>> https://gist.github.com/3059829
> Looking at that program, it seems you build a very large recursive data > structure and then free it in one go, without providing appropriate stack > space for this operation, so the segfault just means "out of memory" > because of the deep recursion. > As such, the problem has nothing to do with Coro, you are just running out > of memory (and Coro detects this because it places guard pages at the end > of stack to catch this problem). > You can verify that by using gdb to get a backtrace on the crash - most > likely you will see hundreds of recursions inside sv_free. > If you free large data structures you need to increase the stack space, > either using ulimit -s and/or using Coro::State::cctx_stacksize. Or change > your program to use less memory on free. https://gist.github.com/3059829#gistcomment-366066 I've just dropped out a lot of unnessesary code and reduced iterations count in test. So now example crashes if it uses 100x(100+0-100) asyncs and uses weaken. But if it doesn't use weakens it never crashes even if it uses 500x(2000+0-100) iterations. ulimit -s and Coro::State::cctx_stacksize don't influence on crashes. -- . ''`. Dmitry E. Oboukhov : :’ : email: un...@debian.org jabber://un...@uvw.ru `. `~’ GPGKey: 1024D / F8E26537 2006-11-21 `- 1B23 D4F8 8EC0 D902 0555 E438 AB8C 00CF F8E2 6537
signature.asc
Description: Digital signature