[Openchrome-devel] Another openchrome drm segfault

[Thomas Hellström]

Xavier Bachelot wrote:
> Hi Thomas,
>
>
> Thomas Hellstr?m wrote:
>   
>> Thomas Hellstr?m wrote:
>>     
>>>>       
>>>>         
>>> Hi, Xavier!
>>>
>>> OK. I'll take a look at the EXA issue when there is time.
>>>   
>>>       
>> I've pushed what I think is a fix now.
>> There are still some problems with EXA on the newer servers, though,
>> In particular, it seems to misrender some pixmaps when a composite
>> manager is running.
>>
>>     
> Yes, I can see some rendering errors, but at least, I don't need the
> ExaNoComposite option to start X now.
>
>   
>> Still interested to hear about 3D problems, though. Does the mesa demo
>> fbo_firecube run?
>>
>>     
> Yesterday I had 3d software running fine but awfully slow (30 fps in
> glxgears). openchrome_dri.so and drm modules were loaded fine though,
> but still I was probably using software rendering somehow.
> I guess the crashes I had the day before was the consequences of the 3D
> engine left in an unclean state after the EXA crashes w/o a reboot in
> between.
>
> Today, the issue with 3D is a bit different. glgears works (370 fps,
> which is about half what I have with the old unichrome_dri), but all 3D
> software are crashing because the openchrome driver tries to use some
> SSE instructions and my CPU can't do that, although the lack of SSE is
> correctly acknowledged in the xorg log...
>
>   
OK. that gives me a place to look.

More questions to help the performance up:

What system is this? CPU? cache size? Chipset?
What kind of cpu usage are you seeing with the new 3D driver and gears? 
system / user?

What happens if you issue
echo "performance" > /sys/devices/system/cpu/cpu0/cpufreq/scaling_governor
as root, and rerun gears?

What happens if you load the openchrome kernel module using
modprobe openchrome disable_verifier=1
before starting X.

I think the remaining EXA problems are in core exa. Need to track them 
down at some point, though.

Thanks,
/Thomas
> Program received signal SIGILL, Illegal instruction.
> 0x006c6487 in via_sse_memcpy (to=0xb11c1000, from=0x83de218, sz=<value
> optimized out>) at via_memcpy.c:120
> 120             SSE_CPY(SSE_PREFETCH, from, to, dummy, lcnt);
> (gdb) bt
> #0  0x006c6487 in via_sse_memcpy (to=0xb11c1000, from=0x83de218,
> sz=<value optimized out>) at via_memcpy.c:120
> #1  0x00754e7d in memcpy_texture (ctx=0x828f260, dimensions=2,
> dstFormat=0x8bd3a0, dstAddr=0xb11c1000, dstXoffset=0, dstYoffset=0,
> dstZoffset=0, dstRowStride=512, dstImageOffsets=0x83ba518,
>     srcWidth=128, srcHeight=512, srcDepth=1, srcFormat=6408,
> srcType=5121, srcAddr=0x83de218, srcPacking=0x829bde8) at
> main/texstore.c:986
> #2  0x00759646 in _mesa_texstore_rgba8888 (ctx=0x828f260, dims=2,
> baseInternalFormat=6408, dstFormat=0x8bd3a0, dstAddr=0xb11c1000,
> dstXoffset=0, dstYoffset=0, dstZoffset=0, dstRowStride=512,
>     dstImageOffsets=0x83ba518, srcWidth=128, srcHeight=512, srcDepth=1,
> srcFormat=6408, srcType=5121, srcAddr=0x83de218, srcPacking=0x829bde8)
> at main/texstore.c:1391
> #3  0x006c617e in viaTexImage (ctx=0x828f260, dims=2, target=3553,
> level=0, internalFormat=6408, width=128, height=512, border=0,
> format=6408, type=5121, pixels=0x83de218, packing=0x829bde8,
>     texObj=0x83ba1f0, texImage=0x83ba488, imageSize=0, compressed=0) at
> via_tex.c:599
> #4  0x006c62c6 in viaTexImage2D (ctx=0x828f260, target=3553, level=0,
> internalFormat=6408, width=128, height=512, border=0, format=6408,
> type=5121, pixels=0x83de218, packing=0x829bde8,
>     texObj=0x83ba1f0, texImage=0x83ba488) at via_tex.c:635
> #5  0x0074dbba in _mesa_TexImage2D (target=3553, level=0,
> internalFormat=6408, width=128, height=512, border=0, format=6408,
> type=5121, pixels=0x83de218) at main/teximage.c:2576
> #6  0x0814f095 in ?? ()
> #7  0x08152e36 in ?? ()
> #8  0x08149d1f in ?? ()
> #9  0x0814a3f2 in ?? ()
> #10 0x081466dc in ?? ()
> #11 0x0807c5f1 in ?? ()
> #12 0x0806accd in ?? ()
> #13 0x00b9f6e5 in __libc_start_main (main=0x806a9e0, argc=1,
> ubp_av=0xbffff3a4, init=0x81bec00, fini=0x81bebf0, rtld_fini=0xb746d0
> <_dl_fini>, stack_end=0xbffff39c) at libc-start.c:220
> #14 0x0804e961 in std::basic_streambuf<char, std::char_traits<char>
>   
>> ::~basic_streambuf ()
>>     
>
>
> Regards,
> Xavier
>
>
>