From: Pavel Tatashin <pasha.tatashin@soleen.com>
To: Michal Hocko <mhocko@kernel.org>
Cc: Vincent Whitchurch <vincent.whitchurch@axis.com>,
"akpm@linux-foundation.org" <akpm@linux-foundation.org>,
"osalvador@suse.de" <osalvador@suse.de>,
"linux-mm@kvack.org" <linux-mm@kvack.org>,
"linux-kernel@vger.kernel.org" <linux-kernel@vger.kernel.org>
Subject: Re: [PATCH] mm/sparse: Consistently do not zero memmap
Date: Wed, 30 Oct 2019 12:53:41 -0400 [thread overview]
Message-ID: <CA+CK2bA3gM4pMSj-wDWgAPNoPtcjwd59_6VivKA2Uf2GriASsw@mail.gmail.com> (raw)
In-Reply-To: <20191030153150.GI31513@dhcp22.suse.cz>
On Wed, Oct 30, 2019 at 11:31 AM Michal Hocko <mhocko@kernel.org> wrote:
>
> On Wed 30-10-19 11:20:44, Pavel Tatashin wrote:
> > On Wed, Oct 30, 2019 at 10:13 AM Michal Hocko <mhocko@kernel.org> wrote:
> > >
> > > [Add Pavel - the email thread starts http://lkml.kernel.org/r/20191030131122.8256-1-vincent.whitchurch@axis.com
> > > but it used your old email address]
> > >
> > > On Wed 30-10-19 15:02:16, Vincent Whitchurch wrote:
> > > > On Wed, Oct 30, 2019 at 02:29:58PM +0100, Michal Hocko wrote:
> > > > > On Wed 30-10-19 14:11:22, Vincent Whitchurch wrote:
> > > > > > (I noticed this because on my ARM64 platform, with 1 GiB of memory the
> > > > > > first [and only] section is allocated from the zeroing path while with
> > > > > > 2 GiB of memory the first 1 GiB section is allocated from the
> > > > > > non-zeroing path.)
> > > > >
> > > > > Do I get it right that sparse_buffer_init couldn't allocate memmap for
> > > > > the full node for some reason and so sparse_init_nid would have to
> > > > > allocate one for each memory section?
> > > >
> > > > Not quite. The sparsemap_buf is successfully allocated with the correct
> > > > size in sparse_buffer_init(), but sparse_buffer_alloc() fails to
> > > > allocate the same size from it.
> > > >
> > > > The reason it fails is that sparse_buffer_alloc() for some reason wants
> > > > to return a pointer which is aligned to the allocation size. But the
> > > > sparsemap_buf was only allocated with PAGE_SIZE alignment so there's not
> > > > enough space to align it.
> > > >
> > > > I don't understand the reason for this alignment requirement since the
> > > > fallback path also allocates with PAGE_SIZE alignment. I'm guessing the
> > > > alignment is for the VMEMAP code which also uses sparse_buffer_alloc()?
> > >
> > > I am not 100% sure TBH. Aligning makes some sense when mapping the
> > > memmaps to page tables but that would suggest that sparse_buffer_init
> > > is using a wrong alignment then. It is quite wasteful to allocate
> > > alarge misaligned block like that.
> > >
> > > Your patch still makes sense but this is something to look into.
> > >
> > > Pavel?
> >
> > I remember thinking about this large alignment, as it looked out of
> > place to me also.
> > It was there to keep memmap in single chunks on larger x86 machines.
> > Perhaps it can be revisited now.
>
> Don't we need 2MB aligned memmaps for their PMD mappings?
Yes, PMD_SIZE should be the alignment here. It just does not make
sense to align to size.
Pasha
next prev parent reply other threads:[~2019-10-30 16:53 UTC|newest]
Thread overview: 19+ messages / expand[flat|nested] mbox.gz Atom feed top
2019-10-30 13:11 [PATCH] mm/sparse: Consistently do not zero memmap Vincent Whitchurch
2019-10-30 13:29 ` Michal Hocko
2019-10-30 14:02 ` Vincent Whitchurch
2019-10-30 14:12 ` Michal Hocko
2019-10-30 15:20 ` Pavel Tatashin
2019-10-30 15:31 ` Michal Hocko
2019-10-30 16:53 ` Pavel Tatashin [this message]
2019-10-30 17:31 ` Michal Hocko
2019-10-30 17:53 ` Pavel Tatashin
2019-10-30 18:01 ` Michal Hocko
2019-10-30 18:23 ` Pavel Tatashin
2019-10-31 7:25 ` Michal Hocko
2019-11-04 15:51 ` Vincent Whitchurch
2019-11-05 8:43 ` Michal Hocko
2019-11-15 23:55 ` Andrew Morton
2019-11-18 11:28 ` Michal Hocko
2019-10-30 13:31 ` David Hildenbrand
2019-10-30 13:38 ` Oscar Salvador
2019-10-30 15:21 ` Pavel Tatashin
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=CA+CK2bA3gM4pMSj-wDWgAPNoPtcjwd59_6VivKA2Uf2GriASsw@mail.gmail.com \
--to=pasha.tatashin@soleen.com \
--cc=akpm@linux-foundation.org \
--cc=linux-kernel@vger.kernel.org \
--cc=linux-mm@kvack.org \
--cc=mhocko@kernel.org \
--cc=osalvador@suse.de \
--cc=vincent.whitchurch@axis.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).