Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S932143AbcDVQGU (ORCPT ); Fri, 22 Apr 2016 12:06:20 -0400 Received: from mailapp01.imgtec.com ([195.59.15.196]:62426 "EHLO mailapp01.imgtec.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S932065AbcDVQGR (ORCPT ); Fri, 22 Apr 2016 12:06:17 -0400 Date: Fri, 22 Apr 2016 17:06:05 +0100 From: "Maciej W. Rozycki" To: Paul Burton CC: , Ralf Baechle , "stable # v4 . 0+" , Adam Buchbinder , , James Hogan Subject: Re: [PATCH 2/2] MIPS: Force CPUs to lose FP context during mode switches In-Reply-To: <1461239038-19991-2-git-send-email-paul.burton@imgtec.com> Message-ID: References: <1461239038-19991-1-git-send-email-paul.burton@imgtec.com> <1461239038-19991-2-git-send-email-paul.burton@imgtec.com> User-Agent: Alpine 2.00 (DEB 1167 2008-08-23) MIME-Version: 1.0 Content-Type: text/plain; charset="US-ASCII" X-Originating-IP: [10.20.78.30] Sender: linux-kernel-owner@vger.kernel.org List-ID: X-Mailing-List: linux-kernel@vger.kernel.org Content-Length: 813 Lines: 17 On Thu, 21 Apr 2016, Paul Burton wrote: > Fix this by broadcasting an IPI if other CPUs may have live FP context > for an affected thread, with a handler causing those CPUs to relinquish > their FPU ownership. Threads will then be allowed to continue running > but will stall on the wait_on_atomic_t in enable_restore_fp_context if > they attempt to use FP again whilst the mode switch is still in > progress. The end result is less fragile poking at scheduler context > switch counts & a more expedient completion of the mode switch. Reviewed-by: Maciej W. Rozycki Thanks, Paul, for your work on this problem! I'll rebase my outstanding NaN interlinking stuff (https://patchwork.linux-mips.org/patch/11491/) on top of your patches -- they address the concern expressed there. Maciej