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: <25456937-8034-2742-4a8a-68fb83b37ea2@kernel.org>
Date:   Mon, 29 Aug 2022 07:23:19 -0700
MIME-Version: 1.0
User-Agent: Mozilla/5.0 (X11; Linux x86_64; rv:102.0) Gecko/20100101
 Thunderbird/102.2.0
Subject: Re: [PATCH v3 3/5] x86/microcode: Avoid any chance of MCE's during
 microcode update
To:     Ashok Raj <ashok.raj@intel.com>, Borislav Petkov <bp@alien8.de>
Cc:     Ingo Molnar <mingo@kernel.org>,
        Thomas Gleixner <tglx@linutronix.de>,
        Tony Luck <tony.luck@intel.com>,
        Dave Hansen <dave.hansen@intel.com>,
        LKML Mailing List <linux-kernel@vger.kernel.org>,
        X86-kernel <x86@kernel.org>,
        Andy Lutomirski <luto@amacapital.net>,
        Tom Lendacky <thomas.lendacky@amd.com>,
        Jacon Jun Pan <jacob.jun.pan@intel.com>
References: <20220817051127.3323755-1-ashok.raj@intel.com>
 <20220817051127.3323755-4-ashok.raj@intel.com> <Yvybq+hYT4tG/yAg@gmail.com>
 <Yvyfi9XC8bu0cOG+@gmail.com> <YvyiHGMbp2MtV0Vr@zn.tnic>
 <YvzXsf0mGEcOlZC5@araj-dh-work> <Yvzay5jOu5XBJpWa@zn.tnic>
 <YvzfebgOWVKYLClJ@araj-dh-work> <Yvz4/ASoX4SiXbhp@zn.tnic>
 <Yv0D88jxFkXcc18o@araj-dh-work>
Content-Language: en-MW
From:   Andy Lutomirski <luto@kernel.org>
In-Reply-To: <Yv0D88jxFkXcc18o@araj-dh-work>
Content-Type: text/plain; charset=UTF-8; format=flowed
Content-Transfer-Encoding: 8bit
Precedence: bulk

On 8/17/22 08:06, Ashok Raj wrote:
> On Wed, Aug 17, 2022 at 04:19:40PM +0200, Borislav Petkov wrote:
>> On Wed, Aug 17, 2022 at 12:30:49PM +0000, Ashok Raj wrote:
>>> You will find out when system returns after reboot and hopefully wasn't
>>> promoted to a cold-boot which will loose MCE banks.
>> Not good enough!
> I probably misread your question.. are you suggesting we add some WARN when
> we initiate late_load? I thought you were asking if the HW must signal
> something and OS should log when an MCE happens if MCIP=1
>
>
>> This should issue a warning in dmesg that a potential MCE while update
>> is running would cause a lockup. That is if we don't disable MCE around
>> it.
>>
>> If we decide to disable MCE, it should say shutdown.
> Ok, that clarifies it.. "IF we choose to set MCIP=1, we should tell users
> that hell can break loose, get under the table" :-)
>
>>> Meaning deal with the effect of a really rare MCE. Rather than trying to
>>> avoid it. Taking the MCE is more important than finishing the update,
>>> and loosing what the error signaled was trying to convey.
>> Right now I'm inclined to not do anything and warn of a potential rare
>> situation.
> Encouraging.. So I'll drop that patch from the list next time around.


If I followed all this correctly, I agree. If we set MCIP to force a 
crash if we get MCE, then we are guaranteed to crash.  If we don't, then 
we might crash.


An imperfect alternative would be to set a (percpu?) flag that we're 
doing a ucode update and then detect that flag early in the MCE handler 
and warn very loudly.  This seems like it will give us the best chance 
of getting a useful diagnostic.