forcing write-back from FS - again
Subsystems (2)
Hi Andrew,
some time ago we were talking about doing write-back from inside a file-system (http://marc.info/?l=linux-kernel&m=119097117713616&w=2). You said that I'm not the only person who needs this, because the same thing is needed for delayed allocation.
The problem is that if we initiate write-back from prepare_write() and we are having a dirty page lock, we deadlock in write_cache_pages() which tries to lock the same page.
You suggested to enhance struct writeback_control and put page that should be skipped.
I tried something like
include/linux/writeback.h +1 -0
@@ -61,6 +61,7 @@ struct writeback_control { unsigned for_reclaim:1; /* Invoked from the page allocator */ unsigned for_writepages:1; /* This is a writepages() call */ unsigned range_cyclic:1; /* range_start is cyclic */ +struct page *skip_pg;/* do not write this page back */ void *fs_private;/* For use by ->writepages() */ };
mm/page-writeback.c +3 -0
@@ -641,6 +641,9 @@ retry:for (i =0 ; i < nr_pages; i++) {struct page *page = pvec.pages[i]; +if (unlikely (page == wbc->skip_pg)) +continue ; +/* * At this point we hold neither mapping->tree_lock nor * lock on the page itself: the page may be truncated
but it does not dot actually work, because if we have two processes forcing write-back from write_page(), they will mutually deadlock (A waits in write_cache_pages() on a page B has locked, B waits on inode or page A has locked).
So this way is not ok, do you have any other ideas?
We could mark page clean temporarily before doing write-back, and mark it dirty again, but this seems to be inefficient (although I'm not sure, need to dig these functions deeper, but they _seem_ to traverse the radix tree and change tags, so marking one page dirty may need to change many tags, but again, I did not really dig tis yet).
I'd appreciate any suggestions. Thanks!
Signature · 2 lines
Best Regards, Artem Bityutskiy (Артём Битюцкий)
Andrew Morton
maintainerQuoted 11 lines · On Sun, 21 Oct 2007 23:19:41 +0300 Artem Bityutskiy…
Hi Andrew,
some time ago we were talking about doing write-back from inside a file-system
(http://marc.info/?l=linux-kernel&m=119097117713616&w=2). You said that I'm not the only person who needs this, because the same thing is needed for delayed allocation.The problem is that if we initiate write-back from
prepare_write()and we are having a dirty page lock, we deadlock inwrite_cache_pages()which tries to lock the same page.
Quoted 8 lines
You suggested to enhance struct
writeback_controland put page that should be skipped....
but it does not dot actually work, because if we have two processes forcing write-back from
write_page(), they will mutually deadlock (A waits inwrite_cache_pages()on a page B has locked, B waits on inode or page A has locked).
Yeah, I was just thinking that as I read this ;)
Quoted 9 lines
So this way is not ok, do you have any other ideas?
We could mark page clean temporarily before doing write-back, and mark it dirty again, but this seems to be inefficient (although I'm not sure, need to dig these functions deeper, but they _seem_ to traverse the radix tree and change tags, so marking one page dirty may need to change many tags, but again, I did not really dig tis yet).
I'd appreciate any suggestions. Thanks!
We could just skip locked pages altogether in writeback. Perhaps in WB_SYNC_NONE mode, or perhaps add a new flag in writeback_control to select this behaviour.
It _should_ be the case that the number of locked-and-dirty pages which writeback encounters is very small, so skipping locked pages during writeback-for-memory-flushing won't have any significant effect. The first step should be to add a new /proc/vmstat field to count these pages and then do broad testing (especially on blocksize<pagesize filesystems) to confirm the theory.
We'll still need to synchronously lock the page in writeback-for-data-integrity mode though.
Quoted 11 lines · Andrew Morton
Quoted 1 line
but it does not dot actually work, because if we have two processes forcing write-back from
write_page(), they will mutually deadlock (A waits inwrite_cache_pages()on a page B has locked, B waits on inode or page A has locked).Yeah, I was just thinking that as I read this ;)
Quoted 3 lines
So this way is not ok, do you have any other ideas?
We could mark page clean temporarily before doing write-back, and mark it dirty again, but this seems to be inefficient (although I'm not sure, need to dig these functions deeper, but they _seem_ to traverse the radix tree and change tags, so marking one page dirty may need to change many tags, but again, I did not really dig tis yet).
We could just skip locked pages altogether in writeback. Perhaps in
WB_SYNC_NONEmode, or perhaps add a new flag inwriteback_controlto select this behaviour.
Yeah, certanly not WB_SYNC_ALL, because this is a deadlocky - the process which forces write-back from the ->prepare_write() is having page X locked, pdflush may have some inode A locked and sleep on page X, while the FS would sleep on inode A.
Quoted 9 lines
It _should_ be the case that the number of locked-and-dirty pages which writeback encounters is very small, so skipping locked pages during writeback-for-memory-flushing won't have any significant effect. The first step should be to add a new
/proc/vmstatfield to count these pages and then do broad testing (especially on blocksize<pagesize filesystems) to confirm the theory.We'll still need to synchronously lock the page in writeback-for-data-integrity mode though.
Thanks for suggestion. It sounds as a separate big job to enhance existing WB_SYNC_NONE. I've just introduced new WB mode, and it seems to work fine:
include/linux/writeback.h +1 -0
@@ -28,6 +28,7 @@ static inline int task_is_pdflush(struct task_struct *task)*/ enum writeback_sync_modes {WB_SYNC_NONE ,/* Don't wait on anything */ +WB_SYNC_NONE_PG ,/* Don't wait on anything, don't touch locked pages */ WB_SYNC_ALL ,/* Wait on every mapping */ WB_SYNC_HOLD ,/* Hold the inode on sb_dirty for sys_sync() */ };
mm/page-writeback.c +4 -0
@@ -641,6 +641,10 @@ retry:for (i =0 ; i < nr_pages; i++) {struct page *page = pvec.pages[i]; +if (wbc->sync_mode ==WB_SYNC_NONE_PG && +PageLocked (page)) +continue ; +
My only concern is - what if the page we skipped because of WB_SYNC_NONE_PG will somehow loose its dirty TAG and will never be written-back? But it is because of my poor knowledge of Linux MM internals. Could you please comment on this?
Thanks!
Signature · 2 lines
Best Regards, Artem Bityutskiy (Артём Битюцкий)
Andrew Morton
maintainerQuoted 15 lines · On Mon, 22 Oct 2007 11:52:33 +0300 Artem Bityutskiy <[email protected]> wrote:
Quoted 9 lines
It _should_ be the case that the number of locked-and-dirty pages which writeback encounters is very small, so skipping locked pages during writeback-for-memory-flushing won't have any significant effect. The first step should be to add a new
/proc/vmstatfield to count these pages and then do broad testing (especially on blocksize<pagesize filesystems) to confirm the theory.We'll still need to synchronously lock the page in writeback-for-data-integrity mode though.
Thanks for suggestion. It sounds as a separate big job to enhance existing
WB_SYNC_NONE. I've just introduced new WB mode, and it seems to work fine:
include/linux/writeback.h static inline int task_is_pdflush(struct task_struct *task) · line 34
Quoted 7 lines
include/linux/writeback.h +1
-0@@ -28,6 +28,7 @@ static inline int task_is_pdflush(struct task_struct *task)*/ enum writeback_sync_modes {WB_SYNC_NONE ,/* Don't wait on anything */ +WB_SYNC_NONE_PG ,/* Don't wait on anything, don't touch locked pages */ WB_SYNC_ALL ,/* Wait on every mapping */ WB_SYNC_HOLD ,/* Hold the inode on sb_dirty for sys_sync() */ };
It would be simpler/safer/saner to add a new bitflag to writeback_control and use that directly. The WB_SYNC_foo flags are a holdover from an earlier time and really should be made to go away, in favour of directly setting up an appropriate writeback_control.
Quoted 10 lines
mm/page-writeback.c +4
-0@@ -641,6 +641,10 @@ retry:for (i =0 ; i < nr_pages; i++) {struct page *page = pvec.pages[i]; +if (wbc->sync_mode ==WB_SYNC_NONE_PG && +PageLocked (page)) +continue ; +
Quoted 4 lines
My only concern is - what if the page we skipped because of
WB_SYNC_NONE_PGwill somehow loose its dirty TAG and will never be written-back? But it is because of my poor knowledge of Linux MM internals. Could you please comment on this?
Well it might lose its dirty tag, if the thread which has a lock on the page is about to write it out or truncate it. But that shouldn't concern you here.
The code you have there looks racy: if someone else locks the page in that little window after the PageLocked() test we'll still block in lock_page().
That's unlikely to happen in your application (apart from a remaining
ab/ba scenario) but we should make it robust:
if (wbc->skip_locked_pages) {
if (TestSetPageLocked(page))
continue;
} else {
lock_page(page);
}Quoted 6 lines · Andrew Morton
Quoted 6 lines
include/linux/writeback.h +1
-0@@ -28,6 +28,7 @@ static inline int task_is_pdflush(struct task_struct *task)*/ enum writeback_sync_modes {WB_SYNC_NONE ,/* Don't wait on anything */ +WB_SYNC_NONE_PG ,/* Don't wait on anything, don't touch locked pages */
Quoted 8 lines
include/linux/writeback.h static inline int task_is_pdflush(struct task_struct *task) · line 34
Quoted 3 lines
include/linux/writeback.h +0
-0WB_SYNC_ALL ,/* Wait on every mapping */ WB_SYNC_HOLD, /* Hold the inode on sb_dirty for sys_sync() */ };It would be simpler/safer/saner to add a new bitflag to
writeback_controland use that directly. TheWB_SYNC_fooflags are a holdover from an earlier time and really should be made to go away, in favour of directly setting up an appropriatewriteback_control.
You mean something like (wbc->flags & WB_SYNC_NONE) etc? But below you used wbc->skip_locked_pages, I'm confused.
Quoted 11 lines
The code you have there looks racy: if someone else locks the page in that little window after the
PageLocked()test we'll still block inlock_page().
That's unlikely to happen in your application (apart from a remaining
ab/ba scenario) but we should make it robust:if (wbc->skip_locked_pages) { if (TestSetPageLocked(page)) continue; } else { lock_page(page); }
Yeah, thanks for the pointer! Thank you for your help on the issue!
Signature · 2 lines
Best Regards, Artem Bityutskiy (Артём Битюцкий)
Andrew Morton
maintainerQuoted 16 lines · On Mon, 22 Oct 2007 12:38:48 +0300 Artem Bityutskiy <[email protected]> wrote:
Quoted 6 lines · Andrew Morton
Quoted 6 lines
include/linux/writeback.h +1
-0@@ -28,6 +28,7 @@ static inline int task_is_pdflush(struct task_struct *task)*/ enum writeback_sync_modes {WB_SYNC_NONE ,/* Don't wait on anything */ +WB_SYNC_NONE_PG ,/* Don't wait on anything, don't touch locked pages */ Quoted 8 lines
include/linux/writeback.h static inline int task_is_pdflush(struct task_struct *task) · line 34
Quoted 3 lines
include/linux/writeback.h +0
-0WB_SYNC_ALL ,/* Wait on every mapping */ WB_SYNC_HOLD, /* Hold the inode on sb_dirty for sys_sync() */ };It would be simpler/safer/saner to add a new bitflag to
writeback_controland use that directly. TheWB_SYNC_fooflags are a holdover from an earlier time and really should be made to go away, in favour of directly setting up an appropriatewriteback_control.
Quoted 2 lines
You mean something like (
wbc->flags&WB_SYNC_NONE) etc? But below you usedwbc->skip_locked_pages, I'm confused.
take a look at struct writeback_control:
unsigned nonblocking:1; /* Don't get stuck on request queues */
unsigned encountered_congestion:1; /* An output: a queue is full */
unsigned for_kupdate:1; /* A kupdate writeback */
unsigned for_reclaim:1; /* Invoked from the page allocator */
unsigned for_writepages:1; /* This is a writepages() call */
unsigned range_cyclic:1; /* range_start is cyclic */
unsigned more_io:1; /* more io to be dispatched */
Add another one...
Quoted 11 lines · Andrew Morton
take a look at struct
writeback_control:unsigned nonblocking:1; /* Don't get stuck on request queues */
unsigned encountered_congestion:1; /* An output: a queue is full */
unsigned for_kupdate:1; /* A kupdate writeback */
unsigned for_reclaim:1; /* Invoked from the page allocator */
unsigned for_writepages:1; /* This is awritepages()call */
unsigned range_cyclic:1; /*range_startis cyclic */
unsigned more_io:1; /* more io to be dispatched */Add another one...
Ups, OK! Thanks again!
Signature · 2 lines
Best Regards, Artem Bityutskiy (Артём Битюцкий)