Commit 9605da8a by Brad Lucier Committed by Richard Henderson

toplev.c: Allow -f[no-]math-errno to set (clear) flag_errno_math

        * toplev.c: Allow -f[no-]math-errno to set (clear) flag_errno_math
        * invoke.texi: Document this change.

From-SVN: r29449
parent fa51b01b
1999-09-15 Brad Lucier <lucier@math.purdue.edu>
* toplev.c: Allow -f[no-]math-errno to set (clear) flag_errno_math
* invoke.texi: Document this change.
Wed Sep 15 17:56:00 1999 Richard Henderson <rth@cygnus.com> Wed Sep 15 17:56:00 1999 Richard Henderson <rth@cygnus.com>
* emit-rtl.c (free_emit_status): Don't check DECL_DEFER_OUTPUT. * emit-rtl.c (free_emit_status): Don't check DECL_DEFER_OUTPUT.
......
...@@ -154,7 +154,7 @@ in the following sections. ...@@ -154,7 +154,7 @@ in the following sections.
-falign-jumps=@var{n} -fbranch-probabilities -falign-jumps=@var{n} -fbranch-probabilities
-fcaller-saves -fcse-follow-jumps -fcse-skip-blocks -fcaller-saves -fcse-follow-jumps -fcse-skip-blocks
-fdelayed-branch -fexpensive-optimizations -fdelayed-branch -fexpensive-optimizations
-ffast-math -ffloat-store -fforce-addr -fforce-mem -ffast-math -ffloat-store -fforce-addr -fforce-mem -fno-math-errno
-fdata-sections -ffunction-sections -fgcse -fdata-sections -ffunction-sections -fgcse
-finline-functions -finline-limit=@var{n} -fkeep-inline-functions -finline-functions -finline-limit=@var{n} -fkeep-inline-functions
-fmove-all-movables -fno-default-inline -fno-defer-pop -fmove-all-movables -fno-default-inline -fno-defer-pop
...@@ -2409,6 +2409,15 @@ This option should never be turned on by any @samp{-O} option since ...@@ -2409,6 +2409,15 @@ This option should never be turned on by any @samp{-O} option since
it can result in incorrect output for programs which depend on it can result in incorrect output for programs which depend on
an exact implementation of IEEE or ANSI rules/specifications for an exact implementation of IEEE or ANSI rules/specifications for
math functions. math functions.
@item -fno-math-errno
Do not set ERRNO after calling math functions that are executed
with a single instruction, e.g., sqrt. A program that relies on
IEEE exceptions for math error handling may want to use this flag
for speed while maintaining IEEE arithmetic compatibility.
The default is @samp{-fmath-errno}. The @samp{-ffast-math} option
sets @samp{-fno-math-errno}.
@end table @end table
@c following causes underfulls.. they don't look great, but we deal. @c following causes underfulls.. they don't look great, but we deal.
......
...@@ -969,7 +969,9 @@ lang_independent_options f_options[] = ...@@ -969,7 +969,9 @@ lang_independent_options f_options[] =
{"ident", &flag_no_ident, 0, {"ident", &flag_no_ident, 0,
"Process #ident directives"}, "Process #ident directives"},
{ "peephole2", &flag_peephole2, 1, { "peephole2", &flag_peephole2, 1,
"Enables an rtl peephole pass run before sched2" } "Enables an rtl peephole pass run before sched2" },
{"math-errno", &flag_errno_math, 1,
"Set errno after built-in math functions"}
}; };
#define NUM_ELEM(a) (sizeof (a) / sizeof ((a)[0])) #define NUM_ELEM(a) (sizeof (a) / sizeof ((a)[0]))
......
Markdown is supported
0% or
You are about to add 0 people to the discussion. Proceed with caution.
Finish editing this message first!
Please register or to comment