From mboxrd@z Thu Jan  1 00:00:00 1970
From: Takashi Iwai <tiwai@suse.de>
Subject: Re: [PATCH 1/2] hda - fix ELD memory leak
Date: Tue, 15 Nov 2011 15:35:41 +0100
Message-ID: <s5h4ny5fo76.wl%tiwai@suse.de>
References: <20111115143155.GA28175@localhost>
Mime-Version: 1.0 (generated by SEMI 1.14.6 - "Maruoka")
Content-Type: text/plain; charset="us-ascii"
Content-Transfer-Encoding: 7bit
Return-path: <alsa-devel-bounces@alsa-project.org>
In-Reply-To: <20111115143155.GA28175@localhost>
List-Unsubscribe: <http://mailman.alsa-project.org/mailman/listinfo/alsa-devel>,
	<mailto:alsa-devel-request@alsa-project.org?subject=unsubscribe>
List-Archive: <http://mailman.alsa-project.org/pipermail/alsa-devel>
List-Post: <mailto:alsa-devel@alsa-project.org>
List-Help: <mailto:alsa-devel-request@alsa-project.org?subject=help>
List-Subscribe: <http://mailman.alsa-project.org/mailman/listinfo/alsa-devel>,
	<mailto:alsa-devel-request@alsa-project.org?subject=subscribe>
Sender: alsa-devel-bounces@alsa-project.org
Errors-To: alsa-devel-bounces@alsa-project.org
To: Wu Fengguang <fengguang.wu@intel.com>
Cc: alsa-devel@alsa-project.org, "Wang, Zhenyu Z" <zhenyu.z.wang@intel.com>, "intel-gfx@lists.freedesktop.org" <intel-gfx@lists.freedesktop.org>, "Barnes, Jesse" <jesse.barnes@intel.com>, Jeremy Bush <contractfrombelow@gmail.com>, Christopher White <c.white@pulseforce.com>, "Bossart,
	Pierre-louis" <pierre-louis.bossart@intel.com>
List-Id: intel-gfx@lists.freedesktop.org

At Tue, 15 Nov 2011 22:31:55 +0800,
Wu Fengguang wrote:
> 
> memset(eld) clears eld->proc_entry which will leak the struct
> snd_info_entry when unloading the module.
> 
> Fix it by
> - remove memset(eld)
> - set eld->eld_valid to true _after_ all eld fields have been filled
> - don't access the other eld fields when (eld->eld_valid == false)
> 
> Signed-off-by: Wu Fengguang <fengguang.wu@intel.com>

This should be send to stable kernel, too, right?
It appears in 3.1, at least.


Takashi

> ---
>  sound/pci/hda/hda_eld.c    |    5 +----
>  sound/pci/hda/patch_hdmi.c |   11 +++++------
>  2 files changed, 6 insertions(+), 10 deletions(-)
> 
> --- linux.orig/sound/pci/hda/hda_eld.c	2011-11-15 21:02:25.000000000 +0800
> +++ linux/sound/pci/hda/hda_eld.c	2011-11-15 21:13:46.000000000 +0800
> @@ -297,10 +297,10 @@ static int hdmi_update_eld(struct hdmi_e
>  					buf + ELD_FIXED_BYTES + mnl + 3 * i);
>  	}
>  
> +	e->eld_valid = true;
>  	return 0;
>  
>  out_fail:
> -	e->eld_ver = 0;
>  	return -EINVAL;
>  }
>  
> @@ -323,9 +323,6 @@ int snd_hdmi_get_eld(struct hdmi_eld *el
>  	 * ELD is valid, actual eld_size is assigned in hdmi_update_eld()
>  	 */
>  
> -	if (!eld->eld_valid)
> -		return -ENOENT;
> -
>  	size = snd_hdmi_get_eld_size(codec, nid);
>  	if (size == 0) {
>  		/* wfg: workaround for ASUS P5E-VM HDMI board */
> --- linux.orig/sound/pci/hda/patch_hdmi.c	2011-11-15 21:02:25.000000000 +0800
> +++ linux/sound/pci/hda/patch_hdmi.c	2011-11-15 21:13:42.000000000 +0800
> @@ -980,20 +980,19 @@ static void hdmi_present_sense(struct hd
>  	 * the unsolicited response to avoid custom WARs.
>  	 */
>  	int present = snd_hda_pin_sense(codec, pin_nid);
> +	bool eld_valid = false;
>  
> -	memset(eld, 0, sizeof(*eld));
> +	eld->eld_valid = false;
>  
>  	eld->monitor_present	= !!(present & AC_PINSENSE_PRESENCE);
>  	if (eld->monitor_present)
> -		eld->eld_valid	= !!(present & AC_PINSENSE_ELDV);
> -	else
> -		eld->eld_valid	= 0;
> +		eld_valid	= !!(present & AC_PINSENSE_ELDV);
>  
>  	printk(KERN_INFO
>  		"HDMI status: Codec=%d Pin=%d Presence_Detect=%d ELD_Valid=%d\n",
> -		codec->addr, pin_nid, eld->monitor_present, eld->eld_valid);
> +		codec->addr, pin_nid, eld->monitor_present, eld_valid);
>  
> -	if (eld->eld_valid)
> +	if (eld_valid)
>  		if (!snd_hdmi_get_eld(eld, codec, pin_nid))
>  			snd_hdmi_show_eld(eld);
>  
>