Please ship proper modaliases for nvidia, fglrx & co

[Trouilliez vincent]

Binary package hint: restricted-manager

My desktop computer has an Nvidia Geforce 6200 video card.
I installed the nvidia-glx via synaptic, to get 3D working.

However the restricted drivers manager refuses to start, and just pops up a message (attached) saying that
"The hardware does not need any restricted drivers".
As much as I would love this to be true, I don't think the 3D Nvidia driver is GPL as of today ! ;-)

ProblemType: Bug
Architecture: i386
Date: Sat Mar 17 22:07:45 2007
DistroRelease: Ubuntu 7.04
Uname: Linux Lotus-Esprit 2.6.20-11-generic #2 SMP Thu Mar 15 08:03:07 UTC 2007 i686 GNU/Linux

[Martin Pitt (pitti)]

Please give me the output of

  lspci | grep VGA; lspci -n | grep 0300

[Trouilliez vincent (vincent-trouilliez-modulonet)]

Yup, here it is:

02:00.0 VGA compatible controller: nVidia Corporation NV44A [GeForce 6200] (rev a1)
02:00.0 0300: 10de:0221 (rev a1)

[Martin Pitt (pitti)]

 restricted-manager (0.11) feisty; urgency=low
 .
   * RestrictedManager/xorg_driver.py: Do not autodetect mouse and monitor when
     reconfiguring xserver-xorg for a new driver. This avoids clobbering
     manually added undetected resolutions. (LP: #93643)
   * po/de.po: Synchronize with Rosetta.
   * RestrictedManager/nvidia.py: Enable NoLogo by default. (LP: #93306)
   * RestrictedManager/nvidia.py: Enable AddARGBVisuals to make Beryl work.
     (LP: #93262)
   * RestrictedManager/xorg_driver.py: Do not mangle xorg.conf if package
     download/installation was aborted. (LP: #92573)
   * RestrictedManager/nvidia.py: Fix passing of driver package name, so that
     the legacy driver actually installs nvidia-glx-legacy. (LP: #92849, part
     one)
   * RestrictedManager/{xorg_driver,nvidia,fglrx}.py: Do not assume that the
     kernel module name always matches the X.org driver module name
     (nvidia_legacy vs. nvidia), but instead pass the X.org driver name as a
     separate attribute. ((LP: #92849, part two)
   * RestrictedManager/modalias.override: Add NV44A [GeForce 6200] (not covered
     by nvidia_supported). (LP: #93209)
   * RestrictedManager/core.py: When loading default or special handlers, check
     that there is actually a kernel module behind. Otherwise we will display
     handlers with overrides on platforms which do not actually provide those
     modules. (LP: #93368)

[Trouilliez vincent (vincent-trouilliez-modulonet)]

I just updated to 0.11, but I still get the same message ?!
do I need to restart Gnome, or the machine, for the change to take effect ?

[Trouilliez vincent (vincent-trouilliez-modulonet)]

I restarted the computer just in case, but still no change. So I have re-opened the bug...

[Martin Pitt (pitti)]

http://people.ubuntu.com/~pitti/tmp/restricted-manager_0.12_all.deb should fix it. Feel free to install this.

I cannot upload this until after the release of Feisty beta.

[Trouilliez vincent (vincent-trouilliez-modulonet)]

Thanks, it does fix the problem, r-m does start now ! :-)
The GUI looks nice, but usability-wise, I am a bit confused : what is the difference between "enabled" and "in use" ? It reports that the Nvidia driver is "in use", but not enabled ? Does that make sense ?
There is no "help" button, so I couldn't figure it out by myself.

[Joey Stanford (joey)]

On my home machine:

Same issue here but different card:

01:00.0 VGA compatible controller: nVidia Corporation NV40 [GeForce 6800 Ultra/GeForce 6800 GT] (rev a2)
01:00.0 0300: 10de:00f9 (rev a2)

[Dean Moreton (dean-moreton)]

same here on my laptop with a nvidia GeForce Go 6100

00:05.0 VGA compatible controller: nVidia Corporation C51 PCI Express Bridge (rev a2)
00:05.0 0300: 10de:0247 (rev a2)

0.12 attached above didn't seem to fix it.

[Johan Kiviniemi (ion)]

We've managed to find a way to grab an accurate list of devices supported by the nVidia driver recently; each card mentioned so far is detected by the bzr version now. The package's maintainer will probably release it soon.

[Matthew Buckett (buckett)]

I was having the same problem with a Nvidia 6200 and after upgrading to 0.12 (from 0.11) I was still getting the same error message.

From the command line I then ran:

buckett@desktop:~$ sudo restricted-manager -c
buckett@desktop:~$ sudo restricted-manager -u

and then the restricted manager ran fine. I'm not sure you need to run both of these command but they seemed to fix it. Should one or both of these commands been in the upgrade scripts for the package?

Enabled: You have requested to use this driver for your system.
In Use: Your system is currently using this driver.

Because some of the drivers require bits of software to be restarted they sometimes cannot be used straight away, hence the reason for the two boxes. A help box or mouseover might be good for this?

[Martin Pitt (pitti)]

 restricted-manager (0.12) feisty; urgency=low
 .
   [ Martin Pitt ]
   * debian/restricted-manager.postinst: Bump version for purging invalid cache
     files on upgrades to 0.11, to clean up after #93368.
   * {applications,autostart}/restricted-manager.desktop.in: Add gettext
     domain. (LP: #94754)
   * restricted-manager, --check-composite: If /var/cache/debconf/config.dat is
     not world-readable (as happens after a crisp clean install for some
     reason), gksu itself and try again.
   * Call programs in /sbin and /usr/sbin/ with absolute paths. (LP: #94516)
 .
   [ Johan Kiviniemi ]
   * RestrictedManager/modalias.{append,override},
     modalias_override/{ath_hal,fglrx,ipw3945.manual,nvidia}:
     Replaced the monolithic modalias pattern files with a directory that
     contains the patterns in separate files. Now manually written patterns
     are in different files than automatically generated ones.
   * debian/rules: Don't install the old patterns.
   * debian/restricted-manager.install: Install the new patterns.
   * RestrictedManager/core.py: Use the new modalias pattern directory
     instead of modalias.{append,override}.
   * modalias_override_scripts/ath_hal_supported: Add a script to generate
     the ath_hal listing from the ath_pci listing.
   * modalias_override_scripts/fglrx_supported:
     - Changed the comments to point to the new pattern directory.
     - Print a comment about the output being automatically generated.
     - Sort the output, so that future changes cause as small diffs as
       possible.
   * modalias_override_scripts/nvidia_supported: Rewrote the whole script.
     Now the IDs are parsed directly from the nvidia modules instead of
     scraping the nVidia website. Now the listing should be complete.
   * modalias_override/nvidia: Generated with the new nvidia_supported
     script. (LP: #93300)
   * debian/restricted-manager.{postinst,postrm}: Always remove cache files
     when upgrading; remove /var/cache/restricted-manager altogether when
     uninstalling.

[revmouse (revmouse)]

I'm running into this same issue. I'm running restricted-manager 0.14 in 7.04 Feisty Fawn beta. I've tried replacing 0.14 with 0.12 and it did not resolve the issue. Now I'm back to 0.14 and the issue is not resolved.

01:00.0 VGA compatible controller: nVidia Corporation GeForce Go 7400 (rev a1) (prog-if 00 [VGA])
        Subsystem: Dell Unknown device 01cc
        Flags: bus master, fast devsel, latency 0, IRQ 9
        Memory at ed000000 (32-bit, non-prefetchable) [size=16M]
        Memory at d0000000 (64-bit, prefetchable) [size=256M]
        Memory at ee000000 (64-bit, non-prefetchable) [size=16M]
        [virtual] Expansion ROM at ef000000 [disabled] [size=128K]
        Capabilities: [60] Power Management version 2
        Capabilities: [68] Message Signalled Interrupts: Mask- 64bit+ Queue=0/0 Enable-
        Capabilities: [78] Express Endpoint IRQ 0

[Martin Pitt (pitti)]

Timothy, please give me the output of those commands:

  dpkg -l 'linux*' restricted-manager | grep ^ii
  uname -a
  lspci -n | grep 0300

Thank you!

[revmouse (revmouse)]

revmouse@ubuntu:~$ dpkg -l 'linux*' restricted-manager | grep ^ii
ii linux-generic 2.6.20.13.10 Complete
Generic Linux kernel
ii linux-headers-2.6.20-12 2.6.20-12.20 Header files
related to Linux kernel version
ii linux-headers-2.6.20-12-generic 2.6.20-12.20 Linux kernel
headers for version 2.6.20 on x
ii linux-headers-2.6.20-13 2.6.20-13.21 Header files
related to Linux kernel version
ii linux-headers-2.6.20-13-generic 2.6.20-13.21 Linux kernel
headers for version 2.6.20 on x
ii linux-headers-generic 2.6.20.13.10 Generic Linux
kernel headers
ii linux-image-2.6.20-12-generic 2.6.20-12.20 Linux kernel
image for version 2.6.20 on x86
ii linux-image-2.6.20-13-generic 2.6.20-13.21 Linux kernel
image for version 2.6.20 on x86
ii linux-image-generic 2.6.20.13.10 Generic Linux
kernel image
ii linux-restricted-modules-2.6.20-12-generic 2.6.20.3-12.11 Non-free
Linux 2.6.20 modules on x86/x86_64
ii linux-restricted-modules-2.6.20-13-generic 2.6.20.4-13.12 Non-free
Linux 2.6.20 modules on x86/x86_64
ii linux-restricted-modules-common 2.6.20.4-13.12 Non-free
Linux 2.6.20 modules helper script
ii linux-restricted-modules-generic 2.6.20.13.10 Restricted
Linux modules for generic kernels
ii linux-sound-base 1.0.13-3ubuntu1 base package
for ALSA and OSS sound systems
ii restricted-manager 0.14 manage
non-free hardware drivers
revmouse@ubuntu:~$ uname -a
Linux ubuntu 2.6.20-13-generic #2 SMP Sun Mar 25 00:21:25 UTC 2007 i686
GNU/Linux
revmouse@ubuntu:~$ lspci -n | grep 0300
01:00.0 0300: 10de:01d8 (rev a1)
revmouse@ubuntu:~$

On 3/28/07, Martin Pitt <email address hidden> wrote:
>
> Timothy, please give me the output of those commands:
>
> dpkg -l 'linux*' restricted-manager | grep ^ii
> uname -a
> lspci -n | grep 0300
>
> Thank you!
>
> --
> Fails to detect Nvidia 3D driver
> https://launchpad.net/bugs/93209
>

[revmouse (revmouse)]

revmouse@ubuntu:~$ dpkg -l 'linux*' restricted-manager | grep ^ii
ii linux-generic 2.6.20.13.10 Complete Generic Linux kernel
ii linux-headers-2.6.20-12 2.6.20-12.20 Header files related to Linux kernel version
ii linux-headers-2.6.20-12-generic 2.6.20-12.20 Linux kernel headers for version 2.6.20 on x
ii linux-headers-2.6.20-13 2.6.20-13.21 Header files related to Linux kernel version
ii linux-headers-2.6.20-13-generic 2.6.20-13.21 Linux kernel headers for version 2.6.20 on x
ii linux-headers-generic 2.6.20.13.10 Generic Linux kernel headers
ii linux-image-2.6.20-12-generic 2.6.20-12.20 Linux kernel image for version 2.6.20 on x86
ii linux-image-2.6.20-13-generic 2.6.20-13.21 Linux kernel image for version 2.6.20 on x86
ii linux-image-generic 2.6.20.13.10 Generic Linux kernel image
ii linux-restricted-modules-2.6.20-12-generic 2.6.20.3-12.11 Non-free Linux 2.6.20 modules on x86/x86_64
ii linux-restricted-modules-2.6.20-13-generic 2.6.20.4-13.12 Non-free Linux 2.6.20 modules on x86/x86_64
ii linux-restricted-modules-common 2.6.20.4-13.12 Non-free Linux 2.6.20 modules helper script
ii linux-restricted-modules-generic 2.6.20.13.10 Restricted Linux modules for generic kernels
ii linux-sound-base 1.0.13-3ubuntu1 base package for ALSA and OSS sound systems
ii restricted-manager 0.14 manage non-free hardware drivers
revmouse@ubuntu:~$ uname -a
Linux ubuntu 2.6.20-13-generic #2 SMP Sun Mar 25 00:21:25 UTC 2007 i686 GNU/Linux
revmouse@ubuntu:~$ lspci -n | grep 0300
01:00.0 0300: 10de:01d8 (rev a1)
revmouse@ubuntu:~$

[R Morrison (rmorrisonva)]

hmm now my nvidia driver no longer works
... a few days ago i updated and rebooted ubuntu (believe it was a kernel upgd.) and my nvidia drivers disappeard in restricted-manager. Ive tried running it "-c" "-u" and my install is up to date (in fact i think i remeber seeing ?.15" as the version of r-m i have). I checked apt-gert and the proper restricted-modules and nvidia-glx packages are installed. but whenever i try to start xorg.conf using the "nvidia" driver i get no screens found. in r-m nvidia doesnt show up even after asking it to "-u"... but lspci shows the hardware is clearly visible on the bus and the nvidia module available in proper version. what should i do... i was hoping restricted-manager meant it'd be eaier to deal with kernel upgrades (not having to run the nvidia driver installer to recompile the modules after each kern upgd) but it seems r-m has actually made it impossible for me to use the nvidia after a kernel upgrade, but alas it seems its a bane of being undder development eh :-)
Thanks,
R Morrison

[Martin Pitt (pitti)]

ID 10de:01d8 is indeed not in the database, I'll add it. Johan, it seems that the driver supports more devices than it claims to...

[R Morrison (rmorrisonva)]

i have a geforce4 go 440... would you like dev ids from my card as well? updated restricted-manager to 0.17 today to no avail... still no nvidia in manager , and issuing it it xorg.cobf x refuses to start: no screens found.... when i first installed the beta restricted-manager worked and correctly configurd my card... im curious why it stopped working. let me be of help if i can.
-Richard

[Loïc Martin (loic-martin3)]

I also have the same chipset (geforce4 go 440) on a Dell Inspiron 8200, and it's detected by restricted-manager who proposes nvidia-legacy (only 7xxx, that has a bug on my system, but that's another bug report).

Could you check by booting a recent daily desktop live CD if the problem still exist for you?

http://cdimage.ubuntu.com/daily-live/current/

Thanks

[R Morrison (rmorrisonva)]

Loic,
 I am writing this from yesterday 3/30 daily build cd for feisty but when i try to open restricted-manager it says "no restricted drivers needed" or something very similar to that. Off of my HD install restricted-manger opens because i have vmware modules installed... and the nvidia card is not listed there of course either.
- Richard

[Martin Pitt (pitti)]

This affects a lot of people, bumping severity. Ben, can this patch be applied for Feisty? Otherwise I have to write a per-architecture solution in restricted-manager, which is pretty ugly.

[Matthew Flower (mattflower)]

I'm running restricted-manager 0.18 and it seems that my Nvidia Go 6600 isn't showing up either.

The results of lspci -n | grep 0300 is:

01:00.0 0300: 10de:0146 (rev a2)

Matt

[robepisc (robepisc)]

Same for my GeForce FX Go5600:
01:00.0 0300: 10de:031a (rev a1)

Thanks!

[Anders H (anders-hem)]

Same for me GeForce4 (NV17) MX440
The results of lspci -n | grep 0300 is:
01:00.0 0300: 10de:0181 (rev a2)

Anders

[SterlingH (sterlinghubbard)]

Same here,

lspci -n | grep 300
00:05.0 0300: 10de:0247 (rev a2)

[Martin Pitt (pitti)]

New l-r-m does ship modaliases which restricted-manager can use. Thanks, Ben!

However, this bug should be kept open for Feisty+1, where the package should be changed to actually override the modaliases, so that they work not just with restricted-manager. Removing the milestone for that.

[PrivateUser132781 (privateuser132781-deactivatedaccount)]

I have a similar issue on Feisty AMD64. When the nvidia-glx package is installed and activated, restricted drivers manager is aware of it. When it is not installed, it does not detect that I have a Nvidia card that can use binary drivers.

lspci lists my card as:
VGA compatible controller: nVidia Corporation C51G [GeForce 6100] (rev a2)

eduard@maanbox:~$ lspci -vvn | grep 300
00:05.0 0300: 10de:0242 (rev a2) (prog-if 00 [VGA])
        Region 3: Memory at f3000000 (64-bit, non-prefetchable) [size=16M]
        [virtual] Expansion ROM at 30000000 [disabled] [size=128K]

[siucdude (siucdude)]

Same Here on a GeForce GO 6600

The results of lspci -n | grep 0300 is:

01:00.0 0300: 10de:0148 (rev a2)

[Magnus Hedemark (viridari)]

I have an nVidia GeForce FX 5200 and have the same problem with feisty.

magnus@grunion:~$ sudo -H lspci -vvn
Password:
00:00.0 0600: 1106:3205
        Subsystem: 1043:80f9
        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: 8
        Region 0: Memory at e0000000 (32-bit, prefetchable) [size=64M]
        Capabilities: [80] AGP version 3.5
                Status: RQ=32 Iso- ArqSz=0 Cal=2 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: [c0] 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-

00:01.0 0604: 1106:b198 (prog-if 00 [Normal decode])
        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: 0
        Bus: primary=00, secondary=01, subordinate=01, sec-latency=0
        Memory behind bridge: e8000000-e9ffffff
        Prefetchable memory behind bridge: e4000000-e7ffffff
        Secondary status: 66MHz+ FastB2B- ParErr- DEVSEL=medium >TAbort- <TAbort- <MAbort+ <SERR- <PERR-
        BridgeCtl: Parity- SERR- NoISA+ VGA+ MAbort- >Reset- FastB2B-
        Capabilities: [80] 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-

00:0b.0 0c00: 1106:3044 (rev 80) (prog-if 10 [OHCI])
        Subsystem: 1043:808a
        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: 32 (8000ns max), Cache Line Size: 32 bytes
        Interrupt: pin A routed to IRQ 16
        Region 0: Memory at ea000000 (32-bit, non-prefetchable) [size=2K]
        Region 1: I/O ports at 9000 [size=128]
        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-

00:0f.0 0101: 1106:3149 (rev 80) (prog-if 8f [Master SecP SecO PriP PriO])
        Subsystem: 1043:80ed
        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: 32
        Interrupt: pin B routed to IRQ 19
        Region 0: I/O ports at 9400 [size=8]
        Region 1: I/O ports at 9800 [size=4]
        Region 2: I/O ports at 9c00 [size=8]
        Region 3: I/O ports at a000 [size=4]
        Region 4: I/O ports at a400 [size=16]
        Region 5: I/O ports at a800 [size=256]
        Capabilities: [c0] Power Management version 2
                Flags: PMEClk- DSI- D1...

[Magnus Hedemark (viridari)]

Very interesting perhaps:

magnus@grunion:~$ gksudo "restricted-manager -l"
X Error: BadDevice, invalid or uninitialized input device 167
  Major opcode: 144
  Minor opcode: 3
  Resource id: 0x0
Failed to open device
X Error: BadDevice, invalid or uninitialized input device 167
  Major opcode: 144
  Minor opcode: 3
  Resource id: 0x0
Failed to open device
modinfo: could not open nvidia_legacy: Permission denied
modinfo: could not open nvidia_new: Permission denied
modinfo: could not open nvidia: Permission denied
modinfo: could not open ipw3945: Permission denied
modinfo: could not open nvidia: Permission denied
magnus@grunion:~$

[Timo Aaltonen (tjaalton)]

The module overrides are in place since 7.10. Closing the bug.