"Proprietary Drivers" shows incorrect version if installed fglrx manually

Bug #291672 reported by dmoyne
2
Affects Status Importance Assigned to Milestone
fglrx-installer (Ubuntu)
Won't Fix
Wishlist
Unassigned

Bug Description

I have an HD 4850 ATI video card ; with Hardy on 64 bit system I could manually install proprietary driver 8.10 from ATI successfully ; I could not reiterate the same process of installation on Intrepid with a different kernel after upgrade from Hardy to Intrepid ; As the app "Proprietary Pilotes" included in Ubuntu package failed to install anything for my video card (probably resulting from upgrade process) I had to remove all fglrx* packages to at last retry installing with the app "Proprietary Pilotes" this time with some success (glxinfo confirms DRI rendering) but I am left with a few questions :
- when looking at window "Proprietary Pilotes" it says that an different version of the pilote as described is installed ; never seen something like this ! ; so what should I do to install correctly what is described ?
- when running Catalyst menu it says that I am running on1680 x 1050 @ 75 Hz resolution whereas it is actually 1680 x 1050 @ 60 Hz
- when looking at the doc attached to package xorg-driver-fglrx installed it does say that ATI 4800 series are supported by such driver package.

So to conclude it works somehow but apparently is not supposed the good driver for all the reasons listed above.

I will keep trying to manually instally the ATI package as I have done on Hardy and post messages if I at last succeed in dooing it.
Regards

Revision history for this message
Bryce Harrington (bryce) wrote :

Hi daniel-moyne,

Please attach the output of `lspci -vvnn`, and attach your /var/log/Xorg.0.log file from after reproducing this issue. If you've made any customizations to your /etc/X11/xorg.conf please attach that as well.

Changed in fglrx-installer:
status: New → Incomplete
Revision history for this message
dmoyne (daniel-moyne) wrote :
Download full text (44.3 KiB)

Here is the desired info :
(1) output of `lspci -vvnn` :
dmoyne@gavarnie:~$ lspci -vvnn
00:00.0 Host bridge [0600]: Intel Corporation 4 Series Chipset DRAM Controller [8086:2e20] (rev 02)
        Subsystem: Giga-byte Technology Device [1458:5000]
        Control: I/O- Mem+ BusMaster+ SpecCycle- MemWINV- VGASnoop- ParErr- Stepping- SERR- FastB2B- DisINTx-
        Status: Cap+ 66MHz- UDF- FastB2B+ ParErr- DEVSEL=fast >TAbort- <TAbort- <MAbort+ >SERR- <PERR- INTx-
        Latency: 0
        Capabilities: <access denied>
        Kernel modules: intel-agp

00:01.0 PCI bridge [0604]: Intel Corporation 4 Series Chipset PCI Express Root Port [8086:2e21] (rev 02)
        Control: I/O+ Mem+ BusMaster+ SpecCycle- MemWINV- VGASnoop- ParErr- Stepping- SERR- FastB2B- DisINTx+
        Status: Cap+ 66MHz- UDF- FastB2B- ParErr- DEVSEL=fast >TAbort- <TAbort- <MAbort- >SERR- <PERR- INTx-
        Latency: 0, Cache Line Size: 32 bytes
        Bus: primary=00, secondary=01, subordinate=01, sec-latency=0
        I/O behind bridge: 0000a000-0000afff
        Memory behind bridge: e4000000-e5ffffff
        Prefetchable memory behind bridge: 00000000d0000000-00000000dfffffff
        Secondary status: 66MHz- FastB2B- ParErr- DEVSEL=fast >TAbort- <TAbort- <MAbort+ <SERR- <PERR-
        BridgeCtl: Parity- SERR- NoISA- VGA+ MAbort- >Reset- FastB2B-
                PriDiscTmr- SecDiscTmr- DiscTmrStat- DiscTmrSERREn-
        Capabilities: <access denied>
        Kernel driver in use: pcieport-driver
        Kernel modules: shpchp

00:1a.0 USB Controller [0c03]: Intel Corporation 82801JI (ICH10 Family) USB UHCI Controller #4 [8086:3a37]
        Subsystem: Giga-byte Technology Device [1458:5004]
        Control: I/O+ Mem- BusMaster+ SpecCycle- MemWINV- VGASnoop- ParErr- Stepping- SERR- FastB2B- DisINTx-
        Status: Cap+ 66MHz- UDF- FastB2B+ ParErr- DEVSEL=medium >TAbort- <TAbort- <MAbort- >SERR- <PERR- INTx-
        Latency: 0
        Interrupt: pin A routed to IRQ 16
        Region 4: I/O ports at e000 [size=32]
        Capabilities: <access denied> ...

Revision history for this message
dmoyne (daniel-moyne) wrote :

Continued !:
my "xorg.conf file" attached.

Revision history for this message
Bryce Harrington (bryce) wrote :

Thanks for attaching the files.

I suspect that the reason the version numbers don't match is because you manually installed the driver, so the packaging files that track your driver version did not get a chance to be properly updated. This probably should be investigated further to see if Jockey or something can detect if a newer fglrx was manually installed.

I don't know what the issue with the catalyst menu's frequency could be. What method are you using for determining that you're actually running at 60Hz?

We would suggest you should use Envy for updating the driver on hardy rather than install manually from the ATI website, since Envy contains checks for issues and updates your xorg.conf with workarounds as appropriate.

Changed in fglrx-installer:
status: Incomplete → Triaged
importance: Undecided → Wishlist
Changed in fglrx-installer:
status: Triaged → Won't Fix
Revision history for this message
dmoyne (daniel-moyne) wrote :

with a resolution of 1680 x 1050 only a 60 Hz frequency can be selected as maximum frequency available (Samsung 22").

I reinstalled the ati driver supplied by the ubuntu distro and the same inconsistency existe on delivered information especially on propretary driver installed ; I will send you next a snapshot of what I get from "proprietary driver".

I have never used Envy I will try first on Hardy before trying it of my second Intrepide machine.
Regards

Revision history for this message
dmoyne (daniel-moyne) wrote :

with the new kernel update to 2-6-27-10 I could recompile the proprietary driver 8.11 supplied by ATI ; with this method as sonn as an update of kernel is activated we loose functionability of previous driver compilation (see bug report #237216) ; in their driver installer package they ptopose either to install drivers directly rhough a combined existing kernel compilation or to build package ; I have not tried the latter alternative but I am not sure it would bring coherency with new kernel update ? ; really lisibility of this window is very confusing :
- on top a line saying that proprietary driver is installed :I wish I could read the version of driver as installed manually by myself !
- a lin that for me does not make sense :
on ebutton (can it be checked ?iunder some particular conditions ) + one comment line saying that a different version of driver is installed (different than what ?) + a button "Activate" (I wonder wether if the correct translation is not "Activé " meaning "Activated" but then why make it anyways selectable for obsure action because as far as I know it is working as is so no need to click on this tricky latter button !
Regards.

By the way some last important points that I want to address here :

(a) posix shared memory
In a technical note linked to ATI driver installation they say that posix shared memory (/dev/shm) must be activated ; I have added in fsab file the following line :
# posix shared memory
tmpfs /dev/shm tmpfs defaults 0 0
2 questions :
why is not such line added in fstab even when using ubuntu fglrx package p?
After doing so /dev/shm directory is still empty !

(b) manual installation of proprietary driver
After manual installation of proprietary driver I still have xorg-driver-fglrx + fglrx-amdccle packages appearing as installed which for me does not make sense as if not mistaken I had removed them before manually installing the proprietary driver (in ubuntu list they say that all *fglrx* packages MUST BE REOVED prior to installing proprietary driver) ; I wonder if when updating kernel from 2-6-27-8 to 2-6-27-10 those packages have not been re-installed at the time because felt as needed ? ; can I now safely remove them ? ; what about vga and vesa packages is it safer to keep them in case ; are they necessary during the boot process before final activation of proprietary driver ?

(c) is it possible to plan in the near future a ubuntu package of ATI driver (with their authorization ) that would be sytematically activated in case of installation of new kernel to anticipate for non installation of new kernel when combined installtion of proprietary driver fails to terminate successfully to avoid a black screen at the next reboot ?

Regards.

To post a comment you must log in.
This report contains Public information  
Everyone can see this information.

Other bug subscribers

Remote bug watches

Bug watches keep track of this bug in other bug trackers.