From owner-aic7xxx  Thu Apr  2 19:25:59 1998
Return-Path: <owner-aic7xxx@FreeBSD.ORG>
Received: (from majordom@localhost)
          by hub.freebsd.org (8.8.8/8.8.8) id TAA07796
          for aic7xxx-outgoing; Thu, 2 Apr 1998 19:25:59 -0800 (PST)
          (envelope-from owner-aic7xxx@FreeBSD.ORG)
Received: from dledford.dialnet.net (root@dledford.dialnet.net [206.65.249.116])
          by hub.freebsd.org (8.8.8/8.8.8) with ESMTP id TAA07785
          for <aic7xxx@FreeBSD.ORG>; Thu, 2 Apr 1998 19:25:55 -0800 (PST)
          (envelope-from dledford@dialnet.net)
Received: from dialnet.net (localhost [127.0.0.1])
          by dledford.dialnet.net (8.8.5/8.8.4) with ESMTP
	  id VAA09597; Thu, 2 Apr 1998 21:25:38 -0600
Message-ID: <352456B2.6FE89D82@dialnet.net>
Date: Thu, 02 Apr 1998 21:25:38 -0600
From: Doug Ledford <dledford@dialnet.net>
X-Mailer: Mozilla 4.04 [en] (X11; I; Linux 2.1.92 i686)
MIME-Version: 1.0
To: Attila Karpati <karpati@cs.elte.hu>
CC: aic7xxx@FreeBSD.ORG, linux-smp@vger.rutgers.edu
Subject: Re: 2.1.91 SMP and aic7xxx
References: <199804021927.VAA01079@neumann.cs.elte.hu>
Content-Type: text/plain; charset=us-ascii
Content-Transfer-Encoding: 7bit
Sender: owner-aic7xxx@FreeBSD.ORG
Precedence: bulk

Attila Karpati wrote:
> 
> Hi,
> 
> Robert HYATT <hyatt@cis.uab.edu> wrote:
> > 2.  try 2.1.92, plus Ingo's patch sent to this list this morning
> > some time.
> 
> I did so and tried the newest aic driver (5.0.10). I got the following kernel
> messages:
> 
> (scsi0) Wide channel, SCSI ID 7, 16/255 SCBs
> (scsi0) BIOS enabled, IO port 0xec00, IRQ 17
> (scsi0) IO Memory at 0xfebff000, MMAP Memory at 0x00000000
> (scsi0) Resetting channel
> (scsi0) Dowloading sequencer code... 413 instructions downloaded
> scsi0: Adaptec AHA274x/284x/294x (EISA/VLB/PCI-Fast SCSI) 5.0.10/3.2.2
>        <Adaptec AIC-7880 Ultra SCSI host adapter>
> scsi: 1 host.
> (scsi0:0:-1:-1) Scanning channel for devices.
> (scsi0:-1:-1:-1) Allocating initial 30 SCB structures
> 
> Here stopped the kernel.
> 
> The place where the hang occurs in the aic7xxx.c source file (i think):
> 
> <Line 8008>
>     scb->flags |= SCB_ACTIVE | SCB_WAITINGQ;
> 
>     save_flags(processor_flags);
>     cli();
>     if (p->delayed_scbs[tindex].head != NULL)
>     {
>         scbq_insert_tail(&p->delayed_scbs[tindex], scb);
>     }
>     else
>     {
>         scbq_insert_tail(&p->waiting_scbs, scb);
>     }
>     if ( (p->flags & (AHC_IN_ISR | AHC_IN_ABORT | AHC_IN_RESET)) == 0)
>     {
>       aic7xxx_run_waiting_queues(p);
>     }
>     restore_flags(processor_flags);
> 
> Can someone correct this with a patch?

It's very unlikely that the code is hanging in the spot you mentioned.  More
likely is that we are sending the command to the sequencer and then never
getting the interrupts we should get from the REQINIT phase of negotiation
or from SELTO interrupts.  Hmmm....maybe we could talk Ingo onto releasing a
patch that would force the IO-APIC code to disabled for testing purposes to
see if the IRQ->IO-APIC mapping on your motherboard has anything to do with
it.

> The machine has 4 PCI slots full with 3 ethernet (tulip) cards and a VGA card.
> The controller is an onboard one. Do you think it would help to rearrange the

Possibly, try to get the VGA card and the Adaptec on the same IRQ, and all
of the Tulip cards on different IRQs (that is, if your BIOS prints out the
nice message at bootup to tell you what IRQs it assiged, otherwise you may
not know).

> 
> cards in this situation? How did you rearrange your cards?
> 
> Attila
> 
> To Unsubscribe: send mail to majordomo@FreeBSD.org
> with "unsubscribe aic7xxx" in the body of the message

-- 

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

To Unsubscribe: send mail to majordomo@FreeBSD.org
with "unsubscribe aic7xxx" in the body of the message