drivers/staging/media/atomisp/pci/atomisp_ioctl.c | 10 +++++++++- 1 file changed, 9 insertions(+), 1 deletion(-)
The function atomisp_set_fmt() modifies shared device state and expects
callers to hold the isp->mutex for synchronization. While most internal
callers correctly lock the mutex before invoking atomisp_set_fmt(), the
V4L2 ioctl handler atomisp_s_fmt_cap() does not.
This results in an unsafe execution path for VIDIOC_S_FMT ioctls
(e.g. via v4l2-ctl), where shared structures such as pipe->pix and
pipe->frame_info may be modified concurrently without proper protection.
- Fix this by explicitly locking isp->mutex in atomisp_s_fmt_cap().
Fixes: 4bdab80981ca ("media: atomisp: Make it possible to call atomisp_set_fmt() without a file handle")
Signed-off-by: Abdelrahman Fekry <abdelrahmanfekry375@gmail.com>
---
v2:
- Add Fixes tag
- use cleanup.h micros instead of explicitly calling mutex_lock/unlock
v1: https://lore.kernel.org/all/20250716014225.15279-1-abdelrahmanfekry375@gmail.com/
drivers/staging/media/atomisp/pci/atomisp_ioctl.c | 10 +++++++++-
1 file changed, 9 insertions(+), 1 deletion(-)
diff --git a/drivers/staging/media/atomisp/pci/atomisp_ioctl.c b/drivers/staging/media/atomisp/pci/atomisp_ioctl.c
index bb8b2f2213b0..d3b8e480065e 100644
--- a/drivers/staging/media/atomisp/pci/atomisp_ioctl.c
+++ b/drivers/staging/media/atomisp/pci/atomisp_ioctl.c
@@ -9,6 +9,7 @@
#include <linux/delay.h>
#include <linux/pci.h>
+#include <linux/cleanup.h>
#include <media/v4l2-ioctl.h>
#include <media/v4l2-event.h>
@@ -416,8 +417,15 @@ static int atomisp_s_fmt_cap(struct file *file, void *fh,
struct v4l2_format *f)
{
struct video_device *vdev = video_devdata(file);
+ struct atomisp_device *isp = video_get_drvdata(vdev);
+
+ int ret;
- return atomisp_set_fmt(vdev, f);
+ scoped_guard(mutex, &isp->mutex)
+ {
+ ret = atomisp_set_fmt(vdev, f);
+ }
+ return ret;
}
/*
--
2.25.1
On Thu, Jul 17, 2025 at 4:30 AM Abdelrahman Fekry <abdelrahmanfekry375@gmail.com> wrote: > > The function atomisp_set_fmt() modifies shared device state and expects > callers to hold the isp->mutex for synchronization. While most internal > callers correctly lock the mutex before invoking atomisp_set_fmt(), the > V4L2 ioctl handler atomisp_s_fmt_cap() does not. > > This results in an unsafe execution path for VIDIOC_S_FMT ioctls > (e.g. via v4l2-ctl), where shared structures such as pipe->pix and > pipe->frame_info may be modified concurrently without proper protection. > > - Fix this by explicitly locking isp->mutex in atomisp_s_fmt_cap(). ... > #include <linux/delay.h> > #include <linux/pci.h> > +#include <linux/cleanup.h> Keep it ordered. ... > + int ret; > > - return atomisp_set_fmt(vdev, f); > + scoped_guard(mutex, &isp->mutex) > + { > + ret = atomisp_set_fmt(vdev, f); > + } > + return ret; As Dan said, this should just add a (one line) guard()() and no other lines being touched. > } -- With Best Regards, Andy Shevchenko
On Thu, Jul 17, 2025 at 04:30:03AM +0300, Abdelrahman Fekry wrote: > diff --git a/drivers/staging/media/atomisp/pci/atomisp_ioctl.c b/drivers/staging/media/atomisp/pci/atomisp_ioctl.c > index bb8b2f2213b0..d3b8e480065e 100644 > --- a/drivers/staging/media/atomisp/pci/atomisp_ioctl.c > +++ b/drivers/staging/media/atomisp/pci/atomisp_ioctl.c > @@ -9,6 +9,7 @@ > > #include <linux/delay.h> > #include <linux/pci.h> > +#include <linux/cleanup.h> > > #include <media/v4l2-ioctl.h> > #include <media/v4l2-event.h> > @@ -416,8 +417,15 @@ static int atomisp_s_fmt_cap(struct file *file, void *fh, > struct v4l2_format *f) > { > struct video_device *vdev = video_devdata(file); > + struct atomisp_device *isp = video_get_drvdata(vdev); > + Delete this blank line. > + int ret; > > - return atomisp_set_fmt(vdev, f); > + scoped_guard(mutex, &isp->mutex) > + { This open curly brace should have gone on the line before: scoped_guard(mutex, &isp->mutex) { But actually just use: guard(mutex)(&isp->mutex); It will hold the lock until the end of the function. regards, dan carpenter > + ret = atomisp_set_fmt(vdev, f); > + } > + return ret; > }
© 2016 - 2025 Red Hat, Inc.