ATI Catalyst

Wildy · 2011-08-17 15:06:36

Hello all, I'm having a few problems getting the ATI drivers working.

I'm using a fully up-to-date system with the 3.0 kernel and a HD 5670 GPU. I was able to get catalyst-utils from the AUR built and installed fine, but when trying to then build catalyst from AUR it comes up with a build() error referring to line 59 of the PKGBUILD which appears to be this part here:

  ## makefile patch to choose arch_compat_alloc_user_space or older compat_alloc_user_space
   ## works only in combination with ati_make.sh script
    patch -Np1 -i ../makefile_compat.patch || return 1

Is it something to do with the 3.0 kernel I'm using?

Any help would be appreciated. Thanks.

Unia · 2011-08-17 15:21:07

This won't actually solve your problem, but it might be interesting for you: why not use the catalyst repo?
Scroll down until you see https://wiki.archlinux.org/index.php/Ca … ock_kernel; the letters are small, though!
https://wiki.archlinux.org/index.php/Ca … ock_kernel

EDIT: I should add that I use it myself and find it much easier than using AUR. It also saves time because you don't have to compile everything.

Last edited by Unia (2011-08-17 15:21:50)

Wildy · 2011-08-17 16:25:59

Thanks for the suggestion, I hadn't spotted that. Well they both installed fine and I generated an Xorg conf. But when starting up X the screen goes white like it is starting up TWM and then goes black and drops the output. Upon inspection of the log file I get:

(EE) fglrx(0): Not enough video memory to allocate primary surface (frame buffer).

[EDIT] When evoking fglrxinfo it would appear there is no display - possibly an xorg.conf problem?

Last edited by Wildy (2011-08-17 16:29:29)

Varg · 2011-08-17 16:55:29

Did you run:

aticonfig --initial

After installing catalyst ?
https://wiki.archlinux.org/index.php/AT … positories

Wildy · 2011-08-17 16:57:55

Yes I did, and I double-checked and my config matches a typical ATI config. I've also managed to install amdstream successfully, but on running the samples it defaults to the CPU as it can't find the GPU - I am guessing the two problems are related. Are there any amdstream log files which might help identify this problem?

[EDIT] Uploading /var/log/Xorg.0.log...
Scratch that, wget is not working.

Last edited by Wildy (2011-08-17 17:14:17)

_artem_ · 2011-08-18 13:30:50

I'm having problems compiling the catalyst kernel module... on kde.
driver seams to be installed (see no errors), but after that console output looks like:
~~~~~~~~
~~~~~~~~~
~~~~~~~ ~~~~~~~~~
broken tabulators... as it should look like:
~~~~~~~~~
~~~~
~~~~~~
but it doesn't. reboot, after kernel message: Starting KDM ........ [ ] I see only black screen.
trying to start kdm from the console tells me that I have no display or something like that...
catalyst from aur compiles without any issues.
I always installed catalyst driver from the *.run script but now I can't do that any more...
(I'm fresh on kde, always used gnome, gnome-shell and never had any compilation problems, but now on kde it just finishes the installation but there is no kernel module)
the only thing it tells me when I start the *.run script is that I'm using glibc 2.14 and OS is unknown (WTF?! o_O)
according to install instructions pdf catalyst needs glibc 2.2 and 2.3 I found them on gnu.org ftp and they are over years old. the latest is 2.14 but catalyst pdf tells me it needs 2.2-2.3 version.

p.s. on the latest 11.8 driver I used that new amdconfig like I always did with aticonfig it successfully generates x config but there is no kernel module. I don't know why, had never such problems before I switched to kde

Last edited by _artem_ (2011-08-18 13:32:39)

nobody44 · 2011-08-25 20:51:10

You should use the Arch Build System:

tar xf catalyst-utils.tar.gz
cd catalyst-utils
makepkg
su
pacman -U catalyst-xxxxxxxxxx.tar.xz

@Wildy:
Could you solve the problem? I have the same problem... a bit different, but very similar:
I have a HD5750, the newest linux kernel and the newest X-Server (started by SLiM Login-Manager). Sometimes, X works very well, without any problems, and sometimes it leaves the screen completely black. I think it's a kernel panic, because i can not switch to other terminals with Alt+F1 and so on. The error message is the same, I get even more:

errors.log:
Aug 26 02:29:39 localhost kernel: [  133.807567] [fglrx:KCL_MEM_VM_MapRegion] *ERROR* remap_pfn_range failed
Aug 26 02:29:39 localhost kernel: [  133.807574] [fglrx:__create_mapping] *ERROR* Can not get virtual address
Aug 26 02:29:39 localhost kernel: [  133.807576] [fglrx:__mc_heap_map_virtual_space] *ERROR* Failed to map the virtual space
Aug 26 02:29:39 localhost kernel: [  133.807578] [fglrx:mc_heap_map_virtual_space] *ERROR* Can not get virtual address
Aug 26 02:29:39 localhost kernel: [  133.807579] [fglrx:MCIL_GetVirtualAddressInDescriptor] *ERROR* Can not get the virtual address
Aug 26 02:29:39 localhost kernel: [  133.846004] BUG: scheduling while atomic: X/1498/0x00000002
Aug 26 02:29:39 localhost kernel: [  133.850733] BUG: scheduling while atomic: X/1498/0x00000002
Aug 26 02:29:39 localhost kernel: [  133.855391] BUG: scheduling while atomic: X/1498/0x00000002

And kernel.log:

Aug 26 02:29:39 localhost kernel: [  133.850733] BUG: scheduling while atomic: X/1498/0x00000002
Aug 26 02:29:39 localhost kernel: [  133.853789] Modules linked in: ipv6 ext3 jbd joydev usbhid hid snd_hda_codec_hdmi vboxdrv cpufreq_ondemand powernow_k8 freq_table mperf snd_hda_codec_realtek evdev ppdev ohci_hcd firewire_ohci serio_raw pcspkr firewire_core edac_core k10temp edac_mce_amd crc_itu_t snd_hda_intel snd_hda_codec sg snd_hwdep radeon ehci_hcd snd_pcm ttm fglrx(P) drm_kms_helper parport_pc drm i2c_piix4 sp5100_tco snd_timer snd i2c_algo_bit r8169 usbcore soundcore parport snd_page_alloc wmi floppy i2c_core mii processor button ext4 mbcache jbd2 crc16 pata_acpi pata_atiixp sr_mod cdrom sd_mod ahci libahci libata scsi_mod
Aug 26 02:29:39 localhost kernel: [  133.854026] Pid: 1498, comm: X Tainted: P        W   3.0-ARCH #1
Aug 26 02:29:39 localhost kernel: [  133.854035] Call Trace:
Aug 26 02:29:39 localhost kernel: [  133.854054]  [<ffffffff813ea533>] __schedule_bug+0x5f/0x64
Aug 26 02:29:39 localhost kernel: [  133.854068]  [<ffffffff813f09e1>] schedule+0x831/0x9f0
Aug 26 02:29:39 localhost kernel: [  133.854154]  [<ffffffffa0191362>] ? firegl_trace+0x72/0x1e0 [fglrx]
Aug 26 02:29:39 localhost kernel: [  133.854170]  [<ffffffff81045128>] ? __wake_up_common+0x58/0x90
Aug 26 02:29:39 localhost kernel: [  133.854249]  [<ffffffffa0191362>] ? firegl_trace+0x72/0x1e0 [fglrx]
Aug 26 02:29:39 localhost kernel: [  133.854266]  [<ffffffff81046e13>] ? __wake_up+0x53/0x70
Aug 26 02:29:39 localhost kernel: [  133.854282]  [<ffffffff8107f470>] ? prepare_to_wait+0x60/0x90
Aug 26 02:29:39 localhost kernel: [  133.854345]  [<ffffffffa016655a>] KAS_Thread_WaitForFinish+0x8a/0xc0 [fglrx]
Aug 26 02:29:39 localhost kernel: [  133.854360]  [<ffffffff8107f680>] ? abort_exclusive_wait+0xb0/0xb0
Aug 26 02:29:39 localhost kernel: [  133.854433]  [<ffffffffa017d527>] MCIL_WorkerThread+0x177/0x230 [fglrx]
Aug 26 02:29:39 localhost kernel: [  133.854507]  [<ffffffffa017ebc6>] ? MCIL_GetExecutionLevel+0x66/0x80 [fglrx]
Aug 26 02:29:39 localhost kernel: [  133.854604]  [<ffffffffa02359e1>] ? PassiveRing_Cleanup+0x61/0xc0 [fglrx]
Aug 26 02:29:39 localhost kernel: [  133.854689]  [<ffffffffa0244628>] DispatchRing_Cleanup+0x48/0x70 [fglrx]
Aug 26 02:29:39 localhost kernel: [  133.854704]  [<ffffffff81141a03>] ? kfree+0x163/0x180
Aug 26 02:29:39 localhost kernel: [  133.854717]  [<ffffffff810bffb6>] ? free_irq+0x56/0xb0
Aug 26 02:29:39 localhost kernel: [  133.854811]  [<ffffffffa0235020>] ? IRQMGR_FreeResources+0x20/0x60 [fglrx]
Aug 26 02:29:39 localhost kernel: [  133.854886]  [<ffffffffa017ffb5>] ? irqmgr_wrap_shutdown+0x125/0x180 [fglrx]
Aug 26 02:29:39 localhost kernel: [  133.854954]  [<ffffffffa01719c6>] ? firegl_takedown+0x7d6/0x8a0 [fglrx]
Aug 26 02:29:39 localhost kernel: [  133.855034]  [<ffffffffa024b442>] ? amdPcsEngine_InitDatabase+0x12/0x20 [fglrx]
Aug 26 02:29:39 localhost kernel: [  133.855103]  [<ffffffffa0170ad5>] ? firegl_release+0x125/0x1c0 [fglrx]
Aug 26 02:29:39 localhost kernel: [  133.855161]  [<ffffffffa0161be8>] ? ip_firegl_fasync+0x78/0x90 [fglrx]
Aug 26 02:29:39 localhost kernel: [  133.855221]  [<ffffffffa0161ee1>] ? ip_firegl_release+0x11/0x20 [fglrx]
Aug 26 02:29:39 localhost kernel: [  133.855235]  [<ffffffff811591ca>] ? fput+0xda/0x210
Aug 26 02:29:39 localhost kernel: [  133.855248]  [<ffffffff81155a66>] ? filp_close+0x66/0x90
Aug 26 02:29:39 localhost kernel: [  133.855262]  [<ffffffff81060000>] ? put_files_struct+0xa0/0x110
Aug 26 02:29:39 localhost kernel: [  133.855276]  [<ffffffff81060132>] ? exit_files+0x52/0x60
Aug 26 02:29:39 localhost kernel: [  133.855288]  [<ffffffff8106063a>] ? do_exit+0x18a/0x8f0
Aug 26 02:29:39 localhost kernel: [  133.855301]  [<ffffffff81175e7f>] ? mntput+0x1f/0x30
Aug 26 02:29:39 localhost kernel: [  133.855314]  [<ffffffff8115925f>] ? fput+0x16f/0x210
Aug 26 02:29:39 localhost kernel: [  133.855326]  [<ffffffff810610f4>] ? do_group_exit+0x44/0xa0
Aug 26 02:29:39 localhost kernel: [  133.855337]  [<ffffffff81061167>] ? sys_exit_group+0x17/0x20
Aug 26 02:29:39 localhost kernel: [  133.855351]  [<ffffffff813f4402>] ? system_call_fastpath+0x16/0x1b

I think these errors are related, but i might be wrong...
This error repeats about 5 times, then it stops.

Anyone knows how to fix it?

Thank you very much

Paul

Arch Linux

#1 2011-08-17 15:06:36

ATI Catalyst

#2 2011-08-17 15:21:07

Re: ATI Catalyst

#3 2011-08-17 16:25:59

Re: ATI Catalyst

#4 2011-08-17 16:55:29

Re: ATI Catalyst

#5 2011-08-17 16:57:55

Re: ATI Catalyst

#6 2011-08-18 13:30:50

Re: ATI Catalyst

#7 2011-08-25 20:51:10

Re: ATI Catalyst

Board footer