This is the mail archive of the
gcc@gcc.gnu.org
mailing list for the GCC project.
Re: expand and truncate and 387
- To: moshier at mediaone dot net
- Subject: Re: expand and truncate and 387
- From: Joern Rennecke <amylaar at cygnus dot co dot uk>
- Date: Tue, 1 Dec 1998 23:17:38 +0000 (GMT)
- Cc: rth at cygnus dot com, hubicka at atrey dot karlin dot mff dot cuni dot cz, law at cygnus dot com, egcs at cygnus dot com, pcg at goof dot com
> Even assuming that there is a subroutine call, how would that allow
> eliminating the convert insn? According to this scenario the insn has
> just been moved to the other side of the subroutine call. The value
> to be converted, which used to be in a register, is now in memory, so
> farther from being a no-op than it was before. The assertion is that
> according to IEEE the conversion is still an operation, it still might
> do something, and it still cannot be deleted.
In order to use the converted value, you have to load it from memory -
which in turn will raise the signal again.
So the only problem arises if the converted value is not used.