On Sun, 31 Jan 1999, Stephen Costaras wrote:
>
> Now that I'm back on the list I wanted to bring this up. For
> the past while running linux 2.2.0pre6, & 2.2.0 release I've
> seen the following on one of my volumes:
>
> --------
> Jan 21 19:49:09 news kernel: raid5: bug: stripe->bh_new[6], sector 2643680
> exists
> Jan 21 19:49:10 news kernel: raid5: bh c27d7080, bh_new c36d5a40
> Jan 23 05:46:18 news kernel: raid5: bug: stripe->bh_new[5], sector 3216552
> exists
> Jan 23 05:46:18 news kernel: raid5: bh c922ec20, bh_new cb11d9c0
> Jan 23 08:18:00 news kernel: raid5: bug: stripe->bh_new[4], sector 3217776
> exists
> Jan 23 08:18:00 news kernel: raid5: bh c922ece0, bh_new ca9735a0
> Jan 25 17:46:20 news kernel: raid5: bug: stripe->bh_new[8], sector 6467032
> exists
> Jan 25 17:46:20 news kernel: raid5: bh c2f88a20, bh_new c123c480
> Jan 26 05:15:14 news kernel: raid5: bug: stripe->bh_new[4], sector 4462040
> exists
> Jan 26 05:15:14 news kernel: raid5: bh c0a3c300, bh_new c32b93e0
> Jan 26 17:46:01 news kernel: raid5: bug: stripe->bh_new[4], sector 3975968
> exists
> Jan 26 17:46:02 news kernel: raid5: bh c4289da0, bh_new c2d5d200
> Jan 26 20:45:18 news kernel: raid5: bug: stripe->bh_new[7], sector 5493120
> exists
> Jan 26 20:45:18 news kernel: raid5: bh c922e2c0, bh_new c8b370e0
> Jan 29 20:45:24 news kernel: raid5: bug: stripe->bh_new[1], sector 5573072
> exists
> Jan 29 20:45:24 news kernel: raid5: bh c2bfb7a0, bh_new ca244c60
> Jan 30 19:45:43 news kernel: raid5: bug: stripe->bh_new[1], sector 5599704
> exists
> Jan 30 19:45:43 news kernel: raid5: bh ccf60860, bh_new c9bdecc0
> Jan 30 22:45:37 news kernel: raid5: bug: stripe->bh_new[6], sector 4089216
> exists
> Jan 30 22:45:38 news kernel: raid5: bh c4da2e00, bh_new c4da2f20
> Jan 31 20:15:25 news kernel: raid5: bug: stripe->bh_new[7], sector 3607576
> exists
> Jan 31 20:15:25 news kernel: raid5: bh c96149c0, bh_new c3086420
> --------
>
> This is my raidtab file for this array:
>
> raiddev /dev/md0
> raid-level 5
> nr-raid-disks 9
> chunk-size 512
> parity-algorithm left-symmetric
> nr-spare-disks 0
> device /dev/sdb1
> raid-disk 0
> device /dev/sdc1
> raid-disk 1
> device /dev/sdd1
> raid-disk 2
> device /dev/sde1
> raid-disk 3
> device /dev/sdf1
> raid-disk 4
> device /dev/sdg1
> raid-disk 5
> device /dev/sdh1
> raid-disk 6
> device /dev/sdi1
> raid-disk 7
> device /dev/sdj1
> raid-disk 8
> --------
>
> Can anyone here shed any light on this? Is this something that I need to
> worry about? Am I corrupting data?
>
> Steve
Probably not. That message means that we detected a problem which would
cause a crash on older kernels, but worked around it (two requests
for the same disk sector were issued simultaneously from two different
buffers).
We still print the message, as it might point to a bug in the higher
levels of the kernel rather than the RAID-5 driver (the work-around
services both requests, where in older versions we serviced just one
of them).
Gadi
