Jakub Jelinek [Fri, 29 Mar 2019 20:10:19 +0000 (21:10 +0100)]
re PR sanitizer/89869 (-fsanitize=undefined miscompilation)
PR sanitizer/89869
* typeck.c: Include gimplify.h.
(cp_build_modify_expr) <case COND_EXPR>: Unshare rhs before using it
for second time. Formatting fixes.
Martin Liska [Fri, 29 Mar 2019 11:45:39 +0000 (12:45 +0100)]
Fix multiple values for -fdbg-cnt.
2019-03-29 Martin Liska <mliska@suse.cz>
* dbgcnt.c (dbg_cnt_process_single_pair): Fix GNU coding style.
(dbg_cnt_process_opt): Parse first tokens aas
dbg_cnt_process_single_pair is also using strtok.
Jakub Jelinek [Thu, 28 Mar 2019 22:33:29 +0000 (23:33 +0100)]
re PR middle-end/89621 (ICE with allocatable character and openmp)
PR middle-end/89621
* tree-inline.h (struct copy_body_data): Add
dont_remap_vla_if_no_change flag.
* tree-inline.c (remap_type_3, remap_type_2): New functions.
(remap_type): Don't remap vla types if id->dont_remap_vla_if_no_change
and remap_type_2 returns false.
* omp-low.c (new_omp_context): Set ctx->cb.dont_remap_vla_if_no_change.
Move ctx->cb.adjust_array_error_bounds setting to the outermost ctx
only from where it is copied to nested contexts.
Marek Polacek [Thu, 28 Mar 2019 20:24:48 +0000 (20:24 +0000)]
PR c++/89612 - ICE with member friend template with noexcept.
* pt.c (maybe_instantiate_noexcept): For function templates, use their
template result (function decl). Don't set up local specializations.
Temporarily turn on processing_template_decl. Update the template type
too.
* g++.dg/cpp0x/noexcept38.C: New test.
* g++.dg/cpp0x/noexcept39.C: New test.
* g++.dg/cpp1z/noexcept-type21.C: New test.
Jakub Jelinek [Thu, 28 Mar 2019 14:47:47 +0000 (15:47 +0100)]
re PR c++/89785 (Incorrect "not a constant expression" error with switch statement that returns)
PR c++/89785
* constexpr.c (struct check_for_return_continue_data): New type.
(check_for_return_continue): New function.
(potential_constant_expression_1) <case SWITCH_STMT>: Walk
SWITCH_STMT_BODY to find RETURN_EXPRs or CONTINUE_STMTs not nested
in loop bodies and set *jump_target to that if found.
* g++.dg/cpp1y/constexpr-89785-1.C: New test.
* g++.dg/cpp1y/constexpr-89785-2.C: New test.
optrecord_json_writer::optinfo_to_json can in theory be called from any
optimization pass, but currently uses get_fnname_from_decl, which
is RTL-specific.
In that PR, Jakub suggested using either DECL_ASSEMBLER_NAME or the
"printable name" (via current_function_name).
This patch makes it use DECL_ASSEMBLER_NAME.
gcc/ChangeLog:
PR middle-end/89725
* optinfo-emit-json.cc (optrecord_json_writer::optinfo_to_json):
Use DECL_ASSEMBLER_NAME rather than get_fnname_from_decl.
* regcprop.c (copyprop_hardreg_forward_1): Remove redundant INSN_P
test.
(cprop_hardreg_bb, cprop_hardreg_debug): New functions.
(pass_cprop_hardreg::execute): Use those. Don't repeat bb processing
immediately after first one with df_analyze in between, but rather
process all bbs, queueing ones that need second pass in a worklist,
df_analyze, process queued debug insn changes and if second pass is
needed, process bbs from worklist, df_analyze, process queued debug
insns again.
Jakub Jelinek [Thu, 28 Mar 2019 14:07:40 +0000 (15:07 +0100)]
rtl.h (NONDEBUG_INSN_P): Define as NONJUMP_INSN_P or JUMP_P or CALL_P instead of INSN_P && !DEBUG_INSN_P.
* rtl.h (NONDEBUG_INSN_P): Define as NONJUMP_INSN_P or JUMP_P
or CALL_P instead of INSN_P && !DEBUG_INSN_P.
(INSN_P): Define using NONDEBUG_INSN_P or DEBUG_INSN_P.
Jason Merrill [Wed, 27 Mar 2019 20:39:19 +0000 (16:39 -0400)]
PR c++/89831 - error with qualified-id in const member function.
Since the fix for 15272 we were remembering the wrong function to use at
instantiation time, because the type of the SCOPE_REF didn't reflect the
cv-quals of 'this'. Conveniently, we can fix this by simplifying the code.
* semantics.c (finish_non_static_data_member): Use object cv-quals
in scoped case, too.
Jason Merrill [Wed, 27 Mar 2019 18:39:20 +0000 (14:39 -0400)]
PR c++/89241 - ICE with lambda in template parameter list.
We were getting confused by a lambda in template definition context that
isn't actually in the scope of any templated entity. Fixed by telling
type_dependent_expression_p that such a lambda is type-dependent even if we
can't tell that from its closure type. I've also restored the error for
defining a non-lambda class in a default template argument, and for a lambda
befor C++20.
* parser.c (cp_parser_lambda_expression): Also reject a lambda in a
template parameter list before C++20.
* pt.c (type_dependent_expression_p): True for LAMBDA_EXPR.
* semantics.c (begin_class_definition): Restore error about defining
non-lambda class in template parm list.
Peter Bergner [Wed, 27 Mar 2019 16:59:15 +0000 (16:59 +0000)]
re PR rtl-optimization/89313 (ICE in process_alt_operands, at lra-constraints.c:2962)
gcc/
PR rtl-optimization/89313
* function.c (matching_constraint_num): New static function.
(match_asm_constraints_1): Use it. Fixup white space and comment.
Don't replace inputs with non-matching constraints which conflict
with early clobber outputs.
gcc/testsuite/
PR rtl-optimization/89313
* gcc.dg/pr89313.c: New test.
Jeff Law [Wed, 27 Mar 2019 16:18:06 +0000 (10:18 -0600)]
re PR rtl-optimization/87761 ([MIPS] New FAIL: gcc.target/mips/fix-r4000-10.c -O1 start with r265398)
PR rtl-optimization/87761
PR rtl-optimization/89826
* regcprop.c (copyprop_hardreg_forward_1): Move may_trap_p test
slightly later.
(pass_cprop_hardreg::execute): Call df_analyze after adding the
note problem to get REG_DEAD/REG_UNUSED notes updated.
PR rtl-optimization/87761
PR rtl-optimization/89826
* gcc.c-torture/execute/pr89826.c: New test.
Jason Merrill [Wed, 27 Mar 2019 14:27:00 +0000 (10:27 -0400)]
PR c++/86932 - missed SFINAE with empty pack.
The issue here was that when processing the explicit template args in
fn_type_unification we added an empty argument pack for the parameter pack,
so we never tried to do any deduction for it, and therefore never looked at
its type. We need that empty pack behavior for partial ordering, but we
don't want it here, so let's make it conditional on tf_partial.
* pt.c (coerce_template_parms): Don't add an empty pack if
tf_partial.
(fn_type_unification): Pass tf_partial to coerce_template_parms.
Paul Thomas [Wed, 27 Mar 2019 12:51:43 +0000 (12:51 +0000)]
re PR fortran/88247 (ICE in get_array_ctor_var_strlen, at fortran/trans-array.c:2068)
2019-03-27 Paul Thomas <pault@gcc.gnu.org>
PR fortran/88247
* expr.c (is_subref_array): Permit substrings to be detected
as subref arrays.
* trans-array.c (get_array_ctor_var_strlen): Obtain the length
of deferred length strings. Handle substrings with a NULL end
expression.
(trans_array_constructor): Remove an unnecessary blank line.
(gfc_conv_scalarized_array_ref): Skip to label 'done' if 'decl'
is a pointer array.
(get_array_charlen): If the expression is an array, convert the
first element of the constructor and use its string length. Get
a new charlen if necessary.
(gfc_conv_expr_descriptor): Call 'get_array_charlen' for array
constructor expressions. If the ss_info string length is
available, use that to set the span of character arrays.
* trans-expr.c (gfc_get_expr_charlen): Handle substrings
* trans-stmt.c (trans_associate_var): Set the pointer array
flag for variable targets and constant array constructors. Take
care not to reset the string length or the span in the case of
expressions that are not converted as direct by reference.
2019-03-27 Paul Thomas <pault@gcc.gnu.org>
PR fortran/88247
* gfortran.dg/associate_47.f90: New test.
Richard Biener [Wed, 27 Mar 2019 09:31:53 +0000 (09:31 +0000)]
re PR debug/89463 (debug information for iterator of an empty loop is gone (at -O3))
2019-03-27 Richard Biener <rguenther@suse.de>
PR tree-optimization/89463
* tree-ssa-dce.c (remove_dead_stmt): Take output vector to
queue edges to remove.
(eliminate_unnecessary_stmts): Remove dead PHIs alongside
dead stmts. Delay edge removal until PHIs are removed to
make debug-stmt creation not confused by seemingly degenerate
PHIs.
Jason Merrill [Tue, 26 Mar 2019 16:02:19 +0000 (12:02 -0400)]
PR c++/86429 - constexpr variable in lambda.
When we refer to a captured variable from a constant-expression context
inside a lambda, the closure (like any function parameter) is not constant
because we aren't in a call, so we don't have an argument. So the capture
is non-constant. But if the captured variable is constant, we might be able
to use it directly in constexpr evaluation.
PR c++/82643
PR c++/87327
* constexpr.c (cxx_eval_constant_expression): In a lambda function,
try evaluating the captured variable directly.
Jonathan Wakely [Tue, 26 Mar 2019 15:28:48 +0000 (15:28 +0000)]
PR libstdc++/85965 delay static assertions until types are complete
The static assertions added for PR libstdc++/48101 were at class scope
and so were evaluated too eagerly, when it might not be possible to
determine whether the function objects are invocable with the key types.
The problematic cases are where the key type is not known to be
convertible to the argument type(s) of the function object until later,
after a type has been completed. Specifically, if the key type is a
pointer to a derived class and the function object's argument type is a
pointer to a base class, then the derived-to-base conversion is only
valid once the derived type is complete.
By moving the static assertions to the destructor they will only be
evaluated when the destructor is instantiated, at which point whether
the key type can be passed to the function object should be knowable.
The ideal place to do the checks would be only when the function objects
are actually invoked, but that would mean adding the checks in numerous
places, so the destructor is used instead.
The tests need to be adjusted because the "required from here" line is
now the location of the destructor, not the point of instantiation in
the test file. For the map and multimap tests which check two
specializations, the dg-error matching the assertion text matches both
cases. Also check the diagnostic output for the template arguments, to
ensure both specializations trigger the assertion.
PR libstdc++/85965
* include/bits/hashtable.h (_Hashtable): Move static assertions to
destructor so they are not evaluated until the _Key type is complete.
* include/bits/stl_tree.h (_Rb_tree): Likewise.
* testsuite/23_containers/set/85965.cc: New test.
* testsuite/23_containers/unordered_set/85965.cc: New test.
* testsuite/23_containers/map/48101_neg.cc: Replace "here" errors
with regexp matching the corresponding _Rb_tree specialization.
* testsuite/23_containers/multimap/48101_neg.cc: Likewise.
* testsuite/23_containers/multiset/48101_neg.cc: Remove "here" error.
* testsuite/23_containers/set/48101_neg.cc: Likewise.
* testsuite/23_containers/unordered_map/48101_neg.cc: Likewise.
* testsuite/23_containers/unordered_multimap/48101_neg.cc: Likewise.
* testsuite/23_containers/unordered_multiset/48101_neg.cc: Likewise.
* testsuite/23_containers/unordered_set/48101_neg.cc: Likewise.
re PR libstdc++/89825 (Jump table for variant visitation could be shortened for never empty variants)
PR libstdc++/89825
Fix based on a suggestion by Antony Polukhin.
* include/std/variant (_Extra_visit_slot_needed): New.
(_Multi_array): Use it.
(_S_apply_all_alts): Likewise.
Jakub Jelinek [Tue, 26 Mar 2019 13:40:43 +0000 (14:40 +0100)]
hash-table.h (hash_table::m_gather_mem_stats): If GATHER_STATISTICS is constant 0...
* hash-table.h (hash_table::m_gather_mem_stats): If GATHER_STATISTICS
is constant 0, turn into static const data member initialized to false.
(hash_table::hash_table): Only initialize m_gather_mem_stats #if
GATHER_STATISTICS. Add ATTRIBUTE_UNUSED to gather_mem_stats param.
re PR libstdc++/89824 (Variant jump table reserves space for __variant_cookie twice)
PR libstdc++/89824
Fix based on a suggestion by Antony Polukhin.
* include/std/variant (__gen_vtable): Don't reserve an
additional table slot, _Multi_array already does that.
re PR libstdc++/89816 (std::variant move construction regressed since GCC 8.3)
PR libstdc++/89816
Fix based on a suggestion by Antony Polukhin.
* include/std/variant (__variant_construct): Capture a pointer
to the storage and visit just one variant.
Bin Cheng [Tue, 26 Mar 2019 11:12:37 +0000 (11:12 +0000)]
re PR tree-optimization/81740 (wrong code at -O3 in both 32-bit and 64-bit modes on x86_64-linux-gnu)
2019-03-26 Bin Cheng <bin.cheng@linux.alibaba.com>
PR tree-optimization/81740
* tree-vect-data-refs.c (vect_analyze_data_ref_dependence):
In case of outer loop vectorization, check for backward dependence
at the inner loop if outer loop dependence is reversed.
* gcc.dg/vect/pr81740-1.c: New testcase.
* gcc.dg/vect/pr81740-2.c: Likewise.
Jeff Law [Mon, 25 Mar 2019 23:33:36 +0000 (17:33 -0600)]
re PR rtl-optimization/88347 (ICE in begin_move_insn, at sched-ebb.c:175)
PR rtl-optimization/88347
PR rtl-optimization/88423
* sched-deps.c (sched_analyze_insn): Take into account that for
tablejumps the barrier appears after a label and a jump_table_data.
* gcc.c-torture/compile/pr88347.c: New test.
* gcc.c-torture/compile/pr88423.c: New test.
Martin Sebor [Mon, 25 Mar 2019 22:56:40 +0000 (22:56 +0000)]
PR c/89812 - incorrect maximum in error: requested alignment '536870912' exceeds maximum 2147483648
gcc/c-family/ChangeLog:
PR c/89812
* c-common.c (check_user_alignment): Rename local. Correct maximum
alignment in diagnostic. Avoid assuming argument fits in SHWI,
convert it to UHWI when it fits.
Johan Karlsson [Mon, 25 Mar 2019 21:19:09 +0000 (21:19 +0000)]
re PR debug/86964 (Too many debug symbols included, especially for extern globals)
PR debug/86964
* dwarf2out.c (premark_used_variables): New function.
(prune_unused_types_walk): Do not mark not premarked external
variables.
(prune_unused_types): Call premark_used_variables.
Vladimir Makarov [Mon, 25 Mar 2019 21:14:40 +0000 (21:14 +0000)]
re PR rtl-optimization/89676 (Redundant moves for long long shift on 32bit x86)
2019-03-25 Vladimir Makarov <vmakarov@redhat.com>
PR rtl-optimization/89676
* lra-constraints.c (curr_insn_transform): Do match reload for
early clobbers when the match was successful only for different
registers.
Jason Merrill [Mon, 25 Mar 2019 18:27:08 +0000 (14:27 -0400)]
PR c++/87748 - substitution failure error with decltype.
This issue is similar to PR 87480; in both cases we were doing non-dependent
substitution with processing_template_decl set, leading to member access
expressions seeming still instantiation-dependent, and therefore decltype
not being simplified to its actual type. And as in that PR, the fix is to
clear processing_template_decl while substituting a default template
argument.
Martin Sebor [Mon, 25 Mar 2019 17:23:31 +0000 (17:23 +0000)]
extend.texi (Common Type Attributes): Document vector_size.
gcc/ChangeLog:
* doc/extend.texi (Common Type Attributes): Document vector_size.
(Common Variable Attributes): Mention size constraint. Correct
quoting and typos.
(Vector Extensions): Use @dfn when defining bas type. Clarify
base type and size constraints.
Thomas Otto [Mon, 25 Mar 2019 13:50:46 +0000 (13:50 +0000)]
dwarf2out.c (comp_dir_string): cached_wd could be set to both a heap string and a gc string...
2019-03-25 Thomas Otto <thomas.otto@pdv-fs.de>
* dwarf2out.c (comp_dir_string): cached_wd could be set to both a
heap string and a gc string, but since this variable is unknown to
ggc the gc string might get reused and corrupted. Fixed by always
using a heap string.
Richard Biener [Mon, 25 Mar 2019 12:18:38 +0000 (12:18 +0000)]
re PR middle-end/89779 (internal compiler error: tree check: expected class ‘type’, have ‘exceptional’ (error_mark) in tree_nop_conversion_p, at tree.c:12798)
2019-03-25 Richard Biener <rguenther@suse.de>
PR tree-optimization/89779
* tree-ssa-loop-ivopts.c (remove_unused_ivs): Return
to remove IV defs, delay actual removal.
(tree_ssa_iv_optimize_loop): Likewise. Avoid SCEV reset.
(tree_ssa_iv_optimize): Remove eliminated IV defs at the
very end, properly also reset loop control IV information.
Kyrylo Tkachov [Mon, 25 Mar 2019 09:25:07 +0000 (09:25 +0000)]
[doc] Fix typos in MINLOC, MAXLOC documentation
I think there's a typo in the documentation of the MINLOC and MAXLOC intrinsics.
It implies that when the BACK argument is true it returns the first occurrence of the min/max value,
but my understanding is that it should return the last occurrence.
This patch fixes the documentation.
* intrinsic.texi (MINLOC): Fix typo in BACK argument documentation.
(MAXLOC): Likewise.
rs6000: Make CSE'ing __tls_get_addr calls possible
CSE does not consider calls, not even const calls. This patch puts a
REG_EQUAL note on the pseudo we assign the __tls_get_addr result to,
so that those pseudos can be CSE'd and the extra calls deleted as dead
code.
CSE should really handle const calls directly, but it is stage 4.
* config/rs6000/rs6000.c (rs6000_legitimize_tls_address): Add REG_EQUAL
notes for the result of the __tls_get_addr calls.
* config/rs6000/rs6000.md (unspec UNSPEC_TLS_GET_ADDR): New.
Jeff Law [Sun, 24 Mar 2019 15:20:52 +0000 (09:20 -0600)]
re PR rtl-optimization/87761 ([MIPS] New FAIL: gcc.target/mips/fix-r4000-10.c -O1 start with r265398)
PR rtl-optimization/87761
* regcprop.c (copyprop_hardreg_forward_1): Check may_trap_p on SET,
not INSN. Also check RTX_FRAME_RELATED_P. Queue insns for DF rescan
as needed.
(pass_cprop_hardreg::execute): Add df note problem and defer insn
rescans. Reprocess blocks as needed, calling df_analyze before
reprocessing. Always call df_analyze before fixing up debug bind
insns.
Thomas Koenig [Sun, 24 Mar 2019 12:51:19 +0000 (12:51 +0000)]
re PR fortran/78865 (ICE in create_tmp_var, at gimple-expr.c:473)
2019-03-24 Thomas Koenig <tkoenig@gcc.gnu.org>
PR fortran/78865
* interface.c (compare_actual_formal): Change errors about
missing or extra to gfc_error_now to make sure they are issued.
Change "spec" to "specifier" in message.
* resolve.c (resolve_global_procedure): Also check for mismatching
interface with global symbols if the namespace has already been
resolved.
2019-03-24 Thomas Koenig <tkoenig@gcc.gnu.org>
PR fortran/78865
* gfortran.dg/altreturn_10.f90: New test.
* gfortran.dg/whole_file_3.f90: Change dg-warning to dg-error.
projects / thirdparty / gcc.git / log
