[Ffmpeg-devel] Re: [PATCH] SIMD accelected SNOW decoding

[Guillaume POIRIER]

Hi,

On 11/28/05, Guillaume POIRIER <poirierg at gmail.com> wrote:
[..]
> commenting out _just_
> inner_add_yblock_mmx in snow.c gives me a grayish picture but no
> segfault.

Good, it looks like I managed to fix 'vertical_compose97i_asm': the
grayish picture is all gone now.

Please find in attachment the fixed files, which should be tested, if
possible, on IA-32 and AMD64.

Now I'm off to find a solution to fix 'inner_add_yblock_mmx' (which
still segfault when it executes the instruction 'movdqa
(%rdi),%xmm3').
I suspected that the problem could be an unaligned memory access and
changed all movdqa to movdqu with no luck.

Then I had a closer look at the machine regs and it does look to me
that rdi being equal to 0xad73ba9000002aba (= -5948205555188225350)
couldn't be right.
I guess fixing the value given to rdi should do the trick.

Thoughts?

Please do keep in mind that current version is work-in-progress: I'll
do my best to address Michael's comments when the SIMD code is fixed.

Guillaume
--
MPlayer's doc is offline. Find some fresh one here:
http://tuxrip.free.fr//MPlayer-DOCS-HTML/en/
http://tuxrip.free.fr//MPlayer-DOCS-HTML/fr/
-------------- next part --------------
A non-text attachment was scrubbed...
Name: snow.c
Type: text/x-csrc
Size: 156958 bytes
Desc: not available
URL: <http://lists.mplayerhq.hu/pipermail/ffmpeg-devel/attachments/20051130/4854537f/attachment.c>
-------------- next part --------------
A non-text attachment was scrubbed...
Name: snow_mmx_sse2.h
Type: text/x-chdr
Size: 47857 bytes
Desc: not available
URL: <http://lists.mplayerhq.hu/pipermail/ffmpeg-devel/attachments/20051130/4854537f/attachment.h>