linux-kernel - Re: [PATCH v2 1/7] KVM: s390: clean up redundant 'kvm

lists.openwall.net		lists / announce owl-users owl-dev john-users john-dev passwdqc-users yescrypt popa3d-users / oss-security kernel-hardening musl sabotage tlsify passwords / crypt-dev xvendor / Bugtraq Full-Disclosure linux-kernel linux-netdev linux-ext4 linux-hardening linux-cve-announce PHC
Open Source and information security mailing list archives
Hash Suite for Android: free password hash cracker in your pocket
[<prev] [next>] [<thread-prev] [day] [month] [year] [list]
Date:   Wed, 29 Apr 2020 10:20:17 +0800
From:   Tianjia Zhang <tianjia.zhang@...ux.alibaba.com>
To:     Thomas Huth <thuth@...hat.com>,
        Christian Borntraeger <borntraeger@...ibm.com>,
        Cornelia Huck <cohuck@...hat.com>
Cc:     pbonzini@...hat.com, tsbogend@...ha.franken.de, paulus@...abs.org,
        mpe@...erman.id.au, benh@...nel.crashing.org,
        frankja@...ux.ibm.com, david@...hat.com, heiko.carstens@...ibm.com,
        gor@...ux.ibm.com, sean.j.christopherson@...el.com,
        vkuznets@...hat.com, wanpengli@...cent.com, jmattson@...gle.com,
        joro@...tes.org, tglx@...utronix.de, mingo@...hat.com,
        bp@...en8.de, x86@...nel.org, hpa@...or.com, maz@...nel.org,
        james.morse@....com, julien.thierry.kdev@...il.com,
        suzuki.poulose@....com, christoffer.dall@....com,
        peterx@...hat.com, kvm@...r.kernel.org,
        linux-arm-kernel@...ts.infradead.org, kvmarm@...ts.cs.columbia.edu,
        linux-mips@...r.kernel.org, kvm-ppc@...r.kernel.org,
        linuxppc-dev@...ts.ozlabs.org, linux-s390@...r.kernel.org,
        linux-kernel@...r.kernel.org
Subject: Re: [PATCH v2 1/7] KVM: s390: clean up redundant 'kvm_run' parameters



On 2020/4/26 20:59, Thomas Huth wrote:
> On 23/04/2020 13.00, Christian Borntraeger wrote:
>>
>>
>> On 23.04.20 12:58, Tianjia Zhang wrote:
>>>
>>>
>>> On 2020/4/23 18:39, Cornelia Huck wrote:
>>>> On Thu, 23 Apr 2020 11:01:43 +0800
>>>> Tianjia Zhang <tianjia.zhang@...ux.alibaba.com> wrote:
>>>>
>>>>> On 2020/4/23 0:04, Cornelia Huck wrote:
>>>>>> On Wed, 22 Apr 2020 17:58:04 +0200
>>>>>> Christian Borntraeger <borntraeger@...ibm.com> wrote:
>>>>>>    
>>>>>>> On 22.04.20 15:45, Cornelia Huck wrote:
>>>>>>>> On Wed, 22 Apr 2020 20:58:04 +0800
>>>>>>>> Tianjia Zhang <tianjia.zhang@...ux.alibaba.com> wrote:
>>>>>>>>       
>>>>>>>>> In the current kvm version, 'kvm_run' has been included in the 'kvm_vcpu'
>>>>>>>>> structure. Earlier than historical reasons, many kvm-related function
>>>>>>>>
>>>>>>>> s/Earlier than/For/ ?
>>>>>>>>       
>>>>>>>>> parameters retain the 'kvm_run' and 'kvm_vcpu' parameters at the same time.
>>>>>>>>> This patch does a unified cleanup of these remaining redundant parameters.
>>>>>>>>>
>>>>>>>>> Signed-off-by: Tianjia Zhang <tianjia.zhang@...ux.alibaba.com>
>>>>>>>>> ---
>>>>>>>>>     arch/s390/kvm/kvm-s390.c | 37 ++++++++++++++++++++++---------------
>>>>>>>>>     1 file changed, 22 insertions(+), 15 deletions(-)
>>>>>>>>>
>>>>>>>>> diff --git a/arch/s390/kvm/kvm-s390.c b/arch/s390/kvm/kvm-s390.c
>>>>>>>>> index e335a7e5ead7..d7bb2e7a07ff 100644
>>>>>>>>> --- a/arch/s390/kvm/kvm-s390.c
>>>>>>>>> +++ b/arch/s390/kvm/kvm-s390.c
>>>>>>>>> @@ -4176,8 +4176,9 @@ static int __vcpu_run(struct kvm_vcpu *vcpu)
>>>>>>>>>         return rc;
>>>>>>>>>     }
>>>>>>>>>     -static void sync_regs_fmt2(struct kvm_vcpu *vcpu, struct kvm_run *kvm_run)
>>>>>>>>> +static void sync_regs_fmt2(struct kvm_vcpu *vcpu)
>>>>>>>>>     {
>>>>>>>>> +    struct kvm_run *kvm_run = vcpu->run;
>>>>>>>>>         struct runtime_instr_cb *riccb;
>>>>>>>>>         struct gs_cb *gscb;
>>>>>>>>>     @@ -4235,7 +4236,7 @@ static void sync_regs_fmt2(struct kvm_vcpu *vcpu, struct kvm_run *kvm_run)
>>>>>>>>>             }
>>>>>>>>>             if (vcpu->arch.gs_enabled) {
>>>>>>>>>                 current->thread.gs_cb = (struct gs_cb *)
>>>>>>>>> -                        &vcpu->run->s.regs.gscb;
>>>>>>>>> +                        &kvm_run->s.regs.gscb;
>>>>>>>>
>>>>>>>> Not sure if these changes (vcpu->run-> => kvm_run->) are really worth
>>>>>>>> it. (It seems they amount to at least as much as the changes advertised
>>>>>>>> in the patch description.)
>>>>>>>>
>>>>>>>> Other opinions?
>>>>>>>
>>>>>>> Agreed. It feels kind of random. Maybe just do the first line (move kvm_run from the
>>>>>>> function parameter list into the variable declaration)? Not sure if this is better.
>>>>>>>    
>>>>>>
>>>>>> There's more in this patch that I cut... but I think just moving
>>>>>> kvm_run from the parameter list would be much less disruptive.
>>>>>>     
>>>>>
>>>>> I think there are two kinds of code(`vcpu->run->` and `kvm_run->`), but
>>>>> there will be more disruptive, not less.
>>>>
>>>> I just fail to see the benefit; sure, kvm_run-> is convenient, but the
>>>> current code is just fine, and any rework should be balanced against
>>>> the cost (e.g. cluttering git annotate).
>>>>
>>>
>>> cluttering git annotate ? Does it mean Fix xxxx ("comment"). Is it possible to solve this problem by splitting this patch?
>>
>> No its about breaking git blame (and bugfix backports) for just a cosmetic improvement.
> 
> It could be slightly more than a cosmetic improvement (depending on the
> smartness of the compiler): vcpu->run-> are two dereferences, while
> kvm_run-> is only one dereference. So it could be slightly more compact
> and faster code.
> 
>   Thomas
> 

If the compiler is smart enough, this place can be automatically 
optimized, but we can't just rely on the compiler, if not? This requires 
a trade-off between code cleanliness readability and breaking git blame.
In addition, I have removed the changes here and sent a v4 patch. Please 
also help review it.

Thanks and best,
Tianjia