The AMD/ATI Bar & Grill

snack · 2014-05-01 12:53:16

@doblerone: your system is too much up-to-date, catalyst-hd234k requires xorg-server<1.13.0 as explained in the pacman output. Use the xorg-112 repo.

doblerone · 2014-05-01 13:02:20

snack wrote:

@doblerone: your system is too much up-to-date, catalyst-hd234k requires xorg-server<1.13.0 as explained in the pacman output. Use the xorg-112 repo.

I think I'm using xorg-112 because I've installed catalyst-total-hd234k as it is explained in the wiki in order to install catalyst drivers.

Vi0L0 · 2014-05-01 13:06:24

ChemBro wrote:

Caldazar wrote:
EDIT: Googling "fstack-protector-strong" revealed that it is a gcc 4.9 feature whereas I am on outdated gcc 4.8.2-8.
4.9 is on testing, so I hope it comes soon. I think I rather wait than switching to the testing repo.
You're right. I opened up testing only to install gcc (and multilib; and I know you shouldn't do it) and that fixed that.

Wait, wha? Why are you guys using this option anyway? Is it set as some global cflag?

doblerone wrote:

[david@archdavid:~]$ sudo pacman -S virtualbox-guest-utils

The question is why you are trying to install this package anyway? Is your system a guest of a hypervisor?

snack · 2014-05-01 13:07:37

doblerone wrote:

snack wrote:
@doblerone: your system is too much up-to-date, catalyst-hd234k requires xorg-server<1.13.0 as explained in the pacman output. Use the xorg-112 repo.
I think I'm using xorg-112 because I've installed catalyst-total-hd234k as it is explained in the wiki in order to install catalyst drivers.

I think you're using xorg-server>=1.13.0, as I can estimate from the pacman output. Moreover, installing catalyst-total-hd234k requires adding [catalyst-hd234k] to pacman.conf because the package is hosted in that repository, but this does not enforce also the usage of [xorg-112]. Are you sure you put the [xorg-112] repo above [extra] in /etc/pacman.conf? Please check this, and eventually post the output of: pacman -Qi xorg-server.

@Vi0l0: given the fact that [xorg-112] is used only for [catalyst-hd234k] compatibility, what do you think about merging [xorg-112] into [catalyst-hd234k]? That would save you some maintenance time and avoid this kind of problem...

doblerone · 2014-05-01 13:10:14

Vi0L0 wrote:

doblerone wrote:
[david@archdavid:~]$ sudo pacman -S virtualbox-guest-utils
The question is why you are trying to install this package anyway? Is your system a guest of a hypervisor?

Hi Vi0L0,

I'm trying to install it to be able to share folders between my host (Archlinux) and my guest (VM with Windows 7).
At least, reading the wiki, this is what I understood.

Is there any other way to do this?

@snack : this is part of my /etc/pacman.conf file

#[testing]
#Include = /etc/pacman.d/mirrorlist

[catalyst-hd234k]
Server = http://catalyst.wirephire.com/repo/catalyst-hd234k/$arch

[xorg112]
Server = http://catalyst.wirephire.com/repo/xorg112/$arch

[core]
Include = /etc/pacman.d/mirrorlist

[extra]
Include = /etc/pacman.d/mirrorlist

is it correct, right?

[david@archdavid:~]$ sudo pacman -Qi xorg-server
Nombre            : xorg-server
Versión           : 1.12.4-3
Descripción       : Xorg X server
Arquitectura      : x86_64
URL               : http://xorg.freedesktop.org
Licencias         : custom
Grupos            : Nada
Provee            : xorg-server=1.12.4  X-ABI-VIDEODRV_VERSION=12
                    X-ABI-XINPUT_VERSION=16  X-ABI-EXTENSION_VERSION=6.0
                    x-server=1.12.4
Depende de        : libxdmcp  libxfont  libpciaccess  libdrm  pixman  libgcrypt
                    libxau  xorg-server-common  xf86-input-evdev
Dep. opcionales   : Nada
Requerido por     : catalyst-total-hd234k
Opcional Para   : Nada
En conflicto con    : xorg-server
Reemplaza a         : Nada
Tamaño instalado : 3819,00 KiB
Empaquetador      : Vi0L0 <vi0l093@gmail.com>
Fecha de compilación : dom 05 ene 2014 13:14:58 CET
Fecha de instalación : sáb 26 abr 2014 21:13:22 CEST
Motivo de la instalación: Instalado como dependencia de otro paquete
Script de instalación: No
Validado Por   : Firma

Last edited by doblerone (2014-05-01 13:14:00)

Vi0L0 · 2014-05-01 14:24:42

snack wrote:

@Vi0l0: given the fact that [xorg-112] is used only for [catalyst-hd234k] compatibility, what do you think about merging [xorg-112] into [catalyst-hd234k]? That would save you some maintenance time and avoid this kind of problem...

I do like the way it looks now, it's logical. Beside it wouldn't help with this kind of problem.

doblerone wrote:

I'm trying to install it to be able to share folders between my host (Archlinux) and my guest (VM with Windows 7).
At least, reading the wiki, this is what I understood.

Ok, the problem is that virtualbox-guest-utils depends on X-ABI-VIDEODRV_VERSION=15.
To solve it you would have to get source of the package (pkgbase=virtualbox), via abs for example or www, change this dependency to X-ABI-VIDEODRV_VERSION=12 (or remove it completelly ;P) and build (compile) the package against xserver 1.12, which you have installed.

AndreaConte · 2014-05-01 14:31:25

Hi!
I've just updated to latest catalyst driver provided by Vi0l0 repository with the --force command as i read in this thread.
I have also catalyst hook installed in my system that now is giving this error when he tries to recompile fglrx module

+ building fglrx module for 3.14.2-1-ARCH kernel ...
Failed!!! Check out log: /var/log/catalyst-install.log

this is my catalyst-install.log

$ cat /var/log/catalyst-install.log

--------
2014-05-01 16:19:38
Building fglrx module for 3.14.2-1-ARCH kernel ...
--------
OsVersion says: SMP=1
file /proc/kallsyms says: SMP=1
file /usr/lib/modules/3.14.2-1-ARCH/build/include/generated/autoconf.h says: SMP=1
file /usr/lib/modules/3.14.2-1-ARCH/build/include/generated/autoconf.h says: MODVERSIONS=1                                                                                                                 
file /usr/lib/modules/3.14.2-1-ARCH/build/arch/x86/include/asm/compat.h says: COMPAT_ALLOC_USER_SPACE=arch_compat_alloc_user_space                                                                         
make: ingresso nella directory "/usr/lib/modules/3.14.2-1-ARCH/build"                                                                                                                                      
Makefile:615: Cannot use CONFIG_CC_STACKPROTECTOR_STRONG: -fstack-protector-strong not supported by compiler                                                                                               
  CC [M]  /tmp/catalyst.F5vjGM/firegl_public.o                                                                                                                                                             
gcc: error: unrecognized command line option ‘-fstack-protector-strong’                                                                                                                                    
scripts/Makefile.build:308: set di istruzioni per l'obiettivo "/tmp/catalyst.F5vjGM/firegl_public.o" non riuscito                                                                                          
make[1]: *** [/tmp/catalyst.F5vjGM/firegl_public.o] Errore 1                                                                                                                                               
Makefile:1274: set di istruzioni per l'obiettivo "_module_/tmp/catalyst.F5vjGM" non riuscito                                                                                                               
make: *** [_module_/tmp/catalyst.F5vjGM] Errore 2                                                                                                                                                          
make: uscita dalla directory "/usr/lib/modules/3.14.2-1-ARCH/build"

i think my problems are inside "gcc: error: unrecognized command line option ‘-fstack-protector-strong’ " but i don't know how to get rid of this error. Anyone?

EDIT:
I noticed that -fstack-protector-strong is only for gcc 4.9 and it is provided only in testing archlinux
so I activated the testing repo, made

# pacman -S libtool gcc

and then just

# pacman -R catalyst-hook
# pacman -S catalyst-hook

to let catalyst-hook proc and finally fglrx was compiled

Last edited by AndreaConte (2014-05-01 14:42:11)

doblerone · 2014-05-01 14:33:58

Vi0L0 wrote:

doblerone wrote:
I'm trying to install it to be able to share folders between my host (Archlinux) and my guest (VM with Windows 7).
At least, reading the wiki, this is what I understood.
Ok, the problem is that virtualbox-guest-utils depends on X-ABI-VIDEODRV_VERSION=15.
To solve it you would have to get source of the package (pkgbase=virtualbox), via abs for example or www, change this dependency to X-ABI-VIDEODRV_VERSION=12 (or remove it completelly ;P) and build (compile) the package against xserver 1.12, which you have installed.

One more time, thanks again Vi0L0!!!

Caldazar · 2014-05-01 14:41:27

Vi0L0 wrote:

ChemBro wrote:
Caldazar wrote:
EDIT: Googling "fstack-protector-strong" revealed that it is a gcc 4.9 feature whereas I am on outdated gcc 4.8.2-8.
4.9 is on testing, so I hope it comes soon. I think I rather wait than switching to the testing repo.
You're right. I opened up testing only to install gcc (and multilib; and I know you shouldn't do it) and that fixed that.
Wait, wha? Why are you guys using this option anyway? Is it set as some global cflag?

Reading the makefile
('/usr/lib/modules/3.14.2-1-ARCH/build' Makefile:615)

# Handle stack protector mode.
ifdef CONFIG_CC_STACKPROTECTOR_REGULAR
  stackp-flag := -fstack-protector
  ifeq ($(call cc-option, $(stackp-flag)),)
    $(warning Cannot use CONFIG_CC_STACKPROTECTOR_REGULAR: \
             -fstack-protector not supported by compiler)
  endif
else
ifdef CONFIG_CC_STACKPROTECTOR_STRONG
  stackp-flag := -fstack-protector-strong
  ifeq ($(call cc-option, $(stackp-flag)),)
    $(warning Cannot use CONFIG_CC_STACKPROTECTOR_STRONG: \
	      -fstack-protector-strong not supported by compiler)
  endif

the problem is CONFIG_CC_STACKPROTECTOR_STRONG being set instead of CONFIG_CC_STACKPROTECTOR_REGULAR.
That led me to this thread, "3.14.2-1 and CONFIG_CC_STACKPROTECTOR_STRONG".

Last edited by Caldazar (2014-05-01 20:05:47)

mid-kid · 2014-05-01 14:51:55

For people with the -fstack-protector-strong error: There's a bug report. I hope it'll get fixed soon.
In the meantime, either switch to [testing], or downgrade your kernel. You'll have to wait for gcc-4.9 to hit [core].

Last edited by mid-kid (2014-05-01 14:54:08)

ChemBro · 2014-05-01 16:52:45

Vi0L0 wrote:

ChemBro wrote:
Caldazar wrote:
EDIT: Googling "fstack-protector-strong" revealed that it is a gcc 4.9 feature whereas I am on outdated gcc 4.8.2-8.
4.9 is on testing, so I hope it comes soon. I think I rather wait than switching to the testing repo.
You're right. I opened up testing only to install gcc (and multilib; and I know you shouldn't do it) and that fixed that.
Wait, wha? Why are you guys using this option anyway? Is it set as some global cflag?

I don't use any option. I updated the kernel today and after that, I wasn't able to use "catalyst_build_module all" until I did the gcc update.

Inkaine · 2014-05-01 18:56:16

mid-kid wrote:

For people with the -fstack-protector-strong error: There's a bug report. I hope it'll get fixed soon.
In the meantime, either switch to [testing], or downgrade your kernel. You'll have to wait for gcc-4.9 to hit [core].

Thanks for the headsup. I was getting the very same problem during my update a few minutes again. Downgrading solved that.

@Violo

Maybe you should update the PKGBUILD to reflect that and increase the required version off gcc >= 4.9?

Vi0L0 · 2014-05-01 19:06:19

Inkaine wrote:

@Violo
Maybe you should update the PKGBUILD to reflect that and increase the required version off gcc >= 4.9?

and force users to use [testing]? Beside that would be a lie since catalyst works perfectly fine with gcc 4.8. This bug came from an Arch dev and imho should be fixed in the arch repo asap.
BTW: wouldn't be safer to downgrade linux and linux-headers to 3.14.1?:

x86_64:
ftp://seblu.net/archlinux/arm/2014/04/2 … pkg.tar.xz
ftp://seblu.net/archlinux/arm/2014/04/2 … pkg.tar.xz

i686:
ftp://seblu.net/archlinux/arm/2014/04/2 … pkg.tar.xz
ftp://seblu.net/archlinux/arm/2014/04/2 … pkg.tar.xz

Last edited by Vi0L0 (2014-05-01 19:24:12)

Joomla12 · 2014-05-02 02:17:56

I'm still experiencing refresh issues. It's really only an annoyance because windows are very jittery when being moved, there's tearing in videos every once in a while, and screenshots still don't capture what's currently on the screen. Everything is installed properly. I'm using "catalyst-total" from the AUR.

~ » fglrxinfo                                                                                               
display: :0.0  screen: 0
OpenGL vendor string: Advanced Micro Devices, Inc.
OpenGL renderer string: AMD Radeon HD 6520G
OpenGL version string: 4.4.12874 Compatibility Profile Context 13.251

Last edited by Joomla12 (2014-05-02 09:15:04)

Steef435 · 2014-05-03 11:16:27

For those who haven't noticed yet, gcc 4.9 is now in [core] (thanks package ninjas) and the module builds like a charm.

Last edited by Steef435 (2014-05-03 11:17:19)

teozkr · 2014-05-03 13:39:01

Just did an update from 14.2 to 14.4, as well as replaced my old Radeon HD 6870 with a Radeon R9 270X, and now X seems to segfault on launch.

xorg.conf
Xorg.0.log

Also, running fglrxinfo seems to return

Error: unable to open display (null)

Last edited by teozkr (2014-05-03 13:40:40)

Vi0L0 · 2014-05-03 19:16:04

But you do know that gdm doesn't work?
fglrxinfo won't run without opened X session.
Check lsmod, is there fglrx and not radeon?
xorg.conf looks fine, was it created after gfx update?
You may just in case remove /etc/ati/amdpcsdb, it would be created anyway.

teozkr · 2014-05-03 20:38:10

Vi0L0 wrote:

But you do know that gdm doesn't work?

KDM doesn't work, startx doesn't work. I assumed GDM doesn't work either, but can give it a shot.

Vi0L0 wrote:

fglrxinfo won't run without opened X session.

Well, that explains that mystery.

Vi0l0 wrote:

Check lsmod, is there fglrx and not radeon?

fglrx is listed, will check for radeon. I blacklisted radeon though, and no longer get those "UMS unavailable" warnings.

Vi0l0 wrote:

Xorg.conf looks fine, was it created after gfx update?

Yes. It was regenerated by aticonfig.

Vi0l0 wrote:

You may just in case remove /etc/ati/amdpcsdb, it would be created anyway.

Tried this, it did nothing.

teozkr · 2014-05-03 20:52:00

Just checked. GDM doesn't work either, radeon isn't listed by lsmod.

Joomla12 · 2014-05-03 21:00:35

Huh...Thought I had already posted this, but whatever. I fixed my lag issues by tweaking compton a bit. Using it for tear free and v-sync handling instead of amdcccle, and animations and whatnot are smooth. Still having an issue with screenshot tools capturing what's not there, though. Seems to have gotten worse since I worked on compton.

http://i.imgur.com/eBmcMjV.png

Definitely shouldn't be capturing SLiM, heh.

brom · 2014-05-04 05:06:51

Was experiencing the exact same problem with GDM and tried using SLiM as a workaround. Also, downgrading to gdm-3.10.0.1-2-x86_64 seems to work, but it is incredibly slow at boot, as well as when restarting the display manager. Has anyone found a solution for this problem yet?

Enverex wrote:

ameer wrote:

Thanks, but that didn't work for me either, I still have the seg fault with other error in the logs. Worth mentioning that I have hybrid graphic with both Intel HD 4400 and AMD Radeon HD 8850M. The error I have is when I downgraded to GDM and libGDM 3.10:
With the latest update I still have seg error with AMD.

Apr 28 22:29:50 encrypted gdm-Xorg-:0[3551]: (EE) Backtrace:
Apr 28 22:29:50 encrypted gdm-Xorg-:0[3551]: (EE) 0: /usr/bin/Xorg (xorg_backtrace+0x48) [0x584b08]
Apr 28 22:29:50 encrypted gdm-Xorg-:0[3551]: (EE) 1: /usr/bin/Xorg (0x400000+0x1887f9) [0x5887f9]
Apr 28 22:29:50 encrypted gdm-Xorg-:0[3551]: (EE) 2: /usr/lib/libpthread.so.0 (0x7f79a6c38000+0xf880) [0x7f79a6c47880]
Apr 28 22:29:50 encrypted gdm-Xorg-:0[3551]: (EE) 3: /usr/lib/xorg/modules/drivers/fglrx_drv.so (0x7f79a0b16000+0x40a7b4) [0x7f79a0f207b4]
Apr 28 22:29:50 encrypted gdm-Xorg-:0[3551]: (EE) 4: /usr/lib/xorg/modules/drivers/fglrx_drv.so (0x7f79a0b16000+0x40a34f) [0x7f79a0f2034f]
Apr 28 22:29:50 encrypted gdm-Xorg-:0[3551]: (EE) 5: /usr/bin/Xorg (0x400000+0x66ecc) [0x466ecc]
Apr 28 22:29:50 encrypted gdm-Xorg-:0[3551]: (EE) 6: /usr/bin/Xorg (xf86LoadModules+0xba) [0x47684a]
Apr 28 22:29:50 encrypted gdm-Xorg-:0[3551]: (EE) 7: /usr/bin/Xorg (InitOutput+0x7bf) [0x47707f]
Apr 28 22:29:50 encrypted gdm-Xorg-:0[3551]: (EE) 8: /usr/bin/Xorg (0x400000+0x398bb) [0x4398bb]
Apr 28 22:29:50 encrypted gdm-Xorg-:0[3551]: (EE) 9: /usr/lib/libc.so.6 (__libc_start_main+0xf5) [0x7f79a58afb05]
Apr 28 22:29:50 encrypted gdm-Xorg-:0[3551]: (EE) 10: /usr/bin/Xorg (0x400000+0x2507e) [0x42507e]
Apr 28 22:29:50 encrypted gdm-Xorg-:0[3551]: (EE)
Apr 28 22:29:50 encrypted gdm-Xorg-:0[3551]: (EE) Segmentation fault at address 0x0
Apr 28 22:29:50 encrypted gdm-Xorg-:0[3551]: (EE)
Apr 28 22:29:50 encrypted gdm-Xorg-:0[3551]: Fatal server error:
Apr 28 22:29:50 encrypted gdm-Xorg-:0[3551]: (EE) Caught signal 11 (Segmentation fault). Server aborting
Apr 28 22:29:50 encrypted gdm-Xorg-:0[3551]: (EE)
Apr 28 22:29:50 encrypted gdm-Xorg-:0[3551]: (EE)
Apr 28 22:29:50 encrypted gdm-Xorg-:0[3551]: Please consult the The X.Org Foundation support
Apr 28 22:29:50 encrypted gdm-Xorg-:0[3551]: at http://wiki.x.org
Apr 28 22:29:50 encrypted gdm-Xorg-:0[3551]: for help.
Apr 28 22:29:50 encrypted gdm-Xorg-:0[3551]: (EE) Please also check the log file at "/dev/null" for additional information.
Apr 28 22:29:50 encrypted gdm-Xorg-:0[3551]: (EE)
Apr 28 22:29:50 encrypted gdm-Xorg-:0[3551]: (EE) Server terminated with error (1). Closing log file.
Apr 28 22:29:50 encrypted gdm[3546]: Child process 3551 was already dead.
Apr 28 22:29:50 encrypted gdm[3546]: GdmDisplay: display lasted 0.388471 seconds

I'm getting this exact same segfault since updating AMD E-350 here (HD6310).

Seems to be GDM specific, SLiM works fine.

Last edited by brom (2014-05-04 05:37:14)

Vi0L0 · 2014-05-04 09:27:36

@teozkr:
could you paste somewhere report generated by atigetsysteminfo.sh, and strace startx?
Did you try to use older catalyst, like catalyst 13.12 + xorg114 ?

teozkr · 2014-05-04 10:53:09

Vi0L0 wrote:

@teozkr:
could you paste somewhere report generated by atigetsysteminfo.sh, and strace startx?

strace
atisysteminfo-report.txt

Vi0L0 wrote:

Did you try to use older catalyst, like catalyst 13.12 + xorg114 ?

I tried Catalyst 14.2 which was what I was using fine before.

Vi0L0 · 2014-05-04 14:45:00

teozkr wrote:

Vi0L0 wrote:
Did you try to use older catalyst, like catalyst 13.12 + xorg114 ?
I tried Catalyst 14.2 which was what I was using fine before.

Was it working fine with your new 270X?

teozkr · 2014-05-04 15:22:23

Vi0L0 wrote:

teozkr wrote:
Vi0L0 wrote:
Did you try to use older catalyst, like catalyst 13.12 + xorg114 ?
I tried Catalyst 14.2 which was what I was using fine before.
Was it working fine with your new 270X?

Nope, same symptoms. Didn't think far ahead enough to grab logs though, hoped the update would fix it.

EDIT: As a side note, I'm now running a separate Ubuntu 14.04 install on the machine with no issues (closed drivers).

Last edited by teozkr (2014-05-04 15:28:55)

Arch Linux

#6526 2014-05-01 12:53:16

Re: The AMD/ATI Bar & Grill

#6527 2014-05-01 13:02:20

Re: The AMD/ATI Bar & Grill

#6528 2014-05-01 13:06:24

Re: The AMD/ATI Bar & Grill

#6529 2014-05-01 13:07:37

Re: The AMD/ATI Bar & Grill

#6530 2014-05-01 13:10:14

Re: The AMD/ATI Bar & Grill

#6531 2014-05-01 14:24:42

Re: The AMD/ATI Bar & Grill

#6532 2014-05-01 14:31:25

Re: The AMD/ATI Bar & Grill

#6533 2014-05-01 14:33:58

Re: The AMD/ATI Bar & Grill

#6534 2014-05-01 14:41:27

Re: The AMD/ATI Bar & Grill

#6535 2014-05-01 14:51:55

Re: The AMD/ATI Bar & Grill

#6536 2014-05-01 16:52:45

Re: The AMD/ATI Bar & Grill

#6537 2014-05-01 18:56:16

Re: The AMD/ATI Bar & Grill

#6538 2014-05-01 19:06:19

Re: The AMD/ATI Bar & Grill

#6539 2014-05-02 02:17:56

Re: The AMD/ATI Bar & Grill

#6540 2014-05-03 11:16:27

Re: The AMD/ATI Bar & Grill

#6541 2014-05-03 13:39:01

Re: The AMD/ATI Bar & Grill

#6542 2014-05-03 19:16:04

Re: The AMD/ATI Bar & Grill

#6543 2014-05-03 20:38:10

Re: The AMD/ATI Bar & Grill

#6544 2014-05-03 20:52:00

Re: The AMD/ATI Bar & Grill

#6545 2014-05-03 21:00:35

Re: The AMD/ATI Bar & Grill

#6546 2014-05-04 05:06:51

Re: The AMD/ATI Bar & Grill

#6547 2014-05-04 09:27:36

Re: The AMD/ATI Bar & Grill

#6548 2014-05-04 10:53:09

Re: The AMD/ATI Bar & Grill

#6549 2014-05-04 14:45:00

Re: The AMD/ATI Bar & Grill

#6550 2014-05-04 15:22:23

Re: The AMD/ATI Bar & Grill

Board footer