| Message ID | 1505921582-26709-5-git-send-email-axboe@kernel.dk (mailing list archive) |
|---|---|
| State | New, archived |
| Headers | show |
On Wed, Sep 20, 2017 at 09:32:59AM -0600, Jens Axboe wrote: > Laptop mode really wants to writeback the number of dirty > pages and inodes. Instead of calculating this in the caller, > just pass in 0 and let wakeup_flusher_threads() handle it. > > Use the new wakeup_flusher_threads_bdi() instead of rolling > our own. This changes the writeback to not be range cyclic, > but that should not matter for laptop mode flush-all > semantics. Looks good, Reviewed-by: Christoph Hellwig <hch@lst.de> While we're at sorting out the laptop_mode_wb_timer mess: can we move initializing and deleting it from the block code to the backing-dev code given that it now doesn't assume anything about block devices any more?
On Wed, Sep 20, 2017 at 09:32:59AM -0600, Jens Axboe wrote: > Use the new wakeup_flusher_threads_bdi() instead of rolling > our own. This changes the writeback to not be range cyclic, > but that should not matter for laptop mode flush-all > semantics. Oh btw - I think this actually is the more important change and should probably be in the subject line.
On Wed 20-09-17 09:32:59, Jens Axboe wrote: > Laptop mode really wants to writeback the number of dirty > pages and inodes. Instead of calculating this in the caller, > just pass in 0 and let wakeup_flusher_threads() handle it. > > Use the new wakeup_flusher_threads_bdi() instead of rolling > our own. This changes the writeback to not be range cyclic, > but that should not matter for laptop mode flush-all > semantics. > > Acked-by: Johannes Weiner <hannes@cmpxchg.org> > Tested-by: Chris Mason <clm@fb.com> > Signed-off-by: Jens Axboe <axboe@kernel.dk> > Reviewed-by: Jan Kara <jack@suse.cz> > Signed-off-by: Jens Axboe <axboe@kernel.dk> The subject of the patch looks stale. Honza
On 09/21/2017 08:59 AM, Christoph Hellwig wrote: > On Wed, Sep 20, 2017 at 09:32:59AM -0600, Jens Axboe wrote: >> Laptop mode really wants to writeback the number of dirty >> pages and inodes. Instead of calculating this in the caller, >> just pass in 0 and let wakeup_flusher_threads() handle it. >> >> Use the new wakeup_flusher_threads_bdi() instead of rolling >> our own. This changes the writeback to not be range cyclic, >> but that should not matter for laptop mode flush-all >> semantics. > > Looks good, > > Reviewed-by: Christoph Hellwig <hch@lst.de> > > While we're at sorting out the laptop_mode_wb_timer mess: > can we move initializing and deleting it from the block code > to the backing-dev code given that it now doesn't assume anything > about block devices any more? Good point, I'll include that in a followup for 4.15.
diff --git a/mm/page-writeback.c b/mm/page-writeback.c index 0b9c5cbe8eba..8d1fc593bce8 100644 --- a/mm/page-writeback.c +++ b/mm/page-writeback.c @@ -1980,23 +1980,8 @@ int dirty_writeback_centisecs_handler(struct ctl_table *table, int write, void laptop_mode_timer_fn(unsigned long data) { struct request_queue *q = (struct request_queue *)data; - int nr_pages = global_node_page_state(NR_FILE_DIRTY) + - global_node_page_state(NR_UNSTABLE_NFS); - struct bdi_writeback *wb; - /* - * We want to write everything out, not just down to the dirty - * threshold - */ - if (!bdi_has_dirty_io(q->backing_dev_info)) - return; - - rcu_read_lock(); - list_for_each_entry_rcu(wb, &q->backing_dev_info->wb_list, bdi_node) - if (wb_has_dirty_io(wb)) - wb_start_writeback(wb, nr_pages, true, - WB_REASON_LAPTOP_TIMER); - rcu_read_unlock(); + wakeup_flusher_threads_bdi(q->backing_dev_info, WB_REASON_LAPTOP_TIMER); } /*