[fglrx] no Visual Effects on a Radeon Xpress 1150 (Dell inspiron 1501, hardy alpha 3)

[harrydb]

I am running hardy alpha3 on a Dell inspriron 1501 laptop that has a radeon Xpress 1150 graphics card [RS485]. Now that the fglrx driver supports compositing I installed it and tried to enable visual effects, which failed. It turns out that it cannot find the 'GLX_EXT_texture_from_pixmap' in 'glxinfo ' in the 'GLX extensions' section (it is there in server and client glx extensions).

I guess it should be capable of doing visual effects, so I am reporting this as a bug (if someone knows otherwise, please let me know)

Harry

[harrydb (harrydeboer)]

I added a 'display' section to xorg.conf, and now the compiz script hangs at "Checking for texture_from_pixmap: "
with cpu: Xorg 66%, glxinfo 34%
which is strange because when i start glxinfo manually it does finish properly.

Output form lspci -vnn
 Subsystem: Dell Unknown device [1028:01f5]
 Flags: bus master, 66MHz, medium devsel, latency 255, IRQ 19
 Memory at c8000000 (32-bit, prefetchable) [size=128M]
 I/O ports at 9000 [size=256]
 Memory at c0100000 (32-bit, non-prefetchable) [size=64K]
 [virtual] Expansion ROM at c0120000 [disabled] [size=128K]
 Capabilities: [50] Power Management version 2

also attaching xorg.conf

[harrydb (harrydeboer)]

whoops, first line of lspci -vnn dropped (I only included the graphics card piece)

01:05.0 VGA compatible controller [0300]: ATI Technologies Inc RS485 [Radeon Xpress 1100 IGP] [1002:5975] (prog-if 00 [VGA controller])

attaching Xorg.0.log

[harrydb (harrydeboer)]

I figured out that the first (direct rendering) test reveals a count of 2 instead of 3 and thus fails, but the indirect test fails for another reason. When I run

LIBGL_ALWAYS_INDIRECT=1 glxinfo

the output is as follows.

name of display: :0.0
X Error of failed request: GLXBadContext
  Major opcode of failed request: 143 (GLX)
  Minor opcode of failed request: 5 (X_GLXMakeCurrent)
  Serial number of failed request: 16
  Current serial number in output stream: 16

[harrydb (harrydeboer)]

I also found this in the dmesg output:
[fglrx:firegl_lock] *ERROR* Process 6088 is using illegal context 0x00000003

[harrydb (harrydeboer)]

and in /var/log/messages
Jan 14 16:40:13 harry-laptop kernel: [ 32.050127] fglrx: module license 'Proprietary. (C) 2002 - ATI Technologies, Starnberg, GERMANY' taints kernel.
Jan 14 16:40:13 harry-laptop kernel: [ 32.101537] [fglrx] Maximum main memory to use for locked dma buffers: 802 MBytes.
Jan 14 16:40:13 harry-laptop kernel: [ 32.101577] [fglrx] ASYNCIO init succeed!
Jan 14 16:40:13 harry-laptop kernel: [ 32.101739] [fglrx] PAT is enabled successfully!
Jan 14 16:40:13 harry-laptop kernel: [ 32.101772] [fglrx] module loaded - fglrx 8.43.2 [Nov 9 2007] on minor 0
Jan 14 16:40:23 harry-laptop kernel: [ 46.125219] [fglrx] GART Table is not in FRAME_BUFFER range
Jan 14 16:40:23 harry-laptop kernel: [ 46.125232] [fglrx] Reserve Block - 0 offset = 0X0 length = 0X40000
Jan 14 16:40:23 harry-laptop kernel: [ 46.125236] [fglrx] Reserve Block - 1 offset = 0X7ff5000 length = 0Xb000

Don't know if they are useful/related, but posting them anyway

[Yann Papouin (yann-papouin)]

Same problem here with a ATI MOBILITY FIRE GL T2/T2e.
Apparently, compiz and compiz.real are counting 'GLX_EXT_texture_from_pixmap' in the grep result.
The result must be 3 and with the fglrx it's only 2

[Yann Papouin (yann-papouin)]

Seems to be the same issue as in Bug #173663