From: Mingming <cmm@us.ibm.com>
To: Curt Wohlgemuth <curtw@google.com>
Cc: tytso@mit.edu, ext4 development <linux-ext4@vger.kernel.org>
Subject: Re: [PATCH 2/4 v2] avoid perform unwritten convertion for direct IO over non fallocated space
Date: Wed, 21 Oct 2009 15:48:01 -0700 [thread overview]
Message-ID: <1256165281.6404.4.camel@mingming-laptop> (raw)
In-Reply-To: <6601abe90910200803n1c499ba0w825c59c2f747dce8@mail.gmail.com>
[-- Attachment #1: Type: text/plain, Size: 5590 bytes --]
On Tue, 2009-10-20 at 08:03 -0700, Curt Wohlgemuth wrote:
> Hi Mingming:
>
> On Thu, Oct 8, 2009 at 6:13 PM, Mingming <cmm@us.ibm.com> wrote:
> > ext4: avoid perform unwritten convertion for direct IO over non fallocated space
> >
> > When direct IO complete, without this patch, it always check and convert
> > the range covered by the completed IO to be written. This is unnecessary
> > if the range doesn't has any preallocated area or holes. This patch add
> > a state flag to alart direct IO code to only run the convertion when
> > there is unwritten extents being covered in this IO.
> >
> > Signed-off-by: Mingming Cao <cmm@us.ibm.com>
> > ---
> > fs/ext4/ext4.h | 1 +
> > fs/ext4/extents.c | 22 +++++++++++++++++-----
> > fs/ext4/inode.c | 4 +++-
> > 3 files changed, 21 insertions(+), 6 deletions(-)
> >
> > Index: linux-2.6.31-rc4/fs/ext4/ext4.h
> > ===================================================================
> > --- linux-2.6.31-rc4.orig/fs/ext4/ext4.h
> > +++ linux-2.6.31-rc4/fs/ext4/ext4.h
> > @@ -299,6 +299,7 @@ static inline __u32 ext4_mask_flags(umod
> > #define EXT4_STATE_XATTR 0x00000004 /* has in-inode xattrs */
> > #define EXT4_STATE_NO_EXPAND 0x00000008 /* No space for expansion */
> > #define EXT4_STATE_DA_ALLOC_CLOSE 0x00000010 /* Alloc DA blks on close */
> > +#define EXT4_STATE_DIO_UNWRITTEN 0x00000020 /* need convert on dio done*/
>
> You'll need to change this value, since EXT4_STATE_EXT_MIGRATE has
> already been assigned 0x20.
>
> Thanks,
> Curt
>
thanks, The patch was originally made on a older kernel, the following
patch is rebased on the new kernel
> >
> > /* Used to pass group descriptor data when online resize is done */
> > struct ext4_new_group_input {
> > Index: linux-2.6.31-rc4/fs/ext4/extents.c
> > ===================================================================
> > --- linux-2.6.31-rc4.orig/fs/ext4/extents.c
> > +++ linux-2.6.31-rc4/fs/ext4/extents.c
> > @@ -3029,12 +3029,18 @@ ext4_ext_handle_uninitialized_extents(ha
> > ret = ext4_split_unwritten_extents(handle,
> > inode, path, iblock,
> > max_blocks, flags);
> > - /* flag the io_end struct that we need convert when IO done */
> > + /*
> > + * Flag the inode(non aio case) or end_io struct (aio case)
> > + * that this IO needs to convertion to written when IO is
> > + * completed
> > + */
> > if (io)
> > io->flag = DIO_AIO_UNWRITTEN;
> > + else
> > + EXT4_I(inode)->i_state |= EXT4_STATE_DIO_UNWRITTEN;
> > goto out;
> > }
> > - /* DIO end_io complete, convert the filled extent to written */
> > + /* async DIO end_io complete, convert the filled extent to written */
> > if (flags == EXT4_GET_BLOCKS_DIO_CONVERT_EXT) {
> > ret = ext4_convert_unwritten_extents_dio(handle, inode,
> > path);
> > @@ -3276,10 +3282,16 @@ int ext4_ext_get_blocks(handle_t *handle
> > * To avoid unecessary convertion for every aio dio rewrite
> > * to the mid of file, here we flag the IO that is really
> > * need the convertion.
> > - *
> > + * For non asycn direct IO case, flag the inode state
> > + * that we need to perform convertion when IO is done.
> > */
> > - if (io && flags == EXT4_GET_BLOCKS_DIO_CREATE_EXT)
> > - io->flag = DIO_AIO_UNWRITTEN;
> > + if (flags == EXT4_GET_BLOCKS_DIO_CREATE_EXT) {
> > + if (io)
> > + io->flag = DIO_AIO_UNWRITTEN;
> > + else
> > + EXT4_I(inode)->i_state |=
> > + EXT4_STATE_DIO_UNWRITTEN;;
> > + }
> > }
> > err = ext4_ext_insert_extent(handle, inode, path, &newex, flags);
> > if (err) {
> > Index: linux-2.6.31-rc4/fs/ext4/inode.c
> > ===================================================================
> > --- linux-2.6.31-rc4.orig/fs/ext4/inode.c
> > +++ linux-2.6.31-rc4/fs/ext4/inode.c
> > @@ -3693,7 +3693,8 @@ static ssize_t ext4_ext_direct_IO(int rw
> > if (ret != -EIOCBQUEUED && ret <= 0 && iocb->private) {
> > ext4_free_io_end(iocb->private);
> > iocb->private = NULL;
> > - } else if (ret > 0){
> > + } else if (ret > 0 && EXT4_I(inode)->i_state &
> > + EXT4_STATE_DIO_UNWRITTEN){
> > int err;
> > /*
> > * for non AIO case, since the IO is already
> > @@ -3703,6 +3704,7 @@ static ssize_t ext4_ext_direct_IO(int rw
> > offset, ret);
> > if (err < 0)
> > ret = err;
> > + EXT4_I(inode)->i_state &= ~EXT4_STATE_DIO_UNWRITTEN;
> > }
> > return ret;
> > }
> >
> >
> >
> --
> To unsubscribe from this list: send the line "unsubscribe linux-ext4" in
> the body of a message to majordomo@vger.kernel.org
> More majordomo info at http://vger.kernel.org/majordomo-info.html
a
[-- Attachment #2: dio-avoid-convertion-for-non-fallocate.patch --]
[-- Type: text/x-patch, Size: 3880 bytes --]
ext4: avoid perform convertion for direct IO over non fallocated space
When direct IO complete, without this patch, it always check and convert
the range covered by the completed IO to be written. This is unnecessary
if the range doesn't has any preallocated area or holes. This patch add
a state flag to alart direct IO code to only run the convertion when
there is unwritten extents being covered in this IO.
Signed-off-by: Mingming Cao <cmm@us.ibm.com>
---
fs/ext4/ext4.h | 1 +
fs/ext4/extents.c | 22 +++++++++++++++++-----
fs/ext4/inode.c | 4 +++-
3 files changed, 21 insertions(+), 6 deletions(-)
Index: linux-2.6.32-rc4/fs/ext4/ext4.h
===================================================================
--- linux-2.6.32-rc4.orig/fs/ext4/ext4.h 2009-10-11 14:43:56.000000000 -0700
+++ linux-2.6.32-rc4/fs/ext4/ext4.h 2009-10-12 15:05:46.000000000 -0700
@@ -322,6 +322,7 @@ static inline __u32 ext4_mask_flags(umod
#define EXT4_STATE_NO_EXPAND 0x00000008 /* No space for expansion */
#define EXT4_STATE_DA_ALLOC_CLOSE 0x00000010 /* Alloc DA blks on close */
#define EXT4_STATE_EXT_MIGRATE 0x00000020 /* Inode is migrating */
+#define EXT4_STATE_DIO_UNWRITTEN 0x00000040 /* need convert on dio done*/
/* Used to pass group descriptor data when online resize is done */
struct ext4_new_group_input {
Index: linux-2.6.32-rc4/fs/ext4/extents.c
===================================================================
--- linux-2.6.32-rc4.orig/fs/ext4/extents.c 2009-10-12 15:03:40.000000000 -0700
+++ linux-2.6.32-rc4/fs/ext4/extents.c 2009-10-12 15:04:07.000000000 -0700
@@ -3048,12 +3048,18 @@ ext4_ext_handle_uninitialized_extents(ha
ret = ext4_split_unwritten_extents(handle,
inode, path, iblock,
max_blocks, flags);
- /* flag the io_end struct that we need convert when IO done */
+ /*
+ * Flag the inode(non aio case) or end_io struct (aio case)
+ * that this IO needs to convertion to written when IO is
+ * completed
+ */
if (io)
io->flag = DIO_AIO_UNWRITTEN;
+ else
+ EXT4_I(inode)->i_state |= EXT4_STATE_DIO_UNWRITTEN;
goto out;
}
- /* DIO end_io complete, convert the filled extent to written */
+ /* async DIO end_io complete, convert the filled extent to written */
if (flags == EXT4_GET_BLOCKS_DIO_CONVERT_EXT) {
ret = ext4_convert_unwritten_extents_dio(handle, inode,
path);
@@ -3295,10 +3301,16 @@ int ext4_ext_get_blocks(handle_t *handle
* To avoid unecessary convertion for every aio dio rewrite
* to the mid of file, here we flag the IO that is really
* need the convertion.
- *
+ * For non asycn direct IO case, flag the inode state
+ * that we need to perform convertion when IO is done.
*/
- if (io && flags == EXT4_GET_BLOCKS_DIO_CREATE_EXT)
- io->flag = DIO_AIO_UNWRITTEN;
+ if (flags == EXT4_GET_BLOCKS_DIO_CREATE_EXT) {
+ if (io)
+ io->flag = DIO_AIO_UNWRITTEN;
+ else
+ EXT4_I(inode)->i_state |=
+ EXT4_STATE_DIO_UNWRITTEN;;
+ }
}
err = ext4_ext_insert_extent(handle, inode, path, &newex, flags);
if (err) {
Index: linux-2.6.32-rc4/fs/ext4/inode.c
===================================================================
--- linux-2.6.32-rc4.orig/fs/ext4/inode.c 2009-10-12 15:03:40.000000000 -0700
+++ linux-2.6.32-rc4/fs/ext4/inode.c 2009-10-12 15:04:07.000000000 -0700
@@ -3771,7 +3771,8 @@ static ssize_t ext4_ext_direct_IO(int rw
if (ret != -EIOCBQUEUED && ret <= 0 && iocb->private) {
ext4_free_io_end(iocb->private);
iocb->private = NULL;
- } else if (ret > 0){
+ } else if (ret > 0 && EXT4_I(inode)->i_state &
+ EXT4_STATE_DIO_UNWRITTEN){
int err;
/*
* for non AIO case, since the IO is already
@@ -3781,6 +3782,7 @@ static ssize_t ext4_ext_direct_IO(int rw
offset, ret);
if (err < 0)
ret = err;
+ EXT4_I(inode)->i_state &= ~EXT4_STATE_DIO_UNWRITTEN;
}
return ret;
}
next prev parent reply other threads:[~2009-10-21 22:48 UTC|newest]
Thread overview: 4+ messages / expand[flat|nested] mbox.gz Atom feed top
2009-10-09 1:13 [PATCH 2/4 v2] avoid perform unwritten convertion for direct IO over non fallocated space Mingming
2009-10-20 15:03 ` Curt Wohlgemuth
2009-10-21 22:48 ` Mingming [this message]
2009-11-04 2:10 ` Theodore Tso
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=1256165281.6404.4.camel@mingming-laptop \
--to=cmm@us.ibm.com \
--cc=curtw@google.com \
--cc=linux-ext4@vger.kernel.org \
--cc=tytso@mit.edu \
/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 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).