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;
MIME-Version: 1.0
References: <20221004103341.12646-1-jirislaby@kernel.org> <MN0PR12MB610136EC41E0D00F0BA7280FE25A9@MN0PR12MB6101.namprd12.prod.outlook.com>
In-Reply-To: <MN0PR12MB610136EC41E0D00F0BA7280FE25A9@MN0PR12MB6101.namprd12.prod.outlook.com>
From:   Chuanhong Guo <gch981213@gmail.com>
Date:   Wed, 5 Oct 2022 12:02:08 +0800
Message-ID: <CAJsYDVJsuYvAyFWwoW9=K_9otJR9Y61Wfk_w2e_K7WiyWf3Xjw@mail.gmail.com>
Subject: Re: [PATCH 1/2] ACPI: resource: do IRQ override on LENOVO IdeaPad
To:     "Limonciello, Mario" <Mario.Limonciello@amd.com>
Cc:     "Jiri Slaby (SUSE)" <jirislaby@kernel.org>,
        "rafael.j.wysocki@intel.com" <rafael.j.wysocki@intel.com>,
        "linux-kernel@vger.kernel.org" <linux-kernel@vger.kernel.org>,
        "Rafael J. Wysocki" <rafael@kernel.org>,
        Len Brown <lenb@kernel.org>,
        "linux-acpi@vger.kernel.org" <linux-acpi@vger.kernel.org>,
        Tighe Donnelly <tighe.donnelly@protonmail.com>,
        Fridrich Strba <fstrba@suse.com>
Content-Type: text/plain; charset="UTF-8"
Precedence: bulk

Hi!

On Wed, Oct 5, 2022 at 5:02 AM Limonciello, Mario
<Mario.Limonciello@amd.com> wrote:
> [...]
> >
> > White-list this specific model in the override_table.
> >
> > For this to work, the ZEN test needs to be put below the table walk.
>
> Unfortunately this is the second case that popped up very recently.
> Another one is listed here:
> https://bugzilla.kernel.org/show_bug.cgi?id=216552

Now I'm really curious how Windows is able to handle all these vendor crap...

> I don't think we have a good solution to cover the intersection of these bugs.  The
> existing heuristic to look at legacy syntax and the IOAPIC doesn't work properly
> on all the Lenovo and ASUS Ryzen 6000 systems, but it does on these other two.

These legacy IRQ declarations are obsolete, but they aren't really wrong.
Meanwhile the two devices popped up until now both got IRQ declarations which
don't match the actual device configuration.

> We're going to be adding more to this table either way.  I /suspect/ the better solution
> is to revert 37c81d9f1d1b and add to the table all those that are broken.

I think we should have a list of only the wrong IRQ declaration and
apply the fix
just for them, instead of applying the fix to all devices and skip it
for selected
devices the fix breaks.

-- 
Regards,
Chuanhong Guo