From mboxrd@z Thu Jan 1 00:00:00 1970 From: Paolo Bonzini Subject: Re: [GIT PULL] Please pull my kvm-ppc-fixes branch Date: Wed, 21 Jun 2017 12:30:51 +0200 Message-ID: <7304ec7d-9eb3-0eec-706f-b7689e64560a@redhat.com> References: <20170620043035.avphtcoght2iiyug@oak.ozlabs.ibm.com> <20170620123907.GC10325@potion> <3bc0aeed-136b-5b7f-03a4-17b3aae31842@redhat.com> <20170620151700.GF10325@potion> <20170620233756.GA3644@fergus.ozlabs.ibm.com> Mime-Version: 1.0 Content-Type: text/plain; charset=utf-8 Content-Transfer-Encoding: 8bit Cc: kvm@vger.kernel.org, David Gibson , kvm-ppc@vger.kernel.org To: Paul Mackerras , =?UTF-8?B?UmFkaW0gS3LEjW3DocWZ?= Return-path: Received: from mx1.redhat.com ([209.132.183.28]:17216 "EHLO mx1.redhat.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1752339AbdFUKaz (ORCPT ); Wed, 21 Jun 2017 06:30:55 -0400 In-Reply-To: <20170620233756.GA3644@fergus.ozlabs.ibm.com> Content-Language: en-US Sender: kvm-owner@vger.kernel.org List-ID: On 21/06/2017 01:37, Paul Mackerras wrote: >> Right, and the arch doesn't even match ... let's forget I asked. :) >> >> Pushed into kvm/master, thanks. > Thanks. > > On a related note, we still have the warning in virt/kvm/kvm_main.c > for CONFIG_SMP=n builds about the variable 'wait' being unused in > kvm_make_all_cpus_request(), because smp_call_function_many() doesn't > reference its 'wait' argument for uniprocessor configs. > > This then causes the build to fail on powerpc because kvm_main.c > actually gets compiled under arch/powerpc, and we use -Werror for > arch/powerpc by default. > > Paolo posted a sketch of a patch to make smp_call_function_many be a > static inline instead of a macro on UP, but it wasn't quite right and > no-one sent it anywhere. Paolo, if you're OK with me cleaning it up > and sending it on, I'll do that. The problem is that it breaks MIPS, which uses an unused variable. I'll send an alternative patch that is confined in KVM. Paolo From mboxrd@z Thu Jan 1 00:00:00 1970 From: Paolo Bonzini Date: Wed, 21 Jun 2017 10:30:51 +0000 Subject: Re: [GIT PULL] Please pull my kvm-ppc-fixes branch Message-Id: <7304ec7d-9eb3-0eec-706f-b7689e64560a@redhat.com> List-Id: References: <20170620043035.avphtcoght2iiyug@oak.ozlabs.ibm.com> <20170620123907.GC10325@potion> <3bc0aeed-136b-5b7f-03a4-17b3aae31842@redhat.com> <20170620151700.GF10325@potion> <20170620233756.GA3644@fergus.ozlabs.ibm.com> In-Reply-To: <20170620233756.GA3644@fergus.ozlabs.ibm.com> MIME-Version: 1.0 Content-Type: text/plain; charset="us-ascii" Content-Transfer-Encoding: 7bit To: Paul Mackerras , =?UTF-8?B?UmFkaW0gS3LEjW3DocWZ?= Cc: kvm@vger.kernel.org, David Gibson , kvm-ppc@vger.kernel.org On 21/06/2017 01:37, Paul Mackerras wrote: >> Right, and the arch doesn't even match ... let's forget I asked. :) >> >> Pushed into kvm/master, thanks. > Thanks. > > On a related note, we still have the warning in virt/kvm/kvm_main.c > for CONFIG_SMP=n builds about the variable 'wait' being unused in > kvm_make_all_cpus_request(), because smp_call_function_many() doesn't > reference its 'wait' argument for uniprocessor configs. > > This then causes the build to fail on powerpc because kvm_main.c > actually gets compiled under arch/powerpc, and we use -Werror for > arch/powerpc by default. > > Paolo posted a sketch of a patch to make smp_call_function_many be a > static inline instead of a macro on UP, but it wasn't quite right and > no-one sent it anywhere. Paolo, if you're OK with me cleaning it up > and sending it on, I'll do that. The problem is that it breaks MIPS, which uses an unused variable. I'll send an alternative patch that is confined in KVM. Paolo