Date: Sun, 08 Jun 2008 09:23:57 +0200
Message-ID: <s5hwsl02xv6.wl%tiwai@suse.de>
From: Takashi Iwai <tiwai@suse.de>
To: Thadeu Lima de Souza Cascardo <cascardo@minaslivre.org>
Cc: linux-kernel@vger.kernel.org
Subject: Re: [PATCH] Correct type and description of	CONFIG_AC97_POWER_SAVE_DEFAULT
In-Reply-To: <20080607194006.GA9516@vespa.holoscopio.com>
References: <20080607162238.GA7893@vespa.holoscopio.com>
	<s5hlk1hp25b.wl%tiwai@suse.de>
	<20080607194006.GA9516@vespa.holoscopio.com>
User-Agent: Wanderlust/2.12.0 (Your Wildest Dreams) SEMI/1.14.6 (Maruoka)
 FLIM/1.14.7 (=?ISO-8859-4?Q?Sanj=F2?=) APEL/10.6 MULE XEmacs/21.5 (beta28)
 (fuki) (x86_64-suse-linux)
MIME-Version: 1.0 (generated by SEMI 1.14.6 - "Maruoka")
Content-Type: text/plain; charset=US-ASCII
Content-Transfer-Encoding: 8BIT
Sender: linux-kernel-owner@vger.kernel.org
Content-Length: 3287
Lines: 86

At Sat, 7 Jun 2008 16:40:07 -0300,
Thadeu Lima de Souza Cascardo wrote:
> 
> On Sat, Jun 07, 2008 at 07:48:16PM +0200, Takashi Iwai wrote:
> > At Sat, 7 Jun 2008 13:22:39 -0300,
> > Thadeu Lima de Souza Cascardo wrote:
> > > 
> > > While CONFIG_HDA_POWER_SAVE_DEFAULT is used as a timeout in seconds,
> > > CONFIG_AC97_POWER_SAVE_DEFAULT simply enables or disables AC97 power
> > > saving.
> > 
> > Thanks for the patch.  However, I can't take this as is.  The reasons
> > are below:
> > 
> > The power_save option takes indeed an integer value, and this kconfig
> > is nothing but its default value.
> > 
> > In your way, it's set always 1 if selected.  This is a very bad choice
> > as power_save value, because you'll turn on/off after one second.
> > This may lead to too frequent click noises. 
> > Thus, even if we need to make it bool, a more sensitive value must be
> > chosen.  And, which value is sensitive is a matter of taste, and you
> > cannot define it alone by yourself.
> > 
> 
> As I said in my comment, that is true for CONFIG_HDA_POWER_SAVE_DEFAULT,
> which is, in fact, used as a number of seconds, and I left that as is.
> The static power_save variable in ac97_codec.c, however, is only used in
> a macro ac97_is_power_save_mode, which, in turn, is only used in two
> if's.

Ah, right.  It's a simply because I forgot to commit the patch
mistakenly.  The power_save option must take integer as its timeout 
value, just like for snd-hda-intel driver, as described in
Documentation/sound/alsa/powersave.txt.

The patch below is already applied to my git tree, thus for the next
release (will be included in linux-next soon).


thanks,

Takashi


>From 92a0a8d2fce86124451bc6d1c1e2d9ce3805df2d Mon Sep 17 00:00:00 2001
From: Takashi Iwai <tiwai@suse.de>
Date: Sun, 8 Jun 2008 09:17:27 +0200
Subject: [ALSA] ac97 - Fix power_save option value as time-out

The power_save option was set as boot although it was meant to be a
timeout value like the same option of snd-hda-intel originally.
Now fixed to the same style.

Signed-off-by: Takashi Iwai <tiwai@suse.de>

diff --git a/sound/pci/ac97/ac97_codec.c b/sound/pci/ac97/ac97_codec.c
index 45fd290..2d2f16e 100644
--- a/sound/pci/ac97/ac97_codec.c
+++ b/sound/pci/ac97/ac97_codec.c
@@ -49,8 +49,9 @@ MODULE_PARM_DESC(enable_loopback, "Enable AC97 ADC/DAC Loopback Control");
 
 #ifdef CONFIG_SND_AC97_POWER_SAVE
 static int power_save = CONFIG_SND_AC97_POWER_SAVE_DEFAULT;
-module_param(power_save, bool, 0644);
-MODULE_PARM_DESC(power_save, "Enable AC97 power-saving control");
+module_param(power_save, int, 0644);
+MODULE_PARM_DESC(power_save, "Automatic power-saving timeout "
+		 "(in second, 0 = disable).");
 #endif
 /*
 
@@ -2362,7 +2363,7 @@ int snd_ac97_update_power(struct snd_ac97 *ac97, int reg, int powerup)
 		 *  that open/close frequently)
 		 */
 		schedule_delayed_work(&ac97->power_work,
-				      msecs_to_jiffies(2000));
+				      msecs_to_jiffies(power_save * 1000));
 	else {
 		cancel_delayed_work(&ac97->power_work);
 		update_power_regs(ac97);
--
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@vger.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html
Please read the FAQ at  http://www.tux.org/lkml/