From: Nick Piggin <npiggin@suse.de>
To: Nishanth Aravamudan <nacc@us.ibm.com>
Cc: akpm@linux-foundation.org, linux-mm@kvack.org,
andi@firstfloor.org, kniht@linux.vnet.ibm.com, abh@cray.com,
wli@holomorphy.com
Subject: Re: [patch 06/18] hugetlb: multi hstate proc files
Date: Fri, 23 May 2008 07:22:15 +0200 [thread overview]
Message-ID: <20080523052215.GF13071@wotan.suse.de> (raw)
In-Reply-To: <20080502195311.GD26273@us.ibm.com>
On Fri, May 02, 2008 at 12:53:11PM -0700, Nishanth Aravamudan wrote:
> On 23.04.2008 [11:53:08 +1000], npiggin@suse.de wrote:
> > Convert /proc output code over to report multiple hstates
> >
> > I chose to just report the numbers in a row, in the hope
> > to minimze breakage of existing software. The "compat" page size
> > is always the first number.
>
> Only if add_huge_hstate() is called first for the compat page size,
> right? That seems bad if we depend on an ordering.
>
> For instance, for power, I think Jon is calling huge_add_hstate() from
> the arch/powerpc/mm/hugetlbpage.c init routine. Which runs before
> hugetlb_init, which means that if he adds hugepages like
>
> huge_add_hstate(64k-order);
> huge_add_hstate(16m-order);
> huge_add_hstate(16g-order);
>
> We'll get 64k as the first field in meminfo.
>
> So perhaps what we should do is:
>
> 1) architectures define HPAGE_* as the default (compat) hugepage values
> 2) architectures have a call into generic code at their init time to
> specify what sizes they support
> 3) the core is the only place that actually does huge_add_hstate() and
> it always does it first for the compat order?
>
> I wonder if this might lead to issues in timing between processing
> hugepagesz= (in arch code) and hugepages= (in generic code). Not sure. I
> guess if we always add all hugepage sizes, we should have all the
> hstates we know about ready to configure and as long as hugetlb_init
> runs before hugepages= processing, we should be fine? Dunno.
You're right I think. The other thing is that we could just have
a small map from the hstate array to reporting order for sysctls.
We could report them in the order specified on the cmdline, with
the default size first if it was not specified on the cmdline.
Hmm, I'll see how that looks.
--
To unsubscribe, send a message with 'unsubscribe linux-mm' in
the body to majordomo@kvack.org. For more info on Linux MM,
see: http://www.linux-mm.org/ .
Don't email: <a href=mailto:"dont@kvack.org"> email@kvack.org </a>
next prev parent reply other threads:[~2008-05-23 5:22 UTC|newest]
Thread overview: 123+ messages / expand[flat|nested] mbox.gz Atom feed top
2008-04-23 1:53 [patch 00/18] multi size, and giant hugetlb page support, 1GB hugetlb for x86 npiggin
2008-04-23 1:53 ` [patch 01/18] hugetlb: fix lockdep spew npiggin
2008-04-23 13:06 ` KOSAKI Motohiro
2008-04-23 1:53 ` [patch 02/18] hugetlb: factor out huge_new_page npiggin
2008-04-24 23:49 ` Nishanth Aravamudan
2008-04-24 23:54 ` Nishanth Aravamudan
2008-04-24 23:58 ` Nishanth Aravamudan
2008-04-25 7:10 ` Andi Kleen
2008-04-25 16:54 ` Nishanth Aravamudan
2008-04-25 19:13 ` Christoph Lameter
2008-04-25 19:29 ` Nishanth Aravamudan
2008-04-30 19:16 ` Christoph Lameter
2008-04-30 20:44 ` Nishanth Aravamudan
2008-05-01 19:23 ` Christoph Lameter
2008-05-01 20:25 ` Nishanth Aravamudan
2008-05-01 20:34 ` Christoph Lameter
2008-05-01 21:01 ` Nishanth Aravamudan
2008-05-23 5:03 ` Nick Piggin
2008-04-23 1:53 ` [patch 03/18] mm: offset align in alloc_bootmem npiggin, Yinghai Lu
2008-04-23 1:53 ` [patch 04/18] hugetlb: modular state npiggin
2008-04-23 15:21 ` Jon Tollefson
2008-04-23 15:38 ` Nick Piggin
2008-04-25 17:13 ` Nishanth Aravamudan
2008-05-23 5:02 ` Nick Piggin
2008-05-23 20:48 ` Nishanth Aravamudan
2008-04-23 1:53 ` [patch 05/18] hugetlb: multiple hstates npiggin
2008-04-25 17:38 ` Nishanth Aravamudan
2008-04-25 17:48 ` Nishanth Aravamudan
2008-04-25 17:55 ` Andi Kleen
2008-04-25 17:52 ` Nishanth Aravamudan
2008-04-25 18:10 ` Andi Kleen
2008-04-28 10:13 ` Andy Whitcroft
2008-05-23 5:18 ` Nick Piggin
2008-04-29 17:27 ` Nishanth Aravamudan
2008-05-23 5:19 ` Nick Piggin
2008-04-23 1:53 ` [patch 06/18] hugetlb: multi hstate proc files npiggin
2008-05-02 19:53 ` Nishanth Aravamudan
2008-05-23 5:22 ` Nick Piggin [this message]
2008-05-23 20:30 ` Nishanth Aravamudan
2008-04-23 1:53 ` [patch 07/18] hugetlbfs: per mount hstates npiggin
2008-04-25 18:09 ` Nishanth Aravamudan
2008-04-25 20:36 ` Nishanth Aravamudan
2008-04-25 22:39 ` Nishanth Aravamudan
2008-04-28 18:20 ` Adam Litke
2008-04-28 18:46 ` Nishanth Aravamudan
2008-05-23 5:24 ` Nick Piggin
2008-05-23 20:34 ` Nishanth Aravamudan
2008-05-23 22:49 ` Nick Piggin
2008-05-23 23:24 ` Nishanth Aravamudan
2008-04-23 1:53 ` [patch 08/18] hugetlb: multi hstate sysctls npiggin
2008-04-25 18:14 ` Nishanth Aravamudan
2008-05-23 5:25 ` Nick Piggin
2008-05-23 20:27 ` Nishanth Aravamudan
2008-04-25 23:35 ` Nishanth Aravamudan
2008-05-23 5:28 ` Nick Piggin
2008-05-23 10:40 ` Andi Kleen
2008-04-23 1:53 ` [patch 09/18] hugetlb: abstract numa round robin selection npiggin
2008-04-23 1:53 ` [patch 10/18] mm: introduce non panic alloc_bootmem npiggin
2008-04-23 1:53 ` [patch 11/18] mm: export prep_compound_page to mm npiggin
2008-04-23 16:12 ` Andrew Hastings
2008-05-23 5:29 ` Nick Piggin
2008-04-23 1:53 ` [patch 12/18] hugetlbfs: support larger than MAX_ORDER npiggin
2008-04-23 16:15 ` Andrew Hastings
2008-04-23 16:25 ` Andi Kleen
2008-04-25 18:55 ` Nishanth Aravamudan
2008-05-23 5:29 ` Nick Piggin
2008-04-30 21:01 ` Dave Hansen
2008-05-23 5:30 ` Nick Piggin
2008-04-23 1:53 ` [patch 13/18] hugetlb: support boot allocate different sizes npiggin
2008-04-23 16:15 ` Andrew Hastings
2008-04-25 18:40 ` Nishanth Aravamudan
2008-04-25 18:50 ` Andi Kleen
2008-04-25 20:05 ` Nishanth Aravamudan
2008-05-23 5:36 ` Nick Piggin
2008-05-23 6:04 ` Nick Piggin
2008-05-23 20:32 ` Nishanth Aravamudan
2008-05-23 22:45 ` Nick Piggin
2008-05-23 22:53 ` Nishanth Aravamudan
2008-04-23 1:53 ` [patch 14/18] hugetlb: printk cleanup npiggin
2008-04-27 3:32 ` Nishanth Aravamudan
2008-05-23 5:37 ` Nick Piggin
2008-04-23 1:53 ` [patch 15/18] hugetlb: introduce huge_pud npiggin
2008-04-23 1:53 ` [patch 16/18] x86: support GB hugepages on 64-bit npiggin
2008-04-23 1:53 ` [patch 17/18] x86: add hugepagesz option " npiggin
2008-04-30 19:34 ` Nishanth Aravamudan
2008-04-30 19:52 ` Andi Kleen
2008-04-30 20:02 ` Nishanth Aravamudan
2008-04-30 20:19 ` Andi Kleen
2008-04-30 20:23 ` Nishanth Aravamudan
2008-04-30 20:45 ` Andi Kleen
2008-04-30 20:51 ` Nishanth Aravamudan
2008-04-30 20:40 ` Jon Tollefson
2008-04-30 20:48 ` Nishanth Aravamudan
2008-05-23 5:41 ` Nick Piggin
2008-05-23 10:43 ` Andi Kleen
2008-05-23 12:34 ` Nick Piggin
2008-05-23 14:29 ` Andi Kleen
2008-05-23 20:43 ` Nishanth Aravamudan
2008-05-23 20:39 ` Nishanth Aravamudan
2008-05-23 22:52 ` Nick Piggin
2008-04-23 1:53 ` [patch 18/18] hugetlb: my fixes 2 npiggin
2008-04-23 10:48 ` Andi Kleen
2008-04-23 15:36 ` Nick Piggin
2008-04-23 18:49 ` Nishanth Aravamudan
2008-04-23 19:37 ` Andi Kleen
2008-04-23 21:11 ` Nishanth Aravamudan
2008-04-23 21:38 ` Nishanth Aravamudan
2008-04-23 22:06 ` Dave Hansen
2008-04-23 15:20 ` Jon Tollefson
2008-04-23 15:44 ` Nick Piggin
2008-04-23 8:05 ` [patch 00/18] multi size, and giant hugetlb page support, 1GB hugetlb for x86 Andi Kleen
2008-04-23 15:34 ` Nick Piggin
2008-04-23 15:46 ` Andi Kleen
2008-04-23 15:53 ` Nick Piggin
2008-04-23 16:02 ` Andi Kleen
2008-04-23 16:02 ` Nick Piggin
2008-04-23 18:54 ` Nishanth Aravamudan
2008-04-23 18:52 ` Nishanth Aravamudan
2008-04-24 2:08 ` Nick Piggin
2008-04-24 6:43 ` Nishanth Aravamudan
2008-04-24 7:06 ` Nick Piggin
2008-04-24 17:08 ` Nishanth Aravamudan
2008-04-23 18:43 ` Nishanth Aravamudan
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=20080523052215.GF13071@wotan.suse.de \
--to=npiggin@suse.de \
--cc=abh@cray.com \
--cc=akpm@linux-foundation.org \
--cc=andi@firstfloor.org \
--cc=kniht@linux.vnet.ibm.com \
--cc=linux-mm@kvack.org \
--cc=nacc@us.ibm.com \
--cc=wli@holomorphy.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).