LKML Archive on lore.kernel.org
help / color / mirror / Atom feed
From: Dan Carpenter <dan.carpenter@oracle.com>
To: Kirill Marinushkin <k.marinushkin@gmail.com>
Cc: Greg Kroah-Hartman <gregkh@linuxfoundation.org>,
	Andy Shevchenko <andy.shevchenko@gmail.com>,
	Eric Anholt <eric@anholt.net>,
	Stefan Wahren <stefan.wahren@i2se.com>,
	Florian Fainelli <f.fainelli@gmail.com>,
	Ray Jui <rjui@broadcom.com>,
	Scott Branden <sbranden@broadcom.com>,
	devel@driverdev.osuosl.org, linux-kernel@vger.kernel.org,
	bcm-kernel-feedback-list@broadcom.com,
	linux-rpi-kernel@lists.infradead.org,
	linux-arm-kernel@lists.infradead.org
Subject: Re: [RESEND PATCH] staging: bcm2835-audio: Disconnect and free vchi_instance on module_exit()
Date: Tue, 24 Apr 2018 14:50:52 +0300	[thread overview]
Message-ID: <20180424115052.pqdblbts2ko36njh@mwanda> (raw)
In-Reply-To: <20180424074459.7215-1-k.marinushkin@gmail.com>

On Tue, Apr 24, 2018 at 09:44:59AM +0200, Kirill Marinushkin wrote:
> In the current implementation, vchi_instance is inited during the first
> call of bcm2835_audio_open_connection(), and is never freed. It causes a
> memory leak when the module `snd_bcm2835` is removed.
> 
> Here is how this commit fixes it:
> 
> * the VCHI context (including vchi_instance) is created once in the
>   platform's devres
> * the VCHI context is allocated and connected once during module_init()
> * all created bcm2835_chips have a pointer to this VCHI context
> * bcm2835_audio_open_connection() can access the VCHI context through the
>   associated bcm2835_chip
> * the VCHI context is disconnected and freed once during module_exit()
> 
> After this commit is applied, I don't see other issues with the module's
> init/exit, so I also remove the associated TODO task.
> 
> Steps to reproduce the memory leak before this commit:
> 
> ~~~~
> root@raspberrypi:/home/pi# aplay test0.wav
> Playing WAVE 'test0.wav' : Signed 16 bit Little Endian, Rate 44100 Hz, Ster
> ^CAborted by signal Interrupt...
> root@raspberrypi:/home/pi# rmmod snd_bcm2835
> root@raspberrypi:/home/pi# modprobe snd_bcm2835
> root@raspberrypi:/home/pi# aplay test0.wav
> Playing WAVE 'test0.wav' : Signed 16 bit Little Endian, Rate 44100 Hz, Ster
> ^CAborted by signal Interrupt...
> root@raspberrypi:/home/pi# echo scan > /sys/kernel/debug/kmemleak
> root@raspberrypi:/home/pi# cat /sys/kernel/debug/kmemleak
> unreferenced object 0xb6794c00 (size 128):
>   comm "aplay", pid 406, jiffies 36870 (age 116.650s)
>   hex dump (first 32 bytes):
>     08 a5 82 81 01 00 00 00 08 4c 79 b6 08 4c 79 b6  .........Ly..Ly.
>     00 00 00 00 00 00 00 00 ad 4e ad de ff ff ff ff  .........N......
>   backtrace:
>     [<802af5e0>] kmem_cache_alloc_trace+0x294/0x3d0
>     [<806ce620>] vchiq_initialise+0x98/0x1b0
>     [<806d0b34>] vchi_initialise+0x24/0x34
>     [<7f1311ec>] 0x7f1311ec
>     [<7f1303bc>] 0x7f1303bc
>     [<7f130590>] 0x7f130590
>     [<7f111fd8>] snd_pcm_open_substream+0x68/0xc4 [snd_pcm]
>     [<7f112108>] snd_pcm_open+0xd4/0x248 [snd_pcm]
>     [<7f112334>] snd_pcm_playback_open+0x4c/0x6c [snd_pcm]
>     [<7f0e250c>] snd_open+0xa8/0x14c [snd]
>     [<802ce590>] chrdev_open+0xac/0x188
>     [<802c57b4>] do_dentry_open+0x10c/0x314
>     [<802c6ba8>] vfs_open+0x5c/0x88
>     [<802d9a68>] path_openat+0x368/0x944
>     [<802dacd4>] do_filp_open+0x70/0xc4
>     [<802c6f70>] do_sys_open+0x110/0x1d4
> ~~~~
> 
> Signed-off-by: Kirill Marinushkin <k.marinushkin@gmail.com>
> Cc: Eric Anholt <eric@anholt.net>
> Cc: Stefan Wahren <stefan.wahren@i2se.com>
> Cc: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
> Cc: Florian Fainelli <f.fainelli@gmail.com>
> Cc: Ray Jui <rjui@broadcom.com>
> Cc: Scott Branden <sbranden@broadcom.com>
> Cc: Andy Shevchenko <andy.shevchenko@gmail.com>
> Cc: bcm-kernel-feedback-list@broadcom.com
> Cc: linux-rpi-kernel@lists.infradead.org
> Cc: linux-arm-kernel@lists.infradead.org
> Cc: devel@driverdev.osuosl.org
> Cc: linux-kernel@vger.kernel.org
> ---
>  .../vc04_services/bcm2835-audio/bcm2835-vchiq.c    | 64 +++++++++++++---------
>  .../staging/vc04_services/bcm2835-audio/bcm2835.c  | 43 ++++++++++++++-
>  .../staging/vc04_services/bcm2835-audio/bcm2835.h  | 12 ++++
>  3 files changed, 91 insertions(+), 28 deletions(-)
> 
> diff --git a/drivers/staging/vc04_services/bcm2835-audio/bcm2835-vchiq.c b/drivers/staging/vc04_services/bcm2835-audio/bcm2835-vchiq.c
> index 3c6f1d91d22d..389a18f9350a 100644
> --- a/drivers/staging/vc04_services/bcm2835-audio/bcm2835-vchiq.c
> +++ b/drivers/staging/vc04_services/bcm2835-audio/bcm2835-vchiq.c
> @@ -33,7 +33,6 @@
>  
>  /* ---- Include Files -------------------------------------------------------- */
>  
> -#include "interface/vchi/vchi.h"
>  #include "vc_vchi_audioserv_defs.h"
>  
>  /* ---- Private Constants and Types ------------------------------------------ */
> @@ -371,14 +370,46 @@ static int vc_vchi_audio_deinit(struct bcm2835_audio_instance *instance)
>  	return 0;
>  }
>  
> +int bcm2835_new_vchi_ctx(struct bcm2835_vchi_ctx *vchi_ctx)
> +{
> +	int ret;
> +
> +	/* Initialize and create a VCHI connection */
> +	ret = vchi_initialise(&vchi_ctx->vchi_instance);
> +	if (ret) {
> +		LOG_ERR("%s: failed to initialise VCHI instance (ret=%d)\n",
> +			__func__, ret);
> +
> +		return -EIO;
> +	}
> +
> +	ret = vchi_connect(NULL, 0, vchi_ctx->vchi_instance);
> +	if (ret) {
> +		LOG_ERR("%s: failed to connect VCHI instance (ret=%d)\n",
> +			__func__, ret);
> +
> +		kfree(vchi_ctx->vchi_instance);
> +		vchi_ctx->vchi_instance = NULL;
> +
> +		return -EIO;
> +	}
> +
> +	return 0;
> +}
> +
> +void bcm2835_free_vchi_ctx(struct bcm2835_vchi_ctx *vchi_ctx)
> +{
> +	/* Close the VCHI connection - it will also free vchi_instance */
> +	WARN_ON(vchi_disconnect(vchi_ctx->vchi_instance));
> +
> +	vchi_ctx->vchi_instance = NULL;
> +}
> +
>  static int bcm2835_audio_open_connection(struct bcm2835_alsa_stream *alsa_stream)
>  {
> -	static VCHI_INSTANCE_T vchi_instance;
> -	static VCHI_CONNECTION_T *vchi_connection;
> -	static int initted;
>  	struct bcm2835_audio_instance *instance =
>  		(struct bcm2835_audio_instance *)alsa_stream->instance;
> -	int ret;
> +	struct bcm2835_vchi_ctx *vhci_ctx = alsa_stream->chip->vchi_ctx;
>  
>  	LOG_INFO("%s: start\n", __func__);
>  	BUG_ON(instance);
> @@ -390,28 +421,9 @@ static int bcm2835_audio_open_connection(struct bcm2835_alsa_stream *alsa_stream
>  		return 0;
>  	}
>  
> -	/* Initialize and create a VCHI connection */
> -	if (!initted) {
> -		ret = vchi_initialise(&vchi_instance);
> -		if (ret) {
> -			LOG_ERR("%s: failed to initialise VCHI instance (ret=%d)\n",
> -				__func__, ret);
> -
> -			return -EIO;
> -		}
> -		ret = vchi_connect(NULL, 0, vchi_instance);
> -		if (ret) {
> -			LOG_ERR("%s: failed to connect VCHI instance (ret=%d)\n",
> -				__func__, ret);
> -
> -			kfree(vchi_instance);
> -			return -EIO;
> -		}
> -		initted = 1;
> -	}
> -
>  	/* Initialize an instance of the audio service */
> -	instance = vc_vchi_audio_init(vchi_instance, &vchi_connection, 1);
> +	instance = vc_vchi_audio_init(vhci_ctx->vchi_instance,
> +				      &vhci_ctx->vchi_connection, 1);
>  
>  	if (IS_ERR(instance)) {
>  		LOG_ERR("%s: failed to initialize audio service\n", __func__);
> diff --git a/drivers/staging/vc04_services/bcm2835-audio/bcm2835.c b/drivers/staging/vc04_services/bcm2835-audio/bcm2835.c
> index 9030d71a3d0b..009c972d93d6 100644
> --- a/drivers/staging/vc04_services/bcm2835-audio/bcm2835.c
> +++ b/drivers/staging/vc04_services/bcm2835-audio/bcm2835.c
> @@ -65,6 +65,36 @@ static int snd_devm_add_child(struct device *dev, struct device *child)
>  	return 0;
>  }
>  
> +static void bcm2835_devm_free_vchi_ctx(struct device *dev, void *res)
> +{
> +	struct bcm2835_vchi_ctx *vchi_ctx = res;
> +
> +	bcm2835_free_vchi_ctx(vchi_ctx);
> +}
> +
> +static int bcm2835_devm_add_vchi_ctx(struct device *dev)
> +{
> +	struct bcm2835_vchi_ctx *vchi_ctx;
> +	int ret;
> +
> +	vchi_ctx = devres_alloc(bcm2835_devm_free_vchi_ctx, sizeof(*vchi_ctx),
> +				GFP_KERNEL);
> +	if (!vchi_ctx)
> +		return -ENOMEM;
> +
> +	memset(vchi_ctx, 0, sizeof(*vchi_ctx));
> +
> +	ret = bcm2835_new_vchi_ctx(vchi_ctx);
> +	if (ret) {
> +		devres_free(vchi_ctx);
> +		return ret;
> +	}
> +
> +	devres_add(dev, vchi_ctx);
> +
> +	return 0;
> +}
> +
>  static void snd_bcm2835_release(struct device *dev)
>  {
>  	struct bcm2835_chip *chip = dev_get_drvdata(dev);
> @@ -106,8 +136,6 @@ static int snd_bcm2835_dev_free(struct snd_device *device)
>  	struct bcm2835_chip *chip = device->device_data;
>  	struct snd_card *card = chip->card;
>  
> -	/* TODO: free pcm, ctl */
> -
>  	snd_device_free(card, chip);
>  
>  	return 0;
> @@ -133,6 +161,13 @@ static int snd_bcm2835_create(struct snd_card *card,
>  
>  	chip->card = card;
>  
> +	chip->vchi_ctx = devres_find(card->dev->parent,
> +				     bcm2835_devm_free_vchi_ctx, NULL, NULL);
> +	if (!chip->vchi_ctx) {
> +		kfree(chip);
> +		return err;

kbuild is complaining that "err" is uninitialized here but for some
reason the line numbers are off.  It's a real bug.

regards,
dan carpenter

  reply	other threads:[~2018-04-24 11:50 UTC|newest]

Thread overview: 18+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2018-04-17  5:00 [PATCH] " Kirill Marinushkin
2018-04-17 19:56 ` kbuild test robot
2018-04-23 13:50 ` Greg Kroah-Hartman
2018-04-24  0:35   ` Kirill Marinushkin
2018-04-24  7:16     ` Greg Kroah-Hartman
2018-04-24  7:44       ` [RESEND PATCH] " Kirill Marinushkin
2018-04-24 11:50         ` Dan Carpenter [this message]
2018-04-24 16:24         ` Andy Shevchenko
2018-04-24 18:27         ` Kirill Marinushkin
2018-04-24 18:35           ` Andy Shevchenko
2018-04-24 18:51             ` Kirill Marinushkin
2018-04-25  6:16               ` Greg Kroah-Hartman
2018-04-24 19:57           ` [PATCH v2] " Kirill Marinushkin
2018-04-25  6:16             ` Greg Kroah-Hartman
2018-04-25 17:35               ` Kirill Marinushkin
2018-04-26  5:38                 ` Dan Carpenter
2018-04-26 17:34               ` [PATCH v3] " Kirill Marinushkin
2018-04-24  8:14     ` [PATCH] " Stefan Wahren

Reply instructions:

You may reply publicly to this message via plain-text email
using any one of the following methods:

* Save the following mbox file, import it into your mail client,
  and reply-to-all from there: mbox

  Avoid top-posting and favor interleaved quoting:
  https://en.wikipedia.org/wiki/Posting_style#Interleaved_style

* Reply using the --to, --cc, and --in-reply-to
  switches of git-send-email(1):

  git send-email \
    --in-reply-to=20180424115052.pqdblbts2ko36njh@mwanda \
    --to=dan.carpenter@oracle.com \
    --cc=andy.shevchenko@gmail.com \
    --cc=bcm-kernel-feedback-list@broadcom.com \
    --cc=devel@driverdev.osuosl.org \
    --cc=eric@anholt.net \
    --cc=f.fainelli@gmail.com \
    --cc=gregkh@linuxfoundation.org \
    --cc=k.marinushkin@gmail.com \
    --cc=linux-arm-kernel@lists.infradead.org \
    --cc=linux-kernel@vger.kernel.org \
    --cc=linux-rpi-kernel@lists.infradead.org \
    --cc=rjui@broadcom.com \
    --cc=sbranden@broadcom.com \
    --cc=stefan.wahren@i2se.com \
    --subject='Re: [RESEND PATCH] staging: bcm2835-audio: Disconnect and free vchi_instance on module_exit()' \
    /path/to/YOUR_REPLY

  https://kernel.org/pub/software/scm/git/docs/git-send-email.html

* If your mail client supports setting the In-Reply-To header
  via mailto: links, try the mailto: link

This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).