| Message ID | 20230816195447.3703954-3-bvanassche@acm.org (mailing list archive) |
|---|---|
| State | Superseded |
| Headers | show |
| Series | Improve performance for zoned UFS devices | expand |
On 8/17/23 04:53, Bart Van Assche wrote: > Make blk_req_needs_zone_write_lock() return false if > q->limits.use_zone_write_lock is false. Inline this function because it > is short and because it is called from the hot path of the mq-deadline > I/O scheduler. Your are not actually inlining the function. Did you forget to move it to blkdev.h ? Otherwise, the change looks OK. > > Cc: Damien Le Moal <dlemoal@kernel.org> > Cc: Christoph Hellwig <hch@lst.de> > Cc: Ming Lei <ming.lei@redhat.com> > Signed-off-by: Bart Van Assche <bvanassche@acm.org> > --- > block/blk-zoned.c | 9 +++++++-- > 1 file changed, 7 insertions(+), 2 deletions(-) > > diff --git a/block/blk-zoned.c b/block/blk-zoned.c > index 112620985bff..d8a80cce832f 100644 > --- a/block/blk-zoned.c > +++ b/block/blk-zoned.c > @@ -53,11 +53,16 @@ const char *blk_zone_cond_str(enum blk_zone_cond zone_cond) > EXPORT_SYMBOL_GPL(blk_zone_cond_str); > > /* > - * Return true if a request is a write requests that needs zone write locking. > + * Return true if a request is a write request that needs zone write locking. > */ > bool blk_req_needs_zone_write_lock(struct request *rq) > { > - if (!rq->q->disk->seq_zones_wlock) > + struct request_queue *q = rq->q; > + > + if (!q->limits.use_zone_write_lock) > + return false; > + > + if (!q->disk->seq_zones_wlock) > return false; > > return blk_rq_is_seq_zoned_write(rq);
On 8/17/23 04:01, Damien Le Moal wrote: > On 8/17/23 04:53, Bart Van Assche wrote: >> Make blk_req_needs_zone_write_lock() return false if >> q->limits.use_zone_write_lock is false. Inline this function because it >> is short and because it is called from the hot path of the mq-deadline >> I/O scheduler. > > Your are not actually inlining the function. Did you forget to move it to > blkdev.h ? I considered inlining but eventually decided not to inline blk_req_needs_zone_write_lock(). I will update the patch description. Thanks, Bart.
diff --git a/block/blk-zoned.c b/block/blk-zoned.c index 112620985bff..d8a80cce832f 100644 --- a/block/blk-zoned.c +++ b/block/blk-zoned.c @@ -53,11 +53,16 @@ const char *blk_zone_cond_str(enum blk_zone_cond zone_cond) EXPORT_SYMBOL_GPL(blk_zone_cond_str); /* - * Return true if a request is a write requests that needs zone write locking. + * Return true if a request is a write request that needs zone write locking. */ bool blk_req_needs_zone_write_lock(struct request *rq) { - if (!rq->q->disk->seq_zones_wlock) + struct request_queue *q = rq->q; + + if (!q->limits.use_zone_write_lock) + return false; + + if (!q->disk->seq_zones_wlock) return false; return blk_rq_is_seq_zoned_write(rq);
Make blk_req_needs_zone_write_lock() return false if q->limits.use_zone_write_lock is false. Inline this function because it is short and because it is called from the hot path of the mq-deadline I/O scheduler. Cc: Damien Le Moal <dlemoal@kernel.org> Cc: Christoph Hellwig <hch@lst.de> Cc: Ming Lei <ming.lei@redhat.com> Signed-off-by: Bart Van Assche <bvanassche@acm.org> --- block/blk-zoned.c | 9 +++++++-- 1 file changed, 7 insertions(+), 2 deletions(-)