Subject: Re: Possible bug in arm32 strongarm optimisations.
To: David Brownlee <abs@netbsd.org>
From: Nicholas Clark <nick@ccl4.org>
List: port-arm32
Date: 10/26/2000 16:31:53
On Thu, Oct 26, 2000 at 05:13:38PM +0100, David Brownlee wrote:

> 	That looks pretty consistent with a quick browse of
> 	dist/gcc/config/arm/arm.[ch] :)
> 
>   {"armv3m",    PROCESSOR_NONE, (FL_CO_PROC | FL_FAST_MULT | FL_MODE32
>                                  | FL_MODE26)},
> 
>   {"strongarm", PROCESSOR_STARM, (FL_FAST_MULT | FL_MODE32 | FL_MODE26
>                                   | FL_ARCH4)},
> 
> 	Hmm... -marmv3m also implies FL_CO_PROC, which by default will
> 	compile for an FPE. Kernel code is always compiled soft-float,
> 	does the same apply for NetBSD/arm32 userland?
> 
> 	Are there any other interesting armv4 instructions?

I know of sign extending byte loads. However, as arm gcc has "char" as
"unsigned char" I'd assume this is only going to benefit code
generation for programs that explicitly specify values with type "signed
char". Or will it also help for signed shorts?

Nicholas Clark