gcc/
2008-08-17 Daniel Jacobowitz <dan@codesourcery.com>
Richard Sandiford <rdsandiford@googlemail.com>
* doc/install.texi (--with-mips-plt): Document.
* doc/invoke.texi (-mplt, -mno-plt): Document.
* config.gcc (mips*-*-*): Add mips-plt to supported_defaults
and handle ${with_mips_plt}.
* config/mips/mips.opt (mplt): New option.
* config/mips/mips.h (TARGET_ABICALLS_PIC0): New macro.
(TARGET_ABICALLS_PIC2): Likewise.
(TARGET_GPWORD): Return false for TARGET_ABSOLUTE_ABICALLS.
(OPTION_DEFAULT_SPECS): Add a mips-plt entry.
(ASM_SPEC): Use !mabi=* instead of !mabi*.
(MIPS_CALL): Use TARGET_ABICALLS_PIC2 instead of TARGET_ABICALLS
to decide whether to output ".option picX" directives.
* config/mips/linux.h (SUBTARGET_ASM_SPEC): Remove -mabi=64 handling.
Pass -call_nonpic rather than -KPIC for -mplt.
(BASE_DRIVER_SELF_SPECS): Remove -mplt if -mno-shared is not present
on the command line. Also remove it when -mabi=64 is used without
-msym32.
* config/mips/linux64.h (SUBTARGET_ASM_SPEC): Delete.
* config/mips/mips.c (mips_use_pic_fn_addr_reg_p): Handle
TARGET_ABICALLS_PIC0.
(mips_classify_symbol): Use TARGET_ABICALLS_PIC2 instead of
TARGET_ABICALLS.
(mips16_build_function_stub): Only output ".option pic" directives
and PIC stubs if TARGET_ABICALLS_PIC2. Call through $25 instead of $1.
(mips16_build_call_stub): Fix comment and remove redundant
".set at"/"set .noat" directives.
(mips_function_rodata_section): Use the default behaviour for
TARGET_ABSOLUTE_ABICALLS.
(mips_file_start): Emit ".option pic0" for TARGET_ABICALLS_PIC0.
(mips_global_pointer): Handle TARGET_ABICALLS_PIC0.
(mips_restore_gp): Do nothing if the current function doesn't use
a global pointer.
(mips_expand_prologue): Only save $gp if the current function uses it.
Use a normal move for TARGET_ABICALLS_PIC0.
(mips_override_options): Only set flag_pic if TARGET_ABICALLS_PIC2.
Co-Authored-By: Richard Sandiford <rdsandiford@googlemail.com>
From-SVN: r139170
Paolo Carlini [Sun, 17 Aug 2008 10:19:34 +0000 (10:19 +0000)]
acinclude.m4 ([GLIBCXX_CHECK_LFS], [...]): Use GCC_TRY_COMPILE_OR_LINK instead of AC_TRY_LINK...
2008-08-17 Paolo Carlini <paolo.carlini@oracle.com>
* acinclude.m4 ([GLIBCXX_CHECK_LFS], [GLIBCXX_CHECK_POLL],
[GLIBCXX_CHECK_S_ISREG_OR_S_IFREG], [GLIBCXX_CHECK_WRITEV]):
Use GCC_TRY_COMPILE_OR_LINK instead of AC_TRY_LINK, in C++,
with -fno-exceptions as flags.
* configure.ac: Run unconditionally GLIBCXX_CHECK_LFS,
GLIBCXX_CHECK_POLL, GLIBCXX_CHECK_S_ISREG_OR_S_IFREG,
GLIBCXX_CHECK_WRITEV and commonized AC_CHECK_HEADERS.
* crossconfig.m4: Adjust; remove AC_CHECK_HEADERS commonized
in configure.ac.
* configure: Regenerate.
Eric Botcazou [Sat, 16 Aug 2008 18:40:57 +0000 (18:40 +0000)]
re PR middle-end/20548 (ACATS c52103x c52104x c52104y segfault)
PR ada/20548
* common.opt (-fstack-check): Do not declare the variable here.
(-fstack-check=): New option variant.
* doc/invoke.texi (Code Gen Options): Document it.
* expr.h (STACK_OLD_CHECK_PROTECT): New macro.
(STACK_CHECK_PROTECT): Bump to 3 pages if DWARF-2 EH is used.
(STACK_CHECK_STATIC_BUILTIN): New macro.
* doc/tm.texi (Stack Checking): Document STACK_CHECK_STATIC_BUILTIN.
* opts.c: Include expr.h.
(common_handle_option) <OPT_fold_stack_check_>: New case.
<OPT_fstack_check>: Likewise.
* calls.c (initialize_argument_information): Use TYPE_SIZE_UNIT
consistently in the test for variable-sized types. Adjust for
new behaviour of flag_stack_check.
* explow.c: Include except.h.
(allocate_dynamic_stack_space): Do not take into account
STACK_CHECK_MAX_FRAME_SIZE for static builtin stack checking.
* function.c (gimplify_parameters): Use DECL_SIZE_UNIT in the test
for variable-sized parameters. Treat all parameters whose size is
greater than STACK_CHECK_MAX_VAR_SIZE as variable-sized if generic
stack checking is enabled.
* gimplify.c (gimplify_decl_expr): Treat non-static objects whose
size is greater than STACK_CHECK_MAX_VAR_SIZE as variable-sized
if generic stack checking is enabled.
(expand_function_end): Adjust for new behaviour of flag_stack_check.
* reload1.c (reload): Likewise.
* stmt.c (expand_decl): Assert that all automatic variables have
fixed size at this point and remove dead code.
* flags.h (stack_check_type): New enumeration type.
(flag_stack_check): Declare.
* toplev.c (flag_stack_check): New global variable.
* Makefile.in (opts.o): Add dependency on EXPR_H.
(explow.o): Add dependency on except.h.
ada/
* gcc-interface/decl.c (gnat_to_gnu_entity): Use DECL_SIZE_UNIT in the
setjmp test consistently. Adjust for new behaviour of flag_stack_check.
* gcc-interface/utils2.c (build_call_alloc_dealloc): Remove redundant
test of flag_stack_check. Adjust for new behaviour of flag_stack_check.
Jerry DeLisle [Sat, 16 Aug 2008 03:38:31 +0000 (03:38 +0000)]
re PR libfortran/35863 ([F2003] Implement ENCODING="UTF-8")
2008-08-15 Jerry DeLisle <jvdelisle@gcc.gnu.org>
PR libfortran/35863
* intrinsics/selected_char_kind.c: Enable iso_10646.
* io/read.c (typedef uchar): New type.
(read_utf8): New function to read a single UTF-8 encoded character.
(read_utf8_char1): New function to read UTF-8 into a KIND=1 string.
(read_default_char1): New functio to read default into KIND=1 string.
(read_utf8_char4): New function to read UTF-8 into a KIND=4 string.
(read_default_char4): New function to read UTF-8 into a KIND=4 string.
(read_a): Modify to use the new functions.
(read_a_char4): Modify to use the new functions.
* io/write.c (error.h): Add include. (typedef uchar): New type.
(write_default_char4): New function to default write KIND=4 string.
(write_utf8_char4): New function to UTF-8 write KIND=4 string.
(write_a_char4): Modify to use new functions.
(write_character): Modify to use new functions.
Eric Botcazou [Fri, 15 Aug 2008 22:34:52 +0000 (22:34 +0000)]
expr.c (expand_expr_real_1): When converting to BLKmode, try to fetch an inner memory reference.
* expr.c (expand_expr_real_1) <VIEW_CONVERT_EXPR>: When converting
to BLKmode, try to fetch an inner memory reference. Use 'mode' in
lieu of TYPE_MODE (type) throughout.
Joseph Myers [Fri, 15 Aug 2008 20:12:01 +0000 (21:12 +0100)]
arm.c (add_minipool_backward_ref): Check for 8-byte-aligned entries in second case of forcing insertion after a...
* config/arm/arm.c (add_minipool_backward_ref): Check for
8-byte-aligned entries in second case of forcing insertion after a
particular entry. Change third case to avoid inserting
non-8-byte-aligned entries before 8-byte-aligned ones.
Thomas Koenig [Thu, 14 Aug 2008 18:31:32 +0000 (18:31 +0000)]
re PR libfortran/36886 (misaligment for cshift of character)
2008-08-14 Thomas Koenig <tkoenig@gcc.gnu.org>
PR libfortran/36886
* Makefile.am: Added $(i_cshift0_c).
Added $(i_cshift0_c) to gfor_built_specific_src.
Add rule to build from cshift0.m4.
* Makefile.in: Regenerated.
* libgfortran.h: Addedd prototypes for cshift0_i1,
cshift0_i2, cshift0_i4, cshift0_i8, cshift0_i16,
cshift0_r4, cshift0_r8, cshift0_r10, cshift0_r16,
cshift0_c4, cshift0_c8, cshift0_c10, cshift0_c16.
Define Macros GFC_UNALIGNED_C4 and GFC_UNALIGNED_C8.
* intrinsics/cshift0.c: Remove helper functions for
the innter shift loop.
(cshift0): Call specific functions depending on type
of array argument. Only call specific functions for
correct alignment for other types.
* m4/cshift0.m4: New file.
* generated/cshift0_i1.c: New file.
* generated/cshift0_i2.c: New file.
* generated/cshift0_i4.c: New file.
* generated/cshift0_i8:.c New file.
* generated/cshift0_i16.c: New file.
* generated/cshift0_r4.c: New file.
* generated/cshift0_r8.c: New file.
* generated/cshift0_r10.c: New file.
* generated/cshift0_r16.c: New file.
* generated/cshift0_c4.c: New file.
* generated/cshift0_c8.c: New file.
* generated/cshift0_c10.c: New file.
* generated/cshift0_c16.c: New file.
2008-08-14 Thomas Koenig <tkoenig@gcc.gnu.org>
PR libfortran/36886
* gfortran.dg/cshift_char_3.f90: New test case.
* gfortran.dg/cshift_nan_1.f90: New test case.
2008-08-14 Rafael Avila de Espindola <espindola@google.com>
* gcc.dg/visibility-14.c: New test.
* gcc.dg/visibility-15.c: New test.
* gcc.dg/visibility-16.c: New test.
* gcc.dg/visibility-17.c: New test.
* gcc.dg/visibility-18.c: New test.
* gcc.dg/visibility-19.c: New test.
re PR c++/28152 (Diagnostic about wrong use _Complex prints __complex__)
2008-08-08 Manuel Lopez-Ibanez <manu@gcc.gnu.org>
PR c/28152
* c-parser.c (c_lex_one_token): Do not store the canonical spelling
for keywords.
testsuite/
* gcc.dg/parser-pr28152.c: New.
* gcc.dg/parser-pr28152-2.c: New.
Dorit Nuzman [Thu, 14 Aug 2008 12:47:56 +0000 (12:47 +0000)]
tree-vect-transform.c (vect_create_epilog_for_reduction): Takes an additional argument.
2008-08-14 Dorit Nuzman <dorit@il.ibm.com>
* tree-vect-transform.c (vect_create_epilog_for_reduction): Takes an
additional argument. Support reduction when duplication is needed due
to data-types of different sizes in the loop.
(get_initial_def_for_induction): Fix printout.
(vect_get_vec_def_for_stmt_copy): Support case where the
vec_stmt_for_operand is a phi node.
(vectorizable_reduction): Support reduction when duplication is needed
due to data-types of different sizes in the loop.
(vectorizable_call): Remove restriction to not vectorize in case we
have data-types of different sizes in the loop.
(vectorizable_conversion): Likewise.
(vectorizable_operation): Likewise.
(vectorizable_type_demotion): Likewise.
(vectorizable_type_promotion): Likewise.
(vectorizable_induction): Add restriction to not vectorize in case
we have data-types of different sizes in the loop.
Jakub Jelinek [Thu, 14 Aug 2008 09:02:46 +0000 (11:02 +0200)]
re PR middle-end/37103 (possible integer codegen bug)
PR middle-end/37103
* fold-const.c (fold_widened_comparison): Do not allow
sign changes that change the result even if shorter type
is wider than arg1_unw's type.
Kazu Hirata [Wed, 13 Aug 2008 23:21:12 +0000 (23:21 +0000)]
arm-g2.c, [...]: Skip if the multilib testing specifies -march that does not agree with the one...
* gcc.dg/arm-g2.c, gcc.dg/arm-mmx-1.c, gcc.dg/arm-scd42-2.c:
Skip if the multilib testing specifies -march that does not
agree with the one specified in the testcase.
Richard Guenther [Wed, 13 Aug 2008 14:22:19 +0000 (14:22 +0000)]
tree.h (maybe_fold_offset_to_address): Declare.
2008-08-13 Richard Guenther <rguenther@suse.de>
* tree.h (maybe_fold_offset_to_address): Declare.
* tree-ssa-ccp.c (surely_varying_stmt_p): Fix typo in last commit.
(ccp_fold): Handle pointer conversions the same as fold_stmt.
Likewise for POINTER_PLUS_EXPR.
(maybe_fold_offset_to_reference): Enable disabled code.
(maybe_fold_offset_to_address): New function.
(fold_stmt_r): Use it.
(fold_gimple_assign): Likewise.
* gimplify.c (gimplify_conversion): Use maybe_fold_offset_to_address.
(gimplify_expr): Likewise.
* gcc.dg/tree-ssa/ssa-ccp-21.c: New testcase.
* gcc.dg/tree-ssa/ssa-ccp-22.c: Likewise.
* gcc.dg/tree-ssa/ssa-ccp-23.c: Likewise.
Samuel Tardieu [Wed, 13 Aug 2008 10:57:43 +0000 (10:57 +0000)]
re PR ada/36777 (Protected type cannot have access taken from its body.)
gcc/ada/
PR ada/36777
* sem_util.ads, sem_util.adb (Is_Protected_Self_Reference): New.
* sem_attr.adb (Check_Type): The current instance of a protected
object is not a type name.
(Analyze_Access_Attribute): Accept instances of protected objects.
(Analyze_Attribute, Attribute_Address clause): Ditto.
* exp_attr.adb (Expand_N_Attribute_Reference): Rewrite
the prefix as being the current instance if needed.
re PR c/15236 (pedantic switch modifies treatment of non-ISO compliant enumerations)
2008-08-13 Manuel Lopez-Ibanez <manu@gcc.gnu.org>
PR c/15236
* diagnostic.c (pedwarn_at): New.
* toplev.h (pedwarn_at): Declare.
* c-tree.h (build_enumerator): Update declaration.
* c-decl.c (finish_enum): Update comment.
(build_enumerator): Take a location parameter. Give a pedwarn but do
not perform any conversion.
* c-parser.c (c_parser_enum_specifier): Set correct location for
enumerator.
testsuite/
* gcc.dg/pr15236.c: New.
* gcc.dg/torture/pr25183.c: Update.
Richard Guenther [Wed, 13 Aug 2008 08:57:20 +0000 (08:57 +0000)]
re PR tree-optimization/15255 ([tree-ssa] a * 2 + a * 2 is not converted to a * 4)
2008-08-13 Richard Guenther <rguenther@suse.de>
PR tree-optimization/15255
* tree-ssa-reassoc.c (linearize_expr_tree): Declare.
(struct oecount_s): New struct and VEC types.
(cvec): New global.
(oecount_hash): New function.
(oecount_eq): Likewise.
(oecount_cmp): Likewise.
(zero_one_operation): New function.
(build_and_add_sum): Likewise.
(undistribute_ops_list): Perform un-distribution of multiplication
and division on the chain of summands.
(should_break_up_subtract): Also break up subtracts for factors.
(reassociate_bb): Delete dead visited statements.
Call undistribute_ops_list. Re-sort and optimize if it did something.
* passes.c (init_optimization_passes): Move DSE before
reassociation.
* tree-ssa-loop-niter.c (stmt_dominates_stmt_p): Correctly handle
PHI nodes.
Janis Johnson [Tue, 12 Aug 2008 21:23:11 +0000 (21:23 +0000)]
pr32000-2.c: Use dg-skip-if for target expression.
* gcc.target/i386/pr32000-2.c: Use dg-skip-if for target expression.
* gcc.target/i386/stackalign/return-3.c: Ditto.
* gcc.target/sparc/ultrasp3.c: Ditto.
* lib/target-supports-dg.exp (dg-require-effective-target): Error
if argument is not a single effective-target keyword.
Paolo Carlini [Tue, 12 Aug 2008 19:38:02 +0000 (19:38 +0000)]
re PR c++/37087 (Segfault on compiling template defined in wrong namespace.)
/cp
2008-08-12 Paolo Carlini <paolo.carlini@oracle.com>
PR c++/37087
* parser.c (cp_parser_class_head): Early return error_mark_node in
case of global qualification of class name or qualified name that
does not name a class.
/testsuite
2008-08-12 Paolo Carlini <paolo.carlini@oracle.com>
* real.h (REAL_MODE_FORMAT): Protect MODE against macro expansion.
(FLOAT_MODE_FORMAT): New macro.
(REAL_MODE_FORMAT_COMPOSITE_P): Remove, replace by ...
(MODE_COMPOSITE_P): ... this new macro.
(MODE_HAS_NANS, MODE_HAS_INFINITIES, MODE_HAS_SIGNED_ZEROS,
MODE_HAS_SIGN_DEPENDENT_ROUNDING): New macros.
* machmode.h (GET_MODE_INNER): Cast result to enum machine_mode.
* flags.h: Include "real.h".
* fold-const.c (const_binop): Use MODE_COMPOSITE_P instead of
REAL_MODE_FORMAT_COMPOSITE_P.
* simplify-rtx.c (simplify_const_binary_operation): Likewise.
* doc/tm.texi (Storage Layout): Remove documentation of
MODE_HAS_NANS, MODE_HAS_INFINITIES, MODE_HAS_SIGNED_ZEROS,
MODE_HAS_SIGN_DEPENDENT_ROUNDING. Update documentation of
ROUND_TOWARDS_ZERO and LARGEST_EXPONENT_IS_NORMAL to clarify
they only apply to libgcc2.a.
Ulrich Weigand [Tue, 12 Aug 2008 13:13:38 +0000 (13:13 +0000)]
real.c (spu_single_format): New variable.
ChangeLog:
* real.c (spu_single_format): New variable.
* real.h (spu_single_format): Declare.
* config/spu/spu.c (spu_override_options): Install SFmode format.
(spu_split_immediate): Use integer mode to operate on pieces of
floating-point values in all cases.
* config/spu/spu.md (UNSPEC_FLOAT_EXTEND, UNSPEC_FLOAT_TRUNCATE): New.
("extendsfdf2"): Use UNSPEC_FLOAT_EXTEND instead of FLOAT_EXTEND.
("truncdfsf2"): Use UNSPEC_FLOAT_TRUNCATE instead of FLOAT_TRUNCATE.
testsuite/ChangeLog:
* gcc.c-torture/execute/ieee/inf-2.c (testf): Skip on the SPU.
Joel Sherrill [Mon, 11 Aug 2008 17:04:07 +0000 (17:04 +0000)]
s-oscons-tmplt.c: RTEMS defines AF_INET6 but does support it.
2008-08-11 Joel Sherrill <joel.sherrill@oarcorp.com>
gcc/ada/
* s-oscons-tmplt.c: RTEMS defines AF_INET6 but does support it.
* gsocket.h, socket.c: Update to support RTEMS.
* gcc-interface/Make-lang.in: Include CFLAGS_FOR_TARGET when cross.
Samuel Tardieu [Sun, 10 Aug 2008 20:13:24 +0000 (20:13 +0000)]
exp_ch4.adb (Expand_N_Op_Expon): Force evaluation of left argument even when right argument is 0.
2008-08-10 Samuel Tardieu <sam@rfc1149.net>
Robert Dewar <dewar@adacore.com>
gcc/ada/
* exp_ch4.adb (Expand_N_Op_Expon): Force evaluation of
left argument even when right argument is 0.
(Expand_N_Op_Mod): Ditto when right argument is 1.
(Expand_N_Op_Multiply): Ditto when any argument is 0.
(Expand_N_Op_Rem): Ditto when right argument is 1.
2008-08-10 Samuel Tardieu <sam@rfc1149.net>
gcc/testsuite/
* gnat.dg/exp0_eval.adb: New.
Co-Authored-By: Robert Dewar <dewar@adacore.com>
From-SVN: r138934
projects / thirdparty / gcc.git / log
