From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: from mx5-phx2.redhat.com ([209.132.183.37]:46255 "EHLO mx5-phx2.redhat.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1752102AbcIRMiN (ORCPT ); Sun, 18 Sep 2016 08:38:13 -0400 Date: Sun, 18 Sep 2016 08:38:09 -0400 (EDT) From: Paolo Bonzini To: Alexander Levin Cc: stable@vger.kernel.org, matt@codeblueprint.co.uk Message-ID: <1290443831.307438.1474202289479.JavaMail.zimbra@redhat.com> In-Reply-To: <20160918004441.GA16340@sasha-lappy> References: <1473976363-22653-1-git-send-email-pbonzini@redhat.com> <20160918004441.GA16340@sasha-lappy> Subject: Re: [PATCH] Revert "KVM: x86: fix missed hardware breakpoints" MIME-Version: 1.0 Content-Type: text/plain; charset=utf-8 Content-Transfer-Encoding: 7bit Sender: stable-owner@vger.kernel.org List-ID: > > The upshot is that KVM_DEBUGREG_RELOAD is always set when returning > > from kvm_arch_vcpu_load() in stable, but not in Linus' tree. > > > > This happened because [0] and [1] are the same patch. [0] and [1] come > > from two > > different merges, and the later merge is trivially resolved; when [2] > > is applied it reverts both of them. Instead, when using the [1][2][0] > > order, patches applies normally but "KVM: x86: fix missed hardware > > breakpoints" is present in the final tree. > > > > Cc: stable kernels # 4.1, 4.4 > > Reported-by: Matt Fleming > > Signed-off-by: Paolo Bonzini > > Hey Paolo, > > Could you please confirm it's indeed a problem in 4.1? The current 4.1 tree > seems to be "correct" (the current code doesn't have the line that this > patch deletes). Hi Sasha, yes, 4.1 is good. It's limited to 4.4. Paolo