linux-kernel - Re: [GIT PULL] Devicetree updates for v5.16

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: <CAL_Jsq+HDXdJL69ybGBM16Nm4giS62ZP+5k3_OGsharrRj-PiQ@mail.gmail.com>
Date:   Mon, 8 Nov 2021 08:57:27 -0600
From:   Rob Herring <robh@...nel.org>
To:     Daniel Palmer <daniel@...f.com>
Cc:     Linus Torvalds <torvalds@...ux-foundation.org>,
        Linux Kernel Mailing List <linux-kernel@...r.kernel.org>,
        DTML <devicetree@...r.kernel.org>,
        Frank Rowand <frowand.list@...il.com>
Subject: Re: [GIT PULL] Devicetree updates for v5.16

On Mon, Nov 8, 2021 at 7:13 AM Daniel Palmer <daniel@...f.com> wrote:
>
> Hi Rob,
>
> On Wed, 3 Nov 2021 at 06:57, Rob Herring <robh@...nel.org> wrote:
> > Rob Herring (22):
> >       kbuild: Enable DT schema checks for %.dtb targets
>
> Sorry for replying to a pull request. I couldn't find the original
> patch email for this commit to reply.

FYI, the link is in the commit and 'b4' makes it easy to get the mbox
to reply to. It's quite nice to be able to reply to *anything*.

> Anyhow, the above commit makes buildroot unhappy.
> I see this when trying to build any of my buildroot environments:
> scripts/dtc/Makefile:23: *** dtc needs libyaml for DT schema
> validation support. Install the necessary libyaml development
> package..  Stop.

If you fix this, then the next thing you will get is the dtschema
python package is needed.

> I think this is some mess with pkg-config from somewhere (the host
> distro, the buildroot host side,..) saying libyaml is available but
> when the build actually happens the header it wants isn't actually
> where it thinks. I think this is a pre-existing problem with
> buildroot.
> buildroot takes a list of the dtbs that you want to build and keep in
> it's config and calls make with that list in the kernel build. So this
> commit causes this issue to trigger when buildroot tries to build the
> configured dtbs.
>
> Not sure what the fix is and I think it's probably a buildroot problem.

I suppose we can keep schema checks optional here, but my intention in
this patch was moving towards the checks being required rather than a
separate target. So we'd be kicking the problem down the road.

Rob