Re: [PATCH v2 6/8] media: platform: amd: isp4 video node and buffers handling added

From: Du, Bin
Date: Tue Jul 29 2025 - 02:13:21 EST


Many thanks Sultan, will try your changes in this and other mails together and let you know the result

On 7/27/2025 5:50 AM, Sultan Alsawaf wrote:
On Sat, Jul 26, 2025 at 02:41:41PM -0700, Sultan Alsawaf wrote:
On Fri, Jul 25, 2025 at 05:22:41PM +0800, Du, Bin wrote:
+ dev_warn(buf->dev, "ignore buffer free, refcount %u > 0",
+ refcount_read(&buf->refcount));

This refcount_read() is a possible use-after-free because `buf` is accessed
after isp4vid_vb2_put() puts its reference to `buf`. So something else could put
the last reference to `buf` and free it after this refcount dec but before the
refcount_read(). Maybe just remove this dev_warn() entirely?

The warning is important to debug mem related issue, plan to keep it but
without accessing buf or buf->refcount here. Do you think it acceptible?

Yes, that sounds good. So something like this:
`dev_warn(buf->dev, "ignore buffer free, refcount > 0");`

Sorry, to fix the dev_warn() we need to make a copy of buf->dev first:

--- a/drivers/media/platform/amd/isp4/isp4_video.c
+++ b/drivers/media/platform/amd/isp4/isp4_video.c
@@ -584,8 +584,9 @@ static void isp4vid_vb2_put(void *buf_priv)
{
struct isp4vid_vb2_buf *buf = (struct isp4vid_vb2_buf *)buf_priv;
struct amdgpu_bo *bo = (struct amdgpu_bo *)buf->bo;
+ struct device *dev = buf->dev;
- dev_dbg(buf->dev,
+ dev_dbg(dev,
"release isp user bo 0x%llx size %ld refcount %d is_expbuf %d",
buf->gpu_addr, buf->size,
buf->refcount.refs.counter, buf->is_expbuf);
@@ -601,8 +602,7 @@ static void isp4vid_vb2_put(void *buf_priv)
kfree(buf);
buf = NULL;
} else {
- dev_warn(buf->dev, "ignore buffer free, refcount %u > 0",
- refcount_read(&buf->refcount));
+ dev_warn(dev, "ignore buffer free, refcount > 0\n");
}
}
--

Sultan