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: <2f3b38fa-8180-a749-a590-2f758f3a479a@redhat.com>
Date:   Fri, 10 Jun 2022 13:14:07 +0200
MIME-Version: 1.0
User-Agent: Mozilla/5.0 (X11; Linux x86_64; rv:91.0) Gecko/20100101
 Thunderbird/91.8.0
Subject: Re: [bugzilla-daemon@kernel.org: [Bug 216109] New: Steam Deck fails
 to boot when E820 entries clipped out of _CRS]
Content-Language: en-US
To:     Bjorn Helgaas <helgaas@kernel.org>, linux-pci@vger.kernel.org
Cc:     "Guilherme G. Piccoli" <gpiccoli@igalia.com>,
        Thorsten Leemhuis <regressions@leemhuis.info>,
        linux-kernel@vger.kernel.org
References: <20220609224339.GA543225@bhelgaas>
From:   Hans de Goede <hdegoede@redhat.com>
In-Reply-To: <20220609224339.GA543225@bhelgaas>
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: 7bit
Precedence: bulk

Hi,

On 6/10/22 00:43, Bjorn Helgaas wrote:
> New regression in v5.19-rc1.
> 
> ----- Forwarded message from bugzilla-daemon@kernel.org -----
> 
> Subject: [Bug 216109] New: Steam Deck fails to boot when E820 entries clipped
> 	out of _CRS
> 
> https://bugzilla.kernel.org/show_bug.cgi?id=216109
> 
>            Summary: Steam Deck fails to boot when E820 entries clipped out
>                     of _CRS
>     Kernel Version: v5.19
> 
> Guilherme G. Piccoli reported that v5.18 boots fine on Steam Deck, but
> v5.19-rc1 does not.  He bisected it to 4c5e242d3e93 ("x86/PCI: Clip only host
> bridge windows for E820 regions") [1].
> 
> A quirk similar to [2] that disables E820 clipping makes v5.19-rc1 work again.
> 
> The reason why v5.18 (which always does E820 clipping by default) works, while
> v5.19-rc1 (which also does E820 clipping on this platform) does not has not
> been explained yet.
> 
> [1] https://git.kernel.org/linus/4c5e242d3e93
> [2] https://git.kernel.org/linus/d341838d776a
> 
> ----- End forwarded message -----

I've added some analysis of the logs to the bug. Summary: I think we may
need to revert 4c5e242d3e93, this will be a non trivial revert moving
the pci_use_e820 check added in later commits to arch_remove_reservations().

See the bug for details and lets discuss this further in the bug.

In case we do eventually decide to go with a revert, I'll prepare
a revert later today and push it to:

https://git.kernel.org/pub/scm/linux/kernel/git/pdx86/platform-drivers-x86.git/log/?h=review-hans

Note I do NOT plan to merge this through the pdx86 tree, the reason
to put it there is to get the builders to play with it, since in the
past: https://lore.kernel.org/linux-pci/20211014183943.27717-2-hdegoede@redhat.com/
we had some build issues with the check inside arch_remove_reservations().

Regards,

Hans