[BZ #4775, BZ #4776]

2007-07-12  Jakub Jelinek  <jakub@redhat.com>
	[BZ #4775]
	* math/tgmath.h (__tgmath_real_type_sub): Formatting.
	(__tgmath_real_type): Fix if expr is const int or other const
	qualified integral type.
	(__TGMATH_UNARY_REAL_ONLY): Rewritten to avoid using statement
	expressions and handle const qualified arguments.
	(__TGMATH_BINARY_FIRST_REAL_ONLY, __TGMATH_UNARY_REAL_IMAG,
	__TGMATH_UNARY_REAL_IMAG_RET_REAL): Likewise.
	(__TGMATH_UNARY_REAL_RET_ONLY): Rewritten to avoid using
	statement expressions.
	(__TGMATH_BINARY_REAL_ONLY, __TGMATH_TERNARY_FIRST_SECOND_REAL_ONLY,
	__TGMATH_TERNARY_REAL_ONLY, __TGMATH_BINARY_REAL_IMAG): Likewise.
	(__TGMATH_UNARY_IMAG): Define.
	(conj, cproj): Use __TGMATH_UNARY_IMAG macro.
	* math/Makefile (tests): Add test-tgmath2.
	(CFLAGS-test-tgmath2.c): Add.
	* math/test-tgmath.c (fy, dy, ly, fz, dz, lz, count_cdouble,
	count_cfloat, count_cldouble): New variables.
	(NCCALLS): Define.
	(main): Check number of complex calls as well.
	(F(compile_test)): Add complex tests and tests with const qualified
	arguments.
	(y, z, ccount): Define.
	(F(cacos), F(casin), F(catan), F(ccos), F(csin), F(ctan), F(cacosh),
	F(casinh), F(catanh), F(ccosh), F(csinh), F(ctanh), F(cexp), F(clog),
	F(csqrt), F(cpow), F(cabs), F(carg), F(creal), F(cimag), F(conj),
	F(cproj)): New functions.
	* math/test-tgmath2.c: New test.

2007-07-11  Jakub Jelinek  <jakub@redhat.com>

	[BZ #4776]
	* elf/dl-load.c (_dl_rtld_di_serinfo): Output / in LD_LIBRARY_PATH,
	RPATH etc. as "/" rather than "", don't segfault on empty paths,
	instead output ".".
	* dlfcn/Makefile (distribute): Add glreflib3.c.
	(module-names): Add glreflib3.
	($(objpfx)tst-dlinfo.out): Depend on glreflib3.so rather than
	glreflib1.so.
	(LDFLAGS_glreflib3.so): New.
	* dlfcn/tst-dlinfo.c (do_test): Load glreflib3.so instead of
	glreflib1.so.
	* dlfcn/glreflib3.c: New file.

	* intl/finddomain.c (_nl_find_domain): If _nl_explode_name
	returned -1, return NULL.
	* intl/explodename.c (_nl_explode_name): Return -1 if
	_nl_normalize_codeset failed.

1 files changed, 488 insertions, 0 deletions

diff --git a/math/test-tgmath2.c b/math/test-tgmath2.c
new file mode 100644
index 0000000000..edb723c95e
--- /dev/null
+++ b/math/test-tgmath2.c
@@ -0,0 +1,488 @@
+/* Test compilation of tgmath macros.
+   Copyright (C) 2007 Free Software Foundation, Inc.
+   This file is part of the GNU C Library.
+   Contributed by Jakub Jelinek <jakub@redhat.com>, 2007.
+
+   The GNU C Library is free software; you can redistribute it and/or
+   modify it under the terms of the GNU Lesser General Public
+   License as published by the Free Software Foundation; either
+   version 2.1 of the License, or (at your option) any later version.
+
+   The GNU C Library is distributed in the hope that it will be useful,
+   but WITHOUT ANY WARRANTY; without even the implied warranty of
+   MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the GNU
+   Lesser General Public License for more details.
+
+   You should have received a copy of the GNU Lesser General Public
+   License along with the GNU C Library; if not, write to the Free
+   Software Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA
+   02111-1307 USA.  */
+
+#ifndef HAVE_MAIN
+#undef __NO_MATH_INLINES
+#define __NO_MATH_INLINES 1
+#include <math.h>
+#include <complex.h>
+#include <stdio.h>
+#include <string.h>
+#include <tgmath.h>
+
+//#define DEBUG
+
+typedef complex float cfloat;
+typedef complex double cdouble;
+#ifndef NO_LONG_DOUBLE
+typedef long double ldouble;
+typedef complex long double cldouble;
+#else
+typedef double ldouble;
+typedef complex double cldouble;
+#endif
+
+float vfloat1, vfloat2, vfloat3;
+double vdouble1, vdouble2, vdouble3;
+ldouble vldouble1, vldouble2, vldouble3;
+cfloat vcfloat1, vcfloat2, vcfloat3;
+cdouble vcdouble1, vcdouble2, vcdouble3;
+cldouble vcldouble1, vcldouble2, vcldouble4;
+int vint1, vint2, vint3;
+long int vlong1, vlong2, vlong3;
+long long int vllong1, vllong2, vllong3;
+const float Vfloat1 = 1, Vfloat2 = 2, Vfloat3 = 3;
+const double Vdouble1 = 1, Vdouble2 = 2, Vdouble3 = 3;
+const ldouble Vldouble1 = 1, Vldouble2 = 2, Vldouble3 = 3;
+const cfloat Vcfloat1 = 1, Vcfloat2 = 2, Vcfloat3 = 3;
+const cdouble Vcdouble1 = 1, Vcdouble2 = 2, Vcdouble3 = 3;
+const cldouble Vcldouble1 = 1, Vcldouble2 = 2, Vcldouble4 = 3;
+const int Vint1 = 1, Vint2 = 2, Vint3 = 3;
+const long int Vlong1 = 1, Vlong2 = 2, Vlong3 = 3;
+const long long int Vllong1 = 1, Vllong2 = 2, Vllong3 = 3;
+enum
+  {
+    Tfloat = 0,
+    Tcfloat,
+    Tdouble,
+    Tcdouble,
+#ifndef NO_LONG_DOUBLE
+    Tldouble,
+    Tcldouble,
+#else
+    Tldouble = Tdouble,
+    Tcldouble = Tcdouble,
+#endif
+    Tlast
+  };
+enum
+  {
+    C_cos = 0,
+    C_fabs,
+    C_cabs,
+    C_conj,
+    C_expm1,
+    C_lrint,
+    C_ldexp,
+    C_atan2,
+    C_remquo,
+    C_pow,
+    C_fma,
+    C_last
+  };
+int count;
+int counts[Tlast][C_last];
+
+int
+test (const int Vint4, const long long int Vllong4)
+{
+  int result = 0;
+  int quo = 0;
+
+#define FAIL(str) \
+  do								\
+    {								\
+      printf ("%s failure on line %d\n", (str), __LINE__);	\
+      result = 1;						\
+    }								\
+  while (0)
+#define TEST_TYPE_ONLY(expr, rettype) \
+  do								\
+    {								\
+      __typeof__ (expr) texpr = 0;				\
+      __typeof__ (rettype) ttype = 0, *ptype;			\
+      if (sizeof (expr) != sizeof (rettype))			\
+	FAIL ("type");						\
+      if (__alignof__ (expr) != __alignof__ (rettype))		\
+	FAIL ("type");						\
+      __asm ("" : "=r" (ptype) : "0" (&ttype), "r" (&texpr));	\
+      if (&texpr == ptype)					\
+	FAIL ("type");						\
+    }								\
+  while (0)
+#define TEST2(expr, type, rettype, fn) \
+  do								\
+    {								\
+      __typeof__ (expr) texpr = 0;				\
+      TEST_TYPE_ONLY (expr, rettype);				\
+      if (count != 0)						\
+	FAIL ("internal error");				\
+      if (counts[T##type][C_##fn] != 0)				\
+	FAIL ("internal error");				\
+      texpr = expr;						\
+      __asm __volatile ("" : : "r" (&texpr));			\
+      if (count != 1 || counts[T##type][C_##fn] != 1)		\
+	{							\
+	  FAIL ("wrong function called");			\
+	  memset (counts, 0, sizeof (counts));			\
+	}							\
+      count = 0;						\
+      counts[T##type][C_##fn] = 0;				\
+    }								\
+  while (0)
+#define TEST(expr, type, fn) TEST2(expr, type, type, fn)
+
+  TEST (cos (vfloat1), float, cos);
+  TEST (cos (vdouble1), double, cos);
+  TEST (cos (vldouble1), ldouble, cos);
+  TEST (cos (vint1), double, cos);
+  TEST (cos (vllong1), double, cos);
+  TEST (cos (vcfloat1), cfloat, cos);
+  TEST (cos (vcdouble1), cdouble, cos);
+  TEST (cos (vcldouble1), cldouble, cos);
+  TEST (cos (Vfloat1), float, cos);
+  TEST (cos (Vdouble1), double, cos);
+  TEST (cos (Vldouble1), ldouble, cos);
+  TEST (cos (Vint1), double, cos);
+  TEST (cos (Vllong1), double, cos);
+  TEST (cos (Vcfloat1), cfloat, cos);
+  TEST (cos (Vcdouble1), cdouble, cos);
+  TEST (cos (Vcldouble1), cldouble, cos);
+
+  TEST (fabs (vfloat1), float, fabs);
+  TEST (fabs (vdouble1), double, fabs);
+  TEST (fabs (vldouble1), ldouble, fabs);
+  TEST (fabs (vint1), double, fabs);
+  TEST (fabs (vllong1), double, fabs);
+  TEST (fabs (vcfloat1), float, cabs);
+  TEST (fabs (vcdouble1), double, cabs);
+  TEST (fabs (vcldouble1), ldouble, cabs);
+  TEST (fabs (Vfloat1), float, fabs);
+  TEST (fabs (Vdouble1), double, fabs);
+  TEST (fabs (Vldouble1), ldouble, fabs);
+#ifndef __OPTIMIZE__
+  /* GCC is too smart to optimize these out.  */
+  TEST (fabs (Vint1), double, fabs);
+  TEST (fabs (Vllong1), double, fabs);
+#else
+  TEST_TYPE_ONLY (fabs (vllong1), double);
+  TEST_TYPE_ONLY (fabs (vllong1), double);
+#endif
+  TEST (fabs (Vint4), double, fabs);
+  TEST (fabs (Vllong4), double, fabs);
+  TEST (fabs (Vcfloat1), float, cabs);
+  TEST (fabs (Vcdouble1), double, cabs);
+  TEST (fabs (Vcldouble1), ldouble, cabs);
+
+  TEST (conj (vfloat1), cfloat, conj);
+  TEST (conj (vdouble1), cdouble, conj);
+  TEST (conj (vldouble1), cldouble, conj);
+  TEST (conj (vint1), cdouble, conj);
+  TEST (conj (vllong1), cdouble, conj);
+  TEST (conj (vcfloat1), cfloat, conj);
+  TEST (conj (vcdouble1), cdouble, conj);
+  TEST (conj (vcldouble1), cldouble, conj);
+  TEST (conj (Vfloat1), cfloat, conj);
+  TEST (conj (Vdouble1), cdouble, conj);
+  TEST (conj (Vldouble1), cldouble, conj);
+  TEST (conj (Vint1), cdouble, conj);
+  TEST (conj (Vllong1), cdouble, conj);
+  TEST (conj (Vcfloat1), cfloat, conj);
+  TEST (conj (Vcdouble1), cdouble, conj);
+  TEST (conj (Vcldouble1), cldouble, conj);
+
+  TEST (expm1 (vfloat1), float, expm1);
+  TEST (expm1 (vdouble1), double, expm1);
+  TEST (expm1 (vldouble1), ldouble, expm1);
+  TEST (expm1 (vint1), double, expm1);
+  TEST (expm1 (vllong1), double, expm1);
+  TEST (expm1 (Vfloat1), float, expm1);
+  TEST (expm1 (Vdouble1), double, expm1);
+  TEST (expm1 (Vldouble1), ldouble, expm1);
+  TEST (expm1 (Vint1), double, expm1);
+  TEST (expm1 (Vllong1), double, expm1);
+
+  TEST2 (lrint (vfloat1), float, long int, lrint);
+  TEST2 (lrint (vdouble1), double, long int, lrint);
+  TEST2 (lrint (vldouble1), ldouble, long int, lrint);
+  TEST2 (lrint (vint1), double, long int, lrint);
+  TEST2 (lrint (vllong1), double, long int, lrint);
+  TEST2 (lrint (Vfloat1), float, long int, lrint);
+  TEST2 (lrint (Vdouble1), double, long int, lrint);
+  TEST2 (lrint (Vldouble1), ldouble, long int, lrint);
+  TEST2 (lrint (Vint1), double, long int, lrint);
+  TEST2 (lrint (Vllong1), double, long int, lrint);
+
+  TEST (ldexp (vfloat1, 6), float, ldexp);
+  TEST (ldexp (vdouble1, 6), double, ldexp);
+  TEST (ldexp (vldouble1, 6), ldouble, ldexp);
+  TEST (ldexp (vint1, 6), double, ldexp);
+  TEST (ldexp (vllong1, 6), double, ldexp);
+  TEST (ldexp (Vfloat1, 6), float, ldexp);
+  TEST (ldexp (Vdouble1, 6), double, ldexp);
+  TEST (ldexp (Vldouble1, 6), ldouble, ldexp);
+  TEST (ldexp (Vint1, 6), double, ldexp);
+  TEST (ldexp (Vllong1, 6), double, ldexp);
+
+#define FIRST(x, y) (y, x)
+#define SECOND(x, y) (x, y)
+#define NON_LDBL_TEST(fn, argm, arg, type, fnt) \
+  TEST (fn argm (arg, vfloat1), type, fnt); \
+  TEST (fn argm (arg, vdouble1), type, fnt); \
+  TEST (fn argm (arg, vint1), type, fnt); \
+  TEST (fn argm (arg, vllong1), type, fnt); \
+  TEST (fn argm (arg, Vfloat1), type, fnt); \
+  TEST (fn argm (arg, Vdouble1), type, fnt); \
+  TEST (fn argm (arg, Vint1), type, fnt); \
+  TEST (fn argm (arg, Vllong1), type, fnt);
+#define NON_LDBL_CTEST(fn, argm, arg, type, fnt) \
+  NON_LDBL_TEST(fn, argm, arg, type, fnt); \
+  TEST (fn argm (arg, vcfloat1), type, fnt); \
+  TEST (fn argm (arg, vcdouble1), type, fnt); \
+  TEST (fn argm (arg, Vcfloat1), type, fnt); \
+  TEST (fn argm (arg, Vcdouble1), type, fnt);
+#define BINARY_TEST(fn, fnt) \
+  TEST (fn (vfloat1, vfloat2), float, fnt); \
+  TEST (fn (Vfloat1, vfloat2), float, fnt); \
+  TEST (fn (vfloat1, Vfloat2), float, fnt); \
+  TEST (fn (Vfloat1, Vfloat2), float, fnt); \
+  TEST (fn (vldouble1, vldouble2), ldouble, fnt); \
+  TEST (fn (Vldouble1, vldouble2), ldouble, fnt); \
+  TEST (fn (vldouble1, Vldouble2), ldouble, fnt); \
+  TEST (fn (Vldouble1, Vldouble2), ldouble, fnt); \
+  NON_LDBL_TEST (fn, FIRST, vldouble2, ldouble, fnt); \
+  NON_LDBL_TEST (fn, SECOND, vldouble2, ldouble, fnt); \
+  NON_LDBL_TEST (fn, FIRST, Vldouble2, ldouble, fnt); \
+  NON_LDBL_TEST (fn, SECOND, Vldouble2, ldouble, fnt); \
+  NON_LDBL_TEST (fn, FIRST, vdouble2, double, fnt); \
+  NON_LDBL_TEST (fn, SECOND, vdouble2, double, fnt); \
+  NON_LDBL_TEST (fn, FIRST, Vdouble2, double, fnt); \
+  NON_LDBL_TEST (fn, SECOND, Vdouble2, double, fnt); \
+  NON_LDBL_TEST (fn, FIRST, vint2, double, fnt); \
+  NON_LDBL_TEST (fn, SECOND, vint2, double, fnt); \
+  NON_LDBL_TEST (fn, FIRST, Vint2, double, fnt); \
+  NON_LDBL_TEST (fn, SECOND, Vint2, double, fnt); \
+  NON_LDBL_TEST (fn, FIRST, vllong2, double, fnt); \
+  NON_LDBL_TEST (fn, SECOND, vllong2, double, fnt); \
+  NON_LDBL_TEST (fn, FIRST, Vllong2, double, fnt); \
+  NON_LDBL_TEST (fn, SECOND, Vllong2, double, fnt);
+#define BINARY_CTEST(fn, fnt) \
+  BINARY_TEST (fn, fnt); \
+  TEST (fn (vcfloat1, vfloat2), cfloat, fnt); \
+  TEST (fn (Vcfloat1, vfloat2), cfloat, fnt); \
+  TEST (fn (vcfloat1, Vfloat2), cfloat, fnt); \
+  TEST (fn (Vcfloat1, Vfloat2), cfloat, fnt); \
+  TEST (fn (vcldouble1, vldouble2), cldouble, fnt); \
+  TEST (fn (Vcldouble1, vldouble2), cldouble, fnt); \
+  TEST (fn (vcldouble1, Vldouble2), cldouble, fnt); \
+  TEST (fn (Vcldouble1, Vldouble2), cldouble, fnt); \
+  TEST (fn (vcfloat1, vfloat2), cfloat, fnt); \
+  TEST (fn (Vcfloat1, vfloat2), cfloat, fnt); \
+  TEST (fn (vcfloat1, Vfloat2), cfloat, fnt); \
+  TEST (fn (Vcfloat1, Vfloat2), cfloat, fnt); \
+  TEST (fn (vcldouble1, vldouble2), cldouble, fnt); \
+  TEST (fn (Vcldouble1, vldouble2), cldouble, fnt); \
+  TEST (fn (vcldouble1, Vldouble2), cldouble, fnt); \
+  TEST (fn (Vcldouble1, Vldouble2), cldouble, fnt); \
+  TEST (fn (vcfloat1, vcfloat2), cfloat, fnt); \
+  TEST (fn (Vcfloat1, vcfloat2), cfloat, fnt); \
+  TEST (fn (vcfloat1, Vcfloat2), cfloat, fnt); \
+  TEST (fn (Vcfloat1, Vcfloat2), cfloat, fnt); \
+  TEST (fn (vcldouble1, vcldouble2), cldouble, fnt); \
+  TEST (fn (Vcldouble1, vcldouble2), cldouble, fnt); \
+  TEST (fn (vcldouble1, Vcldouble2), cldouble, fnt); \
+  TEST (fn (Vcldouble1, Vcldouble2), cldouble, fnt); \
+  NON_LDBL_CTEST (fn, FIRST, vcldouble2, cldouble, fnt); \
+  NON_LDBL_CTEST (fn, SECOND, vcldouble2, cldouble, fnt); \
+  NON_LDBL_CTEST (fn, FIRST, Vcldouble2, cldouble, fnt); \
+  NON_LDBL_CTEST (fn, SECOND, Vcldouble2, cldouble, fnt); \
+  NON_LDBL_CTEST (fn, FIRST, vcdouble2, cdouble, fnt); \
+  NON_LDBL_CTEST (fn, SECOND, vcdouble2, cdouble, fnt); \
+  NON_LDBL_CTEST (fn, FIRST, Vcdouble2, cdouble, fnt); \
+  NON_LDBL_CTEST (fn, SECOND, Vcdouble2, cdouble, fnt);
+
+  BINARY_TEST (atan2, atan2);
+
+#define my_remquo(x, y) remquo (x, y, &quo)
+  BINARY_TEST (my_remquo, remquo);
+#undef my_remquo
+
+  BINARY_CTEST (pow, pow);
+
+  /* Testing all arguments of fma would be just too expensive,
+     so test just some.  */
+#define my_fma(x, y) fma (x, y, vfloat3)
+  BINARY_TEST (my_fma, fma);
+#undef my_fma
+#define my_fma(x, y) fma (x, vfloat3, y)
+  BINARY_TEST (my_fma, fma);
+#undef my_fma
+#define my_fma(x, y) fma (Vfloat3, x, y)
+  BINARY_TEST (my_fma, fma);
+#undef my_fma
+  TEST (fma (vdouble1, Vdouble2, vllong3), double, fma);
+  TEST (fma (vint1, Vint2, vint3), double, fma);
+  TEST (fma (Vldouble1, vldouble2, Vldouble3), ldouble, fma);
+  TEST (fma (vldouble1, vint2, Vdouble3), ldouble, fma);
+
+  return result;
+}
+
+int
+main (void)
+{
+  return test (vint1, vllong1);
+}
+
+/* Now generate the three functions.  */
+#define HAVE_MAIN
+
+#define F(name) name
+#define TYPE double
+#define CTYPE cdouble
+#define T Tdouble
+#define C Tcdouble
+#include "test-tgmath2.c"
+
+#define F(name) name##f
+#define TYPE float
+#define CTYPE cfloat
+#define T Tfloat
+#define C Tcfloat
+#include "test-tgmath2.c"
+
+#ifndef NO_LONG_DOUBLE
+#define F(name) name##l
+#define TYPE ldouble
+#define CTYPE cldouble
+#define T Tldouble
+#define C Tcldouble
+#include "test-tgmath2.c"
+#endif
+
+#else
+
+#ifdef DEBUG
+#define P() puts (__FUNCTION__); count++
+#else
+#define P() count++;
+#endif
+
+TYPE
+(F(cos)) (TYPE x)
+{
+  counts[T][C_cos]++;
+  P ();
+  return x;
+}
+
+CTYPE
+(F(ccos)) (CTYPE x)
+{
+  counts[C][C_cos]++;
+  P ();
+  return x;
+}
+
+TYPE
+(F(fabs)) (TYPE x)
+{
+  counts[T][C_fabs]++;
+  P ();
+  return x;
+}
+
+TYPE
+(F(cabs)) (CTYPE x)
+{
+  counts[T][C_cabs]++;
+  P ();
+  return x;
+}
+
+CTYPE
+(F(conj)) (CTYPE x)
+{
+  counts[C][C_conj]++;
+  P ();
+  return x;
+}
+
+TYPE
+(F(expm1)) (TYPE x)
+{
+  counts[T][C_expm1]++;
+  P ();
+  return x;
+}
+
+long int
+(F(lrint)) (TYPE x)
+{
+  counts[T][C_lrint]++;
+  P ();
+  return x;
+}
+
+TYPE
+(F(ldexp)) (TYPE x, int y)
+{
+  counts[T][C_ldexp]++;
+  P ();
+  return x + y;
+}
+
+TYPE
+(F(atan2)) (TYPE x, TYPE y)
+{
+  counts[T][C_atan2]++;
+  P ();
+  return x + y;
+}
+
+TYPE
+(F(remquo)) (TYPE x, TYPE y, int *z)
+{
+  counts[T][C_remquo]++;
+  P ();
+  return x + y + *z;
+}
+
+TYPE
+(F(pow)) (TYPE x, TYPE y)
+{
+  counts[T][C_pow]++;
+  P ();
+  return x + y;
+}
+
+CTYPE
+(F(cpow)) (CTYPE x, CTYPE y)
+{
+  counts[C][C_pow]++;
+  P ();
+  return x + y;
+}
+
+TYPE
+(F(fma)) (TYPE x, TYPE y, TYPE z)
+{
+  counts[T][C_fma]++;
+  P ();
+  return x + y + z;
+}
+
+#undef F
+#undef TYPE
+#undef CTYPE
+#undef T
+#undef C
+#undef P
+#endif