Hi,
this is patch I ended up comitting after some furhter testing. The difference
to initial
version is that it now eanbles SSE math with -ffast-math too and it does so
outside the
ugly target macro.
Bootstrapped/regtested x86_64-linux, tested with -m32
Honza
* config/i386/i386.c (ix86_option_override_internal): Switch
to SSE math for -ffast-math when target ISA supports SSE2.
Index: config/i386/i386.c
===================================================================
--- config/i386/i386.c (revision 203252)
+++ config/i386/i386.c (working copy)
@@ -3769,6 +3769,19 @@ ix86_option_override_internal (bool main
}
}
}
+ /* For all chips supporting SSE2, -mfpmath=sse performs better than
+ fpmath=387. The second is however default at many targets since the
+ extra 80bit precision of temporaries is considered to be part of ABI.
+ Overwrite the default at least for -ffast-math.
+ TODO: -mfpmath=both seems to produce same performing code with bit
+ smaller binaries. It is however not clear if register allocation is
+ ready for this setting.
+ Also -mfpmath=387 is overall a lot more compact (bout 4-5%) than SSE
+ codegen. We may switch to 387 with -ffast-math for size optimized
+ functions. */
+ else if (fast_math_flags_set_p (&global_options)
+ && TARGET_SSE2)
+ ix86_fpmath = FPMATH_SSE;
else
ix86_fpmath = TARGET_FPMATH_DEFAULT;
