From owner-freebsd-hardware  Tue May 26 19:56:53 1998
Return-Path: <owner-freebsd-hardware@FreeBSD.ORG>
Received: (from majordom@localhost)
          by hub.freebsd.org (8.8.8/8.8.8) id TAA25270
          for freebsd-hardware-outgoing; Tue, 26 May 1998 19:56:53 -0700 (PDT)
          (envelope-from owner-freebsd-hardware@FreeBSD.ORG)
Received: from godzilla.zeta.org.au (godzilla.zeta.org.au [203.15.68.22])
          by hub.freebsd.org (8.8.8/8.8.8) with ESMTP id TAA25238
          for <freebsd-hardware@FreeBSD.ORG>; Tue, 26 May 1998 19:56:38 -0700 (PDT)
          (envelope-from bde@godzilla.zeta.org.au)
Received: (from bde@localhost)
	by godzilla.zeta.org.au (8.8.7/8.8.7) id MAA18228;
	Wed, 27 May 1998 12:56:33 +1000
Date: Wed, 27 May 1998 12:56:33 +1000
From: Bruce Evans <bde@zeta.org.au>
Message-Id: <199805270256.MAA18228@godzilla.zeta.org.au>
To: freebsd-hardware@FreeBSD.ORG, jose@we.lc.ehu.es
Subject: Re: problems with Iwill PIILS motherboard
Sender: owner-freebsd-hardware@FreeBSD.ORG
Precedence: bulk
X-Loop: FreeBSD.org

>First, the serial ports are not detected. I know that this problem
>also happens with other Iwill MBs and some Compaqs. I have found
>that two tests made by the sio driver fail (tests numbered 5 and 8).
>If I disable both tests, the serial ports are detected and work
>fine. However, I would like to know if someone is working on
>a patch for this problem. If there isn't any known solution, a
>new kernel config option (SIO_PROBLEM, for example) would be
>very useful.

Ignoring the results of the failing tests is the best known workaround.
Disabling individual tests in sioprobe is wrong in general because some of
the tests has side effects, but this isn't a problem for tests 5 and 8.
Test 3 may also fail if the system is rebooted in a certain state (e.g.,
the state for tests 5 and 8).

After the probe fails on IWill motherboards, it is normal for leftover
interrupts to cause warnings like "sio0: 64 events for device with no tp".
These just mean that the hardware is weird - interrupts didn't go away
when they were programmed to.  No fix is known.  One theory is that the
problem is related to the 'serialized IRQ protocol' that the ACER UART
claims to support.  It seems to have more to do with IRQs than UARTs.
The sio probe is probably the only thing in FreeBSD that is affected
because it is the only thing that uses old arcane features of the
interrupt controller.

Bruce

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