[alsa-devel] [PATCH 1/1] usb: gadget: f_uac2: Fix pcm sample size selection
The pcm playback and capture sample size format was fixed SNDRV_PCM_FMTBIT_S16_LE.
This patch respects also 16, 24 and 32 bit p_ssize and c_ssize values.
Signed-off-by: Sebastian Reimers sebastian.reimers@gmail.com --- drivers/usb/gadget/f_uac2.c | 24 ++++++++++++++++++++++-- 1 file changed, 22 insertions(+), 2 deletions(-)
diff --git a/drivers/usb/gadget/f_uac2.c b/drivers/usb/gadget/f_uac2.c index 6261db4a..3ed89ec 100644 --- a/drivers/usb/gadget/f_uac2.c +++ b/drivers/usb/gadget/f_uac2.c @@ -348,14 +348,34 @@ static int uac2_pcm_open(struct snd_pcm_substream *substream) if (substream->stream == SNDRV_PCM_STREAM_PLAYBACK) { spin_lock_init(&uac2->p_prm.lock); runtime->hw.rate_min = p_srate; - runtime->hw.formats = SNDRV_PCM_FMTBIT_S16_LE; /* ! p_ssize ! */ + switch (p_ssize) { + case 3: + runtime->hw.formats = SNDRV_PCM_FMTBIT_S24_3LE; + break; + case 4: + runtime->hw.formats = SNDRV_PCM_FMTBIT_S32_LE; + break; + default: + runtime->hw.formats = SNDRV_PCM_FMTBIT_S16_LE; + break; + } runtime->hw.channels_min = num_channels(p_chmask); runtime->hw.period_bytes_min = 2 * uac2->p_prm.max_psize / runtime->hw.periods_min; } else { spin_lock_init(&uac2->c_prm.lock); runtime->hw.rate_min = c_srate; - runtime->hw.formats = SNDRV_PCM_FMTBIT_S16_LE; /* ! c_ssize ! */ + switch (c_ssize) { + case 3: + runtime->hw.formats = SNDRV_PCM_FMTBIT_S24_3LE; + break; + case 4: + runtime->hw.formats = SNDRV_PCM_FMTBIT_S32_LE; + break; + default: + runtime->hw.formats = SNDRV_PCM_FMTBIT_S16_LE; + break; + } runtime->hw.channels_min = num_channels(c_chmask); runtime->hw.period_bytes_min = 2 * uac2->c_prm.max_psize / runtime->hw.periods_min;
At Thu, 3 Jul 2014 20:15:28 +0200, Sebastian Reimers wrote:
The pcm playback and capture sample size format was fixed SNDRV_PCM_FMTBIT_S16_LE.
This patch respects also 16, 24 and 32 bit p_ssize and c_ssize values.
Signed-off-by: Sebastian Reimers sebastian.reimers@gmail.com
In general, this driver should check more on p_ssize. This can be any integer value passed as a module parameter. It should be cut to a sane value or give an error in the probe function when an invalid value is given. Ditto for p_srate and other parameters.
Also, I don't understand the logic of the code runtime->hw.period_bytes_min = 2 * uac2->p_prm.max_psize / runtime->hw.periods_min;
But this is mostly irrelevant with this patch, so for this patch: Reviewed-by: Takashi Iwai tiwai@suse.de
thanks,
Takashi
drivers/usb/gadget/f_uac2.c | 24 ++++++++++++++++++++++-- 1 file changed, 22 insertions(+), 2 deletions(-)
diff --git a/drivers/usb/gadget/f_uac2.c b/drivers/usb/gadget/f_uac2.c index 6261db4a..3ed89ec 100644 --- a/drivers/usb/gadget/f_uac2.c +++ b/drivers/usb/gadget/f_uac2.c @@ -348,14 +348,34 @@ static int uac2_pcm_open(struct snd_pcm_substream *substream) if (substream->stream == SNDRV_PCM_STREAM_PLAYBACK) { spin_lock_init(&uac2->p_prm.lock); runtime->hw.rate_min = p_srate;
runtime->hw.formats = SNDRV_PCM_FMTBIT_S16_LE; /* ! p_ssize ! */
switch (p_ssize) {
case 3:
runtime->hw.formats = SNDRV_PCM_FMTBIT_S24_3LE;
break;
case 4:
runtime->hw.formats = SNDRV_PCM_FMTBIT_S32_LE;
break;
default:
runtime->hw.formats = SNDRV_PCM_FMTBIT_S16_LE;
break;
runtime->hw.channels_min = num_channels(p_chmask); runtime->hw.period_bytes_min = 2 * uac2->p_prm.max_psize / runtime->hw.periods_min; } else { spin_lock_init(&uac2->c_prm.lock); runtime->hw.rate_min = c_srate;}
runtime->hw.formats = SNDRV_PCM_FMTBIT_S16_LE; /* ! c_ssize ! */
switch (c_ssize) {
case 3:
runtime->hw.formats = SNDRV_PCM_FMTBIT_S24_3LE;
break;
case 4:
runtime->hw.formats = SNDRV_PCM_FMTBIT_S32_LE;
break;
default:
runtime->hw.formats = SNDRV_PCM_FMTBIT_S16_LE;
break;
runtime->hw.channels_min = num_channels(c_chmask); runtime->hw.period_bytes_min = 2 * uac2->c_prm.max_psize / runtime->hw.periods_min;}
-- 2.0.1
On 07/04/2014 07:43 AM, Takashi Iwai wrote:
At Thu, 3 Jul 2014 20:15:28 +0200, Sebastian Reimers wrote:
The pcm playback and capture sample size format was fixed SNDRV_PCM_FMTBIT_S16_LE.
This patch respects also 16, 24 and 32 bit p_ssize and c_ssize values.
Signed-off-by: Sebastian Reimers sebastian.reimers@gmail.com
In general, this driver should check more on p_ssize. This can be any integer value passed as a module parameter. It should be cut to a sane value or give an error in the probe function when an invalid value is given. Ditto for p_srate and other parameters.
Also, I don't understand the logic of the code runtime->hw.period_bytes_min = 2 * uac2->p_prm.max_psize / runtime->hw.periods_min;
But this is mostly irrelevant with this patch, so for this patch: Reviewed-by: Takashi Iwai tiwai@suse.de
Felipe, are you fine with this one?
Thanks, Daniel
drivers/usb/gadget/f_uac2.c | 24 ++++++++++++++++++++++-- 1 file changed, 22 insertions(+), 2 deletions(-)
diff --git a/drivers/usb/gadget/f_uac2.c b/drivers/usb/gadget/f_uac2.c index 6261db4a..3ed89ec 100644 --- a/drivers/usb/gadget/f_uac2.c +++ b/drivers/usb/gadget/f_uac2.c @@ -348,14 +348,34 @@ static int uac2_pcm_open(struct snd_pcm_substream *substream) if (substream->stream == SNDRV_PCM_STREAM_PLAYBACK) { spin_lock_init(&uac2->p_prm.lock); runtime->hw.rate_min = p_srate;
runtime->hw.formats = SNDRV_PCM_FMTBIT_S16_LE; /* ! p_ssize ! */
switch (p_ssize) {
case 3:
runtime->hw.formats = SNDRV_PCM_FMTBIT_S24_3LE;
break;
case 4:
runtime->hw.formats = SNDRV_PCM_FMTBIT_S32_LE;
break;
default:
runtime->hw.formats = SNDRV_PCM_FMTBIT_S16_LE;
break;
runtime->hw.channels_min = num_channels(p_chmask); runtime->hw.period_bytes_min = 2 * uac2->p_prm.max_psize / runtime->hw.periods_min; } else { spin_lock_init(&uac2->c_prm.lock); runtime->hw.rate_min = c_srate;}
runtime->hw.formats = SNDRV_PCM_FMTBIT_S16_LE; /* ! c_ssize ! */
switch (c_ssize) {
case 3:
runtime->hw.formats = SNDRV_PCM_FMTBIT_S24_3LE;
break;
case 4:
runtime->hw.formats = SNDRV_PCM_FMTBIT_S32_LE;
break;
default:
runtime->hw.formats = SNDRV_PCM_FMTBIT_S16_LE;
break;
runtime->hw.channels_min = num_channels(c_chmask); runtime->hw.period_bytes_min = 2 * uac2->c_prm.max_psize / runtime->hw.periods_min;}
-- 2.0.1
Alsa-devel mailing list Alsa-devel@alsa-project.org http://mailman.alsa-project.org/mailman/listinfo/alsa-devel
On Tue, Jul 22, 2014 at 07:09:51PM +0200, Daniel Mack wrote:
On 07/04/2014 07:43 AM, Takashi Iwai wrote:
At Thu, 3 Jul 2014 20:15:28 +0200, Sebastian Reimers wrote:
The pcm playback and capture sample size format was fixed SNDRV_PCM_FMTBIT_S16_LE.
This patch respects also 16, 24 and 32 bit p_ssize and c_ssize values.
Signed-off-by: Sebastian Reimers sebastian.reimers@gmail.com
In general, this driver should check more on p_ssize. This can be any integer value passed as a module parameter. It should be cut to a sane value or give an error in the probe function when an invalid value is given. Ditto for p_srate and other parameters.
Also, I don't understand the logic of the code runtime->hw.period_bytes_min = 2 * uac2->p_prm.max_psize / runtime->hw.periods_min;
But this is mostly irrelevant with this patch, so for this patch: Reviewed-by: Takashi Iwai tiwai@suse.de
Felipe, are you fine with this one?
seems like I missed this one, do you mind if we take it during v3.17-rc and add Cc: stable ?
-----BEGIN PGP SIGNED MESSAGE----- Hash: SHA1
On 07/22/2014 07:44 PM, Felipe Balbi wrote:
On Tue, Jul 22, 2014 at 07:09:51PM +0200, Daniel Mack wrote:
On 07/04/2014 07:43 AM, Takashi Iwai wrote:
At Thu, 3 Jul 2014 20:15:28 +0200, Sebastian Reimers wrote:
The pcm playback and capture sample size format was fixed SNDRV_PCM_FMTBIT_S16_LE.
This patch respects also 16, 24 and 32 bit p_ssize and c_ssize values.
Signed-off-by: Sebastian Reimers sebastian.reimers@gmail.com
In general, this driver should check more on p_ssize. This can be any integer value passed as a module parameter. It should be cut to a sane value or give an error in the probe function when an invalid value is given. Ditto for p_srate and other parameters.
Also, I don't understand the logic of the code runtime->hw.period_bytes_min = 2 * uac2->p_prm.max_psize / runtime->hw.periods_min;
But this is mostly irrelevant with this patch, so for this patch: Reviewed-by: Takashi Iwai tiwai@suse.de
Felipe, are you fine with this one?
seems like I missed this one, do you mind if we take it during v3.17-rc and add Cc: stable ?
I guess that was what Sebastian intended :)
Thanks for taking care of it!
Daniel
participants (4)
-
Daniel Mack
-
Felipe Balbi
-
Sebastian Reimers
-
Takashi Iwai