Kazu Hirata [Sat, 8 Oct 2005 18:17:20 +0000 (18:17 +0000)]
configure.in (arm-*-linux-gnueabi): Add to noconfigdirs target-libffi...
/
* configure.in (arm-*-linux-gnueabi): Add to noconfigdirs
target-libffi, target-qthreads, target-libjava, and
targetlibobjc.
* configure: Regenerate.
gcc/
Merge from csl-arm-branch:
2005-09-07 Paul Brook <paul@codesourcery.com>
* config/arm/linux-eabi.h (SUBTARGET_EXTRA_LINK_SPEC): Prepend a
space to string.
2005-03-07 Daniel Jacobowitz <dan@codesourcery.com>
* config/arm/arm.c (arm_all_abis): Add aapcs-linux.
(arm_override_options): Use TARGET_AAPCS_BASED.
* config/arm/arm.h (enum arm_abi_type): Add ARM_ABI_AAPCS_LINUX.
(PTRDIFF_TYPE): Use int for AAPCS.
(DEFAULT_SHORT_ENUMS): Use false for aapcs-linux.
* config/arm/linux-eabi.h (ARM_DEFAULT_ABI, WCHAR_TYPE): Define.
(DEFAULT_SHORT_ENUMS): Delete.
* doc/invoke.texi (ARM Options): Document -mabi=aapcs-linux.
2004-12-15 Daniel Jacobowitz <dan@codesourcery.com>
* config/arm/arm.h (DEFAULT_SHORT_ENUMS): Wrap in #ifndef.
* config/arm/linux-eabi.h (DEFAULT_SHORT_ENUMS): Define to 0.
* config/arm/t-linux-eabi (TARGET_LIBGCC2_CFLAGS): Set to -fPIC.
2004-12-03 Mark Mitchell <mark@codesourcery.com>
* config/arm/linux-eabi.h (LIBGCC_SPEC): Do not define.
2004-11-22 Mark Mitchell <mark@codesourcery.com>
* config.gcc (arm*-*-linux-gnueabi): Use __cxa_atexit.
2004-11-19 Mark Mitchell <mark@codesourcery.com>
* config.gcc (arm*-*-linux-gnueabi): Add it.
* config/arm/bpabi.h (FPUTYPE_DEFAULT): Undefine it before
redefining it.
(TARGET_OS_CPP_BUILTINS): Likeiwse.
* config/arm/linux-eabi.h: New file.
* config/arm/linux-elf.h (LINUX_TARGET_INTERPRETER): New macro.
(LINUX_TARET_LINK_SPEC): Likewise.
(LINK_SPEC): Use it.
* config/arm/t-linux-eabi: New file.
libcpp/
* configure.ac: Require 64-bit int for arm*-*-*eabi*.
* configure: Regenerate.
libstdc++/
Merge from csl-arm-branch:
2004-12-15 Daniel Jacobowitz <dan@codesourcery.com>
* libstdc++-v3/config/linker-map.gnu: Add ARM EABI symbols.
Janne Blomqvist [Fri, 7 Oct 2005 20:02:28 +0000 (23:02 +0300)]
re PR libfortran/16339 (Unformatted i/o on large arrays inefficient)
2005-10-07 Janne Blomqvist <jblomqvi@cc.hut.fi>
PR fortran/16339
PR fortran/23363
* io/io.h: Add read and write members to stream, define access
macros.
* io/transfer.c (read_block_direct): New function.
(write_block_direct): New function.
(unformatted_read): Change to use read_block_direct.
(unformatted_write): Change to use write_block_direct.
* io/unix.c: Remove mmap includes and defines.
(writen): Remove.
(readn): Remove.
(reset_stream): New function.
(do_read): New function.
(do_write): New function.
(fd_flush): Change to use do_write() instead of writen().
(fd_alloc_r_at): Change to use do_read().
(fd_seek): Change return type to try, as the prototype. Add check
to avoid syscall overhead if possible.
(fd_read): New function.
(fd_write): New function.
(fd_open): Set pointers for new functions.
(mem_read): New function.
(mem_write): New function.
(open_internal): Set pointers for new functions.
(is_seekable): Clean up comment.
re PR middle-end/24227 (ICE in compare_values, at tree-vrp.c:415)
2005-10-07 Richard Guenther <rguenther@suse.de>
PR middle-end/24227
* fold-const.c (fold_binary): Fix operand types during folding
of X op (A, Y). Evaluation order of the side-effects of
X and A are frontend-defined, so ensure we honour that even for
tcc_comparison class operands; eased by removing duplicate code.
Jerry DeLisle [Fri, 7 Oct 2005 17:01:48 +0000 (17:01 +0000)]
transfer.c (write_block): Add test for end-of-file condition, removed from mem_alloc_w_at.
2005-10-07 Jerry DeLisle <jvdelisle@verizon.net>
* io/transfer.c (write_block): Add test for end-of-file condition,
removed from mem_alloc_w_at. (next_record_w): Clean up checks for
NULL pointer returns from s_alloc_w.
* io/unix.c (mem_alloc_w_at): Remove call to generate_error end-of-file.
* io/write.c (write_float): Add checks for NULL pointer returns from
write_block calls. (write_integer): Same.
Paolo Carlini [Fri, 7 Oct 2005 15:55:17 +0000 (15:55 +0000)]
Fix libstdc++/24196 for ext/vstring/rc by returning to the behavior of basic_string pre-2003-06-13...
2005-10-07 Paolo Carlini <pcarlini@suse.de>
Fix libstdc++/24196 for ext/vstring/rc by returning to the behavior
of basic_string pre-2003-06-13; remove fully-dynamic-string stuff.
* include/ext/rc_string_base.h (_M_refcopy): Move inside the
_Rep class and remove code in macro.
(__rc_string_base()): Construct with _S_empty_rep()._M_refcopy().
(_M_dispose, _M_leak_hard): Remove code in macro.
(_S_construct): Return _S_empty_rep()._M_refcopy() for empty string.
(_M_grab): Adjust.
* include/ext/rc_string_base.h (_Rep::_M_refdata()): Minor tweak,
mark throw().
Erik Edelmann [Thu, 6 Oct 2005 15:44:01 +0000 (18:44 +0300)]
re PR fortran/18568 (pointers in derived data types do not transmit shape of pointed to arrays - bug or non-standard feature?)
2005-10-07 Erik Edelmann <erik.edelmann@iki.fi>
fortran/
PR 18568
* resolve.c (find_array_spec): Search through the list of
components in the symbol of the type instead of the symbol of the
variable.
testsuite/
PR 18568
* gfortran.dg/der_pointer_3.f90: New test.
re PR middle-end/24049 (compiler error: Segmentation fault In function 'DESX_CBCUpdate')
PR 24049
* passes.c (init_optimization_passes): Move pass_lower_vector_ssa
under pass_vectorize. Clear TODO_ggc_collect from the dce pass
under pass_vectorize.
* include/ext/sso_string_base.h (struct __sso_string_local):
Remove, actually POD types cannot have user defined constructors
(being aggregates) and therefore can always be members of unions.
(class __sso_string_base): Adjust consistently.
Paolo Carlini [Wed, 5 Oct 2005 19:11:18 +0000 (19:11 +0000)]
re PR libstdc++/24198 (Remove from the testsuite gnu_char_type (or fix it))
2005-10-05 Paolo Carlini <pcarlini@suse.de>
PR libstdc++/24198
* testsuite/27_io/basic_filebuf/3.cc: Use __gnu_test::pod_ushort
instead.
* testsuite/27_io/basic_filebuf/seekoff/10132-2.cc: Likewise.
* testsuite/27_io/basic_filebuf/seekpos/10132-3.cc: Likewise.
* testsuite/27_io/basic_fstream/3.cc: Likewise; run the test.
* testsuite/27_io/basic_ifstream/3.cc: Likewise; run the test.
* testsuite/27_io/basic_ios/3.cc: Likewise.
* testsuite/27_io/basic_iostream/3.cc: Likewise; run the test.
* testsuite/27_io/basic_istream/3.cc: Likewise; run the test.
* testsuite/27_io/basic_istringstream/3.cc: Likewise; run the test.
* testsuite/27_io/basic_ofstream/3.cc: Likewise; run the test.
* testsuite/27_io/basic_ostream/3.cc: Likewise; run the test.
* testsuite/27_io/basic_ostringstream/3.cc: Likewise; run the test.
* testsuite/27_io/basic_streambuf/3.cc: Likewise.
* testsuite/27_io/basic_stringbuf/3.cc: Likewise.
* testsuite/27_io/basic_stringstream/3.cc: Likewise; run the test.
re PR target/23602 (1081 test failures in libjava, when configured for i486-linux)
PR target/23602
* toplev.c (process_options): Warn about unsupported combinations
of unwind tables and omit-frame-pointer.
* config/i386/i386.c (override_options): Similarly. Enable
accumulate-outgoing-args if not explicitly disabled.
testsuite/
* g++.dg/eh/omit-frame-pointer.C: Remove i?86 specific options.
* g++.dg/eh/omit-frame-pointer2.C: Likewise.
libjava/
* configure.host (i?86-*): Set -fomit-frame-pointer in libgcj_flags,
but not BACKTRACESPEC.
(x86_64-*): Similarly. Don't set -ffloat-store in 64-bit mode.
Steve Ellcey [Wed, 5 Oct 2005 18:06:37 +0000 (18:06 +0000)]
vect.md (vec_initv2si): Fix typo of V2SF to V2SI.
* vect.md (vec_initv2si): Fix typo of V2SF to V2SI.
Handle big endian vs. small endian.
(vec_initv2sf): Handle big endian vs. small endian.
(*vec_extractv2sf_1): Ditto.
* c-typeck.c (readonly_error): Handle USE being lv_asm.
(build_asm_expr): Call it if outputs are read-only.
* gimplify.c (gimplify_asm_expr): Remove former fix to PR 21419.
Billy Biggs [Wed, 5 Oct 2005 08:27:44 +0000 (08:27 +0000)]
re PR target/23809 (Misleading documentation for -mmmx/msse/...)
2005-10-05 Billy Biggs <billy.biggs@gmail.com>
Paolo Bonzini <bonzini@gnu.org>
PR target/23809
* doc/extend.texi (x86 Built-ins): Document that -msse and friends
enable the instructions and not just the built-ins.
* doc/invoke.texi (x86 Options): Likewise.
Co-Authored-By: Paolo Bonzini <bonzini@gnu.org>
From-SVN: r104979
Paolo Bonzini [Wed, 5 Oct 2005 07:32:13 +0000 (07:32 +0000)]
re PR ada/22340 (Ada build fails with --enable-bootstrap)
2005-10-05 Paolo Bonzini <bonzini@gnu.org>
PR bootstrap/22340
* configure.in (default_target): Remove.
* Makefile.tpl (all): Do not use prerequisites as subroutines
(all) [gcc-bootstrap]: Bootstrap gcc first if it was not done yet.
(do-[+make_target+], check, install, [+bootstrap_target+]): Do not
use prerequisites as subroutines.
(check-host, check-target): New.
(bootstrap configure & all targets): Do not use stage*-start
if the directory layout is already ok.
(non-bootstrap configure & all targets): Prepend a $(unstage).
(stage[+id+]-bubble): Do that here. Do not use NOTPARALLEL.
(NOTPARALLEL): Remove.
(unstage, stage variables): New variables.
(unstage, stage targets): Simply expand to those variables.
James E Wilson [Tue, 4 Oct 2005 23:54:38 +0000 (16:54 -0700)]
This avoids small data/gp reloc overflow errors.
* configure.ac (mudflap_cv_entry_point): Use quadrigraphs to declare
$name as array of characters with unknown bound. Also store into the
array.
* configure: Regenerate.
Janis Johnson [Tue, 4 Oct 2005 21:45:50 +0000 (21:45 +0000)]
target-supports.exp (check_effective_target_fortran_large_real, [...]): Replace cached result when target changes.
* lib/target-supports.exp (check_effective_target_fortran_large_real,
check_effective_target_fortran_large_int): Replace cached result
when target changes.
2005-10-04 Adrian Straetling <straetling@de.ibm.com>
* config/s390/s390.md ("TDSI","DP"): New mode macros.
("TE","tg"): New mode attributes.
("sync_compare_and_swap<mode>"): Replace with a define_expand.
("sync_compare_and_swap<mode>_cc"): Replace GPR with TDSI.
("*sync_compare_and_swap<mode>_cc"): Replace with one pattern for
dword_mode and one for GPRmode.
re PR preprocessor/13726 (cpp -C -dI loses comments on same line as #include directives)
libcpp/
PR preprocessor/13726
* directives.c (check_eol_return_comments): New static function.
(parse_include): Add buf parameter. Change all callers.
(do_include_common): If not discard comments, turn on
save_comments. Pass collected comments to include callback.
* include/cpplib.h (struct cpp_callbacks): Add new parameter to
include callback: cpp_token list.
gcc/
PR preprocessor/13726
* c-ppoutput.c (cb_include): Add comments parameter, and print out
any comments passed in.
gcc/testsuite/
PR preprocessor/13726
* gcc.dg/cpp/cmdlne-dI-C.c: New test.
* gcc.dg/cpp/cmdlne-dI-C.h: New file.
Ulrich Weigand [Tue, 4 Oct 2005 16:14:52 +0000 (16:14 +0000)]
re PR middle-end/19382 (ACATS cxb4005 cxb5002 simple To_COBOL/To_Fortran test fails at runtime on s390-linux)
PR ada/19382
* builtins.c (fold_builtin_memcmp): When constructing the pointer
type used to access data in the inlined length == 1 case, use
build_pointer_type_for_mode with CAN_ALIAS_ALL set to true.
(fold_builtin_strcmp, fold_builtin_strncmp): Likewise.
Uros Bizjak [Tue, 4 Oct 2005 14:07:25 +0000 (16:07 +0200)]
i386.h (TARGET_FISTTP): Enable also for TARGET_SSE3 and only for TARGET_80387.
* config/i386/i386.h (TARGET_FISTTP): Enable also for
TARGET_SSE3 and only for TARGET_80387.
* config/i386/i386.md (fix_trunc<mode>_fisttp_i387_1,
(fix_trunc<mode>_i387_fisttp, fix_trunc<mode>_i387_fisttp_with_temp):
Do not depend on TARGET_80387.
Andrew Pinski [Tue, 4 Oct 2005 13:39:17 +0000 (06:39 -0700)]
re PR middle-end/23125 (OpenBSD's zic.c causes g++ but not gcc to segfault)
2005-10-04 Andrew Pinski <pinskia@physics.uc.edu>
PR middle-end/23125
* decl.c (make_rtl_for_nonlocal_decl): Use set_user_assembler_name
instead of change_decl_assembler_name.
2005-10-04 Andrew Pinski <pinskia@physics.uc.edu>
PR middle-end/23125
* c-decl.c (finish_decl): Use set_user_assembler_name even for
register variables.
* varasm.c (make_decl_rtl): If a register variable does not
have a set user assmbler name, error out.
Decode the asmspec is now name+1 bypassing '*'.
2005-10-04 Andrew Pinski <pinskia@physics.uc.edu>
PR middle-end/23125
* gcc.dg/register-var-3.c: New test.
Steven Bosscher [Tue, 4 Oct 2005 05:57:38 +0000 (05:57 +0000)]
re PR tree-optimization/23049 (ICE with -O3 -ftree-vectorize on 4.1.x)
2005-10-05 Steven Bosscher <stevenb@suse.de>
gcc/
PR tree-optimization/23049
* tree-ssa-dom.c (thread_across_edge): Make sure that the condition
of a COND_EXPR is folded before calling fold on the whole rhs of a
conditional assignment.
* doc/tree-ssa.texi: Update the GIMPLE grammar for a valid rhs to
document that a COND_EXPR may appear there.
Diego Novillo [Tue, 4 Oct 2005 03:02:19 +0000 (03:02 +0000)]
re PR tree-optimization/23445 (ICE with -O1 -ftree-vrp -fdelete-null-pointer-checks)
PR 23445
* tree-vrp.c (extract_range_from_assert): If the new numeric
range created out of the assertion contradicts the existing
numeric range of the ASSERT_EXPR variable, make the new range
varying.
testsuite/
PR 23445
* gcc.c-torture/compile/pr23445.c: New test.
projects / thirdparty / gcc.git / log
