Subject: Re: Possible bug in arm32 strongarm optimisations.
To: Ignatios Souvatzis <is@beverly.kleinbus.org>
From: Richard Earnshaw <rearnsha@arm.com>
List: tech-toolchain
Date: 10/30/2000 12:01:40
> I think there's a mulhi instruction in newer ARMs; gcc could certainly abuse
> that to optimise division by a constant. (It does so on m68020-040).
I think you mean smul and umul (mulhi is a conditionally executed 32-bit
multiply). Yes, -march=armv3m (or armv4) causes the compiler to use these
for division by a constant.
R.