* [patch] mm: task dirty accounting fix
@ 2009-02-10 4:39 Nick Piggin
2009-02-10 9:12 ` Peter Zijlstra
0 siblings, 1 reply; 3+ messages in thread
From: Nick Piggin @ 2009-02-10 4:39 UTC (permalink / raw)
To: Linux Kernel Mailing List, Peter Zijlstra, YAMAMOTO Takashi,
Andrew Morton
Hi,
This is based from an earlier patch from YAMAMOTO-san which I have
somewhat simplified. What do you think?
Thanks,
Nick
--
YAMAMOTO-san noticed that task_dirty_inc doesn't seem to be called properly for
cases where set_page_dirty is not used to dirty a page (eg. mark_buffer_dirty).
Additionally, there is some inconsistency about when task_dirty_inc is
called. It is used for dirty balancing, however it even gets called for
__set_page_dirty_no_writeback.
So rather than increment it in a set_page_dirty wrapper, move it down to
exactly where the dirty page accounting stats are incremented.
Cc: YAMAMOTO Takashi <yamamoto@valinux.co.jp>
Signed-off-by: Nick Piggin <npiggin@suse.de>
---
fs/buffer.c | 1 +
include/linux/mm.h | 1 +
mm/page-writeback.c | 13 +++----------
3 files changed, 5 insertions(+), 10 deletions(-)
Index: linux-2.6/include/linux/mm.h
===================================================================
--- linux-2.6.orig/include/linux/mm.h
+++ linux-2.6/include/linux/mm.h
@@ -1162,6 +1162,7 @@ extern int filemap_fault(struct vm_area_
/* mm/page-writeback.c */
int write_one_page(struct page *page, int wait);
+void task_dirty_inc(struct task_struct *tsk);
/* readahead.c */
#define VM_MAX_READAHEAD 128 /* kbytes */
Index: linux-2.6/mm/page-writeback.c
===================================================================
--- linux-2.6.orig/mm/page-writeback.c
+++ linux-2.6/mm/page-writeback.c
@@ -241,7 +241,7 @@ void bdi_writeout_inc(struct backing_dev
}
EXPORT_SYMBOL_GPL(bdi_writeout_inc);
-static inline void task_dirty_inc(struct task_struct *tsk)
+void task_dirty_inc(struct task_struct *tsk)
{
prop_inc_single(&vm_dirties, &tsk->dirties);
}
@@ -1228,6 +1228,7 @@ int __set_page_dirty_nobuffers(struct pa
__inc_zone_page_state(page, NR_FILE_DIRTY);
__inc_bdi_stat(mapping->backing_dev_info,
BDI_RECLAIMABLE);
+ task_dirty_inc(current);
task_io_account_write(PAGE_CACHE_SIZE);
}
radix_tree_tag_set(&mapping->page_tree,
@@ -1263,7 +1264,7 @@ EXPORT_SYMBOL(redirty_page_for_writepage
* If the mapping doesn't provide a set_page_dirty a_op, then
* just fall through and assume that it wants buffer_heads.
*/
-static int __set_page_dirty(struct page *page)
+int set_page_dirty(struct page *page)
{
struct address_space *mapping = page_mapping(page);
@@ -1281,14 +1282,6 @@ static int __set_page_dirty(struct page
}
return 0;
}
-
-int set_page_dirty(struct page *page)
-{
- int ret = __set_page_dirty(page);
- if (ret)
- task_dirty_inc(current);
- return ret;
-}
EXPORT_SYMBOL(set_page_dirty);
/*
Index: linux-2.6/fs/buffer.c
===================================================================
--- linux-2.6.orig/fs/buffer.c
+++ linux-2.6/fs/buffer.c
@@ -628,6 +628,7 @@ static int __set_page_dirty(struct page
__inc_zone_page_state(page, NR_FILE_DIRTY);
__inc_bdi_stat(mapping->backing_dev_info,
BDI_RECLAIMABLE);
+ task_dirty_inc(current);
task_io_account_write(PAGE_CACHE_SIZE);
}
radix_tree_tag_set(&mapping->page_tree,
^ permalink raw reply [flat|nested] 3+ messages in thread
* Re: [patch] mm: task dirty accounting fix
2009-02-10 4:39 [patch] mm: task dirty accounting fix Nick Piggin
@ 2009-02-10 9:12 ` Peter Zijlstra
2009-02-12 1:43 ` Nick Piggin
0 siblings, 1 reply; 3+ messages in thread
From: Peter Zijlstra @ 2009-02-10 9:12 UTC (permalink / raw)
To: Nick Piggin; +Cc: Linux Kernel Mailing List, YAMAMOTO Takashi, Andrew Morton
On Tue, 2009-02-10 at 05:39 +0100, Nick Piggin wrote:
> Hi,
>
> This is based from an earlier patch from YAMAMOTO-san which I have
> somewhat simplified. What do you think?
Looks nice. Was a bit surprised we never actually merged the previous
one though..
> Thanks,
> Nick
>
> --
> YAMAMOTO-san noticed that task_dirty_inc doesn't seem to be called properly for
> cases where set_page_dirty is not used to dirty a page (eg. mark_buffer_dirty).
>
> Additionally, there is some inconsistency about when task_dirty_inc is
> called. It is used for dirty balancing, however it even gets called for
> __set_page_dirty_no_writeback.
>
> So rather than increment it in a set_page_dirty wrapper, move it down to
> exactly where the dirty page accounting stats are incremented.
>
> Cc: YAMAMOTO Takashi <yamamoto@valinux.co.jp>
> Signed-off-by: Nick Piggin <npiggin@suse.de>
Acked-by: Peter Zijlstra <a.p.zijlstra@chello.nl>
> ---
> fs/buffer.c | 1 +
> include/linux/mm.h | 1 +
> mm/page-writeback.c | 13 +++----------
> 3 files changed, 5 insertions(+), 10 deletions(-)
>
> Index: linux-2.6/include/linux/mm.h
> ===================================================================
> --- linux-2.6.orig/include/linux/mm.h
> +++ linux-2.6/include/linux/mm.h
> @@ -1162,6 +1162,7 @@ extern int filemap_fault(struct vm_area_
>
> /* mm/page-writeback.c */
> int write_one_page(struct page *page, int wait);
> +void task_dirty_inc(struct task_struct *tsk);
>
> /* readahead.c */
> #define VM_MAX_READAHEAD 128 /* kbytes */
> Index: linux-2.6/mm/page-writeback.c
> ===================================================================
> --- linux-2.6.orig/mm/page-writeback.c
> +++ linux-2.6/mm/page-writeback.c
> @@ -241,7 +241,7 @@ void bdi_writeout_inc(struct backing_dev
> }
> EXPORT_SYMBOL_GPL(bdi_writeout_inc);
>
> -static inline void task_dirty_inc(struct task_struct *tsk)
> +void task_dirty_inc(struct task_struct *tsk)
> {
> prop_inc_single(&vm_dirties, &tsk->dirties);
> }
> @@ -1228,6 +1228,7 @@ int __set_page_dirty_nobuffers(struct pa
> __inc_zone_page_state(page, NR_FILE_DIRTY);
> __inc_bdi_stat(mapping->backing_dev_info,
> BDI_RECLAIMABLE);
> + task_dirty_inc(current);
> task_io_account_write(PAGE_CACHE_SIZE);
> }
> radix_tree_tag_set(&mapping->page_tree,
> @@ -1263,7 +1264,7 @@ EXPORT_SYMBOL(redirty_page_for_writepage
> * If the mapping doesn't provide a set_page_dirty a_op, then
> * just fall through and assume that it wants buffer_heads.
> */
> -static int __set_page_dirty(struct page *page)
> +int set_page_dirty(struct page *page)
> {
> struct address_space *mapping = page_mapping(page);
>
> @@ -1281,14 +1282,6 @@ static int __set_page_dirty(struct page
> }
> return 0;
> }
> -
> -int set_page_dirty(struct page *page)
> -{
> - int ret = __set_page_dirty(page);
> - if (ret)
> - task_dirty_inc(current);
> - return ret;
> -}
> EXPORT_SYMBOL(set_page_dirty);
>
> /*
> Index: linux-2.6/fs/buffer.c
> ===================================================================
> --- linux-2.6.orig/fs/buffer.c
> +++ linux-2.6/fs/buffer.c
> @@ -628,6 +628,7 @@ static int __set_page_dirty(struct page
> __inc_zone_page_state(page, NR_FILE_DIRTY);
> __inc_bdi_stat(mapping->backing_dev_info,
> BDI_RECLAIMABLE);
> + task_dirty_inc(current);
> task_io_account_write(PAGE_CACHE_SIZE);
> }
> radix_tree_tag_set(&mapping->page_tree,
^ permalink raw reply [flat|nested] 3+ messages in thread
* Re: [patch] mm: task dirty accounting fix
2009-02-10 9:12 ` Peter Zijlstra
@ 2009-02-12 1:43 ` Nick Piggin
0 siblings, 0 replies; 3+ messages in thread
From: Nick Piggin @ 2009-02-12 1:43 UTC (permalink / raw)
To: Peter Zijlstra; +Cc: Linux Kernel Mailing List, YAMAMOTO Takashi, Andrew Morton
On Tue, Feb 10, 2009 at 10:12:51AM +0100, Peter Zijlstra wrote:
> On Tue, 2009-02-10 at 05:39 +0100, Nick Piggin wrote:
> > Hi,
> >
> > This is based from an earlier patch from YAMAMOTO-san which I have
> > somewhat simplified. What do you think?
>
> Looks nice. Was a bit surprised we never actually merged the previous
> one though..
So was I, it was always in the back of my mind because I didn't
want to lose a good fix. I was going to just resend it, but I
decided it should actually be more consistent to do this accounting
with the other ones, and the patch ended up being much smaller as
a result too.
But credit to YAMAMOTO san for the core fix.
>
> > Thanks,
> > Nick
> >
> > --
> > YAMAMOTO-san noticed that task_dirty_inc doesn't seem to be called properly for
> > cases where set_page_dirty is not used to dirty a page (eg. mark_buffer_dirty).
> >
> > Additionally, there is some inconsistency about when task_dirty_inc is
> > called. It is used for dirty balancing, however it even gets called for
> > __set_page_dirty_no_writeback.
> >
> > So rather than increment it in a set_page_dirty wrapper, move it down to
> > exactly where the dirty page accounting stats are incremented.
> >
> > Cc: YAMAMOTO Takashi <yamamoto@valinux.co.jp>
> > Signed-off-by: Nick Piggin <npiggin@suse.de>
>
> Acked-by: Peter Zijlstra <a.p.zijlstra@chello.nl>
>
> > ---
> > fs/buffer.c | 1 +
> > include/linux/mm.h | 1 +
> > mm/page-writeback.c | 13 +++----------
> > 3 files changed, 5 insertions(+), 10 deletions(-)
> >
> > Index: linux-2.6/include/linux/mm.h
> > ===================================================================
> > --- linux-2.6.orig/include/linux/mm.h
> > +++ linux-2.6/include/linux/mm.h
> > @@ -1162,6 +1162,7 @@ extern int filemap_fault(struct vm_area_
> >
> > /* mm/page-writeback.c */
> > int write_one_page(struct page *page, int wait);
> > +void task_dirty_inc(struct task_struct *tsk);
> >
> > /* readahead.c */
> > #define VM_MAX_READAHEAD 128 /* kbytes */
> > Index: linux-2.6/mm/page-writeback.c
> > ===================================================================
> > --- linux-2.6.orig/mm/page-writeback.c
> > +++ linux-2.6/mm/page-writeback.c
> > @@ -241,7 +241,7 @@ void bdi_writeout_inc(struct backing_dev
> > }
> > EXPORT_SYMBOL_GPL(bdi_writeout_inc);
> >
> > -static inline void task_dirty_inc(struct task_struct *tsk)
> > +void task_dirty_inc(struct task_struct *tsk)
> > {
> > prop_inc_single(&vm_dirties, &tsk->dirties);
> > }
> > @@ -1228,6 +1228,7 @@ int __set_page_dirty_nobuffers(struct pa
> > __inc_zone_page_state(page, NR_FILE_DIRTY);
> > __inc_bdi_stat(mapping->backing_dev_info,
> > BDI_RECLAIMABLE);
> > + task_dirty_inc(current);
> > task_io_account_write(PAGE_CACHE_SIZE);
> > }
> > radix_tree_tag_set(&mapping->page_tree,
> > @@ -1263,7 +1264,7 @@ EXPORT_SYMBOL(redirty_page_for_writepage
> > * If the mapping doesn't provide a set_page_dirty a_op, then
> > * just fall through and assume that it wants buffer_heads.
> > */
> > -static int __set_page_dirty(struct page *page)
> > +int set_page_dirty(struct page *page)
> > {
> > struct address_space *mapping = page_mapping(page);
> >
> > @@ -1281,14 +1282,6 @@ static int __set_page_dirty(struct page
> > }
> > return 0;
> > }
> > -
> > -int set_page_dirty(struct page *page)
> > -{
> > - int ret = __set_page_dirty(page);
> > - if (ret)
> > - task_dirty_inc(current);
> > - return ret;
> > -}
> > EXPORT_SYMBOL(set_page_dirty);
> >
> > /*
> > Index: linux-2.6/fs/buffer.c
> > ===================================================================
> > --- linux-2.6.orig/fs/buffer.c
> > +++ linux-2.6/fs/buffer.c
> > @@ -628,6 +628,7 @@ static int __set_page_dirty(struct page
> > __inc_zone_page_state(page, NR_FILE_DIRTY);
> > __inc_bdi_stat(mapping->backing_dev_info,
> > BDI_RECLAIMABLE);
> > + task_dirty_inc(current);
> > task_io_account_write(PAGE_CACHE_SIZE);
> > }
> > radix_tree_tag_set(&mapping->page_tree,
^ permalink raw reply [flat|nested] 3+ messages in thread
end of thread, other threads:[~2009-02-12 1:44 UTC | newest]
Thread overview: 3+ messages (download: mbox.gz / follow: Atom feed)
-- links below jump to the message on this page --
2009-02-10 4:39 [patch] mm: task dirty accounting fix Nick Piggin
2009-02-10 9:12 ` Peter Zijlstra
2009-02-12 1:43 ` Nick Piggin
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).