From mboxrd@z Thu Jan 1 00:00:00 1970 From: Ingo Molnar Subject: Re: [PATCH 1/1] Speedfreq-SMI call clobbers ECX Date: Tue, 11 Mar 2008 10:39:07 +0100 Message-ID: <20080311093903.GS25110@elte.hu> References: <200803051559.09962.langer_mann@web.de> <200803101605.42251.langer_mann@web.de> <87iqzu8r2q.fsf@basil.nowhere.org> <200803102226.39044.langer_mann@web.de> Mime-Version: 1.0 Return-path: Content-Disposition: inline In-Reply-To: <200803102226.39044.langer_mann@web.de> Sender: linux-kernel-owner@vger.kernel.org List-ID: Content-Type: text/plain; charset="us-ascii" Content-Transfer-Encoding: 7bit To: Stephan Diestelhorst Cc: Andi Kleen , davej@codemonkey.org.uk, cpufreq@lists.linux.org.uk, linux-kernel@vger.kernel.org * Stephan Diestelhorst wrote: > Please also note that these are not clobbers in the strict inline asm > syntax, but rather dummy output values that correspond to actual input > parameters. I'd consider this a serious compiler flaw, if not bug, if > these would not work. But let's not get into GCC vs. > Fancy-inline-asm-hacker flames, like the mplayer folks do ;-) yep, that's my experience as well - output constraints are pretty robust in general, but if one tries the same effect via the clobber list it's easy to run into gcc internal errors. I've applied your patch to x86.git. Do you think it's 2.6.25 material? I'm inclined to have it in the 2.6.26 bucket. We could do your first, minimal fix in 2.6.25 and do this second, full fix in 2.6.26. Ingo