[Ffmpeg-devel] [PATCH] have cs_test check for sigsegv at smaller widths and sigill

[Ivo]

On Friday 13 April 2007 13:45, Michael Niedermayer wrote:
> On Fri, Apr 13, 2007 at 01:03:48PM +0200, Ivo wrote:
> > See $subject. It turns out four rgbxtoy_MMX functions cannot handle
> > small widths:
> >
> > ./cs_test -m
> > .........dst damaged at 134 w:3 src:128 dst:128 rgb32to15
> > .dst damaged at 134 w:3 src:128 dst:128 rgb32to16
> > .............received signal 11 at w:7 src:128 dst:128 rgb24tobgr24
> > .....received signal 11 at w:1 src:128 dst:128 rgb32tobgr32
>
> patch rejected
>
> do i need to comment why installing a segfault handler is not the correct
> solution to a segfault?

No, i get it. I didn't think good enough about what sigsegv actually means 
to the state of the program. So, is the following patch ok and have the 
person that runs cs_test run gdb afterwards and/or fix the bug?

--Ivo
-------------- next part --------------
A non-text attachment was scrubbed...
Name: cs_test.patch
Type: text/x-diff
Size: 490 bytes
Desc: not available
URL: <http://lists.mplayerhq.hu/pipermail/ffmpeg-devel/attachments/20070413/92777194/attachment.patch>