* [PATCH] media: uvcvideo: Use scope-based cleanup helper
@ 2026-01-04 6:45 Chen Changcheng
2026-01-04 10:33 ` Markus Elfring
` (2 more replies)
0 siblings, 3 replies; 4+ messages in thread
From: Chen Changcheng @ 2026-01-04 6:45 UTC (permalink / raw)
To: laurent.pinchart, hansg, mchehab
Cc: linux-media, linux-kernel, Chen Changcheng
Replace the manual kfree() with __free(kfree) annotation for data
references. This aligns the code with the latest kernel style.
No functional change intended.
Signed-off-by: Chen Changcheng <chenchangcheng@kylinos.cn>
---
drivers/media/usb/uvc/uvc_video.c | 19 ++++++-------------
1 file changed, 6 insertions(+), 13 deletions(-)
diff --git a/drivers/media/usb/uvc/uvc_video.c b/drivers/media/usb/uvc/uvc_video.c
index 2094e059d7d3..1ee06d597431 100644
--- a/drivers/media/usb/uvc/uvc_video.c
+++ b/drivers/media/usb/uvc/uvc_video.c
@@ -291,7 +291,7 @@ static int uvc_get_video_ctrl(struct uvc_streaming *stream,
struct uvc_streaming_control *ctrl, int probe, u8 query)
{
u16 size = uvc_video_ctrl_size(stream);
- u8 *data;
+ u8 *data __free(kfree) = NULL;
int ret;
if ((stream->dev->quirks & UVC_QUIRK_PROBE_DEF) &&
@@ -317,8 +317,7 @@ static int uvc_get_video_ctrl(struct uvc_streaming *stream,
"supported. Enabling workaround.\n");
memset(ctrl, 0, sizeof(*ctrl));
ctrl->wCompQuality = le16_to_cpup((__le16 *)data);
- ret = 0;
- goto out;
+ return 0
} else if (query == UVC_GET_DEF && probe == 1 && ret != size) {
/*
* Many cameras don't support the GET_DEF request on their
@@ -328,15 +327,13 @@ static int uvc_get_video_ctrl(struct uvc_streaming *stream,
uvc_warn_once(stream->dev, UVC_WARN_PROBE_DEF, "UVC non "
"compliance - GET_DEF(PROBE) not supported. "
"Enabling workaround.\n");
- ret = -EIO;
- goto out;
+ return -EIO;
} else if (ret != size) {
dev_err(&stream->intf->dev,
"Failed to query (%s) UVC %s control : %d (exp. %u).\n",
uvc_query_name(query), probe ? "probe" : "commit",
ret, size);
- ret = (ret == -EPROTO) ? -EPROTO : -EIO;
- goto out;
+ return (ret == -EPROTO) ? -EPROTO : -EIO;
}
ctrl->bmHint = le16_to_cpup((__le16 *)&data[0]);
@@ -371,18 +368,15 @@ static int uvc_get_video_ctrl(struct uvc_streaming *stream,
* format and frame descriptors.
*/
uvc_fixup_video_ctrl(stream, ctrl);
- ret = 0;
-out:
- kfree(data);
- return ret;
+ return 0;
}
static int uvc_set_video_ctrl(struct uvc_streaming *stream,
struct uvc_streaming_control *ctrl, int probe)
{
u16 size = uvc_video_ctrl_size(stream);
- u8 *data;
+ u8 *data __free(kfree) = NULL;
int ret;
data = kzalloc(size, GFP_KERNEL);
@@ -419,7 +413,6 @@ static int uvc_set_video_ctrl(struct uvc_streaming *stream,
ret = -EIO;
}
- kfree(data);
return ret;
}
base-commit: 805f9a061372164d43ddef771d7cd63e3ba6d845
--
2.25.1
^ permalink raw reply [flat|nested] 4+ messages in thread* Re: [PATCH] media: uvcvideo: Use scope-based cleanup helper
2026-01-04 6:45 [PATCH] media: uvcvideo: Use scope-based cleanup helper Chen Changcheng
@ 2026-01-04 10:33 ` Markus Elfring
2026-01-04 13:57 ` kernel test robot
2026-01-04 19:50 ` Laurent Pinchart
2 siblings, 0 replies; 4+ messages in thread
From: Markus Elfring @ 2026-01-04 10:33 UTC (permalink / raw)
To: Chen Changcheng, linux-media, Hans de Goede, Laurent Pinchart,
Mauro Carvalho Chehab
Cc: LKML
> Replace the manual kfree() with __free(kfree) annotation for data
> references. This aligns the code with the latest kernel style.
Align?
…
> +++ b/drivers/media/usb/uvc/uvc_video.c
> @@ -291,7 +291,7 @@ static int uvc_get_video_ctrl(struct uvc_streaming *stream,
> struct uvc_streaming_control *ctrl, int probe, u8 query)
> {
> u16 size = uvc_video_ctrl_size(stream);
> - u8 *data;
> + u8 *data __free(kfree) = NULL;
> int ret;
…
How do you think about to reduce the scopes also for these local variables?
> @@ -317,8 +317,7 @@ static int uvc_get_video_ctrl(struct uvc_streaming *stream,
> "supported. Enabling workaround.\n");
> memset(ctrl, 0, sizeof(*ctrl));
> ctrl->wCompQuality = le16_to_cpup((__le16 *)data);
> - ret = 0;
> - goto out;
> + return 0
> } else if (query == UVC_GET_DEF && probe == 1 && ret != size) {
…
Did you accidentally overlook the need for another semicolon here?
Regards,
Markus
^ permalink raw reply [flat|nested] 4+ messages in thread* Re: [PATCH] media: uvcvideo: Use scope-based cleanup helper
2026-01-04 6:45 [PATCH] media: uvcvideo: Use scope-based cleanup helper Chen Changcheng
2026-01-04 10:33 ` Markus Elfring
@ 2026-01-04 13:57 ` kernel test robot
2026-01-04 19:50 ` Laurent Pinchart
2 siblings, 0 replies; 4+ messages in thread
From: kernel test robot @ 2026-01-04 13:57 UTC (permalink / raw)
To: Chen Changcheng, laurent.pinchart, hansg, mchehab
Cc: oe-kbuild-all, linux-media, linux-kernel, Chen Changcheng
Hi Chen,
kernel test robot noticed the following build errors:
[auto build test ERROR on 805f9a061372164d43ddef771d7cd63e3ba6d845]
url: https://github.com/intel-lab-lkp/linux/commits/Chen-Changcheng/media-uvcvideo-Use-scope-based-cleanup-helper/20260104-144716
base: 805f9a061372164d43ddef771d7cd63e3ba6d845
patch link: https://lore.kernel.org/r/20260104064520.115462-1-chenchangcheng%40kylinos.cn
patch subject: [PATCH] media: uvcvideo: Use scope-based cleanup helper
config: x86_64-rhel-9.4 (https://download.01.org/0day-ci/archive/20260104/202601041449.BwJ5RZ2a-lkp@intel.com/config)
compiler: gcc-14 (Debian 14.2.0-19) 14.2.0
reproduce (this is a W=1 build): (https://download.01.org/0day-ci/archive/20260104/202601041449.BwJ5RZ2a-lkp@intel.com/reproduce)
If you fix the issue in a separate patch/commit (i.e. not just a new version of
the same patch/commit), kindly add following tags
| Reported-by: kernel test robot <lkp@intel.com>
| Closes: https://lore.kernel.org/oe-kbuild-all/202601041449.BwJ5RZ2a-lkp@intel.com/
All errors (new ones prefixed by >>):
drivers/media/usb/uvc/uvc_video.c: In function 'uvc_get_video_ctrl':
>> drivers/media/usb/uvc/uvc_video.c:320:25: error: expected ';' before '}' token
320 | return 0
| ^
| ;
321 | } else if (query == UVC_GET_DEF && probe == 1 && ret != size) {
| ~
vim +320 drivers/media/usb/uvc/uvc_video.c
289
290 static int uvc_get_video_ctrl(struct uvc_streaming *stream,
291 struct uvc_streaming_control *ctrl, int probe, u8 query)
292 {
293 u16 size = uvc_video_ctrl_size(stream);
294 u8 *data __free(kfree) = NULL;
295 int ret;
296
297 if ((stream->dev->quirks & UVC_QUIRK_PROBE_DEF) &&
298 query == UVC_GET_DEF)
299 return -EIO;
300
301 data = kmalloc(size, GFP_KERNEL);
302 if (data == NULL)
303 return -ENOMEM;
304
305 ret = __uvc_query_ctrl(stream->dev, query, 0, stream->intfnum,
306 probe ? UVC_VS_PROBE_CONTROL : UVC_VS_COMMIT_CONTROL, data,
307 size, uvc_timeout_param);
308
309 if ((query == UVC_GET_MIN || query == UVC_GET_MAX) && ret == 2) {
310 /*
311 * Some cameras, mostly based on Bison Electronics chipsets,
312 * answer a GET_MIN or GET_MAX request with the wCompQuality
313 * field only.
314 */
315 uvc_warn_once(stream->dev, UVC_WARN_MINMAX, "UVC non "
316 "compliance - GET_MIN/MAX(PROBE) incorrectly "
317 "supported. Enabling workaround.\n");
318 memset(ctrl, 0, sizeof(*ctrl));
319 ctrl->wCompQuality = le16_to_cpup((__le16 *)data);
> 320 return 0
321 } else if (query == UVC_GET_DEF && probe == 1 && ret != size) {
322 /*
323 * Many cameras don't support the GET_DEF request on their
324 * video probe control. Warn once and return, the caller will
325 * fall back to GET_CUR.
326 */
327 uvc_warn_once(stream->dev, UVC_WARN_PROBE_DEF, "UVC non "
328 "compliance - GET_DEF(PROBE) not supported. "
329 "Enabling workaround.\n");
330 return -EIO;
331 } else if (ret != size) {
332 dev_err(&stream->intf->dev,
333 "Failed to query (%s) UVC %s control : %d (exp. %u).\n",
334 uvc_query_name(query), probe ? "probe" : "commit",
335 ret, size);
336 return (ret == -EPROTO) ? -EPROTO : -EIO;
337 }
338
339 ctrl->bmHint = le16_to_cpup((__le16 *)&data[0]);
340 ctrl->bFormatIndex = data[2];
341 ctrl->bFrameIndex = data[3];
342 ctrl->dwFrameInterval = le32_to_cpup((__le32 *)&data[4]);
343 ctrl->wKeyFrameRate = le16_to_cpup((__le16 *)&data[8]);
344 ctrl->wPFrameRate = le16_to_cpup((__le16 *)&data[10]);
345 ctrl->wCompQuality = le16_to_cpup((__le16 *)&data[12]);
346 ctrl->wCompWindowSize = le16_to_cpup((__le16 *)&data[14]);
347 ctrl->wDelay = le16_to_cpup((__le16 *)&data[16]);
348 ctrl->dwMaxVideoFrameSize = get_unaligned_le32(&data[18]);
349 ctrl->dwMaxPayloadTransferSize = get_unaligned_le32(&data[22]);
350
351 if (size >= 34) {
352 ctrl->dwClockFrequency = get_unaligned_le32(&data[26]);
353 ctrl->bmFramingInfo = data[30];
354 ctrl->bPreferedVersion = data[31];
355 ctrl->bMinVersion = data[32];
356 ctrl->bMaxVersion = data[33];
357 } else {
358 ctrl->dwClockFrequency = stream->dev->clock_frequency;
359 ctrl->bmFramingInfo = 0;
360 ctrl->bPreferedVersion = 0;
361 ctrl->bMinVersion = 0;
362 ctrl->bMaxVersion = 0;
363 }
364
365 /*
366 * Some broken devices return null or wrong dwMaxVideoFrameSize and
367 * dwMaxPayloadTransferSize fields. Try to get the value from the
368 * format and frame descriptors.
369 */
370 uvc_fixup_video_ctrl(stream, ctrl);
371
372 return 0;
373 }
374
--
0-DAY CI Kernel Test Service
https://github.com/intel/lkp-tests/wiki
^ permalink raw reply [flat|nested] 4+ messages in thread* Re: [PATCH] media: uvcvideo: Use scope-based cleanup helper
2026-01-04 6:45 [PATCH] media: uvcvideo: Use scope-based cleanup helper Chen Changcheng
2026-01-04 10:33 ` Markus Elfring
2026-01-04 13:57 ` kernel test robot
@ 2026-01-04 19:50 ` Laurent Pinchart
2 siblings, 0 replies; 4+ messages in thread
From: Laurent Pinchart @ 2026-01-04 19:50 UTC (permalink / raw)
To: Chen Changcheng; +Cc: hansg, mchehab, linux-media, linux-kernel
On Sun, Jan 04, 2026 at 02:45:20PM +0800, Chen Changcheng wrote:
> Replace the manual kfree() with __free(kfree) annotation for data
> references. This aligns the code with the latest kernel style.
There are more occurrences through the driver, please address them all.
> No functional change intended.
>
> Signed-off-by: Chen Changcheng <chenchangcheng@kylinos.cn>
> ---
> drivers/media/usb/uvc/uvc_video.c | 19 ++++++-------------
> 1 file changed, 6 insertions(+), 13 deletions(-)
>
> diff --git a/drivers/media/usb/uvc/uvc_video.c b/drivers/media/usb/uvc/uvc_video.c
> index 2094e059d7d3..1ee06d597431 100644
> --- a/drivers/media/usb/uvc/uvc_video.c
> +++ b/drivers/media/usb/uvc/uvc_video.c
> @@ -291,7 +291,7 @@ static int uvc_get_video_ctrl(struct uvc_streaming *stream,
> struct uvc_streaming_control *ctrl, int probe, u8 query)
> {
> u16 size = uvc_video_ctrl_size(stream);
> - u8 *data;
> + u8 *data __free(kfree) = NULL;
> int ret;
>
> if ((stream->dev->quirks & UVC_QUIRK_PROBE_DEF) &&
> @@ -317,8 +317,7 @@ static int uvc_get_video_ctrl(struct uvc_streaming *stream,
> "supported. Enabling workaround.\n");
> memset(ctrl, 0, sizeof(*ctrl));
> ctrl->wCompQuality = le16_to_cpup((__le16 *)data);
> - ret = 0;
> - goto out;
> + return 0
> } else if (query == UVC_GET_DEF && probe == 1 && ret != size) {
> /*
> * Many cameras don't support the GET_DEF request on their
> @@ -328,15 +327,13 @@ static int uvc_get_video_ctrl(struct uvc_streaming *stream,
> uvc_warn_once(stream->dev, UVC_WARN_PROBE_DEF, "UVC non "
> "compliance - GET_DEF(PROBE) not supported. "
> "Enabling workaround.\n");
> - ret = -EIO;
> - goto out;
> + return -EIO;
> } else if (ret != size) {
> dev_err(&stream->intf->dev,
> "Failed to query (%s) UVC %s control : %d (exp. %u).\n",
> uvc_query_name(query), probe ? "probe" : "commit",
> ret, size);
> - ret = (ret == -EPROTO) ? -EPROTO : -EIO;
> - goto out;
> + return (ret == -EPROTO) ? -EPROTO : -EIO;
> }
>
> ctrl->bmHint = le16_to_cpup((__le16 *)&data[0]);
> @@ -371,18 +368,15 @@ static int uvc_get_video_ctrl(struct uvc_streaming *stream,
> * format and frame descriptors.
> */
> uvc_fixup_video_ctrl(stream, ctrl);
> - ret = 0;
>
> -out:
> - kfree(data);
> - return ret;
> + return 0;
> }
>
> static int uvc_set_video_ctrl(struct uvc_streaming *stream,
> struct uvc_streaming_control *ctrl, int probe)
> {
> u16 size = uvc_video_ctrl_size(stream);
> - u8 *data;
> + u8 *data __free(kfree) = NULL;
> int ret;
>
> data = kzalloc(size, GFP_KERNEL);
> @@ -419,7 +413,6 @@ static int uvc_set_video_ctrl(struct uvc_streaming *stream,
> ret = -EIO;
You can return -EIO here...
> }
>
> - kfree(data);
> return ret;
... and replace this with return 0. That's a change in behaviour, but as
far as I can tell, none of the callers use the positive return values, so
returning 0 on success should be fine.
> }
>
>
> base-commit: 805f9a061372164d43ddef771d7cd63e3ba6d845
--
Regards,
Laurent Pinchart
^ permalink raw reply [flat|nested] 4+ messages in thread
end of thread, other threads:[~2026-01-04 19:50 UTC | newest]
Thread overview: 4+ messages (download: mbox.gz / follow: Atom feed)
-- links below jump to the message on this page --
2026-01-04 6:45 [PATCH] media: uvcvideo: Use scope-based cleanup helper Chen Changcheng
2026-01-04 10:33 ` Markus Elfring
2026-01-04 13:57 ` kernel test robot
2026-01-04 19:50 ` Laurent Pinchart
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox
all inboxes | Powered by JetHome®