[alsa-devel] [PATCH 1/3] ASoC: Tegra: tegra_pcm_deallocate_dma_buffer: Don't OOPS
Not all PCM devices have all sub-streams. Specifically, the SPDIF driver only supports playback and hence has no capture substream. Check whether a substream exists before dereferencing it, when de-allocating DMA buffers in tegra_pcm_deallocate_dma_buffer.
Signed-off-by: Stephen Warren swarren@nvidia.com --- sound/soc/tegra/tegra_pcm.c | 9 +++++++-- 1 files changed, 7 insertions(+), 2 deletions(-)
diff --git a/sound/soc/tegra/tegra_pcm.c b/sound/soc/tegra/tegra_pcm.c index ff86e5e..c7cfd96 100644 --- a/sound/soc/tegra/tegra_pcm.c +++ b/sound/soc/tegra/tegra_pcm.c @@ -309,9 +309,14 @@ static int tegra_pcm_preallocate_dma_buffer(struct snd_pcm *pcm, int stream)
static void tegra_pcm_deallocate_dma_buffer(struct snd_pcm *pcm, int stream) { - struct snd_pcm_substream *substream = pcm->streams[stream].substream; - struct snd_dma_buffer *buf = &substream->dma_buffer; + struct snd_pcm_substream *substream; + struct snd_dma_buffer *buf; + + substream = pcm->streams[stream].substream; + if (!substream) + return;
+ buf = &substream->dma_buffer; if (!buf->area) return;
Two issues were preventing module snd-soc-tegra-wm8903.ko from being removed and re-inserted:
a) The speaker-enable GPIO is hosted by the WM8903 chip. This GPIO must be freed before snd_soc_unregister_card() is called, because that triggers wm8903.c:wm8903_remove(), which calls gpiochip_remove(), which then fails if any of the GPIOs are in use. To solve this, free all GPIOs first, so the code doesn't care where they come from.
b) We need to call snd_soc_jack_free_gpios() to match the call to snd_soc_jack_add_gpios() during initialization. Without this, the call to snd_soc_jack_add_gpios() fails during any subsequent modprobe and initialization, since the GPIO and IRQ are already registered. In turn, this causes the headphone state not to be monitored, so the headphone is assumed not to be plugged in, and the audio path to it is never enabled.
Signed-off-by: Stephen Warren swarren@nvidia.com --- sound/soc/tegra/tegra_wm8903.c | 15 +++++++++++---- 1 files changed, 11 insertions(+), 4 deletions(-)
diff --git a/sound/soc/tegra/tegra_wm8903.c b/sound/soc/tegra/tegra_wm8903.c index a42e9ac..661373c 100644 --- a/sound/soc/tegra/tegra_wm8903.c +++ b/sound/soc/tegra/tegra_wm8903.c @@ -56,6 +56,7 @@ #define GPIO_HP_MUTE BIT(1) #define GPIO_INT_MIC_EN BIT(2) #define GPIO_EXT_MIC_EN BIT(3) +#define GPIO_HP_DET BIT(4)
struct tegra_wm8903 { struct tegra_asoc_utils_data util_data; @@ -304,6 +305,7 @@ static int tegra_wm8903_init(struct snd_soc_pcm_runtime *rtd) snd_soc_jack_add_gpios(&tegra_wm8903_hp_jack, 1, &tegra_wm8903_hp_jack_gpio); + machine->gpio_requested |= GPIO_HP_DET; }
snd_soc_jack_new(codec, "Mic Jack", SND_JACK_MICROPHONE, @@ -429,10 +431,10 @@ static int __devexit tegra_wm8903_driver_remove(struct platform_device *pdev) struct tegra_wm8903 *machine = snd_soc_card_get_drvdata(card); struct tegra_wm8903_platform_data *pdata = machine->pdata;
- snd_soc_unregister_card(card); - - tegra_asoc_utils_fini(&machine->util_data); - + if (machine->gpio_requested & GPIO_HP_DET) + snd_soc_jack_free_gpios(&tegra_wm8903_hp_jack, + 1, + &tegra_wm8903_hp_jack_gpio); if (machine->gpio_requested & GPIO_EXT_MIC_EN) gpio_free(pdata->gpio_ext_mic_en); if (machine->gpio_requested & GPIO_INT_MIC_EN) @@ -441,6 +443,11 @@ static int __devexit tegra_wm8903_driver_remove(struct platform_device *pdev) gpio_free(pdata->gpio_hp_mute); if (machine->gpio_requested & GPIO_SPKR_EN) gpio_free(pdata->gpio_spkr_en); + machine->gpio_requested = 0; + + snd_soc_unregister_card(card); + + tegra_asoc_utils_fini(&machine->util_data);
kfree(machine);
On Thu, Aug 04, 2011 at 04:44:43PM -0600, Stephen Warren wrote:
Two issues were preventing module snd-soc-tegra-wm8903.ko from being removed and re-inserted:
I've applied all three but please be a bit clearer with your patch descriptions - this isn't a patch for WM8903 at all, it's a patch for the tegra-wm8903 machine driver.
Mark Brown wrote at Monday, August 08, 2011 6:44 PM:
On Thu, Aug 04, 2011 at 04:44:43PM -0600, Stephen Warren wrote:
Two issues were preventing module snd-soc-tegra-wm8903.ko from being removed and re-inserted:
I've applied all three but please be a bit clearer with your patch descriptions - this isn't a patch for WM8903 at all, it's a patch for the tegra-wm8903 machine driver.
OK, but what's the correct tag then? I explicitly wrote Tegra in the commit subject and not just WM8903...
Thanks.
On Tue, Aug 09, 2011 at 08:36:36AM -0700, Stephen Warren wrote:
Mark Brown wrote at Monday, August 08, 2011 6:44 PM:
I've applied all three but please be a bit clearer with your patch descriptions - this isn't a patch for WM8903 at all, it's a patch for the tegra-wm8903 machine driver.
OK, but what's the correct tag then? I explicitly wrote Tegra in the commit subject and not just WM8903...
Something like "Fix reload of Tegra WM8903 machine driver".
Without this, request_irq on subsequent device initialization fails, and the codec cannot be used.
Signed-off-by: Stephen Warren swarren@nvidia.com --- sound/soc/codecs/wm8903.c | 5 +++++ 1 files changed, 5 insertions(+), 0 deletions(-)
diff --git a/sound/soc/codecs/wm8903.c b/sound/soc/codecs/wm8903.c index 43e3d76..4ad8ebd 100644 --- a/sound/soc/codecs/wm8903.c +++ b/sound/soc/codecs/wm8903.c @@ -2046,8 +2046,13 @@ static int wm8903_probe(struct snd_soc_codec *codec) /* power down chip */ static int wm8903_remove(struct snd_soc_codec *codec) { + struct wm8903_priv *wm8903 = snd_soc_codec_get_drvdata(codec); + wm8903_free_gpio(codec); wm8903_set_bias_level(codec, SND_SOC_BIAS_OFF); + if (wm8903->irq) + free_irq(wm8903->irq, codec); + return 0; }
On Fri, 2011-08-05 at 00:44 +0200, Stephen Warren wrote:
Not all PCM devices have all sub-streams. Specifically, the SPDIF driver only supports playback and hence has no capture substream. Check whether a substream exists before dereferencing it, when de-allocating DMA buffers in tegra_pcm_deallocate_dma_buffer.
Signed-off-by: Stephen Warren swarren@nvidia.com
All
Acked-by: Liam Girdwood lrg@ti.com
participants (3)
-
Liam Girdwood -
Mark Brown -
Stephen Warren