Re: [PATCH v2 2/2] vsock: fix possible infinite sleep in vsock_connectible_wait_data()

Hi Dexuan, Stefano,

Tested-by: Frédéric Dalleau <redacted>

Regards,
Frédéric


On Wed, Nov 2, 2022 at 10:42 AM Stefano Garzarella [off-list ref] wrote:

On Wed, Nov 02, 2022 at 10:31:37AM +0100, Stefano Garzarella wrote:

quoted

On Mon, Oct 31, 2022 at 07:17:06PM -0700, Dexuan Cui wrote:

quoted

Currently vsock_connectible_has_data() may miss a wakeup operation
between vsock_connectible_has_data() == 0 and the prepare_to_wait().

Fix the race by adding the process to the wait queue before checking
vsock_connectible_has_data().

Fixes: b3f7fd54881b ("af_vsock: separate wait data loop")
Signed-off-by: Dexuan Cui <decui@microsoft.com>
---

Changes in v2 (Thanks Stefano!):
Fixed a typo in the commit message.
Removed the unnecessary finish_wait() at the end of the loop.

LGTM:

Reviewed-by: Stefano Garzarella <sgarzare@redhat.com>

And I would add

Reported-by: Frédéric Dalleau <redacted>

Since Frédéric posted a similar patch some months ago (I lost it because
netdev and I were not in cc):
https://lore.kernel.org/virtualization/20220824074251.2336997-2-frederic.dalleau@docker.com/ (local)

Thanks,
Stefano