Jump to: navigation, search


6,457 bytes added, 10:53, 6 April 2014
Reverted edits by (talk) to last revision by
When the installer fails to produce results, check the directory where the ati catalyst installer has been placed. If it contains international characters or spaces, weird errors will arise. (The issue has been reported to ATI some time ago, but even in the newest releases they seem to like to prefer to conserve the bug)
===== February 15, 2013 - Catalyst 13.1 Install on Ubuntu 10.04 (lucid) =====
Had much the same problems as the poster below, but found another fix:
dh_install: fglrx missing files (arch/#SRCOTHERARCH#/etc/OpenCL/vendors/*.icd), aborting
Solution found at [].
1. Extract files from the installer:
sh --extract amd
2. Edit the packages/Ubuntu/dists/lucid/rules file as follows:
The following line had to be appended after line 69:
SRC_other_arch := x86_64
and the following line had to be appended after line 151:
-e "s|#SRCOTHERARCH#|$(SRC_other_arch)|g" \
3. Create package:
sudo ./ 9.012 --buildpkg Ubuntu/lucid
===== January 22, 2013 - Catalyst 13.1 Install on Ubuntu 11.04 (natty) =====
Just wanted to document my problems with this install and the workaround I came up with.
1. I got the installer from AMD ( and attempted this:
sh --buildpkg Ubuntu/natty
This failed with:
dh_install: fglrx missing files (arch/#SRCOTHERARCH#/etc/OpenCL/vendors/*.icd), aborting
make: *** [binary-arch] Error 2
dpkg-buildpackage: error: debian/rules binary gave error exit status 2
That '#SRCOTHERARCH#' looked a bit suspicious. I found a directory arch/x86_64/etc/OpenCL/vendors with an "icd" file in it. I guessed that was what the installer was looking for.
2. I then extracted the files from the installer:
sh --extract amd
3. Create sym link for the "#SRCOTHERARCH#" directory that the installer is expecting to find:
cd amd/arch
ln -sf x86_64 '#SRCOTHERARCH#'
4. Back up to the installer directory and retried the "buildpkg" step - this time it worked and created the three deb package files. I was then able to continue with the install as normal.
==X Server Related Issues==
This "free sharing" of information seems too good to be true. Like communism. === No highX Server freeze upon lockscreen/lid-resolution video modes available close when Intel iGPU selected===* Versions On a Haswell system with an Intel HD 4400 as the integrated card ina hybrid system X freezes (moveable mouse cursor but either black screen or frozen image of desktop at time of freeze) often when the fglrx driver following 7.12 had problems with video resolutions higher than 1280 x 1024user locks the screen(inactivity or otherwise) or closes the lid. The modes of the xorgdesktop (in 13.conf are simply ignored, 10 at least)can be recovered by switching to a tty and the server starts up back( e.g. ctrl+alt+fn1, followedby ctl+alt+fn7). In 12.04 I needed to restart completely from tty. This bug occurs because pxpress loads the intel driver with the "AccelMethod" "UXA" option. You can verify this is the causeby using a resolution 1280 x 1024 minimal (even if this resolution is not defined in Intel only) `xorg.conf`(even before fglrx drivers are installed) with just   Section "Device" Identifier "Intel Graphics" Driver "intel" Option "AccelMethod" "UXA" # (mode fglrx uses) I can see the bug # Option "AccelMethod" "SNA" #(default). This is particularly no bug! EndSection   and you should reproduce the bug with a problem on LCD displaysfew attempts at closing lid/locking screen.* [ This Whereas with the change to ` Option "AccelMethod" "SNA"` there is no bug] documents anymore ("SNA" is the problem; it is default anyway). '''fixedHow to fix''' since driver version 8 Create (or edit) the DRI config file `~/.2drirc` containing * Looks like it reappeared again some version above ati-  <device screen="0" driver-installer-8-12-x86="dri2"> <application name="Default"> <option name="vblank_mode" value="0"/> </application> </device>  This turns off "vsync" for the Intel card (the mode that attempts to make your gfx card sync its frameratewith that of the display to avoid tearing, rather than racing ahead at its intrinsic higher framerate) . Version ati- Now even when pxpress loads the intel driver-installer-9.2-x86.x86_64.runin "UXA" mode, and has this the bugshould not be reproducible. At least on integrated HD3200 card It would be a better solution if it were possible to tell pxpress to load inteldriver with SNA mode directly.
=== No XVideo support on 690G integrated graphic chipset ===
That seems to be solved when adding ""load "dri"" to "Section Module" of xorg.conf.
=== [SOLVED] RadeonHD6670 & Catalyst 13.4 : X server crash ( segmentation fault after __driCreateNewScreen in ) / KMS working ===
Debian Wheezy + AMD installer
3.2.46 #1 SMP Wed Aug 21 11:01:42 CEST 2013 x86_64 GNU/Linux
ACPI disabled
AMD TURKS firmwares included
Motherboaord ASUS P5KPL - 4GB RAM
Intel E8500 3.16GHz
BUG trace in syslog
ukiDynamicMajor: found major device number 252
[ 304.303] ukiDynamicMajor: found major device number 252
[ 304.303] ukiOpenByBusid: Searching for BusID PCI:2:0:0
[ 304.303] ukiOpenDevice: node name is /dev/ati/card0
[ 304.303] ukiOpenDevice: open result is 12, (OK)
[ 304.303] ukiOpenByBusid: ukiOpenMinor returns 12
[ 304.303] ukiOpenByBusid: ukiGetBusid reports PCI:2:0:0
[ 304.346]
[ 304.346] Backtrace:
[ 304.346] 0: /usr/bin/Xorg (xorg_backtrace+0x36) [0x7faadd5d1d06]
[ 304.346] 1: /usr/bin/Xorg (0x7faadd453000+0x182859) [0x7faadd5d5859]
[ 304.346] 2: /lib/x86_64-linux-gnu/ (0x7faadc77b000+0xf030) [0x7faadc78a030]
[ 304.346] 3: /lib/x86_64-linux-gnu/ (0x7faadb484000+0x11fc45) [0x7faadb5a3c45]
[ 304.346] 4: /usr/lib/dri/ (__driCreateNewScreen_20050727+0x56) [0x7faad542e0c6]
[ 304.346] 5: /usr/lib/xorg/modules/extensions/ (0x7faad9d4c000+0x1e0ea) [0x7faad9d6a0ea]
[ 304.346] 6: /usr/lib/xorg/modules/extensions/ (0x7faad9d4c000+0x1fb05) [0x7faad9d6bb05]
[ 304.346] 7: /usr/bin/Xorg (InitExtensions+0x99) [0x7faadd510889]
[ 304.346] 8: /usr/bin/Xorg (0x7faadd453000+0x41d42) [0x7faadd494d42]
[ 304.346] 9: /lib/x86_64-linux-gnu/ (__libc_start_main+0xfd) [0x7faadb4a2ead]
[ 304.346] 10: /usr/bin/Xorg (0x7faadd453000+0x4219d) [0x7faadd49519d]
[ 304.346]
[ 304.346] Segmentation fault at address 0x7fff00000014
[ 304.346]
Fatal server error:
[ 304.346] Caught signal 11 (Segmentation fault). Server aborting
The solution is : downgrade AMD driver.
I could these AMD driver working:
AMD Catalyst installer (legacy) 13.1 ( installer says version 8.97.2 )
AMD Catalyst installer (non-legacy) 13.1 ( installer says version 9.012 )
-> both seems to be the same, kernel module version reported in Xorg.0.log is always 8.97.2 !
They both sticks "AMD Unsupported hardware" in the low-right corner of my monitor ( which can be removed easily ).
Seems that from version 13.4, the software was deeply modified, that make it buggy.
KMS works well ( module radeon.ko ). I have frame buffer kernel module ( radeonfb.ko ) built
but I didnt try for instant to load it simultaneously with KMS.
==aticonfig Issues==
void ATI_API_CALL KCL_fpu_begin(void)<BR>
<nowiki>#ifdef CONFIG_X86_64<BR/nowiki> kernel_fpu_begin();<BRnowiki>#else</nowiki><BR><nowiki>#ifndef TS_USEDFPU<BR/nowiki> preempt_disable();<BR> if (__thread_has_fpu(current))<BR> __save_init_fpu(current);<BRnowiki>#else<BR/nowiki> struct thread_info *cur_task = current_thread_info();<BR> preempt_disable();<BR> if (cur_task->status & TS_USEDFPU)<BR> __save_init_fpu(cur_task->task);<BRnowiki>#endif<BR/nowiki> else<BR> clts();<BRnowiki>#endif<BR/nowiki>

Navigation menu