| Message ID | 20230718084522.116952-4-maarten.lankhorst@linux.intel.com (mailing list archive) |
|---|---|
| State | Superseded |
| Headers | show |
| Series | sound: Use -EPROBE_DEFER instead of i915 module loading. | expand |
On Tue, Jul 18, 2023 at 10:45:18AM +0200, Maarten Lankhorst wrote: > Now that we can use -EPROBE_DEFER, it's no longer required to spin off > the snd_hdac_i915_init into a workqueue. It's likely the whole workqueue > can be destroyed, but I don't have the means to test this. Acked-by: Mark Brown <broonie@kernel.org>
On 2023-07-18 10:45 AM, Maarten Lankhorst wrote: > Now that we can use -EPROBE_DEFER, it's no longer required to spin off > the snd_hdac_i915_init into a workqueue. It's likely the whole workqueue > can be destroyed, but I don't have the means to test this. > > Removing the workqueue would simplify init even further, but is left > as exercise for the reviewer. > > Signed-off-by: Maarten Lankhorst <maarten.lankhorst@linux.intel.com> > --- > sound/soc/intel/avs/core.c | 13 +++++++++---- > 1 file changed, 9 insertions(+), 4 deletions(-) > > diff --git a/sound/soc/intel/avs/core.c b/sound/soc/intel/avs/core.c > index 3311a6f142001..d3a7f42387e9b 100644 > --- a/sound/soc/intel/avs/core.c > +++ b/sound/soc/intel/avs/core.c > @@ -191,10 +191,6 @@ static void avs_hda_probe_work(struct work_struct *work) > > pm_runtime_set_active(bus->dev); /* clear runtime_error flag */ > > - ret = snd_hdac_i915_init(bus, true); > - if (ret < 0) > - dev_info(bus->dev, "i915 init unsuccessful: %d\n", ret); > - > snd_hdac_display_power(bus, HDA_CODEC_IDX_CONTROLLER, true); > avs_hdac_bus_init_chip(bus, true); > avs_hdac_bus_probe_codecs(bus); > @@ -465,10 +461,19 @@ static int avs_pci_probe(struct pci_dev *pci, const struct pci_device_id *id) > pci_set_drvdata(pci, bus); > device_disable_async_suspend(dev); > > + ret = snd_hdac_i915_init(bus, false); > + if (ret == -EPROBE_DEFER) > + goto err_unmaster; > + else if (ret < 0) > + dev_info(bus->dev, "i915 init unsuccessful: %d\n", ret); > + While our tests are currently passing I have my doubts about EPROBE_DEFER. We do want to have audio functionality there even if some problems with HDMI arise along the way - some audio is better than no audio. Here, i915 may ruin the day for a platform equipped with hda/hdmi/i2c/dmic chips simultaneously. Also, why call snd_hdac_i915_init() _after_ setting drvdata? > schedule_work(&adev->probe_work); > > return 0; > > +err_unmaster: > + pci_clear_master(pci); > + pci_set_drvdata(pci, NULL); Not a fan. It's hard to grasp entire error-step within name of a single label. Thus I'd suggest "err_<cause>" naming pattern instead. Even here, under "unmaster" you hid clearing master and drvdata both. Let's do "err_i915_init" instead. > err_acquire_irq: > snd_hdac_bus_free_stream_pages(bus); > snd_hdac_ext_stream_free_all(bus);
On Wed, 19 Jul 2023 17:26:24 +0200, Cezary Rojewski wrote: > > On 2023-07-18 10:45 AM, Maarten Lankhorst wrote: > > Now that we can use -EPROBE_DEFER, it's no longer required to spin off > > the snd_hdac_i915_init into a workqueue. It's likely the whole workqueue > > can be destroyed, but I don't have the means to test this. > > > > Removing the workqueue would simplify init even further, but is left > > as exercise for the reviewer. > > > > Signed-off-by: Maarten Lankhorst <maarten.lankhorst@linux.intel.com> > > --- > > sound/soc/intel/avs/core.c | 13 +++++++++---- > > 1 file changed, 9 insertions(+), 4 deletions(-) > > > > diff --git a/sound/soc/intel/avs/core.c b/sound/soc/intel/avs/core.c > > index 3311a6f142001..d3a7f42387e9b 100644 > > --- a/sound/soc/intel/avs/core.c > > +++ b/sound/soc/intel/avs/core.c > > @@ -191,10 +191,6 @@ static void avs_hda_probe_work(struct work_struct *work) > > pm_runtime_set_active(bus->dev); /* clear runtime_error flag > > */ > > - ret = snd_hdac_i915_init(bus, true); > > - if (ret < 0) > > - dev_info(bus->dev, "i915 init unsuccessful: %d\n", ret); > > - > > snd_hdac_display_power(bus, HDA_CODEC_IDX_CONTROLLER, true); > > avs_hdac_bus_init_chip(bus, true); > > avs_hdac_bus_probe_codecs(bus); > > @@ -465,10 +461,19 @@ static int avs_pci_probe(struct pci_dev *pci, const struct pci_device_id *id) > > pci_set_drvdata(pci, bus); > > device_disable_async_suspend(dev); > > + ret = snd_hdac_i915_init(bus, false); > > + if (ret == -EPROBE_DEFER) > > + goto err_unmaster; > > + else if (ret < 0) > > + dev_info(bus->dev, "i915 init unsuccessful: %d\n", ret); > > + > > While our tests are currently passing I have my doubts about > EPROBE_DEFER. We do want to have audio functionality there even if > some problems with HDMI arise along the way - some audio is better > than no audio. Here, i915 may ruin the day for a platform equipped > with hda/hdmi/i2c/dmic chips simultaneously. Yeah, we need some workaround to let user to skip the i915 binding, as discussed in the thread. I guess an option for hdac_i915 (that will be in snd-hda-core module) should suffice. thanks, Takashi
Hey, On 2023-07-19 17:26, Cezary Rojewski wrote: > On 2023-07-18 10:45 AM, Maarten Lankhorst wrote: >> Now that we can use -EPROBE_DEFER, it's no longer required to spin off >> the snd_hdac_i915_init into a workqueue. It's likely the whole workqueue >> can be destroyed, but I don't have the means to test this. >> >> Removing the workqueue would simplify init even further, but is left >> as exercise for the reviewer. >> >> Signed-off-by: Maarten Lankhorst <maarten.lankhorst@linux.intel.com> >> --- >> sound/soc/intel/avs/core.c | 13 +++++++++---- >> 1 file changed, 9 insertions(+), 4 deletions(-) >> >> diff --git a/sound/soc/intel/avs/core.c b/sound/soc/intel/avs/core.c >> index 3311a6f142001..d3a7f42387e9b 100644 >> --- a/sound/soc/intel/avs/core.c >> +++ b/sound/soc/intel/avs/core.c >> @@ -191,10 +191,6 @@ static void avs_hda_probe_work(struct work_struct >> *work) >> pm_runtime_set_active(bus->dev); /* clear runtime_error flag */ >> - ret = snd_hdac_i915_init(bus, true); >> - if (ret < 0) >> - dev_info(bus->dev, "i915 init unsuccessful: %d\n", ret); >> - >> snd_hdac_display_power(bus, HDA_CODEC_IDX_CONTROLLER, true); >> avs_hdac_bus_init_chip(bus, true); >> avs_hdac_bus_probe_codecs(bus); >> @@ -465,10 +461,19 @@ static int avs_pci_probe(struct pci_dev *pci, >> const struct pci_device_id *id) >> pci_set_drvdata(pci, bus); >> device_disable_async_suspend(dev); >> + ret = snd_hdac_i915_init(bus, false); >> + if (ret == -EPROBE_DEFER) >> + goto err_unmaster; >> + else if (ret < 0) >> + dev_info(bus->dev, "i915 init unsuccessful: %d\n", ret); >> + > > While our tests are currently passing I have my doubts about > EPROBE_DEFER. We do want to have audio functionality there even if some > problems with HDMI arise along the way - some audio is better than no > audio. Here, i915 may ruin the day for a platform equipped with > hda/hdmi/i2c/dmic chips simultaneously. > > Also, why call snd_hdac_i915_init() _after_ setting drvdata? I wanted to make the most cautious change. The previous flow called the function immediately from probe_work, so I moved it to right before the scheduling of probe_work >> schedule_work(&adev->probe_work); >> return 0; >> +err_unmaster: >> + pci_clear_master(pci); >> + pci_set_drvdata(pci, NULL); > > Not a fan. It's hard to grasp entire error-step within name of a single > label. Thus I'd suggest "err_<cause>" naming pattern instead. Even here, > under "unmaster" you hid clearing master and drvdata both. Let's do > "err_i915_init" instead. > Ok, will rename. ~Maarten
diff --git a/sound/soc/intel/avs/core.c b/sound/soc/intel/avs/core.c index 3311a6f142001..d3a7f42387e9b 100644 --- a/sound/soc/intel/avs/core.c +++ b/sound/soc/intel/avs/core.c @@ -191,10 +191,6 @@ static void avs_hda_probe_work(struct work_struct *work) pm_runtime_set_active(bus->dev); /* clear runtime_error flag */ - ret = snd_hdac_i915_init(bus, true); - if (ret < 0) - dev_info(bus->dev, "i915 init unsuccessful: %d\n", ret); - snd_hdac_display_power(bus, HDA_CODEC_IDX_CONTROLLER, true); avs_hdac_bus_init_chip(bus, true); avs_hdac_bus_probe_codecs(bus); @@ -465,10 +461,19 @@ static int avs_pci_probe(struct pci_dev *pci, const struct pci_device_id *id) pci_set_drvdata(pci, bus); device_disable_async_suspend(dev); + ret = snd_hdac_i915_init(bus, false); + if (ret == -EPROBE_DEFER) + goto err_unmaster; + else if (ret < 0) + dev_info(bus->dev, "i915 init unsuccessful: %d\n", ret); + schedule_work(&adev->probe_work); return 0; +err_unmaster: + pci_clear_master(pci); + pci_set_drvdata(pci, NULL); err_acquire_irq: snd_hdac_bus_free_stream_pages(bus); snd_hdac_ext_stream_free_all(bus);
Now that we can use -EPROBE_DEFER, it's no longer required to spin off the snd_hdac_i915_init into a workqueue. It's likely the whole workqueue can be destroyed, but I don't have the means to test this. Removing the workqueue would simplify init even further, but is left as exercise for the reviewer. Signed-off-by: Maarten Lankhorst <maarten.lankhorst@linux.intel.com> --- sound/soc/intel/avs/core.c | 13 +++++++++---- 1 file changed, 9 insertions(+), 4 deletions(-)