From mboxrd@z Thu Jan 1 00:00:00 1970 From: David Hildenbrand Subject: Re: [kvm-unit-tests PATCH v1] s390x: add a pgm irq handler and a way to expect them Date: Tue, 30 May 2017 10:24:14 +0200 Message-ID: References: <20170529121736.3055-1-david@redhat.com> <67569aa2-c95f-e7ac-47dd-ed2939a6c3ae@redhat.com> Mime-Version: 1.0 Content-Type: text/plain; charset=utf-8 Content-Transfer-Encoding: 7bit Cc: Paolo Bonzini , =?UTF-8?B?UmFkaW0gS3LEjW3DocWZ?= , Christian Borntraeger To: Thomas Huth , kvm@vger.kernel.org Return-path: Received: from mx1.redhat.com ([209.132.183.28]:46446 "EHLO mx1.redhat.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1751227AbdE3IYS (ORCPT ); Tue, 30 May 2017 04:24:18 -0400 In-Reply-To: <67569aa2-c95f-e7ac-47dd-ed2939a6c3ae@redhat.com> Content-Language: en-US Sender: kvm-owner@vger.kernel.org List-ID: On 30.05.2017 10:03, David Hildenbrand wrote: > On 29.05.2017 18:35, Thomas Huth wrote: >> On 29.05.2017 14:17, David Hildenbrand wrote: >>> The pgm irq handler will detect unexpected pgm irqs and allows to >>> expect pgm irqs + verify that the pgm irq was triggered. >>> >>> We need "-fno-delete-null-pointer-checks", otherwise trying to access the >>> lowcore at address 0 makes GCC generate very weird code. >> >> I wonder whether you could get rid of that by using a global variable >> for lc instead? > > As far as I remember, I tried that and it didn't change a thing. Will do > a quick test. It works as long as that global variable is not static, because then GCC must assume that any other code could set it to !0. -- Thanks, David