RE: [PATCH V8 4/8] mm/fs: add hooks to support cleancache
From: Minchan Kim [mailto:minchan@gmail.com] On Sat, Apr 16, 2011 at 3:53 AM, Dan Magenheimer dan.magenhei...@oracle.com wrote: From: OGAWA Hirofumi [mailto:hirof...@mail.parknet.co.jp] Andrew Morton a...@linux-foundation.org writes: Andrew, I wonder if you would be so kind as to read the following and make a ruling. If you determine a preferable set of names, I will abide by your decision and repost (if necessary). The problem is this: The English language has a limited number of words that can be used to represent data motion and mapping and most/all of them are already used in the kernel, often, to quote Andrew, in confusing ways. Complicating this, I think the semantics of the cleancache operations are different from the semantics of any other kernel operation... intentionally so, because the value of cleancache is a direct result of those differing semantics. And the cleancache semantics are fairly complex (again intentionally) so a single function name can't possibly describe the semantics. The cleancache operations are: - put (page) - get (page) - flush page - flush inode - init fs - flush fs I think these names are reasonable representations of the semantics of the operations performed... but I'm not a kernel expert so there is certainly room for disagreement. Though I absolutely recognize the importance of a name, I am primarily interested in merging the semantics of the operations and would happily accept any name that kernel developers could agree on. However, I fear that there will be NO name that will satisfy all, so would prefer to keep the existing names. If some renaming is eventually agreed upon, this could be done post-merge. Here's a brief description of the semantics: : semantics for other operations elided : The cleancache operation currently known as get has the following semantics: Derive the filesystem-determined handle from this struct page. If cleancache contains a page matching that handle, recreate the page of data from cleancache and place the results in the pageframe referred to by the struct page. Then delete in cleancache any record of the handle and any data associated with it, so that a subsequent get will no longer find a match for the handle; any space used for the data can also be freed. : semantics for other operations elided : At least, I didn't confused your semantics except just flush. That's why I suggested only flush but after seeing your explaining, there is another thing I want to change. The get/put is common semantic of reference counting in kernel but in POV your semantics, it makes sense to me but get has a exclusive semantic so I want to represent it with API name. Maybe cleancache_get_page_exclusive. The summary is that I don't want to change all API name. Just two thing. (I am not sure you and others agree on me. It's just suggestion). 1. cleancache_flush_page - cleancache_[invalidate|remove]_page 2. cleancache_get_page - cleancache_get_page_exclusive Hi Minchan -- Thanks for continuing to be interested in this and sorry for my delayed response. Actually, your comment about get_page_exclusive points out an incompleteness in my description of the semantics for cleancache_get_page. First, I forgot to list cleancache_init_shared_fs, which is the equivalent of cleancache_init_fs but used for clustered filesystems. (Support is included in the patch for ocfs2 but I haven't played with it in quite some time and my focus has been on the other filesystems, so it slipped my mind :-} The cleancache_get_page operation has a slightly different semantics depending on which of the init_fs calls was used. However, the location of the cleancache_get_page hook is the same regardless of the fs, so the name of the operation must represent both semantics. In the case of init_fs (non-shared), the behavior of cleancache_get_page is that the get is destructive; the page is removed from cleancache on a successful get. In the case of a init_shared_fs, however, the get is non-destructive; the page is NOT removed from cleancache. When cleancache contains pages from multiple kernels (e.g. Xen guests or different machines in a RAMster cluster), this semantic difference can make a big performance difference for a clustered filesystem. Since zcache only contains pages for a single kernel, the difference is moot. Because of this, I am hesitant to add exclusive to the end of the name of the operation. BTW, Nice description. Please include it in documentation if we can't reach the conclusion. It will help others to understand semantic of cleancache. Thanks! Nearly all of the description already exists in various places in the patch but I agree that it would be good if I add a new section to the Documentation file with the exact semantics. Dan -- To unsubscribe from this list: send the line unsubscribe linux-btrfs
Re: [PATCH V8 4/8] mm/fs: add hooks to support cleancache
On Sat, Apr 16, 2011 at 3:53 AM, Dan Magenheimer
dan.magenhei...@oracle.com wrote:
From: OGAWA Hirofumi [mailto:hirof...@mail.parknet.co.jp]
Andrew Morton a...@linux-foundation.org writes:
Before I suggested a thing about cleancache_flush_page,
cleancache_flush_inode.
what's the meaning of flush's semantic?
I thought it means invalidation.
AFAIC, how about change flush with invalidate?
I'm not sure the words flush and invalidate are defined
precisely or used consistently everywhere in computer
science, but I think that invalidate is to destroy
a pointer to some data, but not necessarily destroy the
data itself. And flush means to actually remove
the data. So one would invalidate a mapping but one
would flush a cache.
Since cleancache_flush_page and cleancache_flush_inode
semantically remove data from cleancache, I think flush
is a better name than invalidate.
Does that make sense?
nope ;)
Kernel code freely uses flush to refer to both invalidation and to
writeback, sometimes in confusing ways. In this case,
cleancache_flush_inode and cleancache_flush_page rather sound like
they
might write those things to backing store.
I'd like to mention about *_{get,put}_page too. In linux get/put is not
meaning read/write. There is {get,put}_page those are refcount stuff
(Yeah, and I felt those methods does refcount by quick read. But it
seems to be false. There is no xen codes, so I don't know actually
though.).
And I agree, I also think the needing thing is consistency on the linux
codes (term).
Thanks.
--
OGAWA Hirofumi hirof...@mail.parknet.co.jp
Hmmm, yes, that's a point of confusion also. No, cleancache put/get
do not have any relationship with reference counting.
Andrew, I wonder if you would be so kind as to read the following
and make a ruling. If you determine a preferable set of names,
I will abide by your decision and repost (if necessary).
The problem is this: The English language has a limited number
of words that can be used to represent data motion and mapping
and most/all of them are already used in the kernel, often,
to quote Andrew, in confusing ways. Complicating this, I
think the semantics of the cleancache operations are different
from the semantics of any other kernel operation... intentionally
so, because the value of cleancache is a direct result of those
differing semantics. And the cleancache semantics
are fairly complex (again intentionally) so a single function
name can't possibly describe the semantics.
The cleancache operations are:
- put (page)
- get (page)
- flush page
- flush inode
- init fs
- flush fs
I think these names are reasonable representations of the
semantics of the operations performed... but I'm not a kernel
expert so there is certainly room for disagreement. Though I
absolutely recognize the importance of a name, I am primarily
interested in merging the semantics of the operations and
would happily accept any name that kernel developers could
agree on. However, I fear that there will be NO name that
will satisfy all, so would prefer to keep the existing names.
If some renaming is eventually agreed upon, this could be done
post-merge.
Here's a brief description of the semantics:
The cleancache operation currently known as put has the
following semantics: If *possible*, please take the data
contained in the pageframe referred to by this struct page
into cleancache and associate it with the filesystem-determined
handle derived from the struct page.
The cleancache operation currently known as get has the
following semantics: Derive the filesystem-determined handle
from this struct page. If cleancache contains a page matching
that handle, recreate the page of data from cleancache and
place the results in the pageframe referred to by the
struct page. Then delete in cleancache any record of the
handle and any data associated with it, so that a
subsequent get will no longer find a match for the handle;
any space used for the data can also be freed.
(Note that take the data and recreate the page of data are
similar in semantics to copy to and copy from, but since
the cleancache operation may perform an inflight transformation
on the data, and copy usually means a byte-for-byte replication,
the word copy is also misleading.)
The cleancache operation currently known as flush has the
following semantics: Derive the filesystem-determined handle
from this struct page and struct mapping. If cleancache
contains a page matching that handle, delete in cleancache any
record of the handle and any data associated with it, so that a
subsequent get will no longer find a match for the handle;
any space used for the data can also be freed
The cleancache operation currently known as flush inode has
the following semantics: Derive the filesystem-determined filekey
from this struct mapping. If cleancache contains ANY handles
matching
RE: [PATCH V8 4/8] mm/fs: add hooks to support cleancache
Hi Minchan -- First of all, thanks for resolving conflict with my patch. You're welcome! As I pointed out offlist, yours was the first change in MM that caused any semantic changes to the cleancache core hooks patch since before 2.6.18. Before I suggested a thing about cleancache_flush_page, cleancache_flush_inode. what's the meaning of flush's semantic? I thought it means invalidation. AFAIC, how about change flush with invalidate? I'm not sure the words flush and invalidate are defined precisely or used consistently everywhere in computer science, but I think that invalidate is to destroy a pointer to some data, but not necessarily destroy the data itself. And flush means to actually remove the data. So one would invalidate a mapping but one would flush a cache. Since cleancache_flush_page and cleancache_flush_inode semantically remove data from cleancache, I think flush is a better name than invalidate. Does that make sense? Thanks, Dan -- To unsubscribe from this list: send the line unsubscribe linux-btrfs in the body of a message to majord...@vger.kernel.org More majordomo info at http://vger.kernel.org/majordomo-info.html
Re: [PATCH V8 4/8] mm/fs: add hooks to support cleancache
On Fri, 15 Apr 2011 07:47:57 -0700 (PDT) Dan Magenheimer dan.magenhei...@oracle.com wrote: Hi Minchan -- First of all, thanks for resolving conflict with my patch. You're welcome! As I pointed out offlist, yours was the first change in MM that caused any semantic changes to the cleancache core hooks patch since before 2.6.18. Before I suggested a thing about cleancache_flush_page, cleancache_flush_inode. what's the meaning of flush's semantic? I thought it means invalidation. AFAIC, how about change flush with invalidate? I'm not sure the words flush and invalidate are defined precisely or used consistently everywhere in computer science, but I think that invalidate is to destroy a pointer to some data, but not necessarily destroy the data itself. And flush means to actually remove the data. So one would invalidate a mapping but one would flush a cache. Since cleancache_flush_page and cleancache_flush_inode semantically remove data from cleancache, I think flush is a better name than invalidate. Does that make sense? nope ;) Kernel code freely uses flush to refer to both invalidation and to writeback, sometimes in confusing ways. In this case, cleancache_flush_inode and cleancache_flush_page rather sound like they might write those things to backing store. -- To unsubscribe from this list: send the line unsubscribe linux-btrfs in the body of a message to majord...@vger.kernel.org More majordomo info at http://vger.kernel.org/majordomo-info.html
Re: [PATCH V8 4/8] mm/fs: add hooks to support cleancache
Andrew Morton a...@linux-foundation.org writes:
Before I suggested a thing about cleancache_flush_page,
cleancache_flush_inode.
what's the meaning of flush's semantic?
I thought it means invalidation.
AFAIC, how about change flush with invalidate?
I'm not sure the words flush and invalidate are defined
precisely or used consistently everywhere in computer
science, but I think that invalidate is to destroy
a pointer to some data, but not necessarily destroy the
data itself. And flush means to actually remove
the data. So one would invalidate a mapping but one
would flush a cache.
Since cleancache_flush_page and cleancache_flush_inode
semantically remove data from cleancache, I think flush
is a better name than invalidate.
Does that make sense?
nope ;)
Kernel code freely uses flush to refer to both invalidation and to
writeback, sometimes in confusing ways. In this case,
cleancache_flush_inode and cleancache_flush_page rather sound like they
might write those things to backing store.
I'd like to mention about *_{get,put}_page too. In linux get/put is not
meaning read/write. There is {get,put}_page those are refcount stuff
(Yeah, and I felt those methods does refcount by quick read. But it
seems to be false. There is no xen codes, so I don't know actually
though.).
And I agree, I also think the needing thing is consistency on the linux
codes (term).
Thanks.
--
OGAWA Hirofumi hirof...@mail.parknet.co.jp
--
To unsubscribe from this list: send the line unsubscribe linux-btrfs in
the body of a message to majord...@vger.kernel.org
More majordomo info at http://vger.kernel.org/majordomo-info.html
RE: [PATCH V8 4/8] mm/fs: add hooks to support cleancache
From: OGAWA Hirofumi [mailto:hirof...@mail.parknet.co.jp]
Andrew Morton a...@linux-foundation.org writes:
Before I suggested a thing about cleancache_flush_page,
cleancache_flush_inode.
what's the meaning of flush's semantic?
I thought it means invalidation.
AFAIC, how about change flush with invalidate?
I'm not sure the words flush and invalidate are defined
precisely or used consistently everywhere in computer
science, but I think that invalidate is to destroy
a pointer to some data, but not necessarily destroy the
data itself. And flush means to actually remove
the data. So one would invalidate a mapping but one
would flush a cache.
Since cleancache_flush_page and cleancache_flush_inode
semantically remove data from cleancache, I think flush
is a better name than invalidate.
Does that make sense?
nope ;)
Kernel code freely uses flush to refer to both invalidation and to
writeback, sometimes in confusing ways. In this case,
cleancache_flush_inode and cleancache_flush_page rather sound like
they
might write those things to backing store.
I'd like to mention about *_{get,put}_page too. In linux get/put is not
meaning read/write. There is {get,put}_page those are refcount stuff
(Yeah, and I felt those methods does refcount by quick read. But it
seems to be false. There is no xen codes, so I don't know actually
though.).
And I agree, I also think the needing thing is consistency on the linux
codes (term).
Thanks.
--
OGAWA Hirofumi hirof...@mail.parknet.co.jp
Hmmm, yes, that's a point of confusion also. No, cleancache put/get
do not have any relationship with reference counting.
Andrew, I wonder if you would be so kind as to read the following
and make a ruling. If you determine a preferable set of names,
I will abide by your decision and repost (if necessary).
The problem is this: The English language has a limited number
of words that can be used to represent data motion and mapping
and most/all of them are already used in the kernel, often,
to quote Andrew, in confusing ways. Complicating this, I
think the semantics of the cleancache operations are different
from the semantics of any other kernel operation... intentionally
so, because the value of cleancache is a direct result of those
differing semantics. And the cleancache semantics
are fairly complex (again intentionally) so a single function
name can't possibly describe the semantics.
The cleancache operations are:
- put (page)
- get (page)
- flush page
- flush inode
- init fs
- flush fs
I think these names are reasonable representations of the
semantics of the operations performed... but I'm not a kernel
expert so there is certainly room for disagreement. Though I
absolutely recognize the importance of a name, I am primarily
interested in merging the semantics of the operations and
would happily accept any name that kernel developers could
agree on. However, I fear that there will be NO name that
will satisfy all, so would prefer to keep the existing names.
If some renaming is eventually agreed upon, this could be done
post-merge.
Here's a brief description of the semantics:
The cleancache operation currently known as put has the
following semantics: If *possible*, please take the data
contained in the pageframe referred to by this struct page
into cleancache and associate it with the filesystem-determined
handle derived from the struct page.
The cleancache operation currently known as get has the
following semantics: Derive the filesystem-determined handle
from this struct page. If cleancache contains a page matching
that handle, recreate the page of data from cleancache and
place the results in the pageframe referred to by the
struct page. Then delete in cleancache any record of the
handle and any data associated with it, so that a
subsequent get will no longer find a match for the handle;
any space used for the data can also be freed.
(Note that take the data and recreate the page of data are
similar in semantics to copy to and copy from, but since
the cleancache operation may perform an inflight transformation
on the data, and copy usually means a byte-for-byte replication,
the word copy is also misleading.)
The cleancache operation currently known as flush has the
following semantics: Derive the filesystem-determined handle
from this struct page and struct mapping. If cleancache
contains a page matching that handle, delete in cleancache any
record of the handle and any data associated with it, so that a
subsequent get will no longer find a match for the handle;
any space used for the data can also be freed
The cleancache operation currently known as flush inode has
the following semantics: Derive the filesystem-determined filekey
from this struct mapping. If cleancache contains ANY handles
matching that filekey, delete in cleancache any record of
any matching handle and any data associated with those handles;
any space used for the data
Re: [PATCH V8 4/8] mm/fs: add hooks to support cleancache
Hi Dan,
On Fri, Apr 15, 2011 at 6:17 AM, Dan Magenheimer
dan.magenhei...@oracle.com wrote:
[PATCH V8 4/8] mm/fs: add hooks to support cleancache
This fourth patch of eight in this cleancache series provides the
core hooks in VFS for: initializing cleancache per filesystem;
capturing clean pages reclaimed by page cache; attempting to get
pages from cleancache before filesystem read; and ensuring coherency
between pagecache, disk, and cleancache. Note that the placement
of these hooks was stable from 2.6.18 to 2.6.38; a minor semantic
change was required due to a patchset in 2.6.39.
All hooks become no-ops if CONFIG_CLEANCACHE is unset, or become
a check of a boolean global if CONFIG_CLEANCACHE is set but no
cleancache backend has claimed cleancache_ops.
Details and a FAQ can be found in Documentation/vm/cleancache.txt
[v8: minchan@gmail.com: adapt to new remove_from_page_cache function]
Signed-off-by: Chris Mason chris.ma...@oracle.com
Signed-off-by: Dan Magenheimer dan.magenhei...@oracle.com
Reviewed-by: Jeremy Fitzhardinge jer...@goop.org
Reviewed-by: Konrad Rzeszutek Wilk konrad.w...@oracle.com
Cc: Andrew Morton a...@linux-foundation.org
Cc: Al Viro v...@zeniv.linux.org.uk
Cc: Matthew Wilcox matt...@wil.cx
Cc: Nick Piggin npig...@kernel.dk
Cc: Mel Gorman m...@csn.ul.ie
Cc: Rik Van Riel r...@redhat.com
Cc: Jan Beulich jbeul...@novell.com
Cc: Andreas Dilger adil...@sun.com
Cc: Ted Ts'o ty...@mit.edu
Cc: Mark Fasheh mfas...@suse.com
Cc: Joel Becker joel.bec...@oracle.com
Cc: Nitin Gupta ngu...@vflare.org
---
Diffstat:
fs/buffer.c | 5 +
fs/mpage.c | 7 +++
fs/super.c | 3 +++
mm/filemap.c | 11 +++
mm/truncate.c | 6 ++
5 files changed, 32 insertions(+)
--- linux-2.6.39-rc3/fs/super.c 2011-04-11 18:21:51.0 -0600
+++ linux-2.6.39-rc3-cleancache/fs/super.c 2011-04-13 17:08:09.175853426
-0600
@@ -31,6 +31,7 @@
#include linux/mutex.h
#include linux/backing-dev.h
#include linux/rculist_bl.h
+#include linux/cleancache.h
#include internal.h
@@ -112,6 +113,7 @@ static struct super_block *alloc_super(s
s-s_maxbytes = MAX_NON_LFS;
s-s_op = default_op;
s-s_time_gran = 10;
+ s-cleancache_poolid = -1;
}
out:
return s;
@@ -177,6 +179,7 @@ void deactivate_locked_super(struct supe
{
struct file_system_type *fs = s-s_type;
if (atomic_dec_and_test(s-s_active)) {
+ cleancache_flush_fs(s);
fs-kill_sb(s);
/*
* We need to call rcu_barrier so all the delayed rcu free
--- linux-2.6.39-rc3/fs/buffer.c 2011-04-11 18:21:51.0 -0600
+++ linux-2.6.39-rc3-cleancache/fs/buffer.c 2011-04-13 17:07:24.700917174
-0600
@@ -41,6 +41,7 @@
#include linux/bitops.h
#include linux/mpage.h
#include linux/bit_spinlock.h
+#include linux/cleancache.h
static int fsync_buffers_list(spinlock_t *lock, struct list_head *list);
@@ -269,6 +270,10 @@ void invalidate_bdev(struct block_device
invalidate_bh_lrus();
lru_add_drain_all(); /* make sure all lru add caches are flushed */
invalidate_mapping_pages(mapping, 0, -1);
+ /* 99% of the time, we don't need to flush the cleancache on the bdev.
+ * But, for the strange corners, lets be cautious
+ */
+ cleancache_flush_inode(mapping);
}
EXPORT_SYMBOL(invalidate_bdev);
--- linux-2.6.39-rc3/fs/mpage.c 2011-04-11 18:21:51.0 -0600
+++ linux-2.6.39-rc3-cleancache/fs/mpage.c 2011-04-13 17:07:24.706913410
-0600
@@ -27,6 +27,7 @@
#include linux/writeback.h
#include linux/backing-dev.h
#include linux/pagevec.h
+#include linux/cleancache.h
/*
* I/O completion handler for multipage BIOs.
@@ -271,6 +272,12 @@ do_mpage_readpage(struct bio *bio, struc
SetPageMappedToDisk(page);
}
+ if (fully_mapped blocks_per_page == 1 !PageUptodate(page)
+ cleancache_get_page(page) == 0) {
+ SetPageUptodate(page);
+ goto confused;
+ }
+
/*
* This page will go to BIO. Do we need to send this BIO off first?
*/
--- linux-2.6.39-rc3/mm/filemap.c 2011-04-11 18:21:51.0 -0600
+++ linux-2.6.39-rc3-cleancache/mm/filemap.c 2011-04-13 17:09:46.367852002
-0600
@@ -34,6 +34,7 @@
#include linux/hardirq.h /* for BUG_ON(!in_atomic()) only */
#include linux/memcontrol.h
#include linux/mm_inline.h /* for page_is_file_cache() */
+#include linux/cleancache.h
#include internal.h
/*
@@ -118,6 +119,16 @@ void __delete_from_page_cache(struct pag
{
struct address_space *mapping = page-mapping;
+ /*
+ * if we're uptodate, flush out into the
