| Message ID | 1406076938-97633-1-git-send-email-libin.yang@intel.com (mailing list archive) |
|---|---|
| State | Accepted |
| Delegated to: | Takashi Iwai |
| Headers | show |
At Wed, 23 Jul 2014 08:55:37 +0800, libin.yang@intel.com wrote: > > From: Libin Yang <libin.yang@intel.com> > > Add HD Audio Device PCI ID for the Intel Braswell platform. > > Signed-off-by: Libin Yang <libin.yang@intel.com> > --- > sound/pci/hda/hda_intel.c | 3 +++ > 1 file changed, 3 insertions(+) > > diff --git a/sound/pci/hda/hda_intel.c b/sound/pci/hda/hda_intel.c > index 387fc73..f758729 100644 > --- a/sound/pci/hda/hda_intel.c > +++ b/sound/pci/hda/hda_intel.c > @@ -2017,6 +2017,9 @@ static const struct pci_device_id azx_ids[] = { > /* BayTrail */ > { PCI_DEVICE(0x8086, 0x0f04), > .driver_data = AZX_DRIVER_PCH | AZX_DCAPS_INTEL_PCH_NOPM }, > + /* Braswell */ > + { PCI_DEVICE(0x8086, 0x2284), > + .driver_data = AZX_DRIVER_PCH | AZX_DCAPS_INTEL_PCH_NOPM }, Doesn't really this new chip support runtime PM? Takashi > /* ICH */ > { PCI_DEVICE(0x8086, 0x2668), > .driver_data = AZX_DRIVER_ICH | AZX_DCAPS_OLD_SSYNC | > -- > 1.9.1 >
Hi Takashi, > -----Original Message----- > From: Takashi Iwai [mailto:tiwai@suse.de] > Sent: Wednesday, July 23, 2014 6:08 PM > To: Yang, Libin > Cc: alsa-devel@alsa-project.org; Lin, Mengdong > Subject: Re: [PATCH 1/2] ALSA: hda - add PCI IDs for Intel Braswell > > At Wed, 23 Jul 2014 08:55:37 +0800, > libin.yang@intel.com wrote: > > > > From: Libin Yang <libin.yang@intel.com> > > > > Add HD Audio Device PCI ID for the Intel Braswell platform. > > > > Signed-off-by: Libin Yang <libin.yang@intel.com> > > --- > > sound/pci/hda/hda_intel.c | 3 +++ > > 1 file changed, 3 insertions(+) > > > > diff --git a/sound/pci/hda/hda_intel.c b/sound/pci/hda/hda_intel.c > > index 387fc73..f758729 100644 > > --- a/sound/pci/hda/hda_intel.c > > +++ b/sound/pci/hda/hda_intel.c > > @@ -2017,6 +2017,9 @@ static const struct pci_device_id azx_ids[] = { > > /* BayTrail */ > > { PCI_DEVICE(0x8086, 0x0f04), > > .driver_data = AZX_DRIVER_PCH | > AZX_DCAPS_INTEL_PCH_NOPM }, > > + /* Braswell */ > > + { PCI_DEVICE(0x8086, 0x2284), > > + .driver_data = AZX_DRIVER_PCH | > AZX_DCAPS_INTEL_PCH_NOPM }, > > Doesn't really this new chip support runtime PM? Braswell is more like Baytrail, which doesn't support runtime PM. Anyway, I will test the pm-runtime on braswell and update the result later. > > > Takashi > > > /* ICH */ > > { PCI_DEVICE(0x8086, 0x2668), > > .driver_data = AZX_DRIVER_ICH | AZX_DCAPS_OLD_SSYNC | > > -- > > 1.9.1 > > Regards, Libin
Hi Takashi, After checking the HW, the HDMI audio codec doesn't support EPSS. If we force it to support runtime PM, there is risk. Besides, as the codec doesn't support EPSS, if we want to support runtime PM, simply add the AZX_DCAPS_PM_RUNTIME flag doesn't help. We must think out a method to ensure the driver enter pm runtime even the codec doesn't support. What do you think whether we can merge the patch firstly, and on the meantime, I will ask our QA team to do a full test with pm runtime enabled and decide the next step? Regards, Libin > -----Original Message----- > From: Yang, Libin > Sent: Thursday, July 24, 2014 11:57 AM > To: Takashi Iwai > Cc: alsa-devel@alsa-project.org; Lin, Mengdong > Subject: RE: [PATCH 1/2] ALSA: hda - add PCI IDs for Intel Braswell > > Hi Takashi, > > > -----Original Message----- > > From: Takashi Iwai [mailto:tiwai@suse.de] > > Sent: Wednesday, July 23, 2014 6:08 PM > > To: Yang, Libin > > Cc: alsa-devel@alsa-project.org; Lin, Mengdong > > Subject: Re: [PATCH 1/2] ALSA: hda - add PCI IDs for Intel Braswell > > > > At Wed, 23 Jul 2014 08:55:37 +0800, > > libin.yang@intel.com wrote: > > > > > > From: Libin Yang <libin.yang@intel.com> > > > > > > Add HD Audio Device PCI ID for the Intel Braswell platform. > > > > > > Signed-off-by: Libin Yang <libin.yang@intel.com> > > > --- > > > sound/pci/hda/hda_intel.c | 3 +++ > > > 1 file changed, 3 insertions(+) > > > > > > diff --git a/sound/pci/hda/hda_intel.c b/sound/pci/hda/hda_intel.c > > > index 387fc73..f758729 100644 > > > --- a/sound/pci/hda/hda_intel.c > > > +++ b/sound/pci/hda/hda_intel.c > > > @@ -2017,6 +2017,9 @@ static const struct pci_device_id azx_ids[] = { > > > /* BayTrail */ > > > { PCI_DEVICE(0x8086, 0x0f04), > > > .driver_data = AZX_DRIVER_PCH | > > AZX_DCAPS_INTEL_PCH_NOPM }, > > > + /* Braswell */ > > > + { PCI_DEVICE(0x8086, 0x2284), > > > + .driver_data = AZX_DRIVER_PCH | > > AZX_DCAPS_INTEL_PCH_NOPM }, > > > > Doesn't really this new chip support runtime PM? > > Braswell is more like Baytrail, which doesn't support runtime PM. > > Anyway, I will test the pm-runtime on braswell and update the result later. > > > > > > > Takashi > > > > > /* ICH */ > > > { PCI_DEVICE(0x8086, 0x2668), > > > .driver_data = AZX_DRIVER_ICH | AZX_DCAPS_OLD_SSYNC | > > > -- > > > 1.9.1 > > > > > Regards, > Libin
At Thu, 24 Jul 2014 12:59:55 +0000, Yang, Libin wrote: > > Hi Takashi, > > After checking the HW, the HDMI audio codec doesn't support EPSS. If we force it to support runtime PM, there is risk. Besides, as the codec doesn't support EPSS, if we want to support runtime PM, simply add the AZX_DCAPS_PM_RUNTIME flag doesn't help. We must think out a method to ensure the driver enter pm runtime even the codec doesn't support. Wait, is the controller for the generic codecs or dedicated to the embedded HDMI? In the latter case, AZX_DCAPS_INTEL_PCM_NOPM is anyway wrong. For HDMI, we use different flags. Takashi
Hi Takashi, > -----Original Message----- > From: Takashi Iwai [mailto:tiwai@suse.de] > Sent: Thursday, July 24, 2014 9:21 PM > To: Yang, Libin > Cc: 'alsa-devel@alsa-project.org'; Lin, Mengdong > Subject: Re: [PATCH 1/2] ALSA: hda - add PCI IDs for Intel Braswell > > At Thu, 24 Jul 2014 12:59:55 +0000, > Yang, Libin wrote: > > > > Hi Takashi, > > > > After checking the HW, the HDMI audio codec doesn't support EPSS. If we > force it to support runtime PM, there is risk. Besides, as the codec doesn't > support EPSS, if we want to support runtime PM, simply add the > AZX_DCAPS_PM_RUNTIME flag doesn't help. We must think out a method > to ensure the driver enter pm runtime even the codec doesn't support. > > Wait, is the controller for the generic codecs or dedicated to the embedded > HDMI? In the latter case, AZX_DCAPS_INTEL_PCM_NOPM is anyway wrong. > For HDMI, we use different flags. OK, I see. I will add the flags and have a test. > > > Takashi
diff --git a/sound/pci/hda/hda_intel.c b/sound/pci/hda/hda_intel.c index 387fc73..f758729 100644 --- a/sound/pci/hda/hda_intel.c +++ b/sound/pci/hda/hda_intel.c @@ -2017,6 +2017,9 @@ static const struct pci_device_id azx_ids[] = { /* BayTrail */ { PCI_DEVICE(0x8086, 0x0f04), .driver_data = AZX_DRIVER_PCH | AZX_DCAPS_INTEL_PCH_NOPM }, + /* Braswell */ + { PCI_DEVICE(0x8086, 0x2284), + .driver_data = AZX_DRIVER_PCH | AZX_DCAPS_INTEL_PCH_NOPM }, /* ICH */ { PCI_DEVICE(0x8086, 0x2668), .driver_data = AZX_DRIVER_ICH | AZX_DCAPS_OLD_SSYNC |