Gentoo Forums
Gentoo Forums
Gentoo Forums
Quick Search: in
Radeon, blank screen after update
View unanswered posts
View posts from last 24 hours

Goto page Previous  1, 2  
Reply to topic    Gentoo Forums Forum Index Kernel & Hardware
View previous topic :: View next topic  
Author Message
tberger2
n00b
n00b


Joined: 05 Feb 2013
Posts: 69

PostPosted: Tue Mar 28, 2017 5:55 am    Post subject: Reply with quote

This error occurs on three different (radeon) machines (two onboard) using two different account settings and using radeon/radeonsi or radeon/amdgpu in /etc/portage/make.conf.
It's hard to believe that - apart of Berto - I'm the only one having this issue.

Sorry for not having the time to pursue it.
Back to top
View user's profile Send private message
Berto d Sera
n00b
n00b


Joined: 17 Apr 2009
Posts: 20

PostPosted: Tue Mar 28, 2017 7:53 pm    Post subject: Reply with quote

Hi!

NeddySeagoon wrote:
Berto d Sera,
You have two Video cards. Do you have the console on one and Xorg on the other?


The miniport VGA never was connected, the box always used the HDMI output only. I currently connect to it via ssh from a laptop. I just put through an order for a VGA cable on Amazon so tomorrow I can see whether it's simply using the other screen. These lines in Xorg.0.log seemed to exclude that the miniport could be used.
Code:

[    43.624] (II) RADEON(0): Output HDMI-0 connected
[    43.624] (II) RADEON(0): Output DisplayPort-0 disconnected
[    43.624] (II) RADEON(0): Using exact sizes for initial modes
[    43.624] (II) RADEON(0): Output HDMI-0 using initial mode 1920x1080 +0+0


Here is the full log as you asked, after recompiling with
VIDEO_CARDS="radeon"
and disabling both radeonsi r600. fglrx seems to have gone away (still no screen output).

https://paste.pound-python.org/show/iakxLEoHMRChNdpfIAZr/

NeddySeagoon wrote:

You reconfigured the kernel to change the modversions option
Code:
[   21.277485] fglrx: version magic '4.4.39-gentoo SMP mod_unload ' should be '4.4.39-gentoo SMP mod_unload modversions ' 

but did not rebuild the fglrx driver to match the kernel.

I *thought* I did not have fglrx in the first place. Was it brough it by r600 and/or radeonrsi? So it would appear to be, as it seems to be gone now, after removing them from the videocards.
Back to top
View user's profile Send private message
asturm
Developer
Developer


Joined: 05 Apr 2007
Posts: 7088
Location: Austria

PostPosted: Tue Mar 28, 2017 8:39 pm    Post subject: Reply with quote

No, the proprietary fglrx driver was part of the x11-drivers/ati-drivers package that is gone from portage, and should be gone from your system (make sure of that, both from world and xorg.conf). It has nothing to do with the open source drivers.
_________________
backend.cpp:92:2: warning: #warning TODO - this error message is about as useful as a cooling unit in the arctic
Back to top
View user's profile Send private message
Berto d Sera
n00b
n00b


Joined: 17 Apr 2009
Posts: 20

PostPosted: Tue Mar 28, 2017 9:21 pm    Post subject: Reply with quote

asturm wrote:
No, the proprietary fglrx driver was part of the x11-drivers/ati-drivers package that is gone from portage, and should be gone from your system (make sure of that, both from world and xorg.conf). It has nothing to do with the open source drivers.

Such was my impression, and when I check with eix I do not see the package in my system. No mention of it in /var/lib/portage/world either.

I re-checked dmesg and I still have these in dmesg, though :(
Code:

[   21.402043] fglrx: module_layout: kernel tainted.
[   21.402049] Disabling lock debugging due to kernel taint
[   21.402083] fglrx: version magic '4.4.39-gentoo SMP mod_unload ' should be '4.4.39-gentoo SMP mod_unload modversions '                                                                                                 
[   21.402097] fglrx: version magic '4.4.39-gentoo SMP mod_unload ' should be '4.4.39-gentoo SMP mod_unload modversions '   


Now that you mentioned Xorg.conf... I don't configure xorg on this box, but I found an /etc/X11/fglrx.conf which I am removing right away. We DID use fglrx in the past and the file timestamp would make sense, as it's January 2016. The fglrx module is nowhere in lsmod though, and it has not been ever since the problem started.

And indeed there is this
Code:

catlin ~ # ls -lah  /lib/modules/4.4.39-gentoo/video
total 18M
drwxr-xr-x 2 root root 4.0K Dec 27 15:58 .
drwxr-xr-x 4 root root 4.0K Mar 22 23:09 ..
-rw-r--r-- 1 root root  18M Dec 27 15:57 fglrx.ko


which at this point I manually remove.

Reboot and dmesg seems now clean (screen still blank)
https://paste.pound-python.org/show/Y89OxuZpG4ipFwxVFFI6/
Back to top
View user's profile Send private message
Berto d Sera
n00b
n00b


Joined: 17 Apr 2009
Posts: 20

PostPosted: Fri Mar 31, 2017 7:22 pm    Post subject: Reply with quote

Still waiting for a cable to test the second videocard. Disabling it seems impossible, there is no such option in the BIOS (this is the description of the box: http://www.gigabyte.com/Mini-PcBarebone/GB-BXA8G-8890-rev-10#ov there are 16Gb RAM onboard)

In the meantime I tried to upgrade to the lastest kernel. Still no joy

Code:
Linux catlin 4.9.16-gentoo #1 SMP Sat Mar 25 14:07:46 GMT 2017 x86_64 AMD A8-5557M APU with Radeon(tm) HD Graphics AuthenticAMD GNU/Linux


The only odd message during kernel compilation was
Code:

firmware/Makefile:183: target 'firmware/radeon/verde_mc.bin.gen.S' given more than once in the same rule
firmware/Makefile:183: target 'firmware/radeon/verde_me.bin.gen.S' given more than once in the same rule
firmware/Makefile:183: target 'firmware/radeon/TAHITI_uvd.bin.gen.S' given more than once in the same rule
firmware/Makefile:183: target 'firmware/radeon/TAHITI_vce.bin.gen.S' given more than once in the same rule


However, there is no error message in the logs once you boot, just the screen going blank when X starts

dmesg
https://paste.pound-python.org/show/zTyDtlzJKHU0znq2bblO/

Xorg.0.log
https://paste.pound-python.org/show/e4RgCHeMyGOEKl3GtFvK/

sddm.log
https://paste.pound-python.org/show/b1LpYVcFX0RGA2EB4sJr/

/etc/portage/make.conf
https://paste.pound-python.org/show/i7p1VXFKWgI0l0fqXapo/

I don't see anything suspicious at this point. I am honestly rather confused.

Some added info...
Code:

catlin log # eselect mesa list
64bit i915 (Intel 915, 945)
64bit i965 (Intel GMA 965, G/Q3x, G/Q4x, HD)
64bit r300 (Radeon R300-R500)
64bit r600 (Radeon R600-R700, Evergreen, Northern Islands)
  [1]   gallium *
64bit sw (Software renderer)
  [1]   gallium *
32bit i915 (Intel 915, 945)
32bit i965 (Intel GMA 965, G/Q3x, G/Q4x, HD)
32bit r300 (Radeon R300-R500)
32bit r600 (Radeon R600-R700, Evergreen, Northern Islands)
  [1]   gallium *
32bit sw (Software renderer)
  [1]   gallium *

catlin log # eselect opencl list
Available OpenCL implementations:
  [1]   mesa *

catlin log # eselect opengl list
Available OpenGL implementations:
  [1]   xorg-x11 *
Back to top
View user's profile Send private message
Berto d Sera
n00b
n00b


Joined: 17 Apr 2009
Posts: 20

PostPosted: Sun Apr 02, 2017 3:30 pm    Post subject: Reply with quote

Finally the cable arrived, so I unplugged the HDMI card and plugged in the other card on the miniport. And... nothing. Two cards, two blank screens. I do get the startup console, but as soon as X starts the world goes pitch back. Well, at least I can rule out a HW failure.

There is however a thing in dmesg now:
Code:
[   18.922904] vgaarb: device changed decodes: PCI:0000:01:00.0,olddecodes=io+mem,decodes=none:owns=none
[   18.922907] vgaarb: device changed decodes: PCI:0000:00:01.0,olddecodes=io+mem,decodes=none:owns=io+mem
[   24.160445] random: crng init done
[   25.145028] [drm:atom_op_jump [radeon]] *ERROR* atombios stuck in loop for more than 5secs aborting
[   25.145047] [drm:atom_execute_table_locked [radeon]] *ERROR* atombios stuck executing E4DE (len 2650, WS 0, PS 8) @ 0xEAE6
[   25.153452] [drm:radeon_dp_link_train [radeon]] *ERROR* clock recovery reached max voltage
[   25.153477] [drm:radeon_dp_link_train [radeon]] *ERROR* clock recovery failed


Could it be a conflict between the ARUBA and VERDE microcode? I am going to try and re-compile the kernel with ONE firmware only. Let's see what happens
Back to top
View user's profile Send private message
Berto d Sera
n00b
n00b


Joined: 17 Apr 2009
Posts: 20

PostPosted: Sun Apr 02, 2017 3:53 pm    Post subject: Reply with quote

As I just found out (ah, the irony) I had a rather messy string in my .config
Code:
CONFIG_EXTRA_FIRMWARE="radeon/verde_ce.bin radeon/verde_mc.bin radeon/verde_me.bin radeon/verde_pfpradeon/verde_ce.bin radeon/verde_mc.bin radeon/verde_me.bin radeon/verde_pfp.bin radeon/verde_rlc.bin radeon/verde_smc.bin radeon/TAHITI_uvd.bin radeon/TAHITI_vce.bin radeon/ARUBA_me.bin radeon/ARUBA_pfp.bin radeon/ARUBA_rlc.bin radeon/TAHITI_uvd.bin radeon/TAHITI_vce.bin"


Investigating as to how on earth this could even happen... and the thing is that if you edit the string from make menuconfig, and use the backspace to delete the contents of the settings, before inserting a new setting... only the part of the string that is actually on screen gets deleted. The rest stays is, and the only way to see it is to confirm the empty string, then re-edit it, and this is where you find out what was left over.

Re-compiling after clearing the mess. Fingers crossed.
Back to top
View user's profile Send private message
Berto d Sera
n00b
n00b


Joined: 17 Apr 2009
Posts: 20

PostPosted: Sun Apr 02, 2017 5:05 pm    Post subject: Reply with quote

Try as you might, using the miniport VGA leads to this:
Code:
[   25.841838] [drm:atom_op_jump [radeon]] *ERROR* atombios stuck in loop for more than 5secs aborting
[   25.841859] [drm:atom_execute_table_locked [radeon]] *ERROR* atombios stuck executing E4DE (len 2650, WS 0, PS 8) @ 0xEAE6
[   25.850292] [drm:radeon_dp_link_train [radeon]] *ERROR* clock recovery reached max voltage
[   25.850318] [drm:radeon_dp_link_train [radeon]] *ERROR* clock recovery failed


And if I reboot with the HDMI port connected, there is no error message at all, but also nothing on screen (as soon as X starts).
Back to top
View user's profile Send private message
NeddySeagoon
Administrator
Administrator


Joined: 05 Jul 2003
Posts: 43406
Location: 56N 3W

PostPosted: Sun Apr 02, 2017 5:07 pm    Post subject: Reply with quote

Berto d Sera,

Pastebin your /var/log/Xorg.0.log please.
_________________
Regards,

NeddySeagoon

Computer users fall into two groups:-
those that do backups
those that have never had a hard drive fail.
Back to top
View user's profile Send private message
Display posts from previous:   
Reply to topic    Gentoo Forums Forum Index Kernel & Hardware All times are GMT
Goto page Previous  1, 2
Page 2 of 2

 
Jump to:  
You cannot post new topics in this forum
You cannot reply to topics in this forum
You cannot edit your posts in this forum
You cannot delete your posts in this forum
You cannot vote in polls in this forum