[PATCH v2 0/3] ALSA: hda: add link_power op to hdac_bus_ops

[PATCH v2 0/3] ALSA: hda: add link_power op to hdac_bus_ops

[Kai Vehmanen]

Here's the 2nd iteration of the patch to improve link management
granularity between controllers using HDA extended bus and HDA
codec drivers.

Based on feedback to V1, this now adds a new link_power op.
This allows to decouple HDA codec drivers from the hdac_ext core.

I also included two patches to SOF to show how this interface is
used in SOF.

Kai Vehmanen (3):
  ALSA: hda: add link_power op to hdac_bus_ops
  ASoC: SOF: Intel: hda: use hdac_ext fine-grained link management
  ASoC: SOF: Intel: hda: release display power at link_power

 include/sound/hdaudio.h             | 14 ++++-------
 include/sound/hdaudio_ext.h         |  2 ++
 sound/hda/ext/hdac_ext_controller.c | 37 +++++++++++++++++++++++++++++
 sound/hda/hdac_bus.c                | 23 ++++++++++++++++++
 sound/hda/hdac_controller.c         | 14 +++++++++++
 sound/soc/sof/intel/hda-bus.c       | 33 ++++++++++++++++++++++++-
 sound/soc/sof/intel/hda.h           |  3 ++-
 7 files changed, 115 insertions(+), 11 deletions(-)


base-commit: 387740a5e9532db13dfb577bd4b8d1aa496487ab
-- 
2.29.2

[PATCH v2 1/3] ALSA: hda: add link_power op to hdac_bus_ops

[Kai Vehmanen]

The extended HDA bus (hdac_ext) provides interfaces for more
fine-grained control of individual links than what plain HDA
provides for. Links can be powered off when they are not used and if
all links are released, controller can shut down the command DMA.

These interfaces are currently not used by common HDA codec drivers.
When a HDA codec is runtime suspended, it calls snd_hdac_codec_link_down(),
but there is no link to the HDA extended bus, and on controller side
the links are shut down only when all codecs are suspended.

This patch adds link_power() to hdac_bus ops. Controllers using the HDA
extended core, can use this to plug in snd_hdac_ext_bus_link_power() to
implement more fine-grained control of link power.

No change is needed for plain HDA controllers nor to existing HDA
codec drivers.

Co-developed-by: Ranjani Sridharan <ranjani.sridharan@linux.intel.com>
Signed-off-by: Ranjani Sridharan <ranjani.sridharan@linux.intel.com>
Signed-off-by: Kai Vehmanen <kai.vehmanen@linux.intel.com>
---
 include/sound/hdaudio.h             | 14 ++++-------
 include/sound/hdaudio_ext.h         |  2 ++
 sound/hda/ext/hdac_ext_controller.c | 37 +++++++++++++++++++++++++++++
 sound/hda/hdac_bus.c                | 23 ++++++++++++++++++
 sound/hda/hdac_controller.c         | 14 +++++++++++
 5 files changed, 81 insertions(+), 9 deletions(-)

diff --git a/include/sound/hdaudio.h b/include/sound/hdaudio.h
index 6eed61e6cf8a..22af68b01426 100644
--- a/include/sound/hdaudio.h
+++ b/include/sound/hdaudio.h
@@ -241,6 +241,8 @@ struct hdac_bus_ops {
 	/* get a response from the last command */
 	int (*get_response)(struct hdac_bus *bus, unsigned int addr,
 			    unsigned int *res);
+	/* notify of codec link power-up/down */
+	void (*link_power)(struct hdac_device *hdev, bool enable);
 };
 
 /*
@@ -378,15 +380,8 @@ void snd_hdac_bus_exit(struct hdac_bus *bus);
 int snd_hdac_bus_exec_verb_unlocked(struct hdac_bus *bus, unsigned int addr,
 				    unsigned int cmd, unsigned int *res);
 
-static inline void snd_hdac_codec_link_up(struct hdac_device *codec)
-{
-	set_bit(codec->addr, &codec->bus->codec_powered);
-}
-
-static inline void snd_hdac_codec_link_down(struct hdac_device *codec)
-{
-	clear_bit(codec->addr, &codec->bus->codec_powered);
-}
+void snd_hdac_codec_link_up(struct hdac_device *codec);
+void snd_hdac_codec_link_down(struct hdac_device *codec);
 
 int snd_hdac_bus_send_cmd(struct hdac_bus *bus, unsigned int val);
 int snd_hdac_bus_get_response(struct hdac_bus *bus, unsigned int addr,
@@ -400,6 +395,7 @@ void snd_hdac_bus_stop_cmd_io(struct hdac_bus *bus);
 void snd_hdac_bus_enter_link_reset(struct hdac_bus *bus);
 void snd_hdac_bus_exit_link_reset(struct hdac_bus *bus);
 int snd_hdac_bus_reset_link(struct hdac_bus *bus, bool full_reset);
+void snd_hdac_bus_link_power(struct hdac_device *hdev, bool enable);
 
 void snd_hdac_bus_update_rirb(struct hdac_bus *bus);
 int snd_hdac_bus_handle_stream_irq(struct hdac_bus *bus, unsigned int status,
diff --git a/include/sound/hdaudio_ext.h b/include/sound/hdaudio_ext.h
index 7abf74c1c474..a125e3814b58 100644
--- a/include/sound/hdaudio_ext.h
+++ b/include/sound/hdaudio_ext.h
@@ -131,6 +131,8 @@ void snd_hdac_ext_link_clear_stream_id(struct hdac_ext_link *link,
 int snd_hdac_ext_bus_link_get(struct hdac_bus *bus, struct hdac_ext_link *link);
 int snd_hdac_ext_bus_link_put(struct hdac_bus *bus, struct hdac_ext_link *link);
 
+void snd_hdac_ext_bus_link_power(struct hdac_device *codec, bool enable);
+
 /* update register macro */
 #define snd_hdac_updatel(addr, reg, mask, val)		\
 	writel(((readl(addr + reg) & ~(mask)) | (val)), \
diff --git a/sound/hda/ext/hdac_ext_controller.c b/sound/hda/ext/hdac_ext_controller.c
index b0c0ef824d7d..a9bd39b93697 100644
--- a/sound/hda/ext/hdac_ext_controller.c
+++ b/sound/hda/ext/hdac_ext_controller.c
@@ -332,3 +332,40 @@ int snd_hdac_ext_bus_link_put(struct hdac_bus *bus,
 	return ret;
 }
 EXPORT_SYMBOL_GPL(snd_hdac_ext_bus_link_put);
+
+static void hdac_ext_codec_link_up(struct hdac_device *codec)
+{
+	const char *devname = dev_name(&codec->dev);
+	struct hdac_ext_link *hlink =
+		snd_hdac_ext_bus_get_link(codec->bus, devname);
+
+	if (hlink)
+		snd_hdac_ext_bus_link_get(codec->bus, hlink);
+}
+
+static void hdac_ext_codec_link_down(struct hdac_device *codec)
+{
+	const char *devname = dev_name(&codec->dev);
+	struct hdac_ext_link *hlink =
+		snd_hdac_ext_bus_get_link(codec->bus, devname);
+
+	if (hlink)
+		snd_hdac_ext_bus_link_put(codec->bus, hlink);
+}
+
+void snd_hdac_ext_bus_link_power(struct hdac_device *codec, bool enable)
+{
+	struct hdac_bus *bus = codec->bus;
+	bool oldstate = test_bit(codec->addr, &bus->codec_powered);
+
+	if (enable == oldstate)
+		return;
+
+	snd_hdac_bus_link_power(codec, enable);
+
+	if (enable)
+		hdac_ext_codec_link_up(codec);
+	else
+		hdac_ext_codec_link_down(codec);
+}
+EXPORT_SYMBOL_GPL(snd_hdac_ext_bus_link_power);
diff --git a/sound/hda/hdac_bus.c b/sound/hda/hdac_bus.c
index 9766f6af8743..71db8592b33d 100644
--- a/sound/hda/hdac_bus.c
+++ b/sound/hda/hdac_bus.c
@@ -17,6 +17,7 @@ static void snd_hdac_bus_process_unsol_events(struct work_struct *work);
 static const struct hdac_bus_ops default_ops = {
 	.command = snd_hdac_bus_send_cmd,
 	.get_response = snd_hdac_bus_get_response,
+	.link_power = snd_hdac_bus_link_power,
 };
 
 /**
@@ -264,3 +265,25 @@ void snd_hdac_aligned_write(unsigned int val, void __iomem *addr,
 }
 EXPORT_SYMBOL_GPL(snd_hdac_aligned_write);
 #endif /* CONFIG_SND_HDA_ALIGNED_MMIO */
+
+void snd_hdac_codec_link_up(struct hdac_device *codec)
+{
+	struct hdac_bus *bus = codec->bus;
+
+	if (bus->ops->link_power)
+		bus->ops->link_power(codec, true);
+	else
+		snd_hdac_bus_link_power(codec, true);
+}
+EXPORT_SYMBOL_GPL(snd_hdac_codec_link_up);
+
+void snd_hdac_codec_link_down(struct hdac_device *codec)
+{
+	struct hdac_bus *bus = codec->bus;
+
+	if (bus->ops->link_power)
+		bus->ops->link_power(codec, false);
+	else
+		snd_hdac_bus_link_power(codec, false);
+}
+EXPORT_SYMBOL_GPL(snd_hdac_codec_link_down);
diff --git a/sound/hda/hdac_controller.c b/sound/hda/hdac_controller.c
index b98449fd92f3..062da7a7a586 100644
--- a/sound/hda/hdac_controller.c
+++ b/sound/hda/hdac_controller.c
@@ -648,3 +648,17 @@ void snd_hdac_bus_free_stream_pages(struct hdac_bus *bus)
 		snd_dma_free_pages(&bus->posbuf);
 }
 EXPORT_SYMBOL_GPL(snd_hdac_bus_free_stream_pages);
+
+/**
+ * snd_hdac_bus_link_power - power up/down codec link
+ * @codec: HD-audio device
+ * @enable: whether to power-up the link
+ */
+void snd_hdac_bus_link_power(struct hdac_device *codec, bool enable)
+{
+	if (enable)
+		set_bit(codec->addr, &codec->bus->codec_powered);
+	else
+		clear_bit(codec->addr, &codec->bus->codec_powered);
+}
+EXPORT_SYMBOL_GPL(snd_hdac_bus_link_power);
-- 
2.29.2

[PATCH v2 2/3] ASoC: SOF: Intel: hda: use hdac_ext fine-grained link management

[Kai Vehmanen]

Opt-in to use snd_hdac_ext_bus_link_power() to manage HDA link
power up/down events.

This allows to reduce power consumption in cases where some HDA codecs
are suspended, but other child devices (HDA or non-HDA codecs) remain
active and controller itself remains in active state.

By using snd_hdac_ext_bus_link_power(), the individual HDA links can be
powered off and if all HDA codecs are powered down, the command DMA can
also be shut down.

Signed-off-by: Kai Vehmanen <kai.vehmanen@linux.intel.com>
---
 sound/soc/sof/intel/hda-bus.c | 10 +++++++++-
 1 file changed, 9 insertions(+), 1 deletion(-)

diff --git a/sound/soc/sof/intel/hda-bus.c b/sound/soc/sof/intel/hda-bus.c
index 789148e5584b..1ac6e79d7e62 100644
--- a/sound/soc/sof/intel/hda-bus.c
+++ b/sound/soc/sof/intel/hda-bus.c
@@ -19,13 +19,21 @@
 #define sof_hda_ext_ops	NULL
 #endif
 
+#if IS_ENABLED(CONFIG_SND_SOC_SOF_HDA)
+static const struct hdac_bus_ops bus_core_ops = {
+	.command = snd_hdac_bus_send_cmd,
+	.get_response = snd_hdac_bus_get_response,
+	.link_power = snd_hdac_ext_bus_link_power,
+};
+#endif
+
 /*
  * This can be used for both with/without hda link support.
  */
 void sof_hda_bus_init(struct hdac_bus *bus, struct device *dev)
 {
 #if IS_ENABLED(CONFIG_SND_SOC_SOF_HDA)
-	snd_hdac_ext_bus_init(bus, dev, NULL, sof_hda_ext_ops);
+	snd_hdac_ext_bus_init(bus, dev, &bus_core_ops, sof_hda_ext_ops);
 #else /* CONFIG_SND_SOC_SOF_HDA */
 	memset(bus, 0, sizeof(*bus));
 	bus->dev = dev;
-- 
2.29.2

[PATCH v2 3/3] ASoC: SOF: Intel: hda: release display power at link_power

[Kai Vehmanen]

The i915 display power is requested both by controller (for init
and link reset) as well as by codec driver (for codec control).

There's an additional constraint that on some platforms frequent changes
to display power state may cause visible flicker. To avoid this, the SOF
hda controller requests display power whenever it is active and only
releases it when runtime suspended.

This patch utilizes the new hdac_bus link_power op to plug
into HDA link state changes. By monitoring link state changes,
we can keep the controller side display power wakeref until
the codec driver has completed its work, and only release
the wakeref when codec driver is suspended.

Signed-off-by: Kai Vehmanen <kai.vehmanen@linux.intel.com>
---
 sound/soc/sof/intel/hda-bus.c | 25 ++++++++++++++++++++++++-
 sound/soc/sof/intel/hda.h     |  3 ++-
 2 files changed, 26 insertions(+), 2 deletions(-)

diff --git a/sound/soc/sof/intel/hda-bus.c b/sound/soc/sof/intel/hda-bus.c
index 1ac6e79d7e62..30025d3c16b6 100644
--- a/sound/soc/sof/intel/hda-bus.c
+++ b/sound/soc/sof/intel/hda-bus.c
@@ -9,6 +9,7 @@
 
 #include <linux/io.h>
 #include <sound/hdaudio.h>
+#include <sound/hda_i915.h>
 #include "../sof-priv.h"
 #include "hda.h"
 
@@ -20,10 +21,32 @@
 #endif
 
 #if IS_ENABLED(CONFIG_SND_SOC_SOF_HDA)
+static void sof_hda_bus_link_power(struct hdac_device *codec, bool enable)
+{
+	struct hdac_bus *bus = codec->bus;
+	bool oldstate = test_bit(codec->addr, &bus->codec_powered);
+
+	snd_hdac_ext_bus_link_power(codec, enable);
+
+	if (enable == oldstate)
+		return;
+
+	/*
+	 * Both codec driver and controller can hold references to
+	 * display power. To avoid unnecessary power-up/down cycles,
+	 * controller doesn't immediately release its reference.
+	 *
+	 * If the codec driver powers down the link, release
+	 * the controller reference as well.
+	 */
+	if (codec->addr == HDA_IDISP_ADDR && !enable)
+		snd_hdac_display_power(bus, HDA_CODEC_IDX_CONTROLLER, false);
+}
+
 static const struct hdac_bus_ops bus_core_ops = {
 	.command = snd_hdac_bus_send_cmd,
 	.get_response = snd_hdac_bus_get_response,
-	.link_power = snd_hdac_ext_bus_link_power,
+	.link_power = sof_hda_bus_link_power,
 };
 #endif
 
diff --git a/sound/soc/sof/intel/hda.h b/sound/soc/sof/intel/hda.h
index a3b6f3e9121c..1d9b38e6ed40 100644
--- a/sound/soc/sof/intel/hda.h
+++ b/sound/soc/sof/intel/hda.h
@@ -388,7 +388,8 @@
 #define SSP_SET_SFRM_SLAVE	BIT(24)
 #define SSP_SET_SLAVE		(SSP_SET_SCLK_SLAVE | SSP_SET_SFRM_SLAVE)
 
-#define HDA_IDISP_CODEC(x) ((x) & BIT(2))
+#define HDA_IDISP_ADDR		2
+#define HDA_IDISP_CODEC(x) ((x) & BIT(HDA_IDISP_ADDR))
 
 struct sof_intel_dsp_bdl {
 	__le32 addr_l;
-- 
2.29.2

Re: [PATCH v2 0/3] ALSA: hda: add link_power op to hdac_bus_ops

[Takashi Iwai]

On Fri, 05 Feb 2021 19:46:27 +0100,
Kai Vehmanen wrote:
> 
> Here's the 2nd iteration of the patch to improve link management
> granularity between controllers using HDA extended bus and HDA
> codec drivers.
> 
> Based on feedback to V1, this now adds a new link_power op.
> This allows to decouple HDA codec drivers from the hdac_ext core.
> 
> I also included two patches to SOF to show how this interface is
> used in SOF.
> 
> Kai Vehmanen (3):
>   ALSA: hda: add link_power op to hdac_bus_ops
>   ASoC: SOF: Intel: hda: use hdac_ext fine-grained link management
>   ASoC: SOF: Intel: hda: release display power at link_power

The patch set looks good to me.  If Mark can give an ack, I'll merge
them through my tree later.


thanks,

Takashi

Re: [PATCH v2 0/3] ALSA: hda: add link_power op to hdac_bus_ops

[Mark Brown]

[-- Attachment #1: Type: text/plain, Size: 467 bytes --]

On Sat, Feb 06, 2021 at 01:24:38PM +0100, Takashi Iwai wrote:
> Kai Vehmanen wrote:

> > Kai Vehmanen (3):
> >   ALSA: hda: add link_power op to hdac_bus_ops
> >   ASoC: SOF: Intel: hda: use hdac_ext fine-grained link management
> >   ASoC: SOF: Intel: hda: release display power at link_power

> The patch set looks good to me.  If Mark can give an ack, I'll merge
> them through my tree later.

Yeah, looks good to me too

Acked-by: Mark Brown <broonie@kernel.org>

[-- Attachment #2: signature.asc --]
[-- Type: application/pgp-signature, Size: 488 bytes --]

Re: [PATCH v2 0/3] ALSA: hda: add link_power op to hdac_bus_ops

[Takashi Iwai]

On Mon, 08 Feb 2021 12:11:20 +0100,
Mark Brown wrote:
> 
> On Sat, Feb 06, 2021 at 01:24:38PM +0100, Takashi Iwai wrote:
> > Kai Vehmanen wrote:
> 
> > > Kai Vehmanen (3):
> > >   ALSA: hda: add link_power op to hdac_bus_ops
> > >   ASoC: SOF: Intel: hda: use hdac_ext fine-grained link management
> > >   ASoC: SOF: Intel: hda: release display power at link_power
> 
> > The patch set looks good to me.  If Mark can give an ack, I'll merge
> > them through my tree later.
> 
> Yeah, looks good to me too
> 
> Acked-by: Mark Brown <broonie@kernel.org>

OK, thanks.  Now I applied all three patches to for-next branch.


Takashi