Skip site navigation (1)Skip section navigation (2)
Date:      Wed, 26 Jul 2006 14:43:02 -0400
From:      John Baldwin <jhb@freebsd.org>
To:        freebsd-mobile@freebsd.org
Cc:        "Douglas W. Goodall" <douglas_goodall@mac.com>
Subject:   Re: Stray irq7's
Message-ID:  <200607261443.03128.jhb@freebsd.org>
In-Reply-To: <000001c6ae9a$cce42c20$6dce46c0@dougwide>
References:  <000001c6ae9a$cce42c20$6dce46c0@dougwide>

next in thread | previous in thread | raw e-mail | index | archive | help
On Sunday 23 July 2006 16:58, Douglas W. Goodall wrote:
> I looked closely at the source code i386/intr_machdep.c and I can see that
> the interrupt routine deals properly with the 
> Default IR7. The only trouble is, the code considers the stray Ir7 a
> condition worth informing the user about and logging.
> 
> IMHO, it would be better to just take care of it quietly. It isn't an
> important event and isn't worthy of bothering
> the root user. I commented out the logging code and everything works fine.

See my other e-mail.  The IRR check is done in the atpic-specific code in 
i386/isa/atpic.c:atpic_handle_intr() that calls the intr_execute_handlers() 
function you are looking at.  My guess is that he's getting noise from the 
printer that triggers real IRQ7 interrupts, but since the interrupt line is 
masked, those interrupts don't trigger.  Later, a spurious interrupt results 
in an unmasked IRQ7, but since IRQ7 is marked as pending due to the earlier 
noise interrupt from the printer, the IRR check fails and it ends up in 
intr_execute_handlers().  If that is the case, either disconnecting the 
printer or getting the ppc(4) driver working will fix the messages.  Either 
way it's probably harmless and not something the OP needs to worry about.

-- 
John Baldwin



Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?200607261443.03128.jhb>