linux-kernel - Re: [PATCH 4/6] KVM: x86: wean fast IN from emulator_pio

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: Windows password security audit tool. GUI, reports in PDF.

[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]

Date:   Thu, 9 Jun 2022 22:37:26 +0000
From:   Sean Christopherson <seanjc@...gle.com>
To:     Paolo Bonzini <pbonzini@...hat.com>
Cc:     linux-kernel@...r.kernel.org, kvm@...r.kernel.org
Subject: Re: [PATCH 4/6] KVM: x86: wean fast IN from emulator_pio_in

On Wed, Jun 08, 2022, Paolo Bonzini wrote:
> Now that __emulator_pio_in already fills "val" for in-kernel PIO, it

For some reason the "already" confused the heck out of me.  I thought it was
referring to a previous patch, which it kind of is, but then I couldn't figure
out the relevance to this patch.

Ah, I know why I got confused, the in-kernel PIO case has nothing to do with the
usage in complete_fast_pio_in(), e.g. complete_fast_pio_in() could be modified to
call complete_emulator_pio_in() directly even without the previous cleanup in
this series.

Can you split this patch in two?  It's comically trivial, but it makes the
changelogs much easier to understand.

  Use __emulator_pio_in() directly for fast PIO instead of bouncing through
  emulator_pio_in() now that __emulator_pio_in() fills "val" when handling
  in-kernel PIO.  vcpu->arch.pio.count is guaranteed to be '0', so this a
  pure nop.

  No functional change intended.

and

  Use complete_emulator_pio_in() directly when completing fast PIO, there's
  no need to bounce through emulator_pio_in() as the comment about ECX
  changing doesn't apply to fast PIO, which isn't used for string I/O.

  No functional change intended.

> is both simpler and clearer not to use emulator_pio_in.
> Use the appropriate function in kvm_fast_pio_in and complete_fast_pio_in,
> respectively __emulator_pio_in and complete_emulator_pio_in.
> 
> emulator_pio_in_emulated is now the last caller of emulator_pio_in.
> 
> No functional change intended.
> 
> Signed-off-by: Paolo Bonzini <pbonzini@...hat.com>
> ---
>  arch/x86/kvm/x86.c | 8 ++------
>  1 file changed, 2 insertions(+), 6 deletions(-)
> 
> diff --git a/arch/x86/kvm/x86.c b/arch/x86/kvm/x86.c
> index 3b641cd2ff6f..aefcc71a7040 100644
> --- a/arch/x86/kvm/x86.c
> +++ b/arch/x86/kvm/x86.c
> @@ -8692,11 +8692,7 @@ static int complete_fast_pio_in(struct kvm_vcpu *vcpu)
>  	/* For size less than 4 we merge, else we zero extend */
>  	val = (vcpu->arch.pio.size < 4) ? kvm_rax_read(vcpu) : 0;
>  
> -	/*
> -	 * Since vcpu->arch.pio.count == 1 let emulator_pio_in perform
> -	 * the copy and tracing
> -	 */
> -	emulator_pio_in(vcpu, vcpu->arch.pio.size, vcpu->arch.pio.port, &val, 1);
> +	complete_emulator_pio_in(vcpu, &val);
>  	kvm_rax_write(vcpu, val);
>  
>  	return kvm_skip_emulated_instruction(vcpu);
> @@ -8711,7 +8707,7 @@ static int kvm_fast_pio_in(struct kvm_vcpu *vcpu, int size,
>  	/* For size less than 4 we merge, else we zero extend */
>  	val = (size < 4) ? kvm_rax_read(vcpu) : 0;
>  
> -	ret = emulator_pio_in(vcpu, size, port, &val, 1);
> +	ret = __emulator_pio_in(vcpu, size, port, &val, 1);
>  	if (ret) {
>  		kvm_rax_write(vcpu, val);
>  		return ret;
> -- 
> 2.31.1
> 
>