Re: [Xen-devel] [PATCH 2/2] grant_table: convert grant table rwlock to percpu rwlock

On 18/11/15 20:02, Konrad Rzeszutek Wilk wrote: > On Tue, Nov 17, 2015 at 05:30:59PM +, Andrew Cooper wrote: >> On 17/11/15 17:04, Jan Beulich wrote: >> On 03.11.15 at 18:58, wrote: --- a/xen/common/grant_table.c +++ b/xen/common/grant_table.c

Re: [Xen-devel] [PATCH 2/2] grant_table: convert grant table rwlock to percpu rwlock

On 18/11/15 20:02, Konrad Rzeszutek Wilk wrote: > On Tue, Nov 17, 2015 at 05:30:59PM +, Andrew Cooper wrote: >> On 17/11/15 17:04, Jan Beulich wrote: >> On 03.11.15 at 18:58, wrote: --- a/xen/common/grant_table.c +++ b/xen/common/grant_table.c

Re: [Xen-devel] [PATCH 2/2] grant_table: convert grant table rwlock to percpu rwlock

On Tue, Nov 17, 2015 at 05:30:59PM +, Andrew Cooper wrote: > On 17/11/15 17:04, Jan Beulich wrote: > On 03.11.15 at 18:58, wrote: > >> --- a/xen/common/grant_table.c > >> +++ b/xen/common/grant_table.c > >> @@ -178,6 +178,10 @@ struct active_grant_entry { >

Re: [Xen-devel] [PATCH 2/2] grant_table: convert grant table rwlock to percpu rwlock

On 18/11/15 12:07, Ian Campbell wrote: > On Wed, 2015-11-18 at 11:56 +, Malcolm Crossley wrote: >> On 18/11/15 11:50, Ian Campbell wrote: >>> On Wed, 2015-11-18 at 11:23 +, Malcolm Crossley wrote: On 18/11/15 10:54, Jan Beulich wrote: On 18.11.15 at 11:36,

Re: [Xen-devel] [PATCH 2/2] grant_table: convert grant table rwlock to percpu rwlock

On 18/11/15 10:54, Jan Beulich wrote: On 18.11.15 at 11:36, wrote: >> On Tue, 2015-11-17 at 17:53 +, Andrew Cooper wrote: >>> On 17/11/15 17:39, Jan Beulich wrote: >>> On 17.11.15 at 18:30, wrote: > On 17/11/15 17:04, Jan

Re: [Xen-devel] [PATCH 2/2] grant_table: convert grant table rwlock to percpu rwlock

On 18/11/15 11:41, Jan Beulich wrote: On 18.11.15 at 12:23, wrote: >> On 18/11/15 10:54, Jan Beulich wrote: >>> That's not how I understood it, the rwlock isn't per-pCPU (at least not >>> in what this patch does - it remains a per-domain one). The per-pCPU >>>

Re: [Xen-devel] [PATCH 2/2] grant_table: convert grant table rwlock to percpu rwlock

On Wed, 2015-11-18 at 11:23 +, Malcolm Crossley wrote: > On 18/11/15 10:54, Jan Beulich wrote: > > > > > On 18.11.15 at 11:36, wrote: > > > On Tue, 2015-11-17 at 17:53 +, Andrew Cooper wrote: > > > > On 17/11/15 17:39, Jan Beulich wrote: > > > > > > > > On

Re: [Xen-devel] [PATCH 2/2] grant_table: convert grant table rwlock to percpu rwlock

On Tue, 2015-11-17 at 17:53 +, Andrew Cooper wrote: > On 17/11/15 17:39, Jan Beulich wrote: > > > > > On 17.11.15 at 18:30, wrote: > > > On 17/11/15 17:04, Jan Beulich wrote: > > > > > > > On 03.11.15 at 18:58, wrote: > > > > > ---

Re: [Xen-devel] [PATCH 2/2] grant_table: convert grant table rwlock to percpu rwlock

>>> On 18.11.15 at 11:06, wrote: > On 18/11/15 07:45, Jan Beulich wrote: > On 17.11.15 at 18:53, wrote: >>> On 17/11/15 17:39, Jan Beulich wrote: >>> On 17.11.15 at 18:30, wrote: > On 17/11/15 17:04,

Re: [Xen-devel] [PATCH 2/2] grant_table: convert grant table rwlock to percpu rwlock

On 18/11/15 11:50, Ian Campbell wrote: > On Wed, 2015-11-18 at 11:23 +, Malcolm Crossley wrote: >> On 18/11/15 10:54, Jan Beulich wrote: >> On 18.11.15 at 11:36, wrote: On Tue, 2015-11-17 at 17:53 +, Andrew Cooper wrote: > On 17/11/15 17:39, Jan

Re: [Xen-devel] [PATCH 2/2] grant_table: convert grant table rwlock to percpu rwlock

On Wed, 2015-11-18 at 11:56 +, Malcolm Crossley wrote: > On 18/11/15 11:50, Ian Campbell wrote: > > On Wed, 2015-11-18 at 11:23 +, Malcolm Crossley wrote: > > > On 18/11/15 10:54, Jan Beulich wrote: > > > > > > > On 18.11.15 at 11:36, wrote: > > > > > On Tue,

Re: [Xen-devel] [PATCH 2/2] grant_table: convert grant table rwlock to percpu rwlock

On Wed, 2015-11-18 at 13:08 +, Malcolm Crossley wrote: > On 18/11/15 12:07, Ian Campbell wrote: > > On Wed, 2015-11-18 at 11:56 +, Malcolm Crossley wrote: > > > On 18/11/15 11:50, Ian Campbell wrote: > > > > On Wed, 2015-11-18 at 11:23 +, Malcolm Crossley wrote: > > > > > On 18/11/15

Re: [Xen-devel] [PATCH 2/2] grant_table: convert grant table rwlock to percpu rwlock

>>> On 18.11.15 at 14:08, wrote: > On 18/11/15 12:07, Ian Campbell wrote: >> (Maybe I should have said "like a bit lock with 32 or 64 bits, setting any >> of which corresponds to acquiring the lock" ;-)) >> > Not quite, setting the per cpu read area "takes" the read

Re: [Xen-devel] [PATCH 2/2] grant_table: convert grant table rwlock to percpu rwlock

>>> On 03.11.15 at 18:58, wrote: > --- a/xen/common/grant_table.c > +++ b/xen/common/grant_table.c > @@ -178,6 +178,10 @@ struct active_grant_entry { > #define _active_entry(t, e) \ > ((t)->active[(e)/ACGNT_PER_PAGE][(e)%ACGNT_PER_PAGE]) > > +bool_t

Re: [Xen-devel] [PATCH 2/2] grant_table: convert grant table rwlock to percpu rwlock

On 17/11/15 17:04, Jan Beulich wrote: On 03.11.15 at 18:58, wrote: >> --- a/xen/common/grant_table.c >> +++ b/xen/common/grant_table.c >> @@ -178,6 +178,10 @@ struct active_grant_entry { >> #define _active_entry(t, e) \ >>

Re: [Xen-devel] [PATCH 2/2] grant_table: convert grant table rwlock to percpu rwlock

On 17/11/15 17:39, Jan Beulich wrote: On 17.11.15 at 18:30, wrote: >> On 17/11/15 17:04, Jan Beulich wrote: >> On 03.11.15 at 18:58, wrote: --- a/xen/common/grant_table.c +++ b/xen/common/grant_table.c @@ -178,6

Re: [Xen-devel] [PATCH 2/2] grant_table: convert grant table rwlock to percpu rwlock

>>> On 17.11.15 at 18:30, wrote: > On 17/11/15 17:04, Jan Beulich wrote: > On 03.11.15 at 18:58, wrote: >>> --- a/xen/common/grant_table.c >>> +++ b/xen/common/grant_table.c >>> @@ -178,6 +178,10 @@ struct active_grant_entry { >>>

Re: [Xen-devel] [PATCH 2/2] grant_table: convert grant table rwlock to percpu rwlock

>>> On 17.11.15 at 18:53, wrote: > On 17/11/15 17:39, Jan Beulich wrote: > On 17.11.15 at 18:30, wrote: >>> On 17/11/15 17:04, Jan Beulich wrote: >>> On 03.11.15 at 18:58, wrote: > ---

[Xen-devel] [PATCH 2/2] grant_table: convert grant table rwlock to percpu rwlock

The per domain grant table read lock suffers from significant contention when performance multi-queue block or network IO due to the parallel grant map/unmaps/copies occuring on the DomU's grant table. On multi-socket systems, the contention results in the locked compare swap operation failing