For uninit block group, the ondisk bitmap is not initialized. That implies
we cannot depend on the uptodate flag on the bitmap buffer_head to
find bitmap validity. Use a new buffer_head flag which would be set after
we properly initialize the bitmap. This also prevent the initializing
the uninit group bitmap initialization every time we do a
ext4_read_block_bitmap.
Signed-off-by: Aneesh Kumar K.V <[email protected]>
---
fs/ext4/balloc.c | 16 ++++++++++++++--
fs/ext4/ext4.h | 18 ++++++++++++++++++
fs/ext4/ialloc.c | 15 +++++++++++++--
fs/ext4/mballoc.c | 15 +++++++++++++--
include/linux/jbd2.h | 1 +
5 files changed, 59 insertions(+), 6 deletions(-)
diff --git a/fs/ext4/balloc.c b/fs/ext4/balloc.c
index 3cae4c6..b96e49a 100644
--- a/fs/ext4/balloc.c
+++ b/fs/ext4/balloc.c
@@ -320,20 +320,32 @@ ext4_read_block_bitmap(struct super_block *sb, ext4_group_t block_group)
block_group, bitmap_blk);
return NULL;
}
- if (buffer_uptodate(bh) &&
- !(desc->bg_flags & cpu_to_le16(EXT4_BG_BLOCK_UNINIT)))
+
+ if (bitmap_uptodate(bh))
return bh;
lock_buffer(bh);
+ if (bitmap_uptodate(bh)) {
+ unlock_buffer(bh);
+ return bh;
+ }
spin_lock(sb_bgl_lock(EXT4_SB(sb), block_group));
if (desc->bg_flags & cpu_to_le16(EXT4_BG_BLOCK_UNINIT)) {
ext4_init_block_bitmap(sb, bh, block_group, desc);
+ set_bitmap_uptodate(bh);
set_buffer_uptodate(bh);
spin_unlock(sb_bgl_lock(EXT4_SB(sb), block_group));
unlock_buffer(bh);
return bh;
}
spin_unlock(sb_bgl_lock(EXT4_SB(sb), block_group));
+ /*
+ * submit the buffer_head for read. We can
+ * safely mark the bitmap as uptodate now.
+ * We do it here so the bitmap uptodate bit
+ * get set with buffer lock held.
+ */
+ set_bitmap_uptodate(bh);
if (bh_submit_read(bh) < 0) {
put_bh(bh);
ext4_error(sb, __func__,
diff --git a/fs/ext4/ext4.h b/fs/ext4/ext4.h
index 64e535d..4a803b7 100644
--- a/fs/ext4/ext4.h
+++ b/fs/ext4/ext4.h
@@ -19,6 +19,7 @@
#include <linux/types.h>
#include <linux/blkdev.h>
#include <linux/magic.h>
+#include <linux/jbd2.h>
#include "ext4_i.h"
/*
@@ -1350,6 +1351,23 @@ extern int ext4_get_blocks_wrap(handle_t *handle, struct inode *inode,
extern int ext4_fiemap(struct inode *inode, struct fiemap_extent_info *fieinfo,
__u64 start, __u64 len);
+/*
+ * Add new method to test wether block and inode bitmaps are properly
+ * initialized. With uninit_bg reading the block from disk is not enough
+ * to mark the bitmap uptodate. We need to also zero-out the bitmap
+ */
+#define BH_BITMAP_UPTODATE BH_JBD_State_bits_End
+
+static inline int bitmap_uptodate(struct buffer_head *bh)
+{
+ return (buffer_uptodate(bh) &&
+ test_bit(BH_BITMAP_UPTODATE, &(bh)->b_state));
+}
+static inline void set_bitmap_uptodate(struct buffer_head *bh)
+{
+ set_bit(BH_BITMAP_UPTODATE, &(bh)->b_state);
+}
+
#endif /* __KERNEL__ */
#endif /* _EXT4_H */
diff --git a/fs/ext4/ialloc.c b/fs/ext4/ialloc.c
index d1ccae5..742c545 100644
--- a/fs/ext4/ialloc.c
+++ b/fs/ext4/ialloc.c
@@ -115,20 +115,31 @@ ext4_read_inode_bitmap(struct super_block *sb, ext4_group_t block_group)
block_group, bitmap_blk);
return NULL;
}
- if (buffer_uptodate(bh) &&
- !(desc->bg_flags & cpu_to_le16(EXT4_BG_INODE_UNINIT)))
+ if (bitmap_uptodate(bh))
return bh;
lock_buffer(bh);
+ if (bitmap_uptodate(bh)) {
+ unlock_buffer(bh);
+ return bh;
+ }
spin_lock(sb_bgl_lock(EXT4_SB(sb), block_group));
if (desc->bg_flags & cpu_to_le16(EXT4_BG_INODE_UNINIT)) {
ext4_init_inode_bitmap(sb, bh, block_group, desc);
+ set_bitmap_uptodate(bh);
set_buffer_uptodate(bh);
spin_unlock(sb_bgl_lock(EXT4_SB(sb), block_group));
unlock_buffer(bh);
return bh;
}
spin_unlock(sb_bgl_lock(EXT4_SB(sb), block_group));
+ /*
+ * submit the buffer_head for read. We can
+ * safely mark the bitmap as uptodate now.
+ * We do it here so the bitmap uptodate bit
+ * get set with buffer lock held.
+ */
+ set_bitmap_uptodate(bh);
if (bh_submit_read(bh) < 0) {
put_bh(bh);
ext4_error(sb, __func__,
diff --git a/fs/ext4/mballoc.c b/fs/ext4/mballoc.c
index 249d9c7..d7a165c 100644
--- a/fs/ext4/mballoc.c
+++ b/fs/ext4/mballoc.c
@@ -794,15 +794,19 @@ static int ext4_mb_init_cache(struct page *page, char *incore)
if (bh[i] == NULL)
goto out;
- if (buffer_uptodate(bh[i]) &&
- !(desc->bg_flags & cpu_to_le16(EXT4_BG_BLOCK_UNINIT)))
+ if (bitmap_uptodate(bh[i]))
continue;
lock_buffer(bh[i]);
+ if (bitmap_uptodate(bh[i])) {
+ unlock_buffer(bh[i]);
+ continue;
+ }
spin_lock(sb_bgl_lock(EXT4_SB(sb), first_group + i));
if (desc->bg_flags & cpu_to_le16(EXT4_BG_BLOCK_UNINIT)) {
ext4_init_block_bitmap(sb, bh[i],
first_group + i, desc);
+ set_bitmap_uptodate(bh[i]);
set_buffer_uptodate(bh[i]);
spin_unlock(sb_bgl_lock(EXT4_SB(sb), first_group + i));
unlock_buffer(bh[i]);
@@ -810,6 +814,13 @@ static int ext4_mb_init_cache(struct page *page, char *incore)
}
spin_unlock(sb_bgl_lock(EXT4_SB(sb), first_group + i));
get_bh(bh[i]);
+ /*
+ * submit the buffer_head for read. We can
+ * safely mark the bitmap as uptodate now.
+ * We do it here so the bitmap uptodate bit
+ * get set with buffer lock held.
+ */
+ set_bitmap_uptodate(bh[i]);
bh[i]->b_end_io = end_buffer_read_sync;
submit_bh(READ, bh[i]);
mb_debug("read bitmap for group %u\n", first_group + i);
diff --git a/include/linux/jbd2.h b/include/linux/jbd2.h
index 4932b34..6694561 100644
--- a/include/linux/jbd2.h
+++ b/include/linux/jbd2.h
@@ -329,6 +329,7 @@ enum jbd_state_bits {
BH_State, /* Pins most journal_head state */
BH_JournalHead, /* Pins bh->b_private and jh->b_bh */
BH_Unshadow, /* Dummy bit, for BJ_Shadow wakeup filtering */
+ BH_JBD_State_bits_End,
};
BUFFER_FNS(JBD, jbd)
--
1.6.0.4.735.gea4f
On Wed, Nov 26, 2008 at 08:24:10PM +0530, Aneesh Kumar K.V wrote:
> diff --git a/include/linux/jbd2.h b/include/linux/jbd2.h
> index 4932b34..6694561 100644
> --- a/include/linux/jbd2.h
> +++ b/include/linux/jbd2.h
> @@ -329,6 +329,7 @@ enum jbd_state_bits {
> BH_State, /* Pins most journal_head state */
> BH_JournalHead, /* Pins bh->b_private and jh->b_bh */
> BH_Unshadow, /* Dummy bit, for BJ_Shadow wakeup filtering */
> + BH_JBD_State_bits_End,
> };
>
> BUFFER_FNS(JBD, jbd)
Note: this conflicts with a patch by Mark Fasheh which does something
very similar, since OCFS2 also needs some private BH flags. He used
the name BH_JBDPrivateStart, though.
Originally the plan was going to be that Mark was going to send this
to Linus, but given that we need it as well, I suggest that we drop it
into both the OCFS2 and ext4 trees, and whichever one hits Linus's
tree first will win, and in the other case the magic of git's merge
algorithms should make the right thing happen in the second. (Or the
other team can drop the patch before they merge; either will do the
right thing.)
For this path, it would mean dropping this hunk and adding Mark's
patch to the ext4 tree. If this makes sense to everyone, Aneesh, you
don't need to send a patch; I can fix up the one you sent easily
enough.
- Ted
On Wed, Nov 26, 2008 at 10:56:43AM -0500, Theodore Tso wrote:
> On Wed, Nov 26, 2008 at 08:24:10PM +0530, Aneesh Kumar K.V wrote:
> > diff --git a/include/linux/jbd2.h b/include/linux/jbd2.h
> > index 4932b34..6694561 100644
> > --- a/include/linux/jbd2.h
> > +++ b/include/linux/jbd2.h
> > @@ -329,6 +329,7 @@ enum jbd_state_bits {
> > BH_State, /* Pins most journal_head state */
> > BH_JournalHead, /* Pins bh->b_private and jh->b_bh */
> > BH_Unshadow, /* Dummy bit, for BJ_Shadow wakeup filtering */
> > + BH_JBD_State_bits_End,
> > };
> >
> > BUFFER_FNS(JBD, jbd)
>
> Note: this conflicts with a patch by Mark Fasheh which does something
> very similar, since OCFS2 also needs some private BH flags. He used
> the name BH_JBDPrivateStart, though.
>
> Originally the plan was going to be that Mark was going to send this
> to Linus, but given that we need it as well, I suggest that we drop it
> into both the OCFS2 and ext4 trees, and whichever one hits Linus's
> tree first will win, and in the other case the magic of git's merge
> algorithms should make the right thing happen in the second. (Or the
> other team can drop the patch before they merge; either will do the
> right thing.)
>
> For this path, it would mean dropping this hunk and adding Mark's
> patch to the ext4 tree. If this makes sense to everyone, Aneesh, you
> don't need to send a patch; I can fix up the one you sent easily
> enough.
Should be fine.
-aneesh
On Wed, Nov 26, 2008 at 10:56:43AM -0500, Theodore Tso wrote:
> On Wed, Nov 26, 2008 at 08:24:10PM +0530, Aneesh Kumar K.V wrote:
> > diff --git a/include/linux/jbd2.h b/include/linux/jbd2.h
> > index 4932b34..6694561 100644
> > --- a/include/linux/jbd2.h
> > +++ b/include/linux/jbd2.h
> > @@ -329,6 +329,7 @@ enum jbd_state_bits {
> > BH_State, /* Pins most journal_head state */
> > BH_JournalHead, /* Pins bh->b_private and jh->b_bh */
> > BH_Unshadow, /* Dummy bit, for BJ_Shadow wakeup filtering */
> > + BH_JBD_State_bits_End,
> > };
> >
> > BUFFER_FNS(JBD, jbd)
>
> Note: this conflicts with a patch by Mark Fasheh which does something
> very similar, since OCFS2 also needs some private BH flags. He used
> the name BH_JBDPrivateStart, though.
>
> Originally the plan was going to be that Mark was going to send this
> to Linus, but given that we need it as well, I suggest that we drop it
> into both the OCFS2 and ext4 trees, and whichever one hits Linus's
> tree first will win, and in the other case the magic of git's merge
> algorithms should make the right thing happen in the second. (Or the
> other team can drop the patch before they merge; either will do the
> right thing.)
Yeah, that's fine with me. We'll both carry it for now and the merge window
will just get sorted out :)
--Mark
--
Mark Fasheh