[<prev] [next>] [<thread-prev] [day] [month] [year] [list]
Message-ID: <CADYN=9JdZfduPYeM7KV8H18_mbVrKj1aYRSkeWrTprQnffCXNg@mail.gmail.com>
Date: Wed, 27 May 2020 08:28:00 +0200
From: Anders Roxell <anders.roxell@...aro.org>
To: Rob Herring <robh@...nel.org>
Cc: Nathan Chancellor <natechancellor@...il.com>,
Valdis Klētnieks <valdis.kletnieks@...edu>,
Geert Uytterhoeven <geert+renesas@...der.be>,
Magnus Damm <magnus.damm@...il.com>,
"open list:MEDIA DRIVERS FOR RENESAS - FCP"
<linux-renesas-soc@...r.kernel.org>,
Sebastian Reichel <sre@...nel.org>,
"linux-kernel@...r.kernel.org" <linux-kernel@...r.kernel.org>
Subject: Re: [PATCH] power: reset: vexpress: fix build issue
On Tue, 26 May 2020 at 17:41, Rob Herring <robh@...nel.org> wrote:
>
> " On Mon, May 25, 2020 at 11:38 PM Nathan Chancellor
> <natechancellor@...il.com> wrote:
> >
> > On Mon, May 25, 2020 at 07:37:45PM -0400, Valdis Klētnieks wrote:
> > > On Sun, 24 May 2020 15:20:25 -0700, Nathan Chancellor said:
> > >
> > > > arm-linux-gnueabi-ld: drivers/power/reset/vexpress-poweroff.o: in function `vexpress_reset_probe':
> > > > vexpress-poweroff.c:(.text+0x36c): undefined reference to `devm_regmap_init_vexpress_config'
> > >
> > > The part I can't figure out is that git blame tells me there's already an
> > > export:
> > >
> > > 3b9334ac835bb (Pawel Moll 2014-04-30 16:46:29 +0100 154) return regmap;
> > > 3b9334ac835bb (Pawel Moll 2014-04-30 16:46:29 +0100 155) }
> > > b33cdd283bd91 (Arnd Bergmann 2014-05-26 17:25:22 +0200 156) EXPORT_SYMBOL_GPL(devm_regmap_init_vexpress_config);
> > > 3b9334ac835bb (Pawel Moll 2014-04-30 16:46:29 +0100 157)
> > >
> > > but I can't figure out where or if drivers/power/reset/vexpress-poweroff.c gets
> > > a MODULE_LICENSE from...
> >
> > Correct, it is exported but that file is being built as a module whereas
> > the file requiring it is beign builtin. As far as I understand, that
> > will not work, hence the error.
> >
> > The issue with this patch is that ARCH_VEXPRESS still just selects
> > POWER_RESET_VEXPRESS, which ignores "depends on", hence the Kconfig
> > warning and not fixing the error.
> >
> > I am not that much of a Kconfig guru to come up with a solution. I am
> > just reporting it because arm allmodconfig is broken on -next due to
> > this.
>
> Commit "ARM: vexpress: Don't select VEXPRESS_CONFIG" needs to be
> reverted. I've asked Arnd to revert it.
Reverting that patch with my patch will make allmodconfig work for arm too.
>
> Anders patch is still needed for arm64.
Yes, it is still needed for arm64.
Cheers,
Anders
Powered by blists - more mailing lists