Re: Possible Enhancement - Forcing double prec. on Intel FPUs

[Craig Burley <burley at gnu dot org>]

(Thanks for looking into this...it's not easily grappled with!)

>The easiest way to make the Intel FPU behave as it is expected
>to, is to compile the code with -ffloat-store.
>That option ensures that ALL intermediate results are stored in their
>appropriate memory locations and thereby getting rounded properly.
>Using -ffloat-store in turn causes, by stalling the FPU pipeline,
>a large performance decrease. 

I believe this is not quite correct.  -ffloat-store applies only
to variables, i.e. any *explicit* stores, as seen in the source
code being compiled.  It does not apply to computing intermediate
results.

Therefore, I believe it will prevent use of extended precision
in code like

  a = b + c;
  d = a * e;

but not in code like:

  d = (b + c) * e;

Just FYI.  See the egcs/gcc docs for more info.

        tq vm, (burley)