From: Matthew Wilcox <willy@infradead.org>
To: Hugh Dickins <hughd@google.com>
Cc: Andrew Morton <akpm@linux-foundation.org>,
Jan Kara <jack@suse.cz>,
William Kucharski <william.kucharski@oracle.com>,
Linux-FSDevel <linux-fsdevel@vger.kernel.org>,
linux-mm <linux-mm@kvack.org>, Christoph Hellwig <hch@lst.de>,
Johannes Weiner <hannes@cmpxchg.org>,
Yang Shi <yang.shi@linux.alibaba.com>,
dchinner@redhat.com, linux-kernel <linux-kernel@vger.kernel.org>
Subject: Re: [PATCH v4 00/16] Overhaul multi-page lookups for THP
Date: Thu, 26 Nov 2020 20:07:03 +0000 [thread overview]
Message-ID: <20201126200703.GW4327@casper.infradead.org> (raw)
In-Reply-To: <alpine.LSU.2.11.2011261101230.2851@eggly.anvils>
On Thu, Nov 26, 2020 at 11:24:59AM -0800, Hugh Dickins wrote:
> On Thu, 26 Nov 2020, Matthew Wilcox wrote:
> > On Wed, Nov 25, 2020 at 04:11:57PM -0800, Hugh Dickins wrote:
> > > > + index = truncate_inode_partial_page(mapping,
> > > > + page, lstart, lend);
> > > > + if (index > end)
> > > > + end = indices[i] - 1;
> > > > }
> > > > - unlock_page(page);
> > >
> > > The fix needed is here: instead of deleting that unlock_page(page)
> > > line, it needs to be } else { unlock_page(page); }
> >
> > It also needs a put_page(page);
>
> Oh yes indeed, sorry for getting that wrong. I'd misread the
> pagevec_reinit() at the end as the old pagevec_release(). Do you
> really need to do pagevec_remove_exceptionals() there if you're not
> using pagevec_release()?
Oh, good point!
> > That's now taken care of by truncate_inode_partial_page(), so if we're
> > not calling that, we need to put the page as well. ie this:
>
> Right, but I do find it confusing that truncate_inode_partial_page()
> does the unlock_page(),put_page() whereas truncate_inode_page() does
> not: I think you would do better to leave them out of _partial_page(),
> even if including them there saves a couple of lines somewhere else.
I agree; I don't love it. The only reason I moved them in there was
because after calling split_huge_page(), we have to call unlock_page();
put_page(); on the former-tail-page-that-is-now-partial, not on the
head page.
Hm, what I could do is return the struct page which is now the partial
page. Why do I never think of these things before posting? I'll see
how that works out.
> But right now it's the right fix that's important: ack to yours below.
>
> I've not yet worked out the other issues I saw: will report when I have.
> Rebooted this laptop, pretty sure it missed freeing a shmem swap entry,
> not yet reproduced, will study the change there later, but the non-swap
> hang in generic/476 (later seen also in generic/112) more important.
The good news is that I've sorted out the SCRATCH_DEV issue with
running xfstests. The bad news is that (even on an unmodified kernel),
generic/027 takes 19 hours to run. On xfs, it's 4 minutes. Any idea
why it's so slow on tmpfs?
next prev parent reply other threads:[~2020-11-26 20:07 UTC|newest]
Thread overview: 58+ messages / expand[flat|nested] mbox.gz Atom feed top
2020-11-12 21:26 [PATCH v4 00/16] Overhaul multi-page lookups for THP Matthew Wilcox (Oracle)
2020-11-12 21:26 ` [PATCH v4 01/16] mm: Make pagecache tagged lookups return only head pages Matthew Wilcox (Oracle)
2020-11-14 9:53 ` Christoph Hellwig
2020-11-12 21:26 ` [PATCH v4 02/16] mm/shmem: Use pagevec_lookup in shmem_unlock_mapping Matthew Wilcox (Oracle)
2020-11-14 9:53 ` Christoph Hellwig
2020-11-12 21:26 ` [PATCH v4 03/16] mm/swap: Optimise get_shadow_from_swap_cache Matthew Wilcox (Oracle)
2020-11-14 9:53 ` Christoph Hellwig
2020-11-12 21:26 ` [PATCH v4 04/16] mm: Add FGP_ENTRY Matthew Wilcox (Oracle)
2020-11-14 10:00 ` Christoph Hellwig
2020-11-12 21:26 ` [PATCH v4 05/16] mm/filemap: Rename find_get_entry to mapping_get_entry Matthew Wilcox (Oracle)
2020-11-14 10:01 ` Christoph Hellwig
2020-11-12 21:26 ` [PATCH v4 06/16] mm/filemap: Add helper for finding pages Matthew Wilcox (Oracle)
2020-11-14 10:03 ` Christoph Hellwig
2020-11-14 15:15 ` Matthew Wilcox
2020-11-12 21:26 ` [PATCH v4 07/16] mm/filemap: Add mapping_seek_hole_data Matthew Wilcox (Oracle)
2020-11-14 10:04 ` Christoph Hellwig
2020-11-12 21:26 ` [PATCH v4 08/16] iomap: Use mapping_seek_hole_data Matthew Wilcox (Oracle)
2020-11-14 10:06 ` Christoph Hellwig
2020-11-12 21:26 ` [PATCH v4 09/16] mm: Add and use find_lock_entries Matthew Wilcox (Oracle)
2020-11-14 10:07 ` Christoph Hellwig
2020-11-12 21:26 ` [PATCH v4 10/16] mm: Add an 'end' parameter to find_get_entries Matthew Wilcox (Oracle)
2020-11-14 10:08 ` Christoph Hellwig
2020-11-12 21:26 ` [PATCH v4 11/16] mm: Add an 'end' parameter to pagevec_lookup_entries Matthew Wilcox (Oracle)
2020-11-14 10:19 ` Christoph Hellwig
2020-11-12 21:26 ` [PATCH v4 12/16] mm: Remove nr_entries parameter from pagevec_lookup_entries Matthew Wilcox (Oracle)
2020-11-14 10:19 ` Christoph Hellwig
2020-11-12 21:26 ` [PATCH v4 13/16] mm: Pass pvec directly to find_get_entries Matthew Wilcox (Oracle)
2020-11-14 10:21 ` Christoph Hellwig
2020-11-14 15:22 ` Matthew Wilcox
2020-11-12 21:26 ` [PATCH v4 14/16] mm: Remove pagevec_lookup_entries Matthew Wilcox (Oracle)
2020-11-14 10:22 ` Christoph Hellwig
2020-11-12 21:26 ` [PATCH v4 15/16] mm/truncate,shmem: Handle truncates that split THPs Matthew Wilcox (Oracle)
2020-11-12 21:26 ` [PATCH v4 16/16] mm/filemap: Return only head pages from find_get_entries Matthew Wilcox (Oracle)
2020-11-14 10:23 ` Christoph Hellwig
2020-11-16 10:34 ` [PATCH v4 00/16] Overhaul multi-page lookups for THP Hugh Dickins
2020-11-16 15:14 ` Matthew Wilcox
2020-11-16 21:27 ` Hugh Dickins
2020-11-17 15:39 ` Matthew Wilcox
2020-11-17 16:26 ` Hugh Dickins
2020-11-17 19:15 ` Matthew Wilcox
2020-11-17 23:43 ` Matthew Wilcox
2020-11-25 2:32 ` Matthew Wilcox
2020-11-25 2:50 ` Hugh Dickins
2020-11-25 2:56 ` Hugh Dickins
2020-11-25 23:08 ` Andrew Morton
2020-11-26 0:11 ` Hugh Dickins
2020-11-26 12:15 ` Matthew Wilcox
2020-11-26 19:24 ` Hugh Dickins
2020-11-26 20:07 ` Matthew Wilcox [this message]
2020-11-30 19:45 ` Hugh Dickins
2020-12-01 4:52 ` Hugh Dickins
[not found] ` <CGME20201203154604eucas1p200d001d25dd344a1dd1c7da34f35aad0@eucas1p2.samsung.com>
2020-12-03 15:46 ` Marek Szyprowski
[not found] ` <CGME20201203172725eucas1p2fddec1d269c55095859d490942b78b93@eucas1p2.samsung.com>
2020-12-03 17:27 ` Marek Szyprowski
2020-12-03 21:27 ` Qian Cai
2020-12-03 22:19 ` Hugh Dickins
2020-12-03 21:45 ` Hugh Dickins
2021-02-23 22:58 ` Andrew Morton
2021-02-23 23:27 ` Matthew Wilcox
Reply instructions:
You may reply publicly to this message via plain-text email
using any one of the following methods:
* Save the following mbox file, import it into your mail client,
and reply-to-all from there: mbox
Avoid top-posting and favor interleaved quoting:
https://en.wikipedia.org/wiki/Posting_style#Interleaved_style
* Reply using the --to, --cc, and --in-reply-to
switches of git-send-email(1):
git send-email \
--in-reply-to=20201126200703.GW4327@casper.infradead.org \
--to=willy@infradead.org \
--cc=akpm@linux-foundation.org \
--cc=dchinner@redhat.com \
--cc=hannes@cmpxchg.org \
--cc=hch@lst.de \
--cc=hughd@google.com \
--cc=jack@suse.cz \
--cc=linux-fsdevel@vger.kernel.org \
--cc=linux-kernel@vger.kernel.org \
--cc=linux-mm@kvack.org \
--cc=william.kucharski@oracle.com \
--cc=yang.shi@linux.alibaba.com \
/path/to/YOUR_REPLY
https://kernel.org/pub/software/scm/git/docs/git-send-email.html
* If your mail client supports setting the In-Reply-To header
via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line
before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).