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: <bda79e85-c0bf-8d59-2750-d922a59bb859@redhat.com>
Date:   Tue, 18 Jul 2023 13:41:08 +0200
MIME-Version: 1.0
User-Agent: Mozilla/5.0 (X11; Linux x86_64; rv:102.0) Gecko/20100101
 Thunderbird/102.12.0
Content-Language: en-US
To:     Maxim Levitsky <mlevitsk@redhat.com>, kvm@vger.kernel.org
Cc:     "H. Peter Anvin" <hpa@zytor.com>,
        Sean Christopherson <seanjc@google.com>,
        linux-kernel@vger.kernel.org, Thomas Gleixner <tglx@linutronix.de>,
        Ingo Molnar <mingo@redhat.com>,
        Dave Hansen <dave.hansen@linux.intel.com>, x86@kernel.org,
        Borislav Petkov <bp@alien8.de>
References: <20230718091310.119672-1-mlevitsk@redhat.com>
 <20230718091310.119672-2-mlevitsk@redhat.com>
From:   Paolo Bonzini <pbonzini@redhat.com>
Subject: Re: [PATCH 1/3] KVM: x86: VMX: __kvm_apic_update_irr must update the
 IRR atomically
In-Reply-To: <20230718091310.119672-2-mlevitsk@redhat.com>
Content-Type: text/plain; charset=UTF-8; format=flowed
Content-Transfer-Encoding: 7bit
Precedence: bulk

On 7/18/23 11:13, Maxim Levitsky wrote:
> +		irr_val = READ_ONCE(*((u32 *)(regs + APIC_IRR + i * 0x10)));

Let's separate out the complicated arithmetic, as it recurs below too:

	u32 *p_irr = (u32 *)(regs + APIC_IRR + i * 0x10);

> +			while (!try_cmpxchg(((u32 *)(regs + APIC_IRR + i * 0x10)),
> +			       &irr_val, irr_val | pir_val));
> +
>   			prev_irr_val = irr_val;
> -			irr_val |= xchg(&pir[i], 0);
> -			*((u32 *)(regs + APIC_IRR + i * 0x10)) = irr_val;
> -			if (prev_irr_val != irr_val) {
> -				max_updated_irr =
> -					__fls(irr_val ^ prev_irr_val) + vec;
> -			}
> +			irr_val |= pir_val;
> +
> +			if (prev_irr_val != irr_val)
> +				max_updated_irr = __fls(irr_val ^ prev_irr_val) + vec;

We can write this a bit more cleanly too, and avoid unnecessary
try_cmpxchg too:

prev_irr_val = irr_val;
do
	irr_val = prev_irr_val | pir_val;
while (prev_irr_val != irr_val &&
        !try_cmpxchg(p_irr, &prev_irr_val, irr_val));

if (prev_irr_val != irr_val)
	max_updated_irr = __fls(irr_val ^ prev_irr_val) + vec;

If this looks okay to you, I'll queue the patches for -rc3 and also Cc 
them for inclusion in stable kernels.

Paolo