From: Matthew Wilcox <willy@infradead.org>
To: Hugh Dickins <hughd@google.com>
Cc: Zi Yan <ziy@nvidia.com>, Vlastimil Babka <vbabka@suse.cz>,
David Hildenbrand <david@redhat.com>,
linux-mm@kvack.org,
"Kirill A . Shutemov" <kirill.shutemov@linux.intel.com>,
Mike Kravetz <mike.kravetz@oracle.com>,
Michal Hocko <mhocko@kernel.org>,
John Hubbard <jhubbard@nvidia.com>,
linux-kernel@vger.kernel.org
Subject: Re: [RFC PATCH 00/15] Make MAX_ORDER adjustable as a kernel boot time parameter.
Date: Mon, 9 Aug 2021 12:22:09 +0100 [thread overview]
Message-ID: <YREP4ZzFMfkC6N/L@casper.infradead.org> (raw)
In-Reply-To: <c2a4dd1d-c1ac-876b-4a3-c5ea7555498@google.com>
On Sun, Aug 08, 2021 at 09:29:29PM -0700, Hugh Dickins wrote:
> On Sat, 7 Aug 2021, Matthew Wilcox wrote:
> >
> > I am, however, of the opinion that 2MB pages give us so much trouble
> > because they're so very special. Few people exercise those code paths and
> > it's easy to break them without noticing. This is partly why I want to
> > do arbitrary-order pages. If everybody is running with compound pages
> > all the time, we'll see the corner cases often, and people other than
> > Hugh, Kirill and Mike will be able to work on them.
>
> I don't entirely agree. I'm all for your use of compound pages in page
> cache, but don't think its problems are representative of the problems
> in aiming for a PMD (or PUD) bar, with the weird page table transitions
> we expect of "THP" there.
>
> I haven't checked: is your use of compound pages in page cache still
> limited to the HAVE_ARCH_TRANSPARENT_HUGEPAGE architectures? When
> the others could just as well use compound pages in page cache too.
It is no longer gated by whether TRANSPARENT_HUGEPAGE is enabled or not.
That's a relatively recent change, and I can't say that I've tested it.
I'll give it a try today on a PA-RISC system I've booted recently.
One of the followup pieces of work that I hope somebody other than
myself will undertake is using 64KB PTEs on a 4KB PAGE_SIZE ARM/POWER
machine if the stars align.
next prev parent reply other threads:[~2021-08-09 11:23 UTC|newest]
Thread overview: 50+ messages / expand[flat|nested] mbox.gz Atom feed top
2021-08-05 19:02 [RFC PATCH 00/15] Make MAX_ORDER adjustable as a kernel boot time parameter Zi Yan
2021-08-05 19:02 ` [RFC PATCH 01/15] arch: x86: remove MAX_ORDER exceeding SECTION_SIZE check for 32bit vdso Zi Yan
2021-08-05 19:02 ` [RFC PATCH 02/15] arch: mm: rename FORCE_MAX_ZONEORDER to ARCH_FORCE_MAX_ORDER Zi Yan
2021-08-05 19:02 ` [RFC PATCH 03/15] mm: check pfn validity when buddy allocator can merge pages across mem sections Zi Yan
2021-08-05 19:02 ` [RFC PATCH 04/15] mm: prevent pageblock size being larger than section size Zi Yan
2021-08-05 19:02 ` [RFC PATCH 05/15] mm/memory_hotplug: online pages at " Zi Yan
2021-08-05 19:02 ` [RFC PATCH 06/15] mm: use PAGES_PER_SECTION instead for mem_map_offset/next() Zi Yan
2021-08-05 19:02 ` [RFC PATCH 07/15] mm: hugetlb: use PAGES_PER_SECTION to check mem_map discontiguity Zi Yan
2021-08-05 19:02 ` [RFC PATCH 08/15] fs: proc: use PAGES_PER_SECTION for page offline checking period Zi Yan
2021-08-07 10:32 ` Mike Rapoport
2021-08-09 15:45 ` [RFC PATCH 08/15] " Zi Yan
2021-08-05 19:02 ` [RFC PATCH 09/15] virtio: virtio_mem: use PAGES_PER_SECTION instead of MAX_ORDER_NR_PAGES Zi Yan
2021-08-09 7:35 ` David Hildenbrand
2021-08-05 19:02 ` [RFC PATCH 10/15] virtio: virtio_balloon: " Zi Yan
2021-08-09 7:42 ` David Hildenbrand
2021-08-05 19:02 ` [RFC PATCH 11/15] mm/page_reporting: report pages at section size instead of MAX_ORDER Zi Yan
2021-08-09 7:25 ` David Hildenbrand
2021-08-09 14:12 ` Alexander Duyck
2021-08-09 15:08 ` Zi Yan
2021-08-09 16:51 ` Alexander Duyck
2021-08-09 14:08 ` Alexander Duyck
2021-08-05 19:02 ` [RFC PATCH 12/15] mm: Make MAX_ORDER of buddy allocator configurable via Kconfig SET_MAX_ORDER Zi Yan
2021-08-06 15:16 ` Vlastimil Babka
2021-08-06 15:23 ` Zi Yan
2021-08-05 19:02 ` [RFC PATCH 13/15] mm: convert MAX_ORDER sized static arrays to dynamic ones Zi Yan
2021-08-05 19:16 ` Christian König
2021-08-05 19:58 ` Zi Yan
2021-08-06 9:37 ` Christian König
2021-08-06 14:00 ` Zi Yan
2021-08-05 19:02 ` [RFC PATCH 14/15] mm: introduce MIN_MAX_ORDER to replace MAX_ORDER as compile time constant Zi Yan
2021-08-08 8:23 ` Mike Rapoport
2021-08-09 15:35 ` Zi Yan
2021-08-05 19:02 ` [RFC PATCH 15/15] mm: make MAX_ORDER a kernel boot time parameter Zi Yan
2021-08-06 15:36 ` [RFC PATCH 00/15] Make MAX_ORDER adjustable as " Vlastimil Babka
2021-08-06 16:16 ` David Hildenbrand
2021-08-06 16:54 ` Vlastimil Babka
2021-08-06 17:08 ` David Hildenbrand
2021-08-06 18:24 ` Zi Yan
2021-08-09 7:20 ` David Hildenbrand
2021-08-08 7:41 ` Mike Rapoport
2021-08-06 16:32 ` Vlastimil Babka
2021-08-06 17:19 ` Zi Yan
2021-08-06 20:27 ` Hugh Dickins
2021-08-06 21:26 ` Zi Yan
2021-08-09 4:04 ` Hugh Dickins
2021-08-07 1:10 ` Matthew Wilcox
2021-08-07 21:23 ` Matthew Wilcox
2021-08-09 4:29 ` Hugh Dickins
2021-08-09 11:22 ` Matthew Wilcox [this message]
2021-08-09 7:41 ` David Hildenbrand
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=YREP4ZzFMfkC6N/L@casper.infradead.org \
--to=willy@infradead.org \
--cc=david@redhat.com \
--cc=hughd@google.com \
--cc=jhubbard@nvidia.com \
--cc=kirill.shutemov@linux.intel.com \
--cc=linux-kernel@vger.kernel.org \
--cc=linux-mm@kvack.org \
--cc=mhocko@kernel.org \
--cc=mike.kravetz@oracle.com \
--cc=vbabka@suse.cz \
--cc=ziy@nvidia.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).