linux-kernel - Re: [PATCH] pwm,renesas-tpu: Add dependency on HAS

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: <20140226141243.GE29779@ulmo.nvidia.com>
Date:	Wed, 26 Feb 2014 15:12:44 +0100
From:	Thierry Reding <thierry.reding@...il.com>
To:	Richard Weinberger <richard@....at>
Cc:	linux-pwm@...r.kernel.org, linux-kernel@...r.kernel.org
Subject: Re: [PATCH] pwm,renesas-tpu: Add dependency on HAS_IOMEM

On Wed, Feb 26, 2014 at 03:06:31PM +0100, Richard Weinberger wrote:
> -----BEGIN PGP SIGNED MESSAGE-----
> Hash: SHA1
> 
> Am 26.02.2014 13:08, schrieb Thierry Reding:
> > On Fri, Jan 31, 2014 at 01:45:18PM +0100, Richard Weinberger wrote:
> >> On archs like S390 or um this driver cannot build nor work. Make it depend on HAS_IOMEM to bypass build failures.
> >> 
> >> drivers/built-in.o: In function `tpu_probe': drivers/pwm/pwm-renesas-tpu.c:421: undefined reference to `devm_ioremap_resource'
> >> 
> >> Signed-off-by: Richard Weinberger <richard@....at> --- drivers/pwm/Kconfig | 1 + 1 file changed, 1 insertion(+)
> > 
> > In the long run I think it might be better to add a dummy implementation for devm_ioremap_resource() in order to make it build properly even for architectures where it isn't
> > implemented.
> 
> And what should devm_ioremap_resource() or in general ioremap() do on archs without
> iomem?
> BUG()? Returning a random error and hoping that the driver handles it correctly? :)
> 
> IMHO such a dummy implementation does not make much sense.

Well, it could still be useful for drivers that want to support
COMPILE_TEST. If we just make it return any error drivers should simply
refuse to probe just like if the error was genuine.

Thierry

Content of type "application/pgp-signature" skipped