From mboxrd@z Thu Jan 1 00:00:00 1970 From: Laurent Vivier Date: Mon, 29 Feb 2016 16:44:51 +0000 Subject: Re: [kvm-unit-tests PATCH v2 2/2] powerpc: select endianness Message-Id: <56D47583.8010706@redhat.com> List-Id: References: <1456506526-10803-1-git-send-email-lvivier@redhat.com> <1456506526-10803-3-git-send-email-lvivier@redhat.com> <20160226184502.35oadt5jw2ck46la@hawk.localdomain> <56D46C9D.9050908@redhat.com> In-Reply-To: <56D46C9D.9050908@redhat.com> MIME-Version: 1.0 Content-Type: text/plain; charset="us-ascii" Content-Transfer-Encoding: 7bit To: Paolo Bonzini , Andrew Jones Cc: kvm@vger.kernel.org, kvm-ppc@vger.kernel.org, thuth@redhat.com, dgibson@redhat.com, agraf@suse.de On 29/02/2016 17:06, Paolo Bonzini wrote: > > > On 26/02/2016 19:45, Andrew Jones wrote: >>>> + LOAD_REG_ADDR(r10,rtas_return_loc) >>>> + mtlr r10 >>>> + >>>> LOAD_REG_ADDR(r10, rtas_blob) >>>> -//FIXME: change this bctrl to an rtas-prep, rfid, rtas-return sequence >>>> - mtctr r10 >>>> - nop >>>> - bctrl >>>> + B_BE(r10) > > Note that this doesn't apply. I've applied patch 1 though. Yes, Drew has changed this part, and anyway I have to rework this. Thanks, Laurent > Paolo > >>>> + >>>> +rtas_return_loc: >>>> + RETURN_FROM_BE >>>> ld r0, 16(r1) >>>> mtlr r0 >>>> blr >> Besides my earlier comment about always needing rfid to prep for the >> rtas call (which my FIXUP comment here poorly indicated), then this >> looks good to me. Although I'm not sure I like the name RETURN_FROM_BE, >> as sometimes we're returning to LE, and sometimes we're not. In the >> not case, it's a bit confusing.