Discussion: the ati/radeon open source drivers & radeon repository

agapito · 2013-08-06 08:52:39

Ledti wrote:

Though there is an occasional hard lock. As in... everything completely freezes and I have to push the reset button to forcibly restart my computer. I'm not sure if it's related to DPM or the fact that I'm using a prepatch kernel with the latest mesa-git and ATI packages...

I have the same problem, occurs randomly when i watch videos with smplayer and vdpau enabled. I think is more a vdpau related problem than a dpm, because with kernels 3.10.x i'm still getting this bug.

Aldur · 2013-08-06 12:00:03

To report back: I'm now using rc4 and I still have the same issues. I can't find anything strange in dmesg as DPM and the module is just loaded fine. "power_profile" is automatically set to dpm, too.

So, yeah, I'm absolutely lost, especially as it works for others with a Radeon HD 6870. I came to think that it MIGHT be related to my card specifically? Like an issue with its bios?

Ledti · 2013-08-06 20:40:20

@ agapito: I don't use vdpau since it's a little buggy for me, so I'm not sure if it's the same issue.

@ Aldur: If the catalyst drivers work fine for you, then I would think it's an issue with the open source drivers... it just means that your setup is a little unique or something. You should really consider reporting your issue here as the bug report itself is still open.

Last edited by Ledti (2013-08-06 20:40:58)

Ledti · 2013-08-06 22:38:54

Aldur wrote:

To report back: I'm now using rc4 and I still have the same issues. I can't find anything strange in dmesg as DPM and the module is just loaded fine. "power_profile" is automatically set to dpm, too.
So, yeah, I'm absolutely lost, especially as it works for others with a Radeon HD 6870. I came to think that it MIGHT be related to my card specifically? Like an issue with its bios?

You could also try patching 3.11-rc4 against the latest changes from agd5f's drm-fixes-3.11 branch which has a few NI fixes. I posted a patch for you here.

I generated it using:

git diff v3.11-rc4 agd5f/drm-fixes-3.11 -- drivers/gpu/drm/radeon > ../agd5f-latest.patch

... which I think is the correct way.

There are also a few UVD fixes from what I can tell.

Aldur · 2013-08-07 05:33:39

Wow, thanks for your reply *bows*!
I'll give it a try!

I've opened a new bug report as I was not so sure if it's the same issue, given that the Radeon HD 6870 dpm support should be in pretty good shape already, atleast according to Alex Deucher.
So it must be something that's either really unique to my setup or has been overlooked before.

The one "special" thing I have: it's an XFX Black Edition card, then again, that only means slightly increased clock rates.
But yeah, with fglrx I don't have any extra issues, besides the issues everyone has.
It's really hard to debug this issue as my machine is only stable for a few seconds when using DPM, so all I could do was switch to a terminal ASAP, dump dmesg into a file and then it froze again. Before I had to restart it several times until it was able to boot to that point.

https://bugs.freedesktop.org/show_bug.cgi?id=67825

Last edited by Aldur (2013-08-07 05:35:08)

mxx · 2013-08-07 20:24:02

Hi,

Sorry, this might have been asked before, but I can't find any specific information about it after hours of searching.

Could someone please tell me whether the open source radeon driver supports lpcm over hdmi with more than 2 channels on "northern island" (without passthrough)?
All info I find seems to be related to analog surround or ac3 et.c. passthrough.
(When I do a speaker-test -c6 I only get front speakers working )

Now I've found this thread http://forum.xbmc.org/showthread.php?tid=169873&page=2 where they claim that more than 2 channels isn't possible, but it might be only related to the proprietary driver.
I was trying to get sound working with 5.1 over hdmi for the whole day now, so if this is a driver issue, I can at least stop now.

Thank you very much!

Last edited by mxx (2013-08-07 20:25:36)

gee · 2013-08-10 06:25:03

@mxx: I've tried for months, got some help from the official ALSA devs and no go...
I talked to bridgman from AMD too maybe a year ago, and same there was no way to do it, and it was not planned either (yet?)...

xoma · 2013-08-10 21:55:02

I've been using the foss driver and it works great. My only concern is when I try to play Minetest it uses up all 6 cores of my cpu at almost 100% and it uses opengl 2. I don't know if this is the driver having issues or Minetest.

gee · 2013-08-21 03:57:22

mesa 9.2 just hit testing and now vdpau works, yoohoo!

pYrO1v1aniac · 2013-08-25 22:22:05

Just upgraded to Linux-Mainline and Mesa-Git to try out radeon. When I boot into core kernel, everything works fine, but obviously no performance gains. When I try to boot into Linux-Mainline, I get a strange graphical problem after grub, the screen starts all-dark, and then gradually shifts unevenly to bright, then back to all dark again in patches. I've seen this before once when having problems with catalyst, but I don't remember what the issue was then.

Does anybody have any pointers as to what logs, or how I might get this working? I'd really like to try out the new radeon.

I'm using a 5470M

EDIT-
I tried with the linux-adg5f and mesa-git packages from the [mesa-git] repository, same result, screen oscillates between light and dark gradually, and no console or X session is ever shown.

I don't think I'm ever going to get this useless piece of **** card working properly.

Last edited by pYrO1v1aniac (2013-08-29 22:37:06)

Unia · 2013-08-29 16:27:17

So today I got mesa 9.2 in my updates and immediately after I tried to get mplayer running with vdpau. Sadly, it doesn't work for me; mplayer fails with the following error:

Loading extension-related profile 'vo.vdpau'
Inconsistency detected by ld.so: dl-close.c: 771: _dl_close: Assertion `map->l_init_called' failed!

I exported VDPAU_DRIVER:

export VDPAU_DRIVER=r600

And I have this in my ~/.mplayer/config:

vo=vdpau
...
[vo.vdpau]
vc=ffmpeg12vdpau,ffwmv3vdpau,ffvc1vdpau,ffh264vdpau,ffodivxvdpau,
vf-clr=yes

I tried an avi and an mp4 file and strange enough, the mp4 file plays fine even though I get the above mentioned error message. On Google I got to a bug report my Lordheavy but it said it had been resolved. Can someone with more insight help me?

lordheavy · 2013-08-29 16:29:56

Unia wrote:

So today I got mesa 9.2 in my updates and immediately after I tried to get mplayer running with vdpau. Sadly, it doesn't work for me; mplayer fails with the following error:
Loading extension-related profile 'vo.vdpau'
Inconsistency detected by ld.so: dl-close.c: 771: _dl_close: Assertion `map->l_init_called' failed!

Have you tried without exporting VDPAU_DRIVER ?

Unia · 2013-08-29 16:37:03

Just tested (by rebooting, just to be sure) and it doesn't work either. Same error message. I also tried r300, r300g and r600g because I also saw those float around the internet but no go either. (as you can see, I really have no clue what I'm doing here). If it can be of any help, I have an AMD Mobility Radeon HD 5470.

kaszak696 · 2013-08-29 18:39:05

Can you provide a full log from mplayer? The one you posted is not caused by mplayer, but by mesa/libvdpau, i also get it in everything that interfaces with vdpau, but the acceleration works. https://aur.archlinux.org/packages/vdpauinfo output might help debugging the issue, too.

Also, post the codecs your video files use, mp4 and avi are just containers, they can contain virtually any obscure codec imaginable.

Last edited by kaszak696 (2013-08-29 18:40:12)

Unia · 2013-08-29 18:41:16

kaszak696 wrote:

Can you provide a full log from mplayer? The one you posted is not caused by mplayer, but by mesa/libvdpau, i also get it in everything that interfaces with vdpau, but the acceleration works. https://aur.archlinux.org/packages/vdpauinfo output might help debugging the issue, too.

Oke, will do! Give me a minute. Here it is:

display: :0   screen: 0
API version: 1
Information string: G3DVL VDPAU Driver Shared Library version 1.0

Video surface:

name   width height types
-------------------------------------------
420    16384 16384  NV12 
422    16384 16384  NV12 
444    16384 16384  NV12 

Decoder capabilities:

name               level macbs width height
-------------------------------------------
MPEG1                16  9216  2048  1152
MPEG2_SIMPLE         16  9216  2048  1152
MPEG2_MAIN           16  9216  2048  1152
H264_BASELINE        16  9216  2048  1152
H264_MAIN            16  9216  2048  1152
H264_HIGH            16  9216  2048  1152
VC1_SIMPLE           16  9216  2048  1152
VC1_MAIN             16  9216  2048  1152
VC1_ADVANCED         16  9216  2048  1152

Output surface:

name              width height nat types
----------------------------------------------------
B8G8R8A8         16384 16384    y  NV12 
R8G8B8A8         16384 16384    y  NV12 
R10G10B10A2      16384 16384    y  NV12 
B10G10R10A2      16384 16384    y  NV12 

Bitmap surface:

name              width height
------------------------------
B8G8R8A8         16384 16384
R8G8B8A8         16384 16384
R10G10B10A2      16384 16384
B10G10R10A2      16384 16384
A8               16384 16384

Video mixer:

feature name                    sup
------------------------------------
DEINTERLACE_TEMPORAL             -
DEINTERLACE_TEMPORAL_SPATIAL     -
INVERSE_TELECINE                 -
NOISE_REDUCTION                  y
SHARPNESS                        y
LUMA_KEY                         -
HIGH QUALITY SCALING - L1        -
HIGH QUALITY SCALING - L2        -
HIGH QUALITY SCALING - L3        -
HIGH QUALITY SCALING - L4        -
HIGH QUALITY SCALING - L5        -
HIGH QUALITY SCALING - L6        -
HIGH QUALITY SCALING - L7        -
HIGH QUALITY SCALING - L8        -
HIGH QUALITY SCALING - L9        -

parameter name                  sup      min      max
-----------------------------------------------------
VIDEO_SURFACE_WIDTH              y        48     2048
VIDEO_SURFACE_HEIGHT             y        48     1152
CHROMA_TYPE                      y  
LAYERS                           y         0        4

attribute name                  sup      min      max
-----------------------------------------------------
BACKGROUND_COLOR                 y  
CSC_MATRIX                       y  
NOISE_REDUCTION_LEVEL            y      0.00     1.00
SHARPNESS_LEVEL                  y     -1.00     1.00
LUMA_KEY_MIN_LUMA                y  
LUMA_KEY_MAX_LUMA                y  


Inconsistency detected by ld.so: dl-close.c: 771: _dl_close: Assertion `map->l_init_called' failed!

kaszak696 wrote:

Also, post the codecs your video files use, mp4 and avi are just containers, they can contain virtually any obscure codec imaginable.

How would I figure this out?

Last edited by Unia (2013-08-29 18:44:34)

kaszak696 · 2013-08-29 18:43:44

Unia wrote:

How would I figure this out?

Mediainfo tool from the repos should do the trick, it's in the repos, has CLI and GUI versions. I think it's also possible to guess from mplayer log, but mediainfo is nicer ;p
As vdpauinfo output states, your card supports only MPEG1, MPEG2, H264 and VC1, if codecs in these files are anything but this, acceleration is not possible. Bear in mind that 10bit H264 is also unsupported and, as far as i know, never will be.

Last edited by kaszak696 (2013-08-29 18:49:33)

Unia · 2013-08-29 18:48:41

The .avi has this:

Video
ID                                       : 0
Format                                   : MPEG-4 Visual
Format profile                           : Advanced Simple@L5
Format settings, BVOP                    : 4
Format settings, QPel                    : No
Format settings, GMC                     : No warppoints
Format settings, Matrix                  : Default (H.263)
Muxing mode                              : Packed bitstream
Codec ID                                 : XVID
Codec ID/Hint                            : XviD
Writing library                          : XviD 64

The .mp4 has this:

ID                                       : 1
Format                                   : AVC
Format/Info                              : Advanced Video Codec
Format profile                           : High@L4.0
Format settings, CABAC                   : Yes
Format settings, ReFrames                : 4 frames
Codec ID                                 : avc1
Codec ID/Info                            : Advanced Video Coding
Writing library                          : x264 core 130 r2273 b3065e6

kaszak696 · 2013-08-29 18:51:11

Then it's as expected. AVI file contains XVID which cannot be accelerated, and MP4 has H264 HIGH, which can. Works as intended. Too bad that mplayer/ffmpeg does not fallback to unaccelerated codec and crashes instead.

Last edited by kaszak696 (2013-08-29 18:52:45)

Unia · 2013-08-29 18:53:33

Aha, oke. But then why doesn't Mplayer automatically chooses another coded to use? Isn't that what the comma at the end of the vc= line is for?

By the way, thanks for your quick help!

kaszak696 · 2013-08-29 19:00:43

Unia wrote:

Aha, oke. But then why doesn't Mplayer automatically chooses another coded to use? Isn't that what the comma at the end of the vc= line is for?
By the way, thanks for your quick help!

I think it's because technically, the ffodivxvdpau decoder matches the XVID codec in your avi, so mplayer sees no problem there, but it blows up while being handed to vdpau. It's just a wild guess though.

From mplayer wiki page:

Warning: The ffodivxvdpau codec is only supported by the most recent series of NVIDIA hardware. Consider omitting it based on your specific hardware.

Omitting it should do the trick, but it will still blow up on H264 10bit, so be mindful of that.

Last edited by kaszak696 (2013-08-29 19:02:57)

Unia · 2013-08-29 19:02:22

Oh oke, I see. Thanks again for your quick help!

Unia · 2013-08-29 19:47:15

FWIW, mpv does fallback correctly.

Aldur · 2013-08-31 09:09:30

That's really weird, after switching back to the open source drivers I noticed artifacts (especially in the upper left corner of windows) upon logging in and while using the system. Fonts are massively corrupted, too and Thunderbird crashes when I want to click on my inbox, fonts are broken there, too.
Also my system freezes when I use kernel 3.11.
Before, they (the open source drivers) worked quite well, even with kernel 3.11, just dynpm didn't want to work.
I'm quite lost how that could happen all of a sudden, so I expected it to be an issue with the newest ati-dri-git, etc... and I tried mesa 9.2 and ati-dri from the official repos but nothing changed. I also tried deactivating the new shader backend but that didn't fix my issues either.

When I use fglrx everything's working fine: Thunderbird, fonts, kernel 3.11.
Anyone got any idea what's going on and/or is experiencing the same issues?

Ps: I'm assuming it's related to KMS but why would it break, all of a sudden?

"Fun" fact:
when my monitor/GPU went on standby I had no artifacts whatsoever after reactivating it. On reboot I had artifacts once again.
Additionally 3D applications are working flawlessly, without any artifacts.

Last edited by Aldur (2013-08-31 09:56:37)

splashy · 2013-09-02 21:41:21

Have you try this: http://www.botchco.com/agd5f/?p=57 ?

Basically

echo performance > /sys/class/drm/card0/device/power_dpm_state
echo high > /sys/class/drm/card0/device/power_dpm_force_performance_level

Made the trick for me, however I have to do it every time I boot the computer, somebody know how to make these changes permanent?

Last edited by splashy (2013-09-02 21:41:45)

siavashserver · 2013-09-03 02:21:28

splashy wrote:

Have you try this: http://www.botchco.com/agd5f/?p=57 ?
Basically
echo performance > /sys/class/drm/card0/device/power_dpm_state
echo high > /sys/class/drm/card0/device/power_dpm_force_performance_level
Made the trick for me, however I have to do it every time I boot the computer, somebody know how to make these changes permanent?

You can create a new service to do these stuff during boot for you. Create these files on your system:

/etc/systemd/system/powermanagement.service

[Unit]
Description=Power Management Service

[Service]
Type=oneshot
ExecStart=/etc/powermanagement.sh
TimeoutSec=0
StandardInput=tty
RemainAfterExit=yes

[Install]
WantedBy=multi-user.target

/etc/powermanagement.sh

echo performance > /sys/class/drm/card0/device/power_dpm_state
echo high > /sys/class/drm/card0/device/power_dpm_force_performance_level

And run this as root to enable the new service:

systemctl enable powermanagement

Arch Linux

#2526 2013-08-06 08:52:39

Re: Discussion: the ati/radeon open source drivers & radeon repository

#2527 2013-08-06 12:00:03

Re: Discussion: the ati/radeon open source drivers & radeon repository

#2528 2013-08-06 20:40:20

Re: Discussion: the ati/radeon open source drivers & radeon repository

#2529 2013-08-06 22:38:54

Re: Discussion: the ati/radeon open source drivers & radeon repository

#2530 2013-08-07 05:33:39

Re: Discussion: the ati/radeon open source drivers & radeon repository

#2531 2013-08-07 20:24:02

Re: Discussion: the ati/radeon open source drivers & radeon repository

#2532 2013-08-10 06:25:03

Re: Discussion: the ati/radeon open source drivers & radeon repository

#2533 2013-08-10 21:55:02

Re: Discussion: the ati/radeon open source drivers & radeon repository

#2534 2013-08-21 03:57:22

Re: Discussion: the ati/radeon open source drivers & radeon repository

#2535 2013-08-25 22:22:05

Re: Discussion: the ati/radeon open source drivers & radeon repository

#2536 2013-08-29 16:27:17

Re: Discussion: the ati/radeon open source drivers & radeon repository

#2537 2013-08-29 16:29:56

Re: Discussion: the ati/radeon open source drivers & radeon repository

#2538 2013-08-29 16:37:03

Re: Discussion: the ati/radeon open source drivers & radeon repository

#2539 2013-08-29 18:39:05

Re: Discussion: the ati/radeon open source drivers & radeon repository

#2540 2013-08-29 18:41:16

Re: Discussion: the ati/radeon open source drivers & radeon repository

#2541 2013-08-29 18:43:44

Re: Discussion: the ati/radeon open source drivers & radeon repository

#2542 2013-08-29 18:48:41

Re: Discussion: the ati/radeon open source drivers & radeon repository

#2543 2013-08-29 18:51:11

Re: Discussion: the ati/radeon open source drivers & radeon repository

#2544 2013-08-29 18:53:33

Re: Discussion: the ati/radeon open source drivers & radeon repository

#2545 2013-08-29 19:00:43

Re: Discussion: the ati/radeon open source drivers & radeon repository

#2546 2013-08-29 19:02:22

Re: Discussion: the ati/radeon open source drivers & radeon repository

#2547 2013-08-29 19:47:15

Re: Discussion: the ati/radeon open source drivers & radeon repository

#2548 2013-08-31 09:09:30

Re: Discussion: the ati/radeon open source drivers & radeon repository

#2549 2013-09-02 21:41:21

Re: Discussion: the ati/radeon open source drivers & radeon repository

#2550 2013-09-03 02:21:28

Re: Discussion: the ati/radeon open source drivers & radeon repository

Board footer