Received-SPF: pass (google.com: domain of linux-kernel-owner@vger.kernel.org designates 23.128.96.31 as permitted sender) client-ip=23.128.96.31;
MIME-Version: 1.0
References: <20231129172200.430674-1-sjg@chromium.org> <20231129172200.430674-3-sjg@chromium.org>
 <ZWhFtcrZV2FYQ_D6@bergen.fjasle.eu>
In-Reply-To: <ZWhFtcrZV2FYQ_D6@bergen.fjasle.eu>
From:   Masahiro Yamada <masahiroy@kernel.org>
Date:   Fri, 1 Dec 2023 00:12:37 +0900
Message-ID: <CAK7LNARNw0czgf-YE5mvA__tCGi=6wi-03p9LfJVEBfJZ_EYqg@mail.gmail.com>
Subject: Re: [PATCH v7 2/2] arm64: boot: Support Flat Image Tree
To:     Nicolas Schier <nicolas@fjasle.eu>
Cc:     Simon Glass <sjg@chromium.org>,
        linux-arm-kernel@lists.infradead.org,
        Tom Rini <trini@konsulko.com>,
        lkml <linux-kernel@vger.kernel.org>,
        U-Boot Mailing List <u-boot@lists.denx.de>,
        Catalin Marinas <catalin.marinas@arm.com>,
        Nathan Chancellor <nathan@kernel.org>,
        Nick Terrell <terrelln@fb.com>, Will Deacon <will@kernel.org>,
        linux-kbuild@vger.kernel.org
Content-Type: text/plain; charset="UTF-8"
Content-Transfer-Encoding: quoted-printable
Precedence: bulk

On Thu, Nov 30, 2023 at 5:26=E2=80=AFPM Nicolas Schier <nicolas@fjasle.eu> =
wrote:
>
> Simon,
>
> thanks for the patch!  Below are some nitpicks and bike-shedding
> questions.
>
> On Wed 29 Nov 2023 10:21:53 GMT, Simon Glass wrote:
> > Add a script which produces a Flat Image Tree (FIT), a single file
> > containing the built kernel and associated devicetree files.
> > Compression defaults to gzip which gives a good balance of size and
> > performance.
> >
> > The files compress from about 86MB to 24MB using this approach.
> >
> > The FIT can be used by bootloaders which support it, such as U-Boot
> > and Linuxboot. It permits automatic selection of the correct
> > devicetree, matching the compatible string of the running board with
> > the closest compatible string in the FIT. There is no need for
> > filenames or other workarounds.
>
> Have you thought about updating the arch/mips ITB rules to also use the
> new scripts/make_fit.py?  Or is the FIT/ITB format for mips different
> from the one for arm64?


I recommend not touching MIPS at this moment
because this tool simply picks up *.dtb files
that exist under arch/*/boot/dts/, some of which
may be stale files.


Think of this scenario:


[1] Enable CONFIG_ARCH_FOO and build

   foo.dtb

will be created.


[2] Next, disable CONFIG_ARCH_FOO and
    enable CONFIG_ARCH_BAR, and build.

   bar.dtb

will be created.


This script will pick up both foo.dtb and bar.dtb
although foo.dtb is a left-over from the previous build.


Without cleaning, stale *.dtb will accumulate
and unwanted files will be included in image.fit.


Currently, MIPS hard-codes its files.
It always works in a deterministic way.


I do not request Simon to implement everything perfectly
because I know that would require much more effort.

We could do something like modules.order to list out the
dtb files from the current build, but I am not asking for it
in this patchset.


But, you are right. This tool is not arm64-specific at all
(and that is the reason why I think the MAINTAINERS
entry is a little odd)
Perhaps it can be applicable to MIPS after everything
works correctly.


> >  ARM ARCHITECTED TIMER DRIVER
> >  M:   Mark Rutland <mark.rutland@arm.com>
> >  M:   Marc Zyngier <maz@kernel.org>
> > diff --git a/arch/arm64/Makefile b/arch/arm64/Makefile
> > index 1bd4fae6e806..18e092de7cdb 100644
> > --- a/arch/arm64/Makefile
> > +++ b/arch/arm64/Makefile
> > @@ -36,6 +36,8 @@ ifeq ($(CONFIG_BROKEN_GAS_INST),y)
> >  $(warning Detected assembler with broken .inst; disassembly will be un=
reliable)
> >  endif
> >
> > +KBUILD_DTBS      :=3D dtbs
>
> Might you want to use tabs here as in the lines below?


This should not exist in the first place.


  image.fit: dtbs


is better.


--
Best Regards
Masahiro Yamada