2000-07-15 Gabriel Dos Reis <gdr@codesourcery.com>
* bits/valarray_array.h (__valarray_get_storage): New function.
(_Array<>::Array): Use it.
* bits/std_valarray.h (valarray<>::valarray): Likewise.
(valarray<>::resize): Likewise. Tweak.
* src/valarray-inst.cc (__valarray_product): Tweak.
From-SVN: r35053
2000-07-07 Benjamin Kosnik <bkoz@soma.redhat.com>
Clean up configure/build longstanding issues.
* src/Makefile.am (AC_CXXFLAGS): Don't set @SECTION_LDFLAGS@ here.
(CXXLINK): Add here, instead.
* Makefile.am (check-install): Make sure mkcheck has execute privs.
(check): Same here.
* configure.in (AC_OUTPUT_COMMANDS): Remove hacks here.
* configure.in (AC_OUTPUT_COMMANDS): Link c_io_libio.h and
c_io_libio.cc instead of copying.
* acinclude.m4 (GLIBCPP_ENABLE_CSTDIO): Do link here.
* configure.in (AC_OUTPUT_COMMANDS): Link $THREADS.h and
c++threads.h instead of copying.
* acinclude.m4 (GLIBCPP_ENABLE_THREADS): Do link here.
* config/c_io_libio.cc (__basic_file::_M_open_mode): Consolidate.
* bits/basic_file.h: Declare.
* mkcheck.in (TESTS_FILE): Use -v instead of --version.
From-SVN: r34917
2000-07-03 scott snyder <snyder@fnal.gov>
* bits/locale_facets.tcc (_M_extract): Only figure out the base
from the input if base == 0.
* testsuite/27_io/istream_extractor_arith.cc: Test reading a
number with a leading `0' in hex mode.
* shadow/bits/std_cmath.h: Fix typo in _GLIBCPP_HAVE_CEILL test.
* mkinclosure: Change `==' to `=' in test.
2000-07-03 Chip Salzenberg <chip@valinux.com>
* src/Makefile.am (libio_headers): _G_config.h is found in srcdir,
not builddir.
* src/Makefile.in: Regenerate.
From-SVN: r34860
* bits/concept_checks.h
(__less_then_comparable_requirement_violation): Only check for <.
* testsuite/25_algorithms/lower_bound.cc: New file.
From-SVN: r34798
2000-06-29 Branko Cibej <branko.cibej@hermes.si>
* bits/std_memory.h (auto_ptr_ref): Reworked and defined only if
_GLIBCPP_RESOLVE_LIB_DEFECTS.
(auto_ptr): _M_ptr changed to void*.
(suto_ptr::get): Cast _M_ptr to element type.
(auto_ptr::auto_ptr(auto_ptr<Tp1>)): Test implicit convetsion.
(auto_ptr::~auto_ptr): Use this->get() instead of _M_ptr.
(auto_ptr::operator*): Likewise.
(auto_ptr::operator->): Likewise.
(auto_ptr::release): Likewise.
(auto_ptr::reset): Likewise.
(auto_ptr::auto_ptr(auto_ptr_ref)): Initialize from __ref._M_release.
(auto_ptr::operator auto_ptr_rev<_Tp1>): Updated.
Define nested auto_ptr_ref unless _GLIBCPP_RESOLVE_LIB_DEFECTS.
Define operator=(auto_ptr_ref) if _GLIBCPP_RESOLVE_LIB_DEFECTS.
From-SVN: r34792
2000-06-28 Gabriel Dos Reis <dosreis@cmla.ens-cachan.fr>
* bits/valarray_array.h (_Array_default_ctor<>::_S_do_it): Don't
forget the sizeof(_Tp) factor in call to memset().
(_Array_copy_ctor<>::_S_do_it): Likewise for memcpy().
From-SVN: r34757
2000-06-13 Brent Verner <brent@rcfile.org>
* bits/string.tcc (string::rfind): Fix.
* testsuite/21_strings/rfind.cc: New file.
2000-06-26 Anthony Williams <anthony@anthonyw.cjb.net>
* testsuite/21_strings/ctor_copy_dtor.cc: Fixed logic error.
2000-06-26 Branko Cibej <branko.cibej@hermes.si>
* testsuite/27_io/filebuf_members.cc (test_01): Fixed typos.
* mkcheck.in: Make the *.txt and *.tst files writable after
copying them to $TEST_DIR.
* testsuite/27_io/ostream_inserter_arith.cc: Renamed
__TEST_NUMPUT_VERBOSE to TEST_NUMPUT_VERBOSE.
Define TEST_NUMPUT_VERBOSE only if DEBUG_ASSERT.
From-SVN: r34719
2000-06-23 Benjamin Kosnik <bkoz@purist.soma.redhat.com>
* bits/fstream.tcc (basic_filebuf::basic_filebuf(fd)): Use it.
(basic_filebuf::close): Use it.
* bits/std_fstream.h (basic_filebuf): Remove default arguments for
filebuf ctor.
(basic_filebuf): Add _M_fileno_based data member.
As reported by brent verner <brent@rcfile.org>
* testsuite/27_io/filebuf_members.cc: New file.
* testsuite/27_io/filebuf_members-1.tst: New file.
From-SVN: r34673
2000-06-22 Steven King <sxking@uswest.net>
* acinclude.m4: Check for sinl, _sinl using 1 parameter.
2000-06-22 Benjamin Kosnik <bkoz@purist.soma.redhat.com>
* acinclude.m4 (GLIBCPP_CHECK_BUILTIN_MATH_SUPPORT): Redo. Use the
same approach as GLIBCPP_CHECK_MATH_SUPPORT, only don't try to
link against libm and don't use -fno-builtins...
* math/mathconf.h: Change up builtin macros to new schema.
* bits/std_cmath.h: And here. Wheee!
* shadow/bits/std_cmath.h: And here.
* acconfig.h: Change up builtin macros here.
* math/cosf.c: Remove.
* math/fabsf.c: Likewise.
* math/sinf.c: Likewise.
* math/sqrtf.c: Likewise.
* math/Makefile.am (EXTRA_DIST): Remove those four files.
* math/Makefile.in: Regenerate.
From-SVN: r34655
2000-06-20 Anthony Williams <anthony@anthonyw.cjb.net>
* bits/string.tcc: Fix find.
* testsuite/21_strings/find.cc: Patch.
2000-06-20 Benjamin Kosnik <bkoz@purist.soma.redhat.com>
* docs/install.html: Update where to get automake, where to get
binutils.
* acinclude.m4: Need -fno-builtins too, so gcc doesn't recognize
builtins.
From-SVN: r34622
2000-06-13 Steven King <sxking@uswest.net>
* acinclude.m4: Fixup some of the builtin math tests and add tests for
__builtin_fmod* and test libm for fmodf, fmodl or _fmodf, _fmodl.
* acconfig.h: Add entries for fmod*.
2000-06-13 Branko Cibej <branko.cibej@hermes.si>
* bits/std_cmath.h: Fix typos in tests (*_FMODFF -> *_FMODF).
Test *_MODFF not *_MODF for modf(float, float*).
(modf(float, float*)): Remove reference to _C_legacy.
From-SVN: r34539
2000-06-06 Steven King <sxking@uswest.net>
* acconfig.h: Added defines for HAVE_BUILTIN math functs for float,
double and long double. Added defines for HAVE libm math functs for
float and long double.
* aclocal.m4: Added configure checks for builtin math funcs and libm
support for float and long double versions of the math functions.
* config.h.in: Added undefs for math functs.
* configure: Regenerate.
* math/mathconf: Changed the _GLIBCPP_HAS_BUILTIN_* to
_GLIBCPP_HAVE_BUILTIN_*.
* bits/std_cmath.h: Added long double support. Use builtins for
float, double and long if available, otherwise, use libm versions if
availible, otherwise, punt.
* shadow/bits/std_cmath.h: ditto
From-SVN: r34436
2000-05-31 Branko Cibej <branko.cibej@hermes.si>
* bits/limits_generic.h (numeric_limits<wchar_t>): Use WCHAR_MIN
and WCHAR_MAX instead of WCHART_MIN and WCHART_MAX.
From-SVN: r34327
2000-05-10 Benjamin Kosnik <bkoz@redhat.com>
* bits/std_cmath.h: Tweaks.
* math/mathconf.h: Tweaks and fixes for HP-UX 11.
(sqrtf): Define away iff !builtin and !in <math.h>.
(sinf): Same.
(cosf): Same.
(fabsf): Same.
At some point this directory should be converted to c++, the
autoconf tests should be run by the c++ compiler (not c), and
<cmath> should be used instead of math.h.
Move declaration of nan() here.
* math/complex-stub.h (cabsl): Remove nan() declaration.
Finish up FreeBSD4.0 support.
* config/bsd/ctype.cc: Scope out toupper, tolower calls.
* config/generic/ctype.cc: And here.
* testsuite/21_strings/char_traits.cc (test02): Guard with
_GLIBCPP_USE_WCHAR_T.
via Phil Edwards <pme@sourceware.cygnus.com>
* bits/std_cctype.h: Remove _GLIBCPP_USE_NAMESPACES.
* acconfig.h: And here.
* acinclude.m4: Same.
* testsuite/27_io/istream.cc: And here.
* testsuite/27_io/ostream.cc: And here.
From-SVN: r33835
* bits/char_traits.h: use wchar_t utility functions for
char_traits<wchar_t> methods.
* testsuite/21_string/char_traits.cc: New (test02): test
char_traits<wchar_t>
2000-05-08 Benjamin Kosnik <bkoz@cygnus.com>
* acinclude.m4 (GLIBCPP_CXXFLAGS): Add bits for solaris2.8.
From-SVN: r33783
2000-05-08 Benjamin Kosnik <bkoz@cygnus.com>
* bits/fstream.tcc (filebuf::_M_init_filebuf): Don't set
_M_buf_size based on macro, instead use _M_buf_size_opt.
* bits/std_streambuf.h (basic_streambuf): Add _M_buf_size_opt.
(basic_streambuf()): Set _M_buf_size_opt.
* testsuite/27_io/filebuf.cc (filebuf): Use _M_buf_size_opt
instead of _M_buf_size.
From-SVN: r33768
2000-04-26 Nathan C. Myers <ncm@cantrip.org>
* bits/basic_string.h: include <bits/atomicity.h> instead
of <atomicity.h>, and use the uglified names.
* config/cpu/*/atomicity.h: replace with bits/atomicity.h;
uglify names, eliminate dependence on <inttypes.h>.
* src/Makefile.in, src/Makefile.am: refer to correct place
for atomicity.h header.
2000-04-25 Steven King <sxking@uswest.net>
* bits/basic_string.h: Fix syntax error.
2000-04-25 Benjamin Kosnik <bkoz@redhat.com>
* acinclude.m4 (GLIBCPP_CHECK_MATH_SUPPORT): Improve checks for
USE_LONG_DOUBLE.
(GLIBCPP_ENABLE_DEBUG): If alpha, use -gdwarf-2.
* math/Makefile.am (EXTRA_LONG_DOUBLE_yes): Add all the long
versions to this as copysignl is needed anyway for these targets.
* math/Makefile.in: Regenerate.
* bits/streambuf.tcc (streambuf::xsgetn): Tweak.
* bits/fstream.tcc (underflow): Cast to libio types.
* bits/basic_file.h (seekoff): Fixes for alpha: use __c_streamoff.
* config/c_io_libio.cc (seekoff): And here.
From-SVN: r33501
