Hi,
> I guess the downside to this is if a reader is reading a large file, or
> several files, sequentially with a small read size (smaller than
> PAGE_SIZE), the pages will be marked active after just one read pass.
> My gut says the benefits of this patch outweigh the cost. I would
> expect
Hi,
I guess the downside to this is if a reader is reading a large file, or
several files, sequentially with a small read size (smaller than
PAGE_SIZE), the pages will be marked active after just one read pass.
My gut says the benefits of this patch outweigh the cost. I would
expect
[EMAIL PROTECTED] wrote:
On the other hand, Andreas suggested only marking it once every 32 calls,
but that required a helper variable. Statistically, jiffies%32 should
end up about the same as a helper variable %32.
This of course, if just calling mark_page_accessed() is actually expensive
On Thu, 15 Mar 2007 14:35:17 EDT, Rik van Riel said:
> [EMAIL PROTECTED] wrote:
> > On Wed, 14 Mar 2007 22:33:17 BST, Andreas Mohr said:
> >
> >> it'd seem we need some kind of state management here to figure out good
> >> intervals of when to call mark_page_accessed() *again* for this page. E.g.
On Thu, Mar 15, 2007 at 06:15:45PM -0500, Dave Kleikamp wrote:
> On Thu, 2007-03-15 at 23:59 +0100, Andrea Arcangeli wrote:
> > On Thu, Mar 15, 2007 at 05:44:01PM +, Hugh Dickins wrote:
> > > who removed the !offset condition, he should be consulted on its
> > > reintroduction.
> >
> > the
On Thu, 2007-03-15 at 23:59 +0100, Andrea Arcangeli wrote:
> On Thu, Mar 15, 2007 at 05:44:01PM +, Hugh Dickins wrote:
> > who removed the !offset condition, he should be consulted on its
> > reintroduction.
>
> the !offset check looks a pretty broken heuristic indeed, it would
> break random
On Thu, Mar 15, 2007 at 03:06:01PM -0700, Andrew Morton wrote:
> On Thu, 15 Mar 2007 22:49:23 +0100
> Andrea Arcangeli <[EMAIL PROTECTED]> wrote:
>
> > On Thu, Mar 15, 2007 at 11:07:35AM -0800, Andrew Morton wrote:
> > > > On Thu, 15 Mar 2007 01:22:45 -0400 (EDT) Ashif Harji <[EMAIL
> > > >
On Thu, Mar 15, 2007 at 05:44:01PM +, Hugh Dickins wrote:
> who removed the !offset condition, he should be consulted on its
> reintroduction.
the !offset check looks a pretty broken heuristic indeed, it would
break random I/O. The real fix is to add a ra.prev_offset along with
ra.prev_page,
On Thu, 15 Mar 2007 22:49:23 +0100
Andrea Arcangeli <[EMAIL PROTECTED]> wrote:
> On Thu, Mar 15, 2007 at 11:07:35AM -0800, Andrew Morton wrote:
> > > On Thu, 15 Mar 2007 01:22:45 -0400 (EDT) Ashif Harji <[EMAIL PROTECTED]>
> > > wrote:
> > > I still think the simple fix of removing the
> > >
On Thu, Mar 15, 2007 at 11:07:35AM -0800, Andrew Morton wrote:
> > On Thu, 15 Mar 2007 01:22:45 -0400 (EDT) Ashif Harji <[EMAIL PROTECTED]>
> > wrote:
> > I still think the simple fix of removing the
> > condition is the best approach, but I'm certainly open to alternatives.
>
> Yes, the
Hi,
On Thu, Mar 15, 2007 at 09:07:39PM +0100, Nick Piggin wrote:
> Well in general we like to help applications that help themselves. It
> is actually a good heuristic, surprisingly. If an application randomly
> accesses the same page (and there is no write activity going on), then
> it would be
On Thu, Mar 15, 2007 at 03:55:08PM -0400, Ashif Harji wrote:
>
> It sounds like people are happy with the fix suggested by Nick. That fix
> is okay with me as it fixes the problem I am having.
>
> I suspect, however, that by not directly detecting the problematic access
> pattern, where the
On Thu, Mar 15, 2007 at 05:44:01PM +, Hugh Dickins wrote:
> On Thu, 15 Mar 2007, Nick Piggin wrote:
> > On Thu, Mar 15, 2007 at 11:56:59AM -0400, Chuck Ebbert wrote:
> > > Ashif Harji wrote:
> > > >
> > > > This patch unconditionally calls mark_page_accessed to prevent pages,
> > > >
On Thu, 15 Mar 2007, Nick Piggin wrote:
On Thu, Mar 15, 2007 at 11:56:59AM -0400, Chuck Ebbert wrote:
Ashif Harji wrote:
This patch unconditionally calls mark_page_accessed to prevent pages,
especially for small files, from being evicted from the page cache
despite frequent access.
> On Thu, 15 Mar 2007 01:22:45 -0400 (EDT) Ashif Harji <[EMAIL PROTECTED]>
> wrote:
> I still think the simple fix of removing the
> condition is the best approach, but I'm certainly open to alternatives.
Yes, the problem of falsely activating pages when the file is read in small
hunks is worse
[EMAIL PROTECTED] wrote:
On Wed, 14 Mar 2007 22:33:17 BST, Andreas Mohr said:
it'd seem we need some kind of state management here to figure out good
intervals of when to call mark_page_accessed() *again* for this page. E.g.
despite non-changing access patterns you could still call
On Thu, 15 Mar 2007, Nick Piggin wrote:
> On Thu, Mar 15, 2007 at 11:56:59AM -0400, Chuck Ebbert wrote:
> > Ashif Harji wrote:
> > >
> > > This patch unconditionally calls mark_page_accessed to prevent pages,
> > > especially for small files, from being evicted from the page cache
> > > despite
On Wed, 14 Mar 2007 22:33:17 BST, Andreas Mohr said:
> it'd seem we need some kind of state management here to figure out good
> intervals of when to call mark_page_accessed() *again* for this page. E.g.
> despite non-changing access patterns you could still call mark_page_accessed(
)
> every 32
Nick Piggin wrote:
A change to make database style random read() workloads perform better, by
calling mark_page_accessed for some non-page-aligned reads broke the case of
< PAGE_CACHE_SIZE files, which will not get their prev_index moved past the
first page.
Combine both heuristics for marking
On Thu, Mar 15, 2007 at 11:56:59AM -0400, Chuck Ebbert wrote:
> Ashif Harji wrote:
> >
> > This patch unconditionally calls mark_page_accessed to prevent pages,
> > especially for small files, from being evicted from the page cache
> > despite frequent access.
> >
> > Signed-off-by: Ashif Harji
Ashif Harji wrote:
>
> This patch unconditionally calls mark_page_accessed to prevent pages,
> especially for small files, from being evicted from the page cache
> despite frequent access.
>
> Signed-off-by: Ashif Harji <[EMAIL PROTECTED]>
>
I like mine better -- it leaves the comment:
From:
Nick Piggin wrote:
On Thu, Mar 15, 2007 at 11:39:14AM +0100, Peter Zijlstra wrote:
On Wed, 2007-03-14 at 15:58 -0400, Ashif Harji wrote:
This patch unconditionally calls mark_page_accessed to prevent pages,
especially for small files, from being evicted from the page cache despite
frequent
Andreas Mohr wrote:
I've been thinking hard how to avoid the mark_page_accessed() starvation in
case of a fixed, (almost) non-changing access state, but this seems hard since
it'd seem we need some kind of state management here to figure out good
intervals of when to call mark_page_accessed()
On Thu, Mar 15, 2007 at 07:46:56AM -0500, Dave Kleikamp wrote:
> On Thu, 2007-03-15 at 01:22 -0400, Ashif Harji wrote:
>
> > I would tend to agree with David that: "Any application doing many
> > tiny-sized reads isn't exactly asking for great performance." As well,
> > applications concerned
On Thu, 2007-03-15 at 01:22 -0400, Ashif Harji wrote:
> I would tend to agree with David that: "Any application doing many
> tiny-sized reads isn't exactly asking for great performance." As well,
> applications concerned with performance and caching problems can read in a
> file in PAGE_SIZE
On Thu, Mar 15, 2007 at 11:39:14AM +0100, Peter Zijlstra wrote:
> On Wed, 2007-03-14 at 15:58 -0400, Ashif Harji wrote:
> > This patch unconditionally calls mark_page_accessed to prevent pages,
> > especially for small files, from being evicted from the page cache despite
> > frequent access.
>
On Wed, 2007-03-14 at 15:58 -0400, Ashif Harji wrote:
> This patch unconditionally calls mark_page_accessed to prevent pages,
> especially for small files, from being evicted from the page cache despite
> frequent access.
Since we're hackling over the use-once stuff again...
/me brings up:
On Wed, 2007-03-14 at 15:58 -0400, Ashif Harji wrote:
This patch unconditionally calls mark_page_accessed to prevent pages,
especially for small files, from being evicted from the page cache despite
frequent access.
Since we're hackling over the use-once stuff again...
/me brings up:
On Thu, Mar 15, 2007 at 11:39:14AM +0100, Peter Zijlstra wrote:
On Wed, 2007-03-14 at 15:58 -0400, Ashif Harji wrote:
This patch unconditionally calls mark_page_accessed to prevent pages,
especially for small files, from being evicted from the page cache despite
frequent access.
Since
On Thu, 2007-03-15 at 01:22 -0400, Ashif Harji wrote:
I would tend to agree with David that: Any application doing many
tiny-sized reads isn't exactly asking for great performance. As well,
applications concerned with performance and caching problems can read in a
file in PAGE_SIZE
On Thu, Mar 15, 2007 at 07:46:56AM -0500, Dave Kleikamp wrote:
On Thu, 2007-03-15 at 01:22 -0400, Ashif Harji wrote:
I would tend to agree with David that: Any application doing many
tiny-sized reads isn't exactly asking for great performance. As well,
applications concerned with
Andreas Mohr wrote:
I've been thinking hard how to avoid the mark_page_accessed() starvation in
case of a fixed, (almost) non-changing access state, but this seems hard since
it'd seem we need some kind of state management here to figure out good
intervals of when to call mark_page_accessed()
Nick Piggin wrote:
On Thu, Mar 15, 2007 at 11:39:14AM +0100, Peter Zijlstra wrote:
On Wed, 2007-03-14 at 15:58 -0400, Ashif Harji wrote:
This patch unconditionally calls mark_page_accessed to prevent pages,
especially for small files, from being evicted from the page cache despite
frequent
Ashif Harji wrote:
This patch unconditionally calls mark_page_accessed to prevent pages,
especially for small files, from being evicted from the page cache
despite frequent access.
Signed-off-by: Ashif Harji [EMAIL PROTECTED]
I like mine better -- it leaves the comment:
From: Chuck
On Thu, Mar 15, 2007 at 11:56:59AM -0400, Chuck Ebbert wrote:
Ashif Harji wrote:
This patch unconditionally calls mark_page_accessed to prevent pages,
especially for small files, from being evicted from the page cache
despite frequent access.
Signed-off-by: Ashif Harji [EMAIL
Nick Piggin wrote:
A change to make database style random read() workloads perform better, by
calling mark_page_accessed for some non-page-aligned reads broke the case of
PAGE_CACHE_SIZE files, which will not get their prev_index moved past the
first page.
Combine both heuristics for marking
On Wed, 14 Mar 2007 22:33:17 BST, Andreas Mohr said:
it'd seem we need some kind of state management here to figure out good
intervals of when to call mark_page_accessed() *again* for this page. E.g.
despite non-changing access patterns you could still call mark_page_accessed(
)
every 32
On Thu, 15 Mar 2007, Nick Piggin wrote:
On Thu, Mar 15, 2007 at 11:56:59AM -0400, Chuck Ebbert wrote:
Ashif Harji wrote:
This patch unconditionally calls mark_page_accessed to prevent pages,
especially for small files, from being evicted from the page cache
despite frequent access.
[EMAIL PROTECTED] wrote:
On Wed, 14 Mar 2007 22:33:17 BST, Andreas Mohr said:
it'd seem we need some kind of state management here to figure out good
intervals of when to call mark_page_accessed() *again* for this page. E.g.
despite non-changing access patterns you could still call
On Thu, 15 Mar 2007 01:22:45 -0400 (EDT) Ashif Harji [EMAIL PROTECTED]
wrote:
I still think the simple fix of removing the
condition is the best approach, but I'm certainly open to alternatives.
Yes, the problem of falsely activating pages when the file is read in small
hunks is worse than
On Thu, 15 Mar 2007, Nick Piggin wrote:
On Thu, Mar 15, 2007 at 11:56:59AM -0400, Chuck Ebbert wrote:
Ashif Harji wrote:
This patch unconditionally calls mark_page_accessed to prevent pages,
especially for small files, from being evicted from the page cache
despite frequent access.
On Thu, Mar 15, 2007 at 05:44:01PM +, Hugh Dickins wrote:
On Thu, 15 Mar 2007, Nick Piggin wrote:
On Thu, Mar 15, 2007 at 11:56:59AM -0400, Chuck Ebbert wrote:
Ashif Harji wrote:
This patch unconditionally calls mark_page_accessed to prevent pages,
especially for small
On Thu, Mar 15, 2007 at 03:55:08PM -0400, Ashif Harji wrote:
It sounds like people are happy with the fix suggested by Nick. That fix
is okay with me as it fixes the problem I am having.
I suspect, however, that by not directly detecting the problematic access
pattern, where the file is
Hi,
On Thu, Mar 15, 2007 at 09:07:39PM +0100, Nick Piggin wrote:
Well in general we like to help applications that help themselves. It
is actually a good heuristic, surprisingly. If an application randomly
accesses the same page (and there is no write activity going on), then
it would be
On Thu, Mar 15, 2007 at 11:07:35AM -0800, Andrew Morton wrote:
On Thu, 15 Mar 2007 01:22:45 -0400 (EDT) Ashif Harji [EMAIL PROTECTED]
wrote:
I still think the simple fix of removing the
condition is the best approach, but I'm certainly open to alternatives.
Yes, the problem of falsely
On Thu, 15 Mar 2007 22:49:23 +0100
Andrea Arcangeli [EMAIL PROTECTED] wrote:
On Thu, Mar 15, 2007 at 11:07:35AM -0800, Andrew Morton wrote:
On Thu, 15 Mar 2007 01:22:45 -0400 (EDT) Ashif Harji [EMAIL PROTECTED]
wrote:
I still think the simple fix of removing the
condition is the
On Thu, Mar 15, 2007 at 05:44:01PM +, Hugh Dickins wrote:
who removed the !offset condition, he should be consulted on its
reintroduction.
the !offset check looks a pretty broken heuristic indeed, it would
break random I/O. The real fix is to add a ra.prev_offset along with
ra.prev_page,
On Thu, Mar 15, 2007 at 03:06:01PM -0700, Andrew Morton wrote:
On Thu, 15 Mar 2007 22:49:23 +0100
Andrea Arcangeli [EMAIL PROTECTED] wrote:
On Thu, Mar 15, 2007 at 11:07:35AM -0800, Andrew Morton wrote:
On Thu, 15 Mar 2007 01:22:45 -0400 (EDT) Ashif Harji [EMAIL
PROTECTED] wrote:
On Thu, 2007-03-15 at 23:59 +0100, Andrea Arcangeli wrote:
On Thu, Mar 15, 2007 at 05:44:01PM +, Hugh Dickins wrote:
who removed the !offset condition, he should be consulted on its
reintroduction.
the !offset check looks a pretty broken heuristic indeed, it would
break random I/O.
I
On Thu, Mar 15, 2007 at 06:15:45PM -0500, Dave Kleikamp wrote:
On Thu, 2007-03-15 at 23:59 +0100, Andrea Arcangeli wrote:
On Thu, Mar 15, 2007 at 05:44:01PM +, Hugh Dickins wrote:
who removed the !offset condition, he should be consulted on its
reintroduction.
the !offset check
On Thu, 15 Mar 2007 14:35:17 EDT, Rik van Riel said:
[EMAIL PROTECTED] wrote:
On Wed, 14 Mar 2007 22:33:17 BST, Andreas Mohr said:
it'd seem we need some kind of state management here to figure out good
intervals of when to call mark_page_accessed() *again* for this page. E.g.
despite
[EMAIL PROTECTED] wrote:
On the other hand, Andreas suggested only marking it once every 32 calls,
but that required a helper variable. Statistically, jiffies%32 should
end up about the same as a helper variable %32.
This of course, if just calling mark_page_accessed() is actually expensive
On Wed, 14 Mar 2007, Xiaoning Ding wrote:
Dave Kleikamp wrote:
On Wed, 2007-03-14 at 22:33 +0100, Andreas Mohr wrote:
Hi,
On Wed, Mar 14, 2007 at 03:55:41PM -0500, Dave Kleikamp wrote:
On Wed, 2007-03-14 at 15:58 -0400, Ashif Harji wrote:
This patch unconditionally calls
Dave Kleikamp wrote:
On Wed, 2007-03-14 at 22:33 +0100, Andreas Mohr wrote:
Hi,
On Wed, Mar 14, 2007 at 03:55:41PM -0500, Dave Kleikamp wrote:
On Wed, 2007-03-14 at 15:58 -0400, Ashif Harji wrote:
This patch unconditionally calls mark_page_accessed to prevent pages,
especially for small
On Wed, 2007-03-14 at 22:33 +0100, Andreas Mohr wrote:
> Hi,
>
> On Wed, Mar 14, 2007 at 03:55:41PM -0500, Dave Kleikamp wrote:
> > On Wed, 2007-03-14 at 15:58 -0400, Ashif Harji wrote:
> > > This patch unconditionally calls mark_page_accessed to prevent pages,
> > > especially for small files,
Hi,
On Wed, Mar 14, 2007 at 03:55:41PM -0500, Dave Kleikamp wrote:
> On Wed, 2007-03-14 at 15:58 -0400, Ashif Harji wrote:
> > This patch unconditionally calls mark_page_accessed to prevent pages,
> > especially for small files, from being evicted from the page cache despite
> > frequent
On Wed, 2007-03-14 at 15:58 -0400, Ashif Harji wrote:
> This patch unconditionally calls mark_page_accessed to prevent pages,
> especially for small files, from being evicted from the page cache despite
> frequent access.
I guess the downside to this is if a reader is reading a large file, or
On Wed, 2007-03-14 at 15:58 -0400, Ashif Harji wrote:
This patch unconditionally calls mark_page_accessed to prevent pages,
especially for small files, from being evicted from the page cache despite
frequent access.
I guess the downside to this is if a reader is reading a large file, or
Hi,
On Wed, Mar 14, 2007 at 03:55:41PM -0500, Dave Kleikamp wrote:
On Wed, 2007-03-14 at 15:58 -0400, Ashif Harji wrote:
This patch unconditionally calls mark_page_accessed to prevent pages,
especially for small files, from being evicted from the page cache despite
frequent access.
I
On Wed, 2007-03-14 at 22:33 +0100, Andreas Mohr wrote:
Hi,
On Wed, Mar 14, 2007 at 03:55:41PM -0500, Dave Kleikamp wrote:
On Wed, 2007-03-14 at 15:58 -0400, Ashif Harji wrote:
This patch unconditionally calls mark_page_accessed to prevent pages,
especially for small files, from being
Dave Kleikamp wrote:
On Wed, 2007-03-14 at 22:33 +0100, Andreas Mohr wrote:
Hi,
On Wed, Mar 14, 2007 at 03:55:41PM -0500, Dave Kleikamp wrote:
On Wed, 2007-03-14 at 15:58 -0400, Ashif Harji wrote:
This patch unconditionally calls mark_page_accessed to prevent pages,
especially for small
On Wed, 14 Mar 2007, Xiaoning Ding wrote:
Dave Kleikamp wrote:
On Wed, 2007-03-14 at 22:33 +0100, Andreas Mohr wrote:
Hi,
On Wed, Mar 14, 2007 at 03:55:41PM -0500, Dave Kleikamp wrote:
On Wed, 2007-03-14 at 15:58 -0400, Ashif Harji wrote:
This patch unconditionally calls
62 matches
Mail list logo