Re: kern/38376: Hitachi HDD need LBA48-QUIRK

[Martin Seibold <martin.seibold%yahoo.de@localhost>]

The following reply was made to PR kern/38376; it has been noted by GNATS.

From: Martin Seibold <martin.seibold%yahoo.de@localhost>
To: gnats-bugs%NetBSD.org@localhost
Cc: 
Subject: Re: kern/38376: Hitachi HDD need LBA48-QUIRK
Date: Sun, 13 Apr 2008 12:54:11 +0200 (CEST)

 Hi David,
 
 I don't know whether the problem is caused by
 multi-transfairs or not. I only found out that it
 isn't possible to rewrite raidframe parity across
 LBA48 boundary without setting proper treshhold when
 using Hitachi drives.
 
 --- David Laight <david%l8s.co.uk@localhost> schrieb:
 
 >  > >Synopsis:       Hitachi HDD need LBA48-QUIRK
 >  > 
 >  > Cave: Maybe there is another Problem in raidframe
 >  > involved too, since changing the treshhold to 
 >  > 0xfffffe0 didn't work.
 >  > _It had to be 0xfffffd0 when using raidframe_.
 >  
 >  Is the actual problem that multi-sector transfers
 >  that cross the LBA48  boundary fail?
 >  In which case shouldn't the code just look at the
 > 'sector+count' when deciding whether to use LBA48 ?
 
 I am only administrator for NetBSD, i am not a kernel
 developer. Please bear in mind that the disks are
 already in productive use since it's working with my
 custom kernel. I would like to help you, if there is
 anything i can test without manually failing one of my
 disks, changing my kernel or loosing some my data
 don't hesitate to ask.
 
 Have a nice day,
 Martin, Austria
 
 Maybe the log from trying rewriting parity with
 treshhold 0xfffffe0 can help you a bit:
 
 Apr  1 14:24:30 netbsdfs /netbsd: raid5: RAID Level 5
 Apr  1 14:24:30 netbsdfs /netbsd: raid5: Components:
 /dev/wd0a /dev/wd1a[**FAILED**] /dev/wd2a
 
 Apr  1 14:24:30 netbsdfs /netbsd: raid5: Total
 Sectors: 1953546176 (953879 MB)
 Apr  1 14:26:18 netbsdfs /netbsd: dk_lookup on device:
 wd1a failed!
 Apr  1 14:26:29 netbsdfs /netbsd: dk_lookup on device:
 /dev/wd1a failed!
 Apr  1 14:27:15 netbsdfs /netbsd: raid5: initiating
 in-place reconstruction on column 1
 Apr  1 15:42:43 netbsdfs /netbsd: wd0a: error reading
 fsbn 268435424 of 268435424-268435455 (wd0 bn
 268435424; cn 266304 tn 15 sn 47), retrying
 Apr  1 15:42:43 netbsdfs /netbsd: wd0: (aborted
 command)
 Apr  1 15:42:43 netbsdfs /netbsd: wd2a: error reading
 fsbn 268435424 of 268435424-268435455 (wd2 bn
 268435424; cn 266304 tn 15 sn 47), retrying
 Apr  1 15:42:43 netbsdfs /netbsd: wd2: (aborted
 command)
 Apr  1 15:42:44 netbsdfs /netbsd: wd0a: error reading
 fsbn 268435424 of 268435424-268435455 (wd0 bn
 268435424; cn 266304 tn 15 sn 47), retrying
 Apr  1 15:42:44 netbsdfs /netbsd: wd0: (aborted
 command)
 Apr  1 15:42:44 netbsdfs /netbsd: wd2a: error reading
 fsbn 268435424 of 268435424-268435455 (wd2 bn
 268435424; cn 266304 tn 15 sn 47), retrying
 Apr  1 15:42:44 netbsdfs /netbsd: wd2: (aborted
 command)
 Apr  1 15:42:44 netbsdfs /netbsd: wd0a: error reading
 fsbn 268435424 of 268435424-268435455 (wd0 bn
 268435424; cn 266304 tn 15 sn 47), retrying
 Apr  1 15:42:44 netbsdfs /netbsd: wd0: (aborted
 command)
 Apr  1 15:42:44 netbsdfs /netbsd: wd2a: error reading
 fsbn 268435424 of 268435424-268435455 (wd2 bn
 268435424; cn 266304 tn 15 sn 47), retrying
 Apr  1 15:42:44 netbsdfs /netbsd: wd2: (aborted
 command)
 Apr  1 15:42:45 netbsdfs /netbsd: wd0a: error reading
 fsbn 268435424 of 268435424-268435455 (wd0 bn
 268435424; cn 266304 tn 15 sn 47), retrying
 Apr  1 15:42:45 netbsdfs /netbsd: wd0: (aborted
 command)
 Apr  1 15:42:45 netbsdfs /netbsd: wd2a: error reading
 fsbn 268435424 of 268435424-268435455 (wd2 bn
 268435424; cn 266304 tn 15 sn 47), retrying
 Apr  1 15:42:45 netbsdfs /netbsd: wd2: (aborted
 command)
 Apr  1 15:42:45 netbsdfs /netbsd: wd0a: error reading
 fsbn 268435455 of 268435424-268435455 (wd0 bn
 268435455; cn 266305 tn 0 sn 15), retrying
 Apr  1 15:42:45 netbsdfs /netbsd: wd0: (aborted
 command)
 Apr  1 15:42:45 netbsdfs /netbsd: wd2a: error reading
 fsbn 268435455 of 268435424-268435455 (wd2 bn
 268435455; cn 266305 tn 0 sn 15), retrying
 Apr  1 15:42:45 netbsdfs /netbsd: wd2: (aborted
 command)
 Apr  1 15:42:46 netbsdfs /netbsd: wd0a: error reading
 fsbn 268435455 of 268435424-268435455 (wd0 bn
 268435455; cn 266305 tn 0 sn 15)
 Apr  1 15:42:46 netbsdfs /netbsd: wd0: (aborted
 command)
 Apr  1 15:42:46 netbsdfs /netbsd: raid5: Recon read
 failed!
 Apr  1 15:42:46 netbsdfs /netbsd: wd2a: error reading
 fsbn 268435455 of 268435424-268435455 (wd2 bn
 268435455; cn 266305 tn 0 sn 15)
 Apr  1 15:42:46 netbsdfs /netbsd: wd2: (aborted
 command)
 Apr  1 15:42:46 netbsdfs /netbsd: raid5: Recon read
 failed!
 Apr  1 15:42:46 netbsdfs /netbsd: raid5:
 reconstruction failed.
 
 
 
       __________________________________________________________
 Gesendet von Yahoo! Mail.
 Der Lieblings-Mailbox der Welt.
 http://de.overview.mail.yahoo.com