From: Andrew Morton <akpm@linux-foundation.org>
To: "J. Bruce Fields" <bfields@fieldses.org>
Cc: NeilBrown <neilb@suse.de>,
nfs@lists.sourceforge.net, linux-kernel@vger.kernel.org
Subject: Re: [PATCH 007 of 8] knfsd: nfsd4: vary maximum delegation limit based on RAM size
Date: Wed, 27 Jun 2007 19:36:13 -0700 [thread overview]
Message-ID: <20070627193613.746dfbf3.akpm@linux-foundation.org> (raw)
In-Reply-To: <20070628021551.GB20605@fieldses.org>
On Wed, 27 Jun 2007 22:15:52 -0400 "J. Bruce Fields" <bfields@fieldses.org> wrote:
> + /*
> + * Allow at most 4 delegations per megabyte of RAM. Quick
> + * estimates suggest that in the worst case (where every delegation
> + * is for a different inode), a delegation could take about 1.5K,
> + * giving a worst case usage of about 6% of memory.
> + */
> sys.totalram >>= (18 - PAGE_SHIFT);
6% of 32GB is 2GB is too much for a large highmem machine. We should base
this on the amount of kernel-addressable memory (aka lowmem).
How's this?
static void
set_max_delegations(void)
{
/*
* Allow at most 4 delegations per megabyte of RAM. Quick
* estimates suggest that in the worst case (where every delegation
* is for a different inode), a delegation could take about 1.5K,
* giving a worst case usage of about 6% of memory.
*/
max_delegations = nr_free_buffer_pages() >> (20 - 2 - PAGE_SHIFT);
}
fs/nfsd/nfs4state.c | 7 ++-----
mm/page_alloc.c | 1 +
diff -puN fs/nfsd/nfs4state.c~knfsd-nfsd4-vary-maximum-delegation-limit-based-on-ram-size-fix-fix-fix-fix fs/nfsd/nfs4state.c
--- a/fs/nfsd/nfs4state.c~knfsd-nfsd4-vary-maximum-delegation-limit-based-on-ram-size-fix-fix-fix-fix
+++ a/fs/nfsd/nfs4state.c
@@ -49,6 +49,7 @@
#include <linux/nfsd/state.h>
#include <linux/nfsd/xdr4.h>
#include <linux/namei.h>
+#include <linux/swap.h>
#include <linux/mutex.h>
#include <linux/lockd/bind.h>
#include <linux/module.h>
@@ -3210,17 +3211,13 @@ get_nfs4_grace_period(void)
static void
set_max_delegations(void)
{
- struct sysinfo sys;
-
- si_meminfo(&sys);
/*
* Allow at most 4 delegations per megabyte of RAM. Quick
* estimates suggest that in the worst case (where every delegation
* is for a different inode), a delegation could take about 1.5K,
* giving a worst case usage of about 6% of memory.
*/
- sys.totalram >>= (18 - PAGE_SHIFT);
- max_delegations = (unsigned int) sys.totalram;
+ max_delegations = nr_free_buffer_pages() >> (20 - 2 - PAGE_SHIFT);
}
/* initialization to perform when the nfsd service is started: */
diff -puN mm/page_alloc.c~knfsd-nfsd4-vary-maximum-delegation-limit-based-on-ram-size-fix-fix-fix-fix mm/page_alloc.c
--- a/mm/page_alloc.c~knfsd-nfsd4-vary-maximum-delegation-limit-based-on-ram-size-fix-fix-fix-fix
+++ a/mm/page_alloc.c
@@ -1729,6 +1729,7 @@ unsigned int nr_free_buffer_pages(void)
{
return nr_free_zone_pages(gfp_zone(GFP_USER));
}
+EXPORT_SYMBOL_GPL(nr_free_buffer_pages);
/*
* Amount of free RAM allocatable within all zones
_
-------------------------------------------------------------------------
This SF.net email is sponsored by DB2 Express
Download DB2 Express C - the FREE version of DB2 express and take
control of your XML. No limits. Just data. Click to get it now.
http://sourceforge.net/powerbar/db2/
_______________________________________________
NFS maillist - NFS@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/nfs
WARNING: multiple messages have this Message-ID (diff)
From: Andrew Morton <akpm@linux-foundation.org>
To: "J. Bruce Fields" <bfields@fieldses.org>
Cc: NeilBrown <neilb@suse.de>,
nfs@lists.sourceforge.net, linux-kernel@vger.kernel.org
Subject: Re: [PATCH 007 of 8] knfsd: nfsd4: vary maximum delegation limit based on RAM size
Date: Wed, 27 Jun 2007 19:36:13 -0700 [thread overview]
Message-ID: <20070627193613.746dfbf3.akpm@linux-foundation.org> (raw)
In-Reply-To: <20070628021551.GB20605@fieldses.org>
On Wed, 27 Jun 2007 22:15:52 -0400 "J. Bruce Fields" <bfields@fieldses.org> wrote:
> + /*
> + * Allow at most 4 delegations per megabyte of RAM. Quick
> + * estimates suggest that in the worst case (where every delegation
> + * is for a different inode), a delegation could take about 1.5K,
> + * giving a worst case usage of about 6% of memory.
> + */
> sys.totalram >>= (18 - PAGE_SHIFT);
6% of 32GB is 2GB is too much for a large highmem machine. We should base
this on the amount of kernel-addressable memory (aka lowmem).
How's this?
static void
set_max_delegations(void)
{
/*
* Allow at most 4 delegations per megabyte of RAM. Quick
* estimates suggest that in the worst case (where every delegation
* is for a different inode), a delegation could take about 1.5K,
* giving a worst case usage of about 6% of memory.
*/
max_delegations = nr_free_buffer_pages() >> (20 - 2 - PAGE_SHIFT);
}
fs/nfsd/nfs4state.c | 7 ++-----
mm/page_alloc.c | 1 +
diff -puN fs/nfsd/nfs4state.c~knfsd-nfsd4-vary-maximum-delegation-limit-based-on-ram-size-fix-fix-fix-fix fs/nfsd/nfs4state.c
--- a/fs/nfsd/nfs4state.c~knfsd-nfsd4-vary-maximum-delegation-limit-based-on-ram-size-fix-fix-fix-fix
+++ a/fs/nfsd/nfs4state.c
@@ -49,6 +49,7 @@
#include <linux/nfsd/state.h>
#include <linux/nfsd/xdr4.h>
#include <linux/namei.h>
+#include <linux/swap.h>
#include <linux/mutex.h>
#include <linux/lockd/bind.h>
#include <linux/module.h>
@@ -3210,17 +3211,13 @@ get_nfs4_grace_period(void)
static void
set_max_delegations(void)
{
- struct sysinfo sys;
-
- si_meminfo(&sys);
/*
* Allow at most 4 delegations per megabyte of RAM. Quick
* estimates suggest that in the worst case (where every delegation
* is for a different inode), a delegation could take about 1.5K,
* giving a worst case usage of about 6% of memory.
*/
- sys.totalram >>= (18 - PAGE_SHIFT);
- max_delegations = (unsigned int) sys.totalram;
+ max_delegations = nr_free_buffer_pages() >> (20 - 2 - PAGE_SHIFT);
}
/* initialization to perform when the nfsd service is started: */
diff -puN mm/page_alloc.c~knfsd-nfsd4-vary-maximum-delegation-limit-based-on-ram-size-fix-fix-fix-fix mm/page_alloc.c
--- a/mm/page_alloc.c~knfsd-nfsd4-vary-maximum-delegation-limit-based-on-ram-size-fix-fix-fix-fix
+++ a/mm/page_alloc.c
@@ -1729,6 +1729,7 @@ unsigned int nr_free_buffer_pages(void)
{
return nr_free_zone_pages(gfp_zone(GFP_USER));
}
+EXPORT_SYMBOL_GPL(nr_free_buffer_pages);
/*
* Amount of free RAM allocatable within all zones
_
next prev parent reply other threads:[~2007-06-28 2:36 UTC|newest]
Thread overview: 32+ messages / expand[flat|nested] mbox.gz Atom feed top
2007-06-21 4:30 [PATCH 000 of 8] knfsd: Assorted nfsv4 server patches NeilBrown
2007-06-21 4:30 ` NeilBrown
2007-06-21 4:30 ` [PATCH 001 of 8] knfsd: lockd: nfsd4: use same grace period for lockd and nfsd4 NeilBrown
2007-06-21 4:30 ` NeilBrown
2007-06-21 4:30 ` [PATCH 002 of 8] knfsd: nfsd4: fix NFSv4 filehandle size units confusion NeilBrown
2007-06-21 4:30 ` NeilBrown
2007-06-21 4:30 ` [PATCH 003 of 8] knfsd: nfsd4: silence a compiler warning in ACL code NeilBrown
2007-06-21 4:30 ` NeilBrown
2007-06-21 4:30 ` [PATCH 004 of 8] knfsd: nfsd4: fix enc_stateid_sz for nfsd callbacks NeilBrown
2007-06-21 4:30 ` NeilBrown
2007-06-21 4:30 ` [PATCH 005 of 8] knfsd: nfsd4: fix handling of acl errrors NeilBrown
2007-06-21 4:30 ` NeilBrown
2007-06-21 4:31 ` [PATCH 006 of 8] knfsd: nfsd: remove unused header interface.h NeilBrown
2007-06-21 4:31 ` NeilBrown
2007-06-21 4:31 ` [PATCH 007 of 8] knfsd: nfsd4: vary maximum delegation limit based on RAM size NeilBrown
2007-06-21 4:31 ` NeilBrown
2007-06-21 16:15 ` J. Bruce Fields
2007-06-26 3:52 ` Andrew Morton
2007-06-26 3:52 ` Andrew Morton
2007-06-28 2:15 ` J. Bruce Fields
2007-06-28 2:15 ` J. Bruce Fields
2007-06-28 2:36 ` Andrew Morton [this message]
2007-06-28 2:36 ` Andrew Morton
2007-06-28 2:57 ` J. Bruce Fields
2007-06-28 2:57 ` J. Bruce Fields
2007-06-28 3:10 ` Andrew Morton
2007-06-28 3:10 ` Andrew Morton
2007-06-21 4:31 ` [PATCH 008 of 8] knfsd: nfsd4: don't delegate files that have had conflicts NeilBrown
2007-06-21 4:31 ` NeilBrown
2007-06-21 16:28 ` J. Bruce Fields
2007-06-21 16:33 ` J. Bruce Fields
2007-06-21 16:33 ` J. Bruce Fields
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=20070627193613.746dfbf3.akpm@linux-foundation.org \
--to=akpm@linux-foundation.org \
--cc=bfields@fieldses.org \
--cc=linux-kernel@vger.kernel.org \
--cc=neilb@suse.de \
--cc=nfs@lists.sourceforge.net \
/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 an external index of several public inboxes,
see mirroring instructions on how to clone and mirror
all data and code used by this external index.