On 10/21/21 9:18 AM, Jeff Moyer wrote:

Jens Axboe [off-list ref] writes:

quoted

On 10/21/21 8:42 AM, Christoph Hellwig wrote:

quoted

On Thu, Oct 21, 2021 at 08:34:38AM -0600, Jens Axboe wrote:

quoted

Incremental, are you happy with that comment?

Looks fine to me.

OK good, can I add your ack/review? I can send out a v3 if needed, but
seems a bit pointless for that small change.

Jeff, are you happy with this one too?

quoted

diff --git a/drivers/block/loop.c b/drivers/block/loop.c
index 397bfafc4c25..66c6e0c5d638 100644
--- a/drivers/block/loop.c
+++ b/drivers/block/loop.c

@@ -550,7 +550,7 @@ static void lo_rw_aio_do_completion(struct loop_cmd *cmd)
 		blk_mq_complete_request(rq);
 }
 
-static void lo_rw_aio_complete(struct kiocb *iocb, long ret, long ret2)
+static void lo_rw_aio_complete(struct kiocb *iocb, u64 ret)
 {
 	struct loop_cmd *cmd = container_of(iocb, struct loop_cmd, iocb);
 

@@ -623,7 +623,7 @@ static int lo_rw_aio(struct loop_device *lo, struct loop_cmd *cmd,
 	lo_rw_aio_do_completion(cmd);
 
 	if (ret != -EIOCBQUEUED)
-		cmd->iocb.ki_complete(&cmd->iocb, ret, 0);
+		lo_rw_aio_complete(&cmd->iocb, ret);
 	return 0;

I'm not sure why that was part of this patch, but I think it's fine.

Yeah, just came across that one, I can drop this part and make it a
separate patch. Just a bit dumb to not use the function rather than the
indirect call, though maybe the compiler figures it out.

I've still got more testing to do, but you can add:

Reviewed-by: Jeff Moyer <redacted>

I'll follow up if there are issues.

Thanks!

-- 
Jens Axboe