| Message ID | 20240422134354.89291-1-srinivas.kandagatla@linaro.org |
|---|---|
| Headers | show |
| Series | ASoC: qcom: display port changes | expand |
On Mon, Apr 22, 2024 at 02:43:53PM +0100, Srinivas Kandagatla wrote: > From: Srinivas Kandagatla <srinivas.kandagatla@linaro.org> > > Add support to Display Port Jack events, by making use of common helper s/to/for/ drop comma > function. > > Signed-off-by: Srinivas Kandagatla <srinivas.kandagatla@linaro.org> > --- > sound/soc/qcom/sc8280xp.c | 14 ++++++++++++++ > 1 file changed, 14 insertions(+) > > diff --git a/sound/soc/qcom/sc8280xp.c b/sound/soc/qcom/sc8280xp.c > index 878bd50ad4a7..38f97f19add9 100644 > --- a/sound/soc/qcom/sc8280xp.c > +++ b/sound/soc/qcom/sc8280xp.c > @@ -19,6 +19,7 @@ struct sc8280xp_snd_data { > struct snd_soc_card *card; > struct sdw_stream_runtime *sruntime[AFE_PORT_MAX]; > struct snd_soc_jack jack; > + struct snd_soc_jack hdmi_jack[8]; dp_jack > bool jack_setup; > }; > > @@ -27,6 +28,8 @@ static int sc8280xp_snd_init(struct snd_soc_pcm_runtime *rtd) > struct sc8280xp_snd_data *data = snd_soc_card_get_drvdata(rtd->card); > struct snd_soc_dai *cpu_dai = snd_soc_rtd_to_cpu(rtd, 0); > struct snd_soc_card *card = rtd->card; > + struct snd_soc_jack *hdmi_jack = NULL; dp_jack stray whitespace before = > + int hdmi_pcm_id = 0; dp_pcm_id no need to init > switch (cpu_dai->id) { > case WSA_CODEC_DMA_RX_0: > @@ -41,10 +44,21 @@ static int sc8280xp_snd_init(struct snd_soc_pcm_runtime *rtd) > snd_soc_limit_volume(card, "SpkrLeft PA Volume", 17); > snd_soc_limit_volume(card, "SpkrRight PA Volume", 17); > break; > + case DISPLAY_PORT_RX_0: > + hdmi_pcm_id = 0; > + hdmi_jack = &data->hdmi_jack[hdmi_pcm_id]; > + break; > + case DISPLAY_PORT_RX_1 ... DISPLAY_PORT_RX_7: > + hdmi_pcm_id = cpu_dai->id - DISPLAY_PORT_RX_1 + 1; > + hdmi_jack = &data->hdmi_jack[hdmi_pcm_id]; > + break; > default: > break; > } > > + if (hdmi_jack) > + return qcom_snd_dp_jack_setup(rtd, hdmi_jack, hdmi_pcm_id); > + > return qcom_snd_wcd_jack_setup(rtd, &data->jack, &data->jack_setup); > } Johan
On Mon, Apr 22, 2024 at 02:43:50PM +0100, Srinivas Kandagatla wrote: > From: Srinivas Kandagatla <srinivas.kandagatla@linaro.org> > > This patchset adds support for. > 1. parse Display Port module tokens from ASoC topology > 2. add support to DP/HDMI Jack events. > 3. fixes a typo in function name in sm8250 > > Verified these patches on X13s along with changes to tplg in > https://git.codelinaro.org/linaro/qcomlt/audioreach-topology/-/tree/topic/x13s-dp?ref_type=heads > and ucm changes from https://github.com/Srinivas-Kandagatla/alsa-ucm-conf/tree/topic/x13s-dp It looks like your UCM changes are still muxing the speaker and *each* displayport output so that you can only use one device at a time (i.e. only Speaker or DP1 or DP2 can be used). As we discussed off list last week, this seems unnecessarily limited and as far as I understood is mostly needed to work around some implementation details (not sure why DP1 and DP2 can't be used in parallel either). Can you please describe the problem here so that we can discuss this before merging an unnecessarily restricted solution which may later be harder to change (e.g. as kernel, topology and ucm may again need to be updated in lock step). >From what I could tell after a quick look, this series does not necessarily depend on muxing things this way, but please confirm that too. Johan
On Mon, Apr 22, 2024 at 02:43:52PM +0100, Srinivas Kandagatla wrote: > static const struct snd_soc_dapm_widget qcom_jack_snd_widgets[] = { > SND_SOC_DAPM_HP("Headphone Jack", NULL), > SND_SOC_DAPM_MIC("Mic Jack", NULL), > + SND_SOC_DAPM_SPK("HDMI/DP0 Jack", NULL), > + SND_SOC_DAPM_SPK("HDMI/DP1 Jack", NULL), > + SND_SOC_DAPM_SPK("HDMI/DP2 Jack", NULL), Shouldn't these be split in dedicated HDMI and DP jacks too? What if you have a machine with HDMI and DP jacks which would otherwise both claim "HDMI/DP0"? Johan
On 23/04/2024 12:59, Johan Hovold wrote: > On Mon, Apr 22, 2024 at 02:43:50PM +0100, Srinivas Kandagatla wrote: >> From: Srinivas Kandagatla <srinivas.kandagatla@linaro.org> >> >> This patchset adds support for. >> 1. parse Display Port module tokens from ASoC topology >> 2. add support to DP/HDMI Jack events. >> 3. fixes a typo in function name in sm8250 >> >> Verified these patches on X13s along with changes to tplg in >> https://git.codelinaro.org/linaro/qcomlt/audioreach-topology/-/tree/topic/x13s-dp?ref_type=heads >> and ucm changes from https://github.com/Srinivas-Kandagatla/alsa-ucm-conf/tree/topic/x13s-dp > > It looks like your UCM changes are still muxing the speaker and *each* > displayport output so that you can only use one device at a time (i.e. > only Speaker or DP1 or DP2 can be used). that is true. What is the use-case to use more than one audio sink devices at the same time for a laptops? How do you test it? I never tested anything like that on a full desktop setup. May be some manual setup in Wireplumber, but not 100% sure about multiple stream handling. > > As we discussed off list last week, this seems unnecessarily limited and > as far as I understood is mostly needed to work around some > implementation details (not sure why DP1 and DP2 can't be used in > parallel either). It is absolutely possible to run all the streams in parallel from the Audio hardware and DSP point of view. One thing to note is, On Qualcomm DP IP, we can not read/write registers if the DP port is not connected, which means that we can not send data in such cases. This makes it challenging to work with sound-servers like pipewire or pulseaudio as they tend to send silence data at very early stages in the full system boot up, ignoring state of the Jack events. > > Can you please describe the problem here so that we can discuss this > before merging an unnecessarily restricted solution which may later be > harder to change (e.g. as kernel, topology and ucm may again need to be > updated in lock step). > > From what I could tell after a quick look, this series does not > necessarily depend on muxing things this way, but please confirm that > too. These patches have nothing to do with how we model the muxing in UCM or in tplg. so these can go as it is irrespective of how we want to model the DP sinks in the UCM or tplg. --srini > > Johan
On Tue, Apr 23, 2024 at 01:38:18PM +0100, Srinivas Kandagatla wrote: > On 23/04/2024 12:59, Johan Hovold wrote: > > It looks like your UCM changes are still muxing the speaker and *each* > > displayport output so that you can only use one device at a time (i.e. > > only Speaker or DP1 or DP2 can be used). > that is true. > > What is the use-case to use more than one audio sink devices at the same > time for a laptops? I can imagine streaming audio and video to a TV (or audio to a soundbar) over DP while playing systems sounds and doing video conferencing using the internal speakers (or the other DP port). > How do you test it? I never tested anything like that on a full desktop > setup. You can select the sink per application in pavucontrol. Just verified that playing audio over the 3.5 mm jack while playing system sounds using the internal speakers works just fine. > > As we discussed off list last week, this seems unnecessarily limited and > > as far as I understood is mostly needed to work around some > > implementation details (not sure why DP1 and DP2 can't be used in > > parallel either). > > It is absolutely possible to run all the streams in parallel from the > Audio hardware and DSP point of view. > > One thing to note is, On Qualcomm DP IP, we can not read/write registers > if the DP port is not connected, which means that we can not send data > in such cases. > > This makes it challenging to work with sound-servers like pipewire or > pulseaudio as they tend to send silence data at very early stages in the > full system boot up, ignoring state of the Jack events. This bit sounds like it can and should be worked around by the driver to avoid hard-coding policy which would prevent use cases such as the ones mentioned above. > > Can you please describe the problem here so that we can discuss this > > before merging an unnecessarily restricted solution which may later be > > harder to change (e.g. as kernel, topology and ucm may again need to be > > updated in lock step). > > > > From what I could tell after a quick look, this series does not > > necessarily depend on muxing things this way, but please confirm that > > too. > > These patches have nothing to do with how we model the muxing in UCM or > in tplg. > > so these can go as it is irrespective of how we want to model the DP > sinks in the UCM or tplg. Thanks for confirming. Johan
On 23/04/2024 13:02, Johan Hovold wrote: > On Mon, Apr 22, 2024 at 02:43:52PM +0100, Srinivas Kandagatla wrote: > >> static const struct snd_soc_dapm_widget qcom_jack_snd_widgets[] = { >> SND_SOC_DAPM_HP("Headphone Jack", NULL), >> SND_SOC_DAPM_MIC("Mic Jack", NULL), >> + SND_SOC_DAPM_SPK("HDMI/DP0 Jack", NULL), >> + SND_SOC_DAPM_SPK("HDMI/DP1 Jack", NULL), >> + SND_SOC_DAPM_SPK("HDMI/DP2 Jack", NULL), > > Shouldn't these be split in dedicated HDMI and DP jacks too? What if you > have a machine with HDMI and DP jacks which would otherwise both claim > "HDMI/DP0"? These map to the Jack's added as part of qcom_snd_dp_jack_setup and belong to DISPLAY_PORT_RX_0, DISPLAY_PORT_RX_1, DISPLAY_PORT_RX_2. If its going via USB-C DP controller it will be either DP or an HDMI ? This is the most common naming for the USB-C DP/HDMI jack events. Qualcomm LPASS in some older SoCs had a dedicated HDMI interface which is different to this one. Usual Other ways to connect HDMI is via external HDMI Bridge using I2S interface which totally different to this DP interface. So none of these will conflict. hope this clarifies. thanks, Srini > > Johan
On 23/04/2024 15:58, Johan Hovold wrote: >> It is absolutely possible to run all the streams in parallel from the >> Audio hardware and DSP point of view. >> >> One thing to note is, On Qualcomm DP IP, we can not read/write registers >> if the DP port is not connected, which means that we can not send data >> in such cases. >> >> This makes it challenging to work with sound-servers like pipewire or >> pulseaudio as they tend to send silence data at very early stages in the >> full system boot up, ignoring state of the Jack events. > This bit sounds like it can and should be worked around by the driver to > avoid hard-coding policy which would prevent use cases such as the ones > mentioned above. This is not simple as you say. We have to fit these into a proper DPCM. Either we have a dummy Backend connected for each of these pcm sub-devices when DP port is not connected and then switch back to DP when its connected. Or somehow find a way to not let the pipewire talk to devices which are not connected. thanks, srini
On Tue, Apr 23, 2024 at 04:59:56PM +0100, Srinivas Kandagatla wrote: > On 23/04/2024 15:58, Johan Hovold wrote: > >> It is absolutely possible to run all the streams in parallel from the > >> Audio hardware and DSP point of view. > >> > >> One thing to note is, On Qualcomm DP IP, we can not read/write registers > >> if the DP port is not connected, which means that we can not send data > >> in such cases. > >> > >> This makes it challenging to work with sound-servers like pipewire or > >> pulseaudio as they tend to send silence data at very early stages in the > >> full system boot up, ignoring state of the Jack events. > > This bit sounds like it can and should be worked around by the driver to > > avoid hard-coding policy which would prevent use cases such as the ones > > mentioned above. > This is not simple as you say. We have to fit these into a proper DPCM. > Either we have a dummy Backend connected for each of these pcm > sub-devices when DP port is not connected and then switch back to DP > when its connected. I don't know how best to implement it, but we shouldn't necessarily let that determine the user experience. > Or somehow find a way to not let the pipewire talk to devices which are > not connected. Yes, perhaps it requires a change in user space. But it seems the kernel should be able to fake whatever probing user space currently does to determine if the there is a DP jack (even when there is nothing connected). Johan
Hi Xilin, On 23/05/2024 05:09, Xilin Wu wrote: >> >> Srinivas Kandagatla (4): >> ASoC: qcom: q6dsp: parse Display port tokens >> ASoC: qcom: common: add Display port Jack function >> ASoC: qcom: sc8280xp: add Display port Jack >> ASoC: qcom: sm8250: fix a typo in function name >> >> sound/soc/qcom/common.c | 29 +++++++++++++++++++++++++++++ >> sound/soc/qcom/common.h | 3 +++ >> sound/soc/qcom/qdsp6/topology.c | 26 ++++++++++++++++++++++++++ >> sound/soc/qcom/sc8280xp.c | 14 ++++++++++++++ >> sound/soc/qcom/sm8250.c | 4 ++-- >> 5 files changed, 74 insertions(+), 2 deletions(-) >> > > Hi Srini, > > I tested this series on SM8550 with tplg in [1] and ucm in [2]. But the > kernel output errors attached below. Headphone does work properly > without DisplayPort in the ucm. > > What could be the possible cause of this? Is there any significant > change from sc8280xp to sm8550? > > -- > Thanks, > Xilin Wu > > [1] > https://github.com/edk2-porting/audioreach-topology/blob/sakuramist/QCS8550-AYN-ODIN2.m4 > [2] > https://github.com/strongtz/alsa-ucm-conf/blob/odin2/ucm2/Qualcomm/sm8550/HiFi.conf > > [ 1552.313713] qcom-apm gprsvc:service:2:1: Error (1) Processing > 0x01001000 cmd > [ 1552.313730] qcom-apm gprsvc:service:2:1: DSP returned error[1001000] 1 > [ 1552.314455] qcom-apm gprsvc:service:2:1: Error (1) Processing Is the DP cable connected? if its not connected the dsp will throw this error. due to this issue I did workaround this issue by modeling it as conflicting device to Speaker in x13s ucm. I see in your ucm setup its not the case. which is why you might be hitting this issue. Can you try amixer -c 0 cset iface=MIXER,name='DISPLAY_PORT_RX_0 Audio Mixer MultiMedia2' 1 aplay -D plughw:0,1 some-wav-file.wav both with and without display connected. --srini > 0x01001006 cmd > [ 1552.314463] qcom-apm gprsvc:service:2:1: DSP returned error[1001006] 1 > [ 1552.315496] qcom-apm gprsvc:service:2:1: Error (1) Processing > 0x01001006 cmd > [ 1552.315506] qcom-apm gprsvc:service:2:1: DSP returned error[1001006] 1 > [ 1552.316033] qcom-apm gprsvc:service:2:1: Error (1) Processing > 0x01001001 cmd > [ 1552.316042] qcom-apm gprsvc:service:2:1: DSP returned error[1001001] 1 > [ 1552.316045] q6apm-lpass-dais > 30000000.remoteproc:glink-edge:gpr:service@1:bedais: Failed to prepare > Graph -22 > [ 1552.316047] q6apm-lpass-dais > 30000000.remoteproc:glink-edge:gpr:service@1:bedais: ASoC: error at > snd_soc_pcm_dai_prepare on DISPLAY_PORT_RX_0: -22
On Thu, May 09, 2024 at 09:59:40AM +0100, Srinivas Kandagatla wrote: > > > On 29/04/2024 15:54, Johan Hovold wrote: > > On Tue, Apr 23, 2024 at 04:55:32PM +0100, Srinivas Kandagatla wrote: > > > On 23/04/2024 13:02, Johan Hovold wrote: > > > > On Mon, Apr 22, 2024 at 02:43:52PM +0100, Srinivas Kandagatla wrote: > > > > > static const struct snd_soc_dapm_widget qcom_jack_snd_widgets[] = { > > > > > SND_SOC_DAPM_HP("Headphone Jack", NULL), > > > > > SND_SOC_DAPM_MIC("Mic Jack", NULL), > > > > > + SND_SOC_DAPM_SPK("HDMI/DP0 Jack", NULL), > > > > > + SND_SOC_DAPM_SPK("HDMI/DP1 Jack", NULL), > > > > > + SND_SOC_DAPM_SPK("HDMI/DP2 Jack", NULL), > > > > > > > > Shouldn't these be split in dedicated HDMI and DP jacks too? What if you > > > > have a machine with HDMI and DP jacks which would otherwise both claim > > > > "HDMI/DP0"? > > > > > > These map to the Jack's added as part of qcom_snd_dp_jack_setup and > > > belong to DISPLAY_PORT_RX_0, DISPLAY_PORT_RX_1, DISPLAY_PORT_RX_2. > > > > > > If its going via USB-C DP controller it will be either DP or an HDMI ? > > > > It will always be DP out of the machine even if an adapter can convert > > to HDMI internally. > > > > The DRM ports are called "DP-1" and "DP-2" so it seems we should match > > that. > > > > > This is the most common naming for the USB-C DP/HDMI jack events. > > > > It looks like some Intel machines use names like "HDMI/DP, pcm=%d Jack" > > (with a pcm device number), but we also have "DP Jack". Not sure which > > are are used with USB-C, though. (Or if the former actually support HDMI > > altmode.) > > I checked this on my machine which has usb-c and I can confirm using HDMI/DP > naming for these jack. > > Either way I don't mind having any names, but my point here is to be more > consistent across. I fear it is till not consistent. On the Intel laptop I see following jacks: numid=18,iface=CARD,name='HDMI/DP,pcm=3 Jack' numid=24,iface=CARD,name='HDMI/DP,pcm=7 Jack' numid=30,iface=CARD,name='HDMI/DP,pcm=8 Jack' On the other hand Mediatek and RockChip use just 'DP Jack'. I'd suggest settling on the latter option. We are closer to MTK and RockChip rather than Intel. BTW: a platform can easily have 4 (x1e8100) or even 8 (sc8280xp) DP outputs. Could you please point out why there are just 3 jacks? > --srini > > > > > Qualcomm LPASS in some older SoCs had a dedicated HDMI interface which > > > is different to this one. > > > > > > Usual Other ways to connect HDMI is via external HDMI Bridge using I2S > > > interface which totally different to this DP interface. > > > > Sure, but if there's ever a design with such a port then it will be > > called "HDMI Jack" and then the "HDMI in "HDMI/DP0 Jack" is unnecessary > > and confusing when it is always DP out. > > > > Johan
On Tue, Apr 23, 2024 at 01:38:18PM +0100, Srinivas Kandagatla wrote: > > > On 23/04/2024 12:59, Johan Hovold wrote: > > On Mon, Apr 22, 2024 at 02:43:50PM +0100, Srinivas Kandagatla wrote: > > > From: Srinivas Kandagatla <srinivas.kandagatla@linaro.org> > > > > > > This patchset adds support for. > > > 1. parse Display Port module tokens from ASoC topology > > > 2. add support to DP/HDMI Jack events. > > > 3. fixes a typo in function name in sm8250 > > > > > > Verified these patches on X13s along with changes to tplg in > > > https://git.codelinaro.org/linaro/qcomlt/audioreach-topology/-/tree/topic/x13s-dp?ref_type=heads > > > and ucm changes from https://github.com/Srinivas-Kandagatla/alsa-ucm-conf/tree/topic/x13s-dp > > > > It looks like your UCM changes are still muxing the speaker and *each* > > displayport output so that you can only use one device at a time (i.e. > > only Speaker or DP1 or DP2 can be used). > that is true. > > What is the use-case to use more than one audio sink devices at the same > time for a laptops? Consider multi-seat setup, with each monitor having its own set of keyboard, mouse, headphone and user behind it. > > How do you test it? I never tested anything like that on a full desktop > setup. > > May be some manual setup in Wireplumber, but not 100% sure about multiple > stream handling. > > > > > As we discussed off list last week, this seems unnecessarily limited and > > as far as I understood is mostly needed to work around some > > implementation details (not sure why DP1 and DP2 can't be used in > > parallel either). > > It is absolutely possible to run all the streams in parallel from the Audio > hardware and DSP point of view. > > One thing to note is, On Qualcomm DP IP, we can not read/write registers if > the DP port is not connected, which means that we can not send data in such > cases. How is this handled for the native HDMI playback on platforms like Dragonboard 820c? As far as I was able to test, playback fails with -EIO if HDMI output is not enabled or if the DVI monitor is connected. > > This makes it challenging to work with sound-servers like pipewire or > pulseaudio as they tend to send silence data at very early stages in the > full system boot up, ignoring state of the Jack events. > > > > > Can you please describe the problem here so that we can discuss this > > before merging an unnecessarily restricted solution which may later be > > harder to change (e.g. as kernel, topology and ucm may again need to be > > updated in lock step). > > > > From what I could tell after a quick look, this series does not > > necessarily depend on muxing things this way, but please confirm that > > too. > > These patches have nothing to do with how we model the muxing in UCM or in > tplg. > > so these can go as it is irrespective of how we want to model the DP sinks > in the UCM or tplg. > > > --srini > > > > Johan
On 30/05/2024 23:38, Dmitry Baryshkov wrote: >>> It will always be DP out of the machine even if an adapter can convert >>> to HDMI internally. >>> >>> The DRM ports are called "DP-1" and "DP-2" so it seems we should match >>> that. >>> >>>> This is the most common naming for the USB-C DP/HDMI jack events. >>> It looks like some Intel machines use names like "HDMI/DP, pcm=%d Jack" >>> (with a pcm device number), but we also have "DP Jack". Not sure which >>> are are used with USB-C, though. (Or if the former actually support HDMI >>> altmode.) >> I checked this on my machine which has usb-c and I can confirm using HDMI/DP >> naming for these jack. >> >> Either way I don't mind having any names, but my point here is to be more >> consistent across. > I fear it is till not consistent. On the Intel laptop I see following > jacks: > > numid=18,iface=CARD,name='HDMI/DP,pcm=3 Jack' > numid=24,iface=CARD,name='HDMI/DP,pcm=7 Jack' > numid=30,iface=CARD,name='HDMI/DP,pcm=8 Jack' > > On the other hand Mediatek and RockChip use just 'DP Jack'. > > I'd suggest settling on the latter option. We are closer to MTK and > RockChip rather than Intel. that is fine with me. > > BTW: a platform can easily have 4 (x1e8100) or even 8 (sc8280xp) DP > outputs. Could you please point out why there are just 3 jacks? The CRD platform that I have access to has 3 ports which is why I started with 3 ports, but we can add more ports as and when we can really test them. --srini
From: Srinivas Kandagatla <srinivas.kandagatla@linaro.org> This patchset adds support for. 1. parse Display Port module tokens from ASoC topology 2. add support to DP/HDMI Jack events. 3. fixes a typo in function name in sm8250 Verified these patches on X13s along with changes to tplg in https://git.codelinaro.org/linaro/qcomlt/audioreach-topology/-/tree/topic/x13s-dp?ref_type=heads and ucm changes from https://github.com/Srinivas-Kandagatla/alsa-ucm-conf/tree/topic/x13s-dp Thanks, Srini Changes since v1: - Fixed unused variable warning. - fixed warning 'break;' to avoid fall-through Srinivas Kandagatla (4): ASoC: qcom: q6dsp: parse Display port tokens ASoC: qcom: common: add Display port Jack function ASoC: qcom: sc8280xp: add Display port Jack ASoC: qcom: sm8250: fix a typo in function name sound/soc/qcom/common.c | 29 +++++++++++++++++++++++++++++ sound/soc/qcom/common.h | 3 +++ sound/soc/qcom/qdsp6/topology.c | 26 ++++++++++++++++++++++++++ sound/soc/qcom/sc8280xp.c | 14 ++++++++++++++ sound/soc/qcom/sm8250.c | 4 ++-- 5 files changed, 74 insertions(+), 2 deletions(-)