Re: [PATCH] arm, vt8500, LLVMLlinux: Use mcr instead of mcr% for mach-vt8500

[Behan Webster]

On 09/24/14 02:16, Arnd Bergmann wrote:
> On Tuesday 23 September 2014 20:44:44 Behan Webster wrote:
> > The ASM below does not compile with clang and is not the way that the mcr
> > command is used in other parts of the kernel.
> >
> > arch/arm/mach-vt8500/vt8500.c:72:11: error: invalid % escape in inline assembly string
> >          asm("mcr%? p15, 0, %0, c7, c0, 4" : : "r" (0));
> >              ~~~~~^~~~~~~~~~~~~~~~~~~~~~~~
> > 1 error generated.
> >
> > There are other forms that are supported on different ARM instruction sets but
> > generally the kernel just uses mcr as it is supported in all ARM instruction
> > sets.
> Just for confirm: both forms are actually correct and we don't need this
> backported for stable, right?
My understanding is that the %? carries a condition code to the next 
instruction (which in this case is then ignored). So essentially in this 
situation both are equivalent.

> > Signed-off-by: Behan Webster <behanw@xxxxxxxxxxxxxxxxxx>
> > Reviewed-by: Mark Charlebois <charlebm@xxxxxxxxx>
> > Acked-by: Will Deacon <will.deacon@xxxxxxx>
> Acked-by: Arnd Bergmann <arnd@xxxxxxxx>
>
> Tony, would you like to pick this one up and send it in a pull request
> to arm-soc, or should we apply it to fixes-non-critical directly?
Thanks,

Behan

--
Behan Webster
behanw@xxxxxxxxxxxxxxxxxx

--
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at  http://vger.kernel.org/majordomo-info.html
Please read the FAQ at  http://www.tux.org/lkml/