu.ac.jp>
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents: 274
diff changeset
2
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents: 274
diff changeset
3 set_symbol は、もういらないよね?
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents: 274
diff changeset
4
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents: 274
diff changeset
5 list dma 中心の実装にして、もっと細かく read/exec/write
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents: 274
diff changeset
6 した方が良いかも。
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents: 274
diff changeset
7
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents: 274
diff changeset
8 post で、PPE task を渡せると良い。address は parameterとして送る
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents: 274
diff changeset
9
274
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
10 2009-02-13 Wataru MIYAGUNI <gongo@cr.ie.u-ryukyu.ac.jp>
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
11
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
12 * kernel/ppe/Random.cc (reset): fix
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
13 urandom -> random とどれも読めなかったら
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
14 gettimeofday() での時間から seed を求めるように
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
15
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
16 2009-02-12 Wataru MIYAGUNI <gongo@cr.ie.u-ryukyu.ac.jp>
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
17
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
18 * add: kernel/ppe/Random.cc
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
19 乱数生成クラス。
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
20 ゲームだとユーザ使うでしょうきっと。
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
21 一応 /dev/random から seed 取る様にしてます
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
22
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
23 2009-02-04 Wataru MIYAGUNI <gongo@cr.ie.u-ryukyu.ac.jp>
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
24
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
25 * kernel/ppe/TaskManager.cc (TaskManager::allocate): rename
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
26 malloc -> allocate
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
27
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
28 * kernel/main.cc (main): fix
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
29 cerium_main を呼ぶのではなく、TMmain という名前にしました。
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
30 ちょっと SDLmain をパクった感じで。
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
31 まあ TaskManager の main で cerium_* って名前は微妙に変だからね。
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
32
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
33 * kernel/ppe/TaskManager.cc (TaskManager::set_TMend): add
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
34 cerium_main があるんだから、cerium_end があってもいいじゃない。
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
35 もっと言うと、TaskManager に main を隠すって流れなんだけど
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
36 終了を検知できないのはちとやりづらいかなと。
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
37 たとえば測定とか。Task の post_func とかでもやれないことはないけどね。
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
38 というわけで、ユーザが、プログラム終了時に呼ばれる関数を設定できるように。
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
39
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
40 2009-01-20 Wataru MIYAGUNI <gongo@cr.ie.u-ryukyu.ac.jp>
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
41
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
42 * Cell/spe/SchedTask.cc (SchedTask::get_cpuid): add
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
43 printf デバッグ時に、どの CPU かって知りたい時があるので。
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
44
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
45 PPE = 0
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
46 SPE = 0〜spu_num-1;
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
47
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
48 PPE は 0 以外に分かりやすい数字がいいんだけどなー。SPE と被るし。
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
49 -1 とかはエラーっぽいから好かない。まあいいんだけど。
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
50
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
51 User Task では以下の様に使用します
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
52
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
53 int cpuid = smanager->get_cpuid();
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
54
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
55
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
56 * Cell/SpeThreads.cc (SpeThreads::spe_thread_run): fix
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
57 SPE_EXIT が出る時は正常終了だけど、これだと
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
58 エラーでたようなメッセージに見えてしまう(俺がそう見えてしまった)ので
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
59 ここは表示しなくてもいいかな。
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
60
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
61 * kernel/ppe/MailManager.cc (MailManager::destroy): fix
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
62 無限ループになってた。この for() 間違ってるかな。
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
63 なんか TaskQueueInfo.cc とかでも、結局 while() に直してるし。
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
64
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
65 * kernel/ppe/TaskManager.cc (TaskManager::~TaskManager): add
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
66 kernel/main.ccで
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
67
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
68 delete manager;
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
69
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
70 としているのに、TaskManagerImpl::~TaskManagerImpl が呼び出されず
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
71 どうしてかなと思ったら、そもそも ~TaskManager が無かった。あほか
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
72
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
73 2009-01-05 Wataru MIYAGUNI <gongo@cr.ie.u-ryukyu.ac.jp>
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
74
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
75 * all : fix
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
76 Scheduler::curIndex_taskList を削除し、
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
77 SchedTask に持たせる様に変更。(SchedTask::__cur_index)
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
78 それに伴い、SchedTask::__init__() も cur_index を入れる様に変更
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
79
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
80 2008-12-24 Wataru MIYAGUNI <gongo@cr.ie.u-ryukyu.ac.jp>
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
81
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
82 * kernel/schedule/SchedTask.cc (SchedTask::ex_init_renew)
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
83 (SchedTask::ex_init_normal): add
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
84 (SchedTask::__init__): fix
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
85
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
86 init でも ex_init を使える様に。
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
87 あと、コンストラクタで渡していた引数を __init__() に渡す様にした。
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
88 コンストラクタの引数あると、継承する時にいちいち親クラスのも書かないと
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
89 いけなかった。これ省略できないんだよな。めんどくさい。
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
90
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
91 例.
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
92 class Hoge : public SchedTask {
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
93 Hoge(int i) : Task(i) {}
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
94 };
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
95
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
96 なので、今までは Scheduler.h に SchedConstructor ってマクロを書いて
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
97 クラス名入れるだけで上の様な形になるようにしていた。
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
98 でも、例えば
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
99
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
100 SchedTask -> Hoge -> Fuge っていうように Fuge ってタスクを
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
101 作りたいとき、上のままだと SchedTask に引数渡してしまうのでだめ。
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
102 もうめんどくさいってことで、コンストラクタ全てデフォルトにして、
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
103 __init__() の引数に渡す様にしました。
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
104
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
105 (SchedTask::__set_renewFlag): add
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
106
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
107 ここで、PPEで生成されたか(normal)、SPE で生成されたか(renew) の
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
108 判定を行い、ex_xxx の設定もする
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
109
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
110 (SchedTask::get_inputSize, SchedTask::get_outputSize): add
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
111
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
112 アドレスだけじゃなく、そのサイズも取れた方がいいだろう
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
113
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
114
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
115 2008-12-23 Wataru MIYAGUNI <gongo@cr.ie.u-ryukyu.ac.jp>
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
116
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
117 * Cell/spe/SchedTask.cc (SchedTask::get_outputAddr)
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
118 (SchedTask::get_inputAddr): add
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
119
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
120 in/out のデータだけじゃなく、そのアドレスも取れた方がいいだろう
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
121
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
122 2008-12-22 Wataru MIYAGUNI <gongo@cr.ie.u-ryukyu.ac.jp>
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
123
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
124 * Cell/spe/SchedTask.cc (SchedTask::__init__, SchedTask::read)
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
125 (SchedTask::exec, SchedTask::write): fix
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
126 (SchedTask::ex_read_normal, SchedTask::ex_read_renew)
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
127 (SchedTask::ex_exec_normal, SchedTask::ex_exec_renew)
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
128 (SchedTask::ex_write_normal, SchedTask::ex_write_renew): add
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
129
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
130 SPE 内で生成されたタスクは、PPE で生成されたものと違い
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
131
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
132 - add->inData
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
133 : PPE から DMA or SPE 内のものをそのまま使う
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
134 - PPE にタスクが終了したことを知らせる
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
135 : 生成されたタスクを待つ必要があるなら、その時点では送らない
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
136
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
137 とか、まあいろいろ処理が違うわけです。
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
138 そして、タスク内生成タスクの判断をする
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
139
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
140 __flag_renewTask ? 0 = PPE で生成 : 1 = SPE で生成
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
141
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
142 という変数がある。これでいくつか処理を分けてるんだけど、
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
143 今までは
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
144
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
145 if (__flag_renewTask) {
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
146 } else {
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
147 }
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
148
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
149 ってやってた。これではいかんという事で、
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
150 __init__() 内で、関数ポインタに、
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
151
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
152 ex_xxxx_normal: PPE で生成されたタスクに対する処理
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
153 ex_xxxx_renew: SPE で生成されたタスクに対する処理
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
154
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
155 と入れて、if 文無しでやってみた。
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
156 今は ex_write_xxx しか書いてないが、これからread/exec でも
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
157 出てくると思うので、作っておいた
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
158
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
159
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
160 2008-12-19 Wataru MIYAGUNI <gongo@cr.ie.u-ryukyu.ac.jp>
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
161
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
162 * Cell/spe/CellDmaManager.cc (CellDmaManager::dma_wait)
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
163 (CellDmaManager::mail_write, CellDmaManager::mail_read): fix
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
164 writech、readch の関数を、wrap (って言い方でおk?)された関数に変更。
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
165 最適化掛かってるっぽいし、長いよりはわかりやすいし。そのための wrap。
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
166
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
167 例:
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
168 - before
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
169 spu_readch(SPU_RdInMspu_readch(SPU_RdInMbox);
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
170 - after
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
171 spu_read_in_mbox(void);
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
172
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
173 2008-11-05 Wataru MIYAGUNI <gongo@cr.ie.u-ryukyu.ac.jp>
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
174
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
175 * add: Task 内での API
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
176 Task 外での API は、今まで通り manager->create_task とかですが
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
177 タスク内でも、「オブジェクト->関数」の呼び出しがいいんじゃないか
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
178 って話になったので、付け加えました。今のところ、SchedTask.h の
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
179 内部クラスとして
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
180
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
181 STaskManager
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
182
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
183 ってのを加えて、ユーザはそのインスタンスである
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
184
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
185 smanager
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
186
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
187 からAPIにアクセスします。
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
188 今までは __scheduler->dma_load とかいろいろやってたんですが
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
189 これからは全て smanager にしました。
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
190 というわけで、ここに使える API 一覧。いずれゲーム班 wikiの方にも。
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
191
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
192 - get_input, get_output, get_param
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
193 - create_task, wait_task
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
194 - global_alloc, global_get, global_free
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
195 - mainMem_alloc, mainMem_wait, mainMem_get
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
196 - dma_load, dma_store, dma_wait
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
197 - allocate
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
198
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
199 使い方は追々描きますが、
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
200 今のところ上に変更しなくてもそのままの記述で動くはずです。
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
201 いずれは全て移行してもらうことになりますがきっと。
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
202
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
203
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
204 * kernel/schedule/SchedTask.cc:
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
205 いろいろ関数が増えてますが、ラッパーです。
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
206
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
207 2008-11-01 Wataru MIYAGUNI <gongo@cr.ie.u-ryukyu.ac.jp>
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
208
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
209 * add: kernel/main.cc
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
210 main loop をユーザに書かせるのはめんどくさいので、
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
211 ライブラリ側で main() を書く事にしました。
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
212 ユーザ側では main() の代わりに cerium_main() を
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
213 書かせるようにしています。引数は main() のをそのまま渡す感じで。
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
214
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
215 Cerium 標準のオプションとして、-cpu は付けました。
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
216 ゲームフレームワークってことで、-width とか -height は
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
217 標準でつけてもいいかなって話なので、これは後日実装。
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
218 標準オプションで受け取った値にアクセスする方法も考えないと。
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
219 manager->cpu とか manager->width とかは安易か?
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
220
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
221 * add: Cell/PpeScheduler.cc
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
222 MainScheduler をそのまま使うと、
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
223 PPE のタスクで mainMem_alloc で確保した領域がアライメント
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
224 取れていないため、SPE で使うと余裕でバスエラー。
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
225
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
226 Scheduler->allocate で poxis_memalign で使えるように。
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
227
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
228 * move: kernel/schedule/FifoDmaManager.cc, MainScheduler.cc
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
229 kernel というよりは Fifo バージョン用なので Fifo/ に移動。
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
230
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
231
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
232 2008-10-21 Wataru MIYAGUNI <gongo@cr.ie.u-ryukyu.ac.jp>
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
233
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
234 * kernel/ppe/TaskManagerImpl.cc (TaskManagerImpl::systask_init): fix
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
235 下に述べてる SysTask_Finish を regist する部分
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
236
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
237 (TaskManagerImpl::spawn_task):
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
238
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
239 SysTask_Finish に対して、タスクが spawn されるたびに
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
240 wait_for を掛けて、待つようにしている。
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
241
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
242 * add: kernel/systask/
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
243 久々の更新乙
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
244
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
245 プログラム動かすとき、タスクが SPE だけで、
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
246 PPE で待ってるタスクが無いとそのままプログラムが素通りするってことで
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
247 今まではユーザに、全てのタスクを待たせるタスクってのを書かせてた。
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
248 まあもちろんめんどくさいので、いい加減追加した。
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
249
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
250 system task っつーことで、spawn された全てのタスクを待つ
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
251 SysTask_Finish を作った。これでいちいち task_finish とか作らなくておk
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
252
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
253 2008-08-10 Wataru MIYAGUNI <gongo@cr.ie.u-ryukyu.ac.jp>
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
254
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
255 * thinking: add_update() ?
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
256 現在、タスクは input/output があるわけですよ。
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
257 で、例えば
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
258
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
259 - 入力データ : PolygoPpack
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
260 - 出力データ : SpanPack
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
261
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
262 ってなわけですが、別のタスクで
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
263
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
264 - 入力データ : SceneGraphPack (更新前)
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
265 - 出力データ : SceneGraphPack (更新後)
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
266
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
267 ってのがある。つまり Update なわけだ。
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
268 今のところ、同じアドレスを add_inData, add_outData に設定して
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
269 タスク内で memcpy(wbuf, rbuf, sizeof(SceneGraphPack) とかしてる。
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
270 まあそれはそれでいいのかわるいのか。
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
271
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
272 in/out だけじゃなくて update も必要?
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
273
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
274 2008-08-08 Wataru MIYAGUNI <gongo@cr.ie.u-ryukyu.ac.jp>
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
275
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
276 * add: ../include/TaskManager/base.h
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
277 通常の new/delete では、RTTI とか 例外処理とかで
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
278 -fno-exceptions や -fno-rtti をコンパイルオプションにしても
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
279 効かなかったんだけど、operator new/delete をオーバーライドして
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
280 中身を普通の malloc/free にすると、上の処理が無くなるので
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
281 オプションが効くようになる。結果コードサイズも減ると。
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
282 SPE の場合、70〜80KBは減りました。使わない手は無い。
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
283 つーことで、一応動いてる。。。といいたけど動いてないorz
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
284 最適化 (-O2 とか -O9) をかけると止まる。SPE 上でね。
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
285 FIFO バージョンだと問題ない。SPEだとだめだ。
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
286 今わかってる、止まる場所は Scheduler::run() 内の
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
287
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
288 task3->write();
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
289
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
290 だ。task1~3までのnewは(多分)できているんだけど
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
291 そこを呼び出すと SPE 自体が終了してしまう。謎だ
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
292
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
293 一応、俺作の new/delete は base.h に定義してあって、
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
294 通常の API との切り替えは、base.h にある
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
295 BASE_NEW_DELETE を切り替えるだけでおk。
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
296 全てのファイルではなく、現在は SPE で使いそうなところだけやってます。
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
297 いずれは全部やったほうがいいかな〜
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
298
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
299 ライブラリ側の最適化はアウトだけど、ユーザ側では問題ないです。
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
300 なので、今は
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
301
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
302 ライブラリ側(libspemanager.a)は最適化無し(-O0)
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
303 ユーザ側(SchedTaskを継承したやつね)は最適化しても無問題 (-O9)
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
304
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
305 でっす。ここらへん完璧なれば、だいぶ楽になる。
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
306 つーかもう C++ やめ(ry
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
307
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
308
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
309 2008-08-07 Wataru MIYAGUNI <gongo@cr.ie.u-ryukyu.ac.jp>
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
310
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
311 * change: mainMem_set -> mainMem_wait
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
312 allocate を待つんだから、なんとなく wait かな。
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
313 あと、ユーザも使えるので、wait の方がわかりやすいと思ったり。
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
314
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
315 2008-08-05 Wataru MIYAGUNI <gongo@cr.ie.u-ryukyu.ac.jp>
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
316
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
317 * add: mainMem_alloc, mainMem_set, mainMem_get
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
318 SPE から メインメモリの領域に対して allocate できないと
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
319 SceneGraphの生成やら、結構クリティカルな処理を
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
320 全部 PPE でやらないといけなくなるってことで実装しました。
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
321
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
322 流れとして
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
323
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
324 1 タスク中に、mainMem(id,size) を実行する事で、
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
325 メインメモリに対して allocate のコマンドを発行。
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
326
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
327 1.1 Scheduler から PPE に対して
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
328 - commmand (MY_SPE_COMMAND_MALLOC)
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
329 - id (PPEから戻ってくるコマンドに必要)
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
330 - size
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
331 を mailbox で送る
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
332
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
333 1.2 確保した領域はそのタスク内では取得できない(NULL が来ます)
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
334 正確には、返事の mail をここでは read してないから
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
335
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
336 2. PPE では、受信した mail が MY_SPE_COMMAND_MALLOC だったら
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
337 次に来る mail が id と size であるとして read を行い、
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
338 size を元に allocate する。allocate が完了したら
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
339 - id
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
340 - allocate された領域のアドレス
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
341 を SPE に mail で送る
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
342
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
343 3. SPE Scheduler では、SchedTaskList::read で、
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
344 一つ前の TaskList 中で実行された mainMem_alloc の数だけ
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
345 PPE からのメールを待つ。mainMem_set() の処理です。
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
346
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
347 4. create_task されたタスク内で mainMem_get(id) とすると
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
348 allocate したメインメモリ領域のアドレスが得られる。
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
349
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
350 こんな感じ。結構ださい実装なので、もうちょいスマートにいきたいよね。
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
351 例題は Game_project/student/master/gongo/MainMemMalloc にあります。
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
352 README にもおんなじこと書いてます。
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
353
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
354 * memo: The number of available entries of Inbound/Outbound Mailbox
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
355 Outbound (SPE -> PPE) のmailboxデータキュー保持数は
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
356
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
357 /* SPE プログラム中 */
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
358 #include <spu_mfcio.h>
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
359 spu_stat_out_mbox(void);
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
360
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
361 で調べる事が出来る。
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
362
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
363 --- 記述例 ---
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
364 printf("Available capacity of SPU Outbound Mailbox\n");
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
365 printf(" %d\n", spu_stat_out_mbox());
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
366
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
367 --- 実行結果 --
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
368 Available capacity of SPU Outbound Mailbox
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
369 1
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
370
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
371 Inbound (PPE -> SPE) の mailbox データキュー保持数は
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
372
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
373 /* PPE プログラム中 */
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
374 #include <libspe2.h>
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
375 spe_in_mbox_status(spe_context_ptr_t);
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
376
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
377 で調べられます。
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
378
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
379 --- 記述例 ---
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
380 printf("the number of available entries = %d\n",
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
381 spe_in_mbox_status(spe_ctx));
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
382
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
383 --- 実行結果 ---
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
384 the number of available entries = 4
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
385
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
386 Outbound が少ないなー。
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
387 In/Out 共に、キューが MAX の場合、減るまで wait 掛かるんだよな。
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
388 それがどこまで影響あるかは実際にやらないと、ってことか。
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
389
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
390 * fix: ファイル名の変更 (*.cpp -> *.cc)
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
391 前々から先生に直せ言われてたので。
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
392
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
393 cvs のファイル名を変える方法は簡単に二つ(てかこれだけ?)
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
394
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
395 1. cvs rm hoge.cpp; cvs add hoge.cc
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
396 2. リポジトリを直接変更 mv hoge.cpp,v hoge.cc,v
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
397
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
398 めんどくさかったので 2 でやりました。
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
399 Attic (削除されたファイルがあるリポジトリディレクトリ?)にも
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
400 同じ処理を行えば、tag で update かけてもちゃんと反映されました。
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
401
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
402 2008-07-22 Wataru MIYAGUNI <gongo@cr.ie.u-ryukyu.ac.jp>
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
403
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
404 * tag: open-campus-2008
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
405 次やる事は、Cell/spe 以下のコードサイズを減らすために
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
406 new/delete を消して malloc/free で統一する事。
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
407 placement_new ってのを使えば、コンストラクタは呼べて
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
408 new ほどサイズ圧迫しないからこれにしようかな。
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
409 逆の placement_delete ってのは自分で小細工しないと行けないらしい。
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
410 まあ、これが旨く行けば 80KB ほど減るから。やるべきだろう。
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
411
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
412 * Cell/spe/Scheduler.cpp (Scheduler::dma_load): アホなミスその2
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
413 自分で __scheduler->dma_store をやってもデータが送れない。
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
414 そんな馬鹿な。つーことでいろいろ調べてもわからない。
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
415 アドレスやサイズが違うのかと調べても違う。
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
416 こうなったらっつーことでライブラリに printf 加えてみたら表示されない
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
417 あれ、おかしいな。たしかに Connector::dma_store に加えたはz・・
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
418
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
419 Scheduler::dma_store(void *buf, uint32 addr, uint32 size, uint32 mask)
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
420 {
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
421 <<<
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
422 connector->dma_load(buf, addr, size, mask);
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
423 ========
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
424 connector->dma_store(buf, addr, size, mask);
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
425 >>>
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
426 }
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
427
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
428 なぜ store から load を呼んでるのか不思議だった。
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
429 Scheduler::dma_load をコピペして dma_store にした後、
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
430 中の connector->dma_load を変えなかったってオチだな。
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
431 下のミスと合わせて5,6時間費やしたよHAHAHA
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
432
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
433 * Cell/spe/SchedTask.cpp (SchedTask::exec): アホなミスその1
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
434 Test/test_render で、
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
435 SpanPack のデータが時々壊れてる感じがする。
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
436 送る前までは正常だから生成に問題は無いはず。
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
437 つーことでいろいろ調べたがわからず。
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
438 printf デバッグすると動く不思議
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
439 なんだ、printf で遅くなったらできるってことは
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
440 DMA が完了する前に SchedTask::run にきてんのか?
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
441 いやいや、そんなばかな。だってちゃんと wait し・・・
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
442
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
443 <<<
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
444 ============
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
445 __scheduler->dma_wait(DMA_READ);
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
446 >>>
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
447
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
448 はいはい wait し忘れ wait し忘れ
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
449
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
450 2008-07-16 Wataru MIYAGUNI <gongo@cr.ie.u-ryukyu.ac.jp>
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
451
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
452 * memo: if 文消した成果2 & memcpy するかしないかか
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
453 Renew Task では、inListData,outListData は新たに allocate して
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
454 使っているので、SchedTask にそって実行する場合、
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
455
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
456 __scheduler->dma_load(__inListData, (uint32)__task->inData,
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
457 sizeof(ListData), DMA_READ_IN_LIST);
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
458 __scheduler->dma_load(__outListData, (uint32)__task->outData,
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
459 sizeof(ListData), DMA_READ_OUT_LIST);
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
460
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
461 の代わりに
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
462
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
463 memcpy(__inListData, __task->inData, sizeof(ListData));
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
464 memcpy(__outListData, __task->outData, sizeof(ListData));
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
465 free(__task->inData);
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
466 free(__task->outData);
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
467
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
468 もしくは
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
469
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
470 __inListData = __task->inData;
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
471 __outListData = __task->outData;
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
472 (__task->inData と __task->outData は Destructor で free する)
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
473
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
474 とやっています。
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
475 memcpy が重いのはわかるんですが、下の方法では
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
476 Destructor で if 文使って free() しているわけです(このタスクが Renew か否か)。
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
477 ですので、どっちが早いか試してみた。
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
478
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
479 /**
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
480 * memcpy() して、すぐ free() する version
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
481 */
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
482 void
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
483 test_cpy(int flag, int *src)
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
484 {
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
485 if (flag) {
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
486 memcpy(data, src, sizeof(int)*length);
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
487 free(src);
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
488 }
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
489 }
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
490
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
491 /**
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
492 * 参照で扱って、最後に free() する version
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
493 */
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
494 void
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
495 test_nocpy(int flag, int *src)
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
496 {
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
497 if (flag) {
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
498 data = src;
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
499 }
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
500
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
501 // この部分を SchedTask::~SchedTask() と
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
502 // 思ってください
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
503 if (flag) {
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
504 free(data);
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
505 }
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
506 }
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
507
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
508
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
509 これらの関数を10000回ループしました。
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
510 src の allocate は関数の外でやっており、その部分は実行時間に含まれてません
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
511 flag は 1 or 0 の繰り返しです。
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
512
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
513 - 実行結果 (1)
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
514 :no copy
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
515 SPE time by SPU Decrementer: 0.035500
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
516 :copy
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
517 SPE time by SPU Decrementer: 0.057500
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
518
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
519 memcpy しないほうが速いらしいです。
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
520 ためしに、flag を ずっと 1 にしてみました。
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
521
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
522 - 実行結果 (2)
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
523 :no copy
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
524 SPE time by SPU Decrementer: 0.055250
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
525 :copy
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
526 SPE time by SPU Decrementer: 0.053389
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
527
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
528 今度は copy するほうが早いという不思議。
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
529 でもまあ、ずっと 1 ってことはないと思いますし、
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
530 むしろ flag == 1 になるほうが少ないと思うので、
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
531 no_copy version でやったほうがいいかな。
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
532
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
533 おまけで、実行結果 (1) の環境で、test_nocpy を変えてみた
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
534
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
535 void
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
536 test_nocpy(int flag, int *src)
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
537 {
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
538 if (flag) {
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
539 data = src;
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
540 }
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
541
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
542 free((void*)(flag*(int)data));
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
543 }
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
544
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
545 キャストしまくりですが、単純に free(flag*data) だと
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
546 「invalid operands of types 'int' and 'int*' to binary 'operator*'」って
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
547 出るので、キャストで逃げました。
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
548 で、実行結果なんですが
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
549
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
550 - 実行結果 (3)
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
551 :no copy
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
552 SPE time by SPU Decrementer: 0.040375
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
553 :copy
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
554 SPE time by SPU Decrementer: 0.059500
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
555
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
556 遅くなってーら。キャストが悪いのか。乗算が重いのか。
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
557 branch が無い? spe の if 文と対決しても遅いのかー。
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
558 例題が間違ってる可能性もあるが・・・ if 文は使っていくかなー
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
559
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
560
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
561 2008-07-10 Wataru MIYAGUNI <gongo@cr.ie.u-ryukyu.ac.jp>
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
562
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
563 * fix: TaskGroup->group
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
564 今まで slist っていう、ライブラリの単方向リスト構造体?を
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
565 使ってたんだけど、まあいろいろあって、TaskQueue を使うようにしました。
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
566 最初からこれにするつもりではあったけどね。
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
567 RenewTask や static_alloc とかの実装を優先したので
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
568 ライブラリを使いました。といっても、書いてみると
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
569 それほと記述量無いので最初から行っても良かったかなーと思ったり。
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
570
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
571 そんなわけで動いてます。つーか、やめてよかったよ slist。
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
572 slist を使ったやつと使ってない奴のファイルサイズがやばい
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
573
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
574 -rwxr-xr-x 1 gongo gongo 120672 2008-07-10 14:29 spe-main*
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
575 -rwxr-xr-x 1 gongo gongo 180368 2008-07-10 13:40 spe-main.bak*
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
576
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
577 .bak が slist を使ってる、上のやつが使ってないversionです。
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
578 まさか 60k も違ってくるとは思わなかった。
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
579 SPE LS の容量が 256k と考えると、かなりの痛手だったよ。アブねえ。
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
580 インラインとか最適化掛けまくってて、コード量が増えてるからかなー。
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
581
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
582 「SPU C/C++ 言語拡張」とかで、C++ のライブラリがSPUでも使えるよ〜
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
583 って書いてたから入れてみたんだけど。罠だったか。
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
584 おそらく SPU に移植した側の人も「サイズが増えるのを覚悟で使え」って
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
585 ことだったんだろう。なかったら文句言う人も居そうだし。
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
586
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
587 2008-07-09 Wataru MIYAGUNI <gongo@cr.ie.u-ryukyu.ac.jp>
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
588
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
589 * fix: TaskGroup での task の扱い
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
590 下にもかいているけど (直したいところ (1))
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
591 TaskGroup->group が持つ要素は int で持ってて、
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
592 それらは、同じく TaskGroup が持つ cur_id をインクリメントしていって、
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
593 それを要素としていました。つまり、TaskGroup->group は、厳密にいえば
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
594 「どの Task があるか」ではなく、「いくつのタスクがあるか」を
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
595 あらわしているだけでした。slist を使う意味もなかったわけです。
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
596
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
597 そこで、SchedTask が持つ、RenewTaskList の解放のタイミングを
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
598 RenewTaskList の一番最後のタスクが delete されるときにしました。
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
599 これによって、アドレスが被ることがなくなったので
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
600 TaskGroup->group の要素を TaskPtr にできました。
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
601 この方が、TaskGroup の意味的にもしっくりくるのでよかばってん。
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
602
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
603 * memo: if 文消した成果
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
604
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
605 #ifdef FREE_TEST
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
606 free((ListDataPtr)(__flag_renewTask*(int)(__inListData)));
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
607 free((ListDataPtr)(__flag_renewTask*(int)(__outListData)));
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
608 free((TaskListPtr)(__flag_renewTask*(int)(__list)));
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
609 #else
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
610 if (__flag_renewTask) {
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
611 free(__inListData);
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
612 free(__outListData);
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
613 free(__list);
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
614 }
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
615 #endif
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
616
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
617 こんな感じで、いくつかか if 文を消してみた。
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
618 そして、PPE側の main.cc で gettimeofday で計測してみた (各10回)
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
619
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
620
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
621 - if 文消した場合
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
622 time: 1.222000
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
623 time: 1.230000
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
624 time: 1.241000
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
625 time: 1.230000
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
626 time: 1.223000
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
627 time: 1.257000
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
628 time: 1.219000
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
629 time: 1.228000
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
630 time: 1.220000
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
631 time: 1.229000
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
632 avarage: 1.2299
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
633
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
634 - if 文消してない場合
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
635 time: 1.225000
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
636 time: 1.215000
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
637 time: 1.229000
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
638 time: 1.218000
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
639 time: 1.223000
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
640 time: 1.214000
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
641 time: 1.225000
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
642 time: 1.215000
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
643 time: 1.224000
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
644 time: 1.219000
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
645 avarage: 1.2207
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
646
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
647 あまり変わらな(ryむしr(ry
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
648 使い方がまずいのか、もっとと回数を増やせば変わってくるのかね。。。
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
649 PPE でなく、 SPE のほうで計測すべきなのかなーとか思ったり思わなかったり。
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
650
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
651
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
652 2008-07-08 Wataru MIYAGUNI <gongo@cr.ie.u-ryukyu.ac.jp>
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
653
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
654 * add: Renew Task の wait
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
655 Renew Task は今まで「生成されたやつ全部待つ」だったのを
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
656
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
657 void SchedTask::wait_task(TaskPtr task);
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
658
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
659 ってのを作って、任意のタスクに wait 掛けれるようにしました。
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
660 名前が思いつかなかったお。。。
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
661 動作確認済み・・・だと思います。例題・・・誰か例題を!(俺が
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
662
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
663
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
664 * fix: SchedTask の変数名
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
665 ユーザが継承して使う SchedTask クラスなんですが、
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
666 今まで変数は list, task などを使ってました。
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
667 が、これは一般に使われやすい変数名です。
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
668 その証拠に、俺も例題書いている時に task って名前が被ってました。
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
669
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
670 run(r, w)
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
671 {
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
672 ...
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
673
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
674 //TaskPtr task; <= 宣言してないのにエラーにならない
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
675 task = create_task(TASK_EXEC);
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
676 }
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
677
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
678 ってコードを書いてたせいで、Scheduler が使用する task を
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
679 上書きしたせいでバグってました。ってことがありました。
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
680 上のように、宣言してないのに普通に通ってるのを気づきませんでした。
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
681 今のところ変数名は __task とか __list にしてあります。
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
682 private にしてもいいんだけどさ。
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
683
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
684
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
685 2008-07-07 Wataru MIYAGUNI <gongo@cr.ie.u-ryukyu.ac.jp>
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
686
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
687 * fix: if 文を無くしてみた
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
688 下の方に 「if () が多い」って書きましたが、いろいろ小細工を。
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
689 SchedTask をやってみました。例えば
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
690
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
691 if (cmd != 0) {
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
692 delete taskGroup;
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
693 scheduler->mail_write(cmd);
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
694 }
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
695
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
696 ってのがありました。cmd ってのは taskGroup->status で
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
697 もし cmd が 0 でなければ、taskGroup はすでに空っぽで
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
698 待つべきタスクはすべて終了したので、taskGroup を delete し、
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
699 mailbox で cmd を PPE に送ります(cmd にはすでに送るべきコマンドがある)
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
700 でまあ、これくらいなら
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
701
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
702 delete (int*)((cmd != 0)*(int)(taskGroup));
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
703 scheduler->mail_write(cmd);
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
704
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
705 ぐらいに直せました。
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
706 delete や free では NULL を渡しても何もしない(?)って動作なので
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
707 これでも問題ない。つまり、cmd == 0 なら、taskGroup を
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
708 解放する必要は無いので NULL が delete に渡されるわけです
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
709 int* でキャストしてるのは、そのまま 0 を渡すと、
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
710 「int型を delete するのはできない」的なエラーがでるからです。
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
711 。。。だったら int* じゃなくて TaskGroupPtr じゃね?とか思った今。
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
712
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
713 あと、PPE 側で 「mail == 0 なら NOP」 的な処理を入れました。
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
714 これによって、cmd が 0 かその他で if を書く必要がなくなりました。
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
715 問題があるとすれば、 SPE -> PPE の mailbox の queue の長さ。
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
716 NOP コマンドを送って、queue の制限に引っかかって
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
717 mail_write が止まるんじゃないかなーとか少し心配です。
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
718 ここらへんは optimize の時間に考える事かな。
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
719 どうせ PPE では mail しか読んでないし、
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
720 そこまで queue が埋まる事は無いと思いたい。
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
721
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
722
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
723
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
724 あとはこんな感じかな
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
725
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
726 #if 1 // fix
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
727 free((void*)(flag_renewTask*(int)(list)));
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
728 #else
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
729 if (flag_renewTask) {
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
730 free(list);
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
731 }
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
732 #endif
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
733
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
734 動いてるのは確認したし、gdb で x/20i とかしたら
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
735 branch 命令が減ってるのは確認した。
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
736 まあ -O9 とかで最適化掛けるとどっちも同じになるけどな。
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
737
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
738
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
739 * add (API): static_alloc, static_get, static_free
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
740 SchedTask 自身だけが持つ領域ではなく、
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
741 SPE 上に複数のタスクが共有したい領域を作る
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
742 これは task::run() 内で使用する。
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
743
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
744 - void* static_alloc(int id, int size);
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
745 @param [id] 領域ID。現在は 0〜31 まで使用可能 (Scheduler.h で定義)
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
746 @param [size] 領域のサイズ
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
747 @return allocate した領域のポインタ。下の static_get の返り値と同じ
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
748
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
749 - void* static_get(int id);
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
750 @param [id] static_alloc で作った領域 ID。
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
751 @return 領域のポインタ
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
752
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
753 - void static_free(int id);
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
754 @param [id] 解放したい領域の ID
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
755
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
756 こんな感じかなー。
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
757 static_free はさすがにユーザに任せるだろう。
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
758 static_free し忘れると SPE には致命的なので、ここはよく伝える必要有
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
759
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
760 例題は
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
761 cvs: firefly:Game_project/student/master/gongo/Static
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
762
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
763 まあ Renew と大体同じですけどね。
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
764 int 型配列 data を共有にして、各タスクでインクリメントしてる
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
765
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
766 * TODO: TaskGroup の扱い
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
767 通常の Task では、task->self には
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
768 自分が終了した時に PPE に送るコマンド(自分自身)になりますが、
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
769 タスク中に生成されたタスク(もう何度も書くのめんどいんで Renew で)では
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
770 task->self は、task を待っている TaskGroup を表します。
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
771
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
772 self という名前で意味が違うのでこういうことはやめたいんだが。。。
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
773 といいながらやめないのが(ry
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
774
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
775 * memo:
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
776 下の 直したいところ (1) ってやつがよくわからんので、
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
777 現在の状況だけ
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
778
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
779 scheduler->add_groupTask() をするたびに
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
780
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
781 group.insert_front(cur_id++);
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
782
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
783 されます。
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
784 そして、scheduler->remove_groupTask() されると
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
785
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
786 group.remove(--cur_id);
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
787
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
788 されます。要するに、どのタスクでも
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
789 cur_id だけが insert/remove されます。
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
790 「どのタスクがあるか」ではなく「どれだけのタスクがあるか」ですね。
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
791 実際にはしっかりと TaskPtr で管理したかったんですが、
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
792 下にも書いたアドレスが被る云々の問題でそれもできず。
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
793 やり方はあると思うんですが。
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
794
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
795 うーん、うまく説明できないな。
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
796
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
797 * tag: v20080707
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
798 タスク内タスク生成を作りました。
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
799
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
800 [TODO]
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
801 SPE 上で領域を共有する API の
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
802
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
803 - static_alloc
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
804 - static_get
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
805 - static_free
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
806
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
807 を速攻で実装しよう。。
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
808
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
809 * add: タスク内タスク生成
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
810 一応できたんですが、直したい。。。
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
811 仕様としては
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
812
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
813 - 現在のタスク(T1) の中でタスクを生成したとする (Tc = T2, T3, ...)
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
814 - 本来、T1 が終了次第、T1 が終わった事を PPE に伝えるが、
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
815 ここでは、Tc が全て終わってから、T1 の終了を PPE に伝える
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
816 - Tc 内で再びタスクが生成されても(Tcc)、Tcc が終わってから T1 を(ry
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
817
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
818 現在は、生成したタスクすべてに対して wait_for をかけてる感じ。
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
819 しかし、例えば Frame Buffer に書き込む時は待つ必要ない(はず)なので
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
820 タスク毎に wait_for を選べるようにした方がいいだろう。
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
821
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
822 __ 例題
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
823 cvs firefly:Game_project/student/master/gongo/Renew
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
824
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
825 にあります。
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
826 もうちょいちゃんとした例題が欲しいところです。
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
827
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
828
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
829 __ 直したいところ (1)
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
830
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
831 現在、Tc を管理する構造体として、TaskGroup を使ってます
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
832
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
833 class TaskGroup {
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
834 unsigned int command; // T1 が PPE に送るコマンド
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
835 __gnu_cxx::slist<int> group; // Tc がある Linked List
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
836
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
837 // function は省略
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
838 };
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
839
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
840 slist じゃなくて、TaskQueue みたいに自分で作っても良かったんだけど。
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
841 group.empty() == true になったら、command を PPE に送るって感じです。
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
842
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
843 で、slist が持つデータが TaskPtr じゃなくて int の理由。
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
844 まあいろいろあるんだけど(何)、アドレスが重複してしまうことです。
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
845 最初は、create_task で得られた TaskPtr をキーとして使うつもりだったけど
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
846 その TaskPtr は TaskList から取った物で (&list->takss[index] みたいな)
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
847 なんでそれじゃだめなのか。buff_taskList[2] (Scheduler.cpp 参照) を
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
848 使うと、交互に使用するのでアドレスは被る。
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
849 新たに allocate すれば問題は無いが (t1とする)、SPE の LS の問題で
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
850 使わなくなった TaskList は free していかないといけない。
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
851 で、free -> 再び allocate したとき (t2とする)、t1 と t2 の
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
852 アドレスが被ることがあった。当然 TaskPtr も被ると。
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
853 だから、アドレスではなく、TaskGorup が持つ
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
854 unsigned int cur_id を使う事にしました。
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
855
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
856 なんかここまで自分で書いてて、
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
857 なんで出来ないのかまだわからんくなってきた。
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
858
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
859 ので試しに戻してみたら * で * き * ま * し * た *
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
860 わけわからん。まあ勘違いだったのか、いろいろ別のところを直してるうちに
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
861 知らず知らずミスってたところも治ってたのか。まあいいか。
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
862
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
863 と思っていろいろ試したらまた動かなくなった。。もうだめぽ
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
864 とりあえず、また unsigned int に戻しました。
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
865 今のところ、0 <= cur_id <= 0xffff (65535) の範囲のキーを使うように。
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
866
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
867
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
868 __ 直したいところ (2)
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
869 if 文が多い。
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
870 今は、「通常の Task」「タスク内で生成されたタスク」で挙動が違います。
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
871 例えば
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
872
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
873 - SPE で allocate されたデータを使うので、通常 DMA を使うところは
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
874 アドレス参照や memcpy を使う
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
875 - TaskGroup を、上記の Tc や Tcc へ引き継がせるところ
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
876
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
877 なので、flag_renewTask とかいう変数で、ほぼ if 文 で書いてます。
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
878 SPE でこの書き方はかなりまずい気がします。良い書き方はないものか。。。
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
879 「通常の(ry」「タスク内(ry」で新たにインスタンスを作るってのも
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
880 考えはしましたが (SchedTask = 通常、SchedRenewTask = タスク内(ry とか)
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
881 これだと ユーザー側も この二つを選んでやることになります。
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
882 「このタスクは SchedRenewTask 、このタスクは通常」とかやるのは
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
883 かなりめんどくさいと思う。だからライブラリ側で分けるべきか。。。
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
884 多重継承とかってこんなとき役に立つん?
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
885
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
886
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
887 2008-07-03 Wataru MIYAGUNI <gongo@cr.ie.u-ryukyu.ac.jp>
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
888
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
889 * TODO:
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
890 - add_param で渡せるパラメータの数を増やす。15もあればいいんじゃね?
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
891 - 今の実装では、
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
892
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
893 1. PPE でタスク(T1)が生成される
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
894 2. SPE で T1 が実行される
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
895 3. T1 が終わった事を PPE に mailbox で送る
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
896 送る情報は T1 自身。(PPE で生成された時のアドレス)
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
897
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
898 なわけです。しかし、もし T1 から新たにタスクが生成された時はどうするか
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
899 仮に T1 から T2, T3, T4 が作られたとする。
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
900 このとき、
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
901
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
902 1. T1 が終わった時点で、T1 から終了コマンドを送る
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
903 2. T1 だけでなく、T1 内で作られた T2, T3, T4 が終わってから
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
904 終了コマンドを送る
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
905
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
906 の二つが考えられる。
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
907 PPE 側では T1 しか認識していないため、この判定は SPE 内でやることになる
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
908 必要な処理かと言われると微妙だが、欲しくなるのは間違いない。
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
909 つーことで今これを実装中です。
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
910
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
911
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
912 * tag: v20080703
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
913 - タスクに 32 bits パラメータを渡す add_param を実装(現在は3個まで)
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
914 - SPE 内部でタスク生成ができるようになった
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
915
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
916 * add (API): SPE内部での create_task
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
917 今まで、SPE ではタスクを生成する事は出来ず、
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
918 PPE から送られてくるタスクを実行するだけでした。
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
919 それだと不便だってことで SPE 内部でもできるようにしました。
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
920 方法はPPEでやるのと同じく
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
921
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
922 task = create_task(TASK_EXEC);
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
923 task->add_inData(buff, sizeof(Buff));
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
924 task->add_param(data);
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
925
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
926 みたいな感じでいいです。
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
927 spawn() や wait_for() は実装していません。
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
928 SPE 内部で生成するタスク同士で依存関係作るのが
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
929 結構めんどくさいからです。spawn() も、しなくても勝手に実行します。
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
930 PPE とそろえる意味で作ってもいいんだけどね。
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
931 そのためには SPE にも TaskManager が必要になってくるなー。
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
932
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
933
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
934 2008-06-24 Wataru MIYAGUNI <gongo@cr.ie.u-ryukyu.ac.jp>
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
935
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
936 * add (API): add_param, get_param
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
937 DMA で送れないけど、必要になってくる 4 バイトの情報があるとして
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
938 それは今までは
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
939
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
940 add_inData(param, 0);
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
941
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
942 とかして、「サイズ == 0 なら 32 bit のデータ」としていたけど
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
943 それは余りにも変なので(関数の意味的にもおかしい)ので、
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
944
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
945 add_param(parameter);
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
946
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
947 ってのを追加しました。タスク側では
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
948
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
949 get_param(index);
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
950
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
951 とかします。index は、add_param を呼び出した順番で決まります
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
952
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
953 add_param(x);
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
954 add_param(y);
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
955 add_param(z);
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
956
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
957 とあるとき、タスク側では
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
958
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
959 int x = get_param(0);
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
960 int z = get_param(2);
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
961
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
962 とします。
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
963 今のところ parameter は 3つしか送れないことになってますが
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
964 後ほど、上限をあげます。15くらいあれば余裕だと思うんだがどうだい?
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
965 今は、SPE でのタスクの生成のルーチンを書くために、最低限な部分だけ
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
966 ってことで 3 つにしてます。それが出来次第、これもやります。
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
967
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
968
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
969 2008-06-12 Wataru MIYAGUNI <gongo@cr.ie.u-ryukyu.ac.jp>
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
970
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
971 * Cell/CellTaskManagerImpl.cpp (CellTaskManagerImpl::set_runTaskList):
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
972 アホなミス(ry
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
973
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
974 「list が持つ TASK_MAX_SIZE を超えると、次の list へ next を」っていう
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
975 前回直したところがまたミスっててだな。
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
976 簡単に言うと
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
977
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
978 TaskPtr task = &list[list->length++];
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
979 [task の初期化]
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
980
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
981 if (list->length > TASK_MAX_SIZE) {
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
982 [newList 生成]
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
983 newList = append(newList, topList[speid]);
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
984 topList[speid] = newList;
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
985 }
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
986
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
987 ってやってたわけ。これだと、toplist[speid] に
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
988 length = 0 の list が来る可能性があると。
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
989 で、spe に TaskList を送る条件は
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
990
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
991 1. taskList[speid]->length >= 1
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
992 2. speid が次の TaskList を待っている状態
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
993
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
994 で、1 の条件に触れてしまい、TaskList が送られなくなって
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
995 プログラムが終了しないと。アホですね〜
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
996 上の if 文を &list[list->length++]; の前に持って行くだけでおk。
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
997
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
998 2008-06-10 Wataru MIYAGUNI <gongo@cr.ie.u-ryukyu.ac.jp>
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
999
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1000 * Cell/CellTaskManagerImpl.cpp (CellTaskManagerImpl::set_runTaskList):
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1001 アホなミスしてました。
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1002 list が持つ TASK_MAX_SIZE を超えると、次の list へ
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1003 next を繋げるはずなんだけど、speTaskList_bg[speid] とか読む時に
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1004 ちゃんと繋げられてなかったというかなんというか。
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1005 簡単に言うと、タスク多くなると落ち(ry
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1006
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1007 * add (API): set_post
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1008
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1009 create_task(id, 0);
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1010
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1011 とかわざわざ 0 付けるのもアレなので、もうそれように
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1012
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1013 task->set_post(func)
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1014
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1015 を追加しました。func は void (*func)(void) です。
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1016 せっかくだから、引数に void* とか付けてもいいんじゃないかと。
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1017
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1018
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1019 * fix (API): ListDMA API
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1020 タスク側で、ListDMA で指定したデータの取り方
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1021
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1022 run(rbuf, wbuf) として
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1023
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1024 // index は add_inData や add_outData で指定した(順番-1)
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1025 get_input(rbuf, index);
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1026 get_input(wbuf, index);
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1027
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1028 返り値は void* なので、malloc っぽくキャストしてください。
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1029 あと、4バイト以下のデータを送りたい場合、main で
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1030
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1031 add_inData(data, 0)
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1032
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1033 と、アドレスは送りたいデータを則値で、サイズは 0 で指定するとおk。
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1034 get_input で int なりなんなりでキャストすればいいじゃない!
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1035 例題は
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1036
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1037 Game_project/student/master/gongo/arr_cal
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1038
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1039 で複数データ扱ってたり4バイト送ってたりしてます。
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1040
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1041
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1042 * tag: v20080610
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1043 前回との違いは
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1044
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1045 - ListDMA の導入
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1046 - 凡ミスfix
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1047
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1048 とかかな。何気にここには ListDMA の API 書いてなかったな。
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1049
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1050 - task->add_inData(addr, size); // input
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1051 - task->add_outData(addr, size); // output
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1052
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1053 これで Input/Output のデータ領域を指定可能。複数できます。
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1054 詳しくはいずれドキュメントに書く予定だが、
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1055
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1056 - addr は 16 バイトアライメントに取れてないと行けない
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1057 - size は 16 バイト倍数
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1058
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1059 ってのが最低条件。
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1060 16 バイト未満のデータを送りたいとき(整数を2,3個とか)は考え中。
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1061 addr に直接渡すって手法はできるとわかってるので、それでもいいかな。
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1062 まあいろいろ問題はありますが、少しはできたんじゃないかな。
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1063
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1064 次からは SPE 内でのタスク生成(再起動?)を書く予定
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1065
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1066 * Cell/CellTaskManagerImpl.cpp (CellTaskManagerImpl::set_runTaskList):
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1067 if (speid > machineNum) {
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1068 speid %= MAX_USE_SPE_NUM;
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1069 }
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1070
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1071 から
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1072
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1073 if (speid >= machineNum) {
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1074 speid %= machineNum;
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1075 }
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1076
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1077 に。なんという凡ミス
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1078
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1079 * Cell/spe/CellDmaManager.cpp (CellDmaManager::dma_loadList): fix
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1080 ListData が持つ ListElement は
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1081
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1082 class ListElement {
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1083 public:
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1084 int size;
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1085 unsigned int addr;
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1086 };
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1087
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1088 というデータ構造なわけだが、これは、spu_mfcio.h が持っていて
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1089 且つ List DMA で使用される
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1090
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1091 typedef struct mfc_list_element {
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1092 uint64_t notify : 1; /** Stall-and-notify bit */
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1093 uint64_t reserved : 16;
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1094 uint64_t size : 15; /** Transfer size */
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1095 uint64_t eal : 32; /** Lower word of effective address */
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1096 } mfc_list_element_t;
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1097
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1098 と同じである。notify と reserved は 0 となる (ストールは今は
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1099 考えていない)ので、結局は uint が 2 つの 8 バイト のデータ構造であれば
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1100 そのまま mfc_getl とか mfc_putl に遅れるわけである。
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1101 今までは mfc_list_element_t 構造体に for 文でいちいち代入してたが
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1102 まあそれはなくなったっつーことで。dma_storeList もね。
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1103
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1104
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1105 2008-05-30 Wataru MIYAGUNI <gongo@cr.ie.u-ryukyu.ac.jp>
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1106
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1107 * change (API): TaskManager Memory Allocate
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1108 manager->cerium_malloc(&buff, DEFAULT_ALIGNMENT, sizeof(Data))
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1109
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1110 から
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1111
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1112 buff = (Data*)manager->malloc(sizeof(Data));
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1113
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1114 に変更しました。
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1115 alignment の指定は全て TaskManager に埋め込んであります。
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1116 記述は TaskManager.h に書いてあります。
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1117
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1118 void* TaskManager::malloc(int size) {
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1119 return m_impl->allocate(DEFAULT_ALIGNMENT, size);
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1120 }
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1121
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1122 2008-05-29 Wataru MIYAGUNI <gongo@cr.ie.u-ryukyu.ac.jp>
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1123
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1124 * thinking: List DMA (4)
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1125 Cell 版でも動いたのを確認。今、Cell 版で List DMA が動く条件は
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1126
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1127 1. List の各要素の転送サイズが 16 バイト倍数でなければならない
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1128 2. List の各要素の転送するデータのアドレスのアライメントを保証(16or128
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1129
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1130 2に関しては Cell の仕様なんでまあいいんだけど、
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1131 1は、ドキュメント見る分には
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1132
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1133 - Cell Broadband Engine アーキテクチャ version 1.01 より
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1134 - 7.5.3 get list
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1135 > リスト・サイズ・パラメータは、このDMAコマンドの場合は
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1136 > 8バイトの倍数でなければならず、また、リスト・アドレス・パラメータは、
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1137 > ローカルストレージの8バイト境界にアラインされなければなりません。
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1138
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1139 って書いてるんだよな。int が 10 個の配列(40バイト) を送っても
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1140 見事に弾かれたんだよな。おのれバスエラーめ!
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1141 とりあえず、上の条件を満たせば行けました。
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1142 送るデータのアロケートは
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1143
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1144 TaskManager::cerium_allocate(void **buff, int align, int size);
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1145
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1146 ってのを作りました。使い方は別項目で。だいたい posix_memalign 準拠。
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1147
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1148 動くのはいいんだけど、これだとユーザに全部任せる事になります。
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1149 特に、配列をアロケートした後、その途中の部分をリストに入れたい時。
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1150 その配列の要素のサイズが16倍数じゃないとそこでエラーがでると。
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1151 それをユーザに全部任せるのは、まあいけないこともないけどさ。。。
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1152
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1153
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1154 * Cell/CellTaskManagerImpl.cpp (CellTaskManagerImpl::mail_check): fix
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1155 CellTaskManager は FifoTaskManager のオブジェクトを
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1156 ppeManager という変数で持っていて、作業を別々に行っているわけで。
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1157 だけど両方のオブジェクトがもつ waitTaskQueue は同じじゃないと
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1158 ならないので、最初は TaskQueuePtr * とかで渡して
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1159 共有してたわけだけど、よくよく考えると、
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1160
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1161 - waitTaskQueue に task が append される時
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1162 CellTaskManager->append_waitTask()
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1163
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1164 - waitTaskQueue から task が remove されるとき(依存満たした時とか)
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1165 FifoTaskManagerImpl->mail_check() 及び
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1166 CellTaskManagerImpl->mail_check() です。
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1167
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1168 つまり、waitTaskQueue が共有されるのは mail_check だけなので、
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1169 CellTaskManagerImpl の mail_check で
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1170
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1171 ppeManager->mail_check(mail_list, &waitTaskQueue)
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1172
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1173 として、ここで waitTaskQueue を参照渡ししてます。
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1174 ppeManager->mail_check で waitTaskQueue の整理が終わって
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1175 戻ってくる事には waitTaskQueue が更新されていると。
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1176
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1177 なんか文章がおかしいですね。気になる人は俺に直でお願いします。
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1178 要するに、ppe と spe のそれぞれの TaskManagerImpl で
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1179 waitTaskQueue の共有が上手くいったというわけです。
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1180
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1181
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1182 2008-05-24 Wataru MIYAGUNI <gongo@cr.ie.u-ryukyu.ac.jp>
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1183
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1184 * thinking: List DMA (3)
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1185 現在実装中。Fifo 版では動いている模様。
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1186 問題は Cell だよなー。考えないと行けない事がいくつか
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1187
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1188 - Input/Output データはアライメントされている?
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1189 アライメントされていなくても、こっちでアドレスずらして
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1190 DMAしてずらして run() に渡して〜とかもできるんだけど
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1191 かなりめんどくさい。それに、In ならともかく、
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1192 Out は変な領域に書き込みそうなので無理そう。
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1193 これはもうユーザが、送るデータはすべて
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1194 Cerium_malloc 的なものを通したものだけ、っていう
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1195 制約にした方がいいかもしれない。てかそうなんだっけ。
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1196
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1197 - 配列中のデータの指定
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1198 上の項目と少し関連があるんだが、例えば
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1199
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1200 int data[100]; // アライメントは取れてるとする
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1201
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1202 ってのがあって、そのなかの data[0]〜data[49]、
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1203 data[50] 〜 最後まで送りたいとする。
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1204 最初のやつは &data[0] のアドレスは 16 bytes アライメントだけど、
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1205 &data[50] では、sizeof(int)*50 = おそらく 200 ずれて
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1206 16 bytes アライメントではなくなると。これだと DMA できない。
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1207 ユーザがそこまで考えて、例えば data[32] から送る、とかでもいいけど。
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1208 ライブラリ側で、少しは融通効くようにすべきかな。
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1209 やるなら、アドレスずらして取って来て、ユーザが見るデータは
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1210 そのずらした分戻してから見せるって感じ。変な説明だが。
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1211
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1212 うーん。今はとりあえず全てアライメント大丈夫な方向でやってみるか。
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1213
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1214
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1215 2008-05-23 Wataru MIYAGUNI <gongo@cr.ie.u-ryukyu.ac.jp>
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1216
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1217 * Cell/SpeThreads.cpp (SpeThreads::init): スレッドの生成
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1218 今まで作られてたスレッドは
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1219
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1220 - spe_context_run を実行するだけのスレッド (spe_thread_run)
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1221 - 上のスレッドを呼び出し、終了を待つスレッド (frontend_thread_run)
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1222
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1223 2番目に何の意味があるのかということだが、
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1224 SPE 毎にスレッドを立ち上げておいて、
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1225 それぞれのSPEからのメールは、その担当するスレッドが見る、
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1226 って構想で作っていました。だけど、今は mailbox の扱いは
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1227 Cell/CellTaskManagerImpl::mail_check で行っているため
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1228 わざわざ2番目のスレッドを作る必要がなくなりました。
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1229 つーことで、frontend_thread_run ではなく、
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1230 最初から spe_thread_run を起動すればおkとなりました。
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1231
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1232 * Cell/SpeThreads.cpp (SpeThreads::get_mail): if 文排除
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1233 今までは
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1234
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1235 if (spe_out_mbox_read(spe_ctx[speid], &ret, 1) < 0) {
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1236 return ret;
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1237 else
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1238 return -1;
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1239
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1240 とやっていた。これは
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1241
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1242 - データを読めたらそれ(ret)を返す
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1243 - データが無かったら -1 を返す
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1244
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1245 ってことだったんだが、よくよく考えると、spe_out_mbox_read() は
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1246 データがなかった場合 ret の値を変えないので、最初に
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1247
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1248 unsigned int ret = (unsigned int)-1;
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1249
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1250 としておけば、最終的に if 文無しの
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1251
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1252 spe_out_mbox_read(spe_ctx[speid], &ret, 1);
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1253 return ret;
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1254
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1255 だけで済むわけだ。
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1256
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1257 2008-05-22 Wataru MIYAGUNI <gongo@cr.ie.u-ryukyu.ac.jp>
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1258
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1259 * thinking: List DMA (2)
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1260 MFC List DMA read の場合は(少なくともPPEでcreate_taskする時は)
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1261 read size が決まっているので無問題。
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1262 しかし、MFC List DMA write の場合。同じタスクでも
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1263 違うサイズを出力するということはありえるので問題。
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1264 今までも、write の場合は task->run() の返す値が write size として
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1265 使う事にしていた。List DMA write の場合は、おそらく
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1266
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1267 - task->run() 内で write 用の List DMA 構造体を作って Scheduler に
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1268 渡して、task->write() でやってもらう
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1269
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1270 って感じ? でも(上の手法に限らず)、write のサイズが決まってないと
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1271 write 用バッファを生成しておく事ができないので
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1272 書き込めない or あらかじめ多めに取っておくってことが必要になる。
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1273 後者は SPE には痛手(昔は強制16KB確保とかやってたな)なので微妙。
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1274 前者は論外だろう。
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1275
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1276 うーん、どうすっかな。Single DMA write の頃からこれは問題であって。
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1277 最悪、ユーザが「write のサイズが変動しないようなタスクにする」とか?
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1278
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1279 * thinking: List DMA
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1280
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1281 構想としては以下のような考え。
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1282
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1283 class Task {
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1284 int cmd;
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1285 DataListDMA *rlist;
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1286 DataListDMA *wlist;
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1287 };
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1288
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1289 class DataListDMA {
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1290 int length; // リストの数
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1291 unsigned int addr[128]; // アドレス
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1292 int size[128]; // そのアドレスから取得するデータのサイズ
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1293 };
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1294
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1295 128 という数字は、一つのタスクが持てるリストの合計サイズを
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1296 1KB (= 1024B) にしようってことで 4*128+4*128 = 1024 としました。
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1297 ListDMA を使う流れとしては
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1298
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1299 1. Scheduler から cmd にそった Task を生成する
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1300 2. Task のコンストラクタ(もしくは Task を生成する implement 内 )で
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1301 task->rlist, task->wlist を DMA read しておく (ここは通常のDMA)
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1302 3. task->read() で MFC List DMA で List を読む
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1303
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1304 DataListDMA->length に関しては、Task の中に入れるのも有りかと思う。
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1305 その場合は、2 の DMA read で、わざわざ 1KB 全部読む必要は無くなる。
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1306
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1307
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1308 * tag: v20080522
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1309 - PPE 側のタスクも SPE と同じく、クラスオブジェクトとして登録
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1310 - PPE、SPE 側の TaskManagerImpl を整理。見やすくなったと思われ
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1311
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1312 こんなところかなー。
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1313 テストプログラムは
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1314
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1315 Game_project/student/master/gongo/hello
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1316
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1317 にあります。DMA の例題まだだったぜHAHAHA
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1318 ここからは List DMA の処理を入れて行きたいと思います。
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1319
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1320 現在の simple_render のバージョンは
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1321 PPE のタスクが関数ベースだった頃のなのでそのままでは動きません。
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1322 List DMA ができるか、気晴らしに描き直すと思います。
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1323
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1324 * Task 定義について
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1325 PPE も C++ のクラスオブジェクトとしてタスクを定義するようにしました。
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1326 ちゃんとした API を考えたら改めて書きます。
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1327
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1328 * メールチェックから次のタスクリスト生成までの流れの変更
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1329 今までの FifoTaskManagerImpl の mail_check では
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1330
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1331 1. mail_check
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1332 1.1 check_task_finish
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1333 1.1.1 notify_wait_taskQueue
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1334 1.1.1.1 append_activeTask (依存満たしたタスクを)
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1335 1.2 get_runTaskList
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1336
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1337 と、全て mail_check の中で終わってたんですが、これを
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1338
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1339 1. mail_check
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1340 1.1 check_task_finish
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1341 1.1.1 notify_task_finish
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1342 2. wakeup_waitTask (つまり append_activeTask)
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1343 3. get_runTaskList
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1344
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1345 というように分割しました。
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1346 おかげで CellTaskManagerImpl の mail_check もすっきり。
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1347
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1348 2008-05-13 Wataru MIYAGUNI <gongo@cr.ie.u-ryukyu.ac.jp>
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1349
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1350 * Cell/CellTaskManagerImpl.cpp (CellTaskManagerImpl::set_task):
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1351 // set_task って名前やめね?
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1352
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1353 どの SPE に振るかって判定を少し変更。
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1354 cur_anySpeid の宣言場所のコメントにもあるけど、
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1355 これはインクリメントじゃなくて乱数の方が
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1356 より SPE_ANY っぽいのか。むしろ「仕事してない方に割り振る」ってのが
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1357 SPE_ANY の役目な気がするな。ウーム。。。
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1358
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1359 2008-05-05 Wataru MIYAGUNI <gongo@cr.ie.u-ryukyu.ac.jp>
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1360
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1361 * Cell/CellTaskManagerImpl.cpp (CellTaskManagerImpl::mail_check):
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1362 PPE には実行するタスクが一つも無い時の動作がおかしかった。
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1363 要するに、
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1364
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1365 PPE で実行するタスクは全て SPE で実行中のタスク待ち
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1366
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1367 って時。。。よけいわからなくなったな。
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1368 まあなんだ、今まで 必ずタスクが PPE and SPE にあったんだけど
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1369 PPE or SPE ってか、どっちか片方でしか動いてない状況だと
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1370 終了判定というか、それがおかしかったっぽい。
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1371
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1372 Hello World でのタスクは
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1373
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1374 1. "Hello World!!" と表示するタスク (2.) を発行するタスク
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1375 2. 表示するタスク
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1376 3. 2 が全て終わったら実行される、最後のタスク(番兵的な
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1377
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1378 この時、(2) が SPE だけで実行されてると、
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1379 (2) の終了を待つ (3) の判定?というか、それがおかしい
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1380
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1381
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1382 もう眠くてわけわからん。
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1383 一応動いたんだけど、やはり描き直します。
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1384 気持ち悪いほどやっつけな書き方してるので。これはきもい。。。
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1385
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1386 2008-03-09 Wataru MIYAGUNI <gongo@cr.ie.u-ryukyu.ac.jp>
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1387
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1388 * memo: pthread_cond_wait
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1389 この ChangeLog に書くものでもないが、まあメモ。
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1390
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1391 セマフォの P 動作は、基本的に
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1392
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1393 ---------------------
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1394 pthread_mutex_lock(&sem->mutex);
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1395
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1396 while(sem->value == 0) { // 資源が無い
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1397 // 条件付き変数に自分を登録して、ロックを解放して、他の
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1398 // プロセスが資源を解放してくれるのを待つ
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1399 pthread_cond_wait(&sem->cond,&sem->mutex);
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1400 }
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1401 // 自分の分の資源があったので、それを確保する */
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1402 sem->value--;
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1403 // ロックを解放する
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1404 pthread_mutex_unlock(&sem->mutex);
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1405 ----------------------
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1406
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1407 こんな感じ。でコメントにもあるように、
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1408 pthread_cond_wait では、wait の前に unlock する。
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1409 これがよくわかってなくて、「while の外で lock してるのに
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1410 「なんで他のプロセスが lock できるんだろう。」と。
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1411 man 見ろよと思った。てか先生のページのコメントに書いてるよ!
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1412
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1413
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1414 2008-03-08 Wataru MIYAGUNI <gongo@cr.ie.u-ryukyu.ac.jp>
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1415
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1416 * memo: mailbox read の blocking/non-blocking
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1417 spe_out_mbox_read は non-blocking API なので、
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1418 これをぐるぐる回すと busy-wait になるので、
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1419 今の所 ppe 側の Scheduler がトップに戻る?時にメール確認する。
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1420 で、spe_out_intr_mbox_read は blocking API 。
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1421 spe_out_mbox_read との記述の違いは、予め
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1422 spe_event_handler_register で SPE_EVENT_OUT_INTR_MBOX を
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1423 登録しておく。spe 側では、
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1424
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1425 spu_writech(SPU_WrOutMbox, data)
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1426
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1427 じゃなくて
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1428
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1429 spu_writech(SPU_WrOutIntrMbox, data)
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1430
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1431 を使う必要がある。
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1432 両者の mailbox read の速度を調べてみたけど、そんなに違いは感じない。
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1433 まあベンチマークの取り方がへぼいせいかもしれないけど。
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1434 ってことで、こっちの intr の方がいいんじゃないかと思う。
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1435 これと セマフォを組み合わせて mail の処理は簡単になると思う。
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1436 セマフォの処理が重いって話もあるが、どうなんだろうね。
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1437
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1438 * Test/simple_render/task/create_span.cpp (half_triangle): fix
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1439 画面外の span を描画しようとして落ちるので、それの修正。
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1440 polygon->span の時点で外してるけど、span を外すより
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1441 Polygon の時点で修正するべきかな?
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1442
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1443 * kernel/ppe/TaskManagerImpl.cpp (TaskManagerImpl::set_task): fix
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1444 返す TaskList が、mainTaskList の最後尾になってた。
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1445 ってことで、TaskList のトップである bufferManager->mainTaskList を。
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1446
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1447 * kernel/ppe/BufferManager.cpp (BufferManager::clear_taskList): fix
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1448 mainTaskList->length はクリアしてるのに、
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1449 mainTaskList->next をクリアし忘れてた。
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1450 だから空の TaskList が送られてたのか・・・ちくしょう!
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1451
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1452 2008-03-07 Wataru MIYAGUNI <gongo@cr.ie.u-ryukyu.ac.jp>
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1453
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1454 * bug-fix (Test/simple_render): y座標の移動方向
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1455 (1) で、書き込む時に
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1456
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1457 y = height - y
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1458
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1459 としていた。千秋に聞いてみると、
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1460
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1461 「ポリゴンの y を増やす(+)と、画面上に進むようにした」
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1462
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1463 だそうです。なるほどねー。ってことで(2)でもやったら上に進んだよ。
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1464
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1465 しかし、ゲーム的には上が + の方がわかりやすいかもしれんが、
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1466 プログラミング的には、framebuffer ベースでやるので、
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1467 下にいくと y++ ってほうが作りやすいかなーと思いつつ。どっちがいいかね
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1468
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1469 * bug (Test/simple_render): y座標の移動方向
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1470 Viewer::run_draw で、従来の、SpanPack をそのまま描画する方法(1)と、
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1471 SPE に渡すように、8分割したものとして描画する方法(2)で、
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1472 それぞれの y に +0.5 とかしたら、移動する方向が違う。
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1473 (1)では上、(2)では下へ行く。
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1474 送られてくる span には違いが見られず、
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1475 x 方向や 回転は問題ないので、おそらく draw 時の y の計算のバグだろう。
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1476
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1477 1: polygon.cpp Polygon::draw(SPANPACK);
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1478 2: task/span_pack_draw.cpp run();
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1479
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1480 * Test/simple_render/spe/SpuDraw.cpp: ↓の続きの解決
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1481 render_y &= ~7
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1482
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1483 これでおkでした。先生ありがとうございます。
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1484 今はマクロとして
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1485
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1486 #define YTOP(y) (y & ~7)
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1487
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1488 ってやってますわ。
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1489
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1490 2008-03-05 Wataru MIYAGUNI <gongo@cr.ie.u-ryukyu.ac.jp>
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1491
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1492 * memo: MFC List DMA の element の最大値
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1493 「Cell Broadband Engine Architecture Version 1.02」 より
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1494
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1495 P.55
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1496 The maximum number of elements is 2048.
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1497 Each element describes a transfer of up to 16 KB.
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1498
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1499 ってことらしいです。一度の転送での制限は普通のDMAと変わらず16KB。
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1500 mfc_list_element_t は 2048 個まで設定できるってことか。
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1501 テクスチャのロードで、分割しないなら MFC List DMA を使うことになるが、
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1502 2048 個もあれば充分?
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1503
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1504
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1505 * Test/simple_render/spe/SpuDraw.cpp: ↓の続き
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1506 と思ったけど、やっぱりずれるなあ。うーむ。
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1507 とりあえず今は
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1508
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1509 if (render_y < 0) {
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1510 int tmpy = render_y%8;
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1511 render_y -= tmpy + (tmpy != 0)*8;
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1512 } else {
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1513 render_y -= (render_y%8);
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1514 }
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1515 render_y += 1080/2;
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1516
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1517 で落ち着くことに。うーむ。
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1518 もっと良い計算を考えるよりは span の生成時で
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1519 いろいろやるほうがいいのかなー
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1520
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1521 2008-03-04 Wataru MIYAGUNI <gongo@cr.ie.u-ryukyu.ac.jp>
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1522
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1523 * Test/simple_render/spe/SpuDraw.cpp: ↓の続き
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1524 よくよく考えてだな。。。マイナスが気になるなら
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1525
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1526 if (render_y < 0) {
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1527 int tmpy = render_y%8;
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1528 render_y -= tmpy + (tmpy != 0)*8;
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1529 } else {
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1530 render_y -= (render_y%8);
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1531 }
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1532 render_y += 1080/2;
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1533
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1534 じゃなくて
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1535
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1536 render_y += 1080/2;
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1537 render_y -= (render_y%8);
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1538
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1539 これでよくね?ってか元々そのための 1080/2 だった気が。。。
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1540
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1541 * Test/simple_render/spe/SpuDraw.cpp: render_y の計算の修正
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1542 sp->span[0].y (SpanPack に格納されてる最初の Span の y 座標) から
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1543 この SpanPack が描画する範囲の一番上の y 座標を調べる。
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1544
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1545 どういうことかっていうと、例えば SpanPack に入ってる Span が持つ
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1546 y 座標が 1 ~ 8 の時
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1547
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1548 1 -----
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1549 --
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1550 --------
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1551 ----
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1552 ---------
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1553 8 --
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1554
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1555 '-' は描画していると思ってください。
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1556 この場合は、y = 1 がこの SpanPack の一番上、基準となる 座標ってこと。
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1557 framebuffer に書き込むとき、y = 1 から順々に書いて行きます。
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1558
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1559 で、sp->span[0].y ってのが、その基準となる y である保証が無いので、
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1560 sp->span[i].y 、つまりどの y からでも、基準となる y を求める
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1561 必要がある。その計算をミスってました。
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1562
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1563 1 //////////
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1564 <- なぜか書き込まれていない
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1565 //////////
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1566 //////////
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1567
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1568 みたいに、歯抜けした部分があったので、いろいろ調べてみたら
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1569 この render_y がずれてるってことが判明しました。
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1570 今までは
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1571
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1572 render_y = sp->span[0].y;
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1573 render_y += 1080/2;
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1574 render_y = (render_y/8)*8;
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1575
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1576 ってことしてたんだけど、これだと sp->span[0].y が マイナスのとき
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1577 ずれることが判明。なので
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1578
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1579 if (render_y < 0) {
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1580 int tmpy = render_y%8;
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1581 render_y -= tmpy + (tmpy != 0)*8;
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1582 } else {
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1583 render_y -= (render_y%8);
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1584 }
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1585 render_y += 1080/2;
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1586
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1587 こうするとできました。。。が、直したい。
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1588 もう少し奇麗に描けると思うんだけどなー。if 文ぐらいは外したい
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1589
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1590 2008-03-03 Wataru MIYAGUNI <gongo@cr.ie.u-ryukyu.ac.jp>
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1591
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1592 * memo: 最適化の結果
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1593 ppe/spe ともに最適化なしの場合
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1594 263.444 FPS
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1595
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1596 ppe だけ -O9 で最適化
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1597 317.425 FPS
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1598
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1599 spe だけ -O9 で最適化
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1600 812.539 FPS
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1601
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1602 ppe/spe ともに -O9 で最適化
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1603 1610.58 FPS (吹いた
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1604
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1605
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1606 最初、ダブル最適化の画像を見た時の
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1607 あまりの早さにびびった。
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1608 逆に「こ、これはバグか!?」と思った
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1609
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1610
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1611 2008-02-28 Wataru MIYAGUNI <gongo@cr.ie.u-ryukyu.ac.jp>
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1612
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1613 * kernel/ppe/BufferManager.cpp: remove_taskQueue_all()
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1614 taskQueue の create と free が釣り合って無くて、
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1615 queue が足りなくなる -> extend_pool -> 足りなく(ry
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1616 ってのを繰り返してメモリ的なセグメンテーションフォルとが出て
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1617 なんでかなと思ったら、task->wait_me を消去してなかった。
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1618 task->wait_i は notify(ry で削除されるんだけど、
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1619 task->wait_me は、notify(ry に渡した後ほったらかしだった。
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1620 ってことで、wait_me を全消しする関数を作りましたとさ。
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1621 気持ち速度が増した気がする。気ね。
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1622
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1623
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1624 2008-02-17 Wataru MIYAGUNI <gongo@cr.ie.u-ryukyu.ac.jp>
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1625
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1626 * Todo: 悩んでる所
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1627
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1628
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1629 * fix: kernel/ppe/HTask.cpp
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1630 今まで、manager->create_task で生成したタスクは
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1631
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1632 - dependency の設定
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1633 manager->set_task_depend(master, slave) // slave は master を待つ
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1634
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1635 - 実行キューへの追加
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1636 manager->spawn_task(master);
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1637 manager->spawn_task(slave);
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1638
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1639 と、manager を介してやっていました。
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1640 まあそれでもいいんだけど、特に dependency の所は
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1641 どっちがどっちを待つのかってのは、API見るだけじゃわからない。
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1642 そこで、Task (HTask のこと) に、上二つに対応するような
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1643
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1644 void set_depend(HTaskPtr) と void spawn(void) を追加しました。
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1645
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1646 - Usage
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1647 slave->set_depend(master); // slave は master を待つ
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1648 slave->spawn(); // slave をキューへ追加
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1649
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1650 結局は、それぞれの関数の中では、上の set_task_depend とかを
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1651 呼んでるんだけど、ユーザ側からするとこの方がわかりやすいと思います。
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1652
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1653 2008-02-16 Wataru MIYAGUNI <gongo@cr.ie.u-ryukyu.ac.jp>
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1654
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1655 * tag: beta3
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1656 ダブルバッファリングを追加し、まあなんとか動いてるんじゃない?って
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1657 ところまでですが、所詮 Fifo バージョンなので、
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1658 そろそろ Cell を書き上げて、並列にちゃんと動いてるか確かめないとね
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1659
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1660 * add: kernel/ppe/DmaBuffer.cpp
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1661 ダブルバッファリング用に作ったんだけど、
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1662 せっかくなので、DMA は、このオブジェクト(が持つ二つの領域)でしか
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1663 行えないようにする。ってのでどうでしょう。って話を先生としました。
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1664 並列処理だし、ダブルバッファリングがデフォでいいでしょう。
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1665 というか、したくなければ swap_buffer とかしなければおk。
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1666
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1667 -Usage
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1668 DmaBuffer *buffer = manager->allocate(sizeof(SceneGraphPack));
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1669
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1670 今までと違い、create_task の in_addr と out_addr には
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1671 DmaBuffer をいれてください。ユーザが自分で malloc/new したやつは
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1672 エラーを出すようにしてる(seg faultだけどね!)
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1673 汚いソースだが、実際に使ってる様子は Test/simple_render の
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1674 viewer.cpp で使ってます。sgp_buff と pp_buff ってやつね。
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1675
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1676 もうすこしユーザに優しいAPIを作りたい・・・
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1677
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1678 2008-02-11 Wataru MIYAGUNI <gongo@cr.ie.u-ryukyu.ac.jp>
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1679
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1680 * add: Test/simple_render
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1681 chiaki の DataPack を使った Cube の表示プログラム。
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1682 簡単に DataPack を TaskManager の scheduler (SpeManager) に渡して
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1683 処理してコピーして、を繰り返してるだけなんだけど
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1684 まあ動いてる気がするのでいいんじゃないでしょうか。
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1685
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1686
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1687 2008-02-10 Wataru MIYAGUNI <gongo@cr.ie.u-ryukyu.ac.jp>
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1688
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1689 * tag: beta1
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1690 この状況だと、できることよりもできないことを書くべき?違うか。
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1691
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1692 - task (親) 中で task (子) を生成することはできない
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1693 正確には「生成はできる」だけど、その 子task が
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1694 親task に依存している別の task を無視して動くので
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1695 ちゃんとした結果にならないと。
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1696 8日の Todo にも書いてあるけど、今の実装では
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1697 task が task を生成することを想定してない感じなので。
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1698 完全に spe 用にのみ狙いを絞った実装であって
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1699 OS って言えない実装であるからして、書き直すの?
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1700 全ての関数を task しようとするとこうなる訳で、
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1701 ある部分だけやるってのはまあできるんだろうけど、うーん。
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1702
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1703 - chiaki の simple_render が動かない
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1704 (追記) 解決しました
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1705 単に read/write buffer のサイズが足りないだけだった。アホスwww
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1706 まあ辱めの為の下は残しておこう
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1707
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1708 まだ cvs に commit してないけど、chiaki が書いた、
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1709 DataPack 対応の simple_render に TasKManager を組み込んでみた。
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1710 といっても、OSっぽく書いたんじゃなく、今は
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1711 update_sgp と create_pp だけを task 化してみた。
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1712 でまあ動いてるような気はするけど、ものすっごい malloc 系の warning が。
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1713 時々長く動くよねみたいな感じになってしまっている。
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1714 TaskManager が悪いのか、simple_render が悪いのか。
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1715 この TaskManager、ある部分での malloc 系の問題に敏感だなあ。
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1716 まあそうでなかったらバグの探しようも無いし、良いっちゃー良いんだが。
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1717
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1718
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1719 2008-02-08 Wataru MIYAGUNI <gongo@cr.ie.u-ryukyu.ac.jp>
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1720
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1721 * add: kernel/ppe/SymTable.cpp
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1722 今まで func[] = {add, sum, ...}
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1723 とかやってかっこわるい言われまくってたので
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1724 話し合いの通り Symbol Table みたいなものを作ることに
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1725
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1726 // 疑似コードね
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1727 struct sym_table {
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1728 char *sym; // シンボル
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1729 void *address; // シンボルが示すアドレス
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1730 } sym_table[] = {{"Sum", &Sum} , {"Draw", &draw}};
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1731
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1732 int fd = get_fd("Sum");
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1733 void *addr = get_address(fd);
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1734
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1735 table には "Sum" と "Draw" っていう二つのシンボルが登録されている。
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1736 例えば、ユーザ(カーネル?)が "Sum" ってシンボルにアクセスしたい場合、
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1737 まずは get_fd で "Sum" に対する、file descripter を返す。
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1738 ユーザは、その fd に従って get_address を取得することが出来る。
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1739 TaskManager 的な使い方をするなら
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1740
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1741 // 俺は今、Draw 関数を使うタスクを生成したい
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1742 int fd = manager->open("Draw");
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1743 manager->create_task(fd, size, in, out, func);
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1744 manager->open では get_fd と同じ使い方です。
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1745
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1746 まだ改良の余地ありそうですが、今は動いてるってことで。
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1747
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1748
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1749 - 補足
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1750 なぜ file descripter と表すか
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1751
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1752 OS の昨日として、 fopen とかと同じ使い方もできるじゃない!
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1753
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1754
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1755 * Todo: task が task を生成する際の処理
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1756 今まで、 task が行う作業は、演算のみを行うような
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1757 単純な実装に決め打ちされているわけです。
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1758 しかし、OS なんかだと、タスク中から別のタスクを生成するとか
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1759 ありありだと思われる。てか今のテストプログラムでなった。
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1760
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1761 Test/Sum にあるプログラムで使われるタスク
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1762
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1763 - init2 // 貧相な名前ですまない
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1764 演算する数値とかバッファの初期化
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1765
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1766 - sum1
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1767 ある範囲の整数 (i から i+16 とか) の総和
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1768
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1769 - sum2
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1770 sum1 で求められた、複数の範囲の総和を一つにまとめる
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1771 (ex. 複数の sum1 が 1->16, 17->32, 33->48 の総和を計算し、
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1772 sum2 で 上の3つの総和を計算する
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1773 要は 1->48 の総和を分割するっていうプログラムね
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1774
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1775 - finish
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1776 sum2 で求まった値を表示
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1777
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1778 この Sum というプログラム、というか OS と言おう。SumOS ね。
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1779 SumOS は最初に TaskManager (所謂 kernel) を起動し、
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1780 init を起動する。init では、予め決められたタスクである
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1781 init2 と finish の二つのタスクを create して登録する。
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1782 init2 と finish には依存関係がある (init2 が終わったら finish)
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1783 init2 の中で、sum1 と sum2 というタスクが作られる。
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1784 sum1 と sum2 にも依存関係はある (sum1 が終わったら sum2)
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1785
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1786 今の実装だと、タスクが終了して初めて次のタスクへ行く。
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1787 まあ当たり前なんだけど、例えばそのタスクの中で
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1788 新たにタスクが作られた場合、そのタスクが終了するまでは
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1789 実行されなくなってしまう。
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1790 でまあ、今は、manager->create_task される度に
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1791 manager->run とかして、無理やり起動してる訳さ。
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1792 何が無理矢理かっていうと、scheduler の役目をしている
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1793 SpeManager (これも名前変えないと) を2度呼び出してる訳。
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1794 つまり、タスク中でタスクを作る度に、SpeManager オブジェクトを
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1795 new してるわけさ。いいのか?いや、動いてるけどね?
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1796
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1797 ちなみに、Cell version だと spe が勝手に取っていってくれるから
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1798 大丈夫かなと思いつつ、もし spe を1つしか使わない設定だったら微妙。
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1799
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1800 要するに、タスク中でタスクが作られる場合の処理を考えないとね。
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1801
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1802 2008-02-07 Wataru MIYAGUNI <gongo@cr.ie.u-ryukyu.ac.jp>
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1803
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1804 * memo: プログラミングの姿勢
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1805 scheduler とか、task の管理をする部分は
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1806 kernel programing のつもりで、
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1807 example とか、task に割り当てる処理を決めたりする部分は
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1808 user programing のつもりで。
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1809
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1810 それぞれ違った視点で見る必要がある
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1811
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1812 * memo: OS というもの
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1813 OS 起動の流れ
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1814
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1815 - PC の電源を入れる
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1816 - BIOS が立ち上がる (OpenFirmWare, EFI, BIOS)
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1817 - 起動デバイスをチェック (優先度とか種類とか)
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1818 - 起動デバイスから Boot loader を起動
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1819 + BIOS によって、認識できるファイルシステムが違う(だっけ?)
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1820 + ファイルシステムのどこに Boot Loader があるか知っている
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1821 + grub, grub2, lilo, kboot などがある
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1822 - Boot Loader が kernel を起動
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1823 + ネットワークブートの場合、TCP/IP や
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1824 ネットワークデバイス(イーサとか?)のドライバを持ってる必要がある
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1825 - kernel は、最初に scheduler を起動する
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1826 - scheduler の初期化 (init を呼ぶ?)
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1827 - init では、事前?に設定されているスクリプトとかを呼ぶ
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1828 + linux とかだと /etc/rc にあるやつを init が呼ぶ
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1829 - login form が起動
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1830
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1831 補足 こっからユーザ
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1832 - login する
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1833 - shell を呼ぶ
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1834 + login shell かどうか確かめる
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1835 - ユーザに設定されてる起動スクリプト?を実行
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1836 - 晴れてログイン
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1837
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1838 2008-02-06 Wataru MIYAGUNI <gongo@cr.ie.u-ryukyu.ac.jp>
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1839
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1840 * kernel/spe/*.cpp: new と placement new
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1841 現在、spe kernel のタスクは、切り替わる毎に
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1842 new/delete を繰り返しています。今はこれでいいんだけど、
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1843 速度的にも、いずれは直さないといけないと思うわけで。
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1844 で、予め allocate された領域を利用した placement new を使えば
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1845 new よりもそれなりに早くなるっぽい。
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1846 例題として、与えられた回数分 new/delete を繰り返すプログラムと、
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1847 同じ回数分、placement new したときの速度の比較
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1848
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1849 for (int i = 0; i < num; i++) {
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1850
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1851 < task = new Task;
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1852 < task->init(i);
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1853 < task->printID();
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1854 < delete task;
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1855 ---
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1856 > task = new(buff) Task; // buff = malloc(BUFF_SIZE);
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1857 > task->init(id);
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1858 > task->printID(id);
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1859 }
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1860
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1861 placement new では、delete の必要は無い。
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1862 その中で新たに allocate してるなら必要かもしれないが。
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1863 速度比較は以下。no_new が placement new で、ln_new が new/delete 。
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1864
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1865 % ./a.out 10 // 10 回
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1866 no_new: time: 0.012135(msec)
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1867 ln_new: time: 0.003572(msec)
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1868
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1869 % ./a.out 100
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1870 no_new: time: 0.022453(msec)
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1871 ln_new: time: 0.018989(msec)
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1872
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1873 % ./a.out 1000
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1874 no_new: time: 0.115277(msec)
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1875 ln_new: time: 0.136335(msec)
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1876
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1877 % ./a.out 10000
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1878 no_new: time: 1.056156(msec)
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1879 ln_new: time: 1.322709(msec)
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1880
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1881 % ./a.out 100000
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1882 no_new: time: 10.622221(msec)
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1883 ln_new: time: 13.362414(msec)
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1884
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1885 % ./a.out 1000000
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1886 no_new: time: 109.436496(msec)
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1887 ln_new: time: 136.956872(msec)
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1888
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1889 10、100 回だと負けてるが、まあ無視しよう(ぇ
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1890 回数が多くなるにつれて、ほんの少しだが no_new が勝ってる。
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1891 どうなんだろうね。ちなみに printID を無くすと
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1892
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1893 % ./a.out 1000000
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1894 no_new: time: 0.008512(msec)
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1895 ln_new: time: 27.100296(msec)
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1896
Shinji KONO <kono@ie.u-ryukyu.ac.jp>
parents:
diff changeset
1897 I/O に左右され過ぎ。まあそんなもんだろうけどさ。