On Thu, 16 Jun 2005, Geoffrey Keating wrote: > Although it does do some of this, -ffloat-store also has some rather nasty > side-effects, because of what it is actually documented to do: > > > @item -ffloat-store > > @opindex ffloat-store > > Do not store floating point variables in registers, and inhibit other > > options that might change whether a floating point value is taken from a > > register or memory.
But is this actually a useful thing for users to be able to control, other than for its effects in avoiding excess precision? If not - if the real uses of it are for the same purpose as gcc.c-torture/execute/ieee/ieee.exp uses it on x86, x86_64 -m32 and m68k, to avoid excess precision - then we can just repurpose this option and make it a no-op on systems which naturally don't need excess precision, and users using -ffloat-store right now to avoid excess precision will find their code gets improved correctness. -- Joseph S. Myers http://www.srcf.ucam.org/~jsm28/gcc/ [EMAIL PROTECTED] (personal mail) [EMAIL PROTECTED] (CodeSourcery mail) [EMAIL PROTECTED] (Bugzilla assignments and CCs)
