Re: [PATCH] qla2xxx: fix bad locking during eh_abort

[Andrew Vasquez <andrew.vasquez@qlogic.com>]

On Fri, 27 May 2005, James Bottomley wrote:

> On Thu, 2005-05-26 at 16:19 -0700, Andrew Vasquez wrote:
> > Please apply, this should go in before 2.6.12 is released.
> 
> I've got it in the rc-fixes tree for scsi.  However:
> 
> > +	spin_unlock(&ha->hardware_lock);
> 
> Should be spin_unlock_irq(&ha->hardware_lock); shouldn't it?  Otherwise
> we could sleep with interrupts disabled and the kernel now squeaks about
> that.

Yes, with the latest changes being proposed/implemented by Jeff G.,
there would need to be some additional massaging of the driver's
eh_*() routines.

Are you planning on putting the host_lock-free changes in for -rc,
I figured that would be going into the next kernel rev.  I mention
that because I have a block of patches which I have queued-up to add
new chip support to the driver.

Jeff, could you drop the qla2xxx driver from your scrubing
host_lock-free changes.  I'll go ahead and post an updated patch
following the patches in my queue.

Thanks,
Andrew Vasquez