[PATCH 0/3] Fix PM disable depth imbalance on error
The pm_runtime_enable will increase power disable depth. Thus a pairing decrement is needed on the error handling path to keep it balanced according to context. This series of patches fixed it.
Zhang Qilong (3): ASoC: wm8994: Fix PM disable depth imbalance on error ASoC: wm8997: Fix PM disable depth imbalance on error ASoC: wm8998: Fix PM disable depth imbalance on error
sound/soc/codecs/wm8994.c | 6 +++++- sound/soc/codecs/wm8997.c | 9 +++++++-- sound/soc/codecs/wm8998.c | 4 +++- 3 files changed, 15 insertions(+), 4 deletions(-)
The pm_runtime_enable will increase power disable depth. Thus a pairing decrement is needed on the error handling path to keep it balanced according to context.
Fixes: 57e265c8d71fb ("ASoC: wm8994: Move runtime PM init to platform device init") Signed-off-by: Zhang Qilong zhangqilong3@huawei.com --- sound/soc/codecs/wm8994.c | 6 +++++- 1 file changed, 5 insertions(+), 1 deletion(-)
diff --git a/sound/soc/codecs/wm8994.c b/sound/soc/codecs/wm8994.c index fc9ea198ac79..f57884113406 100644 --- a/sound/soc/codecs/wm8994.c +++ b/sound/soc/codecs/wm8994.c @@ -4645,8 +4645,12 @@ static int wm8994_probe(struct platform_device *pdev) pm_runtime_enable(&pdev->dev); pm_runtime_idle(&pdev->dev);
- return devm_snd_soc_register_component(&pdev->dev, &soc_component_dev_wm8994, + ret = devm_snd_soc_register_component(&pdev->dev, &soc_component_dev_wm8994, wm8994_dai, ARRAY_SIZE(wm8994_dai)); + if (ret < 0) + pm_runtime_disable(&pdev->dev); + + return ret; }
static int wm8994_remove(struct platform_device *pdev)
On 11/11/2020 04:13, Zhang Qilong wrote:
The pm_runtime_enable will increase power disable depth. Thus a pairing decrement is needed on the error handling path to keep it balanced according to context.
Fixes: 57e265c8d71fb ("ASoC: wm8994: Move runtime PM init to platform device init") Signed-off-by: Zhang Qilong zhangqilong3@huawei.com
sound/soc/codecs/wm8994.c | 6 +++++- 1 file changed, 5 insertions(+), 1 deletion(-)
diff --git a/sound/soc/codecs/wm8994.c b/sound/soc/codecs/wm8994.c index fc9ea198ac79..f57884113406 100644 --- a/sound/soc/codecs/wm8994.c +++ b/sound/soc/codecs/wm8994.c @@ -4645,8 +4645,12 @@ static int wm8994_probe(struct platform_device *pdev) pm_runtime_enable(&pdev->dev); pm_runtime_idle(&pdev->dev);
- return devm_snd_soc_register_component(&pdev->dev, &soc_component_dev_wm8994,
ret = devm_snd_soc_register_component(&pdev->dev, &soc_component_dev_wm8994, wm8994_dai, ARRAY_SIZE(wm8994_dai));
if (ret < 0)
pm_runtime_disable(&pdev->dev);return ret; }
static int wm8994_remove(struct platform_device *pdev)
Reviewed-by: Richard Fitzgerald rf@opensource.cirrus.com
The pm_runtime_enable will increase power disable depth. Thus a pairing decrement is needed on the error handling path to keep it balanced according to context.
Fixes: 31833ead95c2c ("ASoC: arizona: Move request of speaker IRQs into bus probe") Signed-off-by: Zhang Qilong zhangqilong3@huawei.com --- sound/soc/codecs/wm8997.c | 9 +++++++-- 1 file changed, 7 insertions(+), 2 deletions(-)
diff --git a/sound/soc/codecs/wm8997.c b/sound/soc/codecs/wm8997.c index 37e4bb3dbd8a..744104edb2eb 100644 --- a/sound/soc/codecs/wm8997.c +++ b/sound/soc/codecs/wm8997.c @@ -1163,10 +1163,10 @@ static int wm8997_probe(struct platform_device *pdev)
ret = arizona_init_vol_limit(arizona); if (ret < 0) - return ret; + goto err_pm_disable; ret = arizona_init_spk_irqs(arizona); if (ret < 0) - return ret; + goto err_pm_disable;
ret = devm_snd_soc_register_component(&pdev->dev, &soc_component_dev_wm8997, @@ -1176,9 +1176,14 @@ static int wm8997_probe(struct platform_device *pdev) dev_err(&pdev->dev, "Failed to register component: %d\n", ret); goto err_spk_irqs; } + arizona_free_spk_irqs(arizona); + + return ret;
err_spk_irqs: arizona_free_spk_irqs(arizona); +err_pm_disable: + pm_runtime_disable(&pdev->dev);
return ret; }
On 11/11/2020 04:13, Zhang Qilong wrote:
The pm_runtime_enable will increase power disable depth. Thus a pairing decrement is needed on the error handling path to keep it balanced according to context.
Fixes: 31833ead95c2c ("ASoC: arizona: Move request of speaker IRQs into bus probe") Signed-off-by: Zhang Qilong zhangqilong3@huawei.com
sound/soc/codecs/wm8997.c | 9 +++++++-- 1 file changed, 7 insertions(+), 2 deletions(-)
diff --git a/sound/soc/codecs/wm8997.c b/sound/soc/codecs/wm8997.c index 37e4bb3dbd8a..744104edb2eb 100644 --- a/sound/soc/codecs/wm8997.c +++ b/sound/soc/codecs/wm8997.c @@ -1163,10 +1163,10 @@ static int wm8997_probe(struct platform_device *pdev)
ret = arizona_init_vol_limit(arizona); if (ret < 0)
return ret;
goto err_pm_disable;
return ret;
goto err_pm_disable;ret = devm_snd_soc_register_component(&pdev->dev, &soc_component_dev_wm8997,
@@ -1176,9 +1176,14 @@ static int wm8997_probe(struct platform_device *pdev) dev_err(&pdev->dev, "Failed to register component: %d\n", ret); goto err_spk_irqs; }
- arizona_free_spk_irqs(arizona);
Should not be freeing speaker irqs on successful probe. This looks like an existing bug in the driver where the return 0 was missing so it would always exit through the error path.
return ret;
err_spk_irqs: arizona_free_spk_irqs(arizona);
+err_pm_disable:
pm_runtime_disable(&pdev->dev);
return ret; }
On 11/11/2020 04:13, Zhang Qilong wrote:
The pm_runtime_enable will increase power disable depth. Thus a pairing decrement is needed on the error handling path to keep it balanced according to context.
Fixes: 31833ead95c2c ("ASoC: arizona: Move request of speaker IRQs into bus probe") Signed-off-by: Zhang Qilong zhangqilong3@huawei.com
sound/soc/codecs/wm8997.c | 9 +++++++-- 1 file changed, 7 insertions(+), 2 deletions(-)
diff --git a/sound/soc/codecs/wm8997.c b/sound/soc/codecs/wm8997.c index 37e4bb3dbd8a..744104edb2eb 100644 --- a/sound/soc/codecs/wm8997.c +++ b/sound/soc/codecs/wm8997.c @@ -1163,10 +1163,10 @@ static int wm8997_probe(struct
platform_device
*pdev)
ret = arizona_init_vol_limit(arizona); if (ret < 0)
return ret;
goto err_pm_disable;
return ret;
goto err_pm_disable;ret = devm_snd_soc_register_component(&pdev->dev, &soc_component_dev_wm8997, @@
-1176,9 +1176,14 @@ static
int wm8997_probe(struct platform_device *pdev) dev_err(&pdev->dev, "Failed to register component: %d\n", ret); goto err_spk_irqs; }
- arizona_free_spk_irqs(arizona);
Should not be freeing speaker irqs on successful probe. This looks like an existing bug in the driver where the return 0 was missing so it would always exit through the error path.
Yes, You are right, I had a little doubts here before, it should be freed when it is be removed. I will fix it next version.
Thans, Zhang
return ret;
err_spk_irqs: arizona_free_spk_irqs(arizona);
+err_pm_disable:
pm_runtime_disable(&pdev->dev);
return ret; }
The pm_runtime_enable will increase power disable depth. Thus a pairing decrement is needed on the error handling path to keep it balanced according to context.
Fixes: 31833ead95c2c ("ASoC: arizona: Move request of speaker IRQs into bus probe") Signed-off-by: Zhang Qilong zhangqilong3@huawei.com --- sound/soc/codecs/wm8998.c | 4 +++- 1 file changed, 3 insertions(+), 1 deletion(-)
diff --git a/sound/soc/codecs/wm8998.c b/sound/soc/codecs/wm8998.c index f6c5cc80c970..5413254295b7 100644 --- a/sound/soc/codecs/wm8998.c +++ b/sound/soc/codecs/wm8998.c @@ -1375,7 +1375,7 @@ static int wm8998_probe(struct platform_device *pdev)
ret = arizona_init_spk_irqs(arizona); if (ret < 0) - return ret; + goto err_pm_disable;
ret = devm_snd_soc_register_component(&pdev->dev, &soc_component_dev_wm8998, @@ -1390,6 +1390,8 @@ static int wm8998_probe(struct platform_device *pdev)
err_spk_irqs: arizona_free_spk_irqs(arizona); +err_pm_disable: + pm_runtime_disable(&pdev->dev);
return ret; }
On 11/11/2020 04:13, Zhang Qilong wrote:
The pm_runtime_enable will increase power disable depth. Thus a pairing decrement is needed on the error handling path to keep it balanced according to context.
Fixes: 31833ead95c2c ("ASoC: arizona: Move request of speaker IRQs into bus probe") Signed-off-by: Zhang Qilong zhangqilong3@huawei.com
sound/soc/codecs/wm8998.c | 4 +++- 1 file changed, 3 insertions(+), 1 deletion(-)
diff --git a/sound/soc/codecs/wm8998.c b/sound/soc/codecs/wm8998.c index f6c5cc80c970..5413254295b7 100644 --- a/sound/soc/codecs/wm8998.c +++ b/sound/soc/codecs/wm8998.c @@ -1375,7 +1375,7 @@ static int wm8998_probe(struct platform_device *pdev)
ret = arizona_init_spk_irqs(arizona); if (ret < 0)
return ret;
goto err_pm_disable;ret = devm_snd_soc_register_component(&pdev->dev, &soc_component_dev_wm8998,
@@ -1390,6 +1390,8 @@ static int wm8998_probe(struct platform_device *pdev)
err_spk_irqs: arizona_free_spk_irqs(arizona); +err_pm_disable:
pm_runtime_disable(&pdev->dev);
return ret; }
Reviewed-by: Richard Fitzgerald rf@opensource.cirrus.com
On Wed, 11 Nov 2020 12:13:23 +0800, Zhang Qilong wrote:
The pm_runtime_enable will increase power disable depth. Thus a pairing decrement is needed on the error handling path to keep it balanced according to context. This series of patches fixed it.
Zhang Qilong (3): ASoC: wm8994: Fix PM disable depth imbalance on error ASoC: wm8997: Fix PM disable depth imbalance on error ASoC: wm8998: Fix PM disable depth imbalance on error
[...]
Applied to
https://git.kernel.org/pub/scm/linux/kernel/git/broonie/sound.git for-next
Thanks!
[1/2] ASoC: wm8994: Fix PM disable depth imbalance on error commit: b8161cbe55a1892a19a318eaebbda92438fa708c [2/2] ASoC: wm8998: Fix PM disable depth imbalance on error commit: 193aa0a043645220d2a2f783ba06ae13d4601078
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
participants (4)
-
Mark Brown -
Richard Fitzgerald -
Zhang Qilong -
zhangqilong