public inbox for linux-staging@lists.linux.dev
 help / color / mirror / Atom feed
From: Hans de Goede <hdegoede@redhat.com>
To: Mauro Carvalho Chehab <mchehab@kernel.org>,
	Sakari Ailus <sakari.ailus@linux.intel.com>
Cc: Hans de Goede <hdegoede@redhat.com>,
	Tsuchiya Yuto <kitakar@gmail.com>,
	Andy Shevchenko <andy@kernel.org>,
	Yury Luneff <yury.lunev@gmail.com>,
	Nable <nable.maininbox@googlemail.com>,
	andrey.i.trufanov@gmail.com, Fabio Aiuto <fabioaiuto83@gmail.com>,
	linux-media@vger.kernel.org, linux-staging@lists.linux.dev
Subject: [PATCH 06/20] media: atomisp: Fix deadlock when the /dev/video# node is closed while still streaming
Date: Sun, 20 Nov 2022 23:40:47 +0100	[thread overview]
Message-ID: <20221120224101.746199-7-hdegoede@redhat.com> (raw)
In-Reply-To: <20221120224101.746199-1-hdegoede@redhat.com>

atomisp_release() was taking pipe->vb_queue_mutex + isp->mutex at the
same time. But if the /dev/video# node is closed while still streaming
then vb2_queue_release() will call atomisp_stop_streaming() which takes
isp->mutex itself, leading to a deadlock.

To fix this only take isp->mutex after cleaning up the v4l2_fh /
the vb2_queue. While at it switch to vb2_fop_release() which will take
pipe->vb_queue_mutex for us, which also resolves a FIXME comment.

Signed-off-by: Hans de Goede <hdegoede@redhat.com>
---
 .../staging/media/atomisp/pci/atomisp_fops.c  | 21 +++++--------------
 1 file changed, 5 insertions(+), 16 deletions(-)

diff --git a/drivers/staging/media/atomisp/pci/atomisp_fops.c b/drivers/staging/media/atomisp/pci/atomisp_fops.c
index 21ca276831ce..8cff26d42b82 100644
--- a/drivers/staging/media/atomisp/pci/atomisp_fops.c
+++ b/drivers/staging/media/atomisp/pci/atomisp_fops.c
@@ -861,23 +861,14 @@ static int atomisp_release(struct file *file)
 
 	v4l2_fh_init(&fh.vfh, vdev);
 
-	mutex_lock(&pipe->vb_queue_mutex);
-	mutex_lock(&isp->mutex);
-
 	dev_dbg(isp->dev, "release device %s\n", vdev->name);
 
 	asd->subdev.devnode = vdev;
 
-	/*
-	 * FIXME This if is copied from _vb2_fop_release, this cannot use that
-	 * because that calls v4l2_fh_release() earlier then this function.
-	 * Maybe we can release the fh earlier though, it does not look like
-	 * anything needs it after this.
-	 */
-	if (file->private_data == vdev->queue->owner) {
-		vb2_queue_release(vdev->queue);
-		vdev->queue->owner = NULL;
-	}
+	/* Note file must not be used after this! */
+	vb2_fop_release(file);
+
+	mutex_lock(&isp->mutex);
 
 	pipe->users--;
 	if (pipe->users)
@@ -940,9 +931,7 @@ static int atomisp_release(struct file *file)
 				     V4L2_SEL_TGT_COMPOSE, 0,
 				     &clear_compose);
 	mutex_unlock(&isp->mutex);
-	mutex_unlock(&pipe->vb_queue_mutex);
-
-	return v4l2_fh_release(file);
+	return 0;
 }
 
 const struct v4l2_file_operations atomisp_fops = {
-- 
2.38.1


  parent reply	other threads:[~2022-11-20 22:41 UTC|newest]

Thread overview: 33+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2022-11-20 22:40 [PATCH 00/20] media: atomisp: Misc. cleanups / fixes Hans de Goede
2022-11-20 22:40 ` [PATCH 01/20] media: atomisp: Silence: 'atomisp_q_one_s3a_buffer: drop one s3a stat which has exp_id xx' log messages Hans de Goede
2022-11-21  8:54   ` Andy Shevchenko
2022-11-21 12:27     ` Hans de Goede
2022-11-20 22:40 ` [PATCH 02/20] media: atomisp: Remove accelerator pipe creation code Hans de Goede
2022-11-20 22:40 ` [PATCH 03/20] media: atomisp: Remove unused QOS defines / structure member Hans de Goede
2022-11-20 22:40 ` [PATCH 04/20] media: atomisp: Flush queue on atomisp_css_start() error Hans de Goede
2022-11-20 22:40 ` [PATCH 05/20] media: atomisp: Log an error on failing to alloc private-mem Hans de Goede
2022-11-21  8:58   ` Andy Shevchenko
2022-11-21 12:29     ` Hans de Goede
2022-11-20 22:40 ` Hans de Goede [this message]
2022-11-20 22:40 ` [PATCH 07/20] media: atomisp: Remove 2 unused accelerator mode related functions Hans de Goede
2022-11-20 22:40 ` [PATCH 08/20] media: atomisp: Remove atomisp_css_yuvpp_configure_viewfinder() function Hans de Goede
2022-11-20 22:40 ` [PATCH 09/20] media: atomisp: Remove unused ia_css_frame_*() functions Hans de Goede
2022-11-20 22:40 ` [PATCH 10/20] media: atomisp: Drop userptr support from hmm Hans de Goede
2022-11-21  9:01   ` Andy Shevchenko
2022-11-21 12:31     ` Hans de Goede
2022-11-20 22:40 ` [PATCH 11/20] media: atomisp: Remove double atomisp_mrfld_power_down()/_up() calls from atomisp_reset() Hans de Goede
2022-11-20 22:40 ` [PATCH 12/20] media: atomisp: Remove atomisp_mrfld_power_down()/_up() Hans de Goede
2022-11-20 22:40 ` [PATCH 13/20] media: atomisp: Remove clearing of config from atomisp_css_uninit() Hans de Goede
2022-11-20 22:40 ` [PATCH 14/20] media: atomisp: Remove atomisp_css_suspend()/_resume() Hans de Goede
2022-11-20 22:40 ` [PATCH 15/20] media: atomisp: Remove sw_contex.power_state checks Hans de Goede
2022-11-21  9:09   ` Andy Shevchenko
2022-11-21  9:16     ` Andy Shevchenko
2022-11-20 22:40 ` [PATCH 16/20] media: atomisp: Remove duplication between runtime-pm and normal-pm code Hans de Goede
2022-11-20 22:40 ` [PATCH 17/20] media: atomisp: Move calling of css_[un]init() to power_on()/_off() Hans de Goede
2022-11-21  9:14   ` Andy Shevchenko
2022-11-21 12:33     ` Hans de Goede
2022-11-20 22:40 ` [PATCH 18/20] media: atomisp: Remove atomisp_ospm_dphy_down() call from probe error path Hans de Goede
2022-11-20 22:41 ` [PATCH 19/20] media: atomisp: Remove atomisp_ospm_dphy_up()/_down() functions Hans de Goede
2022-11-20 22:41 ` [PATCH 20/20] media: atomisp_ov2680: Fix 1280x720 -> 1296x736 resolution Hans de Goede
2022-11-21  9:20 ` [PATCH 00/20] media: atomisp: Misc. cleanups / fixes Andy Shevchenko
2022-11-21 12:40   ` Hans de Goede

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=20221120224101.746199-7-hdegoede@redhat.com \
    --to=hdegoede@redhat.com \
    --cc=andrey.i.trufanov@gmail.com \
    --cc=andy@kernel.org \
    --cc=fabioaiuto83@gmail.com \
    --cc=kitakar@gmail.com \
    --cc=linux-media@vger.kernel.org \
    --cc=linux-staging@lists.linux.dev \
    --cc=mchehab@kernel.org \
    --cc=nable.maininbox@googlemail.com \
    --cc=sakari.ailus@linux.intel.com \
    --cc=yury.lunev@gmail.com \
    /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
Be sure your reply has a Subject: header at the top and a blank line before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox