Message-ID: <YpXK6VO8y6ZQlimG@osiris>
Date:   Tue, 31 May 2022 09:59:37 +0200
From:   Heiko Carstens <hca@...ux.ibm.com>
To:     Peter Xu <peterx@...hat.com>
Cc:     linux-kernel@...r.kernel.org, linux-mm@...ck.org,
        Ingo Molnar <mingo@...hat.com>,
        Vasily Gorbik <gor@...ux.ibm.com>,
        linuxppc-dev@...ts.ozlabs.org, Stafford Horne <shorne@...il.com>,
        "David S . Miller" <davem@...emloft.net>,
        Johannes Berg <johannes@...solutions.net>,
        Brian Cain <bcain@...cinc.com>, x86@...nel.org,
        linux-parisc@...r.kernel.org, Richard Henderson <rth@...ddle.net>,
        Christian Borntraeger <borntraeger@...ux.ibm.com>,
        Richard Weinberger <richard@....at>,
        linux-hexagon@...r.kernel.org,
        Benjamin Herrenschmidt <benh@...nel.crashing.org>,
        Thomas Gleixner <tglx@...utronix.de>,
        Janosch Frank <frankja@...ux.ibm.com>,
        Albert Ou <aou@...s.berkeley.edu>,
        Anton Ivanov <anton.ivanov@...bridgegreys.com>,
        Dave Hansen <dave.hansen@...ux.intel.com>,
        linux-ia64@...r.kernel.org, Borislav Petkov <bp@...en8.de>,
        Russell King <linux@...linux.org.uk>,
        Sven Schnelle <svens@...ux.ibm.com>,
        Andrea Arcangeli <aarcange@...hat.com>,
        "James E . J . Bottomley" <James.Bottomley@...senpartnership.com>,
        Al Viro <viro@...iv.linux.org.uk>,
        Alexander Gordeev <agordeev@...ux.ibm.com>,
        Peter Zijlstra <peterz@...radead.org>,
        Alistair Popple <apopple@...dia.com>,
        Jonas Bonn <jonas@...thpole.se>, sparclinux@...r.kernel.org,
        linux-csky@...r.kernel.org, Will Deacon <will@...nel.org>,
        Andrew Morton <akpm@...ux-foundation.org>,
        Vlastimil Babka <vbabka@...e.cz>, linux-um@...ts.infradead.org,
        Michal Simek <monstr@...str.eu>,
        Matt Turner <mattst88@...il.com>,
        linux-m68k@...ts.linux-m68k.org, Paul Mackerras <paulus@...ba.org>,
        linux-xtensa@...ux-xtensa.org,
        Geert Uytterhoeven <geert@...ux-m68k.org>,
        David Hildenbrand <david@...hat.com>,
        openrisc@...ts.librecores.org,
        linux-arm-kernel@...ts.infradead.org,
        Nicholas Piggin <npiggin@...il.com>,
        Palmer Dabbelt <palmer@...belt.com>,
        Michael Ellerman <mpe@...erman.id.au>,
        Stefan Kristiansson <stefan.kristiansson@...nalahti.fi>,
        Paul Walmsley <paul.walmsley@...ive.com>,
        Ivan Kokshaysky <ink@...assic.park.msu.ru>,
        Chris Zankel <chris@...kel.net>,
        Hugh Dickins <hughd@...gle.com>,
        Vineet Gupta <vgupta@...nel.org>,
        Dinh Nguyen <dinguyen@...nel.org>,
        Catalin Marinas <catalin.marinas@....com>,
        Rich Felker <dalias@...c.org>,
        "H . Peter Anvin" <hpa@...or.com>, linux-s390@...r.kernel.org,
        linux-sh@...r.kernel.org, linux-riscv@...ts.infradead.org,
        Johannes Weiner <hannes@...xchg.org>,
        Andy Lutomirski <luto@...nel.org>,
        Max Filippov <jcmvbkbc@...il.com>, Guo Ren <guoren@...nel.org>,
        Thomas Bogendoerfer <tsbogend@...ha.franken.de>,
        linux-snps-arc@...ts.infradead.org, linux-alpha@...r.kernel.org,
        linux-mips@...r.kernel.org, Helge Deller <deller@....de>,
        Yoshinori Sato <ysato@...rs.sourceforge.jp>,
        Ingo Molnar <mingo@...nel.org>
Subject: Re: [PATCH v5] mm: Avoid unnecessary page fault retires on shared
 memory types

On Mon, May 30, 2022 at 02:34:50PM -0400, Peter Xu wrote:
> I observed that for each of the shared file-backed page faults, we're very
> likely to retry one more time for the 1st write fault upon no page.  It's
> because we'll need to release the mmap lock for dirty rate limit purpose
> with balance_dirty_pages_ratelimited() (in fault_dirty_shared_page()).
> 
> Then after that throttling we return VM_FAULT_RETRY.
> 
> We did that probably because VM_FAULT_RETRY is the only way we can return
> to the fault handler at that time telling it we've released the mmap lock.
> 
> However that's not ideal because it's very likely the fault does not need
> to be retried at all since the pgtable was well installed before the
> throttling, so the next continuous fault (including taking mmap read lock,
> walk the pgtable, etc.) could be in most cases unnecessary.
> 
> It's not only slowing down page faults for shared file-backed, but also add
> more mmap lock contention which is in most cases not needed at all.
> 
> To observe this, one could try to write to some shmem page and look at
> "pgfault" value in /proc/vmstat, then we should expect 2 counts for each
> shmem write simply because we retried, and vm event "pgfault" will capture
> that.
> 
> To make it more efficient, add a new VM_FAULT_COMPLETED return code just to
> show that we've completed the whole fault and released the lock.  It's also
> a hint that we should very possibly not need another fault immediately on
> this page because we've just completed it.
> 
> This patch provides a ~12% perf boost on my aarch64 test VM with a simple
> program sequentially dirtying 400MB shmem file being mmap()ed and these are
> the time it needs:
> 
>   Before: 650.980 ms (+-1.94%)
>   After:  569.396 ms (+-1.38%)
> 
> I believe it could help more than that.
> 
> We need some special care on GUP and the s390 pgfault handler (for gmap
> code before returning from pgfault), the rest changes in the page fault
> handlers should be relatively straightforward.
> 
> Another thing to mention is that mm_account_fault() does take this new
> fault as a generic fault to be accounted, unlike VM_FAULT_RETRY.
> 
> I explicitly didn't touch hmm_vma_fault() and break_ksm() because they do
> not handle VM_FAULT_RETRY even with existing code, so I'm literally keeping
> them as-is.
> 
> Acked-by: Geert Uytterhoeven <geert@...ux-m68k.org>
> Acked-by: Peter Zijlstra (Intel) <peterz@...radead.org>
> Acked-by: Johannes Weiner <hannes@...xchg.org>
> Acked-by: Vineet Gupta <vgupta@...nel.org>
> Acked-by: Guo Ren <guoren@...nel.org>
> Acked-by: Max Filippov <jcmvbkbc@...il.com>
> Acked-by: Christian Borntraeger <borntraeger@...ux.ibm.com>
> Acked-by: Michael Ellerman <mpe@...erman.id.au> (powerpc)
> Acked-by: Catalin Marinas <catalin.marinas@....com>
> Reviewed-by: Alistair Popple <apopple@...dia.com>
> Reviewed-by: Ingo Molnar <mingo@...nel.org>
> Signed-off-by: Peter Xu <peterx@...hat.com>
> ---
...
>  arch/s390/mm/fault.c          | 12 ++++++++++++
> diff --git a/arch/s390/mm/fault.c b/arch/s390/mm/fault.c
> index e173b6187ad5..973dcd05c293 100644
> --- a/arch/s390/mm/fault.c
> +++ b/arch/s390/mm/fault.c
> @@ -433,6 +433,17 @@ static inline vm_fault_t do_exception(struct pt_regs *regs, int access)
>  			goto out_up;
>  		goto out;
>  	}
> +
> +	/* The fault is fully completed (including releasing mmap lock) */
> +	if (fault & VM_FAULT_COMPLETED) {
> +		if (gmap) {
> +			mmap_read_lock(mm);
> +			goto out_gmap;
> +		}
> +		fault = 0;
> +		goto out;
> +	}
> +
>  	if (unlikely(fault & VM_FAULT_ERROR))
>  		goto out_up;
>  
> @@ -452,6 +463,7 @@ static inline vm_fault_t do_exception(struct pt_regs *regs, int access)
>  		mmap_read_lock(mm);
>  		goto retry;
>  	}
> +out_gmap:
>  	if (IS_ENABLED(CONFIG_PGSTE) && gmap) {
>  		address =  __gmap_link(gmap, current->thread.gmap_addr,
>  				       address);

FWIW:
Acked-by: Heiko Carstens <hca@...ux.ibm.com>

Powered by blists - more mailing lists