On Wed, Feb 08, 2023 at 05:03:30AM +0000, Shinichiro Kawasaki wrote: > On Feb 07, 2023 / 14:06, Niklas Cassel wrote: > > On Tue, Feb 07, 2023 at 03:37:38PM +0900, Shin'ichiro Kawasaki wrote: > > > The valid data bytes accounting field is initialized at file reset, > > > after each job started. Each job locks zones to check write pointer > > > positions of its write target zones. This can cause zone lock contention > > > with write by other jobs. > > > > > > To avoid the zone lock contention, move the initialization from file > > > reset to file set up before job start. It allows to access the write > > > pointers and the accounting field without locks. Remove the lock and > > > unlock codes which are no longer required. Ensure the locks are not > > > required by checking run-state in the struct thread_data. > > > > > > Signed-off-by: Shin'ichiro Kawasaki <shinichiro.kawasaki@xxxxxxx> > > > --- (snip) > > > @@ -1252,10 +1251,6 @@ void zbd_file_reset(struct thread_data *td, struct fio_file *f) > > > > > > zb = zbd_get_zone(f, f->min_zone); > > > ze = zbd_get_zone(f, f->max_zone); > > > - vdb = zbd_set_vdb(td, f); > > The zbd_set_vbd() call in zbd_file_reset() is removed here. Then the check is > done only at zbd_setup_files() as you expect. Sorry, I missed that you also removed the only other place that called this function. zbd_setup_files() calls functions that are named e.g.: if (!zbd_verify_sizes()) return 1; if (!zbd_verify_bs()) return 1; Perhaps consider renaming this function to something like zbd_verify_and_set_vbd(), since this function is now only called from zbd_setup_files(). Regardless if you rename or not: Reviewed-by: Niklas Cassel <niklas.cassel@xxxxxxx>
