linux-raid.vger.kernel.org archive mirror
 help / color / mirror / Atom feed
* smal mdadm 1.5  bug
@ 2006-08-21 12:25 jwillem
  2006-08-22  6:33 ` Neil Brown
  0 siblings, 1 reply; 2+ messages in thread
From: jwillem @ 2006-08-21 12:25 UTC (permalink / raw)
  To: linux-raid




We have had a raidsystem running for several years 
Created by the Fedora Core 2 installscript. 

Is has bit bigger raidsystem of more than 11 TB
It was installed on 32 bit Fedora Core 2  linux 2.6.10 with mdadm 1.50

After a failure we had one disk that was out sync. 
Couldn't assemble raidset, even with --force -> "Cannot write superblock"

After looking at the sourcecode in the file util.c  from mdadm-1.5.0 we
found out that variable "size" was negative
We changed variable size from "long size" into "unsigned long size" wich
fixed this problem.

Seems that 1.9 already fixed the problem, perhaps even earlyer
But also that 
> Changes Prior to 1.5.0 release
>     -   Fix compiling error with BLKGETSIZE64 and some signed/unsigned 
> comparison warnings.

This was the version we were working on, so something went wrong there. 

Greetings Jan-Willem Michels ( Boy Kentrop working at our company fixed
our problem) 

^ permalink raw reply	[flat|nested] 2+ messages in thread

* Re: smal mdadm 1.5  bug
  2006-08-21 12:25 smal mdadm 1.5 bug jwillem
@ 2006-08-22  6:33 ` Neil Brown
  0 siblings, 0 replies; 2+ messages in thread
From: Neil Brown @ 2006-08-22  6:33 UTC (permalink / raw)
  To: jwillem; +Cc: linux-raid

On Monday August 21, jwillem@stads.net wrote:
> 
> 
> 
> We have had a raidsystem running for several years 
> Created by the Fedora Core 2 installscript. 
> 
> Is has bit bigger raidsystem of more than 11 TB
> It was installed on 32 bit Fedora Core 2  linux 2.6.10 with mdadm 1.50
> 
> After a failure we had one disk that was out sync. 
> Couldn't assemble raidset, even with --force -> "Cannot write superblock"
> 
> After looking at the sourcecode in the file util.c  from mdadm-1.5.0 we
> found out that variable "size" was negative
> We changed variable size from "long size" into "unsigned long size" wich
> fixed this problem.
> 
> Seems that 1.9 already fixed the problem, perhaps even earlyer
> But also that 
> > Changes Prior to 1.5.0 release
> >     -   Fix compiling error with BLKGETSIZE64 and some signed/unsigned 
> > comparison warnings.
> 
> This was the version we were working on, so something went wrong there. 
> 
> Greetings Jan-Willem Michels ( Boy Kentrop working at our company fixed
> our problem) 

I'm a little confused..

Are you just telling us that you found a bug in an old version of
mdadm that was fixed in a newer version of mdadm?  Or are you telling
us that there is still a bug in a current version of mdadm?

If so, which version?
mdadm-1.5.0 is very old, and even mdadm-1.9.0 is fairly old.

NeilBrown

^ permalink raw reply	[flat|nested] 2+ messages in thread

end of thread, other threads:[~2006-08-22  6:33 UTC | newest]

Thread overview: 2+ messages (download: mbox.gz follow: Atom feed
-- links below jump to the message on this page --
2006-08-21 12:25 smal mdadm 1.5 bug jwillem
2006-08-22  6:33 ` Neil Brown

This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).