[tree-ssa] RFC: Never gimplify ASM_EXPRs

[Paul Koning]

>>>>> "Andrew" == Andrew Haley <aph@redhat.com> writes:

 Andrew> Richard Henderson writes:
 >> On Thu, Jun 05, 2003 at 04:35:56PM +0100, Andrew Haley wrote: >
 >> Diego Novillo writes: > > Is this an invalid use of __asm__ then?
 >> 
 >> I don't think so.
 >> 
 >> > I think so.  The asm is asking for a mem operand for something
 >> that it > never intends actually to access.
 >> 
 >> It DID access it.  With a prefetch instruction.

 Andrew> Not really; a prefetch instruction is more of a suggestion
 Andrew> than an access, is it not?

NO!  A prefetch instruction, in those architectures where it is
implemented, is NOT a suggestion, it's an order to do real work.  Just
because other implementations of the same architecture are allowed to
treat it as a NOP does not mean it is semantically equivalent to a
NOP.  We code prefetches because they make a very real -- and often
VERY large -- difference.

     paul