[alsa-devel] regression introduced by commit 70d435ba1cd6 ("ASoC: imx-pcm-dma: simplify pcm_config")
Hi,
the commit 70d435ba1cd6 ("ASoC: imx-pcm-dma: simplify pcm_config") introduced a regression on the Ka-Ro TX series modules (and possibly other hardware as well).
The problem is, that this commit restricts the supported sound formats to S24_LE only rather than S16_LE, S24_LE, S20_3LE without the patch. Therefore aplay fails with the following error messages, when trying to play a sound file with S16_LE format: |ALSA lib alsa-lib-1.1.2/src/pcm/pcm_params.c:2162:(snd1_pcm_hw_refine_slave) Slave PCM not usable |aplay: set_params:1204: Broken configuration for this PCM: no configurations available
This is with the simple-sound-card driver as well as the imx-sgtl5000 driver.
Can someone tell me on what hardware and with what software has this patch been tested?
Lothar Waßmann
Hi Lothar,
Am Mittwoch, den 04.10.2017, 10:36 +0200 schrieb Lothar Waßmann:
Hi,
the commit 70d435ba1cd6 ("ASoC: imx-pcm-dma: simplify pcm_config") introduced a regression on the Ka-Ro TX series modules (and possibly other hardware as well).
The problem is, that this commit restricts the supported sound formats to S24_LE only rather than S16_LE, S24_LE, S20_3LE without the patch. Therefore aplay fails with the following error messages, when trying to play a sound file with S16_LE format:
ALSA lib alsa-lib- 1.1.2/src/pcm/pcm_params.c:2162:(snd1_pcm_hw_refine_slave) Slave PCM not usable aplay: set_params:1204: Broken configuration for this PCM: no configurations available
This is with the simple-sound-card driver as well as the imx-sgtl5000 driver.
Can someone tell me on what hardware and with what software has this patch been tested?
My primary test hardware was the zii-rdu2 board, the DTS for this board can be found in recent mainline kernels.
Userspace is based on pulseaudio, which may mask the regression by resampling to the supported format. Sorry about that.
From my understanding of the code, it should be possible to fix the
regression without reverting this patch by setting SND_DMAENGINE_PCM_DAI_FLAG_PACK in the flags field of the DMA params in fsl_ssi.
Regards, Lucas
Hi,
On Tue, 10 Oct 2017 16:49:04 +0200 Lucas Stach wrote:
Hi Lothar,
Am Mittwoch, den 04.10.2017, 10:36 +0200 schrieb Lothar Waßmann:
Hi,
the commit 70d435ba1cd6 ("ASoC: imx-pcm-dma: simplify pcm_config") introduced a regression on the Ka-Ro TX series modules (and possibly other hardware as well).
The problem is, that this commit restricts the supported sound formats to S24_LE only rather than S16_LE, S24_LE, S20_3LE without the patch. Therefore aplay fails with the following error messages, when trying to play a sound file with S16_LE format:
ALSA lib alsa-lib- 1.1.2/src/pcm/pcm_params.c:2162:(snd1_pcm_hw_refine_slave) Slave PCM not usable aplay: set_params:1204: Broken configuration for this PCM: no configurations available
This is with the simple-sound-card driver as well as the imx-sgtl5000 driver.
Can someone tell me on what hardware and with what software has this patch been tested?
My primary test hardware was the zii-rdu2 board, the DTS for this board can be found in recent mainline kernels.
Userspace is based on pulseaudio, which may mask the regression by resampling to the supported format. Sorry about that.
From my understanding of the code, it should be possible to fix the regression without reverting this patch by setting SND_DMAENGINE_PCM_DAI_FLAG_PACK in the flags field of the DMA params in fsl_ssi.
I tried that, but then I get DMA timeouts.
I circumvented the problem by changing my /etc/asound.conf to use the only supported sample format. So, if nobody else has a problem, there is no need to revert the patch or apply additional modifications to the driver.
Lothar Waßmann
On Tue, Oct 10, 2017 at 05:18:46PM +0200, Lothar Waßmann wrote:
I circumvented the problem by changing my /etc/asound.conf to use the only supported sample format. So, if nobody else has a problem, there is no need to revert the patch or apply additional modifications to the driver.
I've not seen any other reports.
participants (3)
-
Lothar Waßmann
-
Lucas Stach
-
Mark Brown