RE: [PATCH] fix PCI interrupt setting for ia64

From: Liu, Benjamin <>
Date: 2004-03-08 18:44:16

Both ISA and PCI device drivers would call arch/ia64/kernel/irq.c:request_irq()-->arch/ia64/kernel/irq.c:setup_irq() -->arch/ia64/kernel/iosapic.c:iosapic_startup_level_irq() or  arch/ia64/kernel/iosapic.c:iosapic_startup_edge_irq() function to unmask the IRQ. I believe the ISA can be handled gracefully, if any.

ISA is legacy to IA64. The configuration script of 2.4.23 has CONFIG_ISA off explicitly for IA64, 2.6.2 doesn't have this option for IA64. I just wonder whether the legacy probing method still exists on IA64.

Pingping (Benjamin) Liu
Intel China Software Center

>-----Original Message-----
>[] On Behalf Of Grant Grundler
>Sent: 200438 14:31
>To: Kenji Kaneshige
>Subject: Re: [PATCH] fix PCI interrupt setting for ia64
>On Mon, Mar 08, 2004 at 11:49:10AM +0900, Kenji Kaneshige wrote:
>> In ia64 kernel, IOSAPIC's RTEs for PCI interrupts are unmasked at the
>> boot time before installing device drivers. I think it is 
>very dangerous.
>Hi Kenji,
>I think this behavior exists to support "legacy" IRQ probing.
>I'm wondering if it would be sufficient to wrap the patch in
>"#ifndef CONFIG_ISA" or something like that.
>To unsubscribe from this list: send the line "unsubscribe 
>linux-ia64" in
>the body of a message to
>More majordomo info at
To unsubscribe from this list: send the line "unsubscribe linux-ia64" in
the body of a message to
More majordomo info at
Received on Mon Mar 8 02:44:40 2004

This archive was generated by hypermail 2.1.8 : 2005-08-02 09:20:24 EST