Skip to contentSkip to conversation

rpmsg: fix possible refcount leak in rpmsg_register_device_override()

PATCH Applied Hangyu Hua 3 participants 8 messages Jun 24, 2022 – Jun 29, 2022 Remote processor messaging (rpmsg)

lore
Share & tools
lore

Hangyu Hua

author
#0 lore raw

PATCH v2 1/1 rpmsg: fix possible refcount leak in rpmsg_register_device_override()

rpmsg_register_device_override need to call put_device to free vch when driver_set_override fails.

Fix this by adding a put_device() to the error path.

Fixes bb17d110cbf2 ("rpmsg: Fix calling device_lock() on non-initialized device")

---

v2: change the commit log.

 drivers/rpmsg/rpmsg_core.c | 1 +
 1 file changed, 1 insertion(+)
drivers/rpmsg/rpmsg_core.c +1 -0
index 290c1f02da10..5a47cad89fdc 100644
@@ -618,6 +618,7 @@ int rpmsg_register_device_override(struct rpmsg_device *rpdev,
 					  strlen(driver_override));
 		if (ret) {
 			dev_err(dev, "device_set_override failed: %d\n", ret);
+			put_device(dev);
 			return ret;
 		}
 	}
Signature · 1 line
2.25.1
maintainer Applied
#1 lore raw
Quoted 6 lines · Hangyu Hua

rpmsg_register_device_override need to call put_device to free vch when driver_set_override fails.

Fix this by adding a put_device() to the error path.

Fixes bb17d110cbf2 ("rpmsg: Fix calling device_lock() on non-initialized device")

This is funny... Neither Bjorn nor I have reviewed this patch...

Quoted 14 lines

---

v2: change the commit log.

 drivers/rpmsg/rpmsg_core.c | 1 +
 1 file changed, 1 insertion(+)
drivers/rpmsg/rpmsg_core.c +0 -0
index 290c1f02da10..5a47cad89fdc 100644

drivers/rpmsg/rpmsg_core.c int rpmsg_register_device_override(struct rpmsg_device *rpdev, · line 621

Quoted 4 lines
drivers/rpmsg/rpmsg_core.c +1 -0
@@ -618,6 +618,7 @@ int rpmsg_register_device_override(struct rpmsg_device *rpdev,
 					  strlen(driver_override));
 		if (ret) {
 			dev_err(dev, "device_set_override failed: %d\n", ret);
+			put_device(dev);

Applied.

Thanks,
Mathieu

Quoted 6 lines
drivers/rpmsg/rpmsg_core.c +0 -0
 			return ret;
 		}
 	}
Signature · 1 line
2.25.1
#2 lore raw
Quoted 9 lines · Mathieu Poirier
Quoted 6 lines · Hangyu Hua

rpmsg_register_device_override need to call put_device to free vch when driver_set_override fails.

Fix this by adding a put_device() to the error path.

Fixes bb17d110cbf2 ("rpmsg: Fix calling device_lock() on non-initialized device")

This is funny... Neither Bjorn nor I have reviewed this patch...

It was a fix for commit in Greg's tree and Greg's pick it up after a week or something. I am not sure if that's actually funny that Greg has to pick it up without review :(

Best regards,
Krzysztof

maintainer
#3 lore raw
Quoted 8 lines · Krzysztof Kozlowski
Quoted 6 lines · Mathieu Poirier
Quoted 5 lines · Hangyu Hua

rpmsg_register_device_override need to call put_device to free vch when driver_set_override fails.

Fix this by adding a put_device() to the error path.

Quoted 8 lines
Quoted 3 lines
Quoted 1 line
Fixes bb17d110cbf2 ("rpmsg: Fix calling device_lock() on non-initialized device")

This is funny... Neither Bjorn nor I have reviewed this patch...

It was a fix for commit in Greg's tree and Greg's pick it up after a week or something. I am not sure if that's actually funny that Greg has to pick it up without review :(

The patch was sent out on April 19th and committed 3 days later on April 22nd. Is 3 days the new patch review time standard?

Quoted 2 lines

Best regards,
Krzysztof

#4 lore raw
Quoted 14 lines · Mathieu Poirier
Quoted 12 lines · Krzysztof Kozlowski
Quoted 9 lines · Mathieu Poirier
Quoted 6 lines · Hangyu Hua

rpmsg_register_device_override need to call put_device to free vch when driver_set_override fails.

Fix this by adding a put_device() to the error path.

Fixes bb17d110cbf2 ("rpmsg: Fix calling device_lock() on non-initialized device")

This is funny... Neither Bjorn nor I have reviewed this patch...

Quoted 7 lines
Quoted 4 lines

It was a fix for commit in Greg's tree and Greg's pick it up after a week or something. I am not sure if that's actually funny that Greg has to pick it up without review :(

The patch was sent out on April 19th and committed 3 days later on April 22nd. Is 3 days the new patch review time standard?

Neither 19th, nor 22nd are correct. The patch which you set you never reviewed, so commit bb17d110cbf2 was sent on 29th of April:
https://lore.kernel.org/all/[email protected]/

And committed on 6 of May, which gives some time for review. Where did you see the other dates?

Best regards,
Krzysztof

#5 lore raw
Quoted 23 lines · Krzysztof Kozlowski
Quoted 14 lines · Mathieu Poirier
Quoted 12 lines · Krzysztof Kozlowski
Quoted 9 lines · Mathieu Poirier
Quoted 6 lines · Hangyu Hua

rpmsg_register_device_override need to call put_device to free vch when driver_set_override fails.

Fix this by adding a put_device() to the error path.

Fixes bb17d110cbf2 ("rpmsg: Fix calling device_lock() on non-initialized device")

This is funny... Neither Bjorn nor I have reviewed this patch...

Quoted 7 lines
Quoted 4 lines

It was a fix for commit in Greg's tree and Greg's pick it up after a week or something. I am not sure if that's actually funny that Greg has to pick it up without review :(

The patch was sent out on April 19th and committed 3 days later on April 22nd. Is 3 days the new patch review time standard?

Quoted 6 lines

Neither 19th, nor 22nd are correct. The patch which you set you never reviewed, so commit bb17d110cbf2 was sent on 29th of April:
https://lore.kernel.org/all/[email protected]/

And committed on 6 of May, which gives some time for review. Where did you see the other dates?

Maybe you refer to much earlier commit, not mentioned here which was on the list since February and committed indeed around April 22, so it gave around 2 months of time for review. In the meantime the patch was changing although since v6 (beginning of April [1]) it was almost untouched, except one string change.

Therefore this original set, not one being mentioned here bb17d110cbf2, was quite available. And you were Cc-ed since beginning - the version 2 this year (February, [2]). Two months is usually quite enough for review, especially for not so big diff.

[1]
https://lore.kernel.org/all/[email protected]/

[2]
https://lore.kernel.org/all/[email protected]/

Best regards,
Krzysztof

maintainer
#6 lore raw
In reply to Krzysztof Kozlowski
Quoted 20 lines · Krzysztof Kozlowski
Quoted 19 lines · Mathieu Poirier
Quoted 16 lines · Krzysztof Kozlowski
Quoted 9 lines · Mathieu Poirier
Quoted 6 lines · Hangyu Hua

rpmsg_register_device_override need to call put_device to free vch when driver_set_override fails.

Fix this by adding a put_device() to the error path.

Fixes bb17d110cbf2 ("rpmsg: Fix calling device_lock() on non-initialized device")

This is funny... Neither Bjorn nor I have reviewed this patch...

It was a fix for commit in Greg's tree and Greg's pick it up after a week or something. I am not sure if that's actually funny that Greg has to pick it up without review :(

Quoted 7 lines
Quoted 2 lines

The patch was sent out on April 19th and committed 3 days later on April 22nd. Is 3 days the new patch review time standard?

Neither 19th, nor 22nd are correct. The patch which you set you never reviewed, so commit bb17d110cbf2 was sent on 29th of April:
https://lore.kernel.org/all/[email protected]/

Twitchy fingers... Those dates are for commit 42cd402b8fd4, which is referenced by bb17d110cbf2.

The end result is the same, that is patches related to the remoteproc/rpmsg subsystems (or any subsystem) should not be committed before their maintainers have the opportunity to review them.

Quoted 6 lines

And committed on 6 of May, which gives some time for review. Where did you see the other dates?

Best regards,
Krzysztof

#7 lore raw
Quoted 29 lines · Mathieu Poirier
Quoted 20 lines · Krzysztof Kozlowski
Quoted 19 lines · Mathieu Poirier
Quoted 16 lines · Krzysztof Kozlowski
Quoted 9 lines · Mathieu Poirier
Quoted 6 lines · Hangyu Hua

rpmsg_register_device_override need to call put_device to free vch when driver_set_override fails.

Fix this by adding a put_device() to the error path.

Fixes bb17d110cbf2 ("rpmsg: Fix calling device_lock() on non-initialized device")

This is funny... Neither Bjorn nor I have reviewed this patch...

It was a fix for commit in Greg's tree and Greg's pick it up after a week or something. I am not sure if that's actually funny that Greg has to pick it up without review :(

Quoted 7 lines
Quoted 2 lines

The patch was sent out on April 19th and committed 3 days later on April 22nd. Is 3 days the new patch review time standard?

Neither 19th, nor 22nd are correct. The patch which you set you never reviewed, so commit bb17d110cbf2 was sent on 29th of April:
https://lore.kernel.org/all/[email protected]/

Quoted 6 lines

Twitchy fingers... Those dates are for commit 42cd402b8fd4, which is referenced by bb17d110cbf2.

The end result is the same, that is patches related to the remoteproc/rpmsg subsystems (or any subsystem) should not be committed before their maintainers have the opportunity to review them.

I think two months for your involvement was enough. During this two months there was a comment from Bjorn, applied and later quite plenty of time to revise/check/review.

I understand that we are all busy, pretty often working on upstream in spare time etc. So no one here complained that you did not review this patch within two months. But I don't understand what shall we do more if two months are not enough - wait four months? Six months?

Best regards,
Krzysztof