Bug#681056: [mdadm] crash on growing RAID6 (4 to 5 devices) linux-image-3.2.0-4/3.2.32-1

[Dmitry Smirnov]

Package: mdadm
Version: 3.2.5-3

This may be related -- I attempted to grow RAID-6 array of 4 disks
by adding another HDD:

	sudo mdadm --grow /dev/md6 --backup-file /root/md6-critical.mdadm --raid-devices=5

(/root is not on the array that I tried to grow)
Immediately array has become inaccessible and the following appeared
in /var/log/messages:

Nov 23 11:07:59 debstor kernel: [64823.795075] md: reshape of RAID array md6
Nov 23 11:07:59 debstor kernel: [64823.795078] md: minimum _guaranteed_  speed: 16384 KB/sec/disk.
Nov 23 11:07:59 debstor kernel: [64823.795080] md: using maximum available idle IO bandwidth (but not more than 200000 KB/sec) for reshape.
Nov 23 11:07:59 debstor kernel: [64823.795087] md: using 128k window, over a total of 1953382912k.
Nov 23 11:07:59 debstor kernel: [64824.185395] md: md_do_sync() got signal ... exiting
Nov 23 11:11:02 debstor kernel: [65006.642584] md6_raid6       D ffff88045fd93780     0  6772      2 0x00000000
Nov 23 11:11:02 debstor kernel: [65006.642594]  ffff88044b84a280 0000000000000046 ffff880300000000 ffff88044f249750
Nov 23 11:11:02 debstor kernel: [65006.642606]  0000000000013780 ffff880439487fd8 ffff880439487fd8 ffff88044b84a280
Nov 23 11:11:02 debstor kernel: [65006.642617]  ffff8802aee9c180 00000001810709c1 0000000000000046 ffff88044bf0b400
Nov 23 11:11:02 debstor kernel: [65006.642627] Call Trace:
Nov 23 11:11:02 debstor kernel: [65006.642653]  [<ffffffffa016dc54>] ? check_reshape+0x27b/0x51a [raid456]
Nov 23 11:11:02 debstor kernel: [65006.642665]  [<ffffffff8103f411>] ? try_to_wake_up+0x197/0x197
Nov 23 11:11:02 debstor kernel: [65006.642685]  [<ffffffffa0157321>] ? md_check_recovery+0x2a5/0x514 [md_mod]
Nov 23 11:11:02 debstor kernel: [65006.642699]  [<ffffffffa017170a>] ? raid5d+0x1c/0x483 [raid456]
Nov 23 11:11:02 debstor kernel: [65006.642711]  [<ffffffff8134bc6e>] ? schedule_timeout+0x2c/0xdb
Nov 23 11:11:02 debstor kernel: [65006.642720]  [<ffffffff810709c1>] ? arch_local_irq_save+0x11/0x17
Nov 23 11:11:02 debstor kernel: [65006.642727]  [<ffffffff810709c1>] ? arch_local_irq_save+0x11/0x17
Nov 23 11:11:02 debstor kernel: [65006.642743]  [<ffffffffa0151223>] ? md_thread+0x101/0x11f [md_mod]
Nov 23 11:11:02 debstor kernel: [65006.642754]  [<ffffffff8105f7a7>] ? add_wait_queue+0x3c/0x3c
Nov 23 11:11:02 debstor kernel: [65006.642769]  [<ffffffffa0151122>] ? md_rdev_init+0xea/0xea [md_mod]
Nov 23 11:11:02 debstor kernel: [65006.642778]  [<ffffffff8105f155>] ? kthread+0x76/0x7e
Nov 23 11:11:02 debstor kernel: [65006.642787]  [<ffffffff81353c34>] ? kernel_thread_helper+0x4/0x10
Nov 23 11:11:02 debstor kernel: [65006.642796]  [<ffffffff8105f0df>] ? kthread_worker_fn+0x139/0x139
Nov 23 11:11:02 debstor kernel: [65006.642804]  [<ffffffff81353c30>] ? gs_change+0x13/0x13

Fortunately after reboot reshaping continued and finished successfully.

After reboot with linux-image-3.6-trunk-amd64/3.6.6-1~experimental.1
I was able to add another disk to the array and start reshaping without
any problems so perhaps this issue may be reassigned to 
linux-image-3.2.0-4-amd64/3.2.32-1

Also I'm pretty sure few times in the past I successfully grew/reshaped
RAID-6 arrays with older kernels but it's hard to tell when the problem
was introduced.

Thank you.

--- System information. ---
Architecture: amd64
Kernel:       Linux 3.2.0-4-amd64