restricted-manager picks wrong BusID for video cards

[bunz]

Binary package hint: restricted-manager

System: HP Pavilion DV8000

Video Device: (from lspci -v)

01:00.0 VGA compatible controller: nVidia Corporation GeForce Go 7400 (rev a1) (prog-if 00 [VGA])
        Subsystem: Hewlett-Packard Company Unknown device 30a5
        Flags: bus master, fast devsel, latency 0, IRQ 16
        Memory at d1000000 (32-bit, non-prefetchable) [size=16M]
        Memory at c0000000 (64-bit, prefetchable) [size=256M]
        Memory at d0000000 (64-bit, non-prefetchable) [size=16M]
        Capabilities: [60] Power Management version 2
        Capabilities: [68] Message Signalled Interrupts: Mask- 64bit+ Queue=0/0 Enable-
        Capabilities: [78] Express Endpoint IRQ 0
        Capabilities: [100] Virtual Channel
        Capabilities: [128] Power Budgeting

restricted-manager configures my device in xorg.conf as:

Section "Device"
        Identifier "ATI Technologies, Inc. Radeon RV100 QY [Radeon 7000/VE]"
        Driver "nvidia"
        BusID "PCI:0:5:0"
        Option "AddARGBGLXVisuals" "True"
EndSection

Manually changing to the setting below fixes the problem:

Section "Device"
        Identifier "NVIDIA GeForce Go 7400"
        Driver "nvidia"
        BusID "PCI:1:0:0"
        Option "AddARGBGLXVisuals" "True"
EndSection

Also, the xorg.conf created by installing the nvidia driver omits the "1440x900' mode from my screen defs.

[bunz (bunz)]

Forgot to mention - restricted-manager version is 0.4

[John Dong (jdong)]

The same thing happened when I walked a friend thru configuring his fglrx card with restricted-manager.

BusID 0:5:0 got put in when 1:0:1 was the correct ID from lspci.

[John Dong (jdong)]

A sample snippet of lspci:

01:00.0 VGA compatible controller: ATI Technologies Inc RV380 [Radeon X600 (PCIE
)]
01:00.1 Display controller: ATI Technologies Inc RV380 [Radeon X600]

For ATI's that FIRST ONE that shows up (1:0:0) is the one that needs to go into xorg.conf. The second one is a fake, virtual phantom clone one that ATI' Windows drivers use to trick dual-head mode.

[Martin Pitt (pitti)]

restricted-manager does not configure bus IDs, it just sets some debconf keys and calls dexconf. So this needs to be fixed in x11-common.

[Martin Pitt (pitti)]

What does

  grep -A 2 'xserver-xorg/config/device/bus_id$' /var/cache/debconf/config.dat

show for you?

[Martin Pitt (pitti)]

I changed r-m to not call dexconf directly any more, but use the magic in xserver-xorg's postinst script as well. Can you please install the attached r-m and check whether this still happens?

Thank you!

[Jonathan Riddell (jr)]

>grep -A 2 'xserver-xorg/config/device/bus_id$' /var/cache/debconf/config.dat
Name: xserver-xorg/config/device/bus_id
Template: xserver-xorg/config/device/bus_id
Value: PCI:1:5:0
Owners: xserver-xorg

[Jonathan Riddell (jr)]

Using the .deb from comment 6 it now keeps PCI:1:5:0

[Martin Pitt (pitti)]

For the record, this fixes the BusID issue for Jonathan.

[Martin Pitt (pitti)]

Hm, neither dexconf nor dpkg-reconfigure xserver-xorg do the right thing, but let's put this back to r-m for the time being, so that it appears on more of my radars.

[Crashbit (crashbit-gmail)]

$ lspci -b
00:00.0 Host bridge: ALi Corporation M1695 K8 Northbridge [PCI Express and HyperTransport]
00:01.0 PCI bridge: ALi Corporation PCI Express Root Port
00:02.0 PCI bridge: ALi Corporation PCI Express Root Port
00:03.0 PCI bridge: ALi Corporation PCI Express Root Port
00:04.0 Host bridge: ALi Corporation M1689 K8 Northbridge [Super K8 Single Chip]
00:05.0 PCI bridge: ALi Corporation AGP8X Controller
00:06.0 PCI bridge: ALi Corporation M5249 HTT to PCI Bridge
00:07.0 ISA bridge: ALi Corporation M1563 HyperTransport South Bridge (rev 70)
00:07.1 Bridge: ALi Corporation M7101 Power Management Controller [PMU]
00:08.0 Multimedia audio controller: ALi Corporation M5455 PCI AC-Link Controller Audio Device (rev 20)
00:11.0 Ethernet controller: ALi Corporation ULi 1689,1573 integrated ethernet. (rev 40)
00:12.0 IDE interface: ALi Corporation M5229 IDE (rev c7)
00:13.0 USB Controller: ALi Corporation USB 1.1 Controller (rev 03)
00:13.1 USB Controller: ALi Corporation USB 1.1 Controller (rev 03)
00:13.2 USB Controller: ALi Corporation USB 1.1 Controller (rev 03)
00:13.3 USB Controller: ALi Corporation USB 2.0 Controller (rev 01)
00:18.0 Host bridge: Advanced Micro Devices [AMD] K8 [Athlon64/Opteron] HyperTransport Technology Configuration
00:18.1 Host bridge: Advanced Micro Devices [AMD] K8 [Athlon64/Opteron] Address Map
00:18.2 Host bridge: Advanced Micro Devices [AMD] K8 [Athlon64/Opteron] DRAM Controller
00:18.3 Host bridge: Advanced Micro Devices [AMD] K8 [Athlon64/Opteron] Miscellaneous Control
01:00.0 VGA compatible controller: ATI Technologies Inc RV370 [Sapphire X550 Silent]
01:00.1 Display controller: ATI Technologies Inc RV370 secondary [Sapphire X550 Silent]
03:00.0 SATA controller: JMicron Technologies, Inc. JMicron 20360/20363 AHCI Controller

and in the xorg.conf BusID 1:5:0 ----> Agp controller
no BusID 1:0:0 ----> vga compatible controller (PCI-Express)

[Martin Pitt (pitti)]

Adam, thanks for your help so far. Can you please install the attached version and try again? I. e. reset to defaults (vesa, etc.), then enable fglrx in restricted-manager, and then check whether it works as expected. There is a remote chance that this version already fixes the issue we had yesterday (keeping 'vesa' in xorg.conf even after enabling fglrx). If it does not work, can you please attach /var/log/restricted-manger.log and /etc/X11/xorg.conf after that?

Thank you!

[Martin Pitt (pitti)]

Please ignore the previous package, use this one instead. Thank you!

[Martin Pitt (pitti)]

Confirmed by two people that this works now. Thanks all!

[Martin Pitt (pitti)]

 restricted-manager (0.10) feisty; urgency=low
 .
   * RestrictedManager/{fglrx,nvidia}.py: Call dpkg-reconfigure xserver-xorg
     instead of dexconf directly, to take advantage of its special magic. This
     also takes care of writing backup files, so disable our own backup file
     writing. (LP: #91036)
   * RestrictedManager/{fglrx,nvidia}.py: Set the 'seen' flags of the modified
     debconf questions, to reduce the chance of dpkg-reconfigure clobbering
     them.
   * RestrictedManager/{fglrx,nvidia}.py: Add some logging to
     /var/log/restricted-manager.log to debug currently open bugs. This will be
     removed again later.
   * restricted-manager: If there are no applicable handlers at all, display an
     info message box that the hardware does not need any restricted drivers
     and exit. (LP: #92485)
   * restricted-manager, RestrictedManager/core.py: Move reboot notification
     function into core.notify_reboot_required(), so that we can use in the
     handlers, too.
   * RestrictedManager/{fglrx,nvidia}.py: Trigger reboot notification on
     enabling, too. (LP: #92684)
   * Add modalias_override/fglrx_supported: Script to parse supported product
     IDs out of /usr/lib/xorg/modules/drivers/fglrx_drv.so. Thanks to Johan
     Kiviniemi!
   * RestrictedManager/modalias.append: Replace the 'catch all ATI graphics
     cards' line with the product ID list generated by fglrx_supported, so that
     the fglrx driver is not offered any more on models it doesn't support.
     (LP: #92498)
   * Renamed RestrictedManager/nvidia.py to RestrictedManager/xorg_driver.py
     and replaced the few nvidia-specific bits with ctor argument variables.
   * Reimplement nvidia.py and fglrx.py as XorgDriver subclasses, since they
     shared 95% of code.
   * RestrictedManager/xorg_driver.py: Set xserver-xorg/autodetect_video_card
     to false when enabling a driver, so that xserver-xorg's postinst does not
     clobber our freshly selected restricted driver. Enable it again when
     disabling the restricted driver.
   * po/POTFILES.in: Add missing fglrx.py and xorg_driver.py.
   * restricted-manager, RestrictedManager/core.py: Add new DefaultHandler
     method can_change() which checks if r-m can change this driver at all. If
     not, show an error message with an explanation.
   * RestrictedManager/xorg_driver.py: Add can_change() which checks for a
     modified xorg.conf. (LP: #92690)