Low graphics mode in muxless hybrid ATI/Intel GPU systems after fglrx upgrade

Bug #1068404 reported by Marco Vuano on 2012-10-19
438
This bug affects 86 people
Affects Status Importance Assigned to Milestone
xf86-video-intel
Won't Fix
Medium
fglrx-installer (Ubuntu)
Undecided
Unassigned
Nominated for Raring by Christopher M. Penalver
Quantal
Undecided
job
fglrx-installer-updates (Ubuntu)
High
job
Nominated for Raring by Christopher M. Penalver
Quantal
High
Unassigned
xserver-xorg-video-intel (openSUSE)
Won't Fix
Critical

Bug Description

[Impact]

 * Hybrid systems with ATI/intel GPUs using fglrx get a segmentation fault in X when using 12.10. In stock 12.04.1 these systems were working fine.

 * This bug represents a severe regression from precise to quantal, since is crashing on startup

 * The proposed patch updates the driver to version 13.2beta3 of upstream which seems to resolve this issue

[Test Case]

 * Get an hybrid ATI/intel system.
   Install quantal: The system should boot fine
   Install the proprietary fglrx package and reboot: The X server crashes on startup

[Regression Potential]

 * This solution does not need any change on the intel driver (as a previous workaround needed).
 * No regressions have been reported so far using this most recent fglrx upstream, by the people affected by this bug

======

After installing the latest fglrx-updates package on Ubuntu 12.10 (fglrx-updates_9.000-0ubuntu3_amd64) I have the same issue that was present before AMD fixed switchable graphics on PowerXpress 4.0 cards (such as my HD 6470M coupled with the intel HD Graphics 3000 of my intel Core i5-2430M). The X server has a segmentation fault and the low graphics mode is activated.

WORKAROUND: fglrx-installer-13 2:13.101~beta-0ubuntu1~xedgers~quantal1 from:
https://launchpad.net/~xorg-edgers/+archive/ppa?field.series_filter=quantal

WORKAROUND: Follow carefully the instructions here:
https://help.ubuntu.com/community/BinaryDriverHowto/ATI#WORKAROUND

As per Nick Andrik, this is a version regression of xserver-xorg-video-intel from 2.20.2 (working and in below mentioned PPA) and 2.20.3. The regression commit is http://cgit.freedesktop.org/xorg/driver/xf86-video-intel/commit/?id=05dcc5f1699ba90fc14c50882e8d4be89bc4a4f9 .

Upstream fglrx has published a new version (> 13.1 ) of the driver which seems to solve this issue. No need to revert the commit in the Intel driver any more.

ATTENTION:
1. If you choose the integrated GPU (via Catalyst or via amdconfig --px-igpu) and your X server crashes on startup, then your are probably affected by this bug:
https://bugs.launchpad.net/ubuntu/+source/fglrx-installer/+bug/1088220

2. If your system includes an old ATI GPU card (Radeon HD 2x00 3x00 4x00) which is not supported any more by fglrx drivers, then probably you suffer from this bug:
https://bugs.launchpad.net/ubuntu/+source/fglrx-installer/+bug/1058040
Subscribe and comment there please.

If the X server loads (so you see no low graphics mode window) but you get no menu bars, window decorations, etc, then press Ctrl+Shift+T to open a terminal, write in there:
unity --replace &
and post a comment with the result.

The logs from the intial bug report follow:

Here is the Xorg log:
[ 20.924]
X.Org X Server 1.13.0
Release Date: 2012-09-05
[ 20.924] X Protocol Version 11, Revision 0
[ 20.924] Build Operating System: Linux 3.2.0-30-generic x86_64 Ubuntu
[ 20.924] Current Operating System: Linux marco-HP-Pavilion-g6-Notebook-PC 3.5.0-17-generic #28-Ubuntu SMP Tue Oct 9 19:31:23 UTC 2012 x86_64
[ 20.924] Kernel command line: BOOT_IMAGE=/boot/vmlinuz-3.5.0-17-generic root=UUID=5374f779-2ae8-4842-a390-e21317364c6c ro quiet splash vt.handoff=7
[ 20.924] Build Date: 08 October 2012 03:34:01PM
[ 20.924] xorg-server 2:1.13.0-0ubuntu6 (For technical support please see http://www.ubuntu.com/support)
[ 20.924] Current version of pixman: 0.26.0
[ 20.924] Before reporting problems, check http://wiki.x.org
 to make sure that you have the latest version.
[ 20.924] Markers: (--) probed, (**) from config file, (==) default setting,
 (++) from command line, (!!) notice, (II) informational,
 (WW) warning, (EE) error, (NI) not implemented, (??) unknown.
[ 20.924] (==) Log file: "/var/log/Xorg.0.log", Time: Fri Oct 19 02:42:21 2012
[ 20.924] (==) Using config file: "/etc/X11/xorg.conf"
[ 20.924] (==) Using system config directory "/usr/share/X11/xorg.conf.d"
[ 20.924] (==) ServerLayout "aticonfig Layout"
[ 20.924] (**) |-->Screen "aticonfig-Screen[0]-0" (0)
[ 20.924] (**) | |-->Monitor "aticonfig-Monitor[0]-0"
[ 20.925] (**) | |-->Device "aticonfig-Device[0]-0"
[ 20.925] (==) Automatically adding devices
[ 20.925] (==) Automatically enabling devices
[ 20.925] (==) Automatically adding GPU devices
[ 20.925] (WW) The directory "/usr/share/fonts/X11/cyrillic" does not exist.
[ 20.925] Entry deleted from font path.
[ 20.925] (WW) The directory "/usr/share/fonts/X11/100dpi/" does not exist.
[ 20.925] Entry deleted from font path.
[ 20.925] (WW) The directory "/usr/share/fonts/X11/75dpi/" does not exist.
[ 20.925] Entry deleted from font path.
[ 20.925] (WW) The directory "/usr/share/fonts/X11/100dpi" does not exist.
[ 20.925] Entry deleted from font path.
[ 20.925] (WW) The directory "/usr/share/fonts/X11/75dpi" does not exist.
[ 20.925] Entry deleted from font path.
[ 20.925] (==) FontPath set to:
 /usr/share/fonts/X11/misc,
 /usr/share/fonts/X11/Type1,
 /var/lib/defoma/x-ttcidfont-conf.d/dirs/TrueType,
 built-ins
[ 20.925] (==) ModulePath set to "/usr/lib/x86_64-linux-gnu/xorg/extra-modules,/usr/lib/xorg/extra-modules,/usr/lib/xorg/modules"
[ 20.925] (II) The server relies on udev to provide the list of input devices.
 If no devices become available, reconfigure udev or disable AutoAddDevices.
[ 20.925] (II) Loader magic: 0x7fed9d5d2c40
[ 20.925] (II) Module ABI versions:
[ 20.925] X.Org ANSI C Emulation: 0.4
[ 20.925] X.Org Video Driver: 13.0
[ 20.925] X.Org XInput driver : 18.0
[ 20.925] X.Org Server Extension : 7.0
[ 20.925] (II) config/udev: Adding drm device (/dev/dri/card0)
[ 20.926] (--) PCI:*(0:0:2:0) 8086:0116:103c:1670 rev 9, Mem @ 0xc0000000/4194304, 0xb0000000/268435456, I/O @ 0x00005000/64
[ 20.926] (--) PCI: (0:1:0:0) 1002:6760:103c:1670 rev 0, Mem @ 0xa0000000/268435456, 0xc2600000/131072, I/O @ 0x00004000/256, BIOS @ 0x????????/131072
[ 20.926] (II) Open ACPI successful (/var/run/acpid.socket)
[ 20.926] Initializing built-in extension Generic Event Extension
[ 20.926] Initializing built-in extension SHAPE
[ 20.926] Initializing built-in extension MIT-SHM
[ 20.926] Initializing built-in extension XInputExtension
[ 20.926] Initializing built-in extension XTEST
[ 20.926] Initializing built-in extension BIG-REQUESTS
[ 20.926] Initializing built-in extension SYNC
[ 20.926] Initializing built-in extension XKEYBOARD
[ 20.926] Initializing built-in extension XC-MISC
[ 20.926] Initializing built-in extension SECURITY
[ 20.926] Initializing built-in extension XINERAMA
[ 20.926] Initializing built-in extension XFIXES
[ 20.926] Initializing built-in extension RENDER
[ 20.926] Initializing built-in extension RANDR
[ 20.926] Initializing built-in extension COMPOSITE
[ 20.926] Initializing built-in extension DAMAGE
[ 20.926] Initializing built-in extension MIT-SCREEN-SAVER
[ 20.926] Initializing built-in extension DOUBLE-BUFFER
[ 20.926] Initializing built-in extension RECORD
[ 20.926] Initializing built-in extension DPMS
[ 20.926] Initializing built-in extension X-Resource
[ 20.926] Initializing built-in extension XVideo
[ 20.926] Initializing built-in extension XVideo-MotionCompensation
[ 20.926] Initializing built-in extension XFree86-VidModeExtension
[ 20.926] Initializing built-in extension XFree86-DGA
[ 20.926] Initializing built-in extension XFree86-DRI
[ 20.926] Initializing built-in extension DRI2
[ 20.926] (II) "glx" will be loaded by default.
[ 20.926] (II) LoadModule: "glx"
[ 21.055] (II) Loading /usr/lib/x86_64-linux-gnu/xorg/extra-modules/extra-modules.dpkg-tmp/modules/extensions/libglx.so
[ 21.055] (II) Module glx: vendor="Advanced Micro Devices, Inc."
[ 21.055] compiled for 6.9.0, module version = 1.0.0
[ 21.056] Loading extension GLX
[ 21.056] (II) LoadModule: "fglrx"
[ 21.057] (II) Loading /usr/lib/x86_64-linux-gnu/xorg/extra-modules/extra-modules.dpkg-tmp/modules/drivers/fglrx_drv.so
[ 21.562] (II) Module fglrx: vendor="FireGL - AMD Technologies Inc."
[ 21.562] compiled for 1.4.99.906, module version = 9.0.2
[ 21.562] Module class: X.Org Video Driver
[ 21.563] (II) Loading sub module "fglrxdrm"
[ 21.563] (II) LoadModule: "fglrxdrm"
[ 21.563] (II) Loading /usr/lib/x86_64-linux-gnu/xorg/extra-modules/extra-modules.dpkg-tmp/modules/linux/libfglrxdrm.so
[ 21.581] (II) Module fglrxdrm: vendor="FireGL - AMD Technologies Inc."
[ 21.581] compiled for 1.4.99.906, module version = 9.0.2
[ 21.581] (II) AMD Proprietary Linux Driver Version Identifier:9.00.2
[ 21.581] (II) AMD Proprietary Linux Driver Release Identifier: 9.00.11
[ 21.581] (II) AMD Proprietary Linux Driver Build Date: Sep 20 2012 11:56:16
[ 21.581] (++) using VT number 7

[ 21.581] (WW) Falling back to old probe method for fglrx
[ 21.624] (II) Loading PCS database from /etc/ati/amdpcsdb /etc/ati/amdpcsdb.default
[ 21.639] ukiDynamicMajor: found major device number 251
[ 21.639] ukiDynamicMajor: found major device number 251
[ 21.639] ukiOpenByBusid: Searching for BusID PCI:1:0:0
[ 21.639] ukiOpenDevice: node name is /dev/ati/card0
[ 21.639] ukiOpenDevice: open result is 10, (OK)
[ 22.206] ukiOpenByBusid: ukiOpenMinor returns 10
[ 22.206] ukiOpenByBusid: ukiGetBusid reports PCI:1:0:0
[ 22.234] (--) Chipset Supported AMD Graphics Processor (0x6760) found
[ 22.246] (II) fglrx: intel VGA device detected, load intel driver.
[ 22.247] (II) LoadModule: "intel"
[ 22.263] (II) Loading /usr/lib/xorg/modules/drivers/intel_drv.so
[ 22.263] (II) Module intel: vendor="X.Org Foundation"
[ 22.263] compiled for 1.13.0, module version = 2.20.9
[ 22.263] Module class: X.Org Video Driver
[ 22.263] ABI class: X.Org Video Driver, version 13.0
[ 22.264] (II) AMD Video driver is running on a device belonging to a group targeted for this release
[ 22.276] (II) AMD Video driver is signed
[ 22.276] (II) fglrx(0): pEnt->device->identifier=0x7fed9df19450
[ 22.276] (II) intel(1): pEnt->device->identifier=(nil)
[ 22.276] (EE) Screen 1 deleted because of no matching config section.
[ 22.276] (EE)
[ 22.276] (EE) Backtrace:
[ 22.277] (EE) 0: /usr/bin/X (xorg_backtrace+0x36) [0x7fed9d35dac6]
[ 22.277] (EE) 1: /usr/bin/X (0x7fed9d1b5000+0x1ac8f9) [0x7fed9d3618f9]
[ 22.277] (EE) 2: /lib/x86_64-linux-gnu/libpthread.so.0 (0x7fed9c4db000+0xfcb0) [0x7fed9c4eacb0]
[ 22.277] (EE) 3: /usr/lib/xorg/modules/drivers/intel_drv.so (0x7fed9903c000+0x15080) [0x7fed99051080]
[ 22.277] (EE) 4: /usr/lib/xorg/modules/drivers/intel_drv.so (0x7fed9903c000+0x16365) [0x7fed99052365]
[ 22.277] (EE) 5: /usr/bin/X (xf86DeleteScreen+0x84) [0x7fed9d2538b4]
[ 22.277] (EE) 6: /usr/bin/X (xf86BusConfig+0x216) [0x7fed9d23eb16]
[ 22.277] (EE) 7: /usr/bin/X (InitOutput+0x9ea) [0x7fed9d24c9da]
[ 22.277] (EE) 8: /usr/bin/X (0x7fed9d1b5000+0x44386) [0x7fed9d1f9386]
[ 22.277] (EE) 9: /lib/x86_64-linux-gnu/libc.so.6 (__libc_start_main+0xed) [0x7fed9b16076d]
[ 22.277] (EE) 10: /usr/bin/X (0x7fed9d1b5000+0x448ad) [0x7fed9d1f98ad]
[ 22.277] (EE)
[ 22.277] (EE) Segmentation fault at address 0x0
[ 22.277]
Fatal server error:
[ 22.277] Caught signal 11 (Segmentation fault). Server aborting
[ 22.277]
[ 22.277] (EE)
Please consult the The X.Org Foundation support
  at http://wiki.x.org
 for help.
[ 22.277] (EE) Please also check the log file at "/var/log/Xorg.0.log" for additional information.
[ 22.277] (EE)
[ 22.295] Server terminated with error (1). Closing log file.

The system is a Hp g6-1279sl notebook with a discrete AMD HD 6470M card and an Intel Core i5-2430M CPU with Intel HD graphics 3000. I'm running Ubuntu 12.10 desktop 64 bit version. The driver (both the 12.9 beta on AMD site and the 12-9 beta driver from the Quantal Quetzal fglrx-installer-updates repository) worked correctly on Ubuntu 12.4.1 64-bit.
Using fglrx (fglrx_9.000-0ubuntu3_amd64.deb) instead of fglrx-updates doesn't solve this problem.

Potential duplicate https://bugs.launchpad.net/ubuntu/+source/fglrx-installer-updates/+bug/1069199 .

---
.tmp.unity.support.test.0:

ApportVersion: 2.0.1-0ubuntu12
Architecture: amd64
CasperVersion: 1.315
CompizPlugins: [core,composite,opengl,compiztoolbox,decor,vpswitch,snap,mousepoll,resize,place,move,wall,grid,regex,imgpng,session,gnomecompat,animation,fade,unitymtgrabhandles,workarounds,scale,expo,ezoom,unityshell]
CompositorRunning: compiz
DistUpgraded: Fresh install
DistroCodename: precise
DistroRelease: Ubuntu 12.04
DistroVariant: ubuntu
GraphicsCard:
 Intel Corporation 2nd Generation Core Processor Family Integrated Graphics Controller [8086:0116] (rev 09) (prog-if 00 [VGA controller])
   Subsystem: Hewlett-Packard Company Device [103c:1670]
 Advanced Micro Devices [AMD] nee ATI Seymour [Radeon HD 6400M Series] [1002:6760] (prog-if 00 [VGA controller])
   Subsystem: Hewlett-Packard Company Device [103c:1670]
LiveMediaBuild: Ubuntu 12.04.1 LTS "Precise Pangolin" - Release amd64 (20120823.1)
MachineType: Hewlett-Packard HP Pavilion g6 Notebook PC
Package: fglrx-installer (not installed)
ProcEnviron:
 TERM=xterm
 PATH=(custom, no user)
 LANG=en_US.UTF-8
 SHELL=/bin/bash
ProcKernelCmdLine: bootkbd=it console-setup/layoutcode=it console-setup/variantcode=nodeadkeys locale=it_IT noprompt cdrom-detect/try-usb=true file=/cdrom/preseed/username.seed boot=casper initrd=/casper/initrd.lz splash -- maybe-ubiquity
ProcVersionSignature: Ubuntu 3.2.0-29.46-generic 3.2.24
Tags: precise running-unity ubuntu referred-by-support compiz-0.9
Uname: Linux 3.2.0-29-generic x86_64
UpgradeStatus: No upgrade log present (probably fresh install)
UserGroups: adm cdrom dip lpadmin plugdev sambashare sudo
dmi.bios.date: 03/22/2012
dmi.bios.vendor: Insyde
dmi.bios.version: F.63
dmi.board.asset.tag: Base Board Asset Tag
dmi.board.name: 1670
dmi.board.vendor: Hewlett-Packard
dmi.board.version: 09.4B
dmi.chassis.type: 10
dmi.chassis.vendor: Hewlett-Packard
dmi.chassis.version: Chassis Version
dmi.modalias: dmi:bvnInsyde:bvrF.63:bd03/22/2012:svnHewlett-Packard:pnHPPaviliong6NotebookPC:pvr0691120000204610000620100:rvnHewlett-Packard:rn1670:rvr09.4B:cvnHewlett-Packard:ct10:cvrChassisVersion:
dmi.product.name: HP Pavilion g6 Notebook PC
dmi.product.version: 0691120000204610000620100
dmi.sys.vendor: Hewlett-Packard
version.compiz: compiz 1:0.9.7.8-0ubuntu1.4
version.ia32-libs: ia32-libs N/A
version.libdrm2: libdrm2 2.4.32-1ubuntu1
version.libgl1-mesa-dri: libgl1-mesa-dri 8.0.3+8.0.2-0ubuntu3.2
version.libgl1-mesa-dri-experimental: libgl1-mesa-dri-experimental N/A
version.libgl1-mesa-glx: libgl1-mesa-glx 8.0.3+8.0.2-0ubuntu3.2
version.xserver-xorg-core: xserver-xorg-core 2:1.11.4-0ubuntu10.7
version.xserver-xorg-input-evdev: xserver-xorg-input-evdev 1:2.7.0-0ubuntu1.2
version.xserver-xorg-video-ati: xserver-xorg-video-ati 1:6.14.99~git20111219.aacbd629-0ubuntu2
version.xserver-xorg-video-intel: xserver-xorg-video-intel 2:2.17.0-1ubuntu4.1
version.xserver-xorg-video-nouveau: xserver-xorg-video-nouveau 1:0.0.16+git20111201+b5534a1-1build2

Related branches

User-Agent: Mozilla/5.0 (X11; Linux x86_64; rv:15.0) Gecko/20100101 Firefox/15.0

I have an switchable graphic system with AMD and Intel cards.
Everything worked fine until ad update of the Intel Video driver, that cause the segmentation fault in X.
The old driver before that update was xf86-video-driver-2.20.0, the updated driver provided by the Update repo is 2.20.3

Reproducible: Always

Steps to Reproduce:
1.Update system normaly from repo-udate
2.install openSUSE-2012-559 update
3.reboot system
Actual Results:
System start in console mode, X segmentation fault

Downgrade the driver using zypper in -f xf86-video-intel-2.20.0-1.5.1 fix the problem.

*** Bug 780581 has been marked as a duplicate of this bug. ***

Marco Vuano (vuanomarco) on 2012-10-19
summary: - Switchable graphics on 6470M (Powerxpress 4.0) not working again
+ Switchable graphics on 6XXXM (PowerXpress 4.0/Dynamic Switchable
+ Graphics) not working again
Marco Vuano (vuanomarco) on 2012-10-19
description: updated

I'm adding some information to this report (obtained using fglrx
2:9.000-0ubuntu3/fglrx_9.000-0ubuntu3_amd64.deb):
dmesg output(in the dmesgfglrx file)
Xorg log (in the Xorglogfglrx file)
lshw output (in the lshwfglrx file)
uname -a output (in the unamefglrx file)
fglrx package details given by dpkg (in the fglrxpackage file)
The xorg.conf used (the one generated by amdconfig --initial, in the
xorgconffglrx file)

description: updated

Status changed to 'Confirmed' because the bug affects multiple users.

Changed in fglrx-installer (Ubuntu):
status: New → Confirmed
Changed in fglrx-installer-updates (Ubuntu):
status: New → Confirmed
Gareth (gmccumskey) wrote :

I can confirm the same issue is happening here. Obviously trying to use the 8.x drivers would not help because they do not support the new version of xorg.

Marco Vuano, thank you for taking the time to report this bug and helping to make Ubuntu better. Please execute the following command, as it will automatically gather debugging information, in a terminal:
apport-collect -p fglrx-installer 1068404
When reporting bugs in the future please use apport by using 'ubuntu-bug' and the name of the package affected. You can learn more about this functionality at https://wiki.ubuntu.com/ReportingBugs.

Changed in fglrx-installer (Ubuntu):
status: Confirmed → Incomplete

ApportVersion: 2.6.1-0ubuntu3
Architecture: amd64
DistroCodename: quantal
DistroRelease: Ubuntu 12.10
DistroVariant: ubuntu
InstallationMedia: Ubuntu 12.04 LTS "Precise Pangolin" - Release amd64 (20120425)
Package: fglrx-installer (not installed)
ProcVersionSignature: Ubuntu 3.5.0-17.28-generic 3.5.5
Tags: quantal running-unity ubuntu
Uname: Linux 3.5.0-17-generic x86_64
UnreportableReason: Please work this issue through technical support channels first.
UpgradeStatus: Upgraded to quantal on 2012-10-19 (3 days ago)
UserGroups: adm cdrom dip lpadmin plugdev sambashare sudo

tags: added: apport-collected quantal running-unity ubuntu

apport information

Gareth, please do not apport-collect to another person's report. As well, if you have a bug in Ubuntu, could you please file a new report by executing the following in a terminal:
ubuntu-bug fglrx-installer

For more on this, please see the Ubuntu Bug Control and Ubuntu Bug Squad article:
https://wiki.ubuntu.com/Bugs/BestPractices#X.2BAC8-Reporting.Focus_on_One_Issue

and Ubuntu Community article:
https://help.ubuntu.com/community/ReportingBugs

When opening up the new report, please feel free to subscribe me to it. Thank you for your understanding.

tags: removed: apport-collected quantal running-unity ubuntu
Download full text (12.0 KiB)

Sorry about that. Saw the previous email and thought I was helping out. I
am having the exact same issue as this bug reports, however, so creating a
new bug would be duplicitous.

On Mon, Oct 22, 2012 at 3:44 PM, Christopher M. Penalver <
<email address hidden>> wrote:

> Gareth, please do not apport-collect to another person's report. As well,
> if you have a bug in Ubuntu, could you please file a new report by
> executing the following in a terminal:
> ubuntu-bug fglrx-installer
>
> For more on this, please see the Ubuntu Bug Control and Ubuntu Bug Squad
> article:
>
> https://wiki.ubuntu.com/Bugs/BestPractices#X.2BAC8-Reporting.Focus_on_One_Issue
>
> and Ubuntu Community article:
> https://help.ubuntu.com/community/ReportingBugs
>
> When opening up the new report, please feel free to subscribe me to it.
> Thank you for your understanding.
>
> ** Attachment removed: "ProcEnviron.txt"
>
> https://bugs.launchpad.net/ubuntu/+source/fglrx-installer-updates/+bug/1068404/+attachment/3408835/+files/ProcEnviron.txt
>
> ** Tags removed: apport-collected quantal running-unity ubuntu
>
> --
> You received this bug notification because you are subscribed to the bug
> report.
> https://bugs.launchpad.net/bugs/1068404
>
> Title:
> Switchable graphics on 6XXXM (PowerXpress 4.0/Dynamic Switchable
> Graphics) not working again
>
> Status in “fglrx-installer” package in Ubuntu:
> Incomplete
> Status in “fglrx-installer-updates” package in Ubuntu:
> Confirmed
>
> Bug description:
> After installing the latest fglrx-updates package on Ubuntu 12.10
> (fglrx-updates_9.000-0ubuntu3_amd64) I have the same issue that was
> present before AMD fixed switchable graphics on PowerXpress 4.0 cards
> (such as my HD 6470M coupled with the intel HD Graphics 3000 of my
> intel Core i5-2430M). The X server has a segmentation fault and the
> low graphics mode is activated. Here is the Xorg log:
>
> [ 20.924]
> X.Org X Server 1.13.0
> Release Date: 2012-09-05
> [ 20.924] X Protocol Version 11, Revision 0
> [ 20.924] Build Operating System: Linux 3.2.0-30-generic x86_64 Ubuntu
> [ 20.924] Current Operating System: Linux
> marco-HP-Pavilion-g6-Notebook-PC 3.5.0-17-generic #28-Ubuntu SMP Tue Oct 9
> 19:31:23 UTC 2012 x86_64
> [ 20.924] Kernel command line:
> BOOT_IMAGE=/boot/vmlinuz-3.5.0-17-generic
> root=UUID=5374f779-2ae8-4842-a390-e21317364c6c ro quiet splash vt.handoff=7
> [ 20.924] Build Date: 08 October 2012 03:34:01PM
> [ 20.924] xorg-server 2:1.13.0-0ubuntu6 (For technical support please
> see http://www.ubuntu.com/support)
> [ 20.924] Current version of pixman: 0.26.0
> [ 20.924] Before reporting problems, check http://wiki.x.org
> to make sure that you have the latest version.
> [ 20.924] Markers: (--) probed, (**) from config file, (==) default
> setting,
> (++) from command line, (!!) notice, (II) informational,
> (WW) warning, (EE) error, (NI) not implemented, (??) unknown.
> [ 20.924] (==) Log file: "/var/log/Xorg.0.log", Time: Fri Oct 19
> 02:42:21 2012
> [ 20.924] (==) Using config file: "/etc/X11/xorg.conf"
> [ 20.924] (==) Using system config directory
> "/u...

Gareth, regarding your comments https://bugs.launchpad.net/ubuntu/+source/fglrx-installer-updates/+bug/1068404/comments/9 :
>"Sorry about that. Saw the previous email and thought I was helping out."

No problem. :)

>"I am having the exact same issue as this bug reports..."

One would not want to assume this. Xorg issues can be hardware dependent. For example, similar backtraces on different hardware, but different root causes. Hence, this report is only focusing on Marco Vauno's xorg backtrace and hardware vendor ID until a developer notes otherwise. As well, the link https://wiki.ubuntu.com/Bugs/BestPractices#X.2BAC8-Reporting.Focus_on_One_Issue is from the Ubuntu X Team's (maintainers of fglrx-installer) official debugging page. So, if you do not want to follow their directions which advise you what to do, as requested of you in https://bugs.launchpad.net/ubuntu/+source/fglrx-installer-updates/+bug/1068404/comments/8 , then your issue may not get addressed.

>"so creating a new bug would be duplicitous."

See above.

Thank you for your understanding.

Marco Vuano (vuanomarco) wrote :

When running
apport-collect -p fglrx-installer 1068404
I get the following output:

dpkg-query: nessun pacchetto corrispondente a fglrx-installer
Traceback (most recent call last):
  File "/usr/share/apport/apport-gtk", line 588, in <module>
    app.run_argv()
  File "/usr/lib/python2.7/dist-packages/apport/ui.py", line 639, in run_argv
    return self.run_update_report()
  File "/usr/lib/python2.7/dist-packages/apport/ui.py", line 559, in run_update_report
    attachment_comment='apport information')
  File "/usr/lib/python2.7/dist-packages/apport/crashdb_impl/launchpad.py", line 358, in update
    report.write_mime(mime, skip_keys=skip_keys)
  File "/usr/lib/python2.7/dist-packages/problem_report.py", line 504, in write_mime
    attach_value = CompressedValue(v, k).gzipvalue
  File "/usr/lib/python2.7/dist-packages/problem_report.py", line 44, in __init__
    self.set_value(value)
  File "/usr/lib/python2.7/dist-packages/problem_report.py", line 50, in set_value
    gzip.GzipFile(self.name, mode='wb', fileobj=out).write(value)
  File "/usr/lib/python2.7/gzip.py", line 131, in __init__
    self._write_gzip_header()
  File "/usr/lib/python2.7/gzip.py", line 176, in _write_gzip_header
    self.fileobj.write(fname + '\000')
TypeError: 'unicode' does not have the buffer interface

Can you confirm if apport-collect is working?

Marco Vuano, could you please boot into a supported Ubuntu release live environment and then execute:
apport-collect -p fglrx-installer 1068404

apport information

tags: added: apport-collected compiz-0.9 precise referred-by-support running-unity ubuntu
description: updated

apport information

apport information

apport information

apport information

apport information

apport information

apport information

apport information

apport information

apport information

apport information

apport information

apport information

apport information

apport information

apport information

apport information

These apport information was generated with Ubuntu 12.04.1 LTS Desktop version for 64 bit architecture

summary: - Switchable graphics on 6XXXM (PowerXpress 4.0/Dynamic Switchable
- Graphics) not working again
+ 1002:6760 Switchable graphics on Radeon HD 6400M not working again
Changed in fglrx-installer (Ubuntu):
status: Incomplete → New

Status changed to 'Confirmed' because the bug affects multiple users.

Changed in fglrx-installer (Ubuntu):
status: New → Confirmed
Nick Andrik (andrikos) wrote :

Take a look here, probably it is the same issue:
http://forums.opensuse.org/english/get-technical-help-here/hardware/478053-switchable-graphics-card-ati-intel-failed-boot-after-install-ati-drivers.html

I will try to prepare a package of an older version for the intel driver and see what I get

Nick Andrik (andrikos) wrote :

I confirm that the problem is in xserver-xorg-video-intel

I have built and used the 2.20.0 version in my PPA and the problem is gone

https://launchpad.net/~andrikos/+archive/ppa/+packages?field.name_filter=&field.status_filter=published&field.series_filter=quantal

Changed in xserver-xorg-video-intel (Ubuntu):
status: New → Incomplete
Changed in xserver-xorg-video-intel (openSUSE):
importance: Unknown → Critical
status: Unknown → In Progress

I can confirm that the xserver-xorg-video-intel_2.20.0-0~andrik1_amd64.deb from Nick Andrick's PPa is enough to solve the issue with the fglrx-updates_9.000-0ubuntu3_amd64 from the official quantal/restricted repository. The driver is working correctly now, but the discrete GPU has to be used in order to use Unity 3D (the Intel driver has problem with 3d acceleration when the fglrx driver is installed, but this another bug not strictly related to this one).

description: updated
Changed in xserver-xorg-video-intel (Ubuntu):
status: Incomplete → New
Marco Vuano (vuanomarco) wrote :

The xserver-xorg-video-intel_2.20.0-0~andrik1_amd64.deb package is also able to solve the issue with the fglrx drivers installed with the fglrx_9.000-0ubuntu3_amd64.deb package (again, Unity works only with the discrete GPU).
Thank you very much, Nick and Christopher!

Launchpad Janitor (janitor) wrote :

Status changed to 'Confirmed' because the bug affects multiple users.

Changed in xserver-xorg-video-intel (Ubuntu):
status: New → Confirmed
Download full text (4.0 KiB)

Created attachment 69191
Xorg log when unsupported muxless platform is reported

Bug already reported to AMD, but since it seems to involve Intel driver, I thought it might be a good idea to report here as well.

Actually two similar bugs have been reported to AMD :
http://ati.cchtml.com/show_bug.cgi?id=634 (reported by me)
http://ati.cchtml.com/show_bug.cgi?id=580

I am the owner of an ACER Aspire TimelineX 3820 TG featuring a core i5-480M +
Radeon HD6550M.
Configuration (32 bits ONLY installation) :
- kernel 3.4.14
- glibc 2.14.1
- gcc 4.6.1
- Xorg 7.7 with xorg-server 1.12.1 / xf86-video-intel 2.20.10
- Slim 1.3.2
- fglrx 12.10 or 12.11 beta

X works fine with :
- the dGPU only ("discrete" selected in the bios)
- the IGP only ("switchable" selected in the bios and dGPU NOT declared in the
ServerLayout section).

X crashes at startup when :
- "switchable graphics" is selected in the bios
- only the dGPU is declared in the ServerLayout :
excerpt of Xorg.0.log
... snip ...
[ 20.778] (==) ServerLayout "laptop"
[ 20.778] (**) |-->Screen "high performance" (1)
[ 20.778] (**) | |-->Monitor "Acer 3820TG"
[ 20.778] (**) | |-->Device "Radeon HD 6550M"
... snip ...
[ 20.780] (--) PCI:*(0:0:2:0) 8086:0046:1025:051b rev 24, Mem @
0xf0000000/4194304, 0xd0000000/268435456, I/O @ 0x00001800/8
[ 20.780] (--) PCI: (0:2:0:0) 1002:68c1:1025:051b rev 0, Mem @
0xb0000000/268435456, 0xafee0000/131072, I/O @ 0x00002000/256, BIOS @
0x????????/131072
... snip ...
[ 20.865] (--) Chipset Supported AMD Graphics Processor (0x68C1) found
[ 20.866] (WW) fglrx: No matching Device section for instance (BusID
PCI:0@2:0:1) found
[ 20.866] (II) fglrx: intel VGA device detected, load intel driver.
[ 20.866] (II) LoadModule: "intel"
[ 20.866] (II) Loading /opt/X11/lib/xorg/modules/drivers/intel_drv.so
[ 20.866] (II) Module intel: vendor="X.Org Foundation"
[ 20.866] compiled for 1.12.2, module version = 2.20.10
[ 20.866] Module class: X.Org Video Driver
[ 20.866] ABI class: X.Org Video Driver, version 12.0
[ 20.866] (WW) PowerXpress feature is not supported on A+I Mux platform.
Please uninstall fglrx driver.
[ 20.866] (EE) this is a Muxless PX A+I platform, we doesn't supported it
[ 20.866] (EE) No devices detected.
[ 20.866]
Fatal server error:
[ 20.866] no screens found

I am a bit puzzled by :
"this is a Muxless PX A+I platform, we doesn't supported it"

X crashes at startup when :
- "switchable graphics" is selected in the bios
- both dGPU (Radeon HD 6550M) and iGPU (Core i5 IGP) are declared in the
ServerLayout section :

excerpt of Xorg.0.log
... snip ...
[ 205.098] (==) ServerLayout "laptop"
[ 205.098] (**) |-->Screen "low power" (0)
[ 205.098] (**) | |-->Monitor "Acer 3820TG"
[ 205.098] (**) | |-->Device "Intel HD Graphics"
[ 205.098] (**) |-->Screen "high performance" (1)
[ 205.099] (**) | |-->Monitor "Acer 3820TG"
[ 205.099] (**) | |-->Device "Radeon HD 6550M"
... snip ...
[ 142.387] (--) PCI:*(0:0:2:0) 8086:0046:1025:051b rev 24, Mem @
0xf0000000/4194304, 0xd0000000/268435456, I/O @ 0x00001800/8
[ 142.387] (--) PCI: (0:2:0:0) 1002:68c1:1025:051b rev 0, Mem @
0xb0000000/...

Read more...

Marco Vuano (vuanomarco) wrote :

Regarding the first part of the bug (Unity not working after installing fglrx), this is a known bug (when using the integrated GPU after installing fglrx, direct rendering has problems) which is now more apparent becuse it seems that Unity 2d is no longer present on Ubuntu 12.10. There is a solution for this bug, which is described in step 2 of the first message of this thread: http://ubuntuforums.org/showthread.php?t=1930450
I tried that myself and now Unity works even when using the integrated GPU.
Regarding the second part of the bug (low graphics mode after installing fglrx and running sudo aticonfig --initial), it is probably the same bug as the one I reported (after all, PowerXpress 4.0/dynamic switchable graphics is also used on AMD Radeon HD 6770M).

Nick Andrik (andrikos) wrote :

As I understand there are two problems for ati/intel hybrid systems:

a) "Segmentation fault at address 0x0" when using fglrx driver:
This is a regression with the intel drivers, installing version 2.20.0 is a workaround for now

b) Cannot switch on integrated CPU from Catalyst:
Adding :/usr/lib/`uname -m`-linux-gnu/dri in 4th line of /etc/X11/Xsession.d/10fglrx solves the problem

I'm marking this bug as a duplicate of this:
https://bugs.launchpad.net/ubuntu/+source/fglrx-installer-updates/+bug/1068661
since the symptoms and the workarounds are the same.

Nick Andrik (andrikos) on 2012-10-29
Changed in fglrx-installer-updates (Ubuntu):
assignee: nobody → Nick Andrik (andrikos)
Changed in xserver-xorg-video-intel (Ubuntu):
assignee: nobody → Nick Andrik (andrikos)
description: updated
Nick Andrik (andrikos) wrote :

I have performed some bisecting on the actual xf86-video-intel driver and the commit to blame is this one:
http://cgit.freedesktop.org/xorg/driver/xf86-video-intel/commit/?id=05dcc5f1699ba90fc14c50882e8d4be89bc4a4f9

I have prepared a patch which reverts the change above against the current official package of xserver-xorg-video-intel and the resulting package can be found on my PPA:
https://launchpad.net/~andrikos/+archive/ppa/+packages?field.name_filter=&field.status_filter=published&field.series_filter=quantal

I also attach the aforementioned patch

description: updated

Marco Vuano, could you please test Nick's new PPA release and comment on if the issue is resolved via https://launchpad.net/~andrikos/+archive/ppa/+packages?field.name_filter=&field.status_filter=published&field.series_filter=quantal ?

Changed in xserver-xorg-video-intel (Ubuntu):
status: Confirmed → Incomplete
Nick Andrik (andrikos) on 2012-10-30
Changed in fglrx-installer (Ubuntu):
assignee: nobody → Nick Andrik (andrikos)

The attachment "Fixes regression in ati/intel hybrid systems" of this bug report has been identified as being a patch. The ubuntu-reviewers team has been subscribed to the bug report so that they can review the patch. In the event that this is in fact not a patch you can resolve this situation by removing the tag 'patch' from the bug report and editing the attachment so that it is not flagged as a patch. Additionally, if you are member of the ubuntu-reviewers team please also unsubscribe the team from this bug report.

[This is an automated message performed by a Launchpad user owned by Brian Murray. Please contact him regarding any issues with the action taken in this bug report.]

tags: added: patch

There is nothing I can do about this since fglrx plays tricks with the intel hardware without knowledge of the driver.

I can confirm that with the package xserver-xorg-video-intel_2.20.9-0ubuntu2+andrik2_amd64.deb both the fglrx drivers from fglrx and fglrx-updates packages from official quantal/restricted repository work correctly.

Gareth (gmccumskey) wrote :

I was wondering if there is any indication when this fix would hit the live Quantal repo's as at the moment I am running on tri-boot with Precise, Quantal and Windows until the graphics issues are fixed. The software rendering that Quantal has to resort to pushes my CPU temparature way too high to use confidently.

Nick Andrik (andrikos) wrote :

@Gareth
Now that the packages are confirmed to resolve the issue, we will contact the maintainers in order to accept them for the official repos

Nick Andrik (andrikos) on 2012-10-30
Changed in fglrx-installer (Ubuntu):
status: Confirmed → In Progress
Changed in fglrx-installer-updates (Ubuntu):
status: Confirmed → In Progress
Changed in xserver-xorg-video-intel (Ubuntu):
status: Incomplete → In Progress

I thought so... But wanted to report anyway.
I added a comment to my bug report in AMD's bugzilla referencing this one.
'hope this will help.

Thanks a lot for the quality of your driver.

Chris, can you take a look on the workaround proposed on my comment #44 in this bug report?
https://bugs.launchpad.net/ubuntu/+source/fglrx-installer-updates/+bug/1068404

If you revert (or "debug") the mentioned patch then the driver works for intel/ati systems.

Nick Andrik (andrikos) wrote :

I also attach the patch for the fglrx driver (already included in my PPA) that allows to switch to the igpu using the amdccle

Nick Andrik (andrikos) on 2012-11-01
tags: added: quantal
removed: precise

This seems to be a regression of the xf86-vidoe-intel driver.

You can take a look on comment #44 on this bug report in ubuntu for a patch to revert the commit that produces this problem:

https://bugs.launchpad.net/ubuntu/+source/fglrx-installer-updates/+bug/1068404

no longer affects: fglrx
Changed in xserver-xorg-video-intel:
importance: Unknown → Medium
status: Unknown → Won't Fix
bugbot (bugbot) on 2012-11-01
tags: added: precise
description: updated
Nick Andrik (andrikos) wrote :

May I suggest to update the title of the bug to be more generic since this affects all ati/intel hybrid systems, not just the system of the original poster?

Nick Andrik (andrikos) wrote :

I subscribed ubuntu-sponsors so that someone can review and eventually sponsor my updated package

Sam Marley (sam-spreull) wrote :

Thanks Nick

This had been driving me mad, tried Ubuntu 12.04 & 12.10, Mint 13 & 14, lmde, fedora 17 and 18 the intel driver always caused a segmentation fault. Could not get fglrx drivers to work with 7700M until installing xserver-xorg-video-intel_2.20.9-0ubuntu2+andrik2_amd64.deb

Timo Aaltonen (tjaalton) wrote :

this is a bug in fglrx, we're not going to work around issues in it by reverting stuff from the open drivers, and not knowing what side-effects they might bring..

Changed in xserver-xorg-video-intel (Ubuntu):
status: In Progress → Invalid
Nick Andrik (andrikos) wrote :

As I wrote in the merge process:

By using the mentioned patch we enable people to use hybrid intel/fglrx systems with no (at least from my point of view, correct me if I am wrong) disadvantage at all.

There are two official packages in the repositories, i.e. xserver-xorg-video-intel and fglrx(-updates), that make the intel/ATI hybrid systems completely unusable when the two of them are used together.

This issue has to be addressed one why or another:
- My proposal is to revert that specific xorg-intel commit (and use the patch at least in ubuntu, if upstream does not want to include it). If you think reverting the whole change is too intrusive, we can discuss that with upstream, bisect the actual commit and revert just the conflicting (small) change.

- If we decide not to do that, we should include a "Conflicts:" dependency in the packages, and warn/disallow people to use the fglrx driver when the xserver-xorg-video-intel is also used

In either case, something has to be done. This is a real problem and we cannot just ignore it, let people shoot themselves in the foot and blame ubuntu for fglrx drivers.

Just my 2¢

Nick Andrik (andrikos) wrote :

Actually by "Conflicts:" I mean "Breaks:"

Gareth (gmccumskey) wrote :

I am one of those users and Ubuntu 12.10 is completely unusable to me without a fix to the hybrid graphics that works incredibly well on Ubuntu 12.04. Being a developer myself I understand the fact that things can go wrong beyond your control but the average end user doesn't care whose fault it ACTUALLY is and will just say that Ubuntu is crap because it cannot even handle graphics drivers. Hybrid graphics systems are also exceedingly common. Not solving this is essentially giving the finger to a very large potential user base.

@Timo: I understand the frustration, but please reconsider your position! It's the purpose of any driver to make the system work correctly for the users. In such cases where there are closed-source drivers involved you might want to consider those drivers part of the (hardware) system that needs to work. If the bug affecting the open driver would be in the firmware layer you would also accept work-around solutions, why not at this level?
Currently Ubuntu 12.10 is unusable for hybrid fglrx/intel setups, Nick's approach fixes that. I haven't studied the issue deep enough to estimate if a fix on the fglrx side would be easy. (and if we might expect a fix by them any time soon) In the meantime many users are left out in the cold.

BTW, I asked upstream [1] about this specific commit and what they say
is that it is safe to revert it since it just cleans up some public
structures making the Xorg.log cleaner.
The only side-effect they expect is a bit more log information in
Xorg.log which I believe is hardly any annoyance compared to hybrid
systems not working.

Timo, could you please reconsider the choice of not including the patch?

Thanks,
Nick

[1] https://bugs.freedesktop.org/show_bug.cgi?id=56494#c7

Nick Andrik (andrikos) on 2012-12-07
summary: - 1002:6760 Switchable graphics on Radeon HD 6400M not working again
+ Hybrid ATI/Intel GPU systems do not work after fglrx upgrade

This bug was fixed in the package fglrx-installer-updates -
2:9.000-0ubuntu5

---------------
fglrx-installer-updates (2:9.000-0ubuntu5) raring; urgency=low

  * debian/control:
    - Conflict/replace/provide new nvidia flavours.
    - Depend on other flavours of kernel headers.
  * debian/10fglrx.in:
    - Fix path for powerxpress (LP: #1068661). Thanks
      to Nick Andrik and Lu Xingzhe for suggesting
      the fix.

  [ Benjamin Kerensa ]
  * debian/control:
    - Add ${misc:Depends} depends as needed.
 -- Alberto Milone <email address hidden> Fri, 07 Dec 2012 15:55:20 +0100

description: updated
Changed in fglrx-installer-updates (Ubuntu):
status: In Progress → Fix Released
assignee: Nick Andrik (andrikos) → nobody
description: updated
Nick Andrik (andrikos) on 2012-12-07
Changed in fglrx-installer (Ubuntu):
assignee: Nick Andrik (andrikos) → nobody
summary: - Hybrid ATI/Intel GPU systems do not work after fglrx upgrade
+ Low graphics mode in Hybrid ATI/Intel GPU systems after fglrx upgrade
description: updated
Nick Andrik (andrikos) on 2012-12-07
description: updated

The fix that has been merged into fglrx is still incomplete in regards to 32 bit compatibility on a 64 bit host OS.
If you want to run old games like Counter Strike 1.6 on the Intel GPU, you can't.
That's because of 2 reasons:
1) The path to #OTHER_ARCH# DRI libs is missing (/usr/lib/i386-linux-gnu/dri)
2) When doing:

amdconfig --px-igpu
cat /etc/ld.so.conf.d/x86_64-linux-gnu_GL.conf

it shows that ldconfig should count in /usr/i386-linux-gnu/mesa as a valuable path for libraries, but this path does not exist.
Instead, this path should be used: /usr/lib/i386-linux-gnu/mesa/
Symlinking the wrong path to the right one by hand is a nice hack, but it's a hack.
I'm a little disturbed that nobody ever noticed this mistake. Maybe I'm the only one who likes to play old games on the IGPU.

Alfred, it seems you are right, I will prepare a new patch and propose
it for merging, thanks!

Nikos

Awesome, thanks!
Just for reference, this is the workaround for the DRI path problem that I've done
and been using for 2 months now, seems to be working fine.
This goes straight into /etc/X11/Xsession.d/

https://raw.github.com/beidl/amd-indicator/master/11switchable

Commit 96 in my branch should be enough to fix this:
https://code.launchpad.net/~andrikos/ubuntu/quantal/fglrx-installer/fix-switch-to-igpu

We wait for the commit to be reviewed and possibly merged.

Thanks for reporting it.

Nikos

Nick Andrik (andrikos) wrote :

@Alfred:

I have also prepared the appropriate fgrlx-* packages in my PPA:
https://launchpad.net/~andrikos/+archive/ppa/+packages
Could you please check them and report if they fix your problem?

Thanks,
Nick

Tried it on my machine, didn't work right out of the box.
I tried to launch Steam through Wine, but as soon as the login finished, it crashed.
Running it with LIBGL_DEBUG=verbose, it told me that there were no DRI libs.
So that means that having /usr/lib/i386-linux-gnu/dri in the LIBGL_DRIVERS_PATH is really needed, even on amd64 hosts.
Having the path added to 10fglrx makes Steam and CS 1.6 run flawlessly.

I fixed the mesa path, probably I also need the dri path.
I will check it and update my patch, thanks!

Nikos
--
=Do-
N.AND

Nick Andrik (andrikos) wrote :

Now it should be ready, please check the updated fglrx-* packages in my PPA

Thanks,
Nikos

Hi Nick,
I am also going to test on an Alienware mx17r3, which has an hybrid system Intel/ATI Radeon HD 6900M (dual) card.

Kornel Nagy (snowmanx11) wrote :

It does not work for me. I got this for the fglrxinfo:
X Error of failed request: BadRequest (invalid request code or no such operation)
  Major opcode of failed request: 136 (GLX)
  Minor opcode of failed request: 19 (X_GLXQueryServerString)
  Serial number of failed request: 12
  Current serial number in output stream: 12

@Kornel:
Try these commands please:

LIBGL_DEBUG=verbose fglrxinfo
LIBGL_DEBUG=verbose glxinfo

description: updated

Kornel Nagy, if you have a bug in Ubuntu, could you please file a new report by executing the following in a terminal:
ubuntu-bug xorg

For more on this, please see the Ubuntu X.Org Team (maintainers of fglrx-installer and fglrx-installer-updates), Ubuntu Bug Control, and Ubuntu Bug Squad article:
https://wiki.ubuntu.com/Bugs/BestPractices#X.2BAC8-Reporting.Focus_on_One_Issue

and Ubuntu Community article:
https://help.ubuntu.com/community/ReportingBugs#Bug_reporting_etiquette

When opening up the new report, please feel free to subscribe me to it.

Please note, not filing a new report may delay your problem being addressed as quickly as possible.

Thank you for your understanding.

ok, at this time there is only one way that work on my computer running quantal (Dell 7520 SE, i7-3612QM CPU, AMD Radeon HD 7700M Series) and that is beil's method: http://ubuntuforums.org/showpost.php?p=12386383&postcount=536

Thank you, beidl! It took me three days to find this solution (and I tried all of them in this and other discussions). So until fglrx starts playing nice with the intel driver, this is the best way to have a working accelerated Xorg server.

You should also have a look at his amd-indicator: https://github.com/beidl/amd-indicator. great stuff.

cheers, I thought I'd never get it to work (tried on Fedora for months...)

Alfred E. Neumayer (beidl) wrote :

Thank you for the kind words! I'm glad it's working well for you (and that people are actually using my stuff). :)

@Nick, I won't have much time to test the new fglrx packages from your PPA, I'm sorry.
But by just judging the diff, it should be reflecting exactly what I've done in my Xsession script (where it's been working fine),
mostly because simply specifying another DRI path for the Intel drivers should not affect fglrx in any way.
Maybe in a few days I'll have some time, but school is a little more important right now,
much more than dealing with breaking a working (and needed) system.

Marco Vuano (vuanomarco) wrote :

@Nick: I noticed that Freedesktop's bug 56494 affects a system based on AMD Radeon HD 6550M which uses PowerXpress 3.0/Switchable graphics and not PowerXpress 4.0/Dynamic Switchable graphics. It is likely that the bug is not the same.
To clarify the version of PowerXpress supported by the various supported AMD cards, I'll post this short list based on information on AMD site (names are based on information available on AMD site and on the slides which can be viewed on http://www.slideshare.net/AMD/amd-enduro-technology):

PowerXpress 3.0/Switchable Graphics: HD 6300M series, HD 6500M series.
PowerXpress 4.0/Dynamic Switchable Graphics: (HD 6300M series), HD 6400M series, HD 6600M/6700M series, HD 7300M series, HD 7400M series, HD 7500M/7600M series.
PowerXpress 5.0/Enduro: HD 7700M series, HD 7800M series, HD 7900M series

The HD 6300M series is reported as being based both on PowerXpress 3.0 (http://www.amd.com/us/products/technologies/switchable-graphics/Pages/switchable-graphics.aspx) and PowerXpress 4.0 (http://www.amd.com/us/products/notebook/graphics/amd-radeon-6000m/amd-radeon-6300m/Pages/amd-radeon-6300m.aspx), we need to see the report of an user with a GPU based on HD 6300M series to clarify this (apparently the muxless version of PowerXpress 3.0 is not supported on Linux, so if a HD 6300M series GPU uses the muxless version of PowerXpress 3.0 it should have the same issue reported on bug 56494 of Freedesktop).

There are actually 2-3 issues in this bug:
a) There is the incompatibility between the fglrx and the intel
driver. In order to get around that, the workaround for now is to
patch the intel driver (reverting a commit)
b) The fglrx driver did not included some paths and this prevented
applications needing direct rendering from functioning
Both fixes are now included in my PPA.

Muxless hybrid systems with fglrx/intel drivers should work by now if
they select the dgpu.
For issues when using the igpu there is the separate Bug: #1088220

Personally, I have a muxless based on 6630M .

Nikos

2012/12/10 Marco Vuano <email address hidden>:
> @Nick: I noticed that Freedesktop's bug 56494 affects a system based on AMD Radeon HD 6550M which uses PowerXpress 3.0/Switchable graphics and not PowerXpress 4.0/Dynamic Switchable graphics. It is likely that the bug is not the same.
> To clarify the version of PowerXpress supported by the various supported AMD cards, I'll post this short list based on information on AMD site (names are based on information available on AMD site and on the slides which can be viewed on http://www.slideshare.net/AMD/amd-enduro-technology):
>
> PowerXpress 3.0/Switchable Graphics: HD 6300M series, HD 6500M series.
> PowerXpress 4.0/Dynamic Switchable Graphics: (HD 6300M series), HD 6400M series, HD 6600M/6700M series, HD 7300M series, HD 7400M series, HD 7500M/7600M series.
> PowerXpress 5.0/Enduro: HD 7700M series, HD 7800M series, HD 7900M series
>
> The HD 6300M series is reported as being based both on PowerXpress 3.0
> (http://www.amd.com/us/products/technologies/switchable-graphics/Pages
> /switchable-graphics.aspx) and PowerXpress 4.0
> (http://www.amd.com/us/products/notebook/graphics/amd-radeon-6000m/amd-
> radeon-6300m/Pages/amd-radeon-6300m.aspx), we need to see the report of
> an user with a GPU based on HD 6300M series to clarify this (apparently
> the muxless version of PowerXpress 3.0 is not supported on Linux, so if
> a HD 6300M series GPU uses the muxless version of PowerXpress 3.0 it
> should have the same issue reported on bug 56494 of Freedesktop).
>
> --
> You received this bug notification because you are subscribed to the bug
> report.
> https://bugs.launchpad.net/bugs/1068404
>
> Title:
> Low graphics mode in Hybrid ATI/Intel GPU systems after fglrx upgrade
>
> To manage notifications about this bug go to:
> https://bugs.launchpad.net/xserver-xorg-video-intel/+bug/1068404/+subscriptions

Marco Vuano, now that fglrx-updates has been updated as per https://bugs.launchpad.net/ubuntu/+source/fglrx-installer-updates/+bug/1068404/comments/68 , are you able to use fglrx-updates in Quantal without a WORKAROUND?

Actually the version of this package is not released yet.
I marked it as "Fix Released" by error, could you please revert it to
"Fixed Commited"?

Thanks,
Nikos

Nick Andrik (andrikos) wrote :

Also, in order to have this bug fixed without a workaround we also
need the intel driver to be fixed, not only the fglrx

Marking Fix Committed as per Nick Andrik.

Changed in fglrx-installer-updates (Ubuntu):
status: Fix Released → Fix Committed
Kain Centeno (tenshinoneko) wrote :

Experiencing the same problems with integrated RADEON HD 7660D GPU.

^Nick, I followed the steps in the workaround. Still experiencing the same problem, unity won't load, and if try to start compiz it will fail too. Also the workaround shows an AMD watermark in the lower right.

Kain Centeno (tenshinoneko) wrote :

I followed the guide that said to install gnome-classic because of unity now launching, and after running fgrxinfo i got this:

X Error of failed request: BadRequest (invalid request code or no such operation)
  Major opcode of failed request: 155 (ATIFGLEXTENSION)
  Minor opcode of failed request: 66 ()
  Serial number of failed request: 13
  Current serial number in output stream: 13

Make sure you have no drivers installed from other PPAs.
What do you get when you run?
dpkg --list|grep -e intel -e fglrx

@Nick

LIBGL_DEBUG=verbose fglrxinfo:

X Error of failed request: BadRequest (invalid request code or no such operation)
  Major opcode of failed request: 136 (GLX)
  Minor opcode of failed request: 19 (X_GLXQueryServerString)
  Serial number of failed request: 12
  Current serial number in output stream: 12

LIBGL_DEBUG=verbose glxinfo

name of display: :0
X Error of failed request: BadRequest (invalid request code or no such operation)
  Major opcode of failed request: 136 (GLX)
  Minor opcode of failed request: 19 (X_GLXQueryServerString)
  Serial number of failed request: 12
  Current serial number in output stream: 12

Dimitri John Ledkov (xnox) wrote :

@ Everyone
This is quite a large bug report. Ubuntu Sponsors team is subscribed to this bug report & hence it appears at http://reqorts.qa.ubuntu.com/reports/sponsoring/ .

Are there any debdiffs / patches ready that need sponsoring?
If yes: which onces?
If no: ubuntu-sponsors team should be un-subscribed from the bug

(note that bzr branch merge proposals are accounted for separately)

Nick Andrik (andrikos) wrote :

@xnox
You can take a look on comment #44 for the proposed patch

bruise lee (workformydream) wrote :

Nick's PPA did sovled my low-graphic problem on a HP laptop (I7/6770M) for ubuntu 12.10/amd64

However there are 2 steps must to be taken, or-else still low-graphic( segmentation fault shown in Xorg.0.log):

1. run "sudo dpkg-reconfigure Xorg" before reboot

2. when linux kernel is 3.7 or higher, another patch must be apply to Nick's fglrx-updates_9.010-0ubuntu1~andrik7_amd64.deb
for deb file usr/src/fglrx-updates-9.010/firegl_public.c, add follow 4 lines to the head of file

// VM_RESERVED is deleted in 3.7 or higher kernels.
#ifndef VM_RESERVED
#define VM_RESERVED (VM_DONTEXPAND | VM_DONTDUMP)
#endif

then rebuild the deb and install it, which would resolved the error msg

Francesco (francesco-p18) wrote :

This bug affects me and the workaround didnt work to solve it...

Ped (ped) wrote :

@Gareth #65
I'm a developer too, and after all those years *I* don't care who's fault it is either, I feel just like ordinary user: 12.04 worked, 12.10 is crap (took me 4h on work day to revert to working X).

Installing packages from PPA didn't help me, I still get:
"(EE) this is a Muxless..."

My HW:
HP ProBook 4540s, Intel Core i5-2450M, AMD 7500M/7600M

here are steps I tried over time:

"dpkg --list|grep -e intel -e fglrx" output (sorry for shortening it, but I'm copying it by hand on second computer):
fglrx and xserver-xorg-video-intel installed from Andrik's PPA
libdrm-intel1:amd64 2.4.39-0ubuntu1 amd46
libdrm-intel1:i386 2.4.39-0ubuntu1 i386

tried to remove libgl11-mesa-dri, libdrm-intel1 (i386 variants only, as I figured out later :/, so invalid try anyway).
Still no graphics after reboot, only tty1

I tried to disable switchable graphics in BIOS, X will not start due to not finding any ATI device of course.

Removing xorg.conf (with ATI things defined) ... still tty1 (omg)
log X:
(II) Loading /usr/lib/xorg/modules/libfbdevhw.so
..
(EE) Backtrace:
(EE) 0: /usr/bin/X ...

reverting to official repository with video-intel:
- X works (GLX not: (EE) GLX error: Can not get required symbols.")

trying out Andrik's PPA version one more time:
X crashed again after loading libfbdevhw.so

.. so, the proposed packages from PPA don't work for me :/ ... any suggestions what should I try/report?

bruise lee (workformydream) wrote :

@Ped
here's what I did to get my 12.10 work:

1. purge fglrx* radeon*

2. add PPA of xorg-edgers, and "sudo apt-get update & apt-get dist-upgrade & apt-get upgrade"
after reboot, make sure the intel graphic card is working (there should be no error in Xorg.0.log)

3. download deb files from Nick Andrikos' PPA (3 for fglrx-0910, 2 for xserver-xorg-video-intel*)
patch fglrx-updates_9.010-0ubuntu1~andrik7_amd64.deb:

for deb file usr/src/fglrx-updates-9.010/firegl_public.c, add follow 4 lines to the head of file
// VM_RESERVED is deleted in 3.7 or higher kernels.
#ifndef VM_RESERVED
#define VM_RESERVED (VM_DONTEXPAND | VM_DONTDUMP)
#endif

4 intall above patched debs, and run "sudo aticonfig --initial -f" and "sudo dpkg-reconfigure Xorg"
reboot and I found no more segmentation fault in Xorg.0.log, and glxinfo shown corrent render info of AMD

Ped (ped) wrote :

intel-video from xorg-edgers PPA works well too (kernel 3.7.0, etc.. whole PPA), so the only crashing intel-video driver is from Nick's PPA. I thought only some removal of obsolete code was reverted on that one, so how comes it doesn't work? :)

Anyway, I didn't got further in the bruise lee advice, because with xorg-edgers PPA I get crash in xorg synapticsxfg, which makes a bit annoying to use the notebook (some keys don't work, and touchpad is unconfigured), will try it later when touchpad works or when I will have truck loads of patience for crippled inputs, and report here about fglrx.

What's the eta on a fix for this?

--
"Why use Ubuntu? It couldn’t be easier to use" According to this guy, hardware compatability is one of the top 5 reasons to choose Ubuntu 12.10 over Windows 8.

Mike (sugarhigh4242) wrote :

This affects me on an HP Envy 14 (intel + AMD 5650).

After installing fglrx I wasn't able to access my desktop. I purged fglrx and tried Andrikos' PPA, which didn't help. After purging everything, I was able to get LightDM up only after switching to the AMD card via vgaswitcheroo. But then logging into my account dumped me back to lightdm again. The only thing that worked was launching a Guest desktop instance.

No amount of purging and downgrading could salvage my system, I had to re-install 12.10. I never installed fglrx on 12.04 so I can't say whether this is a downgrade. But the opensource radeon drivers on 12.10 seems like a downgrade because HDMI switching doesn't work for me.

I'll make another partition to do more testing, so if anyone wants to see any console output/logs, let me know.

Nick Andrik (andrikos) wrote :

For quantal systems you will need both the intel and the fglrx packages from my quantal PPA.

For raring systems you will need the above intel package and another fglrx package (linux headers changed) already in my raring PPA, but untested (since I don't have raring):
https://launchpad.net/~andrikos/+archive/ppa/+sourcepub/2899856/+listing-archive-extra

If someone has a raring system and is willing to try the fglrx-amdcccle-updates_9.010.11-0ubuntu1~andrik*.deb package, please let me know.

Thanks

description: updated
Timo Aaltonen (tjaalton) wrote :

alrighty, reopening.. will revert the change for now, but not sure about doing it for raring

Changed in xserver-xorg-video-intel (Ubuntu):
assignee: Nick Andrik (andrikos) → Timo Aaltonen (tjaalton)
status: Invalid → Triaged
Changed in xserver-xorg-video-intel (Ubuntu Quantal):
assignee: nobody → Timo Aaltonen (tjaalton)
status: New → Triaged
Ped (ped) wrote :

"* This patch has enabled many users to have their hybrid systems working. No reported regressions so far"

^^ not true, the Nick's PPA packages do *NOT* work for me. (intel driver does crash straight away when vga switching is disabled in BIOS, before fglrx comes into play, X crashes after loading libfbdevhw.so)

Ped, regarding your comments https://bugs.launchpad.net/ubuntu/+source/fglrx-installer-updates/+bug/1068404/comments/107 :
>""* This patch has enabled many users to have their hybrid systems working. No reported regressions so far" ^^ not true, the Nick's PPA packages do *NOT* work for me. (intel driver does crash straight away when vga switching is disabled in BIOS, before fglrx comes into play, X crashes after loading libfbdevhw.so)"

The bug title is too broad, so we are unfortunately getting comments from people who have a hybrid ATI/Intel system, but this bug report's scope does not actually cover their hardware/problem. Based on you and other users who also found the WORKAROUND did not work, this bug report is not about your or their system, but of the original reporter's hardware combination and problem, and duplicate bug reports known to share the same root problem and WORKAROUND.

Hence, if you have a bug in Ubuntu, could you please file a new report by executing the following in a terminal:
ubuntu-bug xorg

For more on this, please see the Ubuntu X.Org Team (maintainers of fglrx-installer and fglrx-installer-updates), Ubuntu Bug Control, and Ubuntu Bug Squad article:
https://wiki.ubuntu.com/Bugs/BestPractices#X.2BAC8-Reporting.Focus_on_One_Issue

and Ubuntu Community article:
https://help.ubuntu.com/community/ReportingBugs#Bug_reporting_etiquette

When opening up the new report, please feel free to subscribe me to it.

Please note, not filing a new report may delay your problem being addressed as quickly as possible.

Thank you for your understanding.

Nick Andrik (andrikos) wrote :

@Timo:
Thanks for including it in quantal!
Before the release of raring, if no other solution is found and the problem still persists, probably we would need to apply it again.
In any case, I have tried 2.20.16 and X server still crashes without the patch.

@Ped:
A regression would be if someone has a working system, applies my patch and his system crashes.
If my patch does not work for you it's not a fix for your case, which means you have a different problem.
If you use 13.04, then please use the appropriate part of my PPA (I have prepared fglrx/-intel also specifically for raring):
https://launchpad.net/~andrikos/+archive/ppa/+packages?field.name_filter=&field.status_filter=published&field.series_filter=raring

Nick Andrik (andrikos) on 2013-01-04
summary: - Low graphics mode in Hybrid ATI/Intel GPU systems after fglrx upgrade
+ Low graphics mode in muxless hybrid ATI/Intel GPU systems after fglrx
+ upgrade
Launchpad Janitor (janitor) wrote :

Status changed to 'Confirmed' because the bug affects multiple users.

Changed in fglrx-installer (Ubuntu Quantal):
status: New → Confirmed
Changed in fglrx-installer-updates (Ubuntu Quantal):
status: New → Confirmed
Luigi R. (xluigi84) wrote :

Please remind that the workaround proposed in #1 by installing xserver-xorg-video-intel from ppa:andrikos/ppa causes again low graphic mode when fglrx is removed and i915/xserver-xorg-video-ati will be used.

Configuration:
Hp Pavillion dv6 6030el
VGA compatible controller: Advanced Micro Devices [AMD] nee ATI Caicos [Radeon HD 6400M/7400M Series] (it should be HD6470M
 i7-2630QM CPU

Nick Andrik (andrikos) wrote :

Luigi, do you still get low graphics mode, if you substitute my xserver-xorg-video-intel package with the official one?
( in order to understand if it is a regression in my package).

Does it happen even if you delete the xorg.conf (or substitute the fglrx driver in there with the radeon one)?

Luigi R. (xluigi84) wrote :

Yes, your package works with fglrx (catalyst 12.11) but as soon as I removed everything to perform a clean installation of the driver following the steps on the unofficial forum :

sudo sh /usr/share/ati/fglrx-uninstall.sh
sudo apt-get remove --purge fglrx fglrx_* fglrx-amdcccle* fglrx-dev*
If you plan on using open-source drivers, you will need to reinstall some packages because Catalyst overwrites or diverts some key 3D libraries with proprietary versions. For more information on this issue, see this Ubuntu wiki page
sudo apt-get remove --purge xserver-xorg-video-ati xserver-xorg-video-radeon
sudo apt-get install xserver-xorg-video-ati
sudo apt-get install --reinstall libgl1-mesa-glx libgl1-mesa-dri xserver-xorg-core
sudo mv /etc/X11/xorg.conf /etc/X11/xorg.conf.backup
sudo rm -rf /etc/ati

but keeping your xserver-xorg-video-intel I started to have a problem getting the low graphics mode windows

In order to see if the problem lies in my patch or in your system you
will have to test your steps with the default intel package (just open
synaptic and choose the appropriate version for
xserver-xorg-driver-intel).
Also, in the instructions, I see you remove the radeon driver, are you
sure you want to do this?

Luigi R. (xluigi84) wrote :
Download full text (16.2 KiB)

By installing the catalyst driver with the default xserver-xorg-driver-intel I have the problem mentioned on this bug. Instead using your patch will solve the problem concerning the fglrx but if the user would like to remove the proprietary driver and restore the previous one, he needs to remove also your patch otherwise he will get low graphics mode. I was not able to login till I restored the original xserver-xorg-driver-intel.

PS: the radeon package will come with xserver-xorg-video-ati

In addition I have a segmentation fault shutting down the machine if igpu is active. Please have a look below. In parallel I will update the bug http://ati.cchtml.com/show_bug.cgi?id=460 because I think it is more a fglrx problem

------------[ cut here ]------------
Jan 5 20:59:37 Houston kernel: [ 20.902685] WARNING: at /build/buildd/linux-3.5.0/drivers/pci/msi.c:828 pci_enable_msi_block+0x2a4/0x2b0()
Jan 5 20:59:37 Houston kernel: [ 20.902686] Hardware name: HP Pavilion dv6 Notebook PC
Jan 5 20:59:37 Houston kernel: [ 20.902687] Modules linked in: fglrx(PO) amd_iommu_v2 parport_pc ppdev snd_hda_codec_hdmi rfcomm bnep snd_hda_codec_idt hp_wmi sparse_keymap binfmt_misc joydev coretemp kvm ghash_clmulni_intel aesni_intel cryptd aes_x86_64 lib80211_crypt_tkip snd_hda_intel snd_hda_codec snd_hwdep snd_pcm snd_seq_midi snd_rawmidi hid_generic wmi snd_seq_midi_event snd_seq snd_timer snd_seq_device uvcvideo videobuf2_core videodev videobuf2_vmalloc videobuf2_memops usbhid mac_hid snd hid microcode wl(PO) btusb hp_accel i915 bluetooth psmouse lis3lv02d input_polldev serio_raw drm_kms_helper lp lib80211 parport drm soundcore snd_page_alloc lpc_ich mei rts_pstor(C) i2c_algo_bit video r8169 [last unloaded: amd_iommu_v2]
Jan 5 20:59:37 Houston kernel: [ 20.902726] Pid: 1253, comm: Xorg Tainted: P C O 3.5.0-21-generic #32-Ubuntu
Jan 5 20:59:37 Houston kernel: [ 20.902727] Call Trace:
Jan 5 20:59:37 Houston kernel: [ 20.902732] [<ffffffff81051c1f>] warn_slowpath_common+0x7f/0xc0
Jan 5 20:59:37 Houston kernel: [ 20.902735] [<ffffffff81051c7a>] warn_slowpath_null+0x1a/0x20
Jan 5 20:59:37 Houston kernel: [ 20.902737] [<ffffffff8136c5c4>] pci_enable_msi_block+0x2a4/0x2b0
Jan 5 20:59:37 Houston kernel: [ 20.902769] [<ffffffffa0bb3b63>] KCL_RequestMSI+0x13/0x20 [fglrx]
Jan 5 20:59:37 Houston kernel: [ 20.902798] [<ffffffffa0bd3c8f>] IRQMGR_initialize+0x24f/0x380 [fglrx]
Jan 5 20:59:37 Houston kernel: [ 20.902827] [<ffffffffa0bd39e7>] ? IRQMGR_alloc_context+0xb7/0x110 [fglrx]
Jan 5 20:59:37 Houston kernel: [ 20.902856] [<ffffffffa0bd2c2e>] ? irqmgr_wrap_initialize+0x1e/0xb0 [fglrx]
Jan 5 20:59:37 Houston kernel: [ 20.902884] [<ffffffffa0bd3885>] ? firegl_irqmgr_init+0x55/0x100 [fglrx]
Jan 5 20:59:37 Houston kernel: [ 20.902914] [<ffffffffa0be4486>] ? hal_init_gpu+0x1d6/0x4b0 [fglrx]
Jan 5 20:59:37 Houston kernel: [ 20.902940] [<ffffffffa0bc07a7>] ? firegl_open+0x2e7/0x320 [fglrx]
Jan 5 20:59:37 Houston kernel: [ 20.902960] [<ffffffffa0bb0b17>] ? ip_firegl_open+0x17/0x20 [fglrx]
Jan 5 20:59:37 Houston kernel: [ 20.902979] [<ffffffffa0bb0145>] ? firegl_stub_open+0xa5/0x120 [fglrx]
Jan 5 20:59:37 Houston ...

Luigi R. (xluigi84) wrote :

I have removed everything and started from scratch but using this time fglrx-updates instead of the beta version from the amd website to see if I can get rid of the segmentation fault. I will let you know after some test.

What I did so far:

sudo apt-get remove --purge fglrx fglrx-amdcccle
sudo apt-get remove --purge fglrx-updates fglrx-amdcccle-updates
sudo apt-get remove --purge xserver-xorg-video-intel
sudo apt-get clean
sudo apt-get install xserver-xorg-video-intel
sudo apt-get remove --purge xserver-xorg-video-ati xserver-xorg-video-radeon
sudo apt-get install xserver-xorg-video-ati
sudo apt-get install --reinstall libgl1-mesa-glx libgl1-mesa-dri xserver-xorg-core
sudo rm /etc/X11/xorg.conf
sudo rm -rf /etc/ati

reboot the machine, then:

sudo apt-get install fglrx-updates fglrx-amdcccle-updates
sudo gedit /etc/X11/Xsession.d/10fglrx and add /usr/lib/x86_64-linux-gnu/dri
sudo dpkg -i xserver-xorg-video-intel_2.20.9-0ubuntu2+andrik2_amd64.deb
sudo aticonfig --initial

neworld (aneworld) on 2013-01-06
Changed in fglrx-installer-updates (Ubuntu):
status: Fix Committed → Opinion
status: Opinion → Fix Committed
Nick Andrik (andrikos) wrote :

The crash when using the igpu is described in this bug:
https://bugs.launchpad.net/ubuntu/+source/fglrx-installer/+bug/1088220

Please comment there about that.

About the crashes when you don't use fglrx but you use the patched intel I will check it again and let you know.

Nick Andrik (andrikos) wrote :

Luigi, you are right. I have tested my patched package without using fglrx and I get a segmenation fault like in your case :(
I will update the bug description in order to delay the SRU till a better solution is found.

BTW, have you checked the official packages from raring to see if the updated versions work ok for quantal?

Thanks for pointing out the regression, much appreciated

Nick Andrik (andrikos) wrote :

@Timo:
Please delay the SRU. There has been a regression found, so we will have to improve the workaround before we commit it in -porposed

Thanks

description: updated
Nick Andrik (andrikos) on 2013-01-06
description: updated
Nick Andrik (andrikos) wrote :

@Ped:
You mentioned:
"intel-video from xorg-edgers PPA works well too (kernel 3.7.0, etc.. whole PPA), so the only crashing intel-video driver is from Nick's PPA."
Do you mean that the intel driver from x-edgers can be used along with fglrx to eliminate the problem of this bug?

Luigi R. (xluigi84) wrote :

@Nick:
 I will test both the raring and x edgers sources this evening and I will let you know.

neworld (aneworld) wrote :

Recently I updated my ubuntu 12.10 kernel to 3.5.0.22. Standard ubuntu fglrx works well. Just Intel video card render console.

Luigi R. (xluigi84) wrote :

Clean installation (everything from quantal repository) + x-edgers packages which include kernel 3.7 and xorg-intel + fglrx-updates = low graphics mode.

Instead by using fglrx my graphics card is not recognized by aticonfig.

Too many reboot recently...I went back to the default configuration (i915+radeon driver used just to turn off the Ati card).

I will try again in the future maybe when my ati will not be supported anymore and my laptop become too old enough to be replaced by other one with nvidia card.

It is absurd, when we buy a card we pay also the support that open-source users never receive :-(

Luigi R. (xluigi84) wrote :

PS: Thanks to all of you for your efforts :-)

Alfred E. Neumayer (beidl) wrote :

Luigi, I believe the problem with those xorg-edgers fglrx drivers is that the are for embedded APUs, at least from what I've read.
Try my package here: http://ubuntuone.com/7D7u8DgDzIAaC4qokYbUzs
I've used that one (had to apply the needed patch) when I was using quantal + kernel 3.7.

Luigi R. (xluigi84) wrote :

Thanks, just one question: is your fglrx based on the catalyst 12.10 or lower? Before 12.10 my card was not supported and maybe also the package on x-edgers is based on an older driver.

Mine was just a test for Nick to verify if the package xorg-intel coming from this ppa was able to run both fglrx-update(12.10) and i915/radeon driver. For the moment just the patch from Nick is working in my case with fglrx-update not considering the problem that introduce with i915/radeon.

Ps: I have noticed that in /etc/X11/Xsession.d/10fglrx we export LIBGL_DRIVERS_PATH as well as LD_LIBRARY_PATH, why after the reboot the first variable is exported properly and LD_LIBRARY_PATH is always empty on Ubuntu by default? Is it unset somewhere?

Ped (ped) wrote :

@Nick:

I haven't tried fglrx with xorg-edgers, only default xorg-video-intel + xorg-video-ati, as the synaptics (touchpad and some keyboard stuff too on my notebook) keeps crashing, I was too lazy to try anything more. But it boots into X, just as original repository version, so the intel+ati works at least as good as Ubuntu quantal.

BTW, after all those experiments on my kubuntu 12.10 install I suddenly got working GLX with intel+ati.
(history of my OS install:
12.04 clean install - desktop works
+ fglrx from Ubuntu repos - 3D games works on ATI, desktop works
* upgrade of distro to 12.10 -> blank screen, several hours of trying to recall all the xorg stuff and find out what works
- ending with intel + ati open source drivers, desktop shows, no GLX working
* trying out different things, including
- Nick's PPA (intel + fglrx = blank screen, intel + ati = blank screen)
- xorg-edgers (intel+ati = desktop works, no GLX)
- purging xorg-edgers PPA, reverting to default Ubuntu intel+ati, where desktop works, and was "no GLX" before, now all of sudden the GLX started to work without warning.

with glxinfo I now get:
"... | direct rendering: Yes | server glx vendor string: SGI | server glx version string: 1.4 | ...
OpenGL vendor string: Intel Open Source Technology Center
OpenGL renderer string: Mesa DRI Intel(R) Sandybridge Mobile
OpenGL version string: 3.0 Mesa 9.0
OpenGL shading language version string: 1.30
..."

By the "no GLX" situations before I did mean situation, where the OS boots up into X, I can log in, desktop works, but I can't run any 3D game, usually ending with some GLX specifix error, which I didn't bother to save, but something along unsupported feature number #x IIRC.

I have muxless HW (no HW switch for VGA, and when I install fglrx from Ubuntu repos, I get blank screen and log/xorg contains that weird "(EE) this is a Muxless PX A+I platform, we doesn't supported it".

I didn't have yet time to look into it, if I could find the difference. Maybe I left the ATI card still switched off in BIOS, so the Intel GLX works for me now? I know it's "just one reboot" away to figure out, but right now.. not.

Summary:
no working fglrx for me (any tested combination ends with blank screen)
intel+ati (ubuntu, xorg edgers): desktop works, now on ubuntu ones even GLX on Intel works (no idea, why/how it got fixed, maybe disabled ATI by BIOS, or reinstall of some core mesa packaged during xorg edgers ppa purge)
intel+ati (nick's PPA): crash in X as mentioned in #100 (this is different crash, not similar to intel+fglrx, which does crash in other stage of loading drivers)

Ped (ped) wrote :

I did check the BIOS, the ATI radeon was enabled, now with "Intel only" setting the GLX/OGL still works, so something got fixed on my Kubuntu installation while I was trying out different PPAs for drivers.

That said, I would still like to have radeon 3D working (like it was in 12.04). :)

bugbot (bugbot) on 2013-01-10
tags: added: resolution
Marco Vuano (vuanomarco) wrote :

For those who were wondering, the new Catlayst 13.1 driver apparently still has this bug. However, those with PowerXpress 3.0 cards (Mobility Radeon HD 5000 series, HD 6500M series cards and probably also HD 6300M cards) and who were having the "this is a Muxless PX A+I platform, we doesn't supported it" Xorg log message should try this new driver, because a similar bug is mentioned in its release notes.

Marco Vuano (vuanomarco) wrote :

I tried the new 13.2 beta driver from AMD (2:12.100-0ubuntu1) and they work correctly with the Intel driver from 12.10 repository (2:2.20.9-0ubuntu2)!
I think it is important to verify whether this problem is finally solved by the new 13.2 beta driver also for other users.

Marco Vuano (vuanomarco) wrote :

As a side note: with the 13.2 beta driver the integrated GPU too works fine without having to apply the patch for direct rendering.
It must be verified, however, if this is true also for systems where the patch was never applied.

Works for me, too - thank you, Marco!
Haven't checked the intel GPU for now, but for me it was critical to be able to use the proprietary driver (OpenCL). Finally AMD reacted...

Ped (ped) wrote :
Download full text (4.6 KiB)

blank screen with 13.2 beta for me, but the problem is now elsewhere.

Xorg.log:
[ 25.961] (II) LoadModule: "fglrx"
[ 25.961] (II) Loading /usr/lib/x86_64-linux-gnu/xorg/extra-modules/extra-modules.dpkg-tmp/modules/drivers/fglrx_drv.so
[ 26.712] (II) Module fglrx: vendor="FireGL - AMD Technologies Inc."
[ 26.712] compiled for 1.4.99.906, module version = 12.10.5
[ 26.712] Module class: X.Org Video Driver
[ 26.713] (II) Loading sub module "fglrxdrm"
[ 26.713] (II) LoadModule: "fglrxdrm"
[ 26.713] (II) Loading /usr/lib/x86_64-linux-gnu/xorg/extra-modules/extra-modules.dpkg-tmp/modules/linux/libfglrxdrm.so
[ 26.861] (II) Module fglrxdrm: vendor="FireGL - AMD Technologies Inc."
[ 26.861] compiled for 1.4.99.906, module version = 12.10.5
[ 26.861] (II) AMD Proprietary Linux Driver Version Identifier:12.10.05
[ 26.861] (II) AMD Proprietary Linux Driver Release Identifier: UNSUPPORTED-12.10.17
[ 26.861] (II) AMD Proprietary Linux Driver Build Date: Jan 28 2013 22:26:03
[ 26.861] (++) using VT number 7
[ 26.861] (WW) Falling back to old probe method for fglrx
[ 27.037] (II) Loading PCS database from /etc/ati/amdpcsdb /etc/ati/amdpcsdb.default
[ 27.070] ukiDynamicMajor: failed to open /proc/ati/major
[ 27.070] ukiDynamicMajor: failed to open /proc/ati/major
[ 27.071] (--) Chipset Supported AMD Graphics Processor (0x6841) found
[ 27.087] (II) AMD Video driver is running on a device belonging to a group targeted for this release
[ 27.099] (II) AMD Video driver is unsigned
[ 27.100] (II) fglrx(0): pEnt->device->identifier=0x7f5e283b40e0
[ 27.100] (II) fglrx(0): === [xdl_xs113_atiddxPreInit] === begin
[ 27.141] (**) fglrx(0): Depth 24, (--) framebuffer bpp 32
[ 27.141] (II) fglrx(0): Pixel depth = 24 bits stored in 4 bytes (32 bpp pixmaps)
[ 27.141] (==) fglrx(0): Default visual is TrueColor
[ 27.141] (**) fglrx(0): Option "DPMS" "true"
[ 27.141] (==) fglrx(0): RGB weight 888
[ 27.141] (II) fglrx(0): Using 8 bits per RGB
[ 27.141] (==) fglrx(0): Buffer Tiling is ON
[ 27.151] (II) Loading sub module "fglrxdrm"
[ 27.151] (II) LoadModule: "fglrxdrm"
[ 27.151] (II) Loading /usr/lib/x86_64-linux-gnu/xorg/extra-modules/extra-modules.dpkg-tmp/modules/linux/libfglrxdrm.so
[ 27.151] (II) Module fglrxdrm: vendor="FireGL - AMD Technologies Inc."
[ 27.151] compiled for 1.4.99.906, module version = 12.10.5
[ 27.152] ukiDynamicMajor: failed to open /proc/ati/major
[ 27.152] ukiDynamicMajor: failed to open /proc/ati/major
[ 27.153] (**) fglrx(0): NoAccel = NO
[ 27.153] (**) fglrx(0): AMD 2D Acceleration Architecture enabled
[ 27.153] (--) fglrx(0): Chipset: "AMD Radeon HD 7500M/7600M Series" (Chipset = 0x6841)
[ 27.153] (--) fglrx(0): (PciSubVendor = 0x103c, PciSubDevice = 0x17f4)
[ 27.153] (==) fglrx(0): board vendor info: third party graphics adapter - NOT original AMD
[ 27.153] (--) fglrx(0): Linear framebuffer (phys) at 0xb0000000
[ 27.153] (--) fglrx(0): MMIO registers at 0xd0900000
[ 27.153] (--) fglrx(0): I/O port at 0x00003000
[ 27.153] (==) fglrx(0): ROM-BIOS at 0x000c0000
[ 27.188] (II) fglrx(0): Inv...

Read more...

Marco Vuano (vuanomarco) wrote :

I think the message is not a real problem, but some kind of workaround. My system too give me this message when I switch to the discrete GPU:

sudo amdconfig --px-dgpu
PowerXpress info: Diagnostic output from /usr/lib64/fglrx/switchlibglx:
update-alternatives: attenzione: viene forzata l'installazione dell'alternativa /usr/lib/fglrx/ld.so.conf poiché il gruppo x86_64-linux-gnu_gl_conf è danneggiato
update-alternatives: attenzione: /usr/lib/x86_64-linux-gnu/xorg/extra-modules non viene sostituito con un collegamento

PowerXpress: Discrete GPU is selected (High-Performance mode), please restart Xserver(s) for changes to take effect!

Nick Andrik (andrikos) wrote :

I have packaged the latest beta version from upstream 13.2-beta3 in my PPA:
https://launchpad.net/~andrikos/+archive/xserver/

In my system, this version resolves the problem without the need to change anything in the intel driver (just use the default 2:2.20.9-0ubuntu2 version shipping in quantal).

Could anyone please test the packages in your systems in order to see if it is a proper fix and if any regressions appear?

Thanks,
Nick

udit mahajan (mahajanudit) wrote :

Did not work for me. I downloaded the tarball from ati website a couple of days ago. I did not really check te error i got this time though.. aticonfig did not give any error, unlike earlier but the system started in low graphics mode.
I just switched to previous ati driver because I highly doubt I will be getting anything extra from new drivers..

Marco Vuano (vuanomarco) wrote :

On my system the driver works fine both with dgpu and the igpu, excpet for a bug which should be the same as Bug #1088220. Basically, when using the iGPU some periodic crashes of the X server with subsequent Low Graphics Mode are experienced. However, the main bug reported here seems to be solved. Thanks Nick for packing this new driver so that we can finally assess if it is the actual fix for this bug!

Marco Vuano (vuanomarco) wrote :

P.S.: the "amd13_2_betacrashdmesg" file contains the dmesg output for one of the periodic crashes so that it is possible to verify if it is the same issue as in Bug #1088220.

It seems that the fglrx 12.3-beta3 driver solves this issus without any patch needed for the intel driver.

Yes, I menat the 13.2, not 12.3, sorry.

I packaged and tried that version for ubuntu and it seems to resolve the problem.

I can confirm that 13.1, solves this issue on openSUSE 12.2 with 2.20.3 intel driver.

But it's still/again broken with fglrx 12.10.5/xf86-video-intel 2.20.19. Verfied
with Intel Ivybridge (8086:0106) / AMD Radeon HD 7500M/7600m Series (1002:6841).

(==) fglrx(0): hotplug detection: "enabled"

Backtrace:
0: /usr/bin/X(xorg_backtrace+0x26) [0x4f4916]
1: /usr/bin/X(xf86SigHandler+0x39) [0x4915b9]
2: /lib64/libc.so.6(+0x32bd0) [0x7f7654562bd0]
3: /usr/lib64/xorg/modules//drivers/fglrx_drv.so(xclIn32+0) [0x7f76524aa8c0]
4: /usr/lib64/xorg/modules//drivers/fglrx_drv.so(xilReadReg32_R600+0xe) [0x7f765265736e]
5: /usr/lib64/xorg/modules//drivers/fglrx_drv.so(xdl_x740_atiddxPxScreenInit+0xa9) [0x7f76524db4d9]
6: /usr/lib64/xorg/modules//drivers/fglrx_drv.so(xdl_x740_atiddxScreenInit+0xdc8) [0x7f76524b83a8]
7: /usr/bin/X(AddScreen+0x1c6) [0x433786]
8: /usr/bin/X(InitOutput+0x241) [0x46eb31]
9: /usr/bin/X(main+0x200) [0x433e90]
10: /lib64/libc.so.6(__libc_start_main+0xe6) [0x7f765454ec16]
11: /usr/bin/X() [0x4334d9]

Fatal server error:
Caught signal 11. Server aborting

Nick Andrik, would you please queue up the new driver that works http://support.amd.com/us/gpudownload/linux/Pages/radeon_linux.aspx into the Proposed repository for fglrx-installer-updates? You may want to connect with Alberto Malone on this one.

Marking Triaged as not assigned to anyone, and nothing queued up in Proposed. Marking Status High as this has a moderate impact on a large portion of Ubuntu users. By moderate it is ment that as in while it is initially angst inducing to not be able to use Quantal with fglrx-installer/fglrx-installer-updates, one could still use FI/FIU in Precise.

Changed in fglrx-installer-updates (Ubuntu):
importance: Undecided → High
status: Fix Committed → Triaged
Changed in fglrx-installer-updates (Ubuntu Quantal):
importance: Undecided → High
status: Confirmed → Triaged
Changed in xserver-xorg-video-intel (Ubuntu Quantal):
importance: Undecided → Wishlist

I think the intel driver is out of this bug and should be marked as invalid.

The solution relies in the fglrx driver. I have tried the 13.2beta3
driver and it solves the problem.
If we don't want to go for this version, a Suse user on the
corresponding attached bug, reported that also 13.1 is ok.

I have already packaged the 13.2beta3 version here:
https://launchpad.net/~andrikos/+archive/xserver/

Should I also prepare an official branch to propose for merging to
quantal-proposed?

--
=Do-
N.AND

Nick Andrik (andrikos) on 2013-02-17
Changed in xserver-xorg-video-intel (Ubuntu):
status: Triaged → Invalid
pcworld (pcworld) wrote :

OT
(sorry for that change, my mouse tends to double-click instead of single-click sometimes)

information type: Public → Public Security
information type: Public Security → Public
Nick Andrik (andrikos) wrote :
no longer affects: xserver-xorg-video-intel (Ubuntu Quantal)
Nick Andrik (andrikos) on 2013-02-18
description: updated
description: updated

Stefan,

The new issue is caused by Intel driver change to map the GTT aperture (bottom half of BAR 1) as WC, while the top half of BAR1 is still Un-Cacheable. Then the conflict happens when fglrx driver maps the whole BAR1 of Intel IGP device. We have already prepared a fix for this, which should be available in Catalyst 13.3. As a workaround before that, I think you can do some change in Intel graphics driver to map GTT aperture as Un-Cacheable.

Some more information about this for the Intel driver change can be found at https://bugs.freedesktop.org/show_bug.cgi?id=55834 and https://patchwork.kernel.org/patch/1309531/

Thanks for explanation, Jammy!

Ped (ped) wrote :

13.2beta6 ( amd-driver-installer-catalyst-13.2-beta6-linux-x86.x86_64.zip ):

still no luck on my HP 4540s, again it fails with invalid BIOS message.
But again it fails reasonably, no crash anymore, just error message in Xorg.log.

I did report it as http://ati.cchtml.com/show_bug.cgi?id=701 ... not sure if somebody from AMD will pay attention or if the issue can be resolved on their side, but I think I'm out of luck at least till fglrx 13.4. :/

Marco Vuano (vuanomarco) wrote :

In the 13.2 beta 6 drivers (2:12.100-0ubuntu1) the situation is more or less the same as that with the beta 3 drivers, the periodic crashes of the Xserver when using the iGPU are there as well.
However, I noted that with the 13.2 beta drivers (both version 3 and 6) XvBA seems to be broken:
when using XvBA for VA API (adding LIBVA_DRIVER_NAME=xvba in /etc/environment) vainfo makes the X server crash and restart with the following message:

libva: VA-API version 0.32.0
libva: User requested driver 'xvba'
libva: Trying to open /usr/lib/x86_64-linux-gnu/dri/xvba_drv_video.so
XIO: fatal IO error 11 (Resource temporarily unavailable) on X server ":0"
      after 11 requests (11 known processed) with 0 events remaining.

Of course also trying to play a video with hardware acceleration makes the X server crash and restart.
If the LIBVA_DRIVER_NAME environment variable is removed from /etc/environment, vainfo works again and it is possible to play videos even when using VA API for video acceleration (independently of the GPU selected). The vainfo output is:

libva: VA-API version 0.32.0
libva: va_getDriverName() returns 0
libva: Trying to open /usr/lib/x86_64-linux-gnu/dri/i965_drv_video.so
libva: va_openDriver() returns 0
vainfo: VA-API version: 0.32 (libva 1.0.15)
vainfo: Driver version: Intel i965 driver - 1.0.17
vainfo: Supported profile and entrypoints
      VAProfileMPEG2Simple : VAEntrypointVLD
      VAProfileMPEG2Main : VAEntrypointVLD
      VAProfileH264Baseline : VAEntrypointVLD
      VAProfileH264Baseline : VAEntrypointEncSlice
      VAProfileH264Main : VAEntrypointVLD
      VAProfileH264Main : VAEntrypointEncSlice
      VAProfileH264High : VAEntrypointVLD
      VAProfileH264High : VAEntrypointEncSlice
      VAProfileVC1Simple : VAEntrypointVLD
      VAProfileVC1Main : VAEntrypointVLD
      VAProfileVC1Advanced : VAEntrypointVLD

Still, this is a minor issue compared to the main bug treated here.

Nick Andrik (andrikos) wrote :

Marco, I think the periodic crashes refer to the Bug #1088220 which still remains.
At least the new (13.1+) drivers seem to solve the initial segmentation fault, is that right?

Marco Vuano (vuanomarco) wrote :

Yes, this is what I was trying to explain.

udit mahajan (mahajanudit) wrote :

Nick, I upgraded my xserver-xorg-video-intel by mistake and now my system doesn't start in full graphics. I could not find the patched deb in your repo now and as mentioned earlier the new ati beta still doesn't work with radeon hd7xxxm. Kindly see if you can send me the file somehow.

I m using windows now, so you can understand my situation....

udit mahajan (mahajanudit) wrote :

Never mind, I found the files in your launchpad archives and now its all fine.
Thank you again for the patch. :)

Timo Aaltonen (tjaalton) on 2013-02-26
no longer affects: xserver-xorg-video-intel (Ubuntu)
Bill (lwlw1988) wrote :
Download full text (18.6 KiB)

@Nick,
I installed the packages in your ppa.
$ dpkg --list|grep -e video-intel -e fglrx
ii fglrx 2:9.000-0ubuntu3+andrik4 amd64 Video driver for the AMD graphics accelerators
ii fglrx-amdcccle 2:9.000-0ubuntu3+andrik4 amd64 Catalyst Control Center for the AMD graphics accelerators
ii xserver-xorg-video-intel 2:2.20.9-0ubuntu2+andrik2 amd64 X.Org X server -- Intel i8xx, i9xx display driver
X works correctly, only not able to switch to DGPU. But the biggest problem is when shutdowning, there is always a crash.

Feb 27 11:26:06 wliao-envymint kernel: [ 5.891430] fglrx_pci 0000:01:00.0: >irq 49 for MSI/MSI-X
Feb 27 11:26:06 wliao-envymint kernel: [ 5.892544] [fglrx] Firegl kernel thread PID: 2334
Feb 27 11:26:06 wliao-envymint kernel: [ 5.892688] [fglrx] Firegl kernel thread PID: 2335
Feb 27 11:26:06 wliao-envymint kernel: [ 5.892754] [fglrx] Firegl kernel thread PID: 2336
Feb 27 11:26:06 wliao-envymint kernel: [ 5.892828] [fglrx] IRQ 49 Enabled
Feb 27 11:26:06 wliao-envymint kernel: [ 5.944549] [fglrx] Gart USWC size:1204 M.
Feb 27 11:26:06 wliao-envymint kernel: [ 5.944556] [fglrx] Gart cacheable size:478 M.
Feb 27 11:26:06 wliao-envymint kernel: [ 5.944565] [fglrx] Reserved FB block: Shared offset:0, size:1000000
Feb 27 11:26:06 wliao-envymint kernel: [ 5.944569] [fglrx] Reserved FB block: Unshared offset:f93a000, size:3c6000
Feb 27 11:26:06 wliao-envymint kernel: [ 5.944573] [fglrx] Reserved FB block: Unshared offset:7fff4000, size:c000
Feb 27 11:26:06 wliao-envymint kernel: [ 6.411819] [fglrx] module unloaded - fglrx 9.0.2 [Sep 20 2012]
Feb 27 11:26:06 wliao-envymint kernel: [ 6.429481] [fglrx] Maximum main memory to use for locked dma buffers: 3696 MBytes.
Feb 27 11:26:06 wliao-envymint kernel: [ 6.429630] [fglrx] vendor: 1002 device: 6840 count: 1
Feb 27 11:26:06 wliao-envymint kernel: [ 6.429800] [fglrx] ioport: bar 4, base 0x3000, size: 0x100
Feb 27 11:26:06 wliao-envymint kernel: [ 6.429871] [fglrx] Kernel PAT support is enabled
Feb 27 11:26:06 wliao-envymint kernel: [ 6.429885] [fglrx] module loaded - fglrx 9.0.2 [Sep 20 2012] with 1 minors
Feb 27 11:26:07 wliao-envymint kernel: [ 7.425490] ------------[ cut here ]------------
Feb 27 11:26:07 wliao-envymint kernel: [ 7.425508] WARNING: at /build/buildd/linux-3.5.0/drivers/pci/msi.c:828 pci_enable_msi_block+0x2a4/0x2b0()
Feb 27 11:26:07 wliao-envymint kernel: [ 7.425512] Hardware name: HP ENVY Sleekbook 4 PC
Feb 27 11:26:07 wliao-envymint kernel: [ 7.425514] Modules linked in: fglrx(PO) bnep rfcomm parport_pc ppdev snd_hda_codec_hdmi snd_hda_codec_idt coretemp kvm ghash_clmulni_intel aesni_intel cryptd aes_x86_64 snd_hda_intel snd_hda_codec snd_hwdep joydev snd_pcm hid_generic hp_wmi sparse_keymap snd_seq_midi snd_rawmidi uvcvideo videobuf2_core videodev videobuf2_vmalloc videobuf2_memops snd_seq_midi_event snd_seq microcode snd_timer snd_seq_device usbhid snd hid psmouse serio_raw arc4 lpc_ich i915 drm_kms_helper ath9k(O) mac80211(O) ath9k_common(O) ath9k_...

Try to install this version 2:12.100.3-0~andrik1 (for both
fglrx-updates and fglrx-amdcccle-updates) and remove the updated intel
driver.
Just the updated fglrx driver should work

--
=Do-
N.AND

Bill (lwlw1988) wrote :

It's works like a charm. The only problem now is that I can't switch to fglrx(high performance) without causing X to crash(segment fault). But, at least I can safely shutdown. Thanks.

Marco Vuano (vuanomarco) wrote :

With Catalyst 13.2 beta7 (2:12.100-0ubuntu1) not much has changed (not even the version number!), so to sum it up:
-2D and 3D acceleration is mostly fine with both the iGPU and dGPU (at least Unity and some OpenGL appplications such as glxgears work correctly)
-XvBA seems to be broken (or the xvba-va-driver package no longer works correctly with the new driver)
-when using the iGPU the system crashes in a reproducible manner (this is described in bug #1088220), if the dGPU is needed, then it is probably better to open a terminal after login and to enter

sudo amdconfig --px-dgpu

After system restart the dGPU should be in use (and probably the system restart itself will result in a crash, because of the aforementioned bug, however next time Ubuntu is booted, the dGPU should be used). Sometimes restarting the X server by logging out after entering the command is enough to switch.

P.S:: @Nick Andrik: I think the last comment of bug #1088220 was wrongly posted there because the same comment is in this bug. Could you please confirm this and eventually add a comment stating that the bug still persists?

Nick Andrik (andrikos) wrote :

Yes, you are probably right, I'm hiding my comment on the other bug

Thanks for pointing it out!

DsXack (dsxack) wrote :

Description: Ubuntu Raring Ringtail (development branch)
Release: 13.04

lspci | grep -i vga:
00:02.0 VGA compatible controller: Intel Corporation 2nd Generation Core Processor Family Integrated Graphics Controller (rev 09)
01:00.0 VGA compatible controller: Advanced Micro Devices [AMD] nee ATI Whistler XT [AMD Radeon HD 6700M Series]

after install any version fglrx driver (from ubuntu repo / from amd website 13.1, 13.2 beta / from ppa repo andrikos 12.10) - xorg crash on start.

DsXack (dsxack) wrote :

I'm back to ubuntu 12.10 and install fglrx-update from andrikos/ppa but xserver don't run again(

Before this (install fglrx from andrikos/ppa) works...

Download full text (8.1 KiB)

 just tried to install the fglrx-update drivers on Ubuntu 13.04, and after the restart, I got X server but no menu bars, window decorations, etc.

lspci | grep VGA
00:02.0 VGA compatible controller: Intel Corporation 2nd Generation Core Processor Family Integrated Graphics Controller (rev 09)
01:00.0 VGA compatible controller: Advanced Micro Devices [AMD] nee ATI Whistler [AMD Radeon HD 6600M Series]

unity --replace & output:

darkhole@devolo:~$ unity --replace &
[1] 2857
darkhole@devolo:~$ compiz (core) - Info: Loading plugin: core
compiz (core) - Info: Starting plugin: core
unity-panel-service: proceso no encontrado
compiz (core) - Info: Loading plugin: ccp
compiz (core) - Info: Starting plugin: ccp
compizconfig - Info: Backend : gsettings
compizconfig - Info: Integration : true
compizconfig - Info: Profile : unity
compiz (core) - Info: Loading plugin: composite
compiz (core) - Info: Starting plugin: composite
compiz (core) - Info: Loading plugin: decor
compiz (core) - Info: Starting plugin: decor
compiz (decor) - Warn: requested a pixmap type decoration when compositing isn't available
compiz (decor) - Warn: requested a pixmap type decoration when compositing isn't available
compiz (decor) - Warn: requested a pixmap type decoration when compositing isn't available
compiz (decor) - Warn: requested a pixmap type decoration when compositing isn't available
compiz (decor) - Warn: requested a pixmap type decoration when compositing isn't available
compiz (decor) - Warn: requested a pixmap type decoration when compositing isn't available
compiz (decor) - Warn: requested a pixmap type decoration when compositing isn't available
compiz (decor) - Warn: requested a pixmap type decoration when compositing isn't available
compiz (decor) - Warn: requested a pixmap type decoration when compositing isn't available
compiz (decor) - Warn: requested a pixmap type decoration when compositing isn't available
compiz (decor) - Warn: requested a pixmap type decoration when compositing isn't available
Compiz (opengl) - Fatal: glXQueryExtensionsString is NULL for screen 0
Compiz (opengl) - Fatal: glXQueryExtensionsString is NULL for screen 0
Compiz (opengl) - Fatal: glXQueryExtensionsString is NULL for screen 0
Compiz (opengl) - Fatal: glXQueryExtensionsString is NULL for screen 0
Compiz (opengl) - Fatal: glXQueryExtensionsString is NULL for screen 0
Compiz (opengl) - Fatal: glXQueryExtensionsString is NULL for screen 0
Compiz (opengl) - Fatal: glXQueryExtensionsString is NULL for screen 0
Compiz (opengl) - Fatal: glXQueryExtensionsString is NULL for screen 0
Compiz (opengl) - Fatal: glXQueryExtensionsString is NULL for screen 0
Compiz (opengl) - Fatal: glXQueryExtensionsString is NULL for screen 0
Compiz (opengl) - Fatal: glXQueryExtensionsString is NULL for screen 0
Compiz (opengl) - Fatal: glXQueryExtensionsString is NULL for screen 0
Compiz (opengl) - Fatal: glXQueryExtensionsString is NULL for screen 0
Compiz (opengl) - Fatal: glXQueryExtensionsString is NULL for screen 0
Compiz (opengl) - Fatal: glXQueryExtensionsString is NULL for screen 0
Compiz (opengl) - Fatal: glXQueryExtensionsString is NULL for screen 0
Compiz (opengl) - Fatal: glXQueryExtens...

Read more...

Marco Vuano (vuanomarco) wrote :

The new 13.3 beta 2 drivers (still 2:12.100-0ubuntu1 !) behave more or less the same as I written in the previous post about the 13.2 beta 7 drivers.

Marco Vuano (vuanomarco) wrote :

P.S.: since I reported this bug with Ubuntu 12.10 I think I'm going to use it until this bug is solved, so unfortunately I cannot help those who have Ubuntu 13.04 and are having issues with the X server different from those I reported.

Marco Vuano (vuanomarco) wrote :

Nothing new with the 13.3 beta 3 drivers regarding this bug, as far as I know (even the version is the same: 2:12:100ubuntu1).

Nick Andrik (andrikos) wrote :

Guys just a small remark.
I have changed laptop and I don't have my intel/ati system any more.
So, I won't be able to troubleshoot any more :(

Nick

udit mahajan (mahajanudit) wrote :

Any update on this bug for 13.04?
I installed it on a different partition and could not use the laptop because of the heating problem. When I installed Ati driver to switch to Intel Graphics, it had the same behaviour.
I thought Ati might have had this bug fixed by now for good, it has been more than 6 months now. If there is something community can help with kindly let us know.
Nick, is it possible to use the previous xorg intel package you created for 13.04 as well (obviously might be with some change)? I know you changed the laptop.. Could you tell me how to do that?

Bryce Harrington (bryce) on 2013-04-09
Changed in fglrx-installer-updates (Ubuntu):
assignee: nobody → Canonical X.org (canonical-x)
Marco Vuano (vuanomarco) wrote :

With the new 13.4 drivers (version 2:12.104-0ubuntu1) the situtaion is the same as reported in my previous post about the 13.2 beta 7 drivers.

P.S.: @Nick Andrik: Even if late, I would like to thank you again for your precious aid in solving this bug!

Marco Vuano (vuanomarco) wrote :

P. P. S.: my message is late, your help was fast!

Jammy Zhou (jammy-zhou) wrote :

Can you check if the fglrx.ko kernel module is loaded successfully? And can you provide Xorg.0.log and dmesg log for analysis?

Marco Vuano (vuanomarco) wrote :

Here is the lsmod output, dmesg log and Xorg log for the XvBa crash
(modules, dmesgXvBAcrash and Xorglogcrash respectively) and the lsmod
output, dmesg log, syslog and Xorg log for the crash while restarting Xorg
with iGPU selected (modulesintel, dmesgintelcrash, syslogintelcrash and
Xorglogintelcrash respectively). Curiously, the Dmesg log is not apparently
stored up to the point of the crash, so I attach the syslog where
information about the crash can be found

Jammy Zhou (jammy-zhou) wrote :

Thanks for your info, Marco. So current problems are different from original one discussed in this thread, right? In this case, I'd suggest to create new bugs to avoid confusion (there are too much information mixed in this thread).

Marco Vuano (vuanomarco) wrote :

This is correct. Regarding the crash when iGPU is selected, since it is similar to Bug #1088220 I didn't report a new bug, but I can if you think it is useful, regarding the bug with XvBA, I reported it as bug 789 on the unofficial ATI Bugzilla, since the new drivers are apparently not available on the Ubuntu 12.10 repositories.
The bug reported here is fixed by the new drivers.

Hello, I am just writing to say that there is also a workaround for raring. I've been searching for this for some time so I thought that maybe this will help somebody in a similar situation: the lastest intel driver apparently causes trouble so reverting to the driver from the quantal repos makes the 13.4 driver work beautifully!
Solution thanks to https://launchpad.net/~andvaranaut from his bugreport #1174278

Chris Sykes (jcgs) wrote :

Hi, what's going on with this bug in raring? There doesn't seem to be much activity on this page, but I think the bug probably still exists. If I install fglrx from the repos, then my X server segfaults.

Sydoruk Yaroslav (lip) wrote :

Solved in Ubuntu 13.04
Need install:
AMD Catalyst™ 13.4 Proprietary Linux x86 Display Driver from ati.com
and http://launchpadlibrarian.net/111621157/xserver-xorg-video-intel_2.20.2-1ubuntu1_amd64.deb
If install xserver-xorg-video-intel_2:2.21.6-0ubuntu4_amd64.deb X server crashes

Juan Pablo Rinaldi (juampi) wrote :

@Sydoruk: I didn't quite understand the fix. Do I just have to install the proprietary driver from ATI? What do you mean by "If install xserver-xorg-video-intel_2:2.21.6-0ubuntu4_amd64.deb X server crashes"?

AG (a-gupta-archit) wrote :

Yes i confirm.. that the above solution by christopher pijarski (kpijarski) works
For this solution it is additionally required to install libudev0 from quantal repo
---
Easy fix method
1. search for 'quantal amd64 xserver-xorg-video-intel' on google or launchpad and download the package
2. similarly search for 'quantal amd64 libudev0' and download the package
3. install both the packages
4. install fglrx drivers if not done already
5. in terminal issue command 'amdconfig --initial -f' to force-initiliaze /etc/X11/xorg.conf
6. reboot and enjoy!

AG (a-gupta-archit) wrote :

download the following packages
http://launchpadlibrarian.net/137095219/xserver-xorg-video-intel_2.20.9-0ubuntu2.1_amd64.deb
http://launchpadlibrarian.net/119461136/libudev0_175-0ubuntu13_amd64.deb

these are the two files that u will find in steps 1 and 2 of my above comment

I confirm that installing these two files we get a working configuation

Now the battery drain issue is gone!

Juan Pablo Rinaldi (juampi) wrote :

I tried installing this package:
http://launchpadlibrarian.net/137095219/xserver-xorg-video-intel_2.20.9-0ubuntu2.1_amd64.deb
but it says I already have a more recent version installed.

Do I still have to install the package? If so, how?

Gareth (gmccumskey) wrote :
Download full text (17.1 KiB)

sudo apt-get remove --purge xserver-xorg-video-intel

Run that before you try to install it

On Mon, May 13, 2013 at 1:50 PM, Juan Pablo Rinaldi <email address hidden>wrote:

> I tried installing this package:
>
> http://launchpadlibrarian.net/137095219/xserver-xorg-video-intel_2.20.9-0ubuntu2.1_amd64.deb
> but it says I already have a more recent version installed.
>
> Do I still have to install the package? If so, how?
>
> --
> You received this bug notification because you are subscribed to the bug
> report.
> https://bugs.launchpad.net/bugs/1068404
>
> Title:
> Low graphics mode in muxless hybrid ATI/Intel GPU systems after fglrx
> upgrade
>
> Status in X.org xf86-video-intel:
> Won't Fix
> Status in “fglrx-installer” package in Ubuntu:
> In Progress
> Status in “fglrx-installer-updates” package in Ubuntu:
> Triaged
> Status in “fglrx-installer” source package in Quantal:
> Confirmed
> Status in “fglrx-installer-updates” source package in Quantal:
> Triaged
> Status in “xserver-xorg-video-intel” package in openSUSE:
> In Progress
>
> Bug description:
> [Impact]
>
> * Hybrid systems with ATI/intel GPUs using fglrx get a segmentation
> fault in X when using 12.10. In stock 12.04.1 these systems were
> working fine.
>
> * This bug represents a severe regression from precise to quantal,
> since is crashing on startup
>
> * The proposed patch updates the driver to version 13.2beta3 of
> upstream which seems to resolve this issue
>
> [Test Case]
>
> * Get an hybrid ATI/intel system.
> Install quantal: The system should boot fine
> Install the proprietary fglrx package and reboot: The X server
> crashes on startup
>
> [Regression Potential]
>
> * This solution does not need any change on the intel driver (as a
> previous workaround needed).
> * No regressions have been reported so far using this most recent fglrx
> upstream, by the people affected by this bug
>
> ======
>
> After installing the latest fglrx-updates package on Ubuntu 12.10
> (fglrx-updates_9.000-0ubuntu3_amd64) I have the same issue that was
> present before AMD fixed switchable graphics on PowerXpress 4.0 cards
> (such as my HD 6470M coupled with the intel HD Graphics 3000 of my
> intel Core i5-2430M). The X server has a segmentation fault and the
> low graphics mode is activated.
>
> WORKAROUND: Follow carefully the instructions here:
> https://help.ubuntu.com/community/BinaryDriverHowto/ATI#WORKAROUND
>
> As per Nick Andrik, this is a version regression of xserver-xorg-
> video-intel from 2.20.2 (working and in below mentioned PPA) and
> 2.20.3. The regression commit is
> http://cgit.freedesktop.org/xorg/driver/xf86-video-
> intel/commit/?id=05dcc5f1699ba90fc14c50882e8d4be89bc4a4f9 .
>
> Upstream fglrx has published a new version (> 13.1 ) of the driver
> which seems to solve this issue. No need to revert the commit in the
> Intel driver any more.
>
> ATTENTION:
> 1. If you choose the integrated GPU (via Catalyst or via amdconfig
> --px-igpu) and your X server crashes on startup, then your are probably
> affected by this bug:
> https://bugs.launchpad.net/ubuntu/+source/fglrx-install...

Juan Pablo Rinaldi (juampi) wrote :

Alright, I can confirm the fix works. According to the battery indicator, the battery life seems to have almost doubled now, but I'll have to test that to be sure.

Great work guys! I'm really grateful.

Juan Pablo Rinaldi (juampi) wrote :

The issue is back. Ubuntu told me there were updates for my system, so I installed them, and now the issue has returned.

I'm pretty sure what happened was that Ubuntu upgraded the old xserver-xorg-video-intel package to the most recent version.

How can I prevent Ubuntu from trying to upgrade that package every time?

Gareth (gmccumskey) wrote :
Download full text (17.5 KiB)

Reinstall that intel driver you downloaded previously. then use the
following command to prevent Ubuntu automatically updating the package

sudo apt-mark hold xserver-xorg-video-intel

Ubuntu will always try to upgrade that package because there are more
recent versions of it. The above command will prevent Ubuntu automatically
upgrading it. It will still appear in your package update dialogue that
pops up but unchecked. Just don't mark it for upgrade yourself in that and
you will be fine.

On Thu, May 16, 2013 at 4:28 PM, Juan Pablo Rinaldi <email address hidden>wrote:

> The issue is back. Ubuntu told me there were updates for my system, so I
> installed them, and now the issue has returned.
>
> I'm pretty sure what happened was that Ubuntu upgraded the old xserver-
> xorg-video-intel package to the most recent version.
>
> How can I prevent Ubuntu from trying to upgrade that package every time?
>
> --
> You received this bug notification because you are subscribed to the bug
> report.
> https://bugs.launchpad.net/bugs/1068404
>
> Title:
> Low graphics mode in muxless hybrid ATI/Intel GPU systems after fglrx
> upgrade
>
> Status in X.org xf86-video-intel:
> Won't Fix
> Status in “fglrx-installer” package in Ubuntu:
> In Progress
> Status in “fglrx-installer-updates” package in Ubuntu:
> Triaged
> Status in “fglrx-installer” source package in Quantal:
> Confirmed
> Status in “fglrx-installer-updates” source package in Quantal:
> Triaged
> Status in “xserver-xorg-video-intel” package in openSUSE:
> In Progress
>
> Bug description:
> [Impact]
>
> * Hybrid systems with ATI/intel GPUs using fglrx get a segmentation
> fault in X when using 12.10. In stock 12.04.1 these systems were
> working fine.
>
> * This bug represents a severe regression from precise to quantal,
> since is crashing on startup
>
> * The proposed patch updates the driver to version 13.2beta3 of
> upstream which seems to resolve this issue
>
> [Test Case]
>
> * Get an hybrid ATI/intel system.
> Install quantal: The system should boot fine
> Install the proprietary fglrx package and reboot: The X server
> crashes on startup
>
> [Regression Potential]
>
> * This solution does not need any change on the intel driver (as a
> previous workaround needed).
> * No regressions have been reported so far using this most recent fglrx
> upstream, by the people affected by this bug
>
> ======
>
> After installing the latest fglrx-updates package on Ubuntu 12.10
> (fglrx-updates_9.000-0ubuntu3_amd64) I have the same issue that was
> present before AMD fixed switchable graphics on PowerXpress 4.0 cards
> (such as my HD 6470M coupled with the intel HD Graphics 3000 of my
> intel Core i5-2430M). The X server has a segmentation fault and the
> low graphics mode is activated.
>
> WORKAROUND: Follow carefully the instructions here:
> https://help.ubuntu.com/community/BinaryDriverHowto/ATI#WORKAROUND
>
> As per Nick Andrik, this is a version regression of xserver-xorg-
> video-intel from 2.20.2 (working and in below mentioned PPA) and
> 2.20.3. The regression commit is
> http://cgit.freedesktop.org/xorg/driver/xf86-vid...

kesten broughton (dathomir) wrote :

I tried the steps by AG. I'm finding it very hard to navigate as windows won't minimize and i can't alt+tab switch between windows. I have to destroy windows to switch around. Is there a better mode i can switch to (gnome instead of compiz?) while im troubleshooting.

Also wondering if there is a way to test if the installs went ok. I still have low-graphics mode. Boot up is fine. I get the desktop wallpaper. But no launchpad and many windowing issues like those mentioned above.

I started having troubles after 12.10 -> 13.04. I upgraded to 13.10. Still problems.
Running on an intel5 with radeon hd5470.

It seems like the upgrade undid all my work to get crysal fglrx working previously. Perhaps there should be a dialog to check before wiping out fglrx for xorg drivers. I want to do some gpu programming and need the fglrx to work with opencl etc packages.

kesten

Marco Vuano (vuanomarco) wrote :

With Catalyst 13.6 beta drivers (version 2:13.101-0ubuntu1) again this bug seems to be solved (as in previous versions), but also the two other bugs I mentioned (the one similar to bug #1088220 and the one reported as bug 789 on the unofficial ATI Bugzilla) are apparently corrected too. Now, if others don't report other important issues or regressions with these drivers on Ubuntu 12.10, they could be nice drivers to have in Ubuntu's 12.10 repository.

description: updated

Is the new workaround valid for Raring?
With Catalyst 13.6 beta (from AMD's website) I still get the low graphics mode error.

I did some extra testing with Raring. The results are below:

fglrx-updates-2:9.012-0ubuntu1 + xserver-xorg-video-intel-2:2.21.6-0ubuntu4 = low graphics mode

fglrx-updates-2:9.012-0ubuntu1 + xserver-xorg-video-intel-2:2.20.9-0ubuntu2.1 = low graphics mode

*Using the xorg-edgers PPA:
fglrx-13 (2:13.101~beta-0ubuntu1~xedgers~raring1) + xserver-xorg-video-intel-2:2.21.8+git20130529.7f7c8087-0ubuntu0sarvatt~raring = low graphics mode

fglrx-13 (2:13.101~beta-0ubuntu1~xedgers~raring1) + xserver-xorg-video-intel-2:2.20.9-0ubuntu2.1 = Solves low graphics mode problem but has issues with external monitor (with different resolution than the laptop screen).

--
$ lspci|grep VGA
00:02.0 VGA compatible controller: Intel Corporation 3rd Gen Core processor Graphics Controller (rev 09)
02:00.0 VGA compatible controller: Advanced Micro Devices [AMD] nee ATI Thames [Radeon 7550M/7570M/7650M]

Jammy Zhou (jammy-zhou) wrote :

For fglrx-13+xserver-xorg-video-intel-2:2.21.6, can you disable Intel SNA acceleration for a try?

@jammy-zhou

Yes, disabling SNA made fglrx-13 (13.6 beta) + xserver-xorg-video-intel-2:2.21.6 work for me.
There is, however, a corruption problem when using an external monitor with the discrete GPU enabled (same problem I have with fglrx-13 + xserver-xorg-video-intel-2:2.20.9-0ubuntu2.1).
The problem happens only when I'm using both screens (laptop and monitor). Everything seems fine when only one of them is enabled.

Luigi R. (xluigi84) wrote :

How to disable SNA, please?

@xluigi84

You need to have the following in /etc/X11/xorg.conf or in a file inside /etc/X11/xorg.conf.d:

Section "Device"
        Identifier "intel"
        Driver "intel"
        Option "AccelMethod" "uxa"
EndSection

Luigi R. (xluigi84) wrote :

Thanks Rafael!!!. Usually I do not have any xorg.conf or xorg.conf.d. xorg.conf is the one that comes from aticonfig. Should I put those lines in the xorg.conf created by the catalyst driver?

ps: I have tried to install the beta version 13.6 using the user interface but the installation does not start complaining about adapter not supported while I have seen it in the list (HD6470M). Maybe I should switch the discrete card on with vgaswitcheroo before doing it.

Jammy Zhou (jammy-zhou) wrote :

Rafael, which Ubuntu version are you using? And what's the kernel version? For the external corruption you mentioned, some fix is needed in Intel kernel driver, and probably you can try the kernel image below.

http://people.canonical.com/~tjaalton/i915-dkms/release/intel-i915-backport-3.8-dkms_3.8.6.0_all.deb

Hi Jammy,

I'm using Ubuntu Raring. The kernel version is 3.8.0.24.40.
I'll try the new driver later today.

bugbot (bugbot) on 2013-06-06
tags: added: edgers

Hi Jammy,

I am getting an error when I try to install the driver. Here's the output:

$ sudo dpkg -i intel-i915-backport-3.8-dkms_3.8.6.0_all.deb
Selecting previously unselected package intel-i915-backport-3.8-dkms.
(Reading database ... 396179 files and directories currently installed.)
Unpacking intel-i915-backport-3.8-dkms (from intel-i915-backport-3.8-dkms_3.8.6.0_all.deb) ...
Setting up intel-i915-backport-3.8-dkms (3.8.6.0) ...
Loading new intel-i915-backport-3.8-dkms-3.8.6.0 DKMS files...
First Installation: checking all kernels...
Building only for 3.8.0-25-generic
Building for architecture x86_64
Building initial module for 3.8.0-25-generic
ERROR (dkms apport): unable to determine source package for intel-i915-backport-3.8-dkms
Error! Bad return status for module build on kernel: 3.8.0-25-generic (x86_64)
Consult /var/lib/dkms/intel-i915-backport-3.8-dkms/3.8.6.0/build/make.log for more information.

---

$ cat /var/lib/dkms/intel-i915-backport-3.8-dkms/3.8.6.0/build/make.log
DKMS make.log for intel-i915-backport-3.8-dkms-3.8.6.0 for kernel 3.8.0-25-generic (x86_64)
Fri Jun 7 21:47:57 BRT 2013
make -C /lib/modules/3.8.0-25-generic/build M=`pwd` modules
make[1]: Entering directory `/usr/src/linux-headers-3.8.0-25-generic'
  CC [M] /var/lib/dkms/intel-i915-backport-3.8-dkms/3.8.6.0/build/i915_drv.o
In file included from /var/lib/dkms/intel-i915-backport-3.8-dkms/3.8.6.0/build/i915_drv.c:33:0:
/var/lib/dkms/intel-i915-backport-3.8-dkms/3.8.6.0/build/i915_drv.h:760:21: error: field 'stolen' has incomplete type
/var/lib/dkms/intel-i915-backport-3.8-dkms/3.8.6.0/build/i915_drv.h:762:21: error: field 'gtt_space' has incomplete type
make[2]: *** [/var/lib/dkms/intel-i915-backport-3.8-dkms/3.8.6.0/build/i915_drv.o] Error 1
make[1]: *** [_module_/var/lib/dkms/intel-i915-backport-3.8-dkms/3.8.6.0/build] Error 2
make[1]: Leaving directory `/usr/src/linux-headers-3.8.0-25-generic'
make: *** [all] Error 2

Marco Vuano (vuanomarco) wrote :

With the latest OpenGL 4.3 beta drivers (version 2:13.150-0ubuntu1) there isn't any apparent regression regarding the three bugs I mentioned in my previous posts.

Marco Vuano (vuanomarco) wrote :

Basically the bug treated here (#1068404), the one similar to bug #1088220 and the problem with XvBA reported as bug 789 on the unofficial ATI Linux Driver Bugzilla are not present in the OpenGL 4.3 beta drivers.

Are you saying that the latest driver fixes the issues?

Marco Vuano (vuanomarco) wrote :

I would like to point out that, since the bug was originally found on Ubuntu Desktop 12.10 64 bit edition, I am still using this version of Ubuntu to verify if it has been solved, so my report is about the state of the bugs on Ubuntu Desktop 12.10. Users of Ubuntu 13.04 may still find that some of the bugs similar to the ones I previously described are still not solved..l

Jonas Schwabe (jonas-schwabe) wrote :

I installed the new beta of fglrx without installing a non default intel
driver on 13.04 and it works, there are some other bugs in this driver
thought. On high performance mode can't use dual head for example but the
initial bug is fixed.

I can confirm as well. I updated the intel drivers and I have a desktop on 13.04. I'm using an HP Pavilion G4 2123 TX (Intel/AMD hybrid AMD Radeon HD 7600M). I can also confirm that dual head does not work at all (glitchy/flickering screen).

Marco Vuano (vuanomarco) wrote :

With the 13.8 beta drivers (version 2:13.200-0ubuntu1) there isn't any apparent regression regarding the aforementioned three bugs on Ubuntu Desktop 12.10 64-bit. What is the situation with Ubuntu 13.04?

In 13.04, drivers install smoothly. The dual head situation is not working (e.g. glitching/flickering desktop).

Marco Vuano (vuanomarco) wrote :

As expected, these new drivers (whose version is the same: 2:13.200-0ubuntu1) show no significant regressions regarding the solved bugs I talked about. Is the dual head situation the same on Ubuntu 13.04?

It is so. Dual head is still broken... I do not know if this is an Ubuntu issue or an fglrx issue at this point. 12.10 did dual head with fglrx for me prior to me switching to 13.04.

Marco Vuano (vuanomarco) wrote :

Is there a bug regarding this dual head problem on Ubuntu 13.04? If not, it could be useful to file a new bug so that Ubuntu developers (and hopefully also AMD's ones) are correctly informed about this issue.

I filed a bug report on AMD's bug tracker. I don't know what else to do about this as it's hard to describe what is happening.

Marco Vuano (vuanomarco) wrote :

Filing on AMD's bug tracker is the right thing to do, if you want you can also file here on Launchpad (try using Apport/ubuntu-bug).

I absolutely have no experience in this and probably would require some more help in the direction, if possible.

Marco Vuano (vuanomarco) wrote :

If you have the dual head bug with the drivers provided by the fglrx or the fglrx-updates package, the command to file a new bug should be
ubuntu-bug fglrx
(or fglrx-updates)
but if you are experiencing this bug only with the drivers downloaded from AMD's site then I think you should not file a new bug here on Launchpad (I apologize for providing incorrect information before).
Anyway, with the 13.9 drivers (2:13.152-0ubuntu1) there is no significant regression reagarding all the three bugs reported by me, and since they are not beta drivers I think they could be useful if offered through the fglrx-updates package, unless they have other important regressions.

Luigi R. (xluigi84) wrote :

Dear Marco why I can find just 13.8beta on amd website? Where can I get 13.9? I would like to test them. Thanks

Marco Vuano (vuanomarco) wrote :

You can find them using this link: http://www2.ati.com/drivers/linux/amd-catalyst-13.9-linux-x86.x86_64.zip
Sometimes AMD allows downloading their drivers before the official page is up.

Hi Marco,

13.9 is supposedly based on an older fglrx driver according to this post: http://devgurus.amd.com/message/1300002

I will not be filing a bug on Ubuntu as I don't use fglrx-updates package from Ubuntu. Last time I tried those drivers on 12.10, the results were disastrous.

A comment: I could manage the full resolution of my Samsung laptop only with the solution here:

http://askubuntu.com/questions/205112/how-do-i-get-amd-intel-hybrid-graphics-drivers-to-work

(one million thanks). But my main problem was that the open-source driver was not giving me the correct screen resolution, otherwise I would have happily stay there. Is anyone on this bug aware of where I can check that one? I am willing to help with the open source drivers testing, if possible.

Marco Vuano (vuanomarco) wrote :

@Romano Giannetti:The bug treated here is related to the binary Catalyst drivers, if you have the bug you mentioned with the open source drivers of the Ubuntu repositories you can file a new bug here on Launchpad, you can also report bugs about open source drivers on https://bugs.freedesktop.org/.

P.S.: of course, the new Catalyst 13.10 beta drivers work fine regarding the main bug of this report and the other bugs I found.

@Marco: you're right --- wrong place. But if anyone would arrive here on trying to solve a problem similar to mine: I can use the free drivers simply removing the nomodesetting flag from the boot command line (which was necessary to boot the installation USB). Thanks.

13.10 beta and still the same problems. Really considering a downgrade to 12.10...

Marco Vuano (vuanomarco) wrote :

No significant regression also with the 13.11 beta drivers (version 2:13.200-0ubuntu1), as expected.

13.11 beta did not fix the dual monitor issues. Switched back to open source radeon drivers which does work.

could anyone point me to the dual head bug on the amd site? and, did maybe anyone come up with a workaround for it? I am using a laptop with a second screen attached and even if the laptop screen is off the secondary screen does not come out of sleep alive, and I get a black screen after logging into the system...

Marco Vuano (vuanomarco) wrote :

@kpijarski: The bug niniendowarrior talked about is this
http://ati.cchtml.com/show_bug.cgi?id=848
I don't know how to solve your problem. I'm using the fglrx drivers version 13.11 beta version 6 (2:13.250-0ubuntu1 on Ubuntu 12.10 64 bit) and only the secondary monitor to have its full resolution and I'm fine, even when resuming from suspend mode.
Have you checked the cable you are using and if you are outputting a supported mode to your monitor? In case you don't see any supported modes on the "Monitor" window you should use xrandr and create a custom mode.

Anyway, no need to say that the new drivers don't show any regression regarding the main bug here and the other bugs I mentioned.

@kpijarski and @vuanomarco

That bug is the right one. There is no workaround. The developer at AMD I spoke to several months back also expressed skepticism in having a fix for that. The only way I could get the second monitor to work is by using mirror mode which is absolutely useless. I'm now on open source radeon drivers which work but the power management stinks, but what can I do? I have not experienced the waking up issue with the second screen, but for what its worth, I don't think it has anything to do with the cable.

@vuanomarco and @niniendowarior

Until upgrading to 13.10 everything worked fine for me - even with an external monitor attached to my dell inspiron 7520 (with a radeon hd 7700m series card). After the upgrade I got a black screen after logging into unity and was able track it back to this bug: https://bugs.launchpad.net/compiz/+bug/1054257 (compiz opengl plugin crashes with Fatal: glXQueryExtensionsString is NULL for screen 0 message). I have no idea why it suddenly cropped up.
I even reinstalled the entire system to see whether it changes anything, all to no avail.
Right now I am running on an UEFI install with GPT partition table and this last problem might be connected to that (although AMD claims to have added UEFI support to their driver), but I can't be sure, because I had the blank screen problem before reformatting the hard disk.

Marco Vuano (vuanomarco) wrote :

@niniendowarrior:if you want to try the new dynamic power management on the open source drivers you should add
radeon.dpm=1
to the
GRUB_CMDLINE_LINUX_DEFAULT
line on the /etc/default/grub file and then run
sudo update-grub
as specified on https://help.ubuntu.com/community/RadeonDriver

You should probably also set the CPUfreq governor to "Performance" if you need the maximum performance from the open source drivers(you can use the cpufreq-set command or install the "indicator-cpufreq" indicator).

@kpijarski: if you're using the drivers from the official Ubuntu repositories, I suggest you to file a new bug here on Launchpad (using ubuntu-bug should help you somewhat). If you are instead using the drivers from AMD site you should file a bug on the unofficial ATI bugzilla (http://ati.cchtml.com/). I think you should also try the open source drivers, installing the glamor-egl package if it is not already instaled and eventually enabling dynamic power management (follow the instructions on https://help.ubuntu.com/community/RadeonDriver, which should be the same as the one I suggested above), and see if the problem persists.

naught101 (naught101) wrote :

I'm hitting this bug, and I can't get any of the workarounds to work. I'm on Kubuntu Saucy, and I've tried fglrx, fglrx-updates, fglrx from x-edgers (https://launchpad.net/~xorg-edgers/+archive/ppa?field.series_filter=saucy), and three versions of the AMD binary driver (13.4, 13.9, and 13.11 beta). Nothing works, I get segfaults with everything, straight after it tries to load GLX (the segfault byte identifier is different each time, but I guess that could be because of un-related code changes). I didn't try the andrikos version (https://launchpad.net/~andrikos/+archive/ppa) because it's not available for saucy. Any ideas what else I can try?

$ lspci -v|grep VGA
00:02.0 VGA compatible controller: Intel Corporation 2nd Generation Core Processor Family Integrated Graphics Controller (rev 09) (prog-if 00 [VGA controller])
01:00.0 VGA compatible controller: Advanced Micro Devices, Inc. [AMD/ATI] Seymour [Radeon HD 6400M/7400M Series] (prog-if 00 [VGA controller])

$ lsb_release -a
LSB Version: core-2.0-amd64:core-2.0-noarch:core-3.0-amd64:core-3.0-noarch:core-3.1-amd64:core-3.1-noarch:core-3.2-amd64:core-3.2-noarch:core-4.0-amd64:core-4.0-noarch:core-4.1-amd64:core-4.1-noarch
Distributor ID: Ubuntu
Description: Ubuntu 13.10
Release: 13.10
Codename: saucy

$ uname -a
Linux naught101-chronos 3.12.0-031200rc6-generic #201310191635 SMP Sat Oct 19 20:36:43 UTC 2013 x86_64 x86_64 x86_64 GNU/Linux

The attached Xorg.1.log is for the binary version, but I get basically the same error with the xedgers version.

Marco Vuano (vuanomarco) wrote :

@naught101: this bug report originally was referring to Ubuntu 12.10 and was later extended to Ubuntu 13.04, however I don't think it has yet been extended to Ubuntu 13.10 (it is not mentioned in the table above). Anyway, the bug is similar to the one I reported, but by watching the Xorg log file, it seems that the intel driver is not found at all. Since you are using a preliminary version of the 3.12 kernel, I would suggest using the stable 3.12 version. Furthermore, you should use an up-to-date xserver-xorg-video-intel package and mesa package (you can try the ones from oibaf's repository) . Anyway, I suspect that the Catalyst driver only supports the kernel and intel driver versions found in the Ubuntu repositories.

naught101 (naught101) wrote :

So, I removed all my PPA, downgraded to the default saucy kernel, added the oibaf PPA, and re-installed the fglrx-updates drivers (also tried the fglrx package). I'm still getting the same problem.

$ uname -a
Linux naught101-chronos 3.11.0-13-generic #20-Ubuntu SMP Wed Oct 23 07:38:26 UTC 2013 x86_64 x86_64 x86_64 GNU/Linux

$ aptitude show fglrx-updates
Package: fglrx-updates
State: installed
Automatically installed: yes
Version: 2:13.101-0ubuntu3
...

$ aptitude show xserver-xorg-video-intel
Package: xserver-xorg-video-intel
State: installed
Automatically installed: no
Version: 2:2.99.905+git1311070905.b796c3~gd~s
...

$ aptitude show xserver-xorg-video-ati
Package: xserver-xorg-video-ati
State: installed
Automatically installed: no
Version: 1:7.2.99~git20130816.d0323622-0ubuntu0smartboyhw
...

$ aptitude show libgl1-mesa-glx
Package: libgl1-mesa-glx
State: installed
Automatically installed: no
Multi-Arch: same
Version: 10.0~git1311070908.110009~gd~s
...

Timo Aaltonen (tjaalton) on 2013-11-11
Changed in fglrx-installer-updates (Ubuntu):
assignee: Canonical X.org (canonical-x) → nobody
Marco Vuano (vuanomarco) wrote :

@naught101:I would suggest using the default intel drivers and mesa version with the default kernel as well, I think the new drivers should be used with the new kernel you were using, this is why I recommended them in the first place.
If you still have this bug with all the packages from the default Ubuntu repository you can file a new bug on Launchpad.

Marco Vuano (vuanomarco) wrote :

No regressions on the 13.11 beta V9.4 drivers (version 2:13.250-0ubuntu1) as well.

Marco Vuano (vuanomarco) wrote :

In case someone was wondering, no regressions on the 13.12 drivers (version 2:13.251-0ubuntu1) either.

Marco Vuano, thank you for your testing. Could you please execute the following in a terminal and post the results to this report:
lsb_release -rd && apt-cache policy fglrx-*

Marco Vuano (vuanomarco) wrote :

The result is
Description: Ubuntu 12.10
Release: 12.10
xorg-driver-fglrx:
  Installato: (nessuno)
  Candidato: (nessuno)
  Tabella versione:
fglrx-kernel-source:
  Installato: (nessuno)
  Candidato: (nessuno)
  Tabella versione:
fglrx-amdcccle-updates:
  Installato: (nessuno)
  Candidato: 2:12.100.3-0~andrik1
  Tabella versione:
     2:12.100.3-0~andrik1 0
        500 http://ppa.launchpad.net/andrikos/xserver/ubuntu/ quantal/main amd64 Packages
     2:9.000-0ubuntu3 0
        500 http://archive.ubuntu.com/ubuntu/ quantal/restricted amd64 Packages
xorg-driver-fglrx-dev:
  Installato: (nessuno)
  Candidato: (nessuno)
  Tabella versione:
fglrx:
  Installato: 2:13.251-0ubuntu1
  Candidato: 2:13.251-0ubuntu1
  Tabella versione:
 *** 2:13.251-0ubuntu1 0
        100 /var/lib/dpkg/status
     2:9.000-0ubuntu3 0
        500 http://archive.ubuntu.com/ubuntu/ quantal/restricted amd64 Packages
fglrx-driver-dev:
  Installato: (nessuno)
  Candidato: (nessuno)
  Tabella versione:
xfree86-driver-fglrx:
  Installato: (nessuno)
  Candidato: (nessuno)
  Tabella versione:
fglrx-updates:
  Installato: (nessuno)
  Candidato: 2:12.100.3-0~andrik1
  Tabella versione:
     2:12.100.3-0~andrik1 0
        500 http://ppa.launchpad.net/andrikos/xserver/ubuntu/ quantal/main amd64 Packages
     2:9.000-0ubuntu3 0
        500 http://archive.ubuntu.com/ubuntu/ quantal/restricted amd64 Packages
glx-alternative-fglrx:
  Installato: (nessuno)
  Candidato: 0.2.2
  Tabella versione:
     0.2.2 0
        500 http://archive.ubuntu.com/ubuntu/ quantal/multiverse amd64 Packages
fglrx-control-qt2:
  Installato: (nessuno)
  Candidato: (nessuno)
  Tabella versione:
fglrx-updates-dev:
  Installato: (nessuno)
  Candidato: 2:12.100.3-0~andrik1
  Tabella versione:
     2:12.100.3-0~andrik1 0
        500 http://ppa.launchpad.net/andrikos/xserver/ubuntu/ quantal/main amd64 Packages
     2:9.000-0ubuntu3 0
        500 http://archive.ubuntu.com/ubuntu/ quantal/restricted amd64 Packages
fglrx-dev:
  Installato: (nessuno)
  Candidato: 2:9.000-0ubuntu3
  Tabella versione:
     2:9.000-0ubuntu3 0
        500 http://archive.ubuntu.com/ubuntu/ quantal/restricted amd64 Packages
fglrx-glx:
  Installato: (nessuno)
  Candidato: (nessuno)
  Tabella versione:
fglrx-modaliases:
  Installato: (nessuno)
  Candidato: (nessuno)
  Tabella versione:
xfree86-driver-fglrx-dev:
  Installato: (nessuno)
  Candidato: (nessuno)
  Tabella versione:
fglrx-amdcccle:
  Installato: 2:13.251-0ubuntu1
  Candidato: 2:13.251-0ubuntu1
  Tabella versione:
 *** 2:13.251-0ubuntu1 0
        100 /var/lib/dpkg/status
     2:9.000-0ubuntu3 0
        500 http://archive.ubuntu.com/ubuntu/ quantal/restricted amd64 Packages
fglrx-driver:
  Installato: (nessuno)
  Candidato: (nessuno)
  Tabella versione:
fglrx-control:
  Installato: (nessuno)
  Candidato: (nessuno)
  Tabella versione:

Changed in fglrx-installer (Ubuntu Quantal):
status: Confirmed → Fix Released

openSUSE 12.2 is no longer supported. Hence the issue won't be fixed for this product. In case the issue still is reproducable with a supported product (openSUSE 12.3/13.1 at the moment) or with openSUSE:Factory, please feel free to reopen. Thanks.

Changed in xserver-xorg-video-intel (openSUSE):
status: In Progress → Won't Fix
Rolf Leggewie (r0lf) wrote :

quantal has seen the end of its life and is no longer receiving any updates. Marking the quantal task for this ticket as "Won't Fix".

Changed in fglrx-installer-updates (Ubuntu Quantal):
status: Triaged → Won't Fix
job (jeppekdahl) on 2015-04-07
Changed in fglrx-installer-updates (Ubuntu):
assignee: nobody → job (jeppekdahl)
Changed in fglrx-installer (Ubuntu Quantal):
assignee: nobody → job (jeppekdahl)
To post a comment you must log in.
This report contains Public information  Edit
Everyone can see this information.

Other bug subscribers

Remote bug watches

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