time2IPL
05-15-2007, 03:29 PM
Hi,
I've posted before regarding my own "daily driver" machine and others; the problem in all cases is that any time I attempt to use ATI's proprietary driver, when X tries to start, the machines go off into the weeds. X starts to start (sorry for that; I had to ...), but then - when I've been able to observe what's going on, which I can do to a point by starting xinit via "nice" CPU utilization (system) is 100%, the screen blanks, monitor can be heard attempting to change mode, and then - nothing. Nada. Zip. Zilch.
I've read of others having similar problems. The 8.32.5 driver works on both my "usual" machine and on the others, too; as the card in mine and several of the others is a R500 series (Radeon X1600 Pro, 512Mb), I can't use the (current) open-source driver; it's the currently proprietary driver, VESA, or nothing.
The last thing Xorg.0.log ever shows - and this is consistent on every one of these machines - is the first few lines that indicate the server is attempting to load libfgrlxdrm.so. Interestingly, if I move libfglrxdrm.so "out of the way", the server comes up, albeit without DRM, DRI support. I've run xinit under strace, which confirmed that X was either dying in libfglrxdrm, or in whatever step comes immediately after that's loaded.
I'm posting this as I am working on yet another machine with the exact same problem. The card in this machine is an "ATI Radeon 9800 Pro (R350) 128Mb 8X AGP Graphics Card"; the machine is a Dell Dimension 8300, P4 @ 2.60Ghz, 512Mb RAM; HDD == 120Gb Hitachi / formerly IBM Deskstar ATA-100 drive. The motherboard is using some variation of the ICH5 chipset; it has 8X AGP & uses the i875 chipset for AGP.
Not a bad box, but getting a bit dated; it strains when running Celestia, or Google Earth. Without the proprietary ATI drivers, the machine is slow enough that neither of those two apps is usable; no great surprise there.
The only thing these machines have in common, as far as I can tell, is that they're all running Fedora Core 6 currently; I tried F7, same problems; also tried FC3. Some are running in x86_64 (AMD64) mode, others are running in "32-bit compatibility mode" (i686 mode). Several of them are multi-core CPU based boxes, several aren't.
I've rebuilt the kernel in various ways to rule out the possibility that, for example, having memory configured as "sparse" might be causing a problem (as opposed to "Flat"; I've turned kernel IRQ balancing on and off; I've turned higmem off completely, and have also tried it w/ the usual split && at 64Gb. To date, nothing I've done has made the least bit of a difference; the stock kernels don't work any better, either.
I'm attaching a list of the machine's configuration along with parts of its log files; if anyone could shed some light on what's going on I would really appreciate it; I simply can't imagine what's going on here. Unless FC6 & the X server it's distributed with and/or some other OS components have some sort of fundamental incompatibility/ies. If for no other reason than my own edification, I would very much like to know why these machines are doing what they are doing.
BTW, has anyone managed to get a driver newer than 8.32.5 running on FC6 with the X server as shipped (eg., RPM install) up and running? Particularly, on a system with an Athlon x2 (socket AM2) 4200, 2Gb DDR2, X1600 (R500 series) graphics card? I've still not yet seen anything to indicate that anyone has been successful with respect to getting that combiniation up and running.
Thanks,
- Larry
(Attachments)
I. dmesg:
[fglrx] Maximum main memory to use for locked dma buffers: 430 MBytes.
[fglrx] module loaded - fglrx 8.32.5 [Dec 12 2006] on minor 0
PCI: Found IRQ 11 for device 0000:01:00.0
PCI: Sharing IRQ 11 with 0000:00:1d.0
PCI: Sharing IRQ 11 with 0000:00:1d.3
[fglrx] Maximum main memory to use for locked dma buffers: 430 MBytes.
[fglrx] AGP detected, AgpState = 0x1f004a1b (hardware caps of chipset)
agpgart: Found an AGP 3.0 compliant device at 0000:00:00.0.
agpgart: Putting AGP V3 device at 0000:00:00.0 into 8x mode
agpgart: Putting AGP V3 device at 0000:01:00.0 into 8x mode
[fglrx] AGP enabled, AgpCommand = 0x1f004312 (selected caps)
[fglrx] total GART = 134217728
[fglrx] free GART = 123727872
[fglrx] max single GART = 123727872
[fglrx] total LFB = 134217728
[fglrx] free LFB = 116387840
[fglrx] max single LFB = 116387840
[fglrx] total Inv = 0
[fglrx] free Inv = 0
[fglrx] max single Inv = 0
[fglrx] total TIM = 0
II. Check of Xorg.0.log for errors, warnings:
$ grep WW /var/log/Xorg.0.log
(WW) warning, (EE) error, (NI) not implemented, (??) unknown.
(WW) fglrx: No matching Device section for instance (BusID PCI:1:0:1) found
$ grep EE /var/log/Xorg.0.log
(WW) warning, (EE) error, (NI) not implemented, (??) unknown.
(II) Loading extension MIT-SCREEN-SAVER
III. /var/log/messages
A. AGP
Linux agpgart interface v0.101 (c) Dave Jones
agpgart: Detected an Intel i875 Chipset.
agpgart: AGP aperture is 128M @ 0xe0000000
B. Oddity noted in /var/log/messages
May 15 13:18:04 cirrus gdm[2787]: gdm_config_parse: Standard X server not found; trying alternatives
May 15 13:18:04 cirrus gdm[2787]: no suitable security token driver could be found
C. Usual fglrx info: what are "Inv" && "TIM"; why do they have a size of
zero (0) bytes?
May 15 13:18:06 cirrus kernel: [fglrx] Maximum main memory to use for locked dma buffers: 430 MBytes.
May 15 13:18:06 cirrus kernel: [fglrx] module loaded - fglrx 8.32.5 [Dec 12 2006] on minor 0
May 15 13:18:06 cirrus kernel: PCI: Found IRQ 11 for device 0000:01:00.0
May 15 13:18:06 cirrus kernel: PCI: Sharing IRQ 11 with 0000:00:1d.0
May 15 13:18:06 cirrus kernel: PCI: Sharing IRQ 11 with 0000:00:1d.3
May 15 13:18:09 cirrus kernel: [fglrx] Maximum main memory to use for locked dma buffers: 430 MBytes.
May 15 13:18:09 cirrus kernel: [fglrx] AGP detected, AgpState = 0x1f004a1b (hardware caps of chipset)
May 15 13:18:09 cirrus kernel: agpgart: Found an AGP 3.0 compliant device at 0000:00:00.0.
May 15 13:18:09 cirrus kernel: agpgart: Putting AGP V3 device at 0000:00:00.0 into 8x mode
May 15 13:18:09 cirrus kernel: agpgart: Putting AGP V3 device at 0000:01:00.0 into 8x mode
May 15 13:18:09 cirrus kernel: [fglrx] AGP enabled, AgpCommand = 0x1f004312 (selected caps)
May 15 13:18:09 cirrus kernel: [fglrx] total GART = 134217728
May 15 13:18:09 cirrus kernel: [fglrx] free GART = 123727872
May 15 13:18:09 cirrus kernel: [fglrx] max single GART = 123727872
May 15 13:18:09 cirrus kernel: [fglrx] total LFB = 134217728
May 15 13:18:09 cirrus kernel: [fglrx] free LFB = 116387840
May 15 13:18:09 cirrus kernel: [fglrx] max single LFB = 116387840
May 15 13:18:09 cirrus kernel: [fglrx] total Inv = 0
May 15 13:18:09 cirrus kernel: [fglrx] free Inv = 0
May 15 13:18:09 cirrus kernel: [fglrx] max single Inv = 0
May 15 13:18:09 cirrus kernel: [fglrx] total TIM = 0
IV. Card ID, lspci
01:00.0 VGA compatible controller: ATI Technologies Inc Radeon R350 [Radeon 9800 Pro] (prog-if 00 [VGA])
Subsystem: ATI Technologies Inc Unknown device 3002
Control: I/O+ Mem+ BusMaster+ SpecCycle- MemWINV- VGASnoop- ParErr- Stepping+ SERR+ FastB2B-
Status: Cap+ 66MHz+ UDF- FastB2B+ ParErr- DEVSEL=medium >TAbort- <TAbort- <MAbort- >SERR- <PERR-
Latency: 255 (2000ns min), Cache Line Size: 64 bytes
Interrupt: pin A routed to IRQ 11
Region 0: Memory at f0000000 (32-bit, prefetchable) [size=128M]
Region 1: I/O ports at de00 [size=256]
Region 2: Memory at fe9e0000 (32-bit, non-prefetchable) [size=64K]
Expansion ROM at fea00000 [disabled] [size=128K]
Capabilities: [58] AGP version 3.0
Status: RQ=256 Iso- ArqSz=0 Cal=0 SBA+ ITACoh- GART64- HTrans- 64bit- FW+ AGP3+ Rate=x4,x8
Command: RQ=1 ArqSz=0 Cal=0 SBA+ AGP- GART64- 64bit- FW- Rate=<none>
Capabilities: [50] Power Management version 2
Flags: PMEClk- DSI- D1+ D2+ AuxCurrent=0mA PME(D0-,D1-,D2-,D3hot-,D3cold-)
Status: D0 PME-Enable- DSel=0 DScale=0 PME-
01:00.1 Display controller: ATI Technologies Inc Radeon R350 [Radeon 9800 Pro] (Secondary)
Subsystem: ATI Technologies Inc Unknown device 3003
Control: I/O+ Mem+ BusMaster+ SpecCycle- MemWINV- VGASnoop- ParErr- Stepping+ SERR- FastB2B-
Status: Cap+ 66MHz+ UDF- FastB2B+ ParErr- DEVSEL=medium >TAbort- <TAbort- <MAbort- >SERR- <PERR-
Latency: 64 (2000ns min), Cache Line Size: 64 bytes
Region 0: Memory at e8000000 (32-bit, prefetchable) [size=128M]
Region 1: Memory at fe9f0000 (32-bit, non-prefetchable) [size=64K]
Capabilities: [50] Power Management version 2
Flags: PMEClk- DSI- D1+ D2+ AuxCurrent=0mA PME(D0-,D1-,D2-,D3hot-,D3cold-)
Status: D0 PME-Enable- DSel=0 DScale=0 PME-
I've posted before regarding my own "daily driver" machine and others; the problem in all cases is that any time I attempt to use ATI's proprietary driver, when X tries to start, the machines go off into the weeds. X starts to start (sorry for that; I had to ...), but then - when I've been able to observe what's going on, which I can do to a point by starting xinit via "nice" CPU utilization (system) is 100%, the screen blanks, monitor can be heard attempting to change mode, and then - nothing. Nada. Zip. Zilch.
I've read of others having similar problems. The 8.32.5 driver works on both my "usual" machine and on the others, too; as the card in mine and several of the others is a R500 series (Radeon X1600 Pro, 512Mb), I can't use the (current) open-source driver; it's the currently proprietary driver, VESA, or nothing.
The last thing Xorg.0.log ever shows - and this is consistent on every one of these machines - is the first few lines that indicate the server is attempting to load libfgrlxdrm.so. Interestingly, if I move libfglrxdrm.so "out of the way", the server comes up, albeit without DRM, DRI support. I've run xinit under strace, which confirmed that X was either dying in libfglrxdrm, or in whatever step comes immediately after that's loaded.
I'm posting this as I am working on yet another machine with the exact same problem. The card in this machine is an "ATI Radeon 9800 Pro (R350) 128Mb 8X AGP Graphics Card"; the machine is a Dell Dimension 8300, P4 @ 2.60Ghz, 512Mb RAM; HDD == 120Gb Hitachi / formerly IBM Deskstar ATA-100 drive. The motherboard is using some variation of the ICH5 chipset; it has 8X AGP & uses the i875 chipset for AGP.
Not a bad box, but getting a bit dated; it strains when running Celestia, or Google Earth. Without the proprietary ATI drivers, the machine is slow enough that neither of those two apps is usable; no great surprise there.
The only thing these machines have in common, as far as I can tell, is that they're all running Fedora Core 6 currently; I tried F7, same problems; also tried FC3. Some are running in x86_64 (AMD64) mode, others are running in "32-bit compatibility mode" (i686 mode). Several of them are multi-core CPU based boxes, several aren't.
I've rebuilt the kernel in various ways to rule out the possibility that, for example, having memory configured as "sparse" might be causing a problem (as opposed to "Flat"; I've turned kernel IRQ balancing on and off; I've turned higmem off completely, and have also tried it w/ the usual split && at 64Gb. To date, nothing I've done has made the least bit of a difference; the stock kernels don't work any better, either.
I'm attaching a list of the machine's configuration along with parts of its log files; if anyone could shed some light on what's going on I would really appreciate it; I simply can't imagine what's going on here. Unless FC6 & the X server it's distributed with and/or some other OS components have some sort of fundamental incompatibility/ies. If for no other reason than my own edification, I would very much like to know why these machines are doing what they are doing.
BTW, has anyone managed to get a driver newer than 8.32.5 running on FC6 with the X server as shipped (eg., RPM install) up and running? Particularly, on a system with an Athlon x2 (socket AM2) 4200, 2Gb DDR2, X1600 (R500 series) graphics card? I've still not yet seen anything to indicate that anyone has been successful with respect to getting that combiniation up and running.
Thanks,
- Larry
(Attachments)
I. dmesg:
[fglrx] Maximum main memory to use for locked dma buffers: 430 MBytes.
[fglrx] module loaded - fglrx 8.32.5 [Dec 12 2006] on minor 0
PCI: Found IRQ 11 for device 0000:01:00.0
PCI: Sharing IRQ 11 with 0000:00:1d.0
PCI: Sharing IRQ 11 with 0000:00:1d.3
[fglrx] Maximum main memory to use for locked dma buffers: 430 MBytes.
[fglrx] AGP detected, AgpState = 0x1f004a1b (hardware caps of chipset)
agpgart: Found an AGP 3.0 compliant device at 0000:00:00.0.
agpgart: Putting AGP V3 device at 0000:00:00.0 into 8x mode
agpgart: Putting AGP V3 device at 0000:01:00.0 into 8x mode
[fglrx] AGP enabled, AgpCommand = 0x1f004312 (selected caps)
[fglrx] total GART = 134217728
[fglrx] free GART = 123727872
[fglrx] max single GART = 123727872
[fglrx] total LFB = 134217728
[fglrx] free LFB = 116387840
[fglrx] max single LFB = 116387840
[fglrx] total Inv = 0
[fglrx] free Inv = 0
[fglrx] max single Inv = 0
[fglrx] total TIM = 0
II. Check of Xorg.0.log for errors, warnings:
$ grep WW /var/log/Xorg.0.log
(WW) warning, (EE) error, (NI) not implemented, (??) unknown.
(WW) fglrx: No matching Device section for instance (BusID PCI:1:0:1) found
$ grep EE /var/log/Xorg.0.log
(WW) warning, (EE) error, (NI) not implemented, (??) unknown.
(II) Loading extension MIT-SCREEN-SAVER
III. /var/log/messages
A. AGP
Linux agpgart interface v0.101 (c) Dave Jones
agpgart: Detected an Intel i875 Chipset.
agpgart: AGP aperture is 128M @ 0xe0000000
B. Oddity noted in /var/log/messages
May 15 13:18:04 cirrus gdm[2787]: gdm_config_parse: Standard X server not found; trying alternatives
May 15 13:18:04 cirrus gdm[2787]: no suitable security token driver could be found
C. Usual fglrx info: what are "Inv" && "TIM"; why do they have a size of
zero (0) bytes?
May 15 13:18:06 cirrus kernel: [fglrx] Maximum main memory to use for locked dma buffers: 430 MBytes.
May 15 13:18:06 cirrus kernel: [fglrx] module loaded - fglrx 8.32.5 [Dec 12 2006] on minor 0
May 15 13:18:06 cirrus kernel: PCI: Found IRQ 11 for device 0000:01:00.0
May 15 13:18:06 cirrus kernel: PCI: Sharing IRQ 11 with 0000:00:1d.0
May 15 13:18:06 cirrus kernel: PCI: Sharing IRQ 11 with 0000:00:1d.3
May 15 13:18:09 cirrus kernel: [fglrx] Maximum main memory to use for locked dma buffers: 430 MBytes.
May 15 13:18:09 cirrus kernel: [fglrx] AGP detected, AgpState = 0x1f004a1b (hardware caps of chipset)
May 15 13:18:09 cirrus kernel: agpgart: Found an AGP 3.0 compliant device at 0000:00:00.0.
May 15 13:18:09 cirrus kernel: agpgart: Putting AGP V3 device at 0000:00:00.0 into 8x mode
May 15 13:18:09 cirrus kernel: agpgart: Putting AGP V3 device at 0000:01:00.0 into 8x mode
May 15 13:18:09 cirrus kernel: [fglrx] AGP enabled, AgpCommand = 0x1f004312 (selected caps)
May 15 13:18:09 cirrus kernel: [fglrx] total GART = 134217728
May 15 13:18:09 cirrus kernel: [fglrx] free GART = 123727872
May 15 13:18:09 cirrus kernel: [fglrx] max single GART = 123727872
May 15 13:18:09 cirrus kernel: [fglrx] total LFB = 134217728
May 15 13:18:09 cirrus kernel: [fglrx] free LFB = 116387840
May 15 13:18:09 cirrus kernel: [fglrx] max single LFB = 116387840
May 15 13:18:09 cirrus kernel: [fglrx] total Inv = 0
May 15 13:18:09 cirrus kernel: [fglrx] free Inv = 0
May 15 13:18:09 cirrus kernel: [fglrx] max single Inv = 0
May 15 13:18:09 cirrus kernel: [fglrx] total TIM = 0
IV. Card ID, lspci
01:00.0 VGA compatible controller: ATI Technologies Inc Radeon R350 [Radeon 9800 Pro] (prog-if 00 [VGA])
Subsystem: ATI Technologies Inc Unknown device 3002
Control: I/O+ Mem+ BusMaster+ SpecCycle- MemWINV- VGASnoop- ParErr- Stepping+ SERR+ FastB2B-
Status: Cap+ 66MHz+ UDF- FastB2B+ ParErr- DEVSEL=medium >TAbort- <TAbort- <MAbort- >SERR- <PERR-
Latency: 255 (2000ns min), Cache Line Size: 64 bytes
Interrupt: pin A routed to IRQ 11
Region 0: Memory at f0000000 (32-bit, prefetchable) [size=128M]
Region 1: I/O ports at de00 [size=256]
Region 2: Memory at fe9e0000 (32-bit, non-prefetchable) [size=64K]
Expansion ROM at fea00000 [disabled] [size=128K]
Capabilities: [58] AGP version 3.0
Status: RQ=256 Iso- ArqSz=0 Cal=0 SBA+ ITACoh- GART64- HTrans- 64bit- FW+ AGP3+ Rate=x4,x8
Command: RQ=1 ArqSz=0 Cal=0 SBA+ AGP- GART64- 64bit- FW- Rate=<none>
Capabilities: [50] Power Management version 2
Flags: PMEClk- DSI- D1+ D2+ AuxCurrent=0mA PME(D0-,D1-,D2-,D3hot-,D3cold-)
Status: D0 PME-Enable- DSel=0 DScale=0 PME-
01:00.1 Display controller: ATI Technologies Inc Radeon R350 [Radeon 9800 Pro] (Secondary)
Subsystem: ATI Technologies Inc Unknown device 3003
Control: I/O+ Mem+ BusMaster+ SpecCycle- MemWINV- VGASnoop- ParErr- Stepping+ SERR- FastB2B-
Status: Cap+ 66MHz+ UDF- FastB2B+ ParErr- DEVSEL=medium >TAbort- <TAbort- <MAbort- >SERR- <PERR-
Latency: 64 (2000ns min), Cache Line Size: 64 bytes
Region 0: Memory at e8000000 (32-bit, prefetchable) [size=128M]
Region 1: Memory at fe9f0000 (32-bit, non-prefetchable) [size=64K]
Capabilities: [50] Power Management version 2
Flags: PMEClk- DSI- D1+ D2+ AuxCurrent=0mA PME(D0-,D1-,D2-,D3hot-,D3cold-)
Status: D0 PME-Enable- DSel=0 DScale=0 PME-