Received-SPF: pass (google.com: domain of linux-kernel-owner@vger.kernel.org designates 2620:137:e000::1:20 as permitted sender) client-ip=2620:137:e000::1:20;
Message-ID: <54de068f-94ff-ce75-333d-7f1f79e2743f@redhat.com>
Date:   Thu, 3 Aug 2023 07:23:59 +0200
MIME-Version: 1.0
User-Agent: Mozilla/5.0 (X11; Linux x86_64; rv:102.0) Gecko/20100101
 Thunderbird/102.13.0
Content-Language: en-US
To:     Sean Christopherson <seanjc@google.com>
Cc:     kvm@vger.kernel.org, linux-kernel@vger.kernel.org,
        Paolo Bonzini <pbonzini@redhat.com>,
        linux-kselftest@vger.kernel.org,
        David Matlack <dmatlack@google.com>
References: <20230712075910.22480-1-thuth@redhat.com>
 <20230712075910.22480-3-thuth@redhat.com> <ZMq0nYYDbOX1cOKN@google.com>
 <ZMrLMYxj3s+vHGrQ@google.com>
From:   Thomas Huth <thuth@redhat.com>
Subject: Re: [PATCH 2/4] KVM: selftests: x86: Use TAP interface in the
 sync_regs test
In-Reply-To: <ZMrLMYxj3s+vHGrQ@google.com>
Content-Type: text/plain; charset=UTF-8; format=flowed
Content-Transfer-Encoding: 7bit
Precedence: bulk

On 02/08/2023 23.31, Sean Christopherson wrote:
> On Wed, Aug 02, 2023, Sean Christopherson wrote:
>> Oh, and no need to post "KVM: selftests: Rename the ASSERT_EQ macro" in the next
>> version, I'm planning on grabbing that one straightaway.
> 
> After paging this all back in...
> 
> I would much prefer that we implement the KVM specific macros[*], e.g. KVM_ONE_VCPU_TEST(),
> and build on top of those.  I'm definitely ok doing a "slow" conversion, i.e. starting
> with a few easy tests.  IIRC at some point I said I strongly preferred an all-or-nothing
> approach, but realistically I don't think we'll make progress anytime soon if we try to
> boil the ocean.

At least I don't have enough spare time to do such a big conversion all at 
once - I'm only occasionally looking at the KVM selftests, mostly for s390x, 
and I also lack the knowledge how to test all those x86 tests. So don't 
expect such a big conversion from me, all I can provide is a small patch 
here or there.

> But I do think we should spend the time to implement the infrastructure right away.  We
> may end up having to tweak the infrastructure down the road, e.g. to convert other tests,
> but I would rather do that then convert some tests twice.
> 
> [*] https://lore.kernel.org/all/Y2v+B3xxYKJSM%2FfH@google.com

Sorry, I somehow completely missed that KVM_ONE_VCPU_TEST suggestion when 
picking up the series up again after working on other stuff for more than 
half a year. I'll try to incorporate this into the next version.

(the other patches don't need a fixture, so I think they shouldn't be 
affected by this?)

  Thomas