ASoC: SOF: topology: Use acpi mach from the machine driver #5665
Conversation
| @@ -2530,8 +2531,12 @@ int snd_sof_load_topology(struct snd_soc_component *scomp, const char *file) | |||
| */ | |||
| bool no_fallback = strstr(file, "dummy"); | |||
|
|
|||
There was a problem hiding this comment.
We use the SOC_SDW_PCH_DMIC quirk to indicate that the PCH DMIC is used. We will set mach_params->dmic_num = DMIC_DEFAULT_CHANNELS; to the mach_params from dev_get_platdata(card->dev) when the quirk is set. But the mach_params of the sof_sdw_get_tplg_files() function is from sdev->pdata. In other words, the mach_params->dmic_num of sof_pdata->machine remains 0 and it will cause the unsupported number of dmics: 0 issue.
There was a problem hiding this comment.
Will the dmic_num override still going to be working?
But in other words we have two places whit dmic_num and they are used in a 'random' way around the code?
ujfalusi
left a comment
ujfalusi
left a comment
There was a problem hiding this comment.
do we want it to be backported to stable kernels? Fixes tag?
sound/soc/sof/topology.c
Outdated
| /* Try to use function topologies if possible */ | ||
| if (!sof_pdata->disable_function_topology && !disable_function_topology && | ||
| sof_pdata->machine && sof_pdata->machine->get_function_tplg_files) { | ||
| struct snd_soc_acpi_mach *card_mach = dev_get_platdata(scomp->card->dev); |
There was a problem hiding this comment.
This never going to be NULL, right?
There was a problem hiding this comment.
Good question @ujfalusi I don't think it will be NULL, but it is not guarantee that it is with the snd_soc_acpi_mach struct. I will move the change to sof_sdw_get_tplg_files() which is specific to Intel SDW machines.
| @@ -2530,8 +2531,12 @@ int snd_sof_load_topology(struct snd_soc_component *scomp, const char *file) | |||
| */ | |||
| bool no_fallback = strstr(file, "dummy"); | |||
|
|
|||
There was a problem hiding this comment.
Will the dmic_num override still going to be working?
But in other words we have two places whit dmic_num and they are used in a 'random' way around the code?
Yes, I will add it. |
bardliao
force-pushed
the
fix-dmic-quirk
branch
from
2f23594 to
3589a5e
Compare
bardliao
requested review from
dbaluta,
lgirdwood and
plbossart
as code owners
The parameters may be changed by the of_sdw achine driver based on the machine driver quirk. We need to use the acpi mach from the machine driver. Fixes: 2fbeff3 ("ASoC: Intel: add sof_sdw_get_tplg_files ops") Signed-off-by: Bard Liao <yung-chuan.liao@linux.intel.com>
bardliao
force-pushed
the
fix-dmic-quirk
branch
from
3589a5e to
64f86c1
Compare
3 participants
The parameters may be changed by the machine driver based on the actual hardware configuration. We need to use the acpi mach from the machine driver.