Hi,
This final part of the O_SYNC patches adds calls to ext2, and to
generic_commit_write, to record dirty buffers against the owning
inode. It also removes most of fs/ext2/fsync.c, which now simply
calls the generic sync code.
--Stephen
2.4.0test11.02.ext2-osync.diff :
--- linux-2.4.0-test11/fs/buffer.c.~1~ Tue Nov 21 15:51:17 2000
+++ linux-2.4.0-test11/fs/buffer.c Tue Nov 21 16:35:35 2000
@@ -1727,6 +1727,7 @@
set_bit(BH_Uptodate, &bh->b_state);
if (!atomic_set_buffer_dirty(bh)) {
__mark_dirty(bh);
+ buffer_insert_inode_queue(bh, inode);
need_balance_dirty = 1;
}
}
--- linux-2.4.0-test11/fs/ext2/fsync.c.~1~ Tue Nov 21 15:47:48 2000
+++ linux-2.4.0-test11/fs/ext2/fsync.c Tue Nov 21 16:01:15 2000
@@ -28,98 +28,6 @@
#include <linux/smp_lock.h>
-#define blocksize (EXT2_BLOCK_SIZE(inode->i_sb))
-#define addr_per_block (EXT2_ADDR_PER_BLOCK(inode->i_sb))
-
-static int sync_indirect(struct inode * inode, u32 * block, int wait)
-{
- struct buffer_head * bh;
-
- if (!*block)
- return 0;
- bh = get_hash_table(inode->i_dev, le32_to_cpu(*block), blocksize);
- if (!bh)
- return 0;
- if (wait && buffer_req(bh) && !buffer_uptodate(bh)) {
- /* There can be a parallell read(2) that started read-I/O
- on the buffer so we can't assume that there's been
- an I/O error without first waiting I/O completation. */
- wait_on_buffer(bh);
- if (!buffer_uptodate(bh))
- {
- brelse (bh);
- return -1;
- }
- }
- if (wait || !buffer_uptodate(bh) || !buffer_dirty(bh)) {
- if (wait)
- /* when we return from fsync all the blocks
- must be _just_ stored on disk */
- wait_on_buffer(bh);
- brelse(bh);
- return 0;
- }
- ll_rw_block(WRITE, 1, &bh);
- atomic_dec(&bh->b_count);
- return 0;
-}
-
-static int sync_iblock(struct inode * inode, u32 * iblock,
- struct buffer_head ** bh, int wait)
-{
- int rc, tmp;
-
- *bh = NULL;
- tmp = le32_to_cpu(*iblock);
- if (!tmp)
- return 0;
- rc = sync_indirect(inode, iblock, wait);
- if (rc)
- return rc;
- *bh = bread(inode->i_dev, tmp, blocksize);
- if (!*bh)
- return -1;
- return 0;
-}
-
-static int sync_dindirect(struct inode * inode, u32 * diblock, int wait)
-{
- int i;
- struct buffer_head * dind_bh;
- int rc, err = 0;
-
- rc = sync_iblock(inode, diblock, &dind_bh, wait);
- if (rc || !dind_bh)
- return rc;
-
- for (i = 0; i < addr_per_block; i++) {
- rc = sync_indirect(inode, ((u32 *) dind_bh->b_data) + i, wait);
- if (rc)
- err = rc;
- }
- brelse(dind_bh);
- return err;
-}
-
-static int sync_tindirect(struct inode * inode, u32 * tiblock, int wait)
-{
- int i;
- struct buffer_head * tind_bh;
- int rc, err = 0;
-
- rc = sync_iblock(inode, tiblock, &tind_bh, wait);
- if (rc || !tind_bh)
- return rc;
-
- for (i = 0; i < addr_per_block; i++) {
- rc = sync_dindirect(inode, ((u32 *) tind_bh->b_data) + i, wait);
- if (rc)
- err = rc;
- }
- brelse(tind_bh);
- return err;
-}
-
/*
* File may be NULL when we are called. Perhaps we shouldn't
* even pass file to fsync ?
@@ -127,34 +35,20 @@
int ext2_sync_file(struct file * file, struct dentry *dentry, int datasync)
{
- int wait, err = 0;
struct inode *inode = dentry->d_inode;
+ return ext2_fsync_inode(inode, datasync);
+}
- lock_kernel();
- if (S_ISLNK(inode->i_mode) && !(inode->i_blocks))
- /*
- * Don't sync fast links!
- */
- goto skip;
-
- err = generic_buffer_fdatasync(inode, 0, ~0UL);
-
- for (wait=0; wait<=1; wait++)
- {
- err |= sync_indirect(inode,
- inode->u.ext2_i.i_data+EXT2_IND_BLOCK,
- wait);
- err |= sync_dindirect(inode,
- inode->u.ext2_i.i_data+EXT2_DIND_BLOCK,
- wait);
- err |= sync_tindirect(inode,
- inode->u.ext2_i.i_data+EXT2_TIND_BLOCK,
- wait);
- }
-skip:
- if ((inode->i_state & I_DIRTY_DATASYNC) ||
- ((inode->i_state & I_DIRTY) && !datasync))
- err |= ext2_sync_inode (inode);
- unlock_kernel();
+int ext2_fsync_inode(struct inode *inode, int datasync)
+{
+ int err;
+
+ err = fsync_inode_buffers(inode);
+ if (!(inode->i_state & I_DIRTY))
+ return err;
+ if (datasync && !(inode->i_state & I_DIRTY_DATASYNC))
+ return err;
+
+ err |= ext2_sync_inode(inode);
return err ? -EIO : 0;
}
--- linux-2.4.0-test11/fs/ext2/inode.c.~1~ Tue Nov 21 15:40:22 2000
+++ linux-2.4.0-test11/fs/ext2/inode.c Tue Nov 21 16:01:15 2000
@@ -404,7 +404,7 @@
branch[n].p = (u32*) bh->b_data + offsets[n];
*branch[n].p = branch[n].key;
mark_buffer_uptodate(bh, 1);
- mark_buffer_dirty(bh);
+ mark_buffer_dirty_inode(bh, inode);
if (IS_SYNC(inode) || inode->u.ext2_i.i_osync) {
ll_rw_block (WRITE, 1, &bh);
wait_on_buffer (bh);
@@ -469,7 +469,7 @@
/* had we spliced it onto indirect block? */
if (where->bh) {
- mark_buffer_dirty(where->bh);
+ mark_buffer_dirty_inode(where->bh, inode);
if (IS_SYNC(inode) || inode->u.ext2_i.i_osync) {
ll_rw_block (WRITE, 1, &where->bh);
wait_on_buffer(where->bh);
@@ -591,7 +591,7 @@
wait_on_buffer(bh);
memset(bh->b_data, 0, inode->i_sb->s_blocksize);
mark_buffer_uptodate(bh, 1);
- mark_buffer_dirty(bh);
+ mark_buffer_dirty_inode(bh, inode);
}
return bh;
}
@@ -907,7 +907,7 @@
if (partial == chain)
mark_inode_dirty(inode);
else
- mark_buffer_dirty(partial->bh);
+ mark_buffer_dirty_inode(partial->bh, inode);
ext2_free_branches(inode, &nr, &nr+1, (chain+n-1) - partial);
}
/* Clear the ends of indirect blocks on the shared branch */
@@ -916,7 +916,7 @@
partial->p + 1,
(u32*)partial->bh->b_data + addr_per_block,
(chain+n-1) - partial);
- mark_buffer_dirty(partial->bh);
+ mark_buffer_dirty_inode(partial->bh, inode);
if (IS_SYNC(inode)) {
ll_rw_block (WRITE, 1, &partial->bh);
wait_on_buffer (partial->bh);
@@ -1208,7 +1208,7 @@
raw_inode->i_block[0] = cpu_to_le32(kdev_t_to_nr(inode->i_rdev));
else for (block = 0; block < EXT2_N_BLOCKS; block++)
raw_inode->i_block[block] = inode->u.ext2_i.i_data[block];
- mark_buffer_dirty(bh);
+ mark_buffer_dirty_inode(bh, inode);
if (do_sync) {
ll_rw_block (WRITE, 1, &bh);
wait_on_buffer (bh);
--- linux-2.4.0-test11/fs/ext2/namei.c.~1~ Tue Nov 21 15:40:22 2000
+++ linux-2.4.0-test11/fs/ext2/namei.c Tue Nov 21 16:01:15 2000
@@ -296,7 +296,7 @@
dir->u.ext2_i.i_flags &= ~EXT2_BTREE_FL;
mark_inode_dirty(dir);
dir->i_version = ++event;
- mark_buffer_dirty(bh);
+ mark_buffer_dirty_inode(bh, dir);
if (IS_SYNC(dir)) {
ll_rw_block (WRITE, 1, &bh);
wait_on_buffer (bh);
@@ -337,7 +337,7 @@
else
de->inode = 0;
dir->i_version = ++event;
- mark_buffer_dirty(bh);
+ mark_buffer_dirty_inode(bh, dir);
if (IS_SYNC(dir)) {
ll_rw_block (WRITE, 1, &bh);
wait_on_buffer (bh);
@@ -449,7 +449,7 @@
strcpy (de->name, "..");
ext2_set_de_type(dir->i_sb, de, S_IFDIR);
inode->i_nlink = 2;
- mark_buffer_dirty(dir_block);
+ mark_buffer_dirty_inode(dir_block, dir);
brelse (dir_block);
inode->i_mode = S_IFDIR | mode;
if (dir->i_mode & S_ISGID)
@@ -755,7 +755,7 @@
EXT2_FEATURE_INCOMPAT_FILETYPE))
new_de->file_type = old_de->file_type;
new_dir->i_version = ++event;
- mark_buffer_dirty(new_bh);
+ mark_buffer_dirty_inode(new_bh, new_dir);
if (IS_SYNC(new_dir)) {
ll_rw_block (WRITE, 1, &new_bh);
wait_on_buffer (new_bh);
@@ -786,7 +786,7 @@
mark_inode_dirty(old_dir);
if (dir_bh) {
PARENT_INO(dir_bh->b_data) = le32_to_cpu(new_dir->i_ino);
- mark_buffer_dirty(dir_bh);
+ mark_buffer_dirty_inode(dir_bh, old_inode);
old_dir->i_nlink--;
mark_inode_dirty(old_dir);
if (new_inode) {
--- linux-2.4.0-test11/include/linux/ext2_fs.h.~1~ Tue Nov 21 15:40:22 2000
+++ linux-2.4.0-test11/include/linux/ext2_fs.h Tue Nov 21 16:01:15 2000
@@ -549,6 +549,7 @@
/* fsync.c */
extern int ext2_sync_file (struct file *, struct dentry *, int);
+extern int ext2_fsync_inode (struct inode *, int);
/* ialloc.c */
extern struct inode * ext2_new_inode (const struct inode *, int, int *);
--- linux-2.4.0-test11/mm/filemap.c.~1~ Tue Nov 21 15:47:48 2000
+++ linux-2.4.0-test11/mm/filemap.c Tue Nov 21 16:01:15 2000
@@ -2521,8 +2521,14 @@
if (cached_page)
page_cache_free(cached_page);
+ /* For now, when the user asks for O_SYNC, we'll actually
+ * provide O_DSYNC. */
+ if ((status >= 0) && (file->f_flags & O_SYNC))
+ status = generic_osync_inode(inode, 1); /* 1 means datasync */
+
err = written ? written : status;
out:
+
up(&inode->i_sem);
return err;
fail_write:
On Wed, Nov 22, 2000 at 11:26:46AM +0000, Stephen C. Tweedie wrote:
> Hi,
>
> This final part of the O_SYNC patches adds calls to ext2, and to
> generic_commit_write, to record dirty buffers against the owning
> inode. It also removes most of fs/ext2/fsync.c, which now simply
> calls the generic sync code.
>
> --Stephen
Stephen,
I have not implemented O_SYNC in NWFS, but it looks like I need to add it
before posting the final patches. This patch appears to force write-through
of only dirty inodes, and allow reads to continue from cache. Is this
assumption correct, and does this comply with TPC rules for ORACLE and
other DBMS benchmarks? It looks like a good patch that will perform
very fast for O_SYNC.
Jeff
>
> 2.4.0test11.02.ext2-osync.diff :
>
>
> --- linux-2.4.0-test11/fs/buffer.c.~1~ Tue Nov 21 15:51:17 2000
> +++ linux-2.4.0-test11/fs/buffer.c Tue Nov 21 16:35:35 2000
> @@ -1727,6 +1727,7 @@
> set_bit(BH_Uptodate, &bh->b_state);
> if (!atomic_set_buffer_dirty(bh)) {
> __mark_dirty(bh);
> + buffer_insert_inode_queue(bh, inode);
> need_balance_dirty = 1;
> }
> }
> --- linux-2.4.0-test11/fs/ext2/fsync.c.~1~ Tue Nov 21 15:47:48 2000
> +++ linux-2.4.0-test11/fs/ext2/fsync.c Tue Nov 21 16:01:15 2000
> @@ -28,98 +28,6 @@
> #include <linux/smp_lock.h>
>
>
> -#define blocksize (EXT2_BLOCK_SIZE(inode->i_sb))
> -#define addr_per_block (EXT2_ADDR_PER_BLOCK(inode->i_sb))
> -
> -static int sync_indirect(struct inode * inode, u32 * block, int wait)
> -{
> - struct buffer_head * bh;
> -
> - if (!*block)
> - return 0;
> - bh = get_hash_table(inode->i_dev, le32_to_cpu(*block), blocksize);
> - if (!bh)
> - return 0;
> - if (wait && buffer_req(bh) && !buffer_uptodate(bh)) {
> - /* There can be a parallell read(2) that started read-I/O
> - on the buffer so we can't assume that there's been
> - an I/O error without first waiting I/O completation. */
> - wait_on_buffer(bh);
> - if (!buffer_uptodate(bh))
> - {
> - brelse (bh);
> - return -1;
> - }
> - }
> - if (wait || !buffer_uptodate(bh) || !buffer_dirty(bh)) {
> - if (wait)
> - /* when we return from fsync all the blocks
> - must be _just_ stored on disk */
> - wait_on_buffer(bh);
> - brelse(bh);
> - return 0;
> - }
> - ll_rw_block(WRITE, 1, &bh);
> - atomic_dec(&bh->b_count);
> - return 0;
> -}
> -
> -static int sync_iblock(struct inode * inode, u32 * iblock,
> - struct buffer_head ** bh, int wait)
> -{
> - int rc, tmp;
> -
> - *bh = NULL;
> - tmp = le32_to_cpu(*iblock);
> - if (!tmp)
> - return 0;
> - rc = sync_indirect(inode, iblock, wait);
> - if (rc)
> - return rc;
> - *bh = bread(inode->i_dev, tmp, blocksize);
> - if (!*bh)
> - return -1;
> - return 0;
> -}
> -
> -static int sync_dindirect(struct inode * inode, u32 * diblock, int wait)
> -{
> - int i;
> - struct buffer_head * dind_bh;
> - int rc, err = 0;
> -
> - rc = sync_iblock(inode, diblock, &dind_bh, wait);
> - if (rc || !dind_bh)
> - return rc;
> -
> - for (i = 0; i < addr_per_block; i++) {
> - rc = sync_indirect(inode, ((u32 *) dind_bh->b_data) + i, wait);
> - if (rc)
> - err = rc;
> - }
> - brelse(dind_bh);
> - return err;
> -}
> -
> -static int sync_tindirect(struct inode * inode, u32 * tiblock, int wait)
> -{
> - int i;
> - struct buffer_head * tind_bh;
> - int rc, err = 0;
> -
> - rc = sync_iblock(inode, tiblock, &tind_bh, wait);
> - if (rc || !tind_bh)
> - return rc;
> -
> - for (i = 0; i < addr_per_block; i++) {
> - rc = sync_dindirect(inode, ((u32 *) tind_bh->b_data) + i, wait);
> - if (rc)
> - err = rc;
> - }
> - brelse(tind_bh);
> - return err;
> -}
> -
> /*
> * File may be NULL when we are called. Perhaps we shouldn't
> * even pass file to fsync ?
> @@ -127,34 +35,20 @@
>
> int ext2_sync_file(struct file * file, struct dentry *dentry, int datasync)
> {
> - int wait, err = 0;
> struct inode *inode = dentry->d_inode;
> + return ext2_fsync_inode(inode, datasync);
> +}
>
> - lock_kernel();
> - if (S_ISLNK(inode->i_mode) && !(inode->i_blocks))
> - /*
> - * Don't sync fast links!
> - */
> - goto skip;
> -
> - err = generic_buffer_fdatasync(inode, 0, ~0UL);
> -
> - for (wait=0; wait<=1; wait++)
> - {
> - err |= sync_indirect(inode,
> - inode->u.ext2_i.i_data+EXT2_IND_BLOCK,
> - wait);
> - err |= sync_dindirect(inode,
> - inode->u.ext2_i.i_data+EXT2_DIND_BLOCK,
> - wait);
> - err |= sync_tindirect(inode,
> - inode->u.ext2_i.i_data+EXT2_TIND_BLOCK,
> - wait);
> - }
> -skip:
> - if ((inode->i_state & I_DIRTY_DATASYNC) ||
> - ((inode->i_state & I_DIRTY) && !datasync))
> - err |= ext2_sync_inode (inode);
> - unlock_kernel();
> +int ext2_fsync_inode(struct inode *inode, int datasync)
> +{
> + int err;
> +
> + err = fsync_inode_buffers(inode);
> + if (!(inode->i_state & I_DIRTY))
> + return err;
> + if (datasync && !(inode->i_state & I_DIRTY_DATASYNC))
> + return err;
> +
> + err |= ext2_sync_inode(inode);
> return err ? -EIO : 0;
> }
> --- linux-2.4.0-test11/fs/ext2/inode.c.~1~ Tue Nov 21 15:40:22 2000
> +++ linux-2.4.0-test11/fs/ext2/inode.c Tue Nov 21 16:01:15 2000
> @@ -404,7 +404,7 @@
> branch[n].p = (u32*) bh->b_data + offsets[n];
> *branch[n].p = branch[n].key;
> mark_buffer_uptodate(bh, 1);
> - mark_buffer_dirty(bh);
> + mark_buffer_dirty_inode(bh, inode);
> if (IS_SYNC(inode) || inode->u.ext2_i.i_osync) {
> ll_rw_block (WRITE, 1, &bh);
> wait_on_buffer (bh);
> @@ -469,7 +469,7 @@
>
> /* had we spliced it onto indirect block? */
> if (where->bh) {
> - mark_buffer_dirty(where->bh);
> + mark_buffer_dirty_inode(where->bh, inode);
> if (IS_SYNC(inode) || inode->u.ext2_i.i_osync) {
> ll_rw_block (WRITE, 1, &where->bh);
> wait_on_buffer(where->bh);
> @@ -591,7 +591,7 @@
> wait_on_buffer(bh);
> memset(bh->b_data, 0, inode->i_sb->s_blocksize);
> mark_buffer_uptodate(bh, 1);
> - mark_buffer_dirty(bh);
> + mark_buffer_dirty_inode(bh, inode);
> }
> return bh;
> }
> @@ -907,7 +907,7 @@
> if (partial == chain)
> mark_inode_dirty(inode);
> else
> - mark_buffer_dirty(partial->bh);
> + mark_buffer_dirty_inode(partial->bh, inode);
> ext2_free_branches(inode, &nr, &nr+1, (chain+n-1) - partial);
> }
> /* Clear the ends of indirect blocks on the shared branch */
> @@ -916,7 +916,7 @@
> partial->p + 1,
> (u32*)partial->bh->b_data + addr_per_block,
> (chain+n-1) - partial);
> - mark_buffer_dirty(partial->bh);
> + mark_buffer_dirty_inode(partial->bh, inode);
> if (IS_SYNC(inode)) {
> ll_rw_block (WRITE, 1, &partial->bh);
> wait_on_buffer (partial->bh);
> @@ -1208,7 +1208,7 @@
> raw_inode->i_block[0] = cpu_to_le32(kdev_t_to_nr(inode->i_rdev));
> else for (block = 0; block < EXT2_N_BLOCKS; block++)
> raw_inode->i_block[block] = inode->u.ext2_i.i_data[block];
> - mark_buffer_dirty(bh);
> + mark_buffer_dirty_inode(bh, inode);
> if (do_sync) {
> ll_rw_block (WRITE, 1, &bh);
> wait_on_buffer (bh);
> --- linux-2.4.0-test11/fs/ext2/namei.c.~1~ Tue Nov 21 15:40:22 2000
> +++ linux-2.4.0-test11/fs/ext2/namei.c Tue Nov 21 16:01:15 2000
> @@ -296,7 +296,7 @@
> dir->u.ext2_i.i_flags &= ~EXT2_BTREE_FL;
> mark_inode_dirty(dir);
> dir->i_version = ++event;
> - mark_buffer_dirty(bh);
> + mark_buffer_dirty_inode(bh, dir);
> if (IS_SYNC(dir)) {
> ll_rw_block (WRITE, 1, &bh);
> wait_on_buffer (bh);
> @@ -337,7 +337,7 @@
> else
> de->inode = 0;
> dir->i_version = ++event;
> - mark_buffer_dirty(bh);
> + mark_buffer_dirty_inode(bh, dir);
> if (IS_SYNC(dir)) {
> ll_rw_block (WRITE, 1, &bh);
> wait_on_buffer (bh);
> @@ -449,7 +449,7 @@
> strcpy (de->name, "..");
> ext2_set_de_type(dir->i_sb, de, S_IFDIR);
> inode->i_nlink = 2;
> - mark_buffer_dirty(dir_block);
> + mark_buffer_dirty_inode(dir_block, dir);
> brelse (dir_block);
> inode->i_mode = S_IFDIR | mode;
> if (dir->i_mode & S_ISGID)
> @@ -755,7 +755,7 @@
> EXT2_FEATURE_INCOMPAT_FILETYPE))
> new_de->file_type = old_de->file_type;
> new_dir->i_version = ++event;
> - mark_buffer_dirty(new_bh);
> + mark_buffer_dirty_inode(new_bh, new_dir);
> if (IS_SYNC(new_dir)) {
> ll_rw_block (WRITE, 1, &new_bh);
> wait_on_buffer (new_bh);
> @@ -786,7 +786,7 @@
> mark_inode_dirty(old_dir);
> if (dir_bh) {
> PARENT_INO(dir_bh->b_data) = le32_to_cpu(new_dir->i_ino);
> - mark_buffer_dirty(dir_bh);
> + mark_buffer_dirty_inode(dir_bh, old_inode);
> old_dir->i_nlink--;
> mark_inode_dirty(old_dir);
> if (new_inode) {
> --- linux-2.4.0-test11/include/linux/ext2_fs.h.~1~ Tue Nov 21 15:40:22 2000
> +++ linux-2.4.0-test11/include/linux/ext2_fs.h Tue Nov 21 16:01:15 2000
> @@ -549,6 +549,7 @@
>
> /* fsync.c */
> extern int ext2_sync_file (struct file *, struct dentry *, int);
> +extern int ext2_fsync_inode (struct inode *, int);
>
> /* ialloc.c */
> extern struct inode * ext2_new_inode (const struct inode *, int, int *);
> --- linux-2.4.0-test11/mm/filemap.c.~1~ Tue Nov 21 15:47:48 2000
> +++ linux-2.4.0-test11/mm/filemap.c Tue Nov 21 16:01:15 2000
> @@ -2521,8 +2521,14 @@
> if (cached_page)
> page_cache_free(cached_page);
>
> + /* For now, when the user asks for O_SYNC, we'll actually
> + * provide O_DSYNC. */
> + if ((status >= 0) && (file->f_flags & O_SYNC))
> + status = generic_osync_inode(inode, 1); /* 1 means datasync */
> +
> err = written ? written : status;
> out:
> +
> up(&inode->i_sem);
> return err;
> fail_write:
> -
> To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
> the body of a message to [email protected]
> Please read the FAQ at http://www.tux.org/lkml/
Jeff,
>>Is this assumption correct, and does this comply with TPC rules for ORACLE and
>>other DBMS benchmarks? It looks like a good patch that will perform
>>very fast for O_SYNC.
Why would that be a problem? If you look at our benchmarks we normally use raw I/O
to bypass buffers altogether (We buffer in the SGA).
That said, we maintain consistency at the database, so as long as the data in the
buffers is consistent with the data on disk, which is only relevant to modified
buffers, which are being written directly to disk, there should not be a problem.
The effect of O_SYNC on the kernel buffer is that there should not be "any" dirty
buffers for that particular file, since all of them are written to disk
"immediately".
Am I missing something?
Regards,
"Jeff V. Merkey" wrote:
> On Wed, Nov 22, 2000 at 11:26:46AM +0000, Stephen C. Tweedie wrote:
> > Hi,
> >
> > This final part of the O_SYNC patches adds calls to ext2, and to
> > generic_commit_write, to record dirty buffers against the owning
> > inode. It also removes most of fs/ext2/fsync.c, which now simply
> > calls the generic sync code.
> >
> > --Stephen
>
> Stephen,
>
> I have not implemented O_SYNC in NWFS, but it looks like I need to add it
> before posting the final patches. This patch appears to force write-through
> of only dirty inodes, and allow reads to continue from cache. Is this
> assumption correct, and does this comply with TPC rules for ORACLE and
> other DBMS benchmarks? It looks like a good patch that will perform
> very fast for O_SYNC.
>
> Jeff
>
> >
> > 2.4.0test11.02.ext2-osync.diff :
> >
> >
> > --- linux-2.4.0-test11/fs/buffer.c.~1~ Tue Nov 21 15:51:17 2000
> > +++ linux-2.4.0-test11/fs/buffer.c Tue Nov 21 16:35:35 2000
> > @@ -1727,6 +1727,7 @@
> > set_bit(BH_Uptodate, &bh->b_state);
> > if (!atomic_set_buffer_dirty(bh)) {
> > __mark_dirty(bh);
> > + buffer_insert_inode_queue(bh, inode);
> > need_balance_dirty = 1;
> > }
> > }
> > --- linux-2.4.0-test11/fs/ext2/fsync.c.~1~ Tue Nov 21 15:47:48 2000
> > +++ linux-2.4.0-test11/fs/ext2/fsync.c Tue Nov 21 16:01:15 2000
> > @@ -28,98 +28,6 @@
> > #include <linux/smp_lock.h>
> >
> >
> > -#define blocksize (EXT2_BLOCK_SIZE(inode->i_sb))
> > -#define addr_per_block (EXT2_ADDR_PER_BLOCK(inode->i_sb))
> > -
> > -static int sync_indirect(struct inode * inode, u32 * block, int wait)
> > -{
> > - struct buffer_head * bh;
> > -
> > - if (!*block)
> > - return 0;
> > - bh = get_hash_table(inode->i_dev, le32_to_cpu(*block), blocksize);
> > - if (!bh)
> > - return 0;
> > - if (wait && buffer_req(bh) && !buffer_uptodate(bh)) {
> > - /* There can be a parallell read(2) that started read-I/O
> > - on the buffer so we can't assume that there's been
> > - an I/O error without first waiting I/O completation. */
> > - wait_on_buffer(bh);
> > - if (!buffer_uptodate(bh))
> > - {
> > - brelse (bh);
> > - return -1;
> > - }
> > - }
> > - if (wait || !buffer_uptodate(bh) || !buffer_dirty(bh)) {
> > - if (wait)
> > - /* when we return from fsync all the blocks
> > - must be _just_ stored on disk */
> > - wait_on_buffer(bh);
> > - brelse(bh);
> > - return 0;
> > - }
> > - ll_rw_block(WRITE, 1, &bh);
> > - atomic_dec(&bh->b_count);
> > - return 0;
> > -}
> > -
> > -static int sync_iblock(struct inode * inode, u32 * iblock,
> > - struct buffer_head ** bh, int wait)
> > -{
> > - int rc, tmp;
> > -
> > - *bh = NULL;
> > - tmp = le32_to_cpu(*iblock);
> > - if (!tmp)
> > - return 0;
> > - rc = sync_indirect(inode, iblock, wait);
> > - if (rc)
> > - return rc;
> > - *bh = bread(inode->i_dev, tmp, blocksize);
> > - if (!*bh)
> > - return -1;
> > - return 0;
> > -}
> > -
> > -static int sync_dindirect(struct inode * inode, u32 * diblock, int wait)
> > -{
> > - int i;
> > - struct buffer_head * dind_bh;
> > - int rc, err = 0;
> > -
> > - rc = sync_iblock(inode, diblock, &dind_bh, wait);
> > - if (rc || !dind_bh)
> > - return rc;
> > -
> > - for (i = 0; i < addr_per_block; i++) {
> > - rc = sync_indirect(inode, ((u32 *) dind_bh->b_data) + i, wait);
> > - if (rc)
> > - err = rc;
> > - }
> > - brelse(dind_bh);
> > - return err;
> > -}
> > -
> > -static int sync_tindirect(struct inode * inode, u32 * tiblock, int wait)
> > -{
> > - int i;
> > - struct buffer_head * tind_bh;
> > - int rc, err = 0;
> > -
> > - rc = sync_iblock(inode, tiblock, &tind_bh, wait);
> > - if (rc || !tind_bh)
> > - return rc;
> > -
> > - for (i = 0; i < addr_per_block; i++) {
> > - rc = sync_dindirect(inode, ((u32 *) tind_bh->b_data) + i, wait);
> > - if (rc)
> > - err = rc;
> > - }
> > - brelse(tind_bh);
> > - return err;
> > -}
> > -
> > /*
> > * File may be NULL when we are called. Perhaps we shouldn't
> > * even pass file to fsync ?
> > @@ -127,34 +35,20 @@
> >
> > int ext2_sync_file(struct file * file, struct dentry *dentry, int datasync)
> > {
> > - int wait, err = 0;
> > struct inode *inode = dentry->d_inode;
> > + return ext2_fsync_inode(inode, datasync);
> > +}
> >
> > - lock_kernel();
> > - if (S_ISLNK(inode->i_mode) && !(inode->i_blocks))
> > - /*
> > - * Don't sync fast links!
> > - */
> > - goto skip;
> > -
> > - err = generic_buffer_fdatasync(inode, 0, ~0UL);
> > -
> > - for (wait=0; wait<=1; wait++)
> > - {
> > - err |= sync_indirect(inode,
> > - inode->u.ext2_i.i_data+EXT2_IND_BLOCK,
> > - wait);
> > - err |= sync_dindirect(inode,
> > - inode->u.ext2_i.i_data+EXT2_DIND_BLOCK,
> > - wait);
> > - err |= sync_tindirect(inode,
> > - inode->u.ext2_i.i_data+EXT2_TIND_BLOCK,
> > - wait);
> > - }
> > -skip:
> > - if ((inode->i_state & I_DIRTY_DATASYNC) ||
> > - ((inode->i_state & I_DIRTY) && !datasync))
> > - err |= ext2_sync_inode (inode);
> > - unlock_kernel();
> > +int ext2_fsync_inode(struct inode *inode, int datasync)
> > +{
> > + int err;
> > +
> > + err = fsync_inode_buffers(inode);
> > + if (!(inode->i_state & I_DIRTY))
> > + return err;
> > + if (datasync && !(inode->i_state & I_DIRTY_DATASYNC))
> > + return err;
> > +
> > + err |= ext2_sync_inode(inode);
> > return err ? -EIO : 0;
> > }
> > --- linux-2.4.0-test11/fs/ext2/inode.c.~1~ Tue Nov 21 15:40:22 2000
> > +++ linux-2.4.0-test11/fs/ext2/inode.c Tue Nov 21 16:01:15 2000
> > @@ -404,7 +404,7 @@
> > branch[n].p = (u32*) bh->b_data + offsets[n];
> > *branch[n].p = branch[n].key;
> > mark_buffer_uptodate(bh, 1);
> > - mark_buffer_dirty(bh);
> > + mark_buffer_dirty_inode(bh, inode);
> > if (IS_SYNC(inode) || inode->u.ext2_i.i_osync) {
> > ll_rw_block (WRITE, 1, &bh);
> > wait_on_buffer (bh);
> > @@ -469,7 +469,7 @@
> >
> > /* had we spliced it onto indirect block? */
> > if (where->bh) {
> > - mark_buffer_dirty(where->bh);
> > + mark_buffer_dirty_inode(where->bh, inode);
> > if (IS_SYNC(inode) || inode->u.ext2_i.i_osync) {
> > ll_rw_block (WRITE, 1, &where->bh);
> > wait_on_buffer(where->bh);
> > @@ -591,7 +591,7 @@
> > wait_on_buffer(bh);
> > memset(bh->b_data, 0, inode->i_sb->s_blocksize);
> > mark_buffer_uptodate(bh, 1);
> > - mark_buffer_dirty(bh);
> > + mark_buffer_dirty_inode(bh, inode);
> > }
> > return bh;
> > }
> > @@ -907,7 +907,7 @@
> > if (partial == chain)
> > mark_inode_dirty(inode);
> > else
> > - mark_buffer_dirty(partial->bh);
> > + mark_buffer_dirty_inode(partial->bh, inode);
> > ext2_free_branches(inode, &nr, &nr+1, (chain+n-1) - partial);
> > }
> > /* Clear the ends of indirect blocks on the shared branch */
> > @@ -916,7 +916,7 @@
> > partial->p + 1,
> > (u32*)partial->bh->b_data + addr_per_block,
> > (chain+n-1) - partial);
> > - mark_buffer_dirty(partial->bh);
> > + mark_buffer_dirty_inode(partial->bh, inode);
> > if (IS_SYNC(inode)) {
> > ll_rw_block (WRITE, 1, &partial->bh);
> > wait_on_buffer (partial->bh);
> > @@ -1208,7 +1208,7 @@
> > raw_inode->i_block[0] = cpu_to_le32(kdev_t_to_nr(inode->i_rdev));
> > else for (block = 0; block < EXT2_N_BLOCKS; block++)
> > raw_inode->i_block[block] = inode->u.ext2_i.i_data[block];
> > - mark_buffer_dirty(bh);
> > + mark_buffer_dirty_inode(bh, inode);
> > if (do_sync) {
> > ll_rw_block (WRITE, 1, &bh);
> > wait_on_buffer (bh);
> > --- linux-2.4.0-test11/fs/ext2/namei.c.~1~ Tue Nov 21 15:40:22 2000
> > +++ linux-2.4.0-test11/fs/ext2/namei.c Tue Nov 21 16:01:15 2000
> > @@ -296,7 +296,7 @@
> > dir->u.ext2_i.i_flags &= ~EXT2_BTREE_FL;
> > mark_inode_dirty(dir);
> > dir->i_version = ++event;
> > - mark_buffer_dirty(bh);
> > + mark_buffer_dirty_inode(bh, dir);
> > if (IS_SYNC(dir)) {
> > ll_rw_block (WRITE, 1, &bh);
> > wait_on_buffer (bh);
> > @@ -337,7 +337,7 @@
> > else
> > de->inode = 0;
> > dir->i_version = ++event;
> > - mark_buffer_dirty(bh);
> > + mark_buffer_dirty_inode(bh, dir);
> > if (IS_SYNC(dir)) {
> > ll_rw_block (WRITE, 1, &bh);
> > wait_on_buffer (bh);
> > @@ -449,7 +449,7 @@
> > strcpy (de->name, "..");
> > ext2_set_de_type(dir->i_sb, de, S_IFDIR);
> > inode->i_nlink = 2;
> > - mark_buffer_dirty(dir_block);
> > + mark_buffer_dirty_inode(dir_block, dir);
> > brelse (dir_block);
> > inode->i_mode = S_IFDIR | mode;
> > if (dir->i_mode & S_ISGID)
> > @@ -755,7 +755,7 @@
> > EXT2_FEATURE_INCOMPAT_FILETYPE))
> > new_de->file_type = old_de->file_type;
> > new_dir->i_version = ++event;
> > - mark_buffer_dirty(new_bh);
> > + mark_buffer_dirty_inode(new_bh, new_dir);
> > if (IS_SYNC(new_dir)) {
> > ll_rw_block (WRITE, 1, &new_bh);
> > wait_on_buffer (new_bh);
> > @@ -786,7 +786,7 @@
> > mark_inode_dirty(old_dir);
> > if (dir_bh) {
> > PARENT_INO(dir_bh->b_data) = le32_to_cpu(new_dir->i_ino);
> > - mark_buffer_dirty(dir_bh);
> > + mark_buffer_dirty_inode(dir_bh, old_inode);
> > old_dir->i_nlink--;
> > mark_inode_dirty(old_dir);
> > if (new_inode) {
> > --- linux-2.4.0-test11/include/linux/ext2_fs.h.~1~ Tue Nov 21 15:40:22 2000
> > +++ linux-2.4.0-test11/include/linux/ext2_fs.h Tue Nov 21 16:01:15 2000
> > @@ -549,6 +549,7 @@
> >
> > /* fsync.c */
> > extern int ext2_sync_file (struct file *, struct dentry *, int);
> > +extern int ext2_fsync_inode (struct inode *, int);
> >
> > /* ialloc.c */
> > extern struct inode * ext2_new_inode (const struct inode *, int, int *);
> > --- linux-2.4.0-test11/mm/filemap.c.~1~ Tue Nov 21 15:47:48 2000
> > +++ linux-2.4.0-test11/mm/filemap.c Tue Nov 21 16:01:15 2000
> > @@ -2521,8 +2521,14 @@
> > if (cached_page)
> > page_cache_free(cached_page);
> >
> > + /* For now, when the user asks for O_SYNC, we'll actually
> > + * provide O_DSYNC. */
> > + if ((status >= 0) && (file->f_flags & O_SYNC))
> > + status = generic_osync_inode(inode, 1); /* 1 means datasync */
> > +
> > err = written ? written : status;
> > out:
> > +
> > up(&inode->i_sem);
> > return err;
> > fail_write:
> > -
> > To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
> > the body of a message to [email protected]
> > Please read the FAQ at http://www.tux.org/lkml/
> -
> To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
> the body of a message to [email protected]
> Please read the FAQ at http://www.tux.org/lkml/
--
=======================================================================
Josue Emmanuel Amaro [email protected]
Linux Products Manager Phone: 650.506.1239
Intel and Linux Technologies Group Fax: 650.413.0167
=======================================================================
Hi,
On Wed, Nov 22, 2000 at 11:54:24AM -0700, Jeff V. Merkey wrote:
>
> I have not implemented O_SYNC in NWFS, but it looks like I need to add it
> before posting the final patches. This patch appears to force write-through
> of only dirty inodes, and allow reads to continue from cache. Is this
> assumption correct
Yes: O_SYNC is not required to force reads to be made from disk.
SingleUnix has an "O_RSYNC" option which does that, but O_SYNC and
O_DSYNC don't imply that.
Cheers,
Stephen
On Thu, Nov 23, 2000 at 12:01:35PM +0000, Stephen C. Tweedie wrote:
> Hi,
>
> On Wed, Nov 22, 2000 at 11:54:24AM -0700, Jeff V. Merkey wrote:
> >
> > I have not implemented O_SYNC in NWFS, but it looks like I need to add it
> > before posting the final patches. This patch appears to force write-through
> > of only dirty inodes, and allow reads to continue from cache. Is this
> > assumption correct
>
> Yes: O_SYNC is not required to force reads to be made from disk.
> SingleUnix has an "O_RSYNC" option which does that, but O_SYNC and
> O_DSYNC don't imply that.
Cool. ORACLE is going to **SMOKE** on EXT2 with this change.
Jeff
>
> Cheers,
> Stephen
On Thu, Nov 23, 2000 at 01:01:25PM -0700, Jeff V. Merkey wrote:
> On Thu, Nov 23, 2000 at 12:01:35PM +0000, Stephen C. Tweedie wrote:
> > Hi,
> >
> > On Wed, Nov 22, 2000 at 11:54:24AM -0700, Jeff V. Merkey wrote:
> > >
> > > I have not implemented O_SYNC in NWFS, but it looks like I need to add it
> > > before posting the final patches. This patch appears to force write-through
> > > of only dirty inodes, and allow reads to continue from cache. Is this
> > > assumption correct
> >
> > Yes: O_SYNC is not required to force reads to be made from disk.
> > SingleUnix has an "O_RSYNC" option which does that, but O_SYNC and
> > O_DSYNC don't imply that.
>
> Cool. ORACLE is going to **SMOKE** on EXT2 with this change.
Note that this is nothing new, linux (say 2.2.18pre23) always used the O_SYNC
semantics Stephen implemented in the 2.4.x O_SYNC showstopper bugfix.
Andrea
"Jeff V. Merkey" <[email protected]> writes:
> Cool. ORACLE is going to **SMOKE** on EXT2 with this change.
<Pessimism>
Hmm I don't see how ORACLE is going to **SMOKE**.
Last I looked ORACLE would need a query optimizer that always
would find the best possible index and much less overhead to **SMOKE**.
Last I looked table reads were 10x slower than file reads.
</Pessimism>
Eric