| Message ID | 20190206100549.GO3837@imbe.wolfsonmicro.main (mailing list archive) |
|---|---|
| State | Not Applicable |
| Headers | show |
| Series | ASoC: dapm: Check for NULL widget in dapm_update_dai_unlocked | expand |
On 2/6/19 11:05, Charles Keepax wrote: > DAIs linked to the dummy will not have an associated playback/capture > widget, so we need to skip the update in that case. > > Fixes: 078a85f2806f ("ASoC: dapm: Only power up active channels from a DAI") > Signed-off-by: Charles Keepax <ckeepax@opensource.cirrus.com> Tested-by: Sylwester Nawrocki <s.nawrocki@samsung.com> > --- > > Ok so that all makes sense, this patch is probably the best fix? It seems so, everything works well with such change, thank you. > sound/soc/soc-dapm.c | 3 +++ > 1 file changed, 3 insertions(+) > > diff --git a/sound/soc/soc-dapm.c b/sound/soc/soc-dapm.c > index 482ddb825fb59..5235d8828758a 100644 > --- a/sound/soc/soc-dapm.c > +++ b/sound/soc/soc-dapm.c > @@ -2570,6 +2570,9 @@ static int dapm_update_dai_unlocked(struct snd_pcm_substream *substream, > else > w = dai->capture_widget; > > + if (!w) > + return 0; > + > dev_dbg(dai->dev, "Update DAI routes for %s %s\n", dai->name, > dir == SNDRV_PCM_STREAM_PLAYBACK ? "playback" : "capture"); >
On Wed, 6 Feb 2019 at 11:05, Charles Keepax <ckeepax@opensource.cirrus.com> wrote: > > DAIs linked to the dummy will not have an associated playback/capture > widget, so we need to skip the update in that case. > > Fixes: 078a85f2806f ("ASoC: dapm: Only power up active channels from a DAI") > Signed-off-by: Charles Keepax <ckeepax@opensource.cirrus.com> > --- > > Ok so that all makes sense, this patch is probably the best fix? > > Thanks, > Charles For this particular issue (NULL-pointer): Reported-by: Krzysztof Kozlowski <krzk@kernel.org> Tested-by: Krzysztof Kozlowski <krzk@kernel.org> However now I see bug sleeping in atomic context: [ 64.000828] BUG: sleeping function called from invalid context at ../kernel/locking/mutex.c:908 [ 64.008483] in_atomic(): 1, irqs_disabled(): 128, pid: 353, name: aplay [ 64.014978] 2 locks held by aplay/353: [ 64.018671] #0: 1fb9b63d (&(&group->lock)->rlock){....}, at: snd_pcm_action_lock_irq+0x18/0x3c [ 64.027337] #1: 8b42bfe8 (&(&pri_dai->spinlock)->rlock){....}, at: i2s_trigger+0x130/0x6c8 [ 64.035654] irq event stamp: 8754 [ 64.038925] hardirqs last enabled at (8753): [<c0a78758>] _raw_spin_unlock_irq+0x24/0x5c [ 64.047094] hardirqs last disabled at (8754): [<c0a785a0>] _raw_spin_lock_irq+0x18/0x50 [ 64.055068] softirqs last enabled at (8096): [<c0102698>] __do_softirq+0x4f0/0x5e4 [ 64.062680] softirqs last disabled at (8083): [<c012ee94>] irq_exit+0x160/0x16c [ 64.069953] Preemption disabled at: [ 64.069956] [<00000000>] (null) [ 64.076700] CPU: 6 PID: 353 Comm: aplay Not tainted 5.0.0-rc5-next-20190206-00001-g101ffa564f78 #348 [ 64.085822] Hardware name: SAMSUNG EXYNOS (Flattened Device Tree) [ 64.091882] [<c0112318>] (unwind_backtrace) from [<c010df2c>] (show_stack+0x10/0x14) [ 64.099601] [<c010df2c>] (show_stack) from [<c0a5626c>] (dump_stack+0x98/0xc4) [ 64.106788] [<c0a5626c>] (dump_stack) from [<c0156ac8>] (___might_sleep+0x264/0x2cc) [ 64.114501] [<c0156ac8>] (___might_sleep) from [<c0a732b4>] (__mutex_lock+0x40/0xa98) [ 64.122300] [<c0a732b4>] (__mutex_lock) from [<c0a73d28>] (mutex_lock_nested+0x1c/0x24) [ 64.130275] [<c0a73d28>] (mutex_lock_nested) from [<c04c2838>] (clk_prepare_lock+0x50/0xf8) [ 64.138596] [<c04c2838>] (clk_prepare_lock) from [<c04c5f48>] (clk_core_get_rate+0xc/0x60) [ 64.146824] [<c04c5f48>] (clk_core_get_rate) from [<c07b419c>] (i2s_trigger+0x444/0x6c8) [ 64.154883] [<c07b419c>] (i2s_trigger) from [<c079cba0>] (soc_pcm_trigger+0x100/0x140) [ 64.162768] [<c079cba0>] (soc_pcm_trigger) from [<c07839c0>] (snd_pcm_action_single+0x38/0x80) [ 64.171349] [<c07839c0>] (snd_pcm_action_single) from [<c0783a5c>] (snd_pcm_action+0x54/0x5c) [ 64.179841] [<c0783a5c>] (snd_pcm_action) from [<c0783bd4>] (snd_pcm_action_lock_irq+0x28/0x3c) [ 64.188508] [<c0783bd4>] (snd_pcm_action_lock_irq) from [<c07860b0>] (snd_pcm_ioctl+0x920/0x123c) [ 64.197350] [<c07860b0>] (snd_pcm_ioctl) from [<c02aa6d4>] (do_vfs_ioctl+0xb0/0x9f8) [ 64.205054] [<c02aa6d4>] (do_vfs_ioctl) from [<c02ab050>] (ksys_ioctl+0x34/0x5c) [ 64.212418] [<c02ab050>] (ksys_ioctl) from [<c0101000>] (ret_fast_syscall+0x0/0x28) [ 64.220045] Exception stack(0xe69dffa8 to 0xe69dfff0) [ 64.225058] ffa0: 004391b8 00001770 00000004 00004142 00439340 00439340 [ 64.233218] ffc0: 004391b8 00001770 00001770 00000036 00001770 00000000 00000000 be8db940 [ 64.241361] ffe0: b6fa382c be8db8ec b6f32a3c b6e42bdc [ 64.246386] [ 64.247823] ====================================================== [ 64.253979] WARNING: possible circular locking dependency detected [ 64.260133] 5.0.0-rc5-next-20190206-00001-g101ffa564f78 #348 Tainted: G W [ 64.268193] ------------------------------------------------------ [ 64.274342] aplay/353 is trying to acquire lock: [ 64.278937] 51044846 (prepare_lock){+.+.}, at: clk_prepare_lock+0x50/0xf8 [ 64.285694] [ 64.285694] but task is already holding lock: [ 64.291500] 8b42bfe8 (&(&pri_dai->spinlock)->rlock){....}, at: i2s_trigger+0x130/0x6c8 [ 64.299387] [ 64.299387] which lock already depends on the new lock. [ 64.299387] [ 64.307534] [ 64.307534] the existing dependency chain (in reverse order) is: [ 64.314985] [ 64.314985] -> #1 (&(&pri_dai->spinlock)->rlock){....}: [ 64.321667] clk_mux_set_parent+0x34/0xb8 [ 64.326162] clk_core_set_parent_nolock+0x21c/0x54c [ 64.331535] clk_set_parent+0x38/0x6c [ 64.335696] of_clk_set_defaults+0x11c/0x384 [ 64.340460] of_clk_add_provider+0x8c/0xc8 [ 64.345054] samsung_i2s_probe+0x484/0x64c [ 64.349651] platform_drv_probe+0x6c/0xa4 [ 64.354153] really_probe+0x280/0x414 [ 64.358311] driver_probe_device+0x78/0x1c4 [ 64.362991] bus_for_each_drv+0x74/0xb8 [ 64.367323] __device_attach+0xd4/0x16c [ 64.371655] bus_probe_device+0x88/0x90 [ 64.375988] deferred_probe_work_func+0x4c/0xd0 [ 64.381017] process_one_work+0x228/0x810 [ 64.385520] worker_thread+0x30/0x570 [ 64.389681] kthread+0x134/0x164 [ 64.393405] ret_from_fork+0x14/0x20 [ 64.397477] (null) [ 64.400249] [ 64.400249] -> #0 (prepare_lock){+.+.}: [ 64.405539] __mutex_lock+0x7c/0xa98 [ 64.409610] mutex_lock_nested+0x1c/0x24 [ 64.414029] clk_prepare_lock+0x50/0xf8 [ 64.418362] clk_core_get_rate+0xc/0x60 [ 64.422695] i2s_trigger+0x444/0x6c8 [ 64.426768] soc_pcm_trigger+0x100/0x140 [ 64.431188] snd_pcm_action_single+0x38/0x80 [ 64.435953] snd_pcm_action+0x54/0x5c [ 64.440112] snd_pcm_action_lock_irq+0x28/0x3c [ 64.445052] snd_pcm_ioctl+0x920/0x123c [ 64.449386] do_vfs_ioctl+0xb0/0x9f8 [ 64.453457] ksys_ioctl+0x34/0x5c [ 64.457269] ret_fast_syscall+0x0/0x28 [ 64.461516] 0xbe8db8ec [ 64.464460] [ 64.464460] other info that might help us debug this: [ 64.464460] [ 64.472438] Possible unsafe locking scenario: [ 64.472438] [ 64.478327] CPU0 CPU1 [ 64.482831] ---- ---- [ 64.487336] lock(&(&pri_dai->spinlock)->rlock); [ 64.492017] lock(prepare_lock); [ 64.497823] lock(&(&pri_dai->spinlock)->rlock); [ 64.505017] lock(prepare_lock); [ 64.508306] [ 64.508306] *** DEADLOCK *** [ 64.508306] [ 64.514203] 2 locks held by aplay/353: [ 64.517935] #0: 1fb9b63d (&(&group->lock)->rlock){....}, at: snd_pcm_action_lock_irq+0x18/0x3c [ 64.526596] #1: 8b42bfe8 (&(&pri_dai->spinlock)->rlock){....}, at: i2s_trigger+0x130/0x6c8 [ 64.534915] [ 64.534915] stack backtrace: [ 64.539246] CPU: 6 PID: 353 Comm: aplay Tainted: G W 5.0.0-rc5-next-20190206-00001-g101ffa564f78 #348 [ 64.549734] Hardware name: SAMSUNG EXYNOS (Flattened Device Tree) [ 64.555802] [<c0112318>] (unwind_backtrace) from [<c010df2c>] (show_stack+0x10/0x14) [ 64.563515] [<c010df2c>] (show_stack) from [<c0a5626c>] (dump_stack+0x98/0xc4) [ 64.570708] [<c0a5626c>] (dump_stack) from [<c01846b0>] (print_circular_bug+0x210/0x334) [ 64.578765] [<c01846b0>] (print_circular_bug) from [<c01874ec>] (__lock_acquire+0x12cc/0x1a5c) [ 64.587344] [<c01874ec>] (__lock_acquire) from [<c01886d8>] (lock_acquire+0xe0/0x268) [ 64.595142] [<c01886d8>] (lock_acquire) from [<c0a732f0>] (__mutex_lock+0x7c/0xa98) [ 64.602768] [<c0a732f0>] (__mutex_lock) from [<c0a73d28>] (mutex_lock_nested+0x1c/0x24) [ 64.610740] [<c0a73d28>] (mutex_lock_nested) from [<c04c2838>] (clk_prepare_lock+0x50/0xf8) [ 64.619059] [<c04c2838>] (clk_prepare_lock) from [<c04c5f48>] (clk_core_get_rate+0xc/0x60) [ 64.627291] [<c04c5f48>] (clk_core_get_rate) from [<c07b419c>] (i2s_trigger+0x444/0x6c8) [ 64.635349] [<c07b419c>] (i2s_trigger) from [<c079cba0>] (soc_pcm_trigger+0x100/0x140) [ 64.643235] [<c079cba0>] (soc_pcm_trigger) from [<c07839c0>] (snd_pcm_action_single+0x38/0x80) [ 64.651815] [<c07839c0>] (snd_pcm_action_single) from [<c0783a5c>] (snd_pcm_action+0x54/0x5c) [ 64.660306] [<c0783a5c>] (snd_pcm_action) from [<c0783bd4>] (snd_pcm_action_lock_irq+0x28/0x3c) [ 64.668972] [<c0783bd4>] (snd_pcm_action_lock_irq) from [<c07860b0>] (snd_pcm_ioctl+0x920/0x123c) [ 64.677811] [<c07860b0>] (snd_pcm_ioctl) from [<c02aa6d4>] (do_vfs_ioctl+0xb0/0x9f8) [ 64.685522] [<c02aa6d4>] (do_vfs_ioctl) from [<c02ab050>] (ksys_ioctl+0x34/0x5c) [ 64.692887] [<c02ab050>] (ksys_ioctl) from [<c0101000>] (ret_fast_syscall+0x0/0x28) [ 64.700510] Exception stack(0xe69dffa8 to 0xe69dfff0) [ 64.705536] ffa0: 004391b8 00001770 00000004 00004142 00439340 00439340 [ 64.713684] ffc0: 004391b8 00001770 00001770 00000036 00001770 00000000 00000000 be8db940 [ 64.721828] ffe0: b6fa382c be8db8ec b6f32a3c b6e42bdc Best regards, Krzysztof
On Wed, Feb 06, 2019 at 11:22:33AM +0100, Krzysztof Kozlowski wrote: > On Wed, 6 Feb 2019 at 11:05, Charles Keepax > <ckeepax@opensource.cirrus.com> wrote: > > > > DAIs linked to the dummy will not have an associated playback/capture > > widget, so we need to skip the update in that case. > > > > Fixes: 078a85f2806f ("ASoC: dapm: Only power up active channels from a DAI") > > Signed-off-by: Charles Keepax <ckeepax@opensource.cirrus.com> > > --- > > > > Ok so that all makes sense, this patch is probably the best fix? > > > > Thanks, > > Charles > > For this particular issue (NULL-pointer): > Reported-by: Krzysztof Kozlowski <krzk@kernel.org> > Tested-by: Krzysztof Kozlowski <krzk@kernel.org> > > However now I see bug sleeping in atomic context: > > [ 64.000828] BUG: sleeping function called from invalid context at > ../kernel/locking/mutex.c:908 Does this probably definitely get fixed by reverting my patch? It's just a bit odd as this seems to be complaining about a clock operation in i2s_trigger and I don't think my patch should have any affect on the trigger callback. It should get run from either the dai_link DAPM event or hw_params, neither of which should happen in an atomic context. > [ 64.307534] the existing dependency chain (in reverse order) is: > [ 64.314985] > [ 64.314985] -> #1 (&(&pri_dai->spinlock)->rlock){....}: > [ 64.321667] clk_mux_set_parent+0x34/0xb8 > [ 64.326162] clk_core_set_parent_nolock+0x21c/0x54c > [ 64.331535] clk_set_parent+0x38/0x6c > [ 64.335696] of_clk_set_defaults+0x11c/0x384 > [ 64.340460] of_clk_add_provider+0x8c/0xc8 > [ 64.345054] samsung_i2s_probe+0x484/0x64c > [ 64.349651] platform_drv_probe+0x6c/0xa4 > [ 64.354153] really_probe+0x280/0x414 > [ 64.358311] driver_probe_device+0x78/0x1c4 > [ 64.362991] bus_for_each_drv+0x74/0xb8 > [ 64.367323] __device_attach+0xd4/0x16c > [ 64.371655] bus_probe_device+0x88/0x90 > [ 64.375988] deferred_probe_work_func+0x4c/0xd0 > [ 64.381017] process_one_work+0x228/0x810 > [ 64.385520] worker_thread+0x30/0x570 > [ 64.389681] kthread+0x134/0x164 > [ 64.393405] ret_from_fork+0x14/0x20 > [ 64.397477] (null) > [ 64.400249] > [ 64.400249] -> #0 (prepare_lock){+.+.}: > [ 64.405539] __mutex_lock+0x7c/0xa98 > [ 64.409610] mutex_lock_nested+0x1c/0x24 > [ 64.414029] clk_prepare_lock+0x50/0xf8 > [ 64.418362] clk_core_get_rate+0xc/0x60 > [ 64.422695] i2s_trigger+0x444/0x6c8 > [ 64.426768] soc_pcm_trigger+0x100/0x140 > [ 64.431188] snd_pcm_action_single+0x38/0x80 > [ 64.435953] snd_pcm_action+0x54/0x5c > [ 64.440112] snd_pcm_action_lock_irq+0x28/0x3c > [ 64.445052] snd_pcm_ioctl+0x920/0x123c > [ 64.449386] do_vfs_ioctl+0xb0/0x9f8 > [ 64.453457] ksys_ioctl+0x34/0x5c > [ 64.457269] ret_fast_syscall+0x0/0x28 > [ 64.461516] 0xbe8db8ec Thanks, Charles
On Wed, 6 Feb 2019 at 12:00, Charles Keepax <ckeepax@opensource.cirrus.com> wrote: > > On Wed, Feb 06, 2019 at 11:22:33AM +0100, Krzysztof Kozlowski wrote: > > On Wed, 6 Feb 2019 at 11:05, Charles Keepax > > <ckeepax@opensource.cirrus.com> wrote: > > > > > > DAIs linked to the dummy will not have an associated playback/capture > > > widget, so we need to skip the update in that case. > > > > > > Fixes: 078a85f2806f ("ASoC: dapm: Only power up active channels from a DAI") > > > Signed-off-by: Charles Keepax <ckeepax@opensource.cirrus.com> > > > --- > > > > > > Ok so that all makes sense, this patch is probably the best fix? > > > > > > Thanks, > > > Charles > > > > For this particular issue (NULL-pointer): > > Reported-by: Krzysztof Kozlowski <krzk@kernel.org> > > Tested-by: Krzysztof Kozlowski <krzk@kernel.org> > > > > However now I see bug sleeping in atomic context: > > > > [ 64.000828] BUG: sleeping function called from invalid context at > > ../kernel/locking/mutex.c:908 > > Does this probably definitely get fixed by reverting my patch? > It's just a bit odd as this seems to be complaining about a clock > operation in i2s_trigger and I don't think my patch should have > any affect on the trigger callback. It should get run from either > the dai_link DAPM event or hw_params, neither of which should > happen in an atomic context. Before this fixup, probably NULL pointer happened before any of this. I tried it now few times and the possible deadlock and sleeping in invalid context did not appear. It might be random/racy or totally unrelated to your change. Best regards, Krzysztof
On Wed, Feb 06, 2019 at 12:14:56PM +0100, Krzysztof Kozlowski wrote: > On Wed, 6 Feb 2019 at 12:00, Charles Keepax > <ckeepax@opensource.cirrus.com> wrote: > > > > On Wed, Feb 06, 2019 at 11:22:33AM +0100, Krzysztof Kozlowski wrote: > > > On Wed, 6 Feb 2019 at 11:05, Charles Keepax > > > <ckeepax@opensource.cirrus.com> wrote: > > > > > > > > DAIs linked to the dummy will not have an associated playback/capture > > > > widget, so we need to skip the update in that case. > > > > > > > > Fixes: 078a85f2806f ("ASoC: dapm: Only power up active channels from a DAI") > > > > Signed-off-by: Charles Keepax <ckeepax@opensource.cirrus.com> > > > > --- > > > > > > > > Ok so that all makes sense, this patch is probably the best fix? > > > > > > > > Thanks, > > > > Charles > > > > > > For this particular issue (NULL-pointer): > > > Reported-by: Krzysztof Kozlowski <krzk@kernel.org> > > > Tested-by: Krzysztof Kozlowski <krzk@kernel.org> > > > > > > However now I see bug sleeping in atomic context: > > > > > > [ 64.000828] BUG: sleeping function called from invalid context at > > > ../kernel/locking/mutex.c:908 > > > > Does this probably definitely get fixed by reverting my patch? > > It's just a bit odd as this seems to be complaining about a clock > > operation in i2s_trigger and I don't think my patch should have > > any affect on the trigger callback. It should get run from either > > the dai_link DAPM event or hw_params, neither of which should > > happen in an atomic context. > > Before this fixup, probably NULL pointer happened before any of this. > I tried it now few times and the possible deadlock and sleeping in > invalid context did not appear. It might be random/racy or totally > unrelated to your change. > Looking through I think this is an unrelated issue. Assuming the driver in question is (sound/soc/samsung/i2s.c). Inside i2s_trigger, there is a call to config_setup(i2s), which in turn will call clk_get_rate if i2s->quirks contains the flag QUIRK_NO_MUXPSR. The trigger callback can be made from an atomic context and clk_get_rate will take the prepare lock of the clock. The clock prepare lock is always a mutex which shouldn't be locked from an atomic context. So it seems like this will fail whenever that QUIRK_NO_MUXPSR flag is set, no idea what causes that to be set. It looks like this bug was introduced by this change: 647d04f8e07a ("ASoC: samsung: i2s: Ensure the RCLK rate is properly determined") Thanks, Charles
On 2/6/19 12:45, Charles Keepax wrote: > Looking through I think this is an unrelated issue. Assuming the > driver in question is (sound/soc/samsung/i2s.c). Inside > i2s_trigger, there is a call to config_setup(i2s), which in turn > will call clk_get_rate if i2s->quirks contains the flag > QUIRK_NO_MUXPSR. > > The trigger callback can be made from an atomic context and > clk_get_rate will take the prepare lock of the clock. The clock > prepare lock is always a mutex which shouldn't be locked from an > atomic context. So it seems like this will fail whenever that > QUIRK_NO_MUXPSR flag is set, no idea what causes that to be set. > > It looks like this bug was introduced by this change: > > 647d04f8e07a ("ASoC: samsung: i2s: Ensure the RCLK rate is properly determined") Thanks or having a look at this. I somehow overlooked it before, there are multiple issues with that clk_get_rate() call. Apart of the problem described above config_setup() is already called with the i2s->lock spinlock held, exactly same spinlock that is passed to the clock API when registering a clock of which we try to get rate. :/ Presumably it works due to clk rate caching. Whether QUIRK_NO_MUXPSR flag is set or not depends on the HW type, it is not set for modern SoCs so most of the time we will hit the problem in I2S master configurations. As we are not using set_sysclk() of the CPU DAI I'm thinking about moving the clk_get_rate() call to the CPU DAI's hw_params() callback. The clock rate may be changed in hw_params() of an ASoC machine driver, and the CPU DAI needs to adjust to those changes. It feels like locking in that driver might need revisiting, there is quite a lot happening while holding a spinlock.
diff --git a/sound/soc/soc-dapm.c b/sound/soc/soc-dapm.c index 482ddb825fb59..5235d8828758a 100644 --- a/sound/soc/soc-dapm.c +++ b/sound/soc/soc-dapm.c @@ -2570,6 +2570,9 @@ static int dapm_update_dai_unlocked(struct snd_pcm_substream *substream, else w = dai->capture_widget; + if (!w) + return 0; + dev_dbg(dai->dev, "Update DAI routes for %s %s\n", dai->name, dir == SNDRV_PCM_STREAM_PLAYBACK ? "playback" : "capture");
DAIs linked to the dummy will not have an associated playback/capture widget, so we need to skip the update in that case. Fixes: 078a85f2806f ("ASoC: dapm: Only power up active channels from a DAI") Signed-off-by: Charles Keepax <ckeepax@opensource.cirrus.com> --- Ok so that all makes sense, this patch is probably the best fix? Thanks, Charles sound/soc/soc-dapm.c | 3 +++ 1 file changed, 3 insertions(+)