FAILED: patch "[PATCH] ASoC: mediatek: mt8192: Check runtime resume during probe" failed to apply to 5.15-stable tree
Subsystems (4)
[email protected]
maintainerThe patch below does not apply to the 5.15-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-5.15.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 '2026072022-fading-available-28ca@gregkh' --subject-prefix 'PATCH 5.15.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.
sound/soc/mediatek/mt8192/mt8192-afe-pcm.c +8 -4
@@ -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);
Sasha Levin
maintainerPATCH 1/5 ASoC: mediatek: mt8192-afe-pcm: Convert to devm_pm_runtime_enable()
From: AngeloGioacchino Del Regno <[email protected]>
[ Upstream commit 7aaaa22de56ce0dae15fd9f42a69a1d1a7a6e078 ]
Switch from pm_runtime_enable() to devm_pm_runtime_enable(), allowing to remove all gotos from the probe function.
Stable-dep-of: e24d5dde56a5 ("ASoC: mediatek: mt8192: Check runtime resume during probe")
--- sound/soc/mediatek/mt8192/mt8192-afe-pcm.c | 46 ++++++++-------------- 1 file changed, 17 insertions(+), 29 deletions(-)
sound/soc/mediatek/mt8192/mt8192-afe-pcm.c +17 -29
@@ -2221,21 +2221,20 @@ static int mt8192_afe_pcm_dev_probe(struct platform_device *pdev)return ret; } -pm_runtime_enable (&pdev->dev); -if (!pm_runtime_enabled (&pdev->dev)) - goto err_pm_disable; + ret =devm_pm_runtime_enable (&pdev->dev); +if (ret) +return ret;/* regmap init */ afe->regmap =syscon_node_to_regmap (dev->parent->of_node);if (IS_ERR (afe->regmap)) {dev_err (dev,"could not get regmap from parent\n" ); - ret =PTR_ERR (afe->regmap); - goto err_pm_disable; +return PTR_ERR (afe->regmap); } ret =regmap_attach_dev (dev, afe->regmap, &mt8192_afe_regmap_config);if (ret) {dev_warn (dev,"regmap_attach_dev fail, ret %d\n" , ret); - goto err_pm_disable; +return ret; }/* enable clock for regcache get default value from hw */
@@ -2245,7 +2244,7 @@ static int mt8192_afe_pcm_dev_probe(struct platform_device *pdev) ret =regmap_reinit_cache (afe->regmap, &mt8192_afe_regmap_config);if (ret) {dev_err (dev,"regmap_reinit_cache fail, ret %d\n" , ret); - goto err_pm_disable; +return ret; }pm_runtime_put_sync (&pdev->dev);
@@ -2258,10 +2257,8 @@ static int mt8192_afe_pcm_dev_probe(struct platform_device *pdev) afe->memif_size =MT8192_MEMIF_NUM ; afe->memif =devm_kcalloc (dev, afe->memif_size,sizeof (*afe->memif),GFP_KERNEL ); -if (!afe->memif) { - ret = -ENOMEM ; - goto err_pm_disable; - } +if (!afe->memif) +return -ENOMEM ;for (i =0 ; i < afe->memif_size; i++) { afe->memif[i].data = &memif_data[i];
@@ -2275,26 +2272,22 @@ static int mt8192_afe_pcm_dev_probe(struct platform_device *pdev) afe->irqs_size =MT8192_IRQ_NUM ; afe->irqs =devm_kcalloc (dev, afe->irqs_size,sizeof (*afe->irqs),GFP_KERNEL ); -if (!afe->irqs) { - ret = -ENOMEM ; - goto err_pm_disable; - } +if (!afe->irqs) +return -ENOMEM ;for (i =0 ; i < afe->irqs_size; i++) afe->irqs[i].irq_data = &irq_data[i];/* request irq */ irq_id =platform_get_irq (pdev,0 ); -if (irq_id <0 ) { - ret = irq_id; - goto err_pm_disable; - } +if (irq_id <0 ) +return irq_id; ret =devm_request_irq (dev, irq_id, mt8192_afe_irq_handler,IRQF_TRIGGER_NONE ,"asys-isr" , (void *)afe);if (ret) {dev_err (dev,"could not request_irq for Afe_ISR_Handle\n" ); - goto err_pm_disable; +return ret; }/* init sub_dais */
@@ -2305,7 +2298,7 @@ static int mt8192_afe_pcm_dev_probe(struct platform_device *pdev)if (ret) {dev_warn (afe->dev,"dai register i %d fail, ret %d\n" , i, ret); - goto err_pm_disable; +return ret; } }
@@ -2314,7 +2307,7 @@ static int mt8192_afe_pcm_dev_probe(struct platform_device *pdev)if (ret) {dev_warn (afe->dev,"mtk_afe_combine_sub_dai fail, ret %d\n" , ret); - goto err_pm_disable; +return ret; }/* others */
@@ -2333,7 +2326,7 @@ static int mt8192_afe_pcm_dev_probe(struct platform_device *pdev) &mt8192_afe_component, NULL, 0);if (ret) {dev_warn (dev,"err_platform\n" ); - goto err_pm_disable; +return ret; } ret = devm_snd_soc_register_component(&pdev->dev,
@@ -2342,15 +2335,10 @@ static int mt8192_afe_pcm_dev_probe(struct platform_device *pdev) afe->num_dai_drivers);if (ret) {dev_warn (dev,"err_dai_component\n" ); - goto err_pm_disable; +return ret; }return 0 ; - -err_pm_disable: -pm_runtime_disable (&pdev->dev); - -return ret; }static int mt8192_afe_pcm_dev_remove (struct platform_device *pdev)
Signature · 1 line
2.53.0
Sasha Levin
maintainerPATCH 2/5 ASoC: mediatek: mt8192-afe-pcm: Simplify with dev_err_probe()
From: AngeloGioacchino Del Regno <[email protected]>
[ Upstream commit 324c603a4efca7d1045e0bf3477ca54970eac72c ]
Simplify the probe function by switching error prints to return dev_err_probe(), lowering the lines count; while at it, also beautify some messages and change some others' level from warn to error.
Stable-dep-of: e24d5dde56a5 ("ASoC: mediatek: mt8192: Check runtime resume during probe")
--- sound/soc/mediatek/mt8192/mt8192-afe-pcm.c | 65 ++++++++-------------- 1 file changed, 22 insertions(+), 43 deletions(-)
sound/soc/mediatek/mt8192/mt8192-afe-pcm.c +22 -43
@@ -2209,17 +2209,12 @@ static int mt8192_afe_pcm_dev_probe(struct platform_device *pdev)/* reset controller to reset audio regs before regmap cache */ rstc =devm_reset_control_get_exclusive (dev,"audiosys" ); -if (IS_ERR (rstc)) { - ret =PTR_ERR (rstc); -dev_err (dev,"could not get audiosys reset:%d\n" , ret); -return ret; - } +if (IS_ERR (rstc)) +return dev_err_probe (dev,PTR_ERR (rstc),"could not get audiosys reset\n" ); ret =reset_control_reset (rstc); -if (ret) { -dev_err (dev,"failed to trigger audio reset:%d\n" , ret); -return ret; - } +if (ret) +return dev_err_probe (dev, ret,"failed to trigger audio reset\n" ); ret =devm_pm_runtime_enable (&pdev->dev);if (ret)
@@ -2227,25 +2222,21 @@ static int mt8192_afe_pcm_dev_probe(struct platform_device *pdev)/* regmap init */ afe->regmap =syscon_node_to_regmap (dev->parent->of_node); -if (IS_ERR (afe->regmap)) { -dev_err (dev,"could not get regmap from parent\n" ); -return PTR_ERR (afe->regmap); - } +if (IS_ERR (afe->regmap)) +return dev_err_probe (dev,PTR_ERR (afe->regmap), +"could not get regmap from parent" ); + ret =regmap_attach_dev (dev, afe->regmap, &mt8192_afe_regmap_config); -if (ret) { -dev_warn (dev,"regmap_attach_dev fail, ret %d\n" , ret); -return ret; - } +if (ret) +return dev_err_probe (dev, ret,"regmap_attach_dev fail\n" );/* enable clock for regcache get default value from hw */ afe_priv->pm_runtime_bypass_reg_ctl = true;pm_runtime_get_sync (&pdev->dev); ret =regmap_reinit_cache (afe->regmap, &mt8192_afe_regmap_config); -if (ret) { -dev_err (dev,"regmap_reinit_cache fail, ret %d\n" , ret); -return ret; - } +if (ret) +return dev_err_probe (dev, ret,"regmap_reinit_cache fail\n" );pm_runtime_put_sync (&pdev->dev); afe_priv->pm_runtime_bypass_reg_ctl = false;
@@ -2285,30 +2276,22 @@ static int mt8192_afe_pcm_dev_probe(struct platform_device *pdev) ret =devm_request_irq (dev, irq_id, mt8192_afe_irq_handler,IRQF_TRIGGER_NONE ,"asys-isr" , (void *)afe); -if (ret) { -dev_err (dev,"could not request_irq for Afe_ISR_Handle\n" ); -return ret; - } +if (ret) +return dev_err_probe (dev, ret,"could not request_irq for Afe_ISR_Handle\n" );/* init sub_dais */ INIT_LIST_HEAD (&afe->sub_dais);for (i =0 ; i <ARRAY_SIZE (dai_register_cbs); i++) { ret = dai_register_cbs[i](afe); -if (ret) { -dev_warn (afe->dev,"dai register i %d fail, ret %d\n" , - i, ret); -return ret; - } +if (ret) +return dev_err_probe (afe->dev, ret,"dai %d register fail" , i); }/* init dai_driver and component_driver */ ret =mtk_afe_combine_sub_dai (afe); -if (ret) { -dev_warn (afe->dev,"mtk_afe_combine_sub_dai fail, ret %d\n" , - ret); -return ret; - } +if (ret) +return dev_err_probe (afe->dev, ret,"mtk_afe_combine_sub_dai fail\n" );/* others */ afe->mtk_afe_hardware = &mt8192_afe_hardware;
@@ -2324,19 +2307,15 @@ static int mt8192_afe_pcm_dev_probe(struct platform_device *pdev)/* register platform */ ret =devm_snd_soc_register_component (&pdev->dev, &mt8192_afe_component,NULL ,0 ); -if (ret) { -dev_warn (dev,"err_platform\n" ); -return ret; - } +if (ret) +return dev_err_probe (dev, ret,"Couldn't register AFE component\n" ); ret =devm_snd_soc_register_component (&pdev->dev, &mt8192_afe_pcm_component, afe->dai_drivers, afe->num_dai_drivers); -if (ret) { -dev_warn (dev,"err_dai_component\n" ); -return ret; - } +if (ret) +return dev_err_probe (dev, ret,"Couldn't register AFE-PCM component\n" );return 0 ; }
Signature · 1 line
2.53.0
Sasha Levin
maintainerPATCH 3/5 ASoC: mediatek: Use common mtk_afe_pcm_platform with common probe cb
In reply to Sasha LevinFrom: AngeloGioacchino Del Regno <[email protected]>
[ Upstream commit 1d2a6b3b5189b6c3e51f29e050aa9b2337b774e8 ]
Since the mtk-afe-platform-driver generic mtk_afe_pcm_platform now has a common .probe() callback, there is no reason to keep duplicating this function over and over in the SoC specific AFE-PCM drivers: switch over to register with the common bits instead.
Note that MT8186 was left out of this because it is registering some extra sinegen controls in the AFE-PCM probe callback and needs extra cleanups to be able to use the common bits.
Stable-dep-of: e24d5dde56a5 ("ASoC: mediatek: mt8192: Check runtime resume during probe")
--- sound/soc/mediatek/mt8192/mt8192-afe-pcm.c | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-)
sound/soc/mediatek/mt8192/mt8192-afe-pcm.c +2 -2
@@ -2232,13 +2232,13 @@ 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);
Signature · 1 line
2.53.0
Sasha Levin
maintainerPATCH 4/5 ASoC: mediatek: mt8192-afe-pcm: Simplify probe() with local dev variable
In reply to Sasha LevinFrom: 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 | 17 ++++++++--------- 1 file changed, 8 insertions(+), 9 deletions(-)
sound/soc/mediatek/mt8192/mt8192-afe-pcm.c +8 -9
@@ -2178,27 +2178,26 @@ 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;/* init audio related clock */ ret =mt8192_init_clock (afe);
@@ -2216,7 +2215,7 @@ 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;
@@ -2305,12 +2304,12 @@ 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, &mt8192_afe_component,NULL ,0 );if (ret)return dev_err_probe (dev, ret,"Couldn't register AFE component\n" ); - ret =devm_snd_soc_register_component (&pdev->dev, + ret =devm_snd_soc_register_component (dev, &mt8192_afe_pcm_component, afe->dai_drivers, afe->num_dai_drivers);
Signature · 1 line
2.53.0
Sasha Levin
maintainerPATCH 5/5 ASoC: mediatek: mt8192: Check runtime resume during probe
In reply to Sasha LevinFrom: 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.
--- 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
@@ -2231,15 +2231,19 @@ 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