Skip site navigation (1)Skip section navigation (2)
Date:      Tue, 6 Mar 2001 22:26:06 +1100 (EST)
From:      Bruce Evans <bde@zeta.org.au>
To:        "Peter S. Housel" <housel@acm.org>
Cc:        John Baldwin <jhb@FreeBSD.org>, cvs-all@FreeBSD.org, cvs-committers@FreeBSD.org
Subject:   Re: cvs commit: src/sys/i386/isa apic_vector.s icu_vector.s
Message-ID:  <Pine.BSF.4.21.0103062214320.10533-100000@besplex.bde.org>
In-Reply-To: <mu97l231sgt.wl@cx281057-a.irvn1.occa.home.com>

next in thread | previous in thread | raw e-mail | index | archive | help
On Mon, 5 Mar 2001, Peter S. Housel wrote:

> At Sun, 04 Mar 2001 20:41:59 -0800 (PST), John Baldwin wrote:
> > This fixed some hard hangs that David was seeing on his machine under heavy
> > load.  It might also fix (or at lesat greatly diminish the number of
> > occurences) the "ltr" panics involving a kernel trap 9 with interrupts disabled.
> > By leaving interrupts disabled in sched_ithd, we don't allow ourselves to be
> > preempted by interrupt B while we are scheduling interrupt A.  Otherwise, we
> > could become too deeply nested in our stack and overwrite various parts of the
> > pcb.  (Some of which probably don't belong in the pcb.)
> 
> Before this change, my Vaio C1VN would panic (trap 9 at the ltr
> instruction) after inserting a PCMCIA card.  If the card was inserted
> during boot, it would panic just after the "pccard: card inserted,
> slot 0" message.  Now it locks up solid in both cases.  (No panic, and
> the DDB key doesn't work.)

This is probably caused by the interrupt not going away for some reason.
This previously would have caused the interrupt handler to keep reentering
itself until the stack gets corrupted enough to cause the ltr panic or
another bad thing.  Now it would cause the system to spin getting unnested
interrupts.  The DDB key would still work if the keyboard interrupt has
higher priority than the card interrupt.

Bruce


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




Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?Pine.BSF.4.21.0103062214320.10533-100000>