This is the mail archive of the
mailing list for the GCC project.
Re: improve unaligned loads with no special insns
On Sun, Nov 07, 2004 at 10:33:36AM +0100, Eric Botcazou wrote:
> It apparently pessimizes on 32-bit targets for 4-byte unaligned loads:
> 2 more insns.
Hum. I'm not sure how to gauge that then. We need both offset
addressing and zero-extending byte loads for this to be shorter.
Which is true for sparc but not for ia64.
> Probably wrong on big-endian. Something like:
> shift = BITS_PER_WORD - shift;
> might be needed.
Looking at it.