Prev: linux-next: build failure after merge of the xfs tree
Next: [PATCH 0/4] memcg reclaim tracepoint v2
From: Stephen Rothwell on 25 Jul 2010 22:00
Hi Al,

[These became irrelevant when I reverted the xfs tree for a build error,
but will be done when the xfs tree is fixed.]

Today's linux-next merge of the vfs tree got a conflict in
fs/xfs/linux-2.6/xfs_aops.c between commit
7346e1197eb76e22199b6b4625f129331e0fd7ac ("xfs simplify and speed up
direct I/O completions") from the xfs tree and commit
b682ce06449c35a85dc5e63fcaab7dba2ba6df9b ("sort out blockdev_direct_IO
variants") from the vfs tree.

Also between commit 6f6b39eb706f5617750cf02952e4e6d7470c40bf ("xfs: use
GFP_NOFS for page cache allocation") from the xfs tree and commit
4478b21470275454f9eb6f590dfe7cc54d643978 ("get rid of
block_write_begin_newtrunc") from the vfs tree.

I fixed them up (I think - see below) and can carry the fix as necessary.
--
Cheers,
Stephen Rothwell sfr(a)canb.auug.org.au

diff --cc fs/xfs/linux-2.6/xfs_aops.c
index d24e78f,77a9ecc..0000000
--- a/fs/xfs/linux-2.6/xfs_aops.c
+++ b/fs/xfs/linux-2.6/xfs_aops.c
@@@ -1471,26 -1655,22 +1471,26 @@@ xfs_vm_direct_IO
loff_t offset,
unsigned long nr_segs)
{
- struct file *file = iocb->ki_filp;
- struct inode *inode = file->f_mapping->host;
- struct block_device *bdev;
- ssize_t ret;
-
- bdev = xfs_find_bdev_for_inode(inode);
-
- iocb->private = xfs_alloc_ioend(inode, rw == WRITE ?
- IO_UNWRITTEN : IO_READ);
-
- ret = __blockdev_direct_IO(rw, iocb, inode, bdev, iov, offset,
- nr_segs, xfs_get_blocks_direct,
- xfs_end_io_direct, NULL, 0);
+ struct inode *inode = iocb->ki_filp->f_mapping->host;
+ struct block_device *bdev = xfs_find_bdev_for_inode(inode);
+ ssize_t ret;
+
+ if (rw & WRITE) {
+ iocb->private = xfs_alloc_ioend(inode, IO_NEW);
+
- ret = blockdev_direct_IO_no_locking(rw, iocb, inode, bdev, iov,
- offset, nr_segs,
- xfs_get_blocks_direct,
- xfs_end_io_direct_write);
++ ret = __blockdev_direct_IO(rw, iocb, inode, bdev, iov,
++ offset, nr_segs,
++ xfs_get_blocks_direct,
++ xfs_end_io_direct_write, NULL, 0);
+ if (ret != -EIOCBQUEUED && iocb->private)
+ xfs_destroy_ioend(iocb->private);
+ } else {
- ret = blockdev_direct_IO_no_locking(rw, iocb, inode, bdev, iov,
- offset, nr_segs,
- xfs_get_blocks_direct,
- NULL);
++ ret = __blockdev_direct_IO(rw, iocb, inode, bdev, iov,
++ offset, nr_segs,
++ xfs_get_blocks_direct,
++ NULL, NULL, 0);
+ }

- if (unlikely(ret != -EIOCBQUEUED && iocb->private))
- xfs_destroy_ioend(iocb->private);
return ret;
}

@@@ -1504,9 -1700,31 +1520,31 @@@ xfs_vm_write_begin
struct page **pagep,
void **fsdata)
{
- *pagep = NULL;
- return block_write_begin(file, mapping, pos, len, flags | AOP_FLAG_NOFS,
- pagep, fsdata, xfs_get_blocks);
+ int ret;
+
- ret = block_write_begin(mapping, pos, len, flags, pagep,
++ ret = block_write_begin(mapping, pos, len, flags | AOP_FLAG_NOFS, pagep,
+ xfs_get_blocks);
+ if (unlikely(ret))
+ xfs_vm_write_failed(mapping, pos + len);
+ return ret;
+ }
+
+ STATIC int
+ xfs_vm_write_end(
+ struct file *file,
+ struct address_space *mapping,
+ loff_t pos,
+ unsigned len,
+ unsigned copied,
+ struct page *page,
+ void *fsdata)
+ {
+ int ret;
+
+ ret = generic_write_end(file, mapping, pos, len, copied, page, fsdata);
+ if (unlikely(ret < len))
+ xfs_vm_write_failed(mapping, pos + len);
+ return ret;
}

STATIC sector_t
--
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo(a)vger.kernel.org
More majordomo info at http://vger.kernel.org/majordomo-info.html
Please read the FAQ at http://www.tux.org/lkml/
From: Dave Chinner on 26 Jul 2010 00:10
On Mon, Jul 26, 2010 at 11:51:16AM +1000, Stephen Rothwell wrote:
> Hi Al,
>
> [These became irrelevant when I reverted the xfs tree for a build error,
> but will be done when the xfs tree is fixed.]
>
> Today's linux-next merge of the vfs tree got a conflict in
> fs/xfs/linux-2.6/xfs_aops.c between commit
> 7346e1197eb76e22199b6b4625f129331e0fd7ac ("xfs simplify and speed up
> direct I/O completions") from the xfs tree and commit
> b682ce06449c35a85dc5e63fcaab7dba2ba6df9b ("sort out blockdev_direct_IO
> variants") from the vfs tree.
>
> Also between commit 6f6b39eb706f5617750cf02952e4e6d7470c40bf ("xfs: use
> GFP_NOFS for page cache allocation") from the xfs tree and commit
> 4478b21470275454f9eb6f590dfe7cc54d643978 ("get rid of
> block_write_begin_newtrunc") from the vfs tree.
>
> I fixed them up (I think - see below) and can carry the fix as necessary.

The fix looks OK. Thanks Stephen.

Cheers,

Dave.
--
Dave Chinner
david(a)fromorbit.com
--
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo(a)vger.kernel.org
More majordomo info at http://vger.kernel.org/majordomo-info.html
Please read the FAQ at http://www.tux.org/lkml/
From: Stephen Rothwell on 26 Jul 2010 00:10
Hi Dave,

On Mon, 26 Jul 2010 13:59:50 +1000 Dave Chinner <david(a)fromorbit.com> wrote:
>
> The fix looks OK. Thanks Stephen.

Thanks for the confirmation.

--
Cheers,
Stephen Rothwell sfr(a)canb.auug.org.au
http://www.canb.auug.org.au/~sfr/
From: Stephen Rothwell on 3 Aug 2010 22:00
Hi Al,

On Mon, 5 Jul 2010 10:02:48 +1000 Stephen Rothwell <sfr(a)canb.auug.org.au> wrote:
>
> Today's linux-next merge of the vfs tree got a conflict in
> fs/xfs/linux-2.6/xfs_super.c between commit
> 04d2da9c682ce2687db29136c60d8679e37e3b35 ("xfs: split xfs_itrace_entry")
> from the xfs tree and commit b43cb885cc49c782dd0734292d3cd95631200e5c
> ("convert remaining ->clear_inode() to ->evict_inode()") from the vfs
> tree.
>
> Just context changes. I fixed it up (see below) and can carry the fix as
> necessary. I guess that the tracing code needs updating for the
> xfs_fs_clear_inode -> xfs_fs_evict_inode name change as well. And I am
> wondering if the tracing wants to go ahead of the truncate_inode_pages()
> call?
>
> In anticipation, I have added this merge fix patch as well:

Linus has now merged the xfs tree, so this patch can be applied to the
vfs tree (assuming it is ok).

From: Stephen Rothwell <sfr(a)canb.auug.org.au>
Date: Mon, 5 Jul 2010 09:43:42 +1000
Subject: [PATCH] xfs: update tracing for clear_inode to evict_inode transition

Signed-off-by: Stephen Rothwell <sfr(a)canb.auug.org.au>
---
fs/xfs/linux-2.6/xfs_super.c | 4 ++--
fs/xfs/linux-2.6/xfs_trace.h | 2 +-
2 files changed, 3 insertions(+), 3 deletions(-)

diff --git a/fs/xfs/linux-2.6/xfs_super.c b/fs/xfs/linux-2.6/xfs_super.c
index 157aaed..3e5052c 100644
--- a/fs/xfs/linux-2.6/xfs_super.c
+++ b/fs/xfs/linux-2.6/xfs_super.c
@@ -1107,10 +1107,10 @@ xfs_fs_evict_inode(
{
xfs_inode_t *ip = XFS_I(inode);

+ trace_xfs_evict_inode(ip);
+
truncate_inode_pages(&inode->i_data, 0);
end_writeback(inode);
- trace_xfs_clear_inode(ip);
-
XFS_STATS_INC(vn_rele);
XFS_STATS_INC(vn_remove);
XFS_STATS_DEC(vn_active);
diff --git a/fs/xfs/linux-2.6/xfs_trace.h b/fs/xfs/linux-2.6/xfs_trace.h
index d506753..76de133 100644
--- a/fs/xfs/linux-2.6/xfs_trace.h
+++ b/fs/xfs/linux-2.6/xfs_trace.h
@@ -576,7 +576,7 @@ DEFINE_INODE_EVENT(xfs_ioctl_setattr);
DEFINE_INODE_EVENT(xfs_file_fsync);
DEFINE_INODE_EVENT(xfs_destroy_inode);
DEFINE_INODE_EVENT(xfs_write_inode);
-DEFINE_INODE_EVENT(xfs_clear_inode);
+DEFINE_INODE_EVENT(xfs_evict_inode);

DEFINE_INODE_EVENT(xfs_dquot_dqalloc);
DEFINE_INODE_EVENT(xfs_dquot_dqdetach);
--
1.7.1

--
Cheers,
Stephen Rothwell sfr(a)canb.auug.org.au
http://www.canb.auug.org.au/~sfr/
--
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo(a)vger.kernel.org
More majordomo info at http://vger.kernel.org/majordomo-info.html
Please read the FAQ at http://www.tux.org/lkml/
 | 
Pages: 1
Prev: linux-next: build failure after merge of the xfs tree
Next: [PATCH 0/4] memcg reclaim tracepoint v2