linux-kernel - Re: [PATCH] arm64: dts: qcom: msm8998: Node ordering, address cleanups

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: <2165bbd3-aa60-bec0-4ee7-dfb7dc1dd1ad@free.fr>
Date:   Wed, 24 Jul 2019 13:16:50 +0200
From:   Marc Gonzalez <marc.w.gonzalez@...e.fr>
To:     Jeffrey Hugo <jeffrey.l.hugo@...il.com>,
        Andy Gross <agross@...nel.org>,
        Bjorn Andersson <bjorn.andersson@...aro.org>
Cc:     Rob Herring <robh+dt@...nel.org>,
        Mark Rutland <mark.rutland@....com>,
        MSM <linux-arm-msm@...r.kernel.org>,
        DT <devicetree@...r.kernel.org>,
        LKML <linux-kernel@...r.kernel.org>
Subject: Re: [PATCH] arm64: dts: qcom: msm8998: Node ordering, address
 cleanups

On 22/07/2019 18:58, Jeffrey Hugo wrote:

> DT nodes should be ordered by address, then node name, and finally label.
> The msm8998 dtsi does not follow this, so clean it up by reordering the
> nodes.  While we are at it, extend the addresses to be fully 32-bits wide
> so that ordering is easy to determine when adding new nodes.  Also, two
> or so nodes had the wrong address value in their node name (did not match
> the reg property), so fix those up as well.
> 
> Hopefully going forward, things can be maintained so that a cleanup like
> this is not needed.
> 
> Signed-off-by: Jeffrey Hugo <jeffrey.l.hugo@...il.com>
> ---
>  arch/arm64/boot/dts/qcom/msm8998.dtsi | 254 +++++++++++++-------------
>  1 file changed, 127 insertions(+), 127 deletions(-)

LGTM.

Reviewed-by: Marc Gonzalez <marc.w.gonzalez@...e.fr>

Rob, Mark: when there are multiple reg properties, why is the convention
to use the *first* address in the node's name, rather than the lowest
address?

e.g.

		spmi_bus: spmi@...f000 {
			compatible = "qcom,spmi-pmic-arb";
			reg =	<0x0800f000 0x1000>,
				<0x08400000 0x1000000>,
				<0x09400000 0x1000000>,
				<0x0a400000 0x220000>,
				<0x0800a000 0x3000>;
			reg-names = "core", "chnls", "obsrvr", "intr", "cnfg";

"spmi@...f000" instead of "spmi@...a000"

Especially, since the reg props could be in any order here, given the
lookup by name.

Regards.