Hi Hans,

On Thu, Jun 27, 2024 at 01:17:55PM +0200, Hans Verkuil wrote:

On 24/06/2024 10:41, Alain Volmat wrote:

...

Correct error handling within the dcmipp_create_subdevs by properly decrementing the i counter when releasing the subdeves.

Fixes: 28e0f3772296 ("media: stm32-dcmipp: STM32 DCMIPP camera interface driver") Cc: stable@vger.kernel.org Signed-off-by: Alain Volmat alain.volmat@foss.st.com

---

drivers/media/platform/st/stm32/stm32-dcmipp/dcmipp-core.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/drivers/media/platform/st/stm32/stm32-dcmipp/dcmipp-core.c b/drivers/media/platform/st/stm32/stm32-dcmipp/dcmipp-core.c index 4acc3b90d03a..4924ee36cfda 100644 --- a/drivers/media/platform/st/stm32/stm32-dcmipp/dcmipp-core.c +++ b/drivers/media/platform/st/stm32/stm32-dcmipp/dcmipp-core.c @@ -202,7 +202,7 @@ static int dcmipp_create_subdevs(struct dcmipp_device *dcmipp) return 0; err_init_entity:

• while (i > 0)

• while (i-- > 0) dcmipp->pipe_cfg->ents[i - 1].release(dcmipp->entity[i - 1]); return ret;

}

I accidentally merged this one, but this patch isn't right.

After this change the [i - 1] indices should be changed to [i]. If i == 1, then the while condition is true, but now i == 0 in the actual statement, so you access out-of-bounds values.

Right. I think the best way to fix this would be to just remove "- 1" inside the array indices. One could think this as a different bug but of course with an unpleasant side effect which you get after fixing the first bug.

I decided to revert it, since it is better to just get stuck in the while loop, then to crash.

But a new patch is needed for this.