real.h (struct real_format): New member round_towards_zero.prereleases/gcc-4.3.2-rc1

ChangeLog:

	* real.h (struct real_format): New member round_towards_zero.
	* real.c (round_for_format): Respect fmt->round_towards_zero.
	(ieee_single_format, mips_single_format, motorola_single_format,
	spu_single_format, ieee_double_format, mips_double_format,
	motorola_double_format, ieee_extended_motorola_format,
	ieee_extended_intel_96_format, ieee_extended_intel_128_format,
	ieee_extended_intel_96_round_53_format, ibm_extended_format,
	mips_extended_format, ieee_quad_format, mips_quad_format,
	vax_f_format, vax_d_format, vax_g_format): Initialize it.
	* config/pdp11/pdp11.c (pdp11_f_format, pdp11_d_format): Likewise.

	* builtins.s (do_mpfr_arg1): Consider round_towards_zero member of
	real_format to choose rounding mode when calling MPFR functions.
	(do_mpfr_arg2, do_mpfr_arg3, do_mpfr_sincos): Likewise.
	(do_mpfr_bessel_n, do_mpfr_remquo, do_mpfr_lgamma_r): Likewise.

	* real.h (real_to_decimal_for_mode): Add prototype.
	* real.c (real_to_decimal_for_mode): Renames old real_to_decimal.
	Respect target rounding mode when generating decimal representation.
	(real_to_decimal): New stub for backwards compatibility.
	* c-cppbuiltin.c (builtin_define_with_hex_fp_value): Use
	real_to_decimal_for_mode instead of real_to_decimal.

	* config/spu/spu.md ("floatdisf2", "floatunsdisf2"): New.

	* config/spu/float_disf.c: New file.
	* config/spu/float_unsdisf.c: New file.
	* config/spu/t-elf (LIB2FUNCS_STATIC_EXTRA): Add them.
	(LIB2FUNCS_EXCLUDE): Define.

testsuite/ChangeLog:

	* gcc.c-torture/execute/20031003-1.c (main): Update test to
	accommodate SPU single-precision rounding mode.
	* gcc.c-torture/execute/conversion.c (test_float_to_integer,
	test_float_to_longlong_integer): Likewise.
	* gcc.c-torture/execute/ieee/rbug.c (main): Likewise.
	* gcc.dg/hex-round-1.c: Skip test on SPU.
	* gcc.dg/hex-round-2.c: Likewise.

	* gcc.dg/torture/fp-int-convert-float.c: Reenable test on SPU.
	* gcc.dg/torture/fp-int-convert-timode.c: Reenable "float" test on SPU.

From-SVN: r139240

1 files changed, 28 insertions, 14 deletions

diff --git a/gcc/builtins.c b/gcc/builtins.c
index 66d55733fe7..d61651e7728 100644
--- a/gcc/builtins.c
+++ b/gcc/builtins.c
@@ -12688,14 +12688,16 @@ do_mpfr_arg1 (tree arg, tree type, int (*func)(mpfr_ptr, mpfr_srcptr, mp_rnd_t),
 	  && (!min || real_compare (inclusive ? GE_EXPR: GT_EXPR , ra, min))
 	  && (!max || real_compare (inclusive ? LE_EXPR: LT_EXPR , ra, max)))
         {
-	  const int prec = REAL_MODE_FORMAT (TYPE_MODE (type))->p;
+	  const struct real_format *fmt = REAL_MODE_FORMAT (TYPE_MODE (type));
+	  const int prec = fmt->p;
+	  const mp_rnd_t rnd = fmt->round_towards_zero? GMP_RNDZ : GMP_RNDN;
 	  int inexact;
 	  mpfr_t m;
 
 	  mpfr_init2 (m, prec);
 	  mpfr_from_real (m, ra, GMP_RNDN);
 	  mpfr_clear_flags ();
-	  inexact = func (m, m, GMP_RNDN);
+	  inexact = func (m, m, rnd);
 	  result = do_mpfr_ckconv (m, type, inexact);
 	  mpfr_clear (m);
 	}
@@ -12730,7 +12732,9 @@ do_mpfr_arg2 (tree arg1, tree arg2, tree type,
 
       if (real_isfinite (ra1) && real_isfinite (ra2))
         {
-	  const int prec = REAL_MODE_FORMAT (TYPE_MODE (type))->p;
+	  const struct real_format *fmt = REAL_MODE_FORMAT (TYPE_MODE (type));
+	  const int prec = fmt->p;
+	  const mp_rnd_t rnd = fmt->round_towards_zero? GMP_RNDZ : GMP_RNDN;
 	  int inexact;
 	  mpfr_t m1, m2;
 
@@ -12738,7 +12742,7 @@ do_mpfr_arg2 (tree arg1, tree arg2, tree type,
 	  mpfr_from_real (m1, ra1, GMP_RNDN);
 	  mpfr_from_real (m2, ra2, GMP_RNDN);
 	  mpfr_clear_flags ();
-	  inexact = func (m1, m1, m2, GMP_RNDN);
+	  inexact = func (m1, m1, m2, rnd);
 	  result = do_mpfr_ckconv (m1, type, inexact);
 	  mpfr_clears (m1, m2, NULL);
 	}
@@ -12776,7 +12780,9 @@ do_mpfr_arg3 (tree arg1, tree arg2, tree arg3, tree type,
 
       if (real_isfinite (ra1) && real_isfinite (ra2) && real_isfinite (ra3))
         {
-	  const int prec = REAL_MODE_FORMAT (TYPE_MODE (type))->p;
+	  const struct real_format *fmt = REAL_MODE_FORMAT (TYPE_MODE (type));
+	  const int prec = fmt->p;
+	  const mp_rnd_t rnd = fmt->round_towards_zero? GMP_RNDZ : GMP_RNDN;
 	  int inexact;
 	  mpfr_t m1, m2, m3;
 
@@ -12785,7 +12791,7 @@ do_mpfr_arg3 (tree arg1, tree arg2, tree arg3, tree type,
 	  mpfr_from_real (m2, ra2, GMP_RNDN);
 	  mpfr_from_real (m3, ra3, GMP_RNDN);
 	  mpfr_clear_flags ();
-	  inexact = func (m1, m1, m2, m3, GMP_RNDN);
+	  inexact = func (m1, m1, m2, m3, rnd);
 	  result = do_mpfr_ckconv (m1, type, inexact);
 	  mpfr_clears (m1, m2, m3, NULL);
 	}
@@ -12819,7 +12825,9 @@ do_mpfr_sincos (tree arg, tree arg_sinp, tree arg_cosp)
 
       if (real_isfinite (ra))
         {
-	  const int prec = REAL_MODE_FORMAT (TYPE_MODE (type))->p;
+	  const struct real_format *fmt = REAL_MODE_FORMAT (TYPE_MODE (type));
+	  const int prec = fmt->p;
+	  const mp_rnd_t rnd = fmt->round_towards_zero? GMP_RNDZ : GMP_RNDN;
 	  tree result_s, result_c;
 	  int inexact;
 	  mpfr_t m, ms, mc;
@@ -12827,7 +12835,7 @@ do_mpfr_sincos (tree arg, tree arg_sinp, tree arg_cosp)
 	  mpfr_inits2 (prec, m, ms, mc, NULL);
 	  mpfr_from_real (m, ra, GMP_RNDN);
 	  mpfr_clear_flags ();
-	  inexact = mpfr_sin_cos (ms, mc, m, GMP_RNDN);
+	  inexact = mpfr_sin_cos (ms, mc, m, rnd);
 	  result_s = do_mpfr_ckconv (ms, type, inexact);
 	  result_c = do_mpfr_ckconv (mc, type, inexact);
 	  mpfr_clears (m, ms, mc, NULL);
@@ -12892,14 +12900,16 @@ do_mpfr_bessel_n (tree arg1, tree arg2, tree type,
 	  && real_isfinite (ra)
 	  && (!min || real_compare (inclusive ? GE_EXPR: GT_EXPR , ra, min)))
         {
-	  const int prec = REAL_MODE_FORMAT (TYPE_MODE (type))->p;
+	  const struct real_format *fmt = REAL_MODE_FORMAT (TYPE_MODE (type));
+	  const int prec = fmt->p;
+	  const mp_rnd_t rnd = fmt->round_towards_zero? GMP_RNDZ : GMP_RNDN;
 	  int inexact;
 	  mpfr_t m;
 
 	  mpfr_init2 (m, prec);
 	  mpfr_from_real (m, ra, GMP_RNDN);
 	  mpfr_clear_flags ();
-	  inexact = func (m, n, m, GMP_RNDN);
+	  inexact = func (m, n, m, rnd);
 	  result = do_mpfr_ckconv (m, type, inexact);
 	  mpfr_clear (m);
 	}
@@ -12933,7 +12943,9 @@ do_mpfr_remquo (tree arg0, tree arg1, tree arg_quo)
 
       if (real_isfinite (ra0) && real_isfinite (ra1))
         {
-	  const int prec = REAL_MODE_FORMAT (TYPE_MODE (type))->p;
+	  const struct real_format *fmt = REAL_MODE_FORMAT (TYPE_MODE (type));
+	  const int prec = fmt->p;
+	  const mp_rnd_t rnd = fmt->round_towards_zero? GMP_RNDZ : GMP_RNDN;
 	  tree result_rem;
 	  long integer_quo;
 	  mpfr_t m0, m1;
@@ -12942,7 +12954,7 @@ do_mpfr_remquo (tree arg0, tree arg1, tree arg_quo)
 	  mpfr_from_real (m0, ra0, GMP_RNDN);
 	  mpfr_from_real (m1, ra1, GMP_RNDN);
 	  mpfr_clear_flags ();
-	  mpfr_remquo (m0, &integer_quo, m0, m1, GMP_RNDN);
+	  mpfr_remquo (m0, &integer_quo, m0, m1, rnd);
 	  /* Remquo is independent of the rounding mode, so pass
 	     inexact=0 to do_mpfr_ckconv().  */
 	  result_rem = do_mpfr_ckconv (m0, type, /*inexact=*/ 0);
@@ -13010,7 +13022,9 @@ do_mpfr_lgamma_r (tree arg, tree arg_sg, tree type)
 	  && ra->cl != rvc_zero
 	  && !(real_isneg(ra) && real_isinteger(ra, TYPE_MODE (type))))
         {
-	  const int prec = REAL_MODE_FORMAT (TYPE_MODE (type))->p;
+	  const struct real_format *fmt = REAL_MODE_FORMAT (TYPE_MODE (type));
+	  const int prec = fmt->p;
+	  const mp_rnd_t rnd = fmt->round_towards_zero? GMP_RNDZ : GMP_RNDN;
 	  int inexact, sg;
 	  mpfr_t m;
 	  tree result_lg;
@@ -13018,7 +13032,7 @@ do_mpfr_lgamma_r (tree arg, tree arg_sg, tree type)
 	  mpfr_init2 (m, prec);
 	  mpfr_from_real (m, ra, GMP_RNDN);
 	  mpfr_clear_flags ();
-	  inexact = mpfr_lgamma (m, &sg, m, GMP_RNDN);
+	  inexact = mpfr_lgamma (m, &sg, m, rnd);
 	  result_lg = do_mpfr_ckconv (m, type, inexact);
 	  mpfr_clear (m);
 	  if (result_lg)