Hi,
commit in $Subject breaks suspend to disk on my box here. Reverting it ontop of 4.7-rc2 fixes the problem. DRM-specific messages in dmesg are:
[ 6.837698] [drm] radeon kernel modesetting enabled. [ 6.871372] [drm] initializing kernel modesetting (RV635 0x1002:0x9598 0x1043:0x01DA 0x00). [ 6.871387] [drm] register mmio base: 0xFEA20000 [ 6.871388] [drm] register mmio size: 65536 [ 6.871497] [drm] Detected VRAM RAM=512M, BAR=256M [ 6.871498] [drm] RAM width 128bits DDR [ 6.875602] [drm] radeon: 512M of VRAM memory ready [ 6.875603] [drm] radeon: 512M of GTT memory ready. [ 6.875619] [drm] Loading RV635 Microcode [ 6.888041] [drm] Internal thermal controller without fan control [ 6.888100] [drm] radeon: power management initialized [ 6.892400] [drm] GART: num cpu pages 131072, num gpu pages 131072 [ 6.893273] [drm] enabling PCIE gen 2 link speeds, disable with radeon.pcie_gen2=0 [ 6.920032] [drm] PCIE GART of 512M enabled (table at 0x0000000000142000). [ 6.920232] [drm] Supports vblank timestamp caching Rev 2 (21.10.2013). [ 6.920232] [drm] Driver supports precise vblank timestamp query. [ 6.920290] [drm] radeon: irq initialized. [ 6.951284] [drm] ring test on 0 succeeded in 0 usecs [ 7.125955] [drm] ring test on 5 succeeded in 1 usecs [ 7.131079] [drm] UVD initialized successfully. [ 7.135817] [drm] ib test on ring 0 succeeded in 0 usecs [ 7.789092] [drm] ib test on ring 5 succeeded [ 7.794502] [drm] Radeon Display Connectors [ 7.798817] [drm] Connector 0: [ 7.801963] [drm] DVI-I-1 [ 7.804825] [drm] HPD1 [ 7.807430] [drm] DDC: 0x7e50 0x7e50 0x7e54 0x7e54 0x7e58 0x7e58 0x7e5c 0x7e5c [ 7.814942] [drm] Encoders: [ 7.817985] [drm] DFP1: INTERNAL_UNIPHY [ 7.822230] [drm] CRT2: INTERNAL_KLDSCP_DAC2 [ 7.826913] [drm] Connector 1: [ 7.830033] [drm] DIN-1 [ 7.832715] [drm] Encoders: [ 7.835751] [drm] TV1: INTERNAL_KLDSCP_DAC2 [ 7.840349] [drm] Connector 2: [ 7.843466] [drm] DVI-I-2 [ 7.846326] [drm] HPD2 [ 7.848926] [drm] DDC: 0x7e40 0x7e40 0x7e44 0x7e44 0x7e48 0x7e48 0x7e4c 0x7e4c [ 7.856437] [drm] Encoders: [ 7.859479] [drm] CRT1: INTERNAL_KLDSCP_DAC1 [ 7.864162] [drm] DFP2: INTERNAL_KLDSCP_LVTMA [ 7.941480] [drm] fb mappable at 0xC0243000 [ 7.945733] [drm] vram apper at 0xC0000000 [ 7.949897] [drm] size 9216000 [ 7.953019] [drm] fb depth is 24 [ 7.956327] [drm] pitch is 7680 [ 7.959930] fbcon: radeondrmfb (fb0) is primary device [ 8.028816] radeon 0000:01:00.0: fb0: radeondrmfb frame buffer device [ 8.045087] [drm] Initialized radeon 2.45.0 20080528 for 0000:01:00.0 on minor 0 [ 28.158851] [drm] enabling PCIE gen 2 link speeds, disable with radeon.pcie_gen2=0 [ 28.160612] [drm] PCIE GART of 512M enabled (table at 0x0000000000142000). [ 28.191669] [drm] ring test on 0 succeeded in 1 usecs [ 28.366319] [drm] ring test on 5 succeeded in 1 usecs [ 28.366323] [drm] UVD initialized successfully. [ 28.366338] [drm] ib test on ring 0 succeeded in 0 usecs [ 28.604362] ata1.00: supports DRM functions and may not be fully accessible [ 28.604405] ata2.00: supports DRM functions and may not be fully accessible [ 28.609305] ata1.00: supports DRM functions and may not be fully accessible [ 28.609410] ata2.00: supports DRM functions and may not be fully accessible [ 29.014440] [drm] ib test on ring 5 succeeded
On Mon, Jun 06, 2016 at 05:40:19PM -0400, Jerome Glisse wrote:
Brokens how ? Symptoms ?
Whoops, sorry, I meant to elaborate...
After doing:
echo "shutdown" > /sys/power/disk echo "disk" > /sys/power/state
screen goes blank but machine remains powered on and doesn't go off. No responses to keyboard presses, etc. The only thing left to do is the reset button.
On Mon, Jun 06, 2016 at 05:40:19PM -0400, Jerome Glisse wrote:
Brokens how ? Symptoms ?
Whoops, sorry, I meant to elaborate...
After doing:
echo "shutdown" > /sys/power/disk echo "disk" > /sys/power/statescreen goes blank but machine remains powered on and doesn't go off. No responses to keyboard presses, etc. The only thing left to do is the reset button.
Ok i don't have too much time to dig into r600 i assume that r700 breaks the same way so could you verify that attached patch fix the issue for you. Note that video decoding is likely broken for you after hibernation but you might have never notice it.
Cheers, Jérôme
Am 07.06.2016 um 23:51 schrieb Jerome Glisse:
On Mon, Jun 06, 2016 at 05:40:19PM -0400, Jerome Glisse wrote:
Brokens how ? Symptoms ?
Whoops, sorry, I meant to elaborate...
After doing:
echo "shutdown" > /sys/power/disk echo "disk" > /sys/power/statescreen goes blank but machine remains powered on and doesn't go off. No responses to keyboard presses, etc. The only thing left to do is the reset button.
Ok i don't have too much time to dig into r600 i assume that r700 breaks the same way so could you verify that attached patch fix the issue for you. Note that video decoding is likely broken for you after hibernation but you might have never notice it.
Patch is Reviewed-by: Christian König christian.koenig@amd.com.
Please keep me CCed if you have time looking into this.
Regards, Christian.
Cheers, Jérôme
On Tue, Jun 07, 2016 at 05:51:52PM -0400, Jerome Glisse wrote:
Ok i don't have too much time to dig into r600 i assume that r700 breaks the same way so could you verify that attached patch fix the issue for you. Note that video decoding is likely broken for you after hibernation but you might have never notice it.
Probably. And that means I really haven't noticed.
How do I test video decoding? mplayer seems to work fine :)
From 8ed42906e430880ce01bc6f175f1c7c180529353 Mon Sep 17 00:00:00 2001 From: =?UTF-8?q?J=C3=A9r=C3=B4me=20Glisse?= jglisse@redhat.com Date: Tue, 7 Jun 2016 17:43:04 -0400 Subject: [PATCH] drm/radeon: do not hard reset GPU while freezing on r600/r700 family MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit
Seems r600/r700 does not like hard reset while freezing for hibernation (regression due to 274ad65c9d02bdcbee9bae045517864c3521d530 which itself is a fix for hibernation on some GPU families). Until i can debug further issue with r600, let just disable this for r600/r700 as they are very similar family and bug affecting one likely affect the other.
Signed-off-by: Jérôme Glisse jglisse@redhat.com
Reported-and-tested-by: Borislav Petkov bp@suse.de
Am 08.06.2016 um 13:36 schrieb Borislav Petkov:
On Tue, Jun 07, 2016 at 05:51:52PM -0400, Jerome Glisse wrote:
Ok i don't have too much time to dig into r600 i assume that r700 breaks the same way so could you verify that attached patch fix the issue for you. Note that video decoding is likely broken for you after hibernation but you might have never notice it.
Probably. And that means I really haven't noticed.
How do I test video decoding? mplayer seems to work fine :)
What's the output of mplayer? Mplayer usually uses video acceleration when it is available.
Regards, Christian.
From 8ed42906e430880ce01bc6f175f1c7c180529353 Mon Sep 17 00:00:00 2001 From: =?UTF-8?q?J=C3=A9r=C3=B4me=20Glisse?= jglisse@redhat.com Date: Tue, 7 Jun 2016 17:43:04 -0400 Subject: [PATCH] drm/radeon: do not hard reset GPU while freezing on r600/r700 family MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit
Seems r600/r700 does not like hard reset while freezing for hibernation (regression due to 274ad65c9d02bdcbee9bae045517864c3521d530 which itself is a fix for hibernation on some GPU families). Until i can debug further issue with r600, let just disable this for r600/r700 as they are very similar family and bug affecting one likely affect the other.
Signed-off-by: Jérôme Glisse jglisse@redhat.com
Reported-and-tested-by: Borislav Petkov bp@suse.de
On Wed, Jun 08, 2016 at 01:50:28PM +0200, Christian König wrote:
What's the output of mplayer? Mplayer usually uses video acceleration when it is available.
Something like this?
libavformat version 56.23.105 (internal) libavformat file format detected. [lavf] stream 0: video (h264), -vid 0 [lavf] stream 1: audio (aac), -aid 0, -alang eng VIDEO: [H264] 1280x720 24bpp 23.976 fps 1702.9 kbps (207.9 kbyte/s) Clip info: major_brand: mp42 minor_version: 0 compatible_brands: isommp42 creation_time: 2015-09-24 14:54:07 Load subtitles in ./ ========================================================================== Opening video decoder: [ffmpeg] FFmpeg's libavcodec codec family libavcodec version 56.26.100 (internal) Selected video codec: [ffh264] vfm: ffmpeg (FFmpeg H.264) ========================================================================== ========================================================================== Opening audio decoder: [ffmpeg] FFmpeg/libavcodec audio decoders AUDIO: 44100 Hz, 2 ch, floatle, 192.0 kbit/6.80% (ratio: 23999->352800) Selected audio codec: [ffaac] afm: ffmpeg (FFmpeg AAC (MPEG-2/MPEG-4 Audio)) ========================================================================== AO: [alsa] 48000Hz 2ch floatle (4 bytes per sample) Starting playback... Movie-Aspect is 1.78:1 - prescaling to correct movie aspect. VO: [xv] 1280x720 => 1280x720 Planar YV12 A: 1.5 V: 1.5 A-V: 0.001 ct: 0.041 0/ 0 8% 2% 1.3% 0 0
Yes, exactly.
VO: [xv] 1280x720 => 1280x720 Planar YV12
Mplayer is using xv without any acceleration (except for color space conversion).
Try forcing mplayer to use VDPAU with "mplayer -vo vdpau $file".
Regards, Christian.
Am 08.06.2016 um 15:26 schrieb Borislav Petkov:
On Wed, Jun 08, 2016 at 01:50:28PM +0200, Christian König wrote:
What's the output of mplayer? Mplayer usually uses video acceleration when it is available.
Something like this?
libavformat version 56.23.105 (internal) libavformat file format detected. [lavf] stream 0: video (h264), -vid 0 [lavf] stream 1: audio (aac), -aid 0, -alang eng VIDEO: [H264] 1280x720 24bpp 23.976 fps 1702.9 kbps (207.9 kbyte/s) Clip info: major_brand: mp42 minor_version: 0 compatible_brands: isommp42 creation_time: 2015-09-24 14:54:07 Load subtitles in ./ ========================================================================== Opening video decoder: [ffmpeg] FFmpeg's libavcodec codec family libavcodec version 56.26.100 (internal) Selected video codec: [ffh264] vfm: ffmpeg (FFmpeg H.264) ========================================================================== ========================================================================== Opening audio decoder: [ffmpeg] FFmpeg/libavcodec audio decoders AUDIO: 44100 Hz, 2 ch, floatle, 192.0 kbit/6.80% (ratio: 23999->352800) Selected audio codec: [ffaac] afm: ffmpeg (FFmpeg AAC (MPEG-2/MPEG-4 Audio)) ========================================================================== AO: [alsa] 48000Hz 2ch floatle (4 bytes per sample) Starting playback... Movie-Aspect is 1.78:1 - prescaling to correct movie aspect. VO: [xv] 1280x720 => 1280x720 Planar YV12 A: 1.5 V: 1.5 A-V: 0.001 ct: 0.041 0/ 0 8% 2% 1.3% 0 0
On Wed, Jun 08, 2016 at 03:30:34PM +0200, Christian König wrote:
Try forcing mplayer to use VDPAU with "mplayer -vo vdpau $file".
All good. Actually, this hw accel thing is much better, I better make it default :-P
libavformat version 56.23.105 (internal) libavformat file format detected. [lavf] stream 0: video (h264), -vid 0 [lavf] stream 1: audio (aac), -aid 0, -alang eng VIDEO: [H264] 1280x720 24bpp 23.976 fps 1702.9 kbps (207.9 kbyte/s) Clip info: major_brand: mp42 minor_version: 0 compatible_brands: isommp42 creation_time: 2015-09-24 14:54:07 Load subtitles in ./ ========================================================================== Opening video decoder: [ffmpeg] FFmpeg's libavcodec codec family libavcodec version 56.26.100 (internal) Selected video codec: [ffh264] vfm: ffmpeg (FFmpeg H.264) ========================================================================== ========================================================================== Opening audio decoder: [ffmpeg] FFmpeg/libavcodec audio decoders AUDIO: 44100 Hz, 2 ch, floatle, 192.0 kbit/6.80% (ratio: 23999->352800) Selected audio codec: [ffaac] afm: ffmpeg (FFmpeg AAC (MPEG-2/MPEG-4 Audio)) ========================================================================== AO: [alsa] 48000Hz 2ch floatle (4 bytes per sample) Starting playback... Movie-Aspect is 1.78:1 - prescaling to correct movie aspect. VO: [vdpau] 1280x720 => 1280x720 Planar YV12 A: 3.5 V: 3.5 A-V: 0.000 ct: 0.042 0/ 0 8% 1% 1.3% 0 0
On Wed, Jun 08, 2016 at 03:47:12PM +0200, Borislav Petkov wrote:
All good. Actually, this hw accel thing is much better, I better make it default :-P
And yes, this is with Jérôme's fix to exclude r600 and r700 from hard reset before hibernation. And after a s2d cycle I did earlier.
On Wed, Jun 08, 2016 at 03:47:12PM +0200, Borislav Petkov wrote:
All good. Actually, this hw accel thing is much better, I better make it default :-P
And yes, this is with Jérôme's fix to exclude r600 and r700 from hard reset before hibernation. And after a s2d cycle I did earlier.
To be clear, you mean that after hibernation video acceleration keeps working ? Can you copy radeon dmesg after hibernation cycle (once you resumed from hibernation).
Cheers, Jérôme
On Wed, Jun 08, 2016 at 10:06:23AM -0400, Jerome Glisse wrote:
To be clear, you mean that after hibernation video acceleration keeps working ?
Apparently. At lest the vdpau output looks fine to me.
Can you copy radeon dmesg after hibernation cycle (once you resumed from hibernation).
$ dmesg | grep -Ei "(radeon|drm)"
...
(boot messages)
... [ 8.149356] fbcon: radeondrmfb (fb0) is primary device [ 8.218800] radeon 0000:01:00.0: fb0: radeondrmfb frame buffer device [ 8.238402] [drm] Initialized radeon 2.45.0 20080528 for 0000:01:00.0 on minor 0
<-- hibernate.
[ 64.711657] [drm] enabling PCIE gen 2 link speeds, disable with radeon.pcie_gen2=0 [ 64.714898] [drm] PCIE GART of 512M enabled (table at 0x0000000000142000). [ 64.714920] radeon 0000:01:00.0: WB enabled [ 64.714923] radeon 0000:01:00.0: fence driver on ring 0 use gpu addr 0x0000000020000c00 and cpu addr 0xffff88042b342c00 [ 64.715123] radeon 0000:01:00.0: fence driver on ring 5 use gpu addr 0x00000000000521d0 and cpu addr 0xffffc900030121d0 [ 64.745988] [drm] ring test on 0 succeeded in 1 usecs [ 64.920633] [drm] ring test on 5 succeeded in 1 usecs [ 64.920637] [drm] UVD initialized successfully. [ 64.920652] [drm] ib test on ring 0 succeeded in 0 usecs [ 65.128759] ata1.00: supports DRM functions and may not be fully accessible [ 65.128799] ata2.00: supports DRM functions and may not be fully accessible [ 65.133700] ata1.00: supports DRM functions and may not be fully accessible [ 65.133811] ata2.00: supports DRM functions and may not be fully accessible [ 65.566744] [drm] ib test on ring 5 succeeded
Let me know if you need more stuff dumped.
On 2016-06-08 15:47, Borislav Petkov wrote:
On Wed, Jun 08, 2016 at 03:30:34PM +0200, Christian König wrote:
Try forcing mplayer to use VDPAU with "mplayer -vo vdpau $file".
All good. Actually, this hw accel thing is much better, I better make it default :-P
Are you sure it is using accelerated decoding? CPU load should be just 1-2%. AFAIR you still have to explicitly enable the hardware accelerated codecs with mplayer. It might have changed at some point. Also try mpv instead, maybe. It's easier to enable decoding with that player, just specify --hwdec=vdpau on the command line and it'll use accelerated decoding whenever it can.
Grigori
On Wed, Jun 08, 2016 at 05:28:53PM +0200, Grigori Goronzy wrote:
Are you sure it is using accelerated decoding? CPU load should be just 1-2%.
Ha, good point. So with mplayer vo=vdpau, CPU load was at something over 12%.
Doing:
$ mpv --vo=vdpau --hwdec=vdpau $file (+) Video --vid=1 (*) (h264) (+) Audio --aid=1 --alang=eng (*) (aac) Trying to use hardware decoding. AO: [alsa] 48000Hz stereo 2ch float VO: [vdpau] 1280x720 vdpau AV: 00:00:50 / 00:02:36 (32%) A-V: 0.000
did bring it down to 5-6ish.
PID USER PR NI VIRT RES SHR S %CPU %MEM TIME+ COMMAND 12482 boris 20 0 1026628 74272 47632 S 5.6 0.5 0:02.77 mpv
so I'm guessing this should be using hw accel now...
In any case, output looks fine again too.
Thanks.
On Wed, Jun 8, 2016 at 12:32 PM, Borislav Petkov bp@alien8.de wrote:
On Wed, Jun 08, 2016 at 05:28:53PM +0200, Grigori Goronzy wrote:
Are you sure it is using accelerated decoding? CPU load should be just 1-2%.
Ha, good point. So with mplayer vo=vdpau, CPU load was at something over 12%.
Doing:
$ mpv --vo=vdpau --hwdec=vdpau $file (+) Video --vid=1 (*) (h264) (+) Audio --aid=1 --alang=eng (*) (aac) Trying to use hardware decoding. AO: [alsa] 48000Hz stereo 2ch float VO: [vdpau] 1280x720 vdpau AV: 00:00:50 / 00:02:36 (32%) A-V: 0.000
did bring it down to 5-6ish.
PID USER PR NI VIRT RES SHR S %CPU %MEM TIME+ COMMAND 12482 boris 20 0 1026628 74272 47632 S 5.6 0.5 0:02.77 mpv
so I'm guessing this should be using hw accel now...
In any case, output looks fine again too.
If the ring and IB tests pass on resume, you should be good to go.
Alex
Thanks.
-- Regards/Gruss, Boris.
ECO tip #101: Trim your mails when you reply. _______________________________________________ dri-devel mailing list dri-devel@lists.freedesktop.org https://lists.freedesktop.org/mailman/listinfo/dri-devel
On Wed, Jun 08, 2016 at 12:44:12PM -0400, Alex Deucher wrote:
If the ring and IB tests pass on resume, you should be good to go.
Yap, they do. I pasted that output earlier but here it is again:
[ 64.745988] [drm] ring test on 0 succeeded in 1 usecs [ 64.920633] [drm] ring test on 5 succeeded in 1 usecs [ 64.920637] [drm] UVD initialized successfully. [ 64.920652] [drm] ib test on ring 0 succeeded in 0 usecs
On Wed, Jun 8, 2016 at 12:51 PM, Borislav Petkov bp@alien8.de wrote:
On Wed, Jun 08, 2016 at 12:44:12PM -0400, Alex Deucher wrote:
If the ring and IB tests pass on resume, you should be good to go.
Yap, they do. I pasted that output earlier but here it is again:
[ 64.745988] [drm] ring test on 0 succeeded in 1 usecs [ 64.920633] [drm] ring test on 5 succeeded in 1 usecs [ 64.920637] [drm] UVD initialized successfully. [ 64.920652] [drm] ib test on ring 0 succeeded in 0 usecs
I've applied Jerome's fix to my tree.
Thanks,
Alex
-- Regards/Gruss, Boris.
ECO tip #101: Trim your mails when you reply.
dri-devel@lists.freedesktop.org
-
Alex Deucher -
Borislav Petkov -
Christian König -
Grigori Goronzy -
Jerome Glisse