Re: [RFC] NUMA balancing: reduce TLB flush via delaying mapping on hint page fault

From: Mel Gorman <mgorman@suse.de>
To: Nadav Amit <nadav.amit@gmail.com>
Cc: "Huang, Ying" <ying.huang@intel.com>,
	Linux-MM <linux-mm@kvack.org>,
	Andrew Morton <akpm@linux-foundation.org>,
	LKML <linux-kernel@vger.kernel.org>,
	Peter Zijlstra <peterz@infradead.org>,
	Peter Xu <peterx@redhat.com>,
	Johannes Weiner <hannes@cmpxchg.org>,
	Vlastimil Babka <vbabka@suse.cz>,
	Matthew Wilcox <willy@infradead.org>,
	Will Deacon <will@kernel.org>,
	Michel Lespinasse <walken@google.com>,
	Arjun Roy <arjunroy@google.com>,
	"Kirill A. Shutemov" <kirill.shutemov@linux.intel.com>
Subject: Re: [RFC] NUMA balancing: reduce TLB flush via delaying mapping on hint page fault
Date: Thu, 1 Apr 2021 09:38:09 +0100	[thread overview]
Message-ID: <20210401083809.GX15768@suse.de> (raw)
In-Reply-To: <C42F16F8-646B-448D-A098-E1A98637E28D@gmail.com>

[-- Attachment #1: Type: text/plain, Size: 2045 bytes --]

On Wed, Mar 31, 2021 at 09:36:04AM -0700, Nadav Amit wrote:
> 
> 
> > On Mar 31, 2021, at 6:16 AM, Mel Gorman <mgorman@suse.de> wrote:
> > 
> > On Wed, Mar 31, 2021 at 07:20:09PM +0800, Huang, Ying wrote:
> >> Mel Gorman <mgorman@suse.de> writes:
> >> 
> >>> On Mon, Mar 29, 2021 at 02:26:51PM +0800, Huang Ying wrote:
> >>>> For NUMA balancing, in hint page fault handler, the faulting page will
> >>>> be migrated to the accessing node if necessary.  During the migration,
> >>>> TLB will be shot down on all CPUs that the process has run on
> >>>> recently.  Because in the hint page fault handler, the PTE will be
> >>>> made accessible before the migration is tried.  The overhead of TLB
> >>>> shooting down is high, so it's better to be avoided if possible.  In
> >>>> fact, if we delay mapping the page in PTE until migration, that can be
> >>>> avoided.  This is what this patch doing.
> >>>> 
> >>> 
> >>> Why would the overhead be high? It was previously inaccessibly so it's
> >>> only parallel accesses making forward progress that trigger the need
> >>> for a flush.
> >> 
> >> Sorry, I don't understand this.  Although the page is inaccessible, the
> >> threads may access other pages, so TLB flushing is still necessary.
> >> 
> > 
> > You assert the overhead of TLB shootdown is high and yes, it can be
> > very high but you also said "the benchmark score has no visible changes"
> > indicating the TLB shootdown cost is not a major problem for the workload.
> > It does not mean we should ignore it though.
> 
> If you are looking for a benchmark that is negatively affected by NUMA
> balancing, then IIRC Parsec???s dedup is such a workload. [1]
> 

Few questions;

Is Parsec imparied due to NUMA balancing in general or due to TLB
shootdowns specifically?

Are you using "gcc-pthreads" for parallelisation and the "native" size
for Parsec?

Is there any specific thread count that matters either in
absolute terms or as a precentage of online CPUs?

-- 
Mel Gorman
SUSE Labs

[-- Attachment #2: Digital signature --]
[-- Type: application/pgp-signature, Size: 488 bytes --]