Re: SCSI lockups in 2.1.101

Doug Ledford (dledford@dialnet.net)
Tue, 19 May 1998 18:16:44 -0500

David Woodhouse wrote:
>
> When trying to use a SCSI scanner under 2.1.101, connected to an Adaptec
> AHA2940UW contrioller, I've twice got a complete lockup.
>
> No helpful messages or anything. Alt-SysRq-P shows EIP values of c010827f,
> which is just after the 'hlt' in cpu_idle:
>
> 0xc010827e <cpu_idle+62>: hlt
> 0xc010827f <cpu_idle+63>: cmpl $0x0,0xc0213578
> 0xc0108286 <cpu_idle+70>: je 0xc0108342 <cpu_idle+258>
>
> It's v5.0.13 of the aic7xxx driver, IIRC. Is this likely to have been fixed in
> 5.0.15?

Doubtful that 5.0.15 will help this. This looks reminiscient of a spin_lock
mis-use/abuse that's locking the machine up. Try 2.1.102 instead as it may
be related to the need to send an abort for a lost command.

-- 


 Doug Ledford  <dledford@dialnet.net>
  Opinions expressed are my own, but
     they should be everybody's.


-
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@vger.rutgers.edu