Skip site navigation (1)Skip section navigation (2)
Date:      Thu, 01 May 2003 11:31:32 -0400 (EDT)
From:      John Baldwin <jhb@FreeBSD.org>
To:        Paul Richards <paul@freebsd-services.com>
Cc:        freebsd-arch@freebsd.org
Subject:   Re: lots of malloc(M_WAITOK)'s in interrupt context from camisr
Message-ID:  <XFMail.20030501113132.jhb@FreeBSD.org>
In-Reply-To: <20030501151409.GD1869@survey.codeburst.net>

next in thread | previous in thread | raw e-mail | index | archive | help

On 01-May-2003 Paul Richards wrote:
> On Thu, May 01, 2003 at 04:31:08PM +1000, Bruce Evans wrote:
>> On Wed, 30 Apr 2003, Andrew Gallatin wrote:
>> 
>> > John Baldwin writes:
>> >
>> >  > If you need to do more work in your interrupt routine than just wakeups
>> >  > and dinking with registers, you can always wake up a software interrupt
>> >  > handler or some other random kthread to do things that take a long amount
>> 
>> (This is about normal interrupt handlers, not INTR_FAST ones.)
>> 
>> > Dumb question: Exactly what is one allowed to do in an INTR_FAST
>> > interrupt context?  Obviously, you can't sleep.  But can you call
>> > wakeup()?
> 
> What exactly defines a INTR_FAST interrupt context in the first
> place. Do we have any rules for when it should be used, it just
> seems to me that all interrupt handlers should be INTR_FAST and
> that we'd then just have interrupt handlers.

Since INTR_FAST handlers can't lock a normal mutex, this would
require all device drivers to use spin locks (which either disable
or defer interrupts) thus greatly increasing latency.  INTR_FAST
are only suitable for things such as sio(4) which need very low
latency between interrupt assertion and handler execution and
cannot deal with that latency for normal interrupt handlers.

-- 

John Baldwin <jhb@FreeBSD.org>  <><  http://www.FreeBSD.org/~jhb/
"Power Users Use the Power to Serve!"  -  http://www.FreeBSD.org/



Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?XFMail.20030501113132.jhb>