md/raid10: fix the 'new' raid10 layout to work correctly.

In Linux 3.9 we introduce a new 'far' layout for RAID10 which was
supposed to rotate the replicas differently and so provide better
resilience.  In particular it could survive more combinations of 2
drive failures.

Unfortunately. due to a coding error, this some did what was wanted,
sometimes improved less than we hoped, and sometimes - in very
unlikely circumstances - put multiple replicas on the same device so
the redundancy was harmed.

No public user-space tool has created arrays using this layout so it
is very unlikely that zero-redundancy arrays actually exist.  Probably
no arrays using any form of the new layout exist.  But we cannot be
certain.

So use another bit in the 'layout' number and introduce a bug-fixed
version of the layout.
Also when assembling an array, if it has a zero-redundancy layout,
give a warning.

Reported-by: Heinz Mauelshagen <heinzm@redhat.com>
Signed-off-by: NeilBrown <neilb@suse.com>

diff --git a/drivers/md/raid10.c b/drivers/md/raid10.c
index 23de2144ee13e90661f1d9fca919eb2e4f312b30..96f36596830696c2f1ba1bd8bc6fbb24d30a43d8 100644 (file)
--- a/drivers/md/raid10.c
+++ b/drivers/md/raid10.c
@@ -39,6 +39,7 @@
  *    far_copies (stored in second byte of layout)
  *    far_offset (stored in bit 16 of layout )
  *    use_far_sets (stored in bit 17 of layout )
+ *    use_far_sets_bugfixed (stored in bit 18 of layout )
  *
  * The data to be stored is divided into chunks using chunksize.  Each device
  * is divided into far_copies sections.   In each section, chunks are laid out
@@ -1497,6 +1498,8 @@ static void status(struct seq_file *seq, struct mddev *mddev)
                        seq_printf(seq, " %d offset-copies", conf->geo.far_copies);
                else
                        seq_printf(seq, " %d far-copies", conf->geo.far_copies);
+               if (conf->geo.far_set_size != conf->geo.raid_disks)
+                       seq_printf(seq, " %d devices per set", conf->geo.far_set_size);
        }
        seq_printf(seq, " [%d/%d] [", conf->geo.raid_disks,
                                        conf->geo.raid_disks - mddev->degraded);
@@ -3394,7 +3397,7 @@ static int setup_geo(struct geom *geo, struct mddev *mddev, enum geo_type new)
                disks = mddev->raid_disks + mddev->delta_disks;
                break;
        }
-       if (layout >> 18)
+       if (layout >> 19)
                return -1;
        if (chunk < (PAGE_SIZE >> 9) ||
            !is_power_of_2(chunk))
@@ -3406,7 +3409,22 @@ static int setup_geo(struct geom *geo, struct mddev *mddev, enum geo_type new)
        geo->near_copies = nc;
        geo->far_copies = fc;
        geo->far_offset = fo;
-       geo->far_set_size = (layout & (1<<17)) ? disks / fc : disks;
+       switch (layout >> 17) {
+       case 0: /* original layout.  simple but not always optimal */
+               geo->far_set_size = disks;
+               break;
+       case 1: /* "improved" layout which was buggy.  Hopefully no-one is
+                * actually using this, but leave code here just in case.*/
+               geo->far_set_size = disks/fc;
+               WARN(geo->far_set_size < fc,
+                    "This RAID10 layout does not provide data safety - please backup and create new array\n");
+               break;
+       case 2: /* "improved" layout fixed to match documentation */
+               geo->far_set_size = fc * nc;
+               break;
+       default: /* Not a valid layout */
+               return -1;
+       }
        geo->chunk_mask = chunk - 1;
        geo->chunk_shift = ffz(~chunk);
        return nc*fc;