SOF idle callback and HDA codec ordering #1003

kv2019i · 2019-06-03T10:27:07Z

Patchset to implement similar runtime PM idle logic as in HDA legacy driver to ensure codec is powered off first before HDA controller and DSP are put to suspend. To impelemnt this, add infrastructure support for the idle callback.

libinyang

LGTM

sound/soc/sof/intel/hda-dsp.c

RanderWang · 2019-06-04T08:26:59Z

sound/soc/sof/intel/hda-dsp.c

here please change %08X to %lX for I got warning with your patch

@RanderWang Thanks! The cast should avoid the warning, but agreed better to solve this without any cast. I'll upload an updated version.

plbossart

@kv2019i wondering why this is for CNL only, this could apply for all SKL+ platforms, no?
Also we should submit this on alsa-devel, this is a change in the SOF core and we should get feedback directly from a wider community.

sound/soc/sof/intel/cnl.c

ranj063 · 2019-06-04T19:34:11Z

Also we should submit this on alsa-devel, this is a change in the SOF core and we should get feedback directly from a wider community.

@plbossart @kv2019i we should run some suspend stress tests with this before we submit it to alsa-devel.
The last time i tested another variant of this, i saw issues with the SOF device not being able to suspend

RanderWang · 2019-06-10T06:13:12Z

I tested again and found one issue: when getting back from system s3, controller invokes idle function and finds a codec is still active, so controller doesn't go to runtime suspend. Then the codec goes to runtime suspend mode. But now the controller is always active until a playback or capture happens. It seems that there is no chance for suspend controller if idle function return false. I also found this issue with legacy hda driver.

RanderWang · 2019-06-10T08:21:19Z

I tested again and found one issue: when getting back from system s3, controller invokes idle function and finds a codec is still active, so controller doesn't go to runtime suspend. Then the codec goes to runtime suspend mode. But now the controller is always active until a playback or capture happens. It seems that there is no chance for suspend controller if idle function return false. I also found this issue with legacy hda driver.

My last comments was wrong. I debugged this issue and found I missed the commit for HDMI.
Now controller could go to suspend mode.

plbossart · 2019-06-10T16:38:47Z

@kv2019i is this PR still candidate for merge? Doesn't look like it's moved in 6 days.

kv2019i · 2019-06-11T07:56:59Z

@kv2019i is this PR still candidate for merge? Doesn't look like it's moved in 6 days.

@ranj063 , @RanderWang and @lyakh what do you think? Based on the validation data we have gathered since last week, it seems this is not strictly needed. That means the change is more about alignment to legacy driver rather than fixing a specific concrete problem. If we don't need the codec_powered based check, then we don't have a real need for the idle callback either.

In any case, this should first be sent to alsa-devel as it changes SOF, so I'll mark it "unclear" in GH.

RanderWang · 2019-06-12T06:55:23Z

@kv2019i is this PR still candidate for merge? Doesn't look like it's moved in 6 days.

@ranj063 , @RanderWang and @lyakh what do you think? Based on the validation data we have gathered since last week, it seems this is not strictly needed. That means the change is more about alignment to legacy driver rather than fixing a specific concrete problem. If we don't need the codec_powered based check, then we don't have a real need for the idle callback either.

In any case, this should first be sent to alsa-devel as it changes SOF, so I'll mark it "unclear" in GH.

It is needed by jack detection. When resuming from S3, controller is suspended before codecs, and WAKEEN feature doesn't work for this case. And I also checked this case with legacy hda driver which
also can't detect jack event. We need this patch to keep correct power sequence. Please refine it, thanks!

lyakh

subject to testing, since I cannot really evaluate the effects of this change. I also agree with Pierre's comment - looks like this should be added to all architectures.

kv2019i · 2019-06-12T09:50:23Z

Patch updated to include support for APL as well, and sent to alsa-devel for review as requested.
UPDATE: patch in alsa-devel https://mailman.alsa-project.org/pipermail/alsa-devel/2019-June/150923.html

RanderWang · 2019-06-12T11:44:14Z

please check ci build failed

kv2019i · 2019-06-12T11:51:58Z

please check ci build failed

Oh, interesting. It seems this has slipped in via commit dec1616 (already merged).

I can make a PR to fix this.
UDPATE: done -> #1030

ranj063 · 2019-06-12T17:48:56Z

@kv2019i I know this has already been sent upstream but just wanted to report I have run the S3 stress test with this patchset and it passed!

RanderWang

LGTM

ranj063 · 2019-06-13T17:43:36Z

@plbossart Takashi has already ack'ed these patches but the ones sent upstream are slightly different because of the ordering of the patches.

kv2019i · 2019-06-18T10:18:29Z

v2 sent to alsa-devel

ranj063

looks good. can be merged when the patches are applied upstream

RanderWang

LGTM

kv2019i · 2019-06-28T09:58:29Z

Still waiting for upstream to take these. I'll send a reminder next week if no activity in 2 weeks.

ranj063 · 2019-07-07T00:06:15Z

@kv2019i merged upstream. do you mind resolving the conflicts?

Report codec power status to the HDA codec bus from runtime pm suspend and resume callbacks. This is required to implement runtime idle logic that relies on 'codec_powered' field of hdac_bus to be maintained for all codecs. Signed-off-by: Kai Vehmanen <kai.vehmanen@linux.intel.com> Reviewed-by: Pierre-Louis Bossart <pierre-louis.bossart@linux.intel.com> Reviewed-by: Takashi Iwai <tiwai@suse.de> Link: https://lore.kernel.org/r/20190702132428.13129-2-kai.vehmanen@linux.intel.com Signed-off-by: Mark Brown <broonie@kernel.org>

Add ability to implement a SOF device level runtime idle callback. Signed-off-by: Kai Vehmanen <kai.vehmanen@linux.intel.com> Reviewed-by: Pierre-Louis Bossart <pierre-louis.bossart@linux.intel.com> Reviewed-by: Takashi Iwai <tiwai@suse.de> Link: https://lore.kernel.org/r/20190702132428.13129-3-kai.vehmanen@linux.intel.com Signed-off-by: Mark Brown <broonie@kernel.org>

Implement runtime idle for CNL/APL devices using similar runtime PM idle logic as the Intel AZX HDA driver. If any HDA codecs are powered when runtime suspend request comes, return -EBUSY. By doing this, strict ordering is enforced between HDA codec and the HDA controller. Signed-off-by: Kai Vehmanen <kai.vehmanen@linux.intel.com> Reviewed-by: Pierre-Louis Bossart <pierre-louis.bossart@linux.intel.com> Reviewed-by: Takashi Iwai <tiwai@suse.de> Link: https://lore.kernel.org/r/20190702132428.13129-4-kai.vehmanen@linux.intel.com Signed-off-by: Mark Brown <broonie@kernel.org>

kv2019i · 2019-07-08T12:54:26Z

@kv2019i merged upstream. do you mind resolving the conflicts?

Ack @ranj063 rebased the versions merged upstream, solved the one conflict and reuploaded.

plbossart · 2019-07-18T15:47:22Z

@kv2019i can you double-check all the patches were added with today's upstream merge and close this PR if we are good?

ranj063 · 2019-07-18T18:10:08Z

@plbossart Kai is on vacation for the next couple of weeks. I can confirm all patches in this PR are merged upstream.

Refactor the NE_FIT_TYPE split case when it comes to an allocation of one extra object. We need it in order to build a remaining space. The preload is done per CPU in non-atomic context with GFP_KERNEL flags. More permissive parameters can be beneficial for systems which are suffer from high memory pressure or low memory condition. For example on my KVM system(4xCPUs, no swap, 256MB RAM) i can simulate the failure of page allocation with GFP_NOWAIT flags. Using "stress-ng" tool and starting N workers spinning on fork() and exit(), i can trigger below trace: <snip> [ 179.815161] stress-ng-fork: page allocation failure: order:0, mode:0x40800(GFP_NOWAIT|__GFP_COMP), nodemask=(null),cpuset=/,mems_allowed=0 [ 179.815168] CPU: 0 PID: 12612 Comm: stress-ng-fork Not tainted 5.2.0-rc3+ thesofproject#1003 [ 179.815170] Hardware name: QEMU Standard PC (i440FX + PIIX, 1996), BIOS 1.10.2-1 04/01/2014 [ 179.815171] Call Trace: [ 179.815178] dump_stack+0x5c/0x7b [ 179.815182] warn_alloc+0x108/0x190 [ 179.815187] __alloc_pages_slowpath+0xdc7/0xdf0 [ 179.815191] __alloc_pages_nodemask+0x2de/0x330 [ 179.815194] cache_grow_begin+0x77/0x420 [ 179.815197] fallback_alloc+0x161/0x200 [ 179.815200] kmem_cache_alloc+0x1c9/0x570 [ 179.815202] alloc_vmap_area+0x32c/0x990 [ 179.815206] __get_vm_area_node+0xb0/0x170 [ 179.815208] __vmalloc_node_range+0x6d/0x230 [ 179.815211] ? _do_fork+0xce/0x3d0 [ 179.815213] copy_process.part.46+0x850/0x1b90 [ 179.815215] ? _do_fork+0xce/0x3d0 [ 179.815219] _do_fork+0xce/0x3d0 [ 179.815226] ? __do_page_fault+0x2bf/0x4e0 [ 179.815229] do_syscall_64+0x55/0x130 [ 179.815231] entry_SYSCALL_64_after_hwframe+0x44/0xa9 [ 179.815234] RIP: 0033:0x7fedec4c738b ... [ 179.815237] RSP: 002b:00007ffda469d730 EFLAGS: 00000246 ORIG_RAX: 0000000000000038 [ 179.815239] RAX: ffffffffffffffda RBX: 00007ffda469d730 RCX: 00007fedec4c738b [ 179.815240] RDX: 0000000000000000 RSI: 0000000000000000 RDI: 0000000001200011 [ 179.815241] RBP: 00007ffda469d780 R08: 00007fededd6e300 R09: 00007ffda47f50a0 [ 179.815242] R10: 00007fededd6e5d0 R11: 0000000000000246 R12: 0000000000000000 [ 179.815243] R13: 0000000000000020 R14: 0000000000000000 R15: 0000000000000000 [ 179.815245] Mem-Info: [ 179.815249] active_anon:12686 inactive_anon:14760 isolated_anon:0 active_file:502 inactive_file:61 isolated_file:70 unevictable:2 dirty:0 writeback:0 unstable:0 slab_reclaimable:2380 slab_unreclaimable:7520 mapped:15069 shmem:14813 pagetables:10833 bounce:0 free:1922 free_pcp:229 free_cma:0 <snip> Link: http://lkml.kernel.org/r/20190606120411.8298-3-urezki@gmail.com Signed-off-by: Uladzislau Rezki (Sony) <urezki@gmail.com> Cc: Hillf Danton <hdanton@sina.com> Cc: Matthew Wilcox <willy@infradead.org> Cc: Michal Hocko <mhocko@suse.com> Cc: Oleksiy Avramchenko <oleksiy.avramchenko@sonymobile.com> Cc: Roman Gushchin <guro@fb.com> Cc: Steven Rostedt <rostedt@goodmis.org> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>

kv2019i requested review from RanderWang, keyonjie, lyakh, plbossart, ranj063 and zhuyingjiang June 3, 2019 10:27

kv2019i requested review from bardliao, juimonen, lgirdwood and libinyang as code owners June 3, 2019 10:27

kv2019i force-pushed the topic/sof-idle-callback branch from ba94c13 to d4d1af0 Compare June 3, 2019 10:27

libinyang previously approved these changes Jun 3, 2019

View reviewed changes

plbossart reviewed Jun 3, 2019

View reviewed changes

sound/soc/sof/intel/hda-dsp.c Outdated Show resolved Hide resolved

RanderWang reviewed Jun 4, 2019

View reviewed changes

kv2019i dismissed libinyang’s stale review via 4ef8b66 June 4, 2019 14:20

kv2019i force-pushed the topic/sof-idle-callback branch from d4d1af0 to 4ef8b66 Compare June 4, 2019 14:20

plbossart requested changes Jun 4, 2019

View reviewed changes

sound/soc/sof/intel/cnl.c Outdated Show resolved Hide resolved

kv2019i added the Unclear No agreement on problem statement and resolution label Jun 11, 2019

RanderWang mentioned this pull request Jun 11, 2019

Enable WAKEEN feature in sof hda driver #983

Merged

lyakh previously approved these changes Jun 12, 2019

View reviewed changes

kv2019i dismissed lyakh’s stale review via f570c84 June 12, 2019 08:37

kv2019i force-pushed the topic/sof-idle-callback branch from 4ef8b66 to f570c84 Compare June 12, 2019 08:37

kv2019i removed the Unclear No agreement on problem statement and resolution label Jun 12, 2019

kv2019i added the upstream Patch has been sent to upstream (e.g. alsa-devel) label Jun 12, 2019

RanderWang previously approved these changes Jun 13, 2019

View reviewed changes

ranj063 mentioned this pull request Jun 13, 2019

[CFL][HDA] "load fw failed ret: -110" after arecord several times. #969

Closed

ranj063 previously approved these changes Jun 13, 2019

View reviewed changes

RanderWang mentioned this pull request Jun 14, 2019

[BUG] No update in sound-setting upon headset hotplug after long time idle #909

Closed

kv2019i dismissed stale reviews from ranj063 and RanderWang via 0372519 June 18, 2019 09:17

kv2019i force-pushed the topic/sof-idle-callback branch from f570c84 to 0372519 Compare June 18, 2019 09:17

ranj063 previously approved these changes Jun 18, 2019

View reviewed changes

RanderWang previously approved these changes Jun 19, 2019

View reviewed changes

kv2019i added 3 commits July 8, 2019 15:41

kv2019i dismissed stale reviews from RanderWang and ranj063 via 1d070fd July 8, 2019 12:53

kv2019i force-pushed the topic/sof-idle-callback branch from 0372519 to 1d070fd Compare July 8, 2019 12:53

ranj063 closed this Jul 18, 2019

SOF idle callback and HDA codec ordering #1003

SOF idle callback and HDA codec ordering #1003

Uh oh!

Conversation

kv2019i commented Jun 3, 2019

Uh oh!

libinyang left a comment

Choose a reason for hiding this comment

Uh oh!

Uh oh!

RanderWang Jun 4, 2019

Choose a reason for hiding this comment

Uh oh!

kv2019i Jun 4, 2019

Choose a reason for hiding this comment

Uh oh!

plbossart left a comment

Choose a reason for hiding this comment

Uh oh!

Uh oh!

ranj063 commented Jun 4, 2019

Uh oh!

RanderWang commented Jun 10, 2019

Uh oh!

RanderWang commented Jun 10, 2019

Uh oh!

plbossart commented Jun 10, 2019

Uh oh!

kv2019i commented Jun 11, 2019

Uh oh!

RanderWang commented Jun 12, 2019 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Uh oh!

lyakh left a comment

Choose a reason for hiding this comment

Uh oh!

kv2019i commented Jun 12, 2019 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Uh oh!

RanderWang commented Jun 12, 2019

Uh oh!

kv2019i commented Jun 12, 2019 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Uh oh!

ranj063 commented Jun 12, 2019

Uh oh!

RanderWang left a comment

Choose a reason for hiding this comment

Uh oh!

ranj063 commented Jun 13, 2019

Uh oh!

kv2019i commented Jun 18, 2019

Uh oh!

ranj063 left a comment

Choose a reason for hiding this comment

Uh oh!

RanderWang left a comment

Choose a reason for hiding this comment

Uh oh!

kv2019i commented Jun 28, 2019

Uh oh!

ranj063 commented Jul 7, 2019

Uh oh!

kv2019i commented Jul 8, 2019

Uh oh!

plbossart commented Jul 18, 2019

Uh oh!

ranj063 commented Jul 18, 2019

Uh oh!

Reviewers

Assignees

Labels

Projects

Milestone

Development

Uh oh!

6 participants

RanderWang commented Jun 12, 2019 •

edited

Loading

kv2019i commented Jun 12, 2019 •

edited

Loading

kv2019i commented Jun 12, 2019 •

edited

Loading