Re: [Devel] [PATCH v3 rh7 2/2] mm/page-writeback: Introduce per-CT dirty memory limit.

2016-01-21 Thread Vladimir Davydov
On Thu, Jan 21, 2016 at 01:55:32PM +0300, Andrey Ryabinin wrote: > > > On 01/21/2016 01:35 PM, Vladimir Davydov wrote: > >> + if (ub_dirty + ub_writeback <= ub_thresh) > >> + break; > >> + if (!writeback_in_progress(bdi)) > >> + bdi_start_backgr

Re: [Devel] [PATCH v3 rh7 2/2] mm/page-writeback: Introduce per-CT dirty memory limit.

2016-01-21 Thread Andrey Ryabinin
On 01/21/2016 01:35 PM, Vladimir Davydov wrote: >> +if (ub_dirty + ub_writeback <= ub_thresh) >> +break; >> +if (!writeback_in_progress(bdi)) >> +bdi_start_background_writeback(bdi); >> + >> +/* >> + * Throttl

Re: [Devel] [PATCH v3 rh7 2/2] mm/page-writeback: Introduce per-CT dirty memory limit.

2016-01-21 Thread Vladimir Davydov
On Thu, Jan 21, 2016 at 11:44:14AM +0300, Andrey Ryabinin wrote: > @@ -1391,6 +1391,101 @@ static inline void bdi_dirty_limits(struct > backing_dev_info *bdi, > } > } > > +static void balance_dirty_pages_ub(struct address_space *mapping, > + unsigned long writ

[Devel] [PATCH v3 rh7 2/2] mm/page-writeback: Introduce per-CT dirty memory limit.

2016-01-21 Thread Andrey Ryabinin
This patch adds balance_dirty_pages_ub() which is implements per-UB logic of balance_dirty_pages() from PCS6. balance_dirty_pages_ub() is invoked right before the original balance_dirty_pages() to issue per-UB writeback. After that balance_dirty_pages() will be executed for global writeback. https