ASoC: SOF: intel: set stream decouple false in stream_free #361

bardliao · 2018-12-04T05:19:22Z

snd_hdac_ext_stream_decouple(bus, stream, true) will be called
on boot up and hw_params. So add snd_hdac_ext_stream_decouple(bus,
stream, false) in hda_dsp_stream_free().

Signed-off-by: Bard liao yung-chuan.liao@linux.intel.com

Fix #299

plbossart

Not completely clear if this is done by symmetry with the inits+hw_params or if this is truly required in the hardware documentation?

plbossart · 2018-12-04T05:33:11Z

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

@@ -659,13 +659,15 @@ void hda_dsp_stream_free(struct snd_sof_dev *sdev)
 #endif

 	list_for_each_entry_safe(s, _s, &bus->stream_list, list) {
+		stream = stream_to_hdac_ext_stream(s);


should stream be defined conditionally? We don't want to use hdac_ext_stream in all cases

move snd_hdac_ext_stream_decouple(bus, stream, false) to hda_link_hw_free() as snd_hdac_ext_stream_decouple(bus,stream, true) is called by hda_link_hw_params(). Signed-off-by: Bard liao <yung-chuan.liao@linux.intel.com>

bardliao · 2018-12-06T06:59:59Z

@plbossart decouple is required from hardware. All streams will be configured as decoupled in init, that's done by hdac_ext_link_stream_assign(). I don't feel comfortable with current code. It is set in hw_params and pcm_trigger which is not symmetry. How about we move snd_hdac_ext_stream_decouple(false) to hw_free? So it will be configured to couple when the stream is not active and we don't need to configure it when the device is removed.

plbossart · 2018-12-06T15:47:28Z

@bardliao can you remind me how it's done in the existing SST driver?

bardliao · 2018-12-07T02:31:50Z

@plbossart Below is what I found, but I don't know if there is anything missed since I didn't test sst driver.

snd_hdac_ext_stream_decouple(true):

skl_tplg_mixer_event() SND_SOC_DAPM_PRE_PMU case -> ... -> snd_hdac_ext_stream_decouple()
skl_pcm_prepare() -> skl_pcm_host_dma_prepare() -> snd_hdac_ext_stream_decouple()
skl_pcm_open() -> snd_hdac_ext_stream_assign() -> ... -> snd_hdac_ext_stream_decouple()
skl_link_hw_params() -> snd_hdac_ext_stream_assign() -> ... -> snd_hdac_ext_stream_decouple()
skl_dsp_prepare() -> snd_hdac_ext_stream_assign() -> ... -> snd_hdac_ext_stream_decouple()

snd_hdac_ext_stream_decouple(false):

skl_shutdown() -> snd_hdac_ext_stream_decouple()
skl_pcm_trigger() SNDRV_PCM_TRIGGER_SUSPEND case -> snd_hdac_ext_stream_decouple(). 
skl_link_pcm_trigger() SNDRV_PCM_TRIGGER_SUSPEND case -> snd_hdac_ext_stream_decouple(). 
skl_remove() -> skl_free() -> snd_hdac_stream_free_all() -> snd_hdac_ext_stream_decouple(). 
skl_dsp_cleanup() -> snd_hdac_ext_stream_release() -> snd_hdac_ext_stream_decouple()
skl_pcm_close() -> snd_hdac_ext_stream_release() -> snd_hdac_ext_stream_decouple() (if the stream type is not _COUPLED)
skl_link_hw_free() -> snd_hdac_ext_stream_release() -> snd_hdac_ext_stream_decouple()

plbossart · 2019-01-15T22:28:53Z

@ranj063 we could use your help here.
The Skylake driver sets the "decouple_stream" to

true on .prepare
false on .trigger(suspend) and .shutdown

@bardliao showed we currently set it to

true on .hw_params
false on .trigger(suspend)

I agree his patch seems more logical to use hw_params/free but I am concerned about the suspend part. What is the flow on resume, would .prepare or .hw_params be reissued?

ranj063 · 2019-01-15T23:59:10Z

@plbossart let me look into the code and get back on this.

ranj063 · 2019-01-16T04:42:58Z

@plbossart yes, hw_params will definitely be called upon resume. I am not sure about prepare as we dont have that callback set in our hda dai link ops. But this change looks good to me.

Also, the suspend trigger was something I never saw only the stop trigger before suspend.

plbossart · 2019-01-16T20:10:51Z

@ranj063 you probably only saw the stop trigger because that's how CRAS operates (streams stopped). We'd still figure out if we can suspend a running stream, that's a valid ALSA use case.
Will apply this for now, let's see if this breaks anything...

When sock_alloc_file fails to allocate a file, it will call sock_release. __sys_socket_file should then not call sock_release again, otherwise there will be a double free. [ 89.319884] ------------[ cut here ]------------ [ 89.320286] kernel BUG at fs/inode.c:1764! [ 89.320656] invalid opcode: 0000 [thesofproject#1] PREEMPT SMP NOPTI [ 89.321051] CPU: 7 PID: 125 Comm: iou-sqp-124 Not tainted 6.2.0+ thesofproject#361 [ 89.321535] RIP: 0010:iput+0x1ff/0x240 [ 89.321808] Code: d1 83 e1 03 48 83 f9 02 75 09 48 81 fa 00 10 00 00 77 05 83 e2 01 75 1f 4c 89 ef e8 fb d2 ba 00 e9 80 fe ff ff c3 cc cc cc cc <0f> 0b 0f 0b e9 d0 fe ff ff 0f 0b eb 8d 49 8d b4 24 08 01 00 00 48 [ 89.322760] RSP: 0018:ffffbdd60068bd50 EFLAGS: 00010202 [ 89.323036] RAX: 0000000000000000 RBX: ffff9d7ad3cacac0 RCX: 0000000000001107 [ 89.323412] RDX: 000000000003af00 RSI: 0000000000000000 RDI: ffff9d7ad3cacb40 [ 89.323785] RBP: ffffbdd60068bd68 R08: ffffffffffffffff R09: ffffffffab606438 [ 89.324157] R10: ffffffffacb3dfa0 R11: 6465686361657256 R12: ffff9d7ad3cacb40 [ 89.324529] R13: 0000000080000001 R14: 0000000080000001 R15: 0000000000000002 [ 89.324904] FS: 00007f7b28516740(0000) GS:ffff9d7aeb1c0000(0000) knlGS:0000000000000000 [ 89.325328] CS: 0010 DS: 0000 ES: 0000 CR0: 0000000080050033 [ 89.325629] CR2: 00007f0af52e96c0 CR3: 0000000002a02006 CR4: 0000000000770ee0 [ 89.326004] PKRU: 55555554 [ 89.326161] Call Trace: [ 89.326298] <TASK> [ 89.326419] __sock_release+0xb5/0xc0 [ 89.326632] __sys_socket_file+0xb2/0xd0 [ 89.326844] io_socket+0x88/0x100 [ 89.327039] ? io_issue_sqe+0x6a/0x430 [ 89.327258] io_issue_sqe+0x67/0x430 [ 89.327450] io_submit_sqes+0x1fe/0x670 [ 89.327661] io_sq_thread+0x2e6/0x530 [ 89.327859] ? __pfx_autoremove_wake_function+0x10/0x10 [ 89.328145] ? __pfx_io_sq_thread+0x10/0x10 [ 89.328367] ret_from_fork+0x29/0x50 [ 89.328576] RIP: 0033:0x0 [ 89.328732] Code: Unable to access opcode bytes at 0xffffffffffffffd6. [ 89.329073] RSP: 002b:0000000000000000 EFLAGS: 00000202 ORIG_RAX: 00000000000001a9 [ 89.329477] RAX: 0000000000000000 RBX: 0000000000000000 RCX: 00007f7b28637a3d [ 89.329845] RDX: 00007fff4e4318a8 RSI: 00007fff4e4318b0 RDI: 0000000000000400 [ 89.330216] RBP: 00007fff4e431830 R08: 00007fff4e431711 R09: 00007fff4e4318b0 [ 89.330584] R10: 0000000000000000 R11: 0000000000000202 R12: 00007fff4e441b38 [ 89.330950] R13: 0000563835e3e725 R14: 0000563835e40d10 R15: 00007f7b28784040 [ 89.331318] </TASK> [ 89.331441] Modules linked in: [ 89.331617] ---[ end trace 0000000000000000 ]--- Fixes: da214a4 ("net: add __sys_socket_file()") Signed-off-by: Thadeu Lima de Souza Cascardo <cascardo@canonical.com> Reviewed-by: Jens Axboe <axboe@kernel.dk> Reviewed-by: Eric Dumazet <edumazet@google.com> Reviewed-by: Kuniyuki Iwashima <kuniyu@amazon.com> Link: https://lore.kernel.org/r/20230307173707.468744-1-cascardo@canonical.com Signed-off-by: Jakub Kicinski <kuba@kernel.org>

bardliao requested review from plbossart and lgirdwood December 4, 2018 05:19

plbossart requested changes Dec 4, 2018

View reviewed changes

ASoC: SOF: intel: set stream decouple false in hw_free

5768813

move snd_hdac_ext_stream_decouple(bus, stream, false) to hda_link_hw_free() as snd_hdac_ext_stream_decouple(bus,stream, true) is called by hda_link_hw_params(). Signed-off-by: Bard liao <yung-chuan.liao@linux.intel.com>

bardliao force-pushed the rmmod branch 2 times, most recently from d28113f to 5768813 Compare December 6, 2018 07:19

plbossart merged commit 6262b88 into thesofproject:topic/sof-dev Jan 16, 2019

Provide feedback

Saved searches

Use saved searches to filter your results more quickly

ASoC: SOF: intel: set stream decouple false in stream_free #361

ASoC: SOF: intel: set stream decouple false in stream_free #361

bardliao commented Dec 4, 2018

plbossart left a comment

plbossart Dec 4, 2018

bardliao commented Dec 6, 2018 •

edited

plbossart commented Dec 6, 2018

bardliao commented Dec 7, 2018

plbossart commented Jan 15, 2019

ranj063 commented Jan 15, 2019

ranj063 commented Jan 16, 2019 •

edited

plbossart commented Jan 16, 2019

ASoC: SOF: intel: set stream decouple false in stream_free #361

ASoC: SOF: intel: set stream decouple false in stream_free #361

Conversation

bardliao commented Dec 4, 2018

plbossart left a comment

Choose a reason for hiding this comment

plbossart Dec 4, 2018

Choose a reason for hiding this comment

bardliao commented Dec 6, 2018 • edited

plbossart commented Dec 6, 2018

bardliao commented Dec 7, 2018

plbossart commented Jan 15, 2019

ranj063 commented Jan 15, 2019

ranj063 commented Jan 16, 2019 • edited

plbossart commented Jan 16, 2019

bardliao commented Dec 6, 2018 •

edited

ranj063 commented Jan 16, 2019 •

edited