On Fri, Sep 29, 2017 at 11:39:03PM +0200, Jens Axboe wrote:
> On 09/29/2017 07:09 PM, weiping zhang wrote:
> > add an option that disable io scheduler for null block device.
> >
> > Signed-off-by: weiping zhang <zhangweip...@didichuxing.com>
> > ---
> > drivers/block/null_blk.c | 6 +++++-
> > 1 file changed, 5 insertions(+), 1 deletion(-)
> >
> > diff --git a/drivers/block/null_blk.c b/drivers/block/null_blk.c
> > index bd92286..3c63863 100644
> > --- a/drivers/block/null_blk.c
> > +++ b/drivers/block/null_blk.c
> > @@ -154,6 +154,10 @@ enum {
> > NULL_Q_MQ = 2,
> > };
> >
> > +static int g_no_sched;
> > +module_param_named(no_sched, g_no_sched, int, S_IRUGO);
> > +MODULE_PARM_DESC(no_sched, "No io scheduler");
> > +
> > static int g_submit_queues = 1;
> > module_param_named(submit_queues, g_submit_queues, int, S_IRUGO);
> > MODULE_PARM_DESC(submit_queues, "Number of submission queues");
> > @@ -1753,7 +1757,7 @@ static int null_init_tag_set(struct nullb *nullb,
> > struct blk_mq_tag_set *set)
> > g_hw_queue_depth;
> > set->numa_node = nullb ? nullb->dev->home_node : g_home_node;
> > set->cmd_size = sizeof(struct nullb_cmd);
> > - set->flags = BLK_MQ_F_SHOULD_MERGE;
> > + set->flags = g_no_sched ? BLK_MQ_F_NO_SCHED : BLK_MQ_F_SHOULD_MERGE;
>
> This should be:
>
> set->flags = BLK_MQ_F_SHOULD_MERGE;
> if (g_no_sched)
> set->flags |= BLK_MQ_F_NO_SCHED;
>
That's right, I go through these two flags, if no io scheduler,
BLK_MQ_F_SHOULD_MERGE can make sw ctx merge happen. I will send V2.
Thanks
weiping
