From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S1752657AbaJ0M3s (ORCPT ); Mon, 27 Oct 2014 08:29:48 -0400 Received: from eddie.linux-mips.org ([148.251.95.138]:50603 "EHLO cvs.linux-mips.org" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1751766AbaJ0M3r (ORCPT ); Mon, 27 Oct 2014 08:29:47 -0400 Date: Mon, 27 Oct 2014 12:29:44 +0000 (GMT) From: "Maciej W. Rozycki" To: Ingo Molnar cc: Thomas Gleixner , Ingo Molnar , "H. Peter Anvin" , x86@kernel.org, linux-kernel@vger.kernel.org Subject: Re: [PATCH] x86: Fix XT-PIC-XT-PIC in /proc/interrupts In-Reply-To: <20141027074748.GA18025@gmail.com> Message-ID: References: <20141027074748.GA18025@gmail.com> User-Agent: Alpine 2.11 (LFD 23 2013-08-11) MIME-Version: 1.0 Content-Type: TEXT/PLAIN; charset=US-ASCII Sender: linux-kernel-owner@vger.kernel.org List-ID: X-Mailing-List: linux-kernel@vger.kernel.org On Mon, 27 Oct 2014, Ingo Molnar wrote: > What's the effect of this change on the output for non-8259A irq > controllers? Good question, I didn't expect any other IRQ controller to be used for ISA interrupts (the APIC is handled elsewhere). I have checked the sources and the only other controller that can be used for `legacy_pic' is `dummy_irq_chip' (cf. `default_legacy_pic' vs `null_legacy_pic' in arch/x86/kernel/i8259.c). That affects `init_ISA_irqs' only (the other place, `make_8259A_irq', has a reference to `i8259A_chip' hardcoded). In this case the output would change from "dummy-dummy" to "dummy", which IMHO has just as much value as the change from "XT-PIC-XT-PIC" to "XT-PIC". However I don't think you'd be able to request such IRQs so they will never show up in /proc/interrupts, making this observation largely irrelevant. If you think I may be wrong here, then can you please find such a system and try this change with it (or can someone else reading this make this check for me by any chance)? Unfortunately all x86 hardware I have does have an 8259A pair. Also please note that both places explicitly refer to the name of the IRQ chip being installed as the name of the handler as well, with `i8259A_chip.name' and `chip->name' respectively, this is hardcoded. So no matter which chip is used there'll be a duplicate name in /proc/interrupts. Does this answer address your concern? Maciej