| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
|
> I tried running FFmpeg's regression tests with it. The compilation went > fine, but I get a segfault on a huge file (snow.c) with a big > always_inline function (add_yblock), that has a variable-length array of > 17920 bytes. The crash happens on _alloca(). It happens with -O1 and up. > Could you help me go further into dissecting this bug so I might send a > bug report? It's hard to debug inside GCC's internals (like _alloca()) > for me. Thank you for this report. The proper way to proceed on these kinds of bugs is to minimize the test case, so you end up with the smallest possible run of code that demonstrates the bug. The general process for runtime failures is to 1) try to all get all relevant code into a single .c file that still demonstrates the problem, 2) preprocess it with -E or -save-temps 3) remove all code you can while retaining the bug and not causing other failures. ------------------------------------------------------------------------- This SF.net email is sponsored by the 2008 JavaOne(SM) Conference Don't miss this year's exciting event. There's still time to save $100. Use priority code J8TL2D2. http://ad.doubleclick.net/clk;198757673;13503038;p?http://java.sun.com/javaone _______________________________________________ MinGW-users mailing list MinGW-users You may change your MinGW Account Options or unsubscribe at: https://lists.sourceforge.net/lists/listinfo/mingw-users
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
© 2004-2008 readlist.com