[PATCH 2/6] xfs: use i_mmaplock on read faults
Brian Foster
bfoster at redhat.com
Mon Feb 9 14:31:21 CST 2015
On Thu, Feb 05, 2015 at 08:04:15AM +1100, Dave Chinner wrote:
> From: Dave Chinner <dchinner at redhat.com>
>
> Take the i_mmaplock over read page faults. These come through the
> ->fault callout, so we need to wrap the generic implementation
> with the i_mmaplock. While there, add tracepoints for the read
> fault as it passes through XFS.
>
> This gives us a lock order of mmap_sem -> i_mmaplock -> page_lock
> -> i_lock.
>
> Signed-off-by: Dave Chinner <dchinner at redhat.com>
> ---
Reviewed-by: Brian Foster <bfoster at redhat.com>
> fs/xfs/xfs_file.c | 28 +++++++++++++++++++++++++++-
> fs/xfs/xfs_trace.h | 2 ++
> 2 files changed, 29 insertions(+), 1 deletion(-)
>
> diff --git a/fs/xfs/xfs_file.c b/fs/xfs/xfs_file.c
> index 712d312..6cc1b7d 100644
> --- a/fs/xfs/xfs_file.c
> +++ b/fs/xfs/xfs_file.c
> @@ -1369,6 +1369,32 @@ xfs_file_llseek(
> }
> }
>
> +/*
> + * Locking for serialisation of IO during page faults. This results in a lock
> + * ordering of:
> + *
> + * mmap_sem (MM)
> + * i_mmap_lock (XFS - truncate serialisation)
> + * page_lock (MM)
> + * i_lock (XFS - extent map serialisation)
> + */
> +STATIC int
> +xfs_filemap_fault(
> + struct vm_area_struct *vma,
> + struct vm_fault *vmf)
> +{
> + struct xfs_inode *ip = XFS_I(vma->vm_file->f_mapping->host);
> + int error;
> +
> + trace_xfs_filemap_fault(ip);
> +
> + xfs_ilock(ip, XFS_MMAPLOCK_SHARED);
> + error = filemap_fault(vma, vmf);
> + xfs_iunlock(ip, XFS_MMAPLOCK_SHARED);
> +
> + return error;
> +}
> +
> const struct file_operations xfs_file_operations = {
> .llseek = xfs_file_llseek,
> .read = new_sync_read,
> @@ -1401,7 +1427,7 @@ const struct file_operations xfs_dir_file_operations = {
> };
>
> static const struct vm_operations_struct xfs_file_vm_ops = {
> - .fault = filemap_fault,
> + .fault = xfs_filemap_fault,
> .map_pages = filemap_map_pages,
> .page_mkwrite = xfs_vm_page_mkwrite,
> .remap_pages = generic_file_remap_pages,
> diff --git a/fs/xfs/xfs_trace.h b/fs/xfs/xfs_trace.h
> index 51372e3..c496153 100644
> --- a/fs/xfs/xfs_trace.h
> +++ b/fs/xfs/xfs_trace.h
> @@ -685,6 +685,8 @@ DEFINE_INODE_EVENT(xfs_inode_set_eofblocks_tag);
> DEFINE_INODE_EVENT(xfs_inode_clear_eofblocks_tag);
> DEFINE_INODE_EVENT(xfs_inode_free_eofblocks_invalid);
>
> +DEFINE_INODE_EVENT(xfs_filemap_fault);
> +
> DECLARE_EVENT_CLASS(xfs_iref_class,
> TP_PROTO(struct xfs_inode *ip, unsigned long caller_ip),
> TP_ARGS(ip, caller_ip),
> --
> 2.0.0
>
> _______________________________________________
> xfs mailing list
> xfs at oss.sgi.com
> http://oss.sgi.com/mailman/listinfo/xfs
More information about the xfs
mailing list