ASoC: soc-pcm: Preserve hw parameters from components in dpcm_runtime_setup_fe
PATCH 1/1 ASoC: soc-pcm: Preserve hw parameters from components in dpcm_runtime_setup_fe
Component drivers can prepare snd_pcm_hardware struct based on the hardware capabilities which information should not be discarded.
Only touch the rates, channels_max and formats if they were left to 0, otherwise keep the provided configuration intact for the parameter cross checking decision.
---
Hi,
Changes since RFC [1]:
- move the hw parameter initialization/preserve code inline to dpcm_runtime_setup_fe
this patch in essence extends the special casing of formats done by 083a25b18d6a ("ASoC: soc-pcm: fix hw->formats cleared by soc_pcm_hw_init() for dpcm")
Other parameters might have been set in the same way as the formats and preserving these are equally important.
A case for this is SOF used with HDA codec (analog or more importantly, HDMI) where the hw-> params are set based on the connected display/device and should be preserved so we can report correct rate, format and channels supported by the equipment.
If the hw-> parameters are left uninitialized then we still need to set the UINT/ULLONG_MAX for the refining code to work.
This applies only for FE setup, in other cases we shall (as before) do a full re-initialization.
I think this makes sense and I cannot think where this might flop, but sent as RFC to see what people think.
Br,
Peter
[1] https://lore.kernel.org/linux-sound/[email protected]/
sound/soc/soc-pcm.c | 18 +++++++++++++----- 1 file changed, 13 insertions(+), 5 deletions(-)
sound/soc/soc-pcm.c +13 -5
@@ -1738,13 +1738,21 @@ static void dpcm_runtime_setup_fe(struct snd_pcm_substream *substream)struct snd_pcm_hardware *hw = &runtime->hw;struct snd_soc_dai *dai;int stream = substream->stream; -u64 formats = hw->formats;int i; -soc_pcm_hw_init (hw); - -if (formats) - hw->formats &= formats; +/* +* Initialize only pcm hardware patameters that has not been initialized +* and preserve the configuration which might have been provided by +* component drivers +*/ +if (!hw->rates) + hw->rates =UINT_MAX ; +if (!hw->rate_max) + hw->rate_max =UINT_MAX ; +if (!hw->channels_max) + hw->channels_max =UINT_MAX ; +if (!hw->formats) + hw->formats =ULLONG_MAX ;for_each_rtd_cpu_dais (fe, i, dai) {const struct snd_soc_pcm_stream *cpu_stream;
Signature · 1 line
2.51.1.dirty
Hi Peter
Thank you for the patch
Quoted 9 lines
Component drivers can prepare
snd_pcm_hardwarestruct based on the hardware capabilities which information should not be discarded.Only touch the rates,
channels_maxand formats if they were left to 0, otherwise keep the provided configuration intact for the parameter cross checking decision.Signed-off-by Peter Ujfalusi <[email protected]>---
(snip)
Quoted 7 lines
sound/soc/soc-pcm.c +3
-4-soc_pcm_hw_init (hw); - -if (formats) - hw->formats &= formats; +/* +* Initialize only pcm hardware patameters that has not been initialized +* and preserve the configuration which might have been provided by
sound/soc/soc-pcm.c static void dpcm_runtime_setup_fe(struct snd_pcm_substream *substream) · line 1755
Quoted 10 lines
sound/soc/soc-pcm.c +10
-0+* component drivers +*/ +if (!hw->rates) + hw->rates =UINT_MAX ; +if (!hw->rate_max) + hw->rate_max =UINT_MAX ; +if (!hw->channels_max) + hw->channels_max =UINT_MAX ; +if (!hw->formats) + hw->formats =ULLONG_MAX ;
soc_pcm_hw_init() is used from only 2 functions
(a) snd_soc_runtime_calc_hw() (b) dpcm_runtime_setup_fe()
And this patch fixed not to use it in (b).
Setting same variable in multiple method is not good idea, IMO.
How about like this ?
static void soc_pcm_hw_init(struct snd_pcm_hardware *hw, boot force)
{
if (force || !hw->rates)
hw->rates = UINT_MAX;
if (force || !hw->rate_max)
hw->rate_max = UINT_MAX;
if (force || !hw->channels_max)
hw->channels_max = UINT_MAX;
if (force || !hw->formats)
hw->formats = ULLONG_MAX;
hw->rate_min = 0;
hw->channels_min = 0;
}
int snd_soc_runtime_calc_hw(...)
{
...
soc_pcm_hw_init(hw, true);
...
}
int dpcm_runtime_setup_fe(...)
{
...
soc_pcm_hw_init(hw, false);
...
}Thank you for your help !!
Best regards
---
Kuninori Morimoto
Hi Morimoto-san,
Quoted 35 lines · Kuninori Morimoto
Hi Peter
Thank you for the patch
Quoted 9 lines
Component drivers can prepare
snd_pcm_hardwarestruct based on the hardware capabilities which information should not be discarded.Only touch the rates,
channels_maxand formats if they were left to 0, otherwise keep the provided configuration intact for the parameter cross checking decision.Signed-off-by Peter Ujfalusi <[email protected]>---
(snip)
Quoted 7 lines
sound/soc/soc-pcm.c +3
-4-soc_pcm_hw_init (hw); - -if (formats) - hw->formats &= formats; +/* +* Initialize only pcm hardware patameters that has not been initialized +* and preserve the configuration which might have been provided by sound/soc/soc-pcm.c static void dpcm_runtime_setup_fe(struct snd_pcm_substream *substream) · line 1755
Quoted 10 lines
sound/soc/soc-pcm.c +10
-0+* component drivers +*/ +if (!hw->rates) + hw->rates =UINT_MAX ; +if (!hw->rate_max) + hw->rate_max =UINT_MAX ; +if (!hw->channels_max) + hw->channels_max =UINT_MAX ; +if (!hw->formats) + hw->formats =ULLONG_MAX ;
soc_pcm_hw_init()is used from only 2 functions
Quoted 7 lines
(a) snd_soc_runtime_calc_hw() (b) dpcm_runtime_setup_fe()And this patch fixed not to use it in (b).
Setting same variable in multiple method is not good idea, IMO.How about like this ?
This is how the RFC was done, but decided to not modify soc_pcm_hw_init() with an added bool.
Other option would be to add a new function which would do only conditional initialization.
https://lore.kernel.org/linux-sound/[email protected]/
Quoted 14 lines
static void soc_pcm_hw_init(struct snd_pcm_hardware *hw, boot force)
{
if (force || !hw->rates)
hw->rates = UINT_MAX;
if (force || !hw->rate_max)
hw->rate_max = UINT_MAX;
if (force || !hw->channels_max)
hw->channels_max = UINT_MAX;
if (force || !hw->formats)
hw->formats = ULLONG_MAX;
hw->rate_min = 0;
hw->channels_min = 0;This way we would loose the rate_min, channel_min from components, see my RFC which covers this, but it is using reverse condition (preserve vs force).
The force boolean might be better than the preserve_config as parameter if the preference is to modify the soc_pcm_hw_init()
Quoted 22 lines
} int snd_soc_runtime_calc_hw(...) { ... soc_pcm_hw_init(hw, true); ... } int dpcm_runtime_setup_fe(...) { ... soc_pcm_hw_init(hw, false); ... }Thank you for your help !!
Best regards
---
Kuninori Morimoto
Signature · 1 line
Péter
Hi Péter
Thank you for your feedback
Quoted 4 lines
This is how the RFC was done, but decided to not modify
soc_pcm_hw_init()with an added bool.
Other option would be to add a new function which would do only conditional initialization.
Yes, I have noticed about this RFC, and you used other metod 083a25b18d6a.
Quoted 5 lines
This way we would loose the
rate_min,channel_minfrom components, see my RFC which covers this, but it is using reverse condition (preserve vs force).
The force boolean might be better than thepreserve_configas parameter if the preference is to modify thesoc_pcm_hw_init()
I guess you can ignore setup xx_min if "force" was set ?
Thank you for your help !!
Best regards
---
Kuninori Morimoto
Hi Morimoto-san,
Quoted 5 lines · Kuninori Morimoto
Hi Péter
Thank you for your feedback
Quoted 7 lines
Quoted 4 lines
This is how the RFC was done, but decided to not modify
soc_pcm_hw_init()with an added bool.
Other option would be to add a new function which would do only conditional initialization.Yes, I have noticed about this RFC, and you used other metod
083a25b18d6a.
I,m not sure what 083a25b18d6a is.
Quoted 7 lines
Quoted 5 lines
This way we would loose the
rate_min,channel_minfrom components, see my RFC which covers this, but it is using reverse condition (preserve vs force).
The force boolean might be better than thepreserve_configas parameter if the preference is to modify thesoc_pcm_hw_init()I guess you can ignore setup
xx_minif "force" was set ?
Yes, in a same way as it was done in RFC under the preserve_config case.
Quoted 6 lines
Thank you for your help !!
Best regards
---
Kuninori Morimoto
Signature · 1 line
Péter
Mark Brown
maintainerQuoted 6 lines · Péter Ujfalusi
Quoted 2 lines · Kuninori Morimoto
Yes, I have noticed about this RFC, and you used other metod
083a25b18d6a.I,m not sure what
083a25b18d6ais.
https://lore.kernel.org/r/[email protected]
(that's a commit ID upstream).