RP1 Audio Out #6668
RP1 Audio Out #6668
Conversation
njhollinghurst
force-pushed
the
audio_out
branch
3 times, most recently
from
aa9d461 to
aec0c90
Compare
|
@pelwell could you cast an eye over the first two commits please (since renamed for rebase), to check they don't break I2S or anything else? Thanks. |
|
Do you need a card to test with? |
popcornmix
force-pushed
the
rpi-6.12.y
branch
from
8d163a9 to
50b8659
Compare
njhollinghurst
force-pushed
the
audio_out
branch
from
ba43740 to
94160db
Compare
|
Successfully tested JustBoom DAC Zero and IQaudIO Pi-Codec+ hats at various sample rates. I could even load and use the rp1aout driver at the same time, but (predictably) I2S stomped on rp1aout's clock when using the (crystal-less) JustBoom hat; they did seem able to coexist with the IQaudIO. Getting 48000Hz and 44100Hz-based rates at the same time is impractical with a single audio PLL. Running two outputs at (48000*32, 48000*3200) ought to be achievable, but the RP1 clk driver doesn't currently oblige. It's not a vital use case. |
njhollinghurst
force-pushed
the
audio_out
branch
from
25c7358 to
c3fe22d
Compare
sound/soc/raspberrypi/rp1_aout.c
Outdated
| if (ret) | ||
| return dev_err_probe(&pdev->dev, ret, "failed to register pcm\n"); | ||
|
|
||
| /* TODO: More PM stuff? */ |
There was a problem hiding this comment.
I don't think this will have an effect unless you define dev_pm_ops with suspend/resume/idle callbacks.
There was a problem hiding this comment.
Yes, that is what I was wondering. I suppose it ought to be done.
| cmd == SNDRV_PCM_TRIGGER_SUSPEND || | ||
| cmd == SNDRV_PCM_TRIGGER_PAUSE_PUSH) { | ||
| /* Push a zero sample (assuming DMA has stopped already?) */ | ||
| aout_reg_wr(aout, AUDIO_OUT_SAMPLE_FIFO, 0); |
There was a problem hiding this comment.
Underflow should be handled gracefully by the hardware (the last sample plays forever) - ASoC decouples DMA ops from driver visibility so you may introduce a glitch here if DMA is still running. I don't know if it synchronously drains the DMA buffers or flushes them.
There was a problem hiding this comment.
In tests it does seem to use the given value quite reliably -- I'll remove the question mark
It wasn't clear to me whether a step to zero would be more or less objectionable than a nonstandard DC value (perhaps followed by a step later). Maybe 'stop' and 'pause' cases should be treated differently?
There was a problem hiding this comment.
After a bit more debug: I haven't actually observed a PAUSE_PUSH or SUSPEND -- when I Ctrl-Z aplay, it does a STOP. I think I will keep the zero as it seems to work, and may be better than sticking at an arbitrary value.
njhollinghurst
force-pushed
the
audio_out
branch
2 times, most recently
from
b81599d to
0ab48b8
Compare
|
Rebooting with the carrier active doesn't result in a clean shutdown - it goes thump. snd_soc_dai_ops needs a shutdown callback... |
I think platform shutdown is the one I want: DAI shutdown happens after every stream, and we decided we wanted to leave the "bias" on in those cases. But I am currently battling a suspected race condition here... |
|
Created a tag https://github.com/njhollinghurst/linux/tree/audio_out_wip with current history. |
popcornmix
force-pushed
the
rpi-6.12.y
branch
from
fd2db7e to
031a382
Compare
Change the standard rate of PLL_AUDIO_SEC from 192MHz to 153.6MHz to suit audio out. Declare audio out hardware and give it a named pin control. Signed-off-by: Nick Hollinghurst <[email protected]>
…rror Connect PLL_AUDIO_SEC to CLK_AUDIO_OUT, which had been commented out to avoid interference with I2S: we expect them never to be enabled at the same time. Work around a rounding error that occurs when the desired rate is exactly the max but not exactly achievable by the PLL. Signed-off-by: Nick Hollinghurst <[email protected]>
njhollinghurst
force-pushed
the
audio_out
branch
2 times, most recently
from
b59584e to
37d7eca
Compare
Only 48000Hz stereo 16-bit output is currently supported. It requires some additional OF plumbing to connect it to a "dummy" codec and generic sound card. Signed-off-by: Nick Hollinghurst <[email protected]>
Since RP1 Audio Out can only work on GPIOs 12, 13 which would previously have needed dtoverlay=audremap, overload it both to enable and pin-map the block (do not enable for other pinouts). At the same time, generate a default "codec" and "sound card". Signed-off-by: Nick Hollinghurst <[email protected]>
njhollinghurst
force-pushed
the
audio_out
branch
from
37d7eca to
d16651b
Compare
Candidate version of an RP1 Audio Out driver
48kHz stereo only. GPIO 12/13 only.
Now rewritten as an ASOC DAI driver; a generic sound card and "dummy" codec must also be enabled via OF.