Hi,

On Wednesday 21 September 2011 02:10 PM, Hiremath, Vaibhav wrote:

-----Original Message-----
From: Taneja, Archit
Sent: Friday, September 16, 2011 3:30 PM
To: Hiremath, Vaibhav
Cc: Valkeinen, Tomi; linux-omap@vger.kernel.org; Semwal, Sumit; linux-
me...@vger.kernel.org; Taneja, Archit
Subject: [PATCH 1/5] [media]: OMAP_VOUT: Fix check in reqbuf&  mmap for
buf_size allocation

The commit 383e4f69879d11c86ebdd38b3356f6d0690fb4cc makes reqbuf and mmap
prevent
requesting a larger size buffer than what is allocated at kernel boot
during
omap_vout_probe.

The requested size is compared with vout->buffer_size, this isn't correct
as
vout->buffer_size is later set to the size requested in reqbuf. When the
video
device is opened the next time, this check will prevent us to allocate a
buffer
which is larger than what we requested the last time.

Don't use vout->buffer_size, always check with the parameters
video1_bufsize
or video2_bufsize.

Signed-off-by: Archit Taneja<arc...@ti.com>
---
  drivers/media/video/omap/omap_vout.c |   10 ++++++++--
  1 files changed, 8 insertions(+), 2 deletions(-)

diff --git a/drivers/media/video/omap/omap_vout.c
b/drivers/media/video/omap/omap_vout.c
index 95daf98..e14c82b 100644
--- a/drivers/media/video/omap/omap_vout.c
+++ b/drivers/media/video/omap/omap_vout.c
@@ -664,10 +664,14 @@ static int omap_vout_buffer_setup(struct
videobuf_queue *q, unsigned int *count,
        u32 phy_addr = 0, virt_addr = 0;
        struct omap_vout_device *vout = q->priv_data;
        struct omapvideo_info *ovid =&vout->vid_info;
+       int vid_max_buf_size;

        if (!vout)
                return -EINVAL;

+       vid_max_buf_size = vout->vid == OMAP_VIDEO1 ? video1_bufsize :
+               video2_bufsize;
+
        if (V4L2_BUF_TYPE_VIDEO_OUTPUT != q->type)
                return -EINVAL;

@@ -690,7 +694,7 @@ static int omap_vout_buffer_setup(struct
videobuf_queue *q, unsigned int *count,
                video1_numbuffers : video2_numbuffers;

        /* Check the size of the buffer */
-       if (*size>  vout->buffer_size) {
+       if (*size>  vid_max_buf_size) {
Good catch !!!

                v4l2_err(&vout->vid_dev->v4l2_dev,
                                "buffer allocation mismatch [%u] [%u]\n",
                                *size, vout->buffer_size);
@@ -865,6 +869,8 @@ static int omap_vout_mmap(struct file *file, struct
vm_area_struct *vma)
        unsigned long size = (vma->vm_end - vma->vm_start);
        struct omap_vout_device *vout = file->private_data;
        struct videobuf_queue *q =&vout->vbq;
+       int vid_max_buf_size = vout->vid == OMAP_VIDEO1 ? video1_bufsize :
+               video2_bufsize;

        v4l2_dbg(1, debug,&vout->vid_dev->v4l2_dev,
                        " %s pgoff=0x%lx, start=0x%lx, end=0x%lx\n", __func__,
@@ -887,7 +893,7 @@ static int omap_vout_mmap(struct file *file, struct
vm_area_struct *vma)
                return -EINVAL;
        }
        /* Check the size of the buffer */
-       if (size>  vout->buffer_size) {
+       if (size>  vid_max_buf_size) {
Don't you think in case of mmap we should still check for the
vout->buffer_size, since this is the size user has requested in req_buf.

Ah, you are right, the check for the maximum size should only be in the reqbuf path. vout->buffer_size would have been updated correctly at time of mmap. I'll change this back to vout->buffer_size.

Thanks,
Archit


Thanks,
Vaibhav


                v4l2_err(&vout->vid_dev->v4l2_dev,
                                "insufficient memory [%lu] [%u]\n",
                                size, vout->buffer_size);
--
1.7.1



--
To unsubscribe from this list: send the line "unsubscribe linux-omap" in
the body of a message to majord...@vger.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html

Reply via email to