Debian uses a different D_FMT string for the zh_TW.UTF-8 locale, which
caused this test to fail. Try to detect the Debian format and adjust
the input being tested.
PR libstdc++/31413
* testsuite/22_locale/time_get/get_date/wchar_t/4.cc: Check D_FMT
string for alternative format.
From-SVN: r263948
2018-08-29 Matthew Malcomson <matthew.malcomson@arm.com>
* gcc.target/aarch64/simd/vect_su_add_sub.c: Use 32 and 64-bit types
where appropriate.
From-SVN: r263946
2018-08-29 Richard Biener <rguenther@suse.de>
PR tree-optimization/87117
* tree-ssa-sccvn.c (eliminate_dom_walker::eliminate_stmt): Only
re-value-number released SSA VDEFs.
* gfortran.dg/pr87117.f90: New testcase.
From-SVN: r263945
This is the gofrontend version of https://golang.org/cl/91796.
This is part of that CL, just the compiler change and required runtime
changes, in preparation for updating libgo to 1.11.
Relevant part of original CL description:
The hmap field in the maptype is only used by the runtime to check the sizes of
the hmap structure created by the compiler and runtime agree.
Comments are already present about the hmap structure definitions in the
compiler and runtime needing to be in sync.
Reviewed-on: https://go-review.googlesource.com/130976
From-SVN: r263941
Previously the logic that turned "a/b/c/../.." into "a/" failed to
preserve an empty path at the end of the iteration sequence, as required
by the trailing slash. That meant the result didn't meet the class
invariants, and that "a/b/c/d/../../.." would remove four components
instead of the three that "../../.." should remove.
PR libstdc++/87116
* src/filesystem/std-path.cc (path::lexically_normal): When handling
a dot-dot filename, preserve an empty final component in the iteration
sequence.
[_GLIBCXX_FILESYSTEM_IS_WINDOWS]: Use preferred-separator for
root-directory.
* testsuite/27_io/filesystem/path/generation/normal.cc: Add new tests
for more than two adjacent dot-dot filenames.
[_GLIBCXX_FILESYSTEM_IS_WINDOWS]: Replace slashes with
preferred-separator in expected normalized strings.
From-SVN: r263922
/cp
2018-08-28 Paolo Carlini <paolo.carlini@oracle.com>
PR c++/86546
* decl.c (finish_case_label): If the type is erroneous early
return error_mark_node.
/testsuite
2018-08-28 Paolo Carlini <paolo.carlini@oracle.com>
PR c++/86546
* g++.dg/other/switch4.C: New.
From-SVN: r263921
gcc/ChangeLog:
* dumpfile.h (ATTRIBUTE_GCC_DUMP_PRINTF): Change version check on
GCC_VERSION for usage of "__gcc_dump_printf__" format from
>= 3005 to >= 9000.
From-SVN: r263920
2018-08-28 Richard Biener <rguenther@suse.de>
PR tree-optimization/87124
* tree-ssa-sccvn.c (vn_lookup_simplify_result): Guard against
constants before looking up avail.
* g++.dg/torture/pr87124.C: New testcase.
From-SVN: r263918
2017-08-28 Paul Thomas <pault@gcc.gnu.org>
PR fortran/80477
* trans-expr.c (gfc_conv_procedure_call): Allocatable class
scalar results being passed to a derived type formal argument
are finalized if possible. Otherwise, rely on existing code for
deallocation. Make the deallocation of allocatable result
components conditional on finalization not taking place. Make
the freeing of data components after finalization conditional
on the data being NULL.
(gfc_trans_arrayfunc_assign): Change the gcc_assert to a
condition to return NULL_TREE.
(gfc_trans_assignment_1): If the assignment is class to class
and the rhs expression must be finalized but the assignment
is not marked as a polymorphic assignment, use the vptr copy
function instead of gfc_trans_scalar_assign.
PR fortran/86481
* trans-expr.c (gfc_conv_expr_reference): Do not add the post
block to the pre block if the expression is to be finalized.
* trans-stmt.c (gfc_trans_allocate): If the expr3 must be
finalized, load the post block into a finalization block and
add it right at the end of the allocation block.
2017-08-28 Paul Thomas <pault@gcc.gnu.org>
PR fortran/80477
* gfortran.dg/class_result_7.f90: New test.
* gfortran.dg/class_result_8.f90: New test.
* gfortran.dg/class_result_9.f90: New test.
PR fortran/86481
* gfortran.dg/allocate_with_source_25.f90: New test.
From-SVN: r263916
PR middle-end/87099
* calls.c (maybe_warn_nonstring_arg): Punt early if
warn_stringop_overflow is zero. Don't call get_range_strlen
on 3rd argument, keep iterating until lenrng[1] is INTEGER_CST.
Swap comparison operands to have constants on rhs. Only use
lenrng[1] if non-NULL and INTEGER_CST. Don't uselessly
increment lenrng[0].
* gcc.dg/pr87099.c: New test.
From-SVN: r263915
Fixes many testsuite failures for SVE.
2018-08-28 Richard Sandiford <richard.sandiford@arm.com>
gcc/
* tree-ssa-sccvn.c (fully_constant_vn_reference_p): Fix unguarded
use of tree_to_shwi. Remove duplicated test for the size being
a whole number of bytes.
From-SVN: r263914
This test was failing for Power 7 due to the lack of hw support
for unaligned accesses.
2018-08-28 Richard Sandiford <richard.sandiford@arm.com>
gcc/testsuite/
PR testsuite/87078
* gcc.dg/vect/slp-37.c: Restrict scan tests to vect_hw_misalign.
From-SVN: r263913
2018-08-28 Richard Biener <rguenther@suse.de>
PR tree-optimization/87117
* tree-ssa-sccvn.c (fully_constant_vn_reference_p): Exclude
void which is is_gimple_reg_type by checking for COMPLETE_TYPE_P.
* gcc.dg/pr87117-1.c: New testcase.
From-SVN: r263911
2018-08-28 Richard Biener <rguenther@suse.de>
PR tree-optimization/87117
* tree-ssa-pre.c (compute_avail): Do not make expressions
with predicated values available.
(get_expr_value_id): Assert we do not run into predicated value
expressions.
* gcc.dg/pr87117-2.c: New testcase.
From-SVN: r263910
2018-08-28 Richard Biener <rguenther@suse.de>
PR tree-optimization/87117
* tree-ssa-operands.c (add_stmt_operand): STRING_CST may
get virtual operands.
(get_expr_operands): Handle STRING_CST like other decls.
* gcc.dg/lvalue-5.c: New testcase.
From-SVN: r263908
gcc/ChangeLog:
PR tree-optimization/87112
* builtins.c (expand_builtin_strnlen): Convert c_strlen result to
the type of the bound argument.
gcc/testsuite/ChangeLog:
PR tree-optimization/87112
* gcc.dg/pr87112.c: New test.
From-SVN: r263900
This patch adds a fix-it hint to missing "typename" errors in the C++
frontend, suggesting the insertion of "typename ".
This addresses part of PR c++/63392; however it does not improve the
error-recovery for such cases.
gcc/cp/ChangeLog:
PR c++/63392
* parser.c (cp_parser_diagnose_invalid_type_name): Add fix-it
hint.
gcc/testsuite/ChangeLog:
PR c++/63392
* g++.dg/diagnostic/missing-typename.C: New test.
From-SVN: r263899
* tree-ssa-dse.c (compute_trims): Handle case where the reference's
type does not have a TYPE_SIZE_UNIT.
* gcc.c-torture/compile/dse.c: New test.
From-SVN: r263896
PR c++/86993
* cp-tree.h (cxx_readonly_error): Add location_t argument.
* typeck2.c (cxx_readonly_error): Add LOC argument, pass it to
ERROR_FOR_ASSIGNMENT macro and readonly_error. Add LOC argument
to ERROR_FOR_ASSIGNMENT macro, use error_at instead of error and
pass LOC to it. Formatting fixes.
* typeck.c (cp_build_unary_op): Pass location to cxx_readonly_error.
(cp_build_modify_expr): Pass loc to cxx_readonly_error.
* semantics.c (finish_asm_stmt): Pass input_location to
cxx_readonly_error.
* g++.dg/diagnostic/pr86993.C: New test.
From-SVN: r263891
2018-08-27 Steve Ellcey <sellcey@cavium.com>
* config/aarch64/aarch64-speculation.cc: Replace include of cfg.h
with include of backend.h.
From-SVN: r263890
This patch tweaks maybe_add_include_fixit so that if we're emitting a note
about adding the header file, the note's primary location will be replaced
by that of the fix-it hint, to avoid repeating a location we've already
emitted (or one close to it).
For example, this simplifies:
../x86_64-pc-linux-gnu/libstdc++-v3/include/vector:87:27: error: msg 1
87 | using vector = std::vector<_Tp, polymorphic_allocator<_Tp>>;
| ^~~~~~
../x86_64-pc-linux-gnu/libstdc++-v3/include/vector:87:22: note: msg 2
73 | # include <debug/vector>
+++ |+#include <vector>
74 | #endif
....
87 | using vector = std::vector<_Tp, polymorphic_allocator<_Tp>>;
| ^~~
to:
../x86_64-pc-linux-gnu/libstdc++-v3/include/vector:87:27: error: msg 1
87 | using vector = std::vector<_Tp, polymorphic_allocator<_Tp>>;
| ^~~~~~
../x86_64-pc-linux-gnu/libstdc++-v3/include/vector:74:1: note: msg 2
73 | # include <debug/vector>
+++ |+#include <vector>
74 | #endif
eliminating the repetition of line 87 in the note.
Doing so requires converting show_caret_p to a tri-state, to avoid
meaninglessly printing a caret for the first column in the next line
(and colorizing it):
../x86_64-pc-linux-gnu/libstdc++-v3/include/vector:74:1: note: msg 2
73 | # include <debug/vector>
+++ |+#include <vector>
74 | #endif
| ^
gcc/c-family/ChangeLog:
PR 87091
* c-common.c (c_cpp_error): Update for conversion of show_caret_p
to a tri-state.
(maybe_suggest_missing_token_insertion): Likewise.
(maybe_add_include_fixit): Add param "override_location". If set,
and source-printing is enabled, then override the rich_location's
primary location with that of the insertion point for the fix-it
hint, marking it with SHOW_LINES_WITHOUT_RANGE.
* c-common.h (extern void maybe_add_include_fixit): Add bool
param.
* c-format.c (selftest::test_type_mismatch_range_labels): Update
for conversion of show_caret_p to a tri-state.
* c-warn.c (warn_for_restrict): Likewise.
* known-headers.cc
(suggest_missing_header::~suggest_missing_header): Update call to
maybe_add_include_fixit to suggest overriding the location, as it
is for a note.
gcc/c/ChangeLog:
PR 87091
* c-decl.c (implicitly_declare): Update call to
maybe_add_include_fixit to suggest overriding the location, as it
is for a note.
* c-objc-common.c (c_tree_printer): Update for conversion of
show_caret_p to a tri-state.
gcc/cp/ChangeLog:
PR 87091
* decl.c (grokdeclarator): Update for conversion of show_caret_p
to a tri-state.
* error.c (cp_printer): Likewise.
* name-lookup.c (maybe_suggest_missing_std_header): Update call to
maybe_add_include_fixit to suggest overriding the location, as it
is for a note.
* parser.c (cp_parser_string_literal): Update for conversion of
show_caret_p to a tri-state.
(cp_parser_elaborated_type_specifier): Likewise.
(set_and_check_decl_spec_loc): Likewise.
* pt.c (listify): Update call to maybe_add_include_fixit to not
override the location, as it is for an error.
* rtti.c (typeid_ok_p): Likewise.
gcc/ChangeLog:
PR 87091
* diagnostic-show-locus.c (class layout_range): Update for
conversion of show_caret_p to a tri-state.
(layout_range::layout_range): Likewise.
(make_range): Likewise.
(layout::maybe_add_location_range): Likewise.
(layout::should_print_annotation_line_p): Don't show annotation
lines for ranges that are SHOW_LINES_WITHOUT_RANGE.
(layout::get_state_at_point): Update for conversion of
show_caret_p to a tri-state. Bail out early for
SHOW_LINES_WITHOUT_RANGE, so that such ranges don't affect
underlining or source colorization.
(gcc_rich_location::add_location_if_nearby): Update for conversion
of show_caret_p to a tri-state.
(selftest::test_one_liner_multiple_carets_and_ranges): Likewise.
(selftest::test_one_liner_fixit_replace_equal_secondary_range):
Likewise.
(selftest::test_one_liner_labels): Likewise.
* gcc-rich-location.c (gcc_rich_location::add_expr): Update for
conversion of show_caret_p to a tri-state.
* pretty-print.c (text_info::set_location): Likewise.
* pretty-print.h (text_info::set_location): Likewise.
* substring-locations.c (format_warning_n_va): Likewise.
* tree-diagnostic.c (default_tree_printer): Likewise.
* tree-pretty-print.c (newline_and_indent): Likewise.
gcc/fortran/ChangeLog:
PR 87091
* error.c (gfc_format_decoder): Update for conversion of
show_caret_p to a tri-state.
gcc/testsuite/ChangeLog:
PR 87091
* gcc.dg/empty.h: New file.
* gcc.dg/fixits-pr84852-1.c: Update for move of fix-it hint to
top of file and removal of redundant second printing of warning
location.
* gcc.dg/fixits-pr84852-2.c: Likewise.
* gcc.dg/missing-header-fixit-3.c: Likewise.
* gcc.dg/missing-header-fixit-4.c: New test.
* gcc.dg/plugin/diagnostic_plugin_test_show_locus.c: Update for
conversion of show_caret_p to a tri-state.
libcpp/ChangeLog:
PR 87091
* include/line-map.h (enum range_display_kind): New enum.
(struct location_range): Replace field "m_show_caret_p" with
"m_range_display_kind", converting from bool to the new enum.
(class rich_location): Add example of line insertion fix-it hint.
(rich_location::add_range): Convert param "show_caret_p" from bool
to enum range_display_kind and rename to "range_display_kind",
giving it a default of SHOW_RANGE_WITHOUT_CARET.
(rich_location::set_range): Likewise, albeit without a default.
* line-map.c (rich_location::rich_location): Update for conversion
of show_caret_p to tri-state enum.
(rich_location::add_range): Likewise.
(rich_location::set_range): Likewise.
From-SVN: r263885
This patch tweaks how we print line-insertion fix-it hints, so that
the line before the insertion point is also printed, to give the user
more context on the proposed change.
For example, it changes:
../x86_64-pc-linux-gnu/libstdc++-v3/include/vector:87:22: note: message
+++ |+#include <vector>
74 | #endif
....
87 | using vector = std::vector<_Tp, polymorphic_allocator<_Tp>>;
| ^~~
to:
../x86_64-pc-linux-gnu/libstdc++-v3/include/vector:87:22: note: message
73 | # include <debug/vector>
+++ |+#include <vector>
74 | #endif
....
87 | using vector = std::vector<_Tp, polymorphic_allocator<_Tp>>;
| ^~~
gcc/ChangeLog:
PR 87091
* diagnostic-show-locus.c (get_line_span_for_fixit_hint): Show the
line above for line-insertion fix-it hints.
(selftest::test_fixit_insert_containing_newline): Update the
expected results, and add a test with line-numbering enabled.
gcc/testsuite/ChangeLog:
PR 87091
* g++.dg/pr85523.C: Extend expected output to show line
before line-insertion fix-it hint.
* gcc.dg/plugin/diagnostic-test-show-locus-bw-line-numbers.c
(test_fixit_insert_newline): Add previous line to expected output.
* gcc.dg/plugin/diagnostic-test-show-locus-bw.c: Likewise.
* gcc.dg/plugin/diagnostic-test-show-locus-color.c: Likewise.
From-SVN: r263884
2018-08-27 Martin Liska <mliska@suse.cz>
PR tree-optimization/86847
* tree-switch-conversion.c (switch_decision_tree::dump_case_nodes):
Dump also subtree probability.
(switch_decision_tree::do_jump_if_equal): New function.
(switch_decision_tree::emit_case_nodes): Handle special
situations in balanced tree that can be emitted much simpler.
Fix calculation of probabilities that happen in tree expansion.
* tree-switch-conversion.h (struct cluster): Add
is_single_value_p.
(struct simple_cluster): Likewise.
(struct case_tree_node): Add new function has_child.
(do_jump_if_equal): New.
2018-08-27 Martin Liska <mliska@suse.cz>
PR tree-optimization/86847
* gcc.dg/tree-ssa/switch-3.c: New test.
* gcc.dg/tree-ssa/vrp105.c: Remove.
From-SVN: r263879
2018-08-27 Martin Liska <mliska@suse.cz>
* tree-switch-conversion.c (bit_test_cluster::find_bit_tests):
Add new argument to bit_test_cluster constructor.
(bit_test_cluster::emit): Set bits really number of values
handlel by a test.
(bit_test_cluster::hoist_edge_and_branch_if_true): Add
probability argument.
* tree-switch-conversion.h (struct bit_test_cluster):
Add m_handles_entire_switch.
2018-08-27 Martin Liska <mliska@suse.cz>
* gcc.dg/tree-ssa/switch-2.c: New test.
From-SVN: r263878
2018-08-27 Martin Liska <mliska@suse.cz>
PR tree-optimization/86702
* tree-switch-conversion.c (jump_table_cluster::emit):
Make probabilities even for values in jump table
according to number of cases handled.
(switch_decision_tree::compute_cases_per_edge): Pass
argument to reset_out_edges_aux function.
(switch_decision_tree::analyze_switch_statement): Likewise.
* tree-switch-conversion.h (switch_decision_tree::reset_out_edges_aux):
Make it static.
From-SVN: r263877
