From mboxrd@z Thu Jan 1 00:00:00 1970 From: Christoffer Dall Subject: Re: [PATCH 3/5] arm/arm64: KVM: Clarify KVM_ARM_VCPU_INIT ABI Date: Tue, 2 Dec 2014 15:47:28 +0100 Message-ID: <20141202144728.GH545@cbox> References: <1417113660-23610-1-git-send-email-christoffer.dall@linaro.org> <1417113660-23610-4-git-send-email-christoffer.dall@linaro.org> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Cc: "kvmarm@lists.cs.columbia.edu" , arm-mail-list , kvm-devel , Ard Biesheuvel , Marc Zyngier To: Peter Maydell Return-path: Received: from mail-la0-f49.google.com ([209.85.215.49]:38034 "EHLO mail-la0-f49.google.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1753500AbaLBOqq (ORCPT ); Tue, 2 Dec 2014 09:46:46 -0500 Received: by mail-la0-f49.google.com with SMTP id hs14so10580585lab.22 for ; Tue, 02 Dec 2014 06:46:44 -0800 (PST) Content-Disposition: inline In-Reply-To: Sender: kvm-owner@vger.kernel.org List-ID: On Thu, Nov 27, 2014 at 10:53:50PM +0000, Peter Maydell wrote: > On 27 November 2014 at 18:40, Christoffer Dall > wrote: > > It is not clear that this ioctl can be called multiple times for a given > > vcpu. Userspace already does this, so clarify the ABI. > > > > Signed-off-by: Christoffer Dall > > --- > > Documentation/virtual/kvm/api.txt | 3 +++ > > 1 file changed, 3 insertions(+) > > > > diff --git a/Documentation/virtual/kvm/api.txt b/Documentation/virtual/kvm/api.txt > > index bb82a90..fc12b4f 100644 > > --- a/Documentation/virtual/kvm/api.txt > > +++ b/Documentation/virtual/kvm/api.txt > > @@ -2453,6 +2453,9 @@ return ENOEXEC for that vcpu. > > Note that because some registers reflect machine topology, all vcpus > > should be created before this ioctl is invoked. > > > > +Userspace can call this function multiple times for a given VCPU, which will > > +reset the VCPU to its initial states. > > How about being a little bit more explicit here with something like: > > "Userspace can call this function multiple times for a given VCPU, including > after the VCPU has been run. This will reset the VCPU to its initial > state." yeah, better. > > (I notice that api.txt is inconsistent about using "vcpu" or "VCPU" > or "vCPU"... do we have a preference for new text?) > I generally try to match whatever the context is, but I clearly failed here. I don't think there's a preference, no. > > + > > Possible features: > > - KVM_ARM_VCPU_POWER_OFF: Starts the CPU in a power-off state. > > Depends on KVM_CAP_ARM_PSCI. If not set, the CPU will be powered on > > Do you have to use the same set of feature flags for second and > subsequent VCPU_INIT calls, or can they be different each time? > That's a good question. Do you have any opinion on the matter? It seems weird to change the target of a Vcpu from some core to another core, but there is not reason why you shouldn't be able to set a vCpU to be powered off when run, just because it wasn't earlier on, is there? Thanks, -Christoffer