Christoph points out that the VFS will always flush out data before calling
nfs_fsync(), so we can dispense with a full call to nfs_wb_all(), and
replace that with a simpler call to nfs_commit_inode().
Signed-off-by: Trond Myklebust <Trond.Myklebust@xxxxxxxxxx>
---
fs/nfs/file.c | 50 +++++++++++++++++++++-----------------------------
fs/nfs/internal.h | 1 +
fs/nfs/write.c | 4 ++--
3 files changed, 24 insertions(+), 31 deletions(-)
diff --git a/fs/nfs/file.c b/fs/nfs/file.c
index 36a5e74..4bf7306 100644
--- a/fs/nfs/file.c
+++ b/fs/nfs/file.c
@@ -202,31 +202,6 @@ static loff_t nfs_file_llseek(struct file *filp, loff_t offset, int origin)
}
/*
- * Helper for nfs_file_flush() and nfs_file_fsync()
- *
- * Notice that it clears the NFS_CONTEXT_ERROR_WRITE before synching to
- * disk, but it retrieves and clears ctx->error after synching, despite
- * the two being set at the same time in nfs_context_set_write_error().
- * This is because the former is used to notify the _next_ call to
- * nfs_file_write() that a write error occured, and hence cause it to
- * fall back to doing a synchronous write.
- */
-static int nfs_do_fsync(struct nfs_open_context *ctx, struct inode *inode)
-{
- int have_error, status;
- int ret = 0;
-
- have_error = test_and_clear_bit(NFS_CONTEXT_ERROR_WRITE, &ctx->flags);
- status = nfs_wb_all(inode);
- have_error |= test_bit(NFS_CONTEXT_ERROR_WRITE, &ctx->flags);
- if (have_error)
- ret = xchg(&ctx->error, 0);
- if (!ret)
- ret = status;
- return ret;
-}
-
-/*
* Flush all dirty pages, and check for write errors.
*/
static int
@@ -245,7 +220,7 @@ nfs_file_flush(struct file *file, fl_owner_t id)
return 0;
/* Flush writes to the server and return any errors */
- return nfs_do_fsync(ctx, inode);
+ return vfs_fsync(file, 0);
}
static ssize_t
@@ -320,6 +295,13 @@ nfs_file_mmap(struct file * file, struct vm_area_struct * vma)
* Flush any dirty pages for this process, and check for write errors.
* The return status from this call provides a reliable indication of
* whether any write errors occurred for this process.
+ *
+ * Notice that it clears the NFS_CONTEXT_ERROR_WRITE before synching to
+ * disk, but it retrieves and clears ctx->error after synching, despite
+ * the two being set at the same time in nfs_context_set_write_error().
+ * This is because the former is used to notify the _next_ call to
+ * nfs_file_write() that a write error occured, and hence cause it to
+ * fall back to doing a synchronous write.
*/
static int
nfs_file_fsync(struct file *file, int datasync)
@@ -327,13 +309,23 @@ nfs_file_fsync(struct file *file, int datasync)
struct dentry *dentry = file->f_path.dentry;
struct nfs_open_context *ctx = nfs_file_open_context(file);
struct inode *inode = dentry->d_inode;
+ int have_error, status;
+ int ret = 0;
+
dprintk("NFS: fsync file(%s/%s) datasync %d\n",
dentry->d_parent->d_name.name, dentry->d_name.name,
datasync);
nfs_inc_stats(inode, NFSIOS_VFSFSYNC);
- return nfs_do_fsync(ctx, inode);
+ have_error = test_and_clear_bit(NFS_CONTEXT_ERROR_WRITE, &ctx->flags);
+ status = nfs_commit_inode(inode, FLUSH_SYNC);
+ have_error |= test_bit(NFS_CONTEXT_ERROR_WRITE, &ctx->flags);
+ if (have_error)
+ ret = xchg(&ctx->error, 0);
+ if (!ret)
+ ret = status;
+ return ret;
}
/*
@@ -639,7 +631,7 @@ static ssize_t nfs_file_write(struct kiocb *iocb, const struct iovec *iov,
/* Return error values for O_DSYNC and IS_SYNC() */
if (result >= 0 && nfs_need_sync_write(iocb->ki_filp, inode)) {
- int err = nfs_do_fsync(nfs_file_open_context(iocb->ki_filp), inode);
+ int err = vfs_fsync(iocb->ki_filp, 0);
if (err < 0)
result = err;
}
@@ -675,7 +667,7 @@ static ssize_t nfs_file_splice_write(struct pipe_inode_info *pipe,
written = ret;
if (ret >= 0 && nfs_need_sync_write(filp, inode)) {
- int err = nfs_do_fsync(nfs_file_open_context(filp), inode);
+ int err = vfs_fsync(filp, 0);
if (err < 0)
ret = err;
}
diff --git a/fs/nfs/internal.h b/fs/nfs/internal.h
index d8bd619..bdaaedf 100644
--- a/fs/nfs/internal.h
+++ b/fs/nfs/internal.h
@@ -251,6 +251,7 @@ extern int nfs4_get_rootfh(struct nfs_server *server, struct nfs_fh *mntfh);
extern void nfs_read_prepare(struct rpc_task *task, void *calldata);
/* write.c */
+extern int nfs_commit_inode(struct inode *inode, int how);
extern void nfs_write_prepare(struct rpc_task *task, void *calldata);
#ifdef CONFIG_MIGRATION
extern int nfs_migrate_page(struct address_space *,
diff --git a/fs/nfs/write.c b/fs/nfs/write.c
index 5eccea1..8fda8a7 100644
--- a/fs/nfs/write.c
+++ b/fs/nfs/write.c
@@ -1382,7 +1382,7 @@ static const struct rpc_call_ops nfs_commit_ops = {
.rpc_release = nfs_commit_release,
};
-static int nfs_commit_inode(struct inode *inode, int how)
+int nfs_commit_inode(struct inode *inode, int how)
{
LIST_HEAD(head);
int may_wait = how & FLUSH_SYNC;
@@ -1446,7 +1446,7 @@ out_mark_dirty:
return ret;
}
#else
-static int nfs_commit_inode(struct inode *inode, int how)
+int nfs_commit_inode(struct inode *inode, int how)
{
return 0;
}
--
To unsubscribe from this list: send the line "unsubscribe linux-nfs" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at http://vger.kernel.org/majordomo-info.html
