Re: [patch 02/14] tmpfs: fix regressions from wider use of ZERO_PAGE

From: Matthew Wilcox <willy@infradead.org>
To: Linus Torvalds <torvalds@linux-foundation.org>
Cc: Andrew Morton <akpm@linux-foundation.org>,
	the arch/x86 maintainers <x86@kernel.org>,
	Peter Zijlstra <peterz@infradead.org>,
	Borislav Petkov <bp@suse.de>,
	patrice.chotard@foss.st.com,
	Mikulas Patocka <mpatocka@redhat.com>,
	markhemm@googlemail.com, Lukas Czerner <lczerner@redhat.com>,
	Christoph Hellwig <hch@lst.de>,
	"Darrick J. Wong" <djwong@kernel.org>,
	Chuck Lever <chuck.lever@oracle.com>,
	Hugh Dickins <hughd@google.com>,
	patches@lists.linux.dev, Linux-MM <linux-mm@kvack.org>,
	mm-commits@vger.kernel.org
Subject: Re: [patch 02/14] tmpfs: fix regressions from wider use of ZERO_PAGE
Date: Fri, 15 Apr 2022 23:21:51 +0100	[thread overview]
Message-ID: <Ylnv/9hiO+g4X9i5@casper.infradead.org> (raw)
In-Reply-To: <CAHk-=wh=Mu_EYhtOmPn6AxoQZyEh-4fo2Zx3G7rBv1g7vwoKiw@mail.gmail.com>

On Fri, Apr 15, 2022 at 03:10:51PM -0700, Linus Torvalds wrote:
> On Thu, Apr 14, 2022 at 7:13 PM Andrew Morton <akpm@linux-foundation.org> wrote:
> >
> > Revert shmem_file_read_iter() to using ZERO_PAGE for holes only when
> > iter_is_iovec(); in other cases, use the more natural iov_iter_zero()
> > instead of copy_page_to_iter().  We would use iov_iter_zero() throughout,
> > but the x86 clear_user() is not nearly so well optimized as copy to user
> > (dd of 1T sparse tmpfs file takes 57 seconds rather than 44 seconds).
> 
> Ugh.
> 
> I've applied this patch, but honestly, the proper course of action
> should just be to improve on clear_user().
> 
> If it really is important enough that we should care about that
> performance, then we just should fix clear_user().
> 
> It's a very odd special thing right now (at least on x86-64) using
> some strange handcrafted inline asm code.
> 
> I assume that 'rep stosb' is the fastest way to clear things on modern
> CPU's that have FSRM, and then we have the usual fallbacks (ie ERMS ->
> "rep stos" except for small areas, and probably that "store zeros by
> hand" for older CPUs).
> 
> Adding PeterZ and Borislav (who seem to be the last ones to have
> worked on the copy and clear_page stuff respectively) and the x86
> maintainers in case somebody gets the urge to just fix this.

Perhaps the x86 maintainers would like to start from
https://lore.kernel.org/lkml/20210523180423.108087-1-sneves@dei.uc.pt/
instead of pushing that work off on the submitter.