[PATCH for-4.15] ASoC: rt5514: don't assume rt5514 component was "attached"

[PATCH for-4.15] ASoC: rt5514: don't assume rt5514 component was "attached"

[Brian Norris]

I've found that on Google's "Kevin" Chromebook, the rt5514 codec might
not be set up completely, yet its device is still present, and therefore
its PM suspend/resume is called. This hits a NULL pointer exception,
since we never had the chance to set our drvdata pointer.

This resolves crashes seen when trying to resume my system.

Fixes: e9c50aa6bd39 ("ASoC: rt5514-spi: check irq status to schedule data copy in resume function")
Signed-off-by: Brian Norris <briannorris@chromium.org>
---
This is a v4.15-rc1 regression

 sound/soc/codecs/rt5514-spi.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/sound/soc/codecs/rt5514-spi.c b/sound/soc/codecs/rt5514-spi.c
index 2df91db765ac..9255afcf2c3a 100644
--- a/sound/soc/codecs/rt5514-spi.c
+++ b/sound/soc/codecs/rt5514-spi.c
@@ -482,7 +482,7 @@ static int __maybe_unused rt5514_resume(struct device *dev)
 	if (device_may_wakeup(dev))
 		disable_irq_wake(irq);
 
-	if (rt5514_dsp->substream) {
+	if (rt5514_dsp && rt5514_dsp->substream) {
 		rt5514_spi_burst_read(RT5514_IRQ_CTRL, (u8 *)&buf, sizeof(buf));
 		if (buf[0] & RT5514_IRQ_STATUS_BIT)
 			rt5514_schedule_copy(rt5514_dsp);
-- 
2.15.1.504.g5279b80103-goog

Re: [PATCH for-4.15] ASoC: rt5514: don't assume rt5514 component was "attached"

[Brian Norris]

+ others

On Fri, Dec 15, 2017 at 05:12:30PM -0800, Brian Norris wrote:
> I've found that on Google's "Kevin" Chromebook, the rt5514 codec might
> not be set up completely, yet its device is still present, and therefore
> its PM suspend/resume is called. This hits a NULL pointer exception,
> since we never had the chance to set our drvdata pointer.
> 
> This resolves crashes seen when trying to resume my system.
> 
> Fixes: e9c50aa6bd39 ("ASoC: rt5514-spi: check irq status to schedule data copy in resume function")
> Signed-off-by: Brian Norris <briannorris@chromium.org>
> ---
> This is a v4.15-rc1 regression

I see that this is probably a bug on its own, which should be fixed on
its own (e.g., with the $subject patch), but FYI I was also pointed at
this patch, which also "resolves" this problem by fixing the rt5514 DAI
setup:

https://patchwork.kernel.org/patch/10067725/
[PATCH] ASoC: rockchip: Use dummy_dai for rt5514 dsp dailink

Would be nice to see it get handled too, possibly in 4.15 as well (I
think that's a regression too?).

Brian

> 
>  sound/soc/codecs/rt5514-spi.c | 2 +-
>  1 file changed, 1 insertion(+), 1 deletion(-)
> 
> diff --git a/sound/soc/codecs/rt5514-spi.c b/sound/soc/codecs/rt5514-spi.c
> index 2df91db765ac..9255afcf2c3a 100644
> --- a/sound/soc/codecs/rt5514-spi.c
> +++ b/sound/soc/codecs/rt5514-spi.c
> @@ -482,7 +482,7 @@ static int __maybe_unused rt5514_resume(struct device *dev)
>  	if (device_may_wakeup(dev))
>  		disable_irq_wake(irq);
>  
> -	if (rt5514_dsp->substream) {
> +	if (rt5514_dsp && rt5514_dsp->substream) {
>  		rt5514_spi_burst_read(RT5514_IRQ_CTRL, (u8 *)&buf, sizeof(buf));
>  		if (buf[0] & RT5514_IRQ_STATUS_BIT)
>  			rt5514_schedule_copy(rt5514_dsp);
> -- 
> 2.15.1.504.g5279b80103-goog
> 

Re: [PATCH for-4.15] ASoC: rt5514: don't assume rt5514 component was "attached"

[Brian Norris]

Hi!

(By the way, your mail is HTML and likely will get rejected by many
mailing lists and/or people reading these mailing lists.)

On Mon, Dec 18, 2017 at 12:23:18PM +0800, Cheng-yi Chiang wrote:
>    Hi Brian,
>      Oder has posted the same fix : [1]https://patchwork.kernel.org/
>    patch/10066257/ and it has been applied.

OK cool. Obviously I'm biased, but I prefer mine, as it has less
needless whitespace, and is appropriately documented ;) But it should be
a fine replacement.

>    Perhaps you can cherry-pick it to v4.15 ?

I have no say in that; that would be up to Mark, I think. It's most
certainly a regression in 4.15-rc1, so IMO it should be given a proper
'Fixes: e9c50aa6bd39 ("ASoC: rt5514-spi: check irq status to schedule
data copy in resume function")' tag and sent for 4.15, not 4.16.

Brian

Re: [PATCH for-4.15] ASoC: rt5514: don't assume rt5514 component was "attached"

[Cheng-yi Chiang]

Hi Brian, I am sorry for not using the plain text mode in the previous mail.
I agree with you on other points.

On Tue, Dec 19, 2017 at 1:42 AM, Brian Norris <briannorris@chromium.org> wrote:
> Hi!
>
> (By the way, your mail is HTML and likely will get rejected by many
> mailing lists and/or people reading these mailing lists.)
>
> On Mon, Dec 18, 2017 at 12:23:18PM +0800, Cheng-yi Chiang wrote:
>>    Hi Brian,
>>      Oder has posted the same fix : [1]https://patchwork.kernel.org/
>>    patch/10066257/ and it has been applied.
>
> OK cool. Obviously I'm biased, but I prefer mine, as it has less
> needless whitespace, and is appropriately documented ;) But it should be
> a fine replacement.
>
>>    Perhaps you can cherry-pick it to v4.15 ?
>
> I have no say in that; that would be up to Mark, I think. It's most
> certainly a regression in 4.15-rc1, so IMO it should be given a proper
> 'Fixes: e9c50aa6bd39 ("ASoC: rt5514-spi: check irq status to schedule
> data copy in resume function")' tag and sent for 4.15, not 4.16.
>
> Brian

Re: [PATCH for-4.15] ASoC: rt5514: don't assume rt5514 component was "attached"

[Mark Brown]

[-- Attachment #1: Type: text/plain, Size: 644 bytes --]

On Mon, Dec 18, 2017 at 09:42:36AM -0800, Brian Norris wrote:
> On Mon, Dec 18, 2017 at 12:23:18PM +0800, Cheng-yi Chiang wrote:

> >    Hi Brian,
> >      Oder has posted the same fix : [1]https://patchwork.kernel.org/
> >    patch/10066257/ and it has been applied.

> >    Perhaps you can cherry-pick it to v4.15 ?

> I have no say in that; that would be up to Mark, I think. It's most
> certainly a regression in 4.15-rc1, so IMO it should be given a proper
> 'Fixes: e9c50aa6bd39 ("ASoC: rt5514-spi: check irq status to schedule
> data copy in resume function")' tag and sent for 4.15, not 4.16.

It's been applied as a fix for some time.

[-- Attachment #2: signature.asc --]
[-- Type: application/pgp-signature, Size: 488 bytes --]

Re: [PATCH for-4.15] ASoC: rt5514: don't assume rt5514 component was "attached"

[Brian Norris]

On Tue, Dec 19, 2017 at 10:58:18AM +0000, Mark Brown wrote:
> It's been applied as a fix for some time.

Indeed it has. Sorry for missing that. I look forward to seeing it in a
release candidate, so my system will again work on mainline :)

Brian

Applied "ASoC: rt5514-spi: Check the validity of drvdata pointer on resume" to the asoc tree

[Mark Brown]

The patch

   ASoC: rt5514-spi: Check the validity of drvdata pointer on resume

has been applied to the asoc tree at

   https://git.kernel.org/pub/scm/linux/kernel/git/broonie/sound.git 

All being well this means that it will be integrated into the linux-next
tree (usually sometime in the next 24 hours) and sent to Linus during
the next merge window (or sooner if it is a bug fix), however if
problems are discovered then the patch may be dropped or reverted.  

You may get further e-mails resulting from automated or manual testing
and review of the tree, please engage with people reporting problems and
send followup patches addressing any issues that are reported if needed.

If any updates are required or you are submitting further changes they
should be sent as incremental updates against current git, existing
patches will not be replaced.

Please add any relevant lists and maintainers to the CCs when replying
to this mail.

Thanks,
Mark

>From 509bf3a7d43ab173abc354df2a859229ede043c0 Mon Sep 17 00:00:00 2001
From: Marc Zyngier <marc.zyngier@arm.com>
Date: Wed, 24 Jan 2018 14:50:00 +0000
Subject: [PATCH] ASoC: rt5514-spi: Check the validity of drvdata pointer on
 resume

The rt5514-spi driver seem to assume the validity of the drvdata pointer
on resume, which it may not be populated, leading to a not-so-nice crash.

This stems from the fact that rt5514_spi_pcm_probe() is never called on
my system (a kevin Chromebook). No idea why, but if it can happen, it
is worth fixing.

Fixes: e9c50aa6bd39 ("ASoC: rt5514-spi: check irq status to schedule data copy in resume function")
Signed-off-by: Marc Zyngier <marc.zyngier@arm.com>
Signed-off-by: Mark Brown <broonie@kernel.org>
Cc: stable@vger.kernel.org
---
 sound/soc/codecs/rt5514-spi.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/sound/soc/codecs/rt5514-spi.c b/sound/soc/codecs/rt5514-spi.c
index 2df91db765ac..9255afcf2c3a 100644
--- a/sound/soc/codecs/rt5514-spi.c
+++ b/sound/soc/codecs/rt5514-spi.c
@@ -482,7 +482,7 @@ static int __maybe_unused rt5514_resume(struct device *dev)
 	if (device_may_wakeup(dev))
 		disable_irq_wake(irq);
 
-	if (rt5514_dsp->substream) {
+	if (rt5514_dsp && rt5514_dsp->substream) {
 		rt5514_spi_burst_read(RT5514_IRQ_CTRL, (u8 *)&buf, sizeof(buf));
 		if (buf[0] & RT5514_IRQ_STATUS_BIT)
 			rt5514_schedule_copy(rt5514_dsp);
-- 
2.15.1

Re: Applied "ASoC: rt5514-spi: Check the validity of drvdata pointer on resume" to the asoc tree

[Mark Brown]

[-- Attachment #1: Type: text/plain, Size: 551 bytes --]

On Wed, Jan 24, 2018 at 03:36:06PM +0000, Mark Brown wrote:
> The patch
> 
>    ASoC: rt5514-spi: Check the validity of drvdata pointer on resume
> 
> has been applied to the asoc tree at
> 
>    https://git.kernel.org/pub/scm/linux/kernel/git/broonie/sound.git 

Dropped again - this was fixed with "ASoC: rt5514-spi: only enable wakeup
when fully initialized" which was already in Linus' tree.  Dunno if that
was part of why it went AWOL, I wouldn't have thought so as I usually
send a "this didn't apply" mail.  No idea what happened.

[-- Attachment #2: signature.asc --]
[-- Type: application/pgp-signature, Size: 488 bytes --]

Re: Applied "ASoC: rt5514-spi: Check the validity of drvdata pointer on resume" to the asoc tree

[Brian Norris]

Hi Marc and Mark,

Mark already noticed this was no longer needed, but I just wanted to
note things for the record, since Marc seemed curious.

On Wed, Jan 24, 2018 at 7:36 AM, Mark Brown <broonie@kernel.org> wrote:
...
>
> From 509bf3a7d43ab173abc354df2a859229ede043c0 Mon Sep 17 00:00:00 2001
> From: Marc Zyngier <marc.zyngier@arm.com>
> Date: Wed, 24 Jan 2018 14:50:00 +0000
> Subject: [PATCH] ASoC: rt5514-spi: Check the validity of drvdata pointer on
>  resume
>
> The rt5514-spi driver seem to assume the validity of the drvdata pointer
> on resume, which it may not be populated, leading to a not-so-nice crash.

This part was fixed with this:

346cccf88319 ASoC: rt5514: Add the sanity check for the driver_data in
the resume function

which is in 4.15-rc6

> This stems from the fact that rt5514_spi_pcm_probe() is never called on
> my system (a kevin Chromebook). No idea why, but if it can happen, it
> is worth fixing.

I believe the rk3399_gru_sound driver currently does not set up the
DAI links properly at the moment. This patch is still waiting, to fix
that:

https://patchwork.kernel.org/patch/10067725/
[PATCH] ASoC: rockchip: Use dummy_dai for rt5514 dsp dailinkgi

Brian

> Fixes: e9c50aa6bd39 ("ASoC: rt5514-spi: check irq status to schedule data copy in resume function")
> Signed-off-by: Marc Zyngier <marc.zyngier@arm.com>
> Signed-off-by: Mark Brown <broonie@kernel.org>
> Cc: stable@vger.kernel.org
> ---
>  sound/soc/codecs/rt5514-spi.c | 2 +-
>  1 file changed, 1 insertion(+), 1 deletion(-)
>
> diff --git a/sound/soc/codecs/rt5514-spi.c b/sound/soc/codecs/rt5514-spi.c
> index 2df91db765ac..9255afcf2c3a 100644
> --- a/sound/soc/codecs/rt5514-spi.c
> +++ b/sound/soc/codecs/rt5514-spi.c
> @@ -482,7 +482,7 @@ static int __maybe_unused rt5514_resume(struct device *dev)
>         if (device_may_wakeup(dev))
>                 disable_irq_wake(irq);
>
> -       if (rt5514_dsp->substream) {
> +       if (rt5514_dsp && rt5514_dsp->substream) {
>                 rt5514_spi_burst_read(RT5514_IRQ_CTRL, (u8 *)&buf, sizeof(buf));
>                 if (buf[0] & RT5514_IRQ_STATUS_BIT)
>                         rt5514_schedule_copy(rt5514_dsp);
> --
> 2.15.1
>