From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S933857AbYDWB1U (ORCPT ); Tue, 22 Apr 2008 21:27:20 -0400 Received: (majordomo@vger.kernel.org) by vger.kernel.org id S1755611AbYDWB1F (ORCPT ); Tue, 22 Apr 2008 21:27:05 -0400 Received: from smtp1.linux-foundation.org ([140.211.169.13]:39347 "EHLO smtp1.linux-foundation.org" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1752315AbYDWB1F (ORCPT ); Tue, 22 Apr 2008 21:27:05 -0400 Date: Tue, 22 Apr 2008 18:26:55 -0700 (PDT) From: Linus Torvalds To: Jeff Chua cc: Ivan Kokshaysky , Jesse Barnes , Greg KH , Linux Kernel Mailing List Subject: Re: [Broken] PCI: clean up resource alignment management In-Reply-To: Message-ID: References: <20080422214050.GA1166@jurassic.park.msu.ru> User-Agent: Alpine 1.10 (LFD 962 2008-03-14) 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 Wed, 23 Apr 2008, Jeff Chua wrote: > On Wed, Apr 23, 2008 at 5:59 AM, Linus Torvalds > wrote: > > On Wed, 23 Apr 2008, Ivan Kokshaysky wrote: > > > Yes, exactly. My fault - I somehow missed the cardbus stuff... > > Ok, this patch looks sane, but .. > > Patch entirely UNTESTED! > > I just tested Linus's patch and it works. Thanks, and I could test it myself (without any actual card, but at least I could see the failure to even set up the bridge, and the fix). So I committed it. When we did the original commit that caused this, we had considered having a separate "alignment" value, but I had discarded it because I didn't think there was any actual hardware that could even use it. But this cardbus thing shows that I was wrong - the two bits may have been clever, and it works no worse than the old setup (and slightly better), but I think the separate alignment field would probably have been better. Anyway, it's probably not worth worrying about now. It's not like we've ever _needed_ the finer-granularity alignment, so I think we're ok with the current setup, but if we ever decide to add the alignment field after all, somebody should remind me/Ink about the cardbus thing. Linus