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: <7403891b-c84c-d0e5-e6e6-caf1bd4fcf51@igalia.com>
Date:   Thu, 6 Oct 2022 12:07:16 -0300
MIME-Version: 1.0
User-Agent: Mozilla/5.0 (X11; Linux x86_64; rv:102.0) Gecko/20100101
 Thunderbird/102.2.2
Subject: Re: [PATCH] x86/split_lock: Restore warn mode (and add a new one) to
 avoid userspace regression
Content-Language: en-US
To:     Pavel Machek <pavel@denx.de>, "Luck, Tony" <tony.luck@intel.com>,
        "x86@kernel.org" <x86@kernel.org>,
        "dave.hansen@linux.intel.com" <dave.hansen@linux.intel.com>
Cc:     "tglx@linutronix.de" <tglx@linutronix.de>,
        "linux-kernel@vger.kernel.org" <linux-kernel@vger.kernel.org>,
        "mingo@redhat.com" <mingo@redhat.com>,
        "bp@alien8.de" <bp@alien8.de>, "hpa@zytor.com" <hpa@zytor.com>,
        "Lutomirski, Andy" <luto@kernel.org>,
        "kernel-dev@igalia.com" <kernel-dev@igalia.com>,
        "kernel@gpiccoli.net" <kernel@gpiccoli.net>,
        "Yu, Fenghua" <fenghua.yu@intel.com>,
        Joshua Ashton <joshua@froggi.es>,
        Paul Gofman <pgofman@codeweavers.com>,
        Pierre-Loup Griffais <pgriffais@valvesoftware.com>,
        Melissa Wen <mwen@igalia.com>
References: <20220928142109.150263-1-gpiccoli@igalia.com>
 <SJ1PR11MB6083113884DD0B3031FE372CFC549@SJ1PR11MB6083.namprd11.prod.outlook.com>
 <20221006090418.GA14960@amd>
From:   "Guilherme G. Piccoli" <gpiccoli@igalia.com>
In-Reply-To: <20221006090418.GA14960@amd>
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: 7bit
Precedence: bulk

On 06/10/2022 06:04, Pavel Machek wrote:
> [...]
>> Why not just use the workaround suggested in that bug report:
>>
>>    "so manual switching from default setting to split_lock_detect=off helps as workaround here"
>>
>> If you add this extra mode, I'm going to argue that the kernel default
>> should be "seq" rather than "warn". So these game players will need
>> to add a split_lock_detect=off (or warn) option.
> 
> Kernel should not cause userland regressions, and this is one. That
> should make it pretty clear what the solution is.
> 
> And no, I don't like CONFIG option, either.

Thanks for your opinion Pavel! Good thing is that seems everybody at
least understands the problem exists and is affecting userspace.

Now, about the action we should take: what if we go with a revert in
this "misery factor" based on Dave's reasoning? I have a patch ready and
tested, let me know otherwise or I'll submit it as a V2 by the end of
this week.

Cheers,


Guilherme