From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S1752359AbcCVWDq (ORCPT ); Tue, 22 Mar 2016 18:03:46 -0400 Received: from eddie.linux-mips.org ([148.251.95.138]:56768 "EHLO cvs.linux-mips.org" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1751898AbcCVWDo (ORCPT ); Tue, 22 Mar 2016 18:03:44 -0400 Date: Tue, 22 Mar 2016 22:03:42 +0000 (GMT) From: "Maciej W. Rozycki" To: Borislav Petkov cc: Linus Torvalds , Ingo Molnar , "Bryan O'Donoghue" , Andy Lutomirski , Andy Shevchenko , "linux-kernel@vger.kernel.org" , "x86@kernel.org" , Fenghua Yu , "H. Peter Anvin" , Thomas Gleixner , Andrew Morton , Dave Hansen , Oleg Nesterov , "Yu, Yu-cheng" Subject: Re: [PATCH] x86/FPU: Fix FPU handling on legacy FPU machines In-Reply-To: <20160311220313.GG4312@pd.tnic> Message-ID: References: <20160310125610.GA26708@pd.tnic> <20160310145940.GB26708@pd.tnic> <20160311090840.GA8486@gmail.com> <20160311094802.GA4312@pd.tnic> <1457694124.2007.12.camel@nexus-software.ie> <20160311112610.GC4312@pd.tnic> <20160311113206.GD4312@pd.tnic> <20160311220313.GG4312@pd.tnic> User-Agent: Alpine 2.20 (LFD 67 2015-01-07) 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 Fri, 11 Mar 2016, Borislav Petkov wrote: > > Obvious Ack to the patch, along with a "how did this ever work > > before?" comment.. > > I had a sarcastic sentence in the commit message which I deleted later: > > "Apparently no one had tried the kernel on a 486er after the FPU > rewrite. Backwards compatibility is overrated." People who care do not always have the resources to make regular tests with their hardware. I for once have meant to check the original series against my 486 EISA box since Ingo posted them back in May last year, and I still have them stashed away for this purpose. Unfortunately I didn't get to wire that machine for remote console access and power supply control (and I have since run out of ports too), so I need to physically get to its location to do any testing. So it's not unusual for bugs in the less common configurations to only surface a bit later on, as people get to making an upgrade. At least we have `git bisect' available and individual changes recorded now, which makes tracking down regressions a lot easier than it used to be in the old days. Maciej