On 05/05/2025 8:51, Greg Kroah-Hartman wrote:
On Sun, May 04, 2025 at 04:47:20PM +0300, Jared Holzman wrote:
On 04/05/2025 15:39, Greg Kroah-Hartman wrote:
On Sun, May 04, 2025 at 02:55:00PM +0300, Jared Holzman wrote:
On 29/04/2025 19:38, Greg Kroah-Hartman wrote:
6.14-stable review patch. If anyone has any objections, please let me know.
From: Ming Lei ming.lei@redhat.com
[ Upstream commit d6aa0c178bf81f30ae4a780b2bca653daa2eb633 ]
We call io_uring_cmd_complete_in_task() to schedule task_work for handling UBLK_U_IO_NEED_GET_DATA.
This way is really not necessary because the current context is exactly the ublk queue context, so call ublk_dispatch_req() directly for handling UBLK_U_IO_NEED_GET_DATA.
Fixes: 216c8f5ef0f2 ("ublk: replace monitor with cancelable uring_cmd") Tested-by: Jared Holzman jholzman@nvidia.com Signed-off-by: Ming Lei ming.lei@redhat.com Link: https://lore.kernel.org/r/20250425013742.1079549-2-ming.lei@redhat.com Signed-off-by: Jens Axboe axboe@kernel.dk Signed-off-by: Sasha Levin sashal@kernel.org
drivers/block/ublk_drv.c | 14 +++----------- 1 file changed, 3 insertions(+), 11 deletions(-)
diff --git a/drivers/block/ublk_drv.c b/drivers/block/ublk_drv.c index 437297022dcfa..c7761a5cfeec0 100644 --- a/drivers/block/ublk_drv.c +++ b/drivers/block/ublk_drv.c @@ -1812,15 +1812,6 @@ static void ublk_mark_io_ready(struct ublk_device *ub, struct ublk_queue *ubq) mutex_unlock(&ub->mutex); } -static void ublk_handle_need_get_data(struct ublk_device *ub, int q_id,
int tag)-{
- struct ublk_queue *ubq = ublk_get_queue(ub, q_id);
- struct request *req = blk_mq_tag_to_rq(ub->tag_set.tags[q_id], tag);
- ublk_queue_cmd(ubq, req);
-}
static inline int ublk_check_cmd_op(u32 cmd_op) { u32 ioc_type = _IOC_TYPE(cmd_op); @@ -1967,8 +1958,9 @@ static int __ublk_ch_uring_cmd(struct io_uring_cmd *cmd, if (!(io->flags & UBLK_IO_FLAG_OWNED_BY_SRV)) goto out; ublk_fill_io_cmd(io, cmd, ub_cmd->addr);
ublk_handle_need_get_data(ub, ub_cmd->q_id, ub_cmd->tag);break;
req = blk_mq_tag_to_rq(ub->tag_set.tags[ub_cmd->q_id], tag);ublk_dispatch_req(ubq, req, issue_flags);return -EIOCBQUEUED;Hi Greg,
Will you also be backporting "ublk: fix race between io_uring_cmd_complete_in_task and ublk_cancel_cmd" to 6.14-stable?
What is the git commit id you are referring to? And was it asked to be included in a stable release?
thanks,
greg k-h
Hi Greg,
The commit is: f40139fde527
It is Part 2 of the same patch series.
It does not apply to the stable tree at all, so no, we will not be adding it unless someone provides a working version of it.
thanks,
greg k-h
Hi Greg,
Happy to provide a version that will apply. I just need to know where to get your working branch to base it on.
Please note that without this patch, our application causes kernel panic 1/20 times it is stopped so we are quite keen to get it into 6.14 stable
Regards,
Jared