| Anonymous | Login | Signup for a new account | 2018-01-04 13:19 MST |
| Main | My View | View Issues | Change Log | Roadmap | Docs |
| Viewing Issue Simple Details [ Jump to Notes ] | [ View Advanced ] [ Issue History ] [ Print ] | ||||||
| ID | Category | Severity | Reproducibility | Date Submitted | Last Update | ||
| 0000547 | [channel: elrepo/el7] kmod-fglrx | major | always | 2015-01-25 12:53 | 2016-02-10 11:09 | ||
| Reporter | ambush | View Status | public | ||||
| Assigned To | wolfy | ||||||
| Priority | normal | Resolution | fixed | ||||
| Status | closed | ||||||
| Summary | 0000547: AIGLX error: failed to open /usr/X11R6/lib64/modules/dri/fglrx_dri.so | ||||||
| Description |
Hi! Thank you for maintaining that drivers! Unfortunately I haven't got luck using package from AMD site so I tried your test packages. I've installed your kmod-fglrx and 32-bit libs from your testing repo. Seems like installation was good but I can't boot into X using that driver. I'm tyred to search that file /usr/X11R6/lib64/modules/dri/fglrx_dri.so but /usr/X11R6/lib and lib64 doesn't contain anything. Xorg log included. |
||||||
| Additional Information | |||||||
| Tags | No tags attached. | ||||||
| Reported upstream | |||||||
| Attached Files |
|
||||||
|
|
|||||||
Notes |
|
|
(0004156) wolfy (developer) 2015-01-26 06:05 |
I am not sure that the error that you've seen ( [ 67.329] (EE) AIGLX error: failed to open /usr/X11R6/lib64/modules/dri/fglrx_dri.so, error[/usr/X11R6/lib64/modules/dri/fglrx_dri.so: cannot open shared object file: No such file or directory ] ) is related to your problem.I have exactly the same error in my logs ( on C6 though ) and the video works just fine, including 3D accel. And , as you can see a bit lower in your logs, DRI is initialized: [ 67.334] ukiOpenDevice: node name is /dev/ati/card0 [ 67.334] ukiOpenDevice: open result is 14, (OK) [ 67.334] ukiOpenByBusid: ukiOpenMinor returns 14 [ 67.334] ukiOpenByBusid: ukiGetBusid reports PCI:1:0:0 [ 67.379] (II) AIGLX: Loaded and initialized OpenGL driver(II) GLX: Initialized DRI GL provider for screen 0 OTOH the kernel command line indicates that stock radeon drivers were not disabled, a succesful & correct instalation of the fglrx-drv package should have appended radeon.modeset=0 rd.driver.blacklist=radeon to the kernel command line. Given the above problem, I'd try to reinstall the fglrx-x11-drv package. Alternatively you could execute manually the sequence below ( it should have been executed in the postinstall sequence, I cannot imagine what happened on your system but I guess that something failed ) : for KERNEL in $(rpm -q --qf '%{v}-%{r}.%{arch}\n' kernel); do VMLINUZ="/boot/vmlinuz-"$KERNEL for KABI in $(find /lib/modules -name radeon.ko | cut -d / -f 4); do if [[ "$KERNEL" == "$KABI" && -e "$VMLINUZ" ]]; then /usr/bin/dracut --add-drivers fglrx -f /boot/initramfs-$KERNEL.img $KERNEL /usr/sbin/grubby --update-kernel="$VMLINUZ" \ --args='radeon.modeset=0 rd.driver.blacklist=radeon' &>/dev/null fi done done |
|
(0004157) ambush (reporter) 2015-01-26 09:22 |
trying to reinstall this driver. yum's error ... Transaction test succeeded Running transaction Установка : kmod-fglrx-14.9-1.el7.elrepo.x86_64 1/4 Working. This may take some time ... Done. WARNING: libglamoregl might conflict with ATI drivers Disable glamoregl by renaming or removing /usr/share/X11/xorg.conf.d/glamor.conf or uninstall xorg-x11-glamor See: http://elrepo.org/tiki/kmod-fglrx [^] (Known Issues) for updated information I've move glamor.conf itno another dir. (I didn't this before) grub.cfg changed disabling radeon driver. will reboot and look. |
|
(0004158) ambush (reporter) 2015-01-26 12:09 |
So I've wait some minutes and I see that log I've sent is from another session. Actually with fglrx X hangs on some minutes and then it crashes. So last attached Xorg.0.log_true is really from fglrx! Excuse me for that mistake. I've also attach messages. Another bug with this driver is it doesn't restores grub.conf after yum remove. |
|
(0004159) wolfy (developer) 2015-01-27 02:09 |
Let's deal with the bug you reported about not restoring grub's configuration after removing the package. The package runs the following code at uninstall time ( but only if there is no other instance of the package left) %preun if [ "${1}" -eq 0 ]; then # Disable the proprietary driver on final uninstall [ -x %{_sbindir}/ati-config-display ] && %{_sbindir}/ati-config-display disable &>/dev/null # Remove init script /sbin/chkconfig --del atieventsd if [ -e /etc/pam.d/amdcccle-su ]; then rm -f /etc/pam.d/amdcccle-su &>/dev/null fi # Clear grub option to disable radeon for all RHEL6 kernels if [ -f %{_sysconfdir}/default/grub ]; then %{__perl} -pi -e 's|(GRUB_CMDLINE_LINUX=.*) radeon\.modeset=0|$1|g' %{_sysconfdir}/default/grub %{__perl} -pi -e 's|(GRUB_CMDLINE_LINUX=.*) rd\.driver\.blacklist=radeon|$1|g' %{_sysconfdir}/default/grub fi if [[ -x /sbin/grubby && -e /boot/grub/grub.conf ]]; then # get installed kernels for KERNEL in $(rpm -q --qf '%{v}-%{r}.%{arch}\n' kernel); do VMLINUZ="/boot/vmlinuz-"$KERNEL if [[ -e "$VMLINUZ" ]]; then /sbin/grubby --update-kernel="$VMLINUZ" \ --remove-args='radeon.modeset=0' &>/dev/null fi done fi # Backup and remove xorg.conf [ -f %{_sysconfdir}/X11/xorg.conf ] && \ mv %{_sysconfdir}/X11/xorg.conf %{_sysconfdir}/X11/xorg.conf.uninstalled-fglrx &>/dev/null fi || : I cannot imagine any reasonable case when grub would not be restored, unless the uninstall scripts are not run, in which case you should see some sort of error message. Now, about your main issue: the Xorg.log you posted shows a "SIG11 (Segmentation fault)" while your log/messages file shows that X died ( which was expected, given the crash ). I am afraid I cannot help with that as it does not look to be a packaging issue. The only thing that I can do is to provide an advanced copy of the new packages I am working on ( release 14.12 of the AMD drivers ). They still have one issue which, for the time being I was not able to solve in a polite manner: one file must be manually "moved out of way". If you are willing to test those, let me know, please |
|
(0004160) ambush (reporter) 2015-01-27 05:34 |
May be way I've removed it can be the reason why grub.cfg wasn't restored? yum remove kmod-fglrx fglrx* Yep, I can try your new packages. |
|
(0004161) wolfy (developer) 2015-01-27 06:49 |
As far as I can see that yum comnmnand is perfectly OK, I fail to understand why grub was not restored to its original configuration. Unless yum faced some error... You can download the test packages from http://wdl.lug.ro/fglrx. [^] Note that they are NOT signed and will be available there for a short time. You will probably need to move /usr/lib64/xorg/modules/extensions/libglx.so out of way ( for instance rename it so that you can easily restore it in case of need ). I have to identify ( and fix ) the reason why the library shipped as /usr/lib64/xorg/modules/extensions/fglrx/libglx.so does not always supersede the one provided by the Xorg server but unfortunately I am once again without the proper hardware to test on... |
|
(0004162) ambush (reporter) 2015-01-27 07:40 |
renamed libgls.so but after installation I see that: [root@tan test_drv]# ls /usr/lib64/xorg/modules/extensions/ fglrx libglx~.~so~ ^^^^^^^^^^^^ this is renamed file. there is nothing instead but I've got new 'fglrx' folder with libglx.so: [root@tan test_drv]# ls /usr/lib64/xorg/modules/extensions/fglrx/ fglrx-libglx.so libglx.so Gonna reboot. |
|
(0004163) wolfy (developer) 2015-01-27 07:44 edited on: 2015-01-27 07:45 |
/usr/lib64/xorg/modules/extensions/fglrx/libglx.so is the one that should be loaded ( as reported in Xorg.0.log). On my test machines sometimes it was loaded, othertimes X preferred to load the library provided by xorg-x11-server-Xorg. Beats me why. And fiddling with ModulePath did not help ( yet ) |
|
(0004166) ambush (reporter) 2015-01-27 14:12 |
Sooo still no luck. messages_test and Xorg.0.log_test added. |
|
(0004167) ambush (reporter) 2015-01-27 14:15 |
I'll try to use drv from site and see what happend with it. Tomorrow. |
|
(0004168) wolfy (developer) 2015-01-27 14:32 |
I am no expert but I am afraid that "Segmentation fault at address 0x0" from the excerpt below indicates a bug in the driver itself. I am all ears if anyone has a solution which I could implement. [ 437.879] (EE) Backtrace: [ 438.058] (EE) 0: /usr/bin/Xorg (xorg_backtrace+0x55) [0x591da5] [ 438.058] (EE) 1: /usr/bin/Xorg (0x400000+0x195e49) [0x595e49] [ 438.058] (EE) 2: /lib64/libc.so.6 (0x7fa7abf3a000+0x35640) [0x7fa7abf6f640] [ 438.058] (EE) 3: /lib64/libpciaccess.so.0 (0x7fa7ad31b000+0x51b2) [0x7fa7ad3201b2] [ 438.058] (EE) 4: /usr/lib64/xorg/modules/drivers/fglrx_drv.so (amd_xs113_int10_x_inb+0x46) [0x7fa7aa2d45e6] [ 438.058] (EE) 5: /usr/lib64/xorg/modules/drivers/fglrx_drv.so (0x7fa7a97df000+0xaee2c5) [0x7fa7aa2cd2c5] [ 438.059] (EE) 6: /usr/lib64/xorg/modules/drivers/fglrx_drv.so (X86EMU_exec+0xa5) [0x7fa7aa2c0a15] [ 438.059] (EE) 7: /usr/lib64/xorg/modules/drivers/fglrx_drv.so (amd_xs113_int10_xf86ExecX86int10+0x46) [0x7fa7aa2d5746] [ 438.059] (EE) 8: /usr/lib64/xorg/modules/drivers/fglrx_drv.so (xf86ExecX86int10+0xd) [0x7fa7a9c684ed] [ 438.059] (EE) 9: /usr/lib64/xorg/modules/libvbe.so (VBESetVBEMode+0x9d) [0x7fa7a918256d] [ 438.059] (EE) 10: /usr/lib64/xorg/modules/drivers/fglrx_drv.so (0x7fa7a97df000+0x4bf963) [0x7fa7a9c9e963] [ 438.059] (EE) 11: /usr/lib64/xorg/modules/drivers/fglrx_drv.so (atiddxVBESetConsoleMode+0x5b) [0x7fa7a9c9e78b] [ 438.060] (EE) 12: /usr/lib64/xorg/modules/drivers/fglrx_drv.so (xdl_xs115_atiddxFreeScreen+0x76f) [0x7fa7a9e3d83f] [ 438.060] (EE) 13: /usr/lib64/xorg/modules/drivers/fglrx_drv.so (xdl_xs115_atiddxCloseScreen+0x321) [0x7fa7a9e3ce31] [ 438.060] (EE) 14: /usr/lib64/xorg/modules/drivers/fglrx_drv.so (0x7fa7a97df000+0xb65206) [0x7fa7aa344206] [ 438.060] (EE) 15: /usr/bin/Xorg (0x400000+0xc9706) [0x4c9706] [ 438.061] (EE) 16: /usr/bin/Xorg (0x400000+0x1135c4) [0x5135c4] [ 438.061] (EE) 17: /usr/bin/Xorg (0x400000+0x114f23) [0x514f23] [ 438.061] (EE) 18: /usr/bin/Xorg (0x400000+0x3c437) [0x43c437] [ 438.061] (EE) 19: /lib64/libc.so.6 (__libc_start_main+0xf5) [0x7fa7abf5baf5] [ 438.061] (EE) 20: /usr/bin/Xorg (0x400000+0x26801) [0x426801] [ 438.061] (EE) [ 438.061] (EE) Segmentation fault at address 0x0 [ 438.061] (EE) Fatal server error: [ 438.061] (EE) Caught signal 11 (Segmentation fault). Server aborting |
|
(0004673) wolfy (developer) 2016-02-10 11:09 |
fglrx-15.11 is now available in the ElRepo testing repository. The error reported initially ( AIGLX error: failed to open /usr/X11R6/lib64/modules/dri/fglrx_dri.so ) is benign and can be ignored. The ""Segmentation fault at address 0x0" should be solved. Please open a new bug if the problem reappears. |
| Mantis 1.1.8[^] Copyright © 2000 - 2009 Mantis Group |