Re: [PATCH v2 2/3] ASoC: codecs: add new pm4125 audio codec driver

["Alexey Klimov" <alexey.klimov@linaro.org>]

On Fri Jul 11, 2025 at 2:32 PM BST, Srinivas Kandagatla wrote:
> On 7/11/25 4:00 AM, Alexey Klimov wrote:
>> The audio codec is found in Qualcomm PM2250/PM4125 PMICs and is used on
>> platforms like Qualcomm QCM2290. It has soundwire interface and
>> corresponding RX and TX slave devices.
>> 
>> It has only two input channels: HPH left and right. The line output (LO)
>> is linked to HPHL so the hardware has some limitations regarding concurrent
>> playback via HPH and LO for instance.
>> 
>> The codec driver also uses WCD MBCH framework. The MBHC functionality is
>> implemented in a minimalistic way to enable IRQs and avoid different
>> issues with IRQs.
>> 
>> Co-developed-by: Srinivas Kandagatla <srinivas.kandagatla@oss.qualcomm.com>
>> Signed-off-by: Srinivas Kandagatla <srinivas.kandagatla@oss.qualcomm.com>
>> Signed-off-by: Alexey Klimov <alexey.klimov@linaro.org>
>
> Thanks for doing this Alexey,
>
> I have tested Headset playback, headset recording and Lineout.
>
> While testing I have noticed one issue with the places where enable_irq
> is being used, there is a possiblity of things going wrong in some cases.
>
>
>> ---
>>  sound/soc/codecs/Kconfig      |   18 +
>>  sound/soc/codecs/Makefile     |    8 +
>>  sound/soc/codecs/pm4125-sdw.c |  546 +++++++++++++
>>  sound/soc/codecs/pm4125.c     | 1767 +++++++++++++++++++++++++++++++++++++++++
>>  sound/soc/codecs/pm4125.h     |  314 ++++++++
>>  5 files changed, 2653 insertions(+)
>> 
>
> ...
>> +static int pm4125_codec_enable_lo_pa(struct snd_soc_dapm_widget *w,
>> +				     struct snd_kcontrol *kcontrol, int event)
>> +{
>> +	struct snd_soc_component *component = snd_soc_dapm_to_component(w->dapm);
>> +	struct pm4125_priv *pm4125 = snd_soc_component_get_drvdata(component);
>> +
>> +	switch (event) {
>> +	case SND_SOC_DAPM_PRE_PMU:
>> +		snd_soc_component_update_bits(component, PM4125_ANA_COMBOPA_CTL_5, 0x04, 0x00);
>> +		usleep_range(1000, 1010);
>> +		snd_soc_component_update_bits(component, PM4125_ANA_COMBOPA_CTL_4, 0x0F, 0x0F);
>> +		usleep_range(1000, 1010);
>> +		snd_soc_component_write_field(component, PM4125_ANA_COMBOPA_CTL,
>> +					      PM4125_ANA_COMBO_PA_SELECT_MASK,
>> +					      PM4125_ANA_COMBO_PA_SELECT_LO);
>> +		snd_soc_component_write_field(component, PM4125_DIG_SWR_PDM_WD_CTL0,
>> +					      PM4125_WDT_ENABLE_MASK,
>> +					      (PM4125_WDT_ENABLE_RX0_M | PM4125_WDT_ENABLE_RX0_L));
>> +		break;
>> +	case SND_SOC_DAPM_POST_PMU:
>> +		usleep_range(5000, 5010);
>> +		snd_soc_component_update_bits(component, PM4125_ANA_COMBOPA_CTL_4, 0x0F, 0x04);
>> +		enable_irq(pm4125->hphl_pdm_wd_int);
>
> if headset playback dapm widgets are powered up or active and you try to
> play on lineout you would get below splat:
>
> [ 1181.352762] ------------[ cut here ]------------
> [ 1181.357435] Unbalanced enable for IRQ 156
> [ 1181.361489] WARNING: CPU: 1 PID: 877 at kernel/irq/manage.c:753
> __enable_irq+0x4c/0x7c
> ...
>
> [ 1181.483541] pstate: 600000c5 (nZCv daIF -PAN -UAO -TCO -DIT -SSBS
> BTYPE=--)
> [ 1181.490534] pc : __enable_irq+0x4c/0x7c
> [ 1181.494405] lr : __enable_irq+0x4c/0x7c
> [ 1181.498278] sp : ffff8000840ab8a0
> [ 1181.501618] x29: ffff8000840ab8a0 x28: ffffd79b9725d588 x27:
> ffff8000840ab948
> [ 1181.508796] x26: 000000000000309b x25: 0000000000000080 x24:
> 0000000000000080
> [ 1181.515980] x23: ffffd79b9698f250 x22: 0000000000000000 x21:
> 0000000000000000
> [ 1181.523153] x20: 000000000000009c x19: ffff23c347fd9400 x18:
> 0000000000000006
> [ 1181.530335] x17: 0000000000000001 x16: ffffd79b94dbcce8 x15:
> ffff8000840ab46f
> [ 1181.537508] x14: 0000000000000000 x13: 3635312051524920 x12:
> ffffd79b96d96768
> [ 1181.544691] x11: 0000000000000058 x10: 0000000000000018 x9 :
> ffffd79b96d96768
> [ 1181.551862] x8 : 000000000000024f x7 : ffffd79b96dee768 x6 :
> ffffd79b96dee768
> [ 1181.559044] x5 : ffff23c3bbbb8508 x4 : 0000000000000000 x3 :
> 0000000000000027
> [ 1181.566217] x2 : 0000000000000000 x1 : 0000000000000000 x0 :
> ffff23c35a0a4600
> [ 1181.573392] Call trace:
> [ 1181.575871]  __enable_irq+0x4c/0x7c (P)
> [ 1181.579748]  enable_irq+0x60/0xd8
> [ 1181.583098]  pm4125_codec_enable_lo_pa+0x6c/0x1a0 [snd_soc_pm4125]
> [ 1181.589319]  dapm_seq_check_event+0x11c/0x144
> [ 1181.593720]  dapm_seq_run_coalesced+0x130/0x1d0
> [ 1181.598286]  dapm_seq_run+0x264/0x374
> [ 1181.601991]  dapm_power_widgets+0x664/0x900
> [ 1181.606212]  snd_soc_dapm_stream_event+0xf4/0x178
> [ 1181.610957]  __soc_pcm_prepare+0x64/0x120
> [ 1181.614998]  dpcm_be_dai_prepare+0x120/0x168
> [ 1181.619299]  dpcm_fe_dai_prepare+0x94/0x178
> [ 1181.623517]  snd_pcm_do_prepare+0x30/0x50
> [ 1181.627567]  snd_pcm_action_single+0x48/0xa4
> [ 1181.631872]  snd_pcm_action_nonatomic+0x94/0xb0
> [ 1181.636435]  snd_pcm_prepare+0x7c/0xe0
> [ 1181.640222]  snd_pcm_common_ioctl+0xca4/0x16e0
> [ 1181.644705]  snd_pcm_ioctl+0x34/0x4c
> [ 1181.648315]  __arm64_sys_ioctl+0xac/0x104
> [ 1181.652355]  invoke_syscall+0x48/0x104
> [ 1181.656147]  el0_svc_common.constprop.0+0x40/0xe0
> [ 1181.660892]  do_el0_svc+0x1c/0x28
> [ 1181.664242]  el0_svc+0x30/0xcc
> [ 1181.667340]  el0t_64_sync_handler+0x10c/0x138
> [ 1181.671737]  el0t_64_sync+0x198/0x19c
> [ 1181.675430] ---[ end trace 0000000000000000 ]---


Yep, I noticed that one time only but was puzzled how to reproduce.
As you suggested during out of email list discussion I will move handling
of IRQs into separate widget.

Thank you,
Alexey