On 10/25/21 10:16 AM, Darrick J. Wong wrote:

On Mon, Oct 25, 2021 at 09:33:02AM -0600, Jens Axboe wrote:

quoted

The second argument was only used by the USB gadget code, yet everyone
pays the overhead of passing a zero to be passed into aio, where it
ends up being part of the aio res2 value.

Now that everybody is passing in zero, kill off the extra argument.

Signed-off-by: Jens Axboe <axboe@kernel.dk>

The fs/ parts look reasonable to me.
Reviewed-by: Darrick J. Wong <djwong@kernel.org>

Thanks!

So, uh, do you get even more iops by avoiding the extra register? :)

Heh, well if you're register starved, it'll likely make a bigger
difference, but even on my CPU it is actually noticeable. Obviously not
a 10% gain, and the change is more about cleaning up the API a bit.

-- 
Jens Axboe