[Dri-devel] [BUG] DRI-CVS doesn't work on Radeon VE

[Nick K. <nic...@ma...>]

Hello!

I've checkout DRI-CVS and got the next error:

glxgears

disabling TCL support
Floating point exception (core dumped)

gdb glxgears core
GNU gdb 5.2.1
[snip]

(gdb) bt
#0  0x4046cb6a in _mesa_sse_transform_points3_general ()
   from /usr/X11R6/lib/modules/dri/radeon_dri.so
#1  0x081d0728 in ?? ()
#2  0x40371bd4 in _mesa_validate_all_lighting_tables (ctx=0x638be)
    at light.c:1116
#3  0x245c8b53 in ?? ()
Cannot access memory at address 0x5614ec83

(gdb) disass $eip $eip+15
Dump of assembler code from 0x4046cb6a to 0x4046cb79:
0x4046cb6a <_mesa_sse_transform_points3_general+138>:   addps  %xmm3,%xmm4
0x4046cb6d <_mesa_sse_transform_points3_general+141>:   movaps %xmm4,0x0(%edi)
0x4046cb71 <_mesa_sse_transform_points3_general+145>:   add    $0x10,%edi
0x4046cb74 <_mesa_sse_transform_points3_general+148>:   add    %eax,%esi
0x4046cb76 <_mesa_sse_transform_points3_general+150>:   cmp    %ecx,%edi
0x4046cb78 <_mesa_sse_transform_points3_general+152>:
    jne    0x4046cb40 <_mesa_sse_transform_points3_general+96>
End of assembler dump.

(gdb) info all-registers
eax            0x10     16
ecx            0x8289420        136877088
edx            0x8072a70        134687344
ebx            0x827c984        136825220
esp            0xbfffed98       0xbfffed98
ebp            0x81d0168        0x81d0168
esi            0x81d7c68        136150120
edi            0x8288720        136873760
eip            0x4046cb6a       0x4046cb6a
eflags         0x10202  66050
cs             0x23     35
ss             0x2b     43
ds             0x2b     43
es             0x2b     43
fs             0x2b     43
gs             0x2b     43
st0            0        (raw 0x00000000000000000000)
st1            9.15698587894439697265625        (raw 0x4002928303a000000000)
st2            -0.39477355778217315673828125    (raw 0xbffdca1fc28000000000)
st3            1.0389819958971111546475185605231673     (raw 0x3fff84fd5caec1652000)
st4            1        (raw 0x3fff8000000000000000)
st5            1        (raw 0x3fff8000000000000000)
st6            0        (raw 0x00000000000000000000)
st7            0.81921756267547607421875        (raw 0x3ffed1b83e0000000000)
fctrl          0x7f     127
fstat          0x0      0
ftag           0x0      0
fiseg          0x0      0
fioff          0x0      0
foseg          0x20     32
fooff          0x0      0
fop            0x0      0
xmm0           {f = {0x0, 0x0, 0x0, 0x0}}       {f = {-nan(0x7fffff),
    -nan(0x7fffff), -nan(0x7fffff), -nan(0x7fffff)}}
xmm1           {f = {0x0, 0x0, 0x0, 0x0}}       {f = {-nan(0x7fffff),
    -nan(0x7fffff), -nan(0x7fffff), -nan(0x7fffff)}}
xmm2           {f = {0x0, 0x0, 0x0, 0x0}}       {f = {-nan(0x7fffff),
    -nan(0x7fffff), -nan(0x7fffff), -nan(0x7fffff)}}
xmm3           {f = {0x0, 0x0, 0x0, 0x0}}       {f = {-nan(0x7fffff),
    -nan(0x7fffff), -nan(0x7fffff), -nan(0x7fffff)}}
xmm4           {f = {0x0, 0x0, 0x0, 0x0}}       {f = {-nan(0x7fffff),
    -nan(0x7fffff), -nan(0x7fffff), -nan(0x7fffff)}}
xmm5           {f = {0x0, 0x0, 0x0, 0x0}}       {f = {-nan(0x7fffff),
    -nan(0x7fffff), -nan(0x7fffff), -nan(0x7fffff)}}
xmm6           {f = {0x0, 0x0, 0x0, 0x0}}       {f = {-nan(0x7fffff),
    -nan(0x7fffff), -nan(0x7fffff), -nan(0x7fffff)}}
xmm7           {f = {0x0, 0x0, 0x0, 0x0}}       {f = {-nan(0x7fffff),
    -nan(0x7fffff), -nan(0x7fffff), -nan(0x7fffff)}}
mxcsr          0x1f80   8064

My CPU is DuronHD with SSE support!

Btw, maybe it would be better to optimize more functions with 3DNow/SSE.
IMHO it could help a lot for Radeon VE users since Radeon VE lacks TCL 
and this job is performed by CPU. I can help to do that if someone
will mark functions which require to be optimized.

Best regards! Nick