On 09/05/11 11:17, rafael ff1 wrote:

2011/5/6 Sven-Hendrik Haase:

...

On 07.05.2011 01:18, rafael ff1 wrote:

...

HI there!

I'm trying to update PCSX2 with some help of its dev team [1], but it is crashing all the time. According to 'gdb' output, it is somehow related to lib32-glibc, but it is omitting some information. I was hoping to be able to activate more verbosity, which AFAIK I can get by compiling it with some debug flags. [2]

Is it correct or there is another better way to do it ?

Thanks,

Rafael.

[1] http://code.google.com/p/pcsx2/issues/detail?id=1019 [2] https://wiki.archlinux.org/index.php/Debug_-_Getting_Traces No, that's exactly the way to do it to make gdb happy. Get it from abs as always and enable debug flags.

-- Sven-Hendrik

I added '-g -O1' to the CFLAGS in the PKGBUILD [1] and it seems to provide zero information [2], just like before compiling with debug flags [3]. Did I do something wrong? Any other ideas to debug this binary?

Not that the glibc PKGBUILD manually strips its files. Commment out all that stuff at the end. Allan

2011/5/9 Rémy Oudompheng :

I don't think recompiling glibc with debugging flags is going to help you in any way. Are you sure that your pcsx executable is compiled with debugging symbols? A debug version of glibc is usually of no use except to debug glibc itself.

In the backtrace linked at  http://pastebin.com/eCWRNd2X, it is *Thread1* that segfaulted.

Rémy.

Well, I know it won't fix the problem, but might help understanding what's wrong. And the pcsx2 executable was not enable due to a typo. Glibc stripped lines had to be commented as well, as mentioned by Allan. I fixed and it provided a lot of information with gdb. I'm still not able to get information on "Thread1", only that it's stack is corrupted. My guess is the whole problem started around gtk3 or gcc 4.6 appear and I'd like to test this compilation with GCC 4.5.. Would it brake my system or can I downgrade it? Thanks