Age | Commit message (Collapse) | Author | Files | Lines |
|
2001-02-02 Phil Edwards <pme@sources.redhat.com>
* mkcheck.in: Also limit virtual memory size, for mmap-based mallocs.
From-SVN: r39416
|
|
gcc_gxx_include_dir.
* Makefile.in (gcc_gxx_target_include_dir): Generate dynamically
as for gcc_gxx_include_dir.
From-SVN: r39415
|
|
From-SVN: r39414
|
|
2001-02-02 Phil Edwards <pme@sources.redhat.com>
* COPYING.LIB: Update to LGPL 2.1 from the FSF.
From-SVN: r39413
|
|
* cp-tree.h (build_expr_ptr_wrapper, can_free): Remove.
* tree.c (build_expr_ptr_wrapper, can_free, permanent_obstack):
Remove.
* call.c (convert_class_to_reference, build_user_type_conversion_1,
add_warning): Change build_expr_ptr_wrapper to build_ptr_wrapper.
From-SVN: r39412
|
|
arm-*-pe
From-SVN: r39411
|
|
* config/sh/sh.c (output_far_jump): Don't use braf on SH1. Emit
label before alignment to be used as the braf base address.
* config/sh/sh.md (length): Use longer lengths for SH1 PIC far
branches.
(casesi_jump_2): Require at least TARGET_SH2.
From-SVN: r39410
|
|
-m3e and -m4-single-only.
* config/float-sh.h (LDBL_EPSILON, LDBL_MIN, LDBL_MAX): Prevent
truncation to 32-bit doubles on -m3e and -m4-single-only.
From-SVN: r39409
|
|
* gcc.c (init_gcc_specs): New function. Make -shared-libgcc
the default when building a shared object.
(init_spec): Use it.
* testsuite/lib/g++.exp: Include the directory where libgcc
is located to the LD_LIBRARY_PATH list.
* inovke.texi (-shared-libgcc): Document the cases in which
* Make-lang.in (g++spec.o): Add DRIVER_DEFINES to the list
of macros used when compiling g++spec.c.
* g++spec.c (lang_specific_driver): Link with the shared
libgcc by default.
From-SVN: r39408
|
|
* config/fp-bit.h (MAX_USI_INT, MAX_SI_INT): Don't assume
unsigned is wide enough.
(BITS_PER_SI): New.
* config/fp-bit.c (_fpmul_parts): Don't assume 32-bits SI.
(si_to_float): Likewise.
(float_to_si, float_to_usi): Use BITS_PER_SI.
From-SVN: r39407
|
|
From-SVN: r39406
|
|
2001-02-01 Jeffrey Oldham <oldham@codesourcery.com>
* dwarf2out.c: Added explanation of abbreviations.
(def_cfa_1): Added comments listing DWARF2 instructions.
(cfa_temp_reg): Removed in favor of cfa_temp.
(cfa_temp_value): Removed in favor of cfa_temp.
(cfa_temp): New global variable.
(dwarf2out_frame_debug_expr): Added extensive introductory
comments explaining the function's transformations. Revised to
use cfa_temp. Added some rtx checking. Generalize IOR case.
(dwarf2out_frame_debug): Revised to use cfa_temp.
(output_aranges): Cast as "unsigned" to avoid warning.
* rtl.texi (RTX_FRAME_RELATED_P): Revise entry to emphasize better
explain which instructions must be marked.
From-SVN: r39405
|
|
* local-alloc.c (update_equiv_regs): Copy INSN_CODE to the
new instruction.
From-SVN: r39400
|
|
* cp-demangle.c (NAMESPACE_SEPARATOR): New define.
(struct demangling_def): Add `style' field.
(demangling_new): New parameter `style'. Set it in demangling_t.
(demangle_prefix): Use NAMESPACE_SEPARATOR.
(demangle_type_ptr): Don't emit pointer symbol if doing Java output.
(cp_demangle): New parameter `style'. Pass it to demangling_new().
(main): Call cp_demangle with extra parameter.
(java_demangle_v3): New function.
(java_builtin_type_names): New. Table of primitive type names used
for Java demangling.
(demangle_builtin_type): Look up in java_builtin_type_names if doing
Java output.
* cplus-dem.c (cplus_demangle): Use java_demangle_v3 to do Java
demangling.
(long_options): Remove obsolete `java' option.
(main): Remove explicit handling of `java' option. Instead, pass style
parameter in cplus_demangle flags as gdb does.
* testsuite/demangle.expected: Add some Java test cases.
From-SVN: r39399
|
|
* jvgenmain.c (class_mangling_suffix): Remove unused string.
(error): Remove unused function.
(main): Don't use "__attribute__ alias" on generated class symbol.
From-SVN: r39397
|
|
* config/sh/crt1.asm (main_k): Modified so that start calls
___setup_argv_and_call_main.
From-SVN: r39396
|
|
* configure.in (toplevel_srcdir, auxdir): Set.
* acinclude.m4 (glibcpp_basedir): Set based on auxdir.
* aclocal.m4, configure, Makefile.in: Rebuilt.
* libio/Makefile.in, math/Makefile.in, src/Makefile.in: Likewise.
* libsupc++-v3/Makefile.am (INCLUDES): Use toplevel_srcdir.
* libsupc++-v3/Makefile.in: Rebuilt.
* libsupc++-v3/configure.in, libsupc++-v3/configure: Removed.
* libsupc++-v3/aclocal.m4, libsupc++-v3/config.h.in: Likewise.
From-SVN: r39395
|
|
From-SVN: r39394
|
|
* cpphash.h (struct cpp_buffer): Move saved_flags from cpp_reader.
* cpplex.c (_cpp_lex_token): New token picks up the saved flags,
and AVOID_LPASTE is cleared on meeting an unescaped newline.
* cppmacro.c (builtin_macro): Set builtin flags here.
(paste_all_tokens): Preserve AVOID_LPASTE on pasted token.
(replace_args): Clarify intent.
(cpp_get_token): Macro expansion tokens get the saved flags.
Update.
* cppmain.c (scan_buffer): Remove now-redundant print.printed
check.
From-SVN: r39393
|
|
2001-02-01 Jeffrey Oldham <oldham@codesourcery.com>
* config/mips/iris6.h (SUPPORTS_INIT_PRIORITY): Reverse change of
2001-Jan-29.
From-SVN: r39392
|
|
Pmode pseudo, but we need ptr_mode.
* builtins.c (expand_builtin_alloca): allocate_dynamic_stack_space
returns Pmode pseudo, but we need ptr_mode.
* explow.c (allocate_dynamic_stack_space): Use plus_constant.
Remove bogus conversions; use HOST_WIDE_INT for size.
Don't use TARGET if wrong mode.
From-SVN: r39390
|
|
non-default even if size is zero.
* config/i386/i386.c (ix86_compute_frame_size): Allow
stack_alignment_needed to be non-default even if size is zero.
From-SVN: r39389
|
|
2001-01-30 Alexandre Petit-Bianco <apbianco@cygnus.com>
* jcf-parse.c (init_jcf_parse): Added cast to ggc_add_root's last
argument.
* parse.y (finish_method_declaration): Code accounting for WFLed
method DECL_NAMEs deleted.
(check_abstract_method_definitions): Likewise.
(resolve_type_during_patch): Layout resolved type.
* typeck.c (lookup_do): Removed unused local.
(http://gcc.gnu.org/ml/gcc-patches/2001-02/msg00005.html)
From-SVN: r39388
|
|
From-SVN: r39387
|
|
* (hppa_init_pic_save): Emit the pic offset table
reg save after last_parm_insn.
* pa.c (hppa_init_pic_save): New function.
* pa.h (hppa_init_pic_save): Declare.
* pa.md (call, call_value, sibcall, sibcall_value): Use
the above instead of duplicated code.
From-SVN: r39386
|
|
* pa.c (hppa_encode_label): Correct size of alloca buffer
so we don't overrun it. Correct leading `*' case.
* pa.h (STRIP_NAME_ENCODING): Simplify now that we don't
need to handle `*@'.
(FUNCTION_NAME_P): Likewise.
From-SVN: r39385
|
|
From-SVN: r39384
|
|
* config/float-i128.h (FLT_EVAL_METHOD, DECIMAL_DIG): Define.
* config/float-i32.h: Likewise.
* config/float-i64.h: Likewise.
* config/float-sh.h: Likewise.
* config/float-sparc.h: Likewise.
From-SVN: r39383
|
|
From-SVN: r39382
|
|
a whole register from itself.
* expmed.c (extract_bit_field): allow non-integral modes if we
want to extract a whole register from itself.
From-SVN: r39381
|
|
not a range at all.
* c-typeck.c (set_init_index): If first is equal to last, assume as
if it was not a range at all.
* gcc.dg/gnu99-init-1.c: Add test for [0 ... 0] range.
From-SVN: r39380
|
|
* Makefile.in: Rebuilt.
* Makefile.am (LTCXXCOMPILE): New macro.
From-SVN: r39379
|
|
2001-01-31 Jeffrey Oldham <oldham@codesourcery.com>
* gcc.dg/c99-float-1.c: Back out "XFAIL FLT_EVAL_METHOD and
DECIMAL_DIG cases for Irix."
From-SVN: r39378
|
|
2001-01-31 Jeffrey Oldham <oldham@codesourcery.com>
* gcc.dg/c99-float-1.c: Back out "XFAIL FLT_EVAL_METHOD and
DECIMAL_DIG cases for Irix."
From-SVN: r39377
|
|
From-SVN: r39376
|
|
From-SVN: r39375
|
|
From-SVN: r39374
|
|
* config/fp-bit.c: Include hconfig.h
(float_to_si): Don't assume SImode is 32-bits wide.
(float_to_usi): Likewise.
From-SVN: r39373
|
|
* c-parse.in (structsp): Pedwarn when "enum foo" refers to an
incomplete type.
(typename): Call pending_xref_error after parsing typed_typespecs.
* c-decl.c (lookup_tag): Give error immediately rather than
leaving it pending if the tag of the wrong type is in the same
binding level.
(xref_tag): Don't pedwarn for forward declarations of enum types
here.
* gcc.texi (Actual Bugs): Remove entry for misuse of struct, union
and enum tags.
testsuite:
* gcc.dg/c99-tag-1.c: New test.
From-SVN: r39372
|
|
From-SVN: r39371
|
|
From-SVN: r39370
|
|
From-SVN: r39369
|
|
* cpplib.c (T_ELIF): Move to STDC89, not that it matters.
(_cpp_handle_directive): Suggest not using #elif with -Wtraditional.
From-SVN: r39368
|
|
* gcc.dg/cpp/avoidpaste1.c: Update.
* gcc.dg/cpp/paste4.c: Update.
From-SVN: r39367
|
|
* cpphash.h (struct cpp_reader): New saved_flags.
* cppmacro.c (cpp_get_token): Use saved_flags to remember
to avoid a paste after a pasted token.
From-SVN: r39366
|
|
* libgcc2.h (Wtype_MAX, Wtype_MIN): Define.
* libgcc2.c (__fixunssfSI, __fixunsdfSI, __fixunsxfSI): Use
Wtype_MIN instead of LONG_MIN.
From-SVN: r39365
|
|
From-SVN: r39363
|
|
2001-01-30 Jeffrey Oldham <oldham@codesourcery.com>
* fixinc/fixincl.tpl: Add 2001 to copyright. Change name of
README file.
* fixinc/fixincl.x: Regenerated.
* fixinc/inclhack.def (irix_limits_const): New definition to
convert "const" to "__const" in limits.h.
From-SVN: r39362
|
|
2001-01-30 Jeffrey Oldham <oldham@codesourcery.com>
* fixinc/genfixes (AG): Fix "autogen4" thinko.
From-SVN: r39361
|
|
2001-01-30 Jeffrey Oldham <oldham@codesourcery.com>
* gcc.dg/c99-float-1.c: XFAIL FLT_EVAL_METHOD and DECIMAL_DIG
cases for Irix.
From-SVN: r39360
|