kvm.vger.kernel.org archive mirror
 help / color / mirror / Atom feed
From: Sean Christopherson <seanjc@google.com>
To: Liam Ni <zhiguangni01@gmail.com>
Cc: linux-kernel@vger.kernel.org, kvm@vger.kernel.org,
	x86@kernel.org, pbonzini@redhat.com, tglx@linutronix.de,
	dave.hansen@linux.intel.com
Subject: Re: [PATCH] KVM: Reduce the execution of one instruction
Date: Tue, 6 Sep 2022 16:14:23 +0000	[thread overview]
Message-ID: <Yxdx36BHlClCq52J@google.com> (raw)
In-Reply-To: <20220906153357.1362555-1-zhiguangni01@zhaoxin.com>

"KVM: x86:" for the shortlog.

On Tue, Sep 06, 2022, Liam Ni wrote:
> From: Liam Ni <zhiguangni01@gmail.com>
> 
> If the condition is met,

Please describe this specific code change, "If the condition is met" is extremely
generic and doesn't help the reader understand what change is being made.

> reduce the execution of one instruction.

This is highly speculative, e.g. clang will generate identical output since it's
trivial for the compiler to observe that ctxt->modrm_reg doesn't need to be read.

And similar to the above "If the condition is met", the shortlog is too generic
even if it were 100% accurate.

I do think this change is a net positive, but it's beneficial only in making the
code easier to read.  Shaving a single cheap instruction in a relatively slow path
isn't sufficient justification even if the compiler isn't clever enough to optimize
away the load in the first place.

E.g. something like:

  KVM: x86: Clean up ModR/M "reg" initialization in reg op decoding

  Refactor decode_register_operand() to get the ModR/M register if and
  only if the instruction uses a ModR/M encoding to make it more obvious
  how the register operand is retrieved.

> Signed-off-by: Liam Ni <zhiguangni01@gmail.com>
> ---
>  arch/x86/kvm/emulate.c | 4 +++-
>  1 file changed, 3 insertions(+), 1 deletion(-)
> 
> diff --git a/arch/x86/kvm/emulate.c b/arch/x86/kvm/emulate.c
> index f8382abe22ff..ebb95f3f9862 100644
> --- a/arch/x86/kvm/emulate.c
> +++ b/arch/x86/kvm/emulate.c
> @@ -1139,10 +1139,12 @@ static int em_fnstsw(struct x86_emulate_ctxt *ctxt)
>  static void decode_register_operand(struct x86_emulate_ctxt *ctxt,
>  				    struct operand *op)
>  {
> -	unsigned reg = ctxt->modrm_reg;
> +	unsigned int reg;
>  
>  	if (!(ctxt->d & ModRM))
>  		reg = (ctxt->b & 7) | ((ctxt->rex_prefix & 1) << 3);
> +	else
> +		reg = ctxt->modrm_reg;

I'd prefer to write this as:

	unsigned int reg;

	if (ctxt->d & ModRM)
		reg = ctxt->modrm_reg;
	else
		reg = (ctxt->b & 7) | ((ctxt->rex_prefix & 1) << 3);

so that "is ModRM" check is immediately followed by "get ModRM".

>  
>  	if (ctxt->d & Sse) {
>  		op->type = OP_XMM;
> -- 
> 2.25.1

> 

  reply	other threads:[~2022-09-06 16:37 UTC|newest]

Thread overview: 3+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2022-09-06 15:33 [PATCH] KVM: Reduce the execution of one instruction Liam Ni
2022-09-06 16:14 ` Sean Christopherson [this message]
2022-09-07 16:03   ` Liam Ni

Reply instructions:

You may reply publicly to this message via plain-text email
using any one of the following methods:

* Save the following mbox file, import it into your mail client,
  and reply-to-all from there: mbox

  Avoid top-posting and favor interleaved quoting:
  https://en.wikipedia.org/wiki/Posting_style#Interleaved_style

* Reply using the --to, --cc, and --in-reply-to
  switches of git-send-email(1):

  git send-email \
    --in-reply-to=Yxdx36BHlClCq52J@google.com \
    --to=seanjc@google.com \
    --cc=dave.hansen@linux.intel.com \
    --cc=kvm@vger.kernel.org \
    --cc=linux-kernel@vger.kernel.org \
    --cc=pbonzini@redhat.com \
    --cc=tglx@linutronix.de \
    --cc=x86@kernel.org \
    --cc=zhiguangni01@gmail.com \
    /path/to/YOUR_REPLY

  https://kernel.org/pub/software/scm/git/docs/git-send-email.html

* If your mail client supports setting the In-Reply-To header
  via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).