Re: [alsa-devel] [PATCH 2/2] ALSA: hda - Add power-welll support for haswell HDA

16 May 2013

At Thu, 16 May 2013 03:51:16 +0000,
Wang, Xingchao wrote:
...
Hi Takashi,
Thanks your quick feedback, please see my comments below.
...
-----Original Message-----
From: Takashi Iwai [mailto:tiwai@suse.de]
Sent: Tuesday, May 14, 2013 8:15 PM
To: Wang Xingchao
Cc: daniel@ffwll.ch; Girdwood, Liam R; david.henningsson@canonical.com; Lin,
Mengdong; Li, Jocelyn; alsa-devel@alsa-project.org;
intel-gfx@lists.freedesktop.org; Barnes, Jesse; Zanoni, Paulo R; Wang,
Xingchao
Subject: Re: [PATCH 2/2] ALSA: hda - Add power-welll support for haswell HDA
At Tue, 14 May 2013 19:44:19 +0800,
Wang Xingchao wrote:
...
For Intel Haswell chip, HDA controller and codec have power well
dependency from GPU side. This patch added support to request/release
power well in audio driver. Power save feature should be enabled to
get runtime power saving.
Signed-off-by: Wang Xingchao xingchao.wang@linux.intel.com
sound/pci/hda/Kconfig     |    8 ++++
 sound/pci/hda/Makefile    |    3 ++
 sound/pci/hda/hda_i915.c  |   92
+++++++++++++++++++++++++++++++++++++++++++++
...
sound/pci/hda/hda_i915.h  |   35 +++++++++++++++++
 sound/pci/hda/hda_intel.c |   33 ++++++++++++++--
 5 files changed, 168 insertions(+), 3 deletions(-)  create mode
100644 sound/pci/hda/hda_i915.c  create mode 100644
sound/pci/hda/hda_i915.h

diff --git a/sound/pci/hda/Kconfig b/sound/pci/hda/Kconfig index
80a7d44..d9e71e4 100644
--- a/sound/pci/hda/Kconfig
+++ b/sound/pci/hda/Kconfig
@@ -152,6 +152,14 @@ config SND_HDA_CODEC_HDMI
     snd-hda-codec-hdmi.
     This module is automatically loaded at probing.
+config SND_HDA_I915

bool "Build Display HD-audio controller/codec power well support for i915

cards"
...

depends on DRM_I915
default y
help
 Say Y here to include full HDMI and DisplayPort HD-audio



controller/codec
...

 support for Intel Haswell graphics cards based on the i915 driver.



Do we need to make this selectable?
If you have i915 driver, this can be always set.
Well, for Non-Haswell chips with SND_HDA_I915 compiled in, it do nothing as it has no " AZX_DCAPS_I915_POWERWELL" set, so no power-well API called.
For Haswell chips, SND_HDA_I915 must be enabled otherwise display audio would fail. Anyway I added a note message in the choice to let
user know this option is a *must* one for Haswell chips.
Yes, but the problem is if the driver without CONFIG_SND_HDA_I915 is
used on Haswell.  Then the driver is loaded as if everything is OK
even though you know it's broken.
If you make this selectable, you should give an error from
hda_i915_init() in the case CONFIG_SND_HDA_i915=n, so that the driver
load shall fail.
...
...
...
config SND_HDA_CODEC_CIRRUS
   bool "Build Cirrus Logic codec support"
   default y
diff --git a/sound/pci/hda/Makefile b/sound/pci/hda/Makefile index
24a2514..4b0a4bc 100644
--- a/sound/pci/hda/Makefile
+++ b/sound/pci/hda/Makefile
@@ -6,6 +6,9 @@ snd-hda-codec-$(CONFIG_PROC_FS) += hda_proc.o
 snd-hda-codec-$(CONFIG_SND_HDA_HWDEP) += hda_hwdep.o
 snd-hda-codec-$(CONFIG_SND_HDA_INPUT_BEEP) += hda_beep.o
+# for haswell power well
+snd-hda-intel-$(CONFIG_SND_HDA_I915) +=	hda_i915.o



# for trace-points
 CFLAGS_hda_codec.o := -I$(src)
 CFLAGS_hda_intel.o := -I$(src)
diff --git a/sound/pci/hda/hda_i915.c b/sound/pci/hda/hda_i915.c new
file mode 100644 index 0000000..96ca9e7
--- /dev/null
+++ b/sound/pci/hda/hda_i915.c
@@ -0,0 +1,92 @@
+/*


hda_i915.c - routines for Haswell HDA controller power well



+support






This program is free software; you can redistribute it and/or



+modify it


under the terms of the GNU General Public License as published by



+the Free


Software Foundation; either version 2 of the License, or (at your



+option)


any later version.







This program is distributed in the hope that it will be useful,



+but


WITHOUT ANY WARRANTY; without even the implied warranty of



+MERCHANTABILITY


or FITNESS FOR A PARTICULAR PURPOSE.  See the GNU General Public



+License


for more details.







You should have received a copy of the GNU General Public License



along with this program; if not, write to the Free Software



+Foundation,


Inc., 59 Temple Place - Suite 330, Boston, MA  02111-1307, USA.


*/


+#include <linux/init.h>
+#include <linux/module.h>
+#include <sound/core.h>
+#include <drm/i915_powerwell.h>
+#include "hda_i915.h"



+static const char *module_name = "i915"; static struct module *i915;
+static void (*get_power)(void); static void (*put_power)(void);



+void hda_display_power(bool enable)
+{

if (!get_power || !put_power)
return;



snd_printk(KERN_DEBUG "HDA display power %s \n",
	enable ? "Enable" : "Disable");



Use snd_printdd() for such a frequent debug message.
We don't want to see this message at each time you open/close the audio
stream.
Fixed.
...
...

if (enable)
get_power();


else
put_power();



+}



+/* in case i915 not loaded, try load i915 first */ static int
+load_i915(void) {

int err;
mutex_lock(&module_mutex);
i915 = find_module(module_name);
mutex_unlock(&module_mutex);

if (!i915) {
err = request_module(module_name);


if (err < 0)


	return -ENODEV;



Why not passing the returned error code?
Fixed.
...
Also, you need to get i915 module again after the successful request_module().
request_module() doesn't give the module reference.
Symbol_get/put will call try_module_get() internally so it think it's not necessary get module reference again.
So in above case, if request_module() finished, we can call symbol_get/put safely, Do you think so?
Looking back at the code, it looks so.  But you can reduce these open
codes via simple symbol_request().
...
...
Also, we can skip the whole load_i915() if i915 driver is built-in, i.e.
#ifdef CONFIG_DRM_i915_MODULE
static int load_i915(void)
{
   ....
}
#else
#define load_i915(void)		(-ENODEV)
#endif
Is it really necessary here? I assume if the module is live, symbol_get() would be successful.
If the symbol_get fail, the module doesnot exist. Is there any other case? i.e. module is live but symbol_get fail?
You already know that i915 is never a module but only built-in in the
above.  Thus request_module() and retry is simply superfluous.
...
...
...

}

return 0;

+}



+int hda_i915_init(void)
+{

get_power = symbol_get(i915_request_power_well);
if (!get_power) {
snd_printk(KERN_WARNING "hda-i915: get symbol fail, try



again!\n");
...



Doesn't make sense to give a warning at this point.  The fact that no symbol is
taken at this point is no real error.  The error should be reported when the
module load failed or the symbol_get() failed twice.
Yes, fixed in next version.
...
...

if (load_i915() < 0)


	return -ENODEV;



get_power = symbol_get(i915_request_power_well);



if (!get_power)


	return -ENODEV;



The i915 module should be unreferenced in the error path.
...

}

put_power = symbol_get(i915_release_power_well);
if (!put_power)
return -ENODEV;



Ditto.
Yeah, please correct me if my understanding was wrong: 
Symbol_get should be safe here as it would request module reference.
Symbol_put is needed in exit() case, but no module_put() needed.
Right, I thought wrongly.  So you just need to call symbol_put() for
the leftover ones.
Takashi
...
if we donot use symbol finding way, we should use try_module_get() and module_put() here.
...
...

snd_printk(KERN_INFO "HDA driver get symbol successfully from i915

+module\n");
Use snd_printd() as a debug message, if any.
Fixed.
...
...


return 0;

+}



+int hda_i915_exit(void)
+{

if (get_power)
symbol_put(i915_request_power_well);


if (put_power)
symbol_put(i915_release_power_well);



i915 module should be unreferenced at exit, too.
...


return 0;

+}
diff --git a/sound/pci/hda/hda_i915.h b/sound/pci/hda/hda_i915.h new
file mode 100644 index 0000000..cad4971
--- /dev/null
+++ b/sound/pci/hda/hda_i915.h
@@ -0,0 +1,35 @@
+/*


This program is free software; you can redistribute it and/or



+modify it


under the terms of the GNU General Public License as published by



+the Free


Software Foundation; either version 2 of the License, or (at your



+option)


any later version.







This program is distributed in the hope that it will be useful,



+but WITHOUT


ANY WARRANTY; without even the implied warranty of



+MERCHANTABILITY or


FITNESS FOR A PARTICULAR PURPOSE.  See the GNU General Public



+License for


more details.







You should have received a copy of the GNU General Public License



+along with


this program; if not, write to the Free Software Foundation,



+Inc., 59


Temple Place - Suite 330, Boston, MA  02111-1307, USA.


*/

+#ifndef __SOUND_HDA_I915_H
+#define __SOUND_HDA_I915_H



+#ifdef CONFIG_SND_HDA_I915
+void hda_display_power(bool enable);
+int hda_i915_init(void);
+int hda_i915_exit(void);
+#else
+static inline void hda_display_power(bool enable) {} static inline
+int hda_i915_init(void) {

return 0;

+}
+static inline int hda_i915_exit(void) {

return 0;

+}
+#endif



+#endif
diff --git a/sound/pci/hda/hda_intel.c b/sound/pci/hda/hda_intel.c
index 418bfc0..6854a7e 100644
--- a/sound/pci/hda/hda_intel.c
+++ b/sound/pci/hda/hda_intel.c
@@ -62,6 +62,7 @@
 #include <linux/vga_switcheroo.h>
 #include <linux/firmware.h>
 #include "hda_codec.h"
+#include "hda_i915.h"
static int index[SNDRV_CARDS] = SNDRV_DEFAULT_IDX; @@ -594,6 +595,7
@@ enum {
 #define AZX_DCAPS_4K_BDLE_BOUNDARY (1 << 23)	/* BDLE in 4k
boundary */
...
#define AZX_DCAPS_COUNT_LPIB_DELAY  (1 << 25)	/* Take LPIB as
delay */
...
#define AZX_DCAPS_PM_RUNTIME	(1 << 26)	/* runtime PM support */
+#define AZX_DCAPS_I915_POWERWELL (1 << 27)	/* HSW i915 power well
support */
...
/* quirks for Intel PCH */
 #define AZX_DCAPS_INTEL_PCH_NOPM \
@@ -2869,6 +2871,8 @@ static int azx_suspend(struct device *dev)
   pci_disable_device(pci);
   pci_save_state(pci);
   pci_set_power_state(pci, PCI_D3hot);

if (chip->driver_caps & AZX_DCAPS_I915_POWERWELL)
hda_display_power(false);

return 0;

}
@@ -2881,6 +2885,8 @@ static int azx_resume(struct device *dev)
   if (chip->disabled)
   	return 0;

if (chip->driver_caps & AZX_DCAPS_I915_POWERWELL)
hda_display_power(true);

pci_set_power_state(pci, PCI_D0);
pci_restore_state(pci);
if (pci_enable_device(pci) < 0) {

@@ -2913,6 +2919,8 @@ static int azx_runtime_suspend(struct device
*dev)
azx_stop_chip(chip);
   azx_clear_irq_pending(chip);

if (chip->driver_caps & AZX_DCAPS_I915_POWERWELL)
hda_display_power(false);

return 0;

}
@@ -2921,6 +2929,8 @@ static int azx_runtime_resume(struct device *dev)
   struct snd_card *card = dev_get_drvdata(dev);
   struct azx *chip = card->private_data;

if (chip->driver_caps & AZX_DCAPS_I915_POWERWELL)
hda_display_power(true);

azx_init_pci(chip);
azx_init_chip(chip, 1);
return 0;

@@ -3700,6 +3710,15 @@ static int azx_probe(struct pci_dev *pci,
   	chip->disabled = true;
   }

if (chip->driver_caps & AZX_DCAPS_I915_POWERWELL) {
err = hda_i915_init();


if (err < 0) {


	snd_printk(KERN_ERR "Error request power-well from i915\n");


	goto out_free;


}


hda_display_power(true);


}
probe_now = !chip->disabled;
 if (probe_now) {
  err = azx_first_init(chip);

Missing hda_display_power(false) and hda_i915_exit() in the error path.
Okay, fixed.
...
...
@@ -3799,6 +3818,7 @@ out_free:
 static void azx_remove(struct pci_dev *pci)  {
   struct snd_card *card = pci_get_drvdata(pci);

struct azx *chip = card->private_data;
if (pci_dev_run_wake(pci))
pm_runtime_get_noresume(&pci->dev);


@@ -3806,6 +3826,10 @@ static void azx_remove(struct pci_dev *pci)
   if (card)
   	snd_card_free(card);
   pci_set_drvdata(pci, NULL);

if (chip->driver_caps & AZX_DCAPS_I915_POWERWELL) {
hda_display_power(false);


hda_i915_exit();


}

}
/* PCI IDs */
@@ -3835,11 +3859,14 @@ static DEFINE_PCI_DEVICE_TABLE(azx_ids) = {
     .driver_data = AZX_DRIVER_PCH | AZX_DCAPS_INTEL_PCH },
   /* Haswell */
   { PCI_DEVICE(0x8086, 0x0a0c),

 .driver_data = AZX_DRIVER_SCH | AZX_DCAPS_INTEL_PCH },




 .driver_data = AZX_DRIVER_SCH | AZX_DCAPS_INTEL_PCH |


 AZX_DCAPS_I915_POWERWELL},

{ PCI_DEVICE(0x8086, 0x0c0c),


 .driver_data = AZX_DRIVER_SCH | AZX_DCAPS_INTEL_PCH },




 .driver_data = AZX_DRIVER_SCH | AZX_DCAPS_INTEL_PCH |


 AZX_DCAPS_I915_POWERWELL},

{ PCI_DEVICE(0x8086, 0x0d0c),


 .driver_data = AZX_DRIVER_SCH | AZX_DCAPS_INTEL_PCH },




 .driver_data = AZX_DRIVER_SCH | AZX_DCAPS_INTEL_PCH |


 AZX_DCAPS_I915_POWERWELL},



Try to follow the indentation / spacing of the original code.
Fixed too.
Thanks
--xingchao

    