Re: [RFC-PROTOTYPE 1/1] mm: Add __GFP_FAST_TRY flag

2020-08-04 Thread Uladzislau Rezki
On Tue, Aug 04, 2020 at 07:34:18PM +0200, Vlastimil Babka wrote: > On 8/4/20 7:12 PM, Matthew Wilcox wrote: > > On Tue, Aug 04, 2020 at 07:02:14PM +0200, Vlastimil Babka wrote: > >> > 2) There was a proposal from Matthew Wilcox: > >> > https://lkml.org/lkml/2020/7/31/1015 > >> > > >> > > >> >

Re: [RFC-PROTOTYPE 1/1] mm: Add __GFP_FAST_TRY flag

2020-08-04 Thread Uladzislau Rezki
On Tue, Aug 04, 2020 at 06:12:03PM +0100, Matthew Wilcox wrote: > On Tue, Aug 04, 2020 at 07:02:14PM +0200, Vlastimil Babka wrote: > > > 2) There was a proposal from Matthew Wilcox: > > > https://lkml.org/lkml/2020/7/31/1015 > > > > > > > > > On non-RT, we could make that lock a raw spinlock.

Re: [RFC-PROTOTYPE 1/1] mm: Add __GFP_FAST_TRY flag

2020-08-04 Thread Uladzislau Rezki
On Tue, Aug 04, 2020 at 07:02:14PM +0200, Vlastimil Babka wrote: > On 8/3/20 6:30 PM, Uladzislau Rezki (Sony) wrote: > > Some background and kfree_rcu() > > === > > The pointers to be freed are stored in the per-cpu array to improve > > performance, to enable an

Re: [RFC-PROTOTYPE 1/1] mm: Add __GFP_FAST_TRY flag

2020-08-04 Thread Vlastimil Babka
On 8/4/20 7:12 PM, Matthew Wilcox wrote: > On Tue, Aug 04, 2020 at 07:02:14PM +0200, Vlastimil Babka wrote: >> > 2) There was a proposal from Matthew Wilcox: >> > https://lkml.org/lkml/2020/7/31/1015 >> > >> > >> > On non-RT, we could make that lock a raw spinlock. On RT, we could >> > decline

Re: [RFC-PROTOTYPE 1/1] mm: Add __GFP_FAST_TRY flag

2020-08-04 Thread Matthew Wilcox
On Tue, Aug 04, 2020 at 07:02:14PM +0200, Vlastimil Babka wrote: > > 2) There was a proposal from Matthew Wilcox: > > https://lkml.org/lkml/2020/7/31/1015 > > > > > > On non-RT, we could make that lock a raw spinlock. On RT, we could > > decline to take the lock. We'd need to abstract the

Re: [RFC-PROTOTYPE 1/1] mm: Add __GFP_FAST_TRY flag

2020-08-04 Thread Vlastimil Babka
On 8/3/20 6:30 PM, Uladzislau Rezki (Sony) wrote: > Some background and kfree_rcu() > === > The pointers to be freed are stored in the per-cpu array to improve > performance, to enable an easier-to-use API, to accommodate vmalloc > memmory and to support a single

[RFC-PROTOTYPE 1/1] mm: Add __GFP_FAST_TRY flag

2020-08-03 Thread Uladzislau Rezki (Sony)
Some background and kfree_rcu() === The pointers to be freed are stored in the per-cpu array to improve performance, to enable an easier-to-use API, to accommodate vmalloc memmory and to support a single argument of the kfree_rcu() when only a pointer is passed. More