linux-kernel - Re: [PATCH] arm64: dts: marvell: add DT for ESPRESSObin-Ultra

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] [day] [month] [year] [list]

Message-Id: <20201023161802.265575-1-vladimir.vid@sartura.hr>
Date:   Fri, 23 Oct 2020 18:18:02 +0200
From:   Vladimir Vid <vladimir.vid@...tura.hr>
To:     pali@...nel.org, a.heider@...il.com
Cc:     devicetree@...r.kernel.org, linux-kernel@...r.kernel.org,
        linux-arm-kernel@...ts.infradead.org, tmn505@...il.com,
        sebastian.hesselbarth@...il.com, gregory.clement@...tlin.com,
        andrew@...n.ch, jason@...edaemon.net, robh+dt@...nel.org,
        Vladimir Vid <vladimir.vid@...tura.hr>
Subject: Re: [PATCH] arm64: dts: marvell: add DT for ESPRESSObin-Ultra

On Fri, Oct 9, 2020 at 5:37 PM Pali Rohár wrote:
> Cannot you in this case include armada-3720-espressobin.dtsi and write
> only differences into DTS ultra file to reduce definitions?

> Currently there are pending patches for espressobin DTS files to
> simplify them and use common include file as much as possible to
> de-duplicate same parts.

Updated in v2.

> Looks like this is incorrect as on Armada 3720 at the beginning of the
> nor must be stored Marvell wtmi/secure firmware which is running on
> Cortex M3 processor. For sure there cannot be u-boot. Also on A3720 is
> u-boot loaded by ARM trusted firmware therefore I doubt that u-boot is
> stored at offset zero.

> I guess that on offset zero is stored concatenated binary of Marvell
> secure firmware for M3, ARM trusted firmware (for main CPU) and U-Boot.
> Can you verify it? In this case I would just call this partition as
> "firmware" to indicate what is stored here.

Correct. Concatenated binary you get from building u-boot and ATF, which
includes wtmi, is flashed to offset zero.
I agree that firmware is more appropriate name for the partition,
updated in v2.


On Mon, Oct 12, 2020 at 9:08 AM Andre Heider wrote:
> the other espressobin boards use "Globalscale Marvell ESPRESSOBin
> Board", so maybe "Globalscale Marvell ESPRESSOBin Ultra Board"?

Updated in v2.

>> > +                     partition@...000 {
>> > +                             label = "hw-info";
> any idea what's in there?

SPI is used insted of the EEPROM to save some hardware related info,
currently there are MAC addresses and serial number on that partition.

>> > +&usb3 {
>> > +     status = "okay";
>> > +     usb-phy = <&usb3_phy>;
>
> That node comes from armada-37xx.dtsi and has:
>
>        phys = <&comphy0 0>, <&usb2_utmi_otg_phy>;
>        phy-names = "usb3-phy", "usb2-utmi-otg-phy";
>
> Are those supposed to be mixed?

No, I don't think they should. I originally added dts for the 4.19
and there weren't usb2 related nodes there.
I'm disabling usb3 for the moment until I see what's going on,
as I'm currently working on adding u-boot support and a couple
of other things.

For the other comments, I think they are resolved by including
the espressobin dtsi instead of armada-37xx.