Skip to contentSkip to conversation

FAILED: patch "[PATCH] ASoC: mediatek: mt8192: Check runtime resume during probe" failed to apply to 6.12-stable tree

PATCH New [email protected] 2 participants 3 messages Jul 20 – Jul 23

Subsystems (4)
lore
Share & tools
lore

[email protected]

author maintainer
#0 lore raw

The patch below does not apply to the 6.12-stable tree.
If someone wants it applied there, or to any other stable or longterm tree, then please email the backport, including the original git commit id to <[email protected]>.

To reproduce the conflict and resubmit, you may use the following commands:

git fetch https://git.kernel.org/pub/scm/linux/kernel/git/stable/linux.git/ linux-6.12.y
git checkout FETCH_HEAD
git cherry-pick -x e24d5dde56a50946020b134fa8448869093db76a
# <resolve conflicts, build, test, etc.>
git commit -s
git send-email --to '<[email protected]>' --in-reply-to '2026072021-slicer-fleshed-2403@gregkh' --subject-prefix 'PATCH 6.12.y' 'HEAD^..'

Possible dependencies:

thanks,

greg k-h

------------------ original commit in Linus's tree ------------------

From e24d5dde56a50946020b134fa8448869093db76a Mon Sep 17 00:00:00 2001
From: =?UTF-8?q?C=C3=A1ssio=20Gabriel?= <[email protected]>
Date: Wed, 27 May 2026 10:55:47 -0300
Subject: [PATCH] ASoC: mediatek: mt8192: Check runtime resume during probe
MIME-Version: 1.0
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: 8bit

The MT8192 AFE probe enables runtime PM temporarily while reinitializing the regmap cache from hardware, but it uses pm_runtime_get_sync() without checking the return value. If runtime resume fails, probe keeps going without the device necessarily being accessible, and pm_runtime_get_sync() may leave the PM usage count incremented.

The regmap_reinit_cache() failure path also returns before dropping the temporary PM reference and before clearing pm_runtime_bypass_reg_ctl.

Use pm_runtime_resume_and_get() so resume failures do not leak a usage count, and clear the temporary bypass flag after dropping the probe PM reference on all regmap_reinit_cache() outcomes.

Fixes 125ab5d588b0 ("ASoC: mediatek: mt8192: add platform driver")
sound/soc/mediatek/mt8192/mt8192-afe-pcm.c +8 -4
index 9f5057eeeff9..db0ae44a86af 100644
@@ -2227,15 +2227,19 @@ static int mt8192_afe_pcm_dev_probe(struct platform_device *pdev)
 
 	/* enable clock for regcache get default value from hw */
 	afe_priv->pm_runtime_bypass_reg_ctl = true;
-	pm_runtime_get_sync(dev);
+	ret = pm_runtime_resume_and_get(dev);
+	if (ret) {
+		afe_priv->pm_runtime_bypass_reg_ctl = false;
+		return dev_err_probe(dev, ret, "failed to resume device\n");
+	}
 
 	ret = regmap_reinit_cache(afe->regmap, &mt8192_afe_regmap_config);
-	if (ret)
-		return dev_err_probe(dev, ret, "regmap_reinit_cache fail\n");
-
 	pm_runtime_put_sync(dev);
 	afe_priv->pm_runtime_bypass_reg_ctl = false;
 
+	if (ret)
+		return dev_err_probe(dev, ret, "regmap_reinit_cache fail\n");
+
 	regcache_cache_only(afe->regmap, true);
 	regcache_mark_dirty(afe->regmap);
 
maintainer
#1 lore raw

PATCH 1/2 ASoC: mediatek: mt8192-afe-pcm: Simplify probe() with local dev variable

From: Tang Bin <[email protected]>

[ Upstream commit 01981565c764c554cc96e2d30a71c42975171416 ]

Simplify the function mt8192_afe_pcm_dev_probe() by using local 'dev' instead of '&pdev->dev'.

Stable-dep-of: e24d5dde56a5 ("ASoC: mediatek: mt8192: Check runtime resume during probe")

---
 sound/soc/mediatek/mt8192/mt8192-afe-pcm.c | 19 +++++++++----------
 1 file changed, 9 insertions(+), 10 deletions(-)
sound/soc/mediatek/mt8192/mt8192-afe-pcm.c +9 -10
index 69ed34495d0f73..6638307935ee9c 100644
@@ -2159,27 +2159,26 @@ static int mt8192_afe_pcm_dev_probe(struct platform_device *pdev)
 {
 	struct mtk_base_afe *afe;
 	struct mt8192_afe_private *afe_priv;
-	struct device *dev;
+	struct device *dev = &pdev->dev;
 	struct reset_control *rstc;
 	int i, ret, irq_id;
 
-	ret = dma_set_mask_and_coherent(&pdev->dev, DMA_BIT_MASK(34));
+	ret = dma_set_mask_and_coherent(dev, DMA_BIT_MASK(34));
 	if (ret)
 		return ret;
 
-	afe = devm_kzalloc(&pdev->dev, sizeof(*afe), GFP_KERNEL);
+	afe = devm_kzalloc(dev, sizeof(*afe), GFP_KERNEL);
 	if (!afe)
 		return -ENOMEM;
 	platform_set_drvdata(pdev, afe);
 
-	afe->platform_priv = devm_kzalloc(&pdev->dev, sizeof(*afe_priv),
+	afe->platform_priv = devm_kzalloc(dev, sizeof(*afe_priv),
 					  GFP_KERNEL);
 	if (!afe->platform_priv)
 		return -ENOMEM;
 	afe_priv = afe->platform_priv;
 
-	afe->dev = &pdev->dev;
-	dev = afe->dev;
+	afe->dev = dev;
 
 	ret = of_reserved_mem_device_init(dev);
 	if (ret) {
@@ -2203,7 +2202,7 @@ static int mt8192_afe_pcm_dev_probe(struct platform_device *pdev)
 	if (ret)
 		return dev_err_probe(dev, ret, "failed to trigger audio reset\n");
 
-	ret = devm_pm_runtime_enable(&pdev->dev);
+	ret = devm_pm_runtime_enable(dev);
 	if (ret)
 		return ret;
 
@@ -2219,13 +2218,13 @@ static int mt8192_afe_pcm_dev_probe(struct platform_device *pdev)
 
 	/* enable clock for regcache get default value from hw */
 	afe_priv->pm_runtime_bypass_reg_ctl = true;
-	pm_runtime_get_sync(&pdev->dev);
+	pm_runtime_get_sync(dev);
 
 	ret = regmap_reinit_cache(afe->regmap, &mt8192_afe_regmap_config);
 	if (ret)
 		return dev_err_probe(dev, ret, "regmap_reinit_cache fail\n");
 
-	pm_runtime_put_sync(&pdev->dev);
+	pm_runtime_put_sync(dev);
 	afe_priv->pm_runtime_bypass_reg_ctl = false;
 
 	regcache_cache_only(afe->regmap, true);
@@ -2292,7 +2291,7 @@ static int mt8192_afe_pcm_dev_probe(struct platform_device *pdev)
 	afe->runtime_suspend = mt8192_afe_runtime_suspend;
 
 	/* register platform */
-	ret = devm_snd_soc_register_component(&pdev->dev,
+	ret = devm_snd_soc_register_component(dev,
 					      &mtk_afe_pcm_platform,
 					      afe->dai_drivers,
 					      afe->num_dai_drivers);
Signature · 1 line
2.53.0
maintainer
#2 lore raw

PATCH 2/2 ASoC: mediatek: mt8192: Check runtime resume during probe

From: Cássio Gabriel <[email protected]>

[ Upstream commit e24d5dde56a50946020b134fa8448869093db76a ]

The MT8192 AFE probe enables runtime PM temporarily while reinitializing the regmap cache from hardware, but it uses pm_runtime_get_sync() without checking the return value. If runtime resume fails, probe keeps going without the device necessarily being accessible, and pm_runtime_get_sync() may leave the PM usage count incremented.

The regmap_reinit_cache() failure path also returns before dropping the temporary PM reference and before clearing pm_runtime_bypass_reg_ctl.

Use pm_runtime_resume_and_get() so resume failures do not leak a usage count, and clear the temporary bypass flag after dropping the probe PM reference on all regmap_reinit_cache() outcomes.

Fixes 125ab5d588b0 ("ASoC: mediatek: mt8192: add platform driver")
---
 sound/soc/mediatek/mt8192/mt8192-afe-pcm.c | 12 ++++++++----
 1 file changed, 8 insertions(+), 4 deletions(-)
sound/soc/mediatek/mt8192/mt8192-afe-pcm.c +8 -4
index 6638307935ee9c..52939d481fa136 100644
@@ -2218,15 +2218,19 @@ static int mt8192_afe_pcm_dev_probe(struct platform_device *pdev)
 
 	/* enable clock for regcache get default value from hw */
 	afe_priv->pm_runtime_bypass_reg_ctl = true;
-	pm_runtime_get_sync(dev);
+	ret = pm_runtime_resume_and_get(dev);
+	if (ret) {
+		afe_priv->pm_runtime_bypass_reg_ctl = false;
+		return dev_err_probe(dev, ret, "failed to resume device\n");
+	}
 
 	ret = regmap_reinit_cache(afe->regmap, &mt8192_afe_regmap_config);
-	if (ret)
-		return dev_err_probe(dev, ret, "regmap_reinit_cache fail\n");
-
 	pm_runtime_put_sync(dev);
 	afe_priv->pm_runtime_bypass_reg_ctl = false;
 
+	if (ret)
+		return dev_err_probe(dev, ret, "regmap_reinit_cache fail\n");
+
 	regcache_cache_only(afe->regmap, true);
 	regcache_mark_dirty(afe->regmap);
 
Signature · 1 line
2.53.0