On Sunday 28 June 2009 21:37:32 Jarkko Nikula wrote:
On Sun, 28 Jun 2009 00:21:05 +0200 Janusz Krzysztofik jkrzyszt@tis.icnet.pl wrote:
This patch tries to work around the problem of broken OMAP1510 PCM playback pointer calculation by replacing DMA function call that incorrectly tries to read the value form DMA hardware with a value computed locally from an already maintained variable omap_runtime_data.period_index.
Tested on OMAP5910 based Amstrad Delta (E3) using work in progress ASoC driver.
Based on linux-2.6-asoc.git v2.6.31-rc1.
Signed-off-by: Janusz Krzysztofik jkrzyszt@tis.icnet.pl
It seems that on OMAP1510, DMA Channel Progress Counter registers (DMA_CPC_CH[0-8]) always contain values derived from DMA channels destination port address, even if constant, and there are no DMA registers available that would follow DMA channels source port address. Because of this limitation, current implementation of omap_get_dma_src_pos() for OMAP1510 is broken and doesn't seem to be easy correctable.
Hi
Before going into workaround, did you try to change function omap_get_dma_src_pos to read CSSA_L instead of CPC in arch/arm/plat-omap/dma.c that I was speculating in mail below?
http://mailman.alsa-project.org/pipermail/alsa-devel/2009-June/018569.html
While CPC seems to be correct [1], you mentioned before that patch below broke the older ALSA driver so it's worth to find out if the problem can be corrected by simple one line fix (or two if omap_get_dma_dst_pos need to be fixed as well).
Jarkko,
AFAIK, both CSSA_L and CDSA_L DMA registers are static. Loaded by CPU with 16 LSB of initial source and destination port addresses respectively, they are never updated by the DMA engine itself. That's why they can't be used for transfer progress indication unless updated by CPU.
The old omap-alsa driver was just updating them, intentionally or not, by reprogramming and restarting DMA every PCM period. That's why calculating PCM pointers from CSSA_L/CDSA_L worked.
ASoC OMAP driver transfers whole PCM buffer with single DMA transfer, so it doesn't need to update DMA source/destination port address after initial playback/capture setup, even if restarting DMA, and actually never does this. Calculating PCM pointers from CSSA_L/CDSA_L registers without updating them every period would then be wrong.
For capture, reading CPC, that follows destination port address progress, just works fine (for both old and new driver). For playback, similar hardware functionality seems to be missing, so it has to be emulated in software if required.
Thanks, Janusz