From mboxrd@z Thu Jan  1 00:00:00 1970
From: Heiko Carstens <heiko.carstens@de.ibm.com>
Subject: Re: [PATCH RFC 0/2] KVM: s390: avoid having to enable vm.alloc_pgste
Date: Fri, 2 Jun 2017 09:06:18 +0200
Message-ID: <20170602070618.GB4221@osiris>
References: <20170529163202.13077-1-david@redhat.com>
 <20170601124651.3e7969ab@mschwideX1>
 <a02736a2-f257-36c4-0c62-2cd57ae8c141@redhat.com>
Mime-Version: 1.0
Content-Type: text/plain; charset=us-ascii
Cc: Martin Schwidefsky <schwidefsky@de.ibm.com>, kvm@vger.kernel.org,
        linux-kernel@vger.kernel.org, Thomas Huth <thuth@redhat.com>,
        Christian Borntraeger <borntraeger@de.ibm.com>
To: David Hildenbrand <david@redhat.com>
Return-path: <linux-kernel-owner@vger.kernel.org>
Content-Disposition: inline
In-Reply-To: <a02736a2-f257-36c4-0c62-2cd57ae8c141@redhat.com>
Sender: linux-kernel-owner@vger.kernel.org
List-Id: kvm.vger.kernel.org

On Thu, Jun 01, 2017 at 01:27:28PM +0200, David Hildenbrand wrote:
> An alternative: Have some process that enables PGSTE for all of its
> future children. Fork+execv qemu. However, such a process in between
> will most likely confuse tooling like libvirt when it comes to process ids.

That would be something like sys_personality() vs setarch, e.g. adding a
new s390 specific personality flag and then do something like

setarch s390x --kvm qemu...

However I'm afraid that this is also not acceptable from a usability point
of view.