- Alsa-devel - mailman.alsa-project.org

[PATCH AUTOSEL 5.10 04/18] ASoC: rt5668: do not block workqueue if card is unbound
by Sasha Levin 23 Feb '22

23 Feb '22

From: Kai Vehmanen <kai.vehmanen(a)linux.intel.com> [ Upstream commit a6d78661dc903d90a327892bbc34268f3a5f4b9c ] The current rt5668_jack_detect_handler() assumes the component and card will always show up and implements an infinite usleep loop waiting for them to show up. This does not hold true if a codec interrupt (or other event) occurs when the card is unbound. The codec driver's remove or shutdown functions cannot cancel the workqueue due to the wait loop. As a result, code can either end up blocking the workqueue, or hit a kernel oops when the card is freed. Fix the issue by rescheduling the jack detect handler in case the card is not ready. In case card never shows up, the shutdown/remove/suspend calls can now cancel the detect task. Signed-off-by: Kai Vehmanen <kai.vehmanen(a)linux.intel.com> Reviewed-by: Bard Liao <yung-chuan.liao(a)linux.intel.com> Reviewed-by: Ranjani Sridharan <ranjani.sridharan(a)linux.intel.com> Reviewed-by: Pierre-Louis Bossart <pierre-louis.bossart(a)linux.intel.com> Reviewed-by: Péter Ujfalusi <peter.ujfalusi(a)linux.intel.com> Reviewed-by: Shuming Fan <shumingf(a)realtek.com> Link: https://lore.kernel.org/r/20220207153000.3452802-2-kai.vehmanen@linux.intel… Signed-off-by: Mark Brown <broonie(a)kernel.org> Signed-off-by: Sasha Levin <sashal(a)kernel.org> --- sound/soc/codecs/rt5668.c | 12 +++++++----- 1 file changed, 7 insertions(+), 5 deletions(-) diff --git a/sound/soc/codecs/rt5668.c b/sound/soc/codecs/rt5668.c index bc69adc9c8b70..e625df57c69e5 100644 --- a/sound/soc/codecs/rt5668.c +++ b/sound/soc/codecs/rt5668.c @@ -1022,11 +1022,13 @@ static void rt5668_jack_detect_handler(struct work_struct *work) container_of(work, struct rt5668_priv, jack_detect_work.work); int val, btn_type; - while (!rt5668->component) - usleep_range(10000, 15000); - - while (!rt5668->component->card->instantiated) - usleep_range(10000, 15000); + if (!rt5668->component || !rt5668->component->card || + !rt5668->component->card->instantiated) { + /* card not yet ready, try later */ + mod_delayed_work(system_power_efficient_wq, + &rt5668->jack_detect_work, msecs_to_jiffies(15)); + return; + } mutex_lock(&rt5668->calibrate_mutex); -- 2.34.1

1 0

[PATCH AUTOSEL 5.15 05/28] ASoC: rt5682: do not block workqueue if card is unbound
by Sasha Levin 23 Feb '22

23 Feb '22

From: Kai Vehmanen <kai.vehmanen(a)linux.intel.com> [ Upstream commit 4c33de0673ced9c7c37b3bbd9bfe0fda72340b2a ] The current rt5682_jack_detect_handler() assumes the component and card will always show up and implements an infinite usleep loop waiting for them to show up. This does not hold true if a codec interrupt (or other event) occurs when the card is unbound. The codec driver's remove or shutdown functions cannot cancel the workqueue due to the wait loop. As a result, code can either end up blocking the workqueue, or hit a kernel oops when the card is freed. Fix the issue by rescheduling the jack detect handler in case the card is not ready. In case card never shows up, the shutdown/remove/suspend calls can now cancel the detect task. Signed-off-by: Kai Vehmanen <kai.vehmanen(a)linux.intel.com> Reviewed-by: Bard Liao <yung-chuan.liao(a)linux.intel.com> Reviewed-by: Ranjani Sridharan <ranjani.sridharan(a)linux.intel.com> Reviewed-by: Pierre-Louis Bossart <pierre-louis.bossart(a)linux.intel.com> Reviewed-by: Péter Ujfalusi <peter.ujfalusi(a)linux.intel.com> Reviewed-by: Shuming Fan <shumingf(a)realtek.com> Link: https://lore.kernel.org/r/20220207153000.3452802-3-kai.vehmanen@linux.intel… Signed-off-by: Mark Brown <broonie(a)kernel.org> Signed-off-by: Sasha Levin <sashal(a)kernel.org> --- sound/soc/codecs/rt5682.c | 12 +++++++----- 1 file changed, 7 insertions(+), 5 deletions(-) diff --git a/sound/soc/codecs/rt5682.c b/sound/soc/codecs/rt5682.c index 757d1362b5f48..6ad3159eceb98 100644 --- a/sound/soc/codecs/rt5682.c +++ b/sound/soc/codecs/rt5682.c @@ -1090,11 +1090,13 @@ void rt5682_jack_detect_handler(struct work_struct *work) container_of(work, struct rt5682_priv, jack_detect_work.work); int val, btn_type; - while (!rt5682->component) - usleep_range(10000, 15000); - - while (!rt5682->component->card->instantiated) - usleep_range(10000, 15000); + if (!rt5682->component || !rt5682->component->card || + !rt5682->component->card->instantiated) { + /* card not yet ready, try later */ + mod_delayed_work(system_power_efficient_wq, + &rt5682->jack_detect_work, msecs_to_jiffies(15)); + return; + } mutex_lock(&rt5682->calibrate_mutex); -- 2.34.1

1 0

[PATCH AUTOSEL 5.15 04/28] ASoC: rt5668: do not block workqueue if card is unbound
by Sasha Levin 23 Feb '22

23 Feb '22

From: Kai Vehmanen <kai.vehmanen(a)linux.intel.com> [ Upstream commit a6d78661dc903d90a327892bbc34268f3a5f4b9c ] The current rt5668_jack_detect_handler() assumes the component and card will always show up and implements an infinite usleep loop waiting for them to show up. This does not hold true if a codec interrupt (or other event) occurs when the card is unbound. The codec driver's remove or shutdown functions cannot cancel the workqueue due to the wait loop. As a result, code can either end up blocking the workqueue, or hit a kernel oops when the card is freed. Fix the issue by rescheduling the jack detect handler in case the card is not ready. In case card never shows up, the shutdown/remove/suspend calls can now cancel the detect task. Signed-off-by: Kai Vehmanen <kai.vehmanen(a)linux.intel.com> Reviewed-by: Bard Liao <yung-chuan.liao(a)linux.intel.com> Reviewed-by: Ranjani Sridharan <ranjani.sridharan(a)linux.intel.com> Reviewed-by: Pierre-Louis Bossart <pierre-louis.bossart(a)linux.intel.com> Reviewed-by: Péter Ujfalusi <peter.ujfalusi(a)linux.intel.com> Reviewed-by: Shuming Fan <shumingf(a)realtek.com> Link: https://lore.kernel.org/r/20220207153000.3452802-2-kai.vehmanen@linux.intel… Signed-off-by: Mark Brown <broonie(a)kernel.org> Signed-off-by: Sasha Levin <sashal(a)kernel.org> --- sound/soc/codecs/rt5668.c | 12 +++++++----- 1 file changed, 7 insertions(+), 5 deletions(-) diff --git a/sound/soc/codecs/rt5668.c b/sound/soc/codecs/rt5668.c index 6ab1a8bc37352..1186ceb5a978e 100644 --- a/sound/soc/codecs/rt5668.c +++ b/sound/soc/codecs/rt5668.c @@ -1022,11 +1022,13 @@ static void rt5668_jack_detect_handler(struct work_struct *work) container_of(work, struct rt5668_priv, jack_detect_work.work); int val, btn_type; - while (!rt5668->component) - usleep_range(10000, 15000); - - while (!rt5668->component->card->instantiated) - usleep_range(10000, 15000); + if (!rt5668->component || !rt5668->component->card || + !rt5668->component->card->instantiated) { + /* card not yet ready, try later */ + mod_delayed_work(system_power_efficient_wq, + &rt5668->jack_detect_work, msecs_to_jiffies(15)); + return; + } mutex_lock(&rt5668->calibrate_mutex); -- 2.34.1

1 0

[PATCH AUTOSEL 5.16 07/30] ASoC: rt5682: do not block workqueue if card is unbound
by Sasha Levin 23 Feb '22

23 Feb '22

From: Kai Vehmanen <kai.vehmanen(a)linux.intel.com> [ Upstream commit 4c33de0673ced9c7c37b3bbd9bfe0fda72340b2a ] The current rt5682_jack_detect_handler() assumes the component and card will always show up and implements an infinite usleep loop waiting for them to show up. This does not hold true if a codec interrupt (or other event) occurs when the card is unbound. The codec driver's remove or shutdown functions cannot cancel the workqueue due to the wait loop. As a result, code can either end up blocking the workqueue, or hit a kernel oops when the card is freed. Fix the issue by rescheduling the jack detect handler in case the card is not ready. In case card never shows up, the shutdown/remove/suspend calls can now cancel the detect task. Signed-off-by: Kai Vehmanen <kai.vehmanen(a)linux.intel.com> Reviewed-by: Bard Liao <yung-chuan.liao(a)linux.intel.com> Reviewed-by: Ranjani Sridharan <ranjani.sridharan(a)linux.intel.com> Reviewed-by: Pierre-Louis Bossart <pierre-louis.bossart(a)linux.intel.com> Reviewed-by: Péter Ujfalusi <peter.ujfalusi(a)linux.intel.com> Reviewed-by: Shuming Fan <shumingf(a)realtek.com> Link: https://lore.kernel.org/r/20220207153000.3452802-3-kai.vehmanen@linux.intel… Signed-off-by: Mark Brown <broonie(a)kernel.org> Signed-off-by: Sasha Levin <sashal(a)kernel.org> --- sound/soc/codecs/rt5682.c | 12 +++++++----- 1 file changed, 7 insertions(+), 5 deletions(-) diff --git a/sound/soc/codecs/rt5682.c b/sound/soc/codecs/rt5682.c index 7e6e2f26accd0..e3643ae6de66d 100644 --- a/sound/soc/codecs/rt5682.c +++ b/sound/soc/codecs/rt5682.c @@ -1092,11 +1092,13 @@ void rt5682_jack_detect_handler(struct work_struct *work) struct snd_soc_dapm_context *dapm; int val, btn_type; - while (!rt5682->component) - usleep_range(10000, 15000); - - while (!rt5682->component->card->instantiated) - usleep_range(10000, 15000); + if (!rt5682->component || !rt5682->component->card || + !rt5682->component->card->instantiated) { + /* card not yet ready, try later */ + mod_delayed_work(system_power_efficient_wq, + &rt5682->jack_detect_work, msecs_to_jiffies(15)); + return; + } dapm = snd_soc_component_get_dapm(rt5682->component); -- 2.34.1

1 0

[PATCH AUTOSEL 5.16 06/30] ASoC: rt5668: do not block workqueue if card is unbound
by Sasha Levin 23 Feb '22

23 Feb '22

From: Kai Vehmanen <kai.vehmanen(a)linux.intel.com> [ Upstream commit a6d78661dc903d90a327892bbc34268f3a5f4b9c ] The current rt5668_jack_detect_handler() assumes the component and card will always show up and implements an infinite usleep loop waiting for them to show up. This does not hold true if a codec interrupt (or other event) occurs when the card is unbound. The codec driver's remove or shutdown functions cannot cancel the workqueue due to the wait loop. As a result, code can either end up blocking the workqueue, or hit a kernel oops when the card is freed. Fix the issue by rescheduling the jack detect handler in case the card is not ready. In case card never shows up, the shutdown/remove/suspend calls can now cancel the detect task. Signed-off-by: Kai Vehmanen <kai.vehmanen(a)linux.intel.com> Reviewed-by: Bard Liao <yung-chuan.liao(a)linux.intel.com> Reviewed-by: Ranjani Sridharan <ranjani.sridharan(a)linux.intel.com> Reviewed-by: Pierre-Louis Bossart <pierre-louis.bossart(a)linux.intel.com> Reviewed-by: Péter Ujfalusi <peter.ujfalusi(a)linux.intel.com> Reviewed-by: Shuming Fan <shumingf(a)realtek.com> Link: https://lore.kernel.org/r/20220207153000.3452802-2-kai.vehmanen@linux.intel… Signed-off-by: Mark Brown <broonie(a)kernel.org> Signed-off-by: Sasha Levin <sashal(a)kernel.org> --- sound/soc/codecs/rt5668.c | 12 +++++++----- 1 file changed, 7 insertions(+), 5 deletions(-) diff --git a/sound/soc/codecs/rt5668.c b/sound/soc/codecs/rt5668.c index fb09715bf9328..5b12cbf2ba215 100644 --- a/sound/soc/codecs/rt5668.c +++ b/sound/soc/codecs/rt5668.c @@ -1022,11 +1022,13 @@ static void rt5668_jack_detect_handler(struct work_struct *work) container_of(work, struct rt5668_priv, jack_detect_work.work); int val, btn_type; - while (!rt5668->component) - usleep_range(10000, 15000); - - while (!rt5668->component->card->instantiated) - usleep_range(10000, 15000); + if (!rt5668->component || !rt5668->component->card || + !rt5668->component->card->instantiated) { + /* card not yet ready, try later */ + mod_delayed_work(system_power_efficient_wq, + &rt5668->jack_detect_work, msecs_to_jiffies(15)); + return; + } mutex_lock(&rt5668->calibrate_mutex); -- 2.34.1

1 0

[PATCH AUTOSEL 5.16 05/30] ASoC: rt5682s: do not block workqueue if card is unbound
by Sasha Levin 23 Feb '22

23 Feb '22

From: Kai Vehmanen <kai.vehmanen(a)linux.intel.com> [ Upstream commit d7b530fdc45e75a54914a194c4becd9672a4e24f ] The current rt5682s_jack_detect_handler() assumes the component and card will always show up and implements an infinite usleep loop waiting for them to show up. This does not hold true if a codec interrupt (or other event) occurs when the card is unbound. The codec driver's remove or shutdown functions cannot cancel the workqueue due to the wait loop. As a result, code can either end up blocking the workqueue, or hit a kernel oops when the card is freed. Fix the issue by rescheduling the jack detect handler in case the card is not ready. In case card never shows up, the shutdown/remove/suspend calls can now cancel the detect task. Signed-off-by: Kai Vehmanen <kai.vehmanen(a)linux.intel.com> Reviewed-by: Bard Liao <yung-chuan.liao(a)linux.intel.com> Reviewed-by: Ranjani Sridharan <ranjani.sridharan(a)linux.intel.com> Reviewed-by: Pierre-Louis Bossart <pierre-louis.bossart(a)linux.intel.com> Reviewed-by: Péter Ujfalusi <peter.ujfalusi(a)linux.intel.com> Reviewed-by: Shuming Fan <shumingf(a)realtek.com> Link: https://lore.kernel.org/r/20220207153000.3452802-1-kai.vehmanen@linux.intel… Signed-off-by: Mark Brown <broonie(a)kernel.org> Signed-off-by: Sasha Levin <sashal(a)kernel.org> --- sound/soc/codecs/rt5682s.c | 12 +++++++----- 1 file changed, 7 insertions(+), 5 deletions(-) diff --git a/sound/soc/codecs/rt5682s.c b/sound/soc/codecs/rt5682s.c index d49a4f68566d2..d79b548d23fa4 100644 --- a/sound/soc/codecs/rt5682s.c +++ b/sound/soc/codecs/rt5682s.c @@ -824,11 +824,13 @@ static void rt5682s_jack_detect_handler(struct work_struct *work) container_of(work, struct rt5682s_priv, jack_detect_work.work); int val, btn_type; - while (!rt5682s->component) - usleep_range(10000, 15000); - - while (!rt5682s->component->card->instantiated) - usleep_range(10000, 15000); + if (!rt5682s->component || !rt5682s->component->card || + !rt5682s->component->card->instantiated) { + /* card not yet ready, try later */ + mod_delayed_work(system_power_efficient_wq, + &rt5682s->jack_detect_work, msecs_to_jiffies(15)); + return; + } mutex_lock(&rt5682s->jdet_mutex); mutex_lock(&rt5682s->calibrate_mutex); -- 2.34.1

1 0

[PATCH 00/10] ASoC: codec: add pm runtime support for Qualcomm codecs
by Srinivas Kandagatla 22 Feb '22

22 Feb '22

This patchset adds support for runtime pm on tx/rx/wsa/wcd lpass macro, wsa881x and wcd938x codecs that are wired up on SoundWire bus. During pm testing it was also found that soundwire clks enabled by lpass macros are not enabling all the required clocks correctly, so last 3 patches corrects them. Tested this on SM8250 MTP along SoundWire In band Headset Button wakeup interrupts. Srinivas Kandagatla (10): ASoC: codecs: va-macro: add runtime pm support ASoC: codecs: wsa-macro: add runtime pm support ASoC: codecs: rx-macro: add runtime pm support ASoC: codecs: tx-macro: add runtime pm support ASoC: codecs: wsa881x: add runtime pm support ASoC: codecs: wcd938x: add simple clk stop support ASoC: codecs: wcd-mbhc: add runtime pm support ASoC: codecs: wsa-macro: setup soundwire clks correctly ASoC: codecs: tx-macro: setup soundwire clks correctly ASoC: codecs: rx-macro: setup soundwire clks correctly sound/soc/codecs/lpass-rx-macro.c | 49 +++++++++++++++++++++++++-- sound/soc/codecs/lpass-tx-macro.c | 45 ++++++++++++++++++++++++- sound/soc/codecs/lpass-va-macro.c | 36 ++++++++++++++++++++ sound/soc/codecs/lpass-wsa-macro.c | 43 +++++++++++++++++++++++- sound/soc/codecs/wcd-mbhc-v2.c | 26 ++++++++++++++ sound/soc/codecs/wcd938x-sdw.c | 1 + sound/soc/codecs/wsa881x.c | 54 ++++++++++++++++++++++++++++++ 7 files changed, 250 insertions(+), 4 deletions(-) -- 2.21.0

3 17

[PATCH v8 0/7] Add pin control support for lpass sc7280
by Srinivasa Rao Mandadapu 22 Feb '22

22 Feb '22

This patch series is to split lpass variant common pin control functions and SoC specific functions and to add lpass sc7280 pincontrol support. It also Adds dt-bindings for lpass sc7280 lpass lpi pincontrol. Changes Since V7: -- Update optional clock voting with conditional check. -- Add const to lpi_pinctrl_variant_data structure. -- Update required headers and remove redundant. -- Change EXPORT_SYMBOL to EXPORT_SYMBOL_GPL -- Fix typo errors. Changes Since V6: -- Update conditional clock voting to optional clock voting. -- Update Kconfig depends on field with select. -- Fix typo errors. Changes Since V5: -- Create new patch by updating macro name to lpi specific. -- Create new patch by updating lpi pin group structure with core group_desc structure. -- Fix typo errors. -- Sort macros in the make file and configuration file. Changes Since V4: -- Update commit message and description of the chip specific extraction patch. -- Sort macros in kconfig and makefile. -- Update optional clock voting to conditional clock voting. -- Fix typo errors. -- Move to quicinc domain email id's. Changes Since V3: -- Update separate Kconfig fields for sm8250 and sc7280. -- Update module license and description. -- Move static variables to corresponding .c files from header file. Changes Since V2: -- Add new dt-bindings for sc7280 lpi driver. -- Make clock voting change as separate patch. -- Split existing pincontrol driver and make common functions as part of separate file. -- Rename lpass pincontrol lpi dt-bindings to sm8250 specific dt-bindings Changes Since V1: -- Make lpi pinctrl variant data structure as constant -- Add appropriate commit message -- Change signedoff by sequence. Srinivasa Rao Mandadapu (7): dt-bindings: pinctrl: qcom: Update lpass lpi file name to SoC specific dt-bindings: pinctrl: qcom: Add sc7280 lpass lpi pinctrl bindings pinctrl: qcom: Update macro name to LPI specific pinctrl: qcom: Update lpi pin group structure pinctrl: qcom: Extract chip specific LPASS LPI code pinctrl: qcom: Add SC7280 lpass pin configuration pinctrl: qcom: Update clock voting as optional Tested this on SM8250 MTP with WSA and WCD codecs. Tested-by: Srinivas Kandagatla <srinivas.kandagatla(a)linaro.org> .../bindings/pinctrl/qcom,lpass-lpi-pinctrl.yaml | 133 ----------- .../pinctrl/qcom,sc7280-lpass-lpi-pinctrl.yaml | 115 ++++++++++ .../pinctrl/qcom,sm8250-lpass-lpi-pinctrl.yaml | 133 +++++++++++ drivers/pinctrl/qcom/Kconfig | 16 ++ drivers/pinctrl/qcom/Makefile | 2 + drivers/pinctrl/qcom/pinctrl-lpass-lpi.c | 255 ++------------------- drivers/pinctrl/qcom/pinctrl-lpass-lpi.h | 86 +++++++ drivers/pinctrl/qcom/pinctrl-sc7280-lpass-lpi.c | 169 ++++++++++++++ drivers/pinctrl/qcom/pinctrl-sm8250-lpass-lpi.c | 166 ++++++++++++++ 9 files changed, 705 insertions(+), 370 deletions(-) delete mode 100644 Documentation/devicetree/bindings/pinctrl/qcom,lpass-lpi-pinctrl.yaml create mode 100644 Documentation/devicetree/bindings/pinctrl/qcom,sc7280-lpass-lpi-pinctrl.yaml create mode 100644 Documentation/devicetree/bindings/pinctrl/qcom,sm8250-lpass-lpi-pinctrl.yaml create mode 100644 drivers/pinctrl/qcom/pinctrl-lpass-lpi.h create mode 100644 drivers/pinctrl/qcom/pinctrl-sc7280-lpass-lpi.c create mode 100644 drivers/pinctrl/qcom/pinctrl-sm8250-lpass-lpi.c -- 2.7.4

2 8

Asking for information about ALSA's kselftest
by Giacomo Guiduzzi 22 Feb '22

22 Feb '22

[RESENDING BECAUSE BOUNCED BY SPAMASSASSIN] Greetings Mr. Brown and users of the alsa-devel mailing list, I’m a student at the University of Modena and Reggio Emilia. I am working on the Kernel Hacking course’s exam, a course kept by Paolo Valente which introduces students to the world of the Linux kernel and its magics. I was taking a look at the kernel’s kselftests from Linus Torvald’s repository to find some test to fix for the exam and I stumbled upon the ALSA’s selftest, which I’ve passed the last few days on, debugging and exploring. During the execution the test goes as it should except for two or three tests. Until now I’ve tried understanding test 71, which sometimes succeeds and sometimes fails, and test 72. I have no previous knowledge of how ALSA works and I’m completely new to the kselftests and the kernel in general. Test 71 gives the following output: # Surround Playback Volume.0 expected 0 but read 1, is_volatile 0 # Surround Playback Volume.1 expected 1 but read 0, is_volatile 0 not ok 71 write_default.0.22 I wanted to ask you if it is normal or not that this test fails on my hardware, and if not, what I could do to debug the problem and give you useful information to help understand what the issue is. I am currently testing on an Ubuntu Desktop VirtualBox VM on kernel 5.17.0-rc3. The audio hardware (as the output of the lshw command, multimedia section) is: description: Multimedia audio controller product: 82801AA AC'97 Audio Controller vendor: Intel Corporation physical id: 5 bus info: pci@0000:00:05.0 version: 01 width: 32 bits clock: 33MHz capabilities: bus_master configuration: driver=snd_intel8x0 latency=64 resources: irq:21 ioport:d100(size=256) ioport:d200(size=64) From what I have seen until now it doesn’t look like a bug from the test nor an issue from the kernel, so I’m not really sure where to look at: I have compiled alsalib’s source code from the Git repo with debug symbols and I’ve followed the execution of the test until I reached a call to the ioctl syscall, which doesn’t return an error code so it doesn’t look like the problem relies in the kernel; I’m no expert though, I’ll let you confirm this or not. I am at your complete disposal to give you any further useful information as I can’t do much more from my side alone. Thank you in advance, looking forward to your kind response. Best Regards, Giacomo Guiduzzi

2 1

[PATCH] ASoC: samsung: i2s: check the return value of kstrdup()
by xkernel.wang＠foxmail.com 22 Feb '22

22 Feb '22

From: Xiaoke Wang <xkernel.wang(a)foxmail.com> kstrdup() is a memory allocation function which can return NULL when some internal memory errors happen. It is better to check the return value of it to catch the error in time. Signed-off-by: Xiaoke Wang <xkernel.wang(a)foxmail.com> --- sound/soc/samsung/i2s.c | 4 ++++ 1 file changed, 4 insertions(+) diff --git a/sound/soc/samsung/i2s.c b/sound/soc/samsung/i2s.c index 309badc..70c8271 100644 --- a/sound/soc/samsung/i2s.c +++ b/sound/soc/samsung/i2s.c @@ -1349,6 +1349,10 @@ static int i2s_create_secondary_device(struct samsung_i2s_priv *priv) return -ENOMEM; pdev_sec->driver_override = kstrdup("samsung-i2s", GFP_KERNEL); + if (!pdev_sec->driver_override) { + platform_device_put(pdev_sec); + return -ENOMEM; + } ret = platform_device_add(pdev_sec); if (ret < 0) { --

2 1