Currently we fall through to nfs4_async_handle_error when we get a bad stateid error back from layoutget. nfs4_async_handle_error with a NULL state argument will never retry the operations but return the error to higher layer, causing an avoiable fallback to MDS I/O. Signed-off-by: Christoph Hellwig <hch@xxxxxx> --- fs/nfs/nfs4proc.c | 9 +++++++-- 1 file changed, 7 insertions(+), 2 deletions(-) diff --git a/fs/nfs/nfs4proc.c b/fs/nfs/nfs4proc.c index 75ae8d2..4ca5880 100644 --- a/fs/nfs/nfs4proc.c +++ b/fs/nfs/nfs4proc.c @@ -7571,11 +7571,16 @@ static void nfs4_layoutget_done(struct rpc_task *task, void *calldata) } else { LIST_HEAD(head); + /* + * Mark the bad layout state as invalid, then retry + * with the current stateid. + */ pnfs_mark_matching_lsegs_invalid(lo, &head, NULL); spin_unlock(&inode->i_lock); - /* Mark the bad layout state as invalid, then - * retry using the open stateid. */ pnfs_free_lseg_list(&head); + + task->tk_status = 0; + rpc_restart_call_prepare(task); } } if (nfs4_async_handle_error(task, server, state) == -EAGAIN) -- 1.9.1 -- 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