Linux-Fsdevel Archive on lore.kernel.org
help / color / mirror / Atom feed
From: Matthew Wilcox <willy@infradead.org>
To: linux-fsdevel@vger.kernel.org
Cc: linux-mm@kvack.org, linux-kernel@vger.kernel.org,
	Hugh Dickins <hughd@google.com>
Subject: Re: [RFC v6 00/51] Large pages in the page cache
Date: Sun, 14 Jun 2020 09:26:50 -0700	[thread overview]
Message-ID: <20200614162650.GP8681@bombadil.infradead.org> (raw)
In-Reply-To: <20200610201345.13273-1-willy@infradead.org>

On Wed, Jun 10, 2020 at 01:12:54PM -0700, Matthew Wilcox wrote:
> Another fortnight, another dump of my current large pages work.

The generic/127 test has pointed out to me that range writeback is
broken by this patchset.  Here's how (may not be exactly what's going on,
but it's close):

page cache allocates an order-2 page covering indices 40-43.
bytes are written, page is dirtied
test then calls fallocate(FALLOC_FL_COLLAPSE_RANGE) for a range which
starts in page 41.
XFS calls filemap_write_and_wait_range() which calls
__filemap_fdatawrite_range() which calls
do_writepages() which calls
iomap_writepages() which calls
write_cache_pages() which calls
tag_pages_for_writeback() which calls
xas_for_each_marked() starting at page 41.  Which doesn't find page
  41 because when we dirtied pages 40-43, we only marked index 40 as
  being dirty.

Annoyingly, the XArray actually handles this just fine ... if we were
using multi-order entries, we'd find it.  But we're still storing 2^N
entries for an order N page.

I can see two ways to fix this.  One is to bite the bullet and do the
conversion of the page cache to use multi-order entries.  The second
is to set and clear the marks on all entries.  I'm concerned about the
performance of the latter solution.  Not so bad for order-2 pages, but for
an order-9 page we have 520 bits to set, spread over 9 non-consecutive
cachelines.  Also, I'm unenthusiastic about writing code that I want to
throw away as quickly as possible.

So unless somebody has a really good alternative idea, I'm going to
convert the page cache over to multi-order entries.  This will have
several positive effects:

 - Get DAX and regular page cache using the xarray in a more similar way
 - Saves about 4.5kB of memory for every 2MB page in tmpfs/shmem
 - Prep work for converting hugetlbfs to use the page cache the same
   way as tmpfs

      parent reply	other threads:[~2020-06-14 16:26 UTC|newest]

Thread overview: 60+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2020-06-10 20:12 Matthew Wilcox
2020-06-10 20:12 ` [PATCH v6 01/51] mm: Print head flags in dump_page Matthew Wilcox
2020-06-10 20:12 ` [PATCH v6 02/51] mm: Print the inode number " Matthew Wilcox
2020-06-10 20:12 ` [PATCH v6 03/51] mm: Print hashed address of struct page Matthew Wilcox
2020-06-10 20:12 ` [PATCH v6 04/51] mm: Move PageDoubleMap bit Matthew Wilcox
2020-06-11 15:03   ` Zi Yan
2020-06-10 20:12 ` [PATCH v6 05/51] mm: Simplify PageDoubleMap with PF_SECOND policy Matthew Wilcox
2020-06-11 15:14   ` Zi Yan
2020-06-10 20:13 ` [PATCH v6 06/51] mm: Store compound_nr as well as compound_order Matthew Wilcox
2020-06-10 20:13 ` [PATCH v6 07/51] mm: Move page-flags include to top of file Matthew Wilcox
2020-06-10 20:13 ` [PATCH v6 08/51] mm: Add thp_order Matthew Wilcox
2020-06-10 20:13 ` [PATCH v6 09/51] mm: Add thp_size Matthew Wilcox
2020-06-10 20:13 ` [PATCH v6 10/51] mm: Replace hpage_nr_pages with thp_nr_pages Matthew Wilcox
2020-06-10 20:13 ` [PATCH v6 11/51] mm: Add thp_head Matthew Wilcox
2020-06-10 20:13 ` [PATCH v6 12/51] mm: Introduce offset_in_thp Matthew Wilcox
2020-06-10 20:13 ` [PATCH v6 13/51] mm: Support arbitrary THP sizes Matthew Wilcox
2020-06-10 20:13 ` [PATCH v6 14/51] fs: Add a filesystem flag for THPs Matthew Wilcox
2020-06-10 20:13 ` [PATCH v6 15/51] fs: Do not update nr_thps for mappings which support THPs Matthew Wilcox
2020-06-10 20:13 ` [PATCH v6 16/51] fs: Introduce i_blocks_per_page Matthew Wilcox
2020-06-10 20:13 ` [PATCH v6 17/51] fs: Make page_mkwrite_check_truncate thp-aware Matthew Wilcox
2020-06-10 20:13 ` [PATCH v6 18/51] mm: Support THPs in zero_user_segments Matthew Wilcox
2020-06-10 20:13 ` [PATCH v6 19/51] mm: Zero the head page, not the tail page Matthew Wilcox
2020-06-10 20:13 ` [PATCH v6 20/51] block: Add bio_for_each_thp_segment_all Matthew Wilcox
2020-06-11 18:20   ` Matthew Wilcox
2020-06-10 20:13 ` [PATCH v6 21/51] block: Support THPs in page_is_mergeable Matthew Wilcox
2020-06-12 16:17   ` Matthew Wilcox
2020-06-10 20:13 ` [PATCH v6 22/51] iomap: Support arbitrarily many blocks per page Matthew Wilcox
2020-06-10 20:13 ` [PATCH v6 23/51] iomap: Support THPs in iomap_adjust_read_range Matthew Wilcox
2020-06-10 20:13 ` [PATCH v6 24/51] iomap: Support THPs in invalidatepage Matthew Wilcox
2020-06-10 20:13 ` [PATCH v6 25/51] iomap: Support THPs in read paths Matthew Wilcox
2020-06-10 20:13 ` [PATCH v6 26/51] iomap: Convert iomap_write_end types Matthew Wilcox
2020-06-10 20:13 ` [PATCH v6 27/51] iomap: Change calling convention for zeroing Matthew Wilcox
2020-06-10 20:13 ` [PATCH v6 28/51] iomap: Change iomap_write_begin calling convention Matthew Wilcox
2020-06-10 20:13 ` [PATCH v6 29/51] iomap: Support THPs in write paths Matthew Wilcox
2020-06-10 20:13 ` [PATCH v6 30/51] iomap: Inline data shouldn't see THPs Matthew Wilcox
2020-06-10 20:13 ` [PATCH v6 31/51] iomap: Handle tail pages in iomap_page_mkwrite Matthew Wilcox
2020-06-10 20:13 ` [PATCH v6 32/51] xfs: Support THPs Matthew Wilcox
2020-06-10 20:13 ` [PATCH v6 33/51] mm: Make prep_transhuge_page return its argument Matthew Wilcox
2020-06-10 20:13 ` [PATCH v6 34/51] mm: Add __page_cache_alloc_order Matthew Wilcox
2020-06-10 20:13 ` [PATCH v6 35/51] mm: Allow THPs to be added to the page cache Matthew Wilcox
2020-06-10 20:13 ` [PATCH v6 36/51] mm: Allow THPs to be removed from " Matthew Wilcox
2020-06-10 20:13 ` [PATCH v6 37/51] mm: Remove page fault assumption of compound page size Matthew Wilcox
2020-06-10 20:13 ` [PATCH v6 38/51] mm: Fix total_mapcount assumption of " Matthew Wilcox
2020-06-10 20:13 ` [PATCH v6 39/51] mm: Remove assumptions of THP size Matthew Wilcox
2020-06-10 20:13 ` [PATCH v6 40/51] mm: Avoid splitting THPs Matthew Wilcox
2020-06-10 20:13 ` [PATCH v6 41/51] mm: Fix truncation for pages of arbitrary size Matthew Wilcox
2020-06-10 20:13 ` [PATCH v6 42/51] mm: Handle truncates that split THPs Matthew Wilcox
2020-06-10 20:13 ` [PATCH v6 43/51] mm: Support storing shadow entries for THPs Matthew Wilcox
2020-06-10 20:13 ` [PATCH v6 44/51] mm: Support retrieving tail pages from the page cache Matthew Wilcox
2020-06-10 20:13 ` [PATCH v6 45/51] mm: Support tail pages in wait_for_stable_page Matthew Wilcox
2020-06-10 20:13 ` [PATCH v6 46/51] mm: Add DEFINE_READAHEAD Matthew Wilcox
2020-06-10 20:13 ` [PATCH v6 47/51] mm: Make page_cache_readahead_unbounded take a readahead_control Matthew Wilcox
2020-06-10 20:13 ` [PATCH v6 48/51] mm: Make __do_page_cache_readahead " Matthew Wilcox
2020-06-10 20:13 ` [PATCH v6 49/51] mm: Allow PageReadahead to be set on head pages Matthew Wilcox
2020-06-10 20:13 ` [PATCH v6 50/51] mm: Add THP readahead Matthew Wilcox
2020-06-10 20:13 ` [PATCH v6 51/51] mm: Align THP mappings for non-DAX Matthew Wilcox
2020-06-11  6:59 ` [RFC v6 00/51] Large pages in the page cache Christoph Hellwig
2020-06-11 11:24   ` Matthew Wilcox
2020-06-15 13:32     ` Christoph Hellwig
2020-06-14 16:26 ` Matthew Wilcox [this message]

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=20200614162650.GP8681@bombadil.infradead.org \
    --to=willy@infradead.org \
    --cc=hughd@google.com \
    --cc=linux-fsdevel@vger.kernel.org \
    --cc=linux-kernel@vger.kernel.org \
    --cc=linux-mm@kvack.org \
    --subject='Re: [RFC v6 00/51] Large pages in the page cache' \
    /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

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).