netdev - Re: [PATCH] brcmfmac: of: Use board compatible string for board type

lists.openwall.net		lists / announce owl-users owl-dev john-users john-dev passwdqc-users yescrypt popa3d-users / oss-security kernel-hardening musl sabotage tlsify passwords / crypt-dev xvendor / Bugtraq Full-Disclosure linux-kernel linux-netdev linux-ext4 linux-hardening linux-cve-announce PHC
Open Source and information security mailing list archives

Hash Suite: Windows password security audit tool. GUI, reports in PDF.

[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]

Message-ID: <2711b084-5937-7e0f-26d8-67510da3939c@marcan.st>
Date:   Fri, 6 Jan 2023 21:13:32 +0900
From:   Hector Martin <marcan@...can.st>
To:     "Ivan T. Ivanov" <iivanov@...e.de>, aspriel@...il.com
Cc:     franky.lin@...adcom.com, hante.meuleman@...adcom.com,
        rmk+kernel@...linux.org.uk, kvalo@...nel.org, davem@...emloft.net,
        devicetree@...r.kernel.org, edumazet@...gle.com,
        krzysztof.kozlowski+dt@...aro.org, kuba@...nel.org,
        pabeni@...hat.com, robh+dt@...nel.org,
        linux-arm-kernel@...ts.infradead.org,
        linux-wireless@...r.kernel.org, netdev@...r.kernel.org,
        brcm80211-dev-list.pdl@...adcom.com,
        SHA-cyfmac-dev-list@...ineon.com
Subject: Re: [PATCH] brcmfmac: of: Use board compatible string for board type

On 2023/01/06 18:27, Hector Martin wrote:
> On 2023/01/06 16:27, Ivan T. Ivanov wrote:
>> When "brcm,board-type" is not explicitly set in devicetree
>> fallback to board compatible string for board type.
>>
>> Some of the existing devices rely on the most compatible device
>> string to find best firmware files, including Raspberry PI's[1].
>>
>> Fixes: 7682de8b3351 ("wifi: brcmfmac: of: Fetch Apple properties")
>>
>> [1] https://bugzilla.opensuse.org/show_bug.cgi?id=1206697#c13
>>
>> Signed-off-by: Ivan T. Ivanov <iivanov@...e.de>
> 
> The existing code already falls back to the compatible string, *as long
> as there is no board_type set already*.
> 
> As far as I can tell, the only way the board_type can get another value
> first is if it comes from DMI. This behavior was inadvertently changed
> by commit 7682de8b3351 (since I was not expecting platforms to have
> *both* DT and DMI information).
> 
> I'm guessing the Raspberry Pi is one such platform, and
> `/sys/devices/virtual/dmi` exists? Hybrid UEFI+ACPI+DT platform I take it?
> 
> If so, your commit description should probably be something like:
> 
> ===
> brcmfmac: Prefer DT board type over DMI board type
> 
> The introduction of support for Apple board types inadvertently changed
> the precedence order, causing hybrid ACPI+DT platforms to look up the
> firmware using the DMI information instead of the device tree compatible
> to generate the board type. Revert back to the old behavior,
> as affected platforms use firmwares named after the DT compatible.
> 
> Fixes: 7682de8b3351 ("wifi: brcmfmac: of: Fetch Apple properties")
> ===
> 
> An also add a Cc: stable@...r.kernel.org to make sure this gets backported.
> 
> With the fixed description,
> 
> Reviewed-by: Hector Martin <marcan@...can.st>
> 
> - Hector

Looking into this a bit more from what was mentioned in the linked bug,
the DMI data comes from the SMBIOS table. We don't have that on Apple
platforms even though we also boot via U-Boot+EFI, but I'm guessing you
build U-Boot with CONFIG_GENERATE_SMBIOS_TABLE and provide that stuff in
the DT? So s/ACPI/SMBIOS/ would be more accurate in the commit message.

- Hector