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;
From:   John Ogness <john.ogness@linutronix.de>
To:     Borislav Petkov <bp@alien8.de>
Cc:     linux-kernel@vger.kernel.org
Subject: Re: [PATCH] x86/microcode: move @microcode_mutex definition near usage
In-Reply-To: <20230324120019.GAZB2Q09ODVq0iYz5l@fat_crate.local>
References: <20230324114720.1756466-1-john.ogness@linutronix.de>
 <20230324120019.GAZB2Q09ODVq0iYz5l@fat_crate.local>
Date:   Fri, 24 Mar 2023 14:17:10 +0106
Message-ID: <87edpenvkh.fsf@jogness.linutronix.de>
MIME-Version: 1.0
Content-Type: text/plain
Precedence: bulk

On 2023-03-24, Borislav Petkov <bp@alien8.de> wrote:
>> If CONFIG_MICROCODE_LATE_LOADING is not enabled, the compiler warns:
>> 
>> 'microcode_mutex' defined but not used
>
> How do you trigger this?

I was doing some tests with CONFIG_PREEMPT_RT. I did not think that
mattered, since the mutex is obviously defined but not used for
!CONFIG_MICROCODE_LATE_LOADING.

Digging deeper I see that initializing @wait_list in
__MUTEX_INITIALIZER() is what is allowing unused global mutexes to go
unnoticed. Since with CONFIG_PREEMPT_RT the mutex is different and
initialized differently, it is (correctly) detected as unused.

CONFIG_PREEMPT_RT cannot be enabled yet, so this patch is not urgent for
mainline. But at some point it will need fixing.

John Ogness