NFSv4.x/pnfs: Fix a race between layoutget and pnfs_destroy_layout

If the server reboots while there is a layoutget outstanding, then
the call to pnfs_choose_layoutget_stateid() will fail with an EAGAIN
error, which causes an infinite loop in send_layoutget(). The reason
why we never break out of the loop is that the layout 'plh_block_lgets'
field is never cleared.

Fix is to replace plh_block_lgets with NFS_LAYOUT_INVALID_STID, which
can be reset after a new layoutget.

Fixes: ab7d763e477c5 ("pNFS: Ensure nfs4_layoutget_prepare returns...")
Signed-off-by: Trond Myklebust <trond.myklebust@primarydata.com>

1 files changed, 22 insertions, 2 deletions

diff --git a/fs/nfs/pnfs.c b/fs/nfs/pnfs.c
index 5064f23ca8cc..8a41538f1155 100644
--- a/fs/nfs/pnfs.c
+++ b/fs/nfs/pnfs.c
@@ -252,6 +252,27 @@ pnfs_put_layout_hdr(struct pnfs_layout_hdr *lo)
         }
 }
 
+/*
+ * Mark a pnfs_layout_hdr and all associated layout segments as invalid
+ *
+ * In order to continue using the pnfs_layout_hdr, a full recovery
+ * is required.
+ * Note that caller must hold inode->i_lock.
+ */
+static void
+pnfs_mark_layout_stateid_invalid(struct pnfs_layout_hdr *lo,
+                struct list_head *lseg_list)
+{
+        struct pnfs_layout_range range = {
+                .iomode = IOMODE_ANY,
+                .offset = 0,
+                .length = NFS4_MAX_UINT64,
+        };
+
+        set_bit(NFS_LAYOUT_INVALID_STID, &lo->plh_flags);
+        pnfs_mark_matching_lsegs_invalid(lo, lseg_list, &range);
+}
+
 static int
 pnfs_iomode_to_fail_bit(u32 iomode)
 {
@@ -554,9 +575,8 @@ pnfs_destroy_layout(struct nfs_inode *nfsi)
         spin_lock(&nfsi->vfs_inode.i_lock);
         lo = nfsi->layout;
         if (lo) {
-                lo->plh_block_lgets++; /* permanently block new LAYOUTGETs */
-                pnfs_mark_matching_lsegs_invalid(lo, &tmp_list, NULL);
                 pnfs_get_layout_hdr(lo);
+                pnfs_mark_layout_stateid_invalid(lo, &tmp_list);
                 pnfs_layout_clear_fail_bit(lo, NFS_LAYOUT_RO_FAILED);
                 pnfs_layout_clear_fail_bit(lo, NFS_LAYOUT_RW_FAILED);
                 spin_unlock(&nfsi->vfs_inode.i_lock);