so don't try to do tests with it.
2018-08-16 Iain Sandoe <iain@sandoe.co.uk>
gcc/testsuite/
* c-c++-common/asan/pointer-subtract-3.c: Skip for Darwin.
* c-c++-common/asan/pointer-subtract-4.c: Likewise.
* g++.dg/torture/pr44295.C : Likewise.
From-SVN: r263585
The original issue comes from the fact that the code does
... foo (... bar)
{
return bar;
}
The expansion of the return statement causes GCC to try to return the value in
a register. GCC will try to emit the move then, from MEM to REG (due to the SSA
temporary.). It checks for a mov optab for this which isn't available and
then tries to do the move in bits using emit_move_multi_word.
emit_move_multi_word will split the move into sub parts, but then needs to get
the sub parts and does this using subregs, but it's told it can't do subregs!
The compiler is now stuck in an infinite loop.
The way this is worked around in the back-end is that we have move patterns in
neon.md that usually just force the register instead of checking with the
back-end. This prevents emit_move_multi_word from being needed. However the
pattern for V4HF and V8HF were guarded by TARGET_NEON && TARGET_FP16.
I don't believe the TARGET_FP16 guard to be needed, because the pattern doesn't
actually generate code and requires another pattern for that, and a reg to reg move
should always be possible anyway. So allowing the force to register here is safe
and it allows the compiler to generate a correct error instead of ICEing in an
infinite loop.
gcc/
2018-08-16 Tamar Christina <tamar.christina@arm.com>
PR target/84711
* config/arm/arm.c (arm_can_change_mode_class): Disallow subreg.
* config/arm/neon.md (movv4hf, movv8hf): Refactored to..
(mov<mov>): ..this and enable unconditionally.
From-SVN: r263584
About 13 years ago the reg-to-reg patterns were split up, before that time
output_move_double could actually handle this case.
After the split was done most patterns were updated except for *neon_mov
which incorrectly retained reg,reg as a valid alternative.
However output_move_double can not handle this and simply returns ""
and asserts.
This pattern is essentially dead and I'm removing it for clarity.
gcc/
2018-08-16 Tamar Christina <tamar.christina@arm.com>
* config/arm/neon.md (*neon_mov<mode>): Remove reg-to-reg alternative.
From-SVN: r263582
This patch replaces the "Common" attribute in some aarch64 options with "Target"
to make them target-dependent. This doesn't affect their behaviour or how they
are documented with "--help", but does mean that they follow the option
specification that target-specific options are given the "Target" attribute.
Tested by ensuring the changed options are still accepted as before and that
they aren't invokable from an invalid target. Bootstrapped and regression tested
on aarch64-none-elf with ... regressions.
2018-08-16 Sam Tebbs <sam.tebbs@arm.com>
* config/aarch64/aarch64.opt (mlow-precision-recip-sqrt)
(mlow-precision-sqrt, mlow-precision-div, mverbose-cost-dump): Replace
"Common" with "Target".
From-SVN: r263578
[gcc]
2018-08-07 Will Schmidt <will_schmidt@vnet.ibm.com>
* config/rs6000/rs600.c (rs6000_gimple_fold_builtin): Add entries to
allow folding of mergeh() and mergel() for the float and double types.
(fold_mergehl_helper): Rework to handle building a permute tree
for float vectors.
From-SVN: r263569
The typedefs in <experimental/regex> and <experimental/string> don't
need to be in the __cxx11 namespace, because they are only aliases and
so will have the same mangled name as the underlying types.
* include/experimental/regex: Remove begin/end macros for namespace.
* include/experimental/string: Likewise.
* testsuite/experimental/polymorphic_allocator/pmr_typedefs_deque.cc:
New test.
* testsuite/experimental/polymorphic_allocator/
pmr_typedefs_forward_list.cc: New test.
* testsuite/experimental/polymorphic_allocator/pmr_typedefs_list.cc:
New test.
* testsuite/experimental/polymorphic_allocator/pmr_typedefs_map.cc:
New test.
* testsuite/experimental/polymorphic_allocator/pmr_typedefs_match.cc:
New test.
* testsuite/experimental/polymorphic_allocator/
pmr_typedefs_multimap.cc: New test.
* testsuite/experimental/polymorphic_allocator/
pmr_typedefs_multiset.cc: New test.
* testsuite/experimental/polymorphic_allocator/pmr_typedefs_set.cc:
New test.
* testsuite/experimental/polymorphic_allocator/pmr_typedefs_string.cc:
New test.
* testsuite/experimental/polymorphic_allocator/
pmr_typedefs_unordered_map.cc: New test.
* testsuite/experimental/polymorphic_allocator/
pmr_typedefs_unordered_multimap.cc: New test.
* testsuite/experimental/polymorphic_allocator/
pmr_typedefs_unordered_multiset.cc: New test.
* testsuite/experimental/polymorphic_allocator/
pmr_typedefs_unordered_set.cc: New test.
* testsuite/experimental/polymorphic_allocator/pmr_typedefs_vector.cc:
New test.
From-SVN: r263568
Remove duplicated logic in experimental::pmr::polymorphic_allocator by
calling the __uses_allocator_construct helper.
Fix bugs in std::pmr::polymorphic_allocator with incorrect SFINAE
constraint and incorrect argument order.
* include/bits/uses_allocator.h (__uses_allocator_construct): Qualify
calls to __uses_allocator_construct_impl and __use_alloc.
* include/experimental/memory_resource
(polymorphic_allocator::_M_construct): Remove.
(polymorphic_allocator::construct): Call __uses_allocator_construct.
Qualify calls to __use_alloc.
* include/std/memory_resource (polymorphic_allocator::construct): Fix
type in SFINAE constraint. Use constexpr if instead of tag dispatching
to _S_construct overloads.
(polymorphic_allocator::construct(pair<T1, T2>*, ...)): Fix order of
arguments to _S_construct_p.
(polymorphic_allocator::_S_construct): Remove.
(polymorphic_allocator::_S_construct_p): Return allocators by value
not by reference.
* include/std/scoped_allocator (scoped_allocator_adaptor::construct):
Qualify calls to __use_alloc.
* testsuite/20_util/polymorphic_allocator/construct_pair.cc: New test,
copied from testsuite/20_util/scoped_allocator/construct_pair.cc.
* testsuite/experimental/polymorphic_allocator/1.cc: New test.
* testsuite/experimental/polymorphic_allocator/construct_pair.cc:
New test.
From-SVN: r263566
This patch adds the ability to label source ranges within a rich_location,
to be printed by diagnostic_show_locus.
For example:
pr69554-1.c:11:18: error: invalid operands to binary + (have 'const char *' and 'const char *')
11 | return (p + 1) + (q + 1);
| ~~~~~~~ ^ ~~~~~~~
| | |
| | const char *
| const char *
The patch implements labels for various type mismatch errors in the C and
C++ frontends, and in -Wformat. I implemented it wherever accurate location
information was guaranteed (there are other places that could benefit, but
we need better location information in those places).
The labels can be disabled via -fno-diagnostics-show-labels.
Similarly:
param-type-mismatch.C: In function 'int test_1(int, int, float)':
param-type-mismatch.C:11:27: error: invalid conversion from 'int' to 'const char*' [-fpermissive]
11 | return callee_1 (first, second, third);
| ^~~~~~
| |
| int
param-type-mismatch.C:7:43: note: initializing argument 2 of 'int callee_1(int, const char*, float)'
7 | extern int callee_1 (int one, const char *two, float three);
| ~~~~~~~~~~~~^~~
where the first "error" describing the bad argument gets a label
describing the type inline (since it's non-obvious from "second").
The "note" describing the type of the param of the callee *doesn't*
get a label, since that information is explicit there in the
source ("const char *two").
The idea is that in any diagnostic where two aspects of the source aren't
in sync it ought to be easier for the user if we directly show them the
mismatching aspects inline (e.g. types).
As well as type mismatch errors, perhaps labels could also be used for
buffer overflow warnings, for describing the capacity of the destination
buffer vs the size of what's being written:
sprintf (buf, "filename: %s\n", file);
^~~ ~~~~~~~~~~~^~~
| |
capacity: 32 10 + strlen(file) + 2
or somesuch. Another idea might be for macro expansion warnings:
warning: repeated side effects in macro expansion...
x = MIN (p++, q++);
~~~~^~~~~~~~~~
note: ...expanded here as
#define MIN(X,Y) (X<Y?X:Y)
^~~ ~ ~ ~ ~ ~ ~
| | | | | |
| | | | | q++
| | | | p++
| | | q++
| q++ p++
p++
The patch removes some logic from multiline.exp which special-cased
lines ending with a '|' character (thus complicating testing of this
patch). I believe that this was a vestige from experiments I did to
support strippng dg directives from the output; it was present in the
earliest version of multiline.exp I posted:
"[RFC, stage1] Richer source location information for gcc 6 (location ranges etc)"
https://gcc.gnu.org/ml/gcc-patches/2015-03/msg00837.html
and I believe was neved used.
gcc/c-family/ChangeLog:
* c-format.c: Include "selftest-diagnostic.h" and
"gcc-rich-location.h".
(format_warning_at_char): Pass NULL for new label params of
format_warning_va.
(class indirection_suffix): New class.
(class range_label_for_format_type_mismatch): New class.
(format_type_warning): Move logic for generating "*" suffix to
class indirection_suffix. Create "fmt_label" and "param_label"
to show their types, and pass them to the
format_warning_at_substring calls.
(selftest::test_type_mismatch_range_labels): New test.
(selftest::c_format_c_tests): Call it.
gcc/c/ChangeLog:
* c-objc-common.c: Include "gcc-rich-location.h".
(c_tree_printer): Move implemenation of '%T' to...
(print_type): ...this new function.
(range_label_for_type_mismatch::get_text): New function.
* c-typeck.c (convert_for_assignment): Add type labels to the rhs
range for the various ic_argpass cases.
(class maybe_range_label_for_tree_type_mismatch): New class.
(build_binary_op): Use it when calling binary_op_error.
gcc/cp/ChangeLog:
* call.c: Include "gcc-rich-location.h".
(convert_like_real): Add range label for "invalid conversion"
diagnostic.
(perform_implicit_conversion_flags): Add type label to the
"could not convert" error.
* error.c: Include "gcc-rich-location.h".
(range_label_for_type_mismatch::get_text): New function.
* typeck.c (convert_for_assignment): Add type label to
the "cannot convert" error if a location is available.
gcc/ChangeLog:
* common.opt (fdiagnostics-show-labels): New option.
* diagnostic-show-locus.c (class layout_range): Add field
"m_label".
(class layout): Add field "m_show_labels_p".
(layout_range::layout_range): Add param "label" and use it to
initialize m_label.
(make_range): Pass in NULL for new "label" param of layout_range's
ctor.
(layout::layout): Initialize m_show_labels_p.
(layout::maybe_add_location_range): Pass in loc_range->m_label
when constructing layout_range instances.
(struct line_label): New struct.
(layout::print_any_labels): New member function.
(layout::print_line): Call it if label-printing is enabled.
(selftest::test_one_liner_labels): New test.
(selftest::test_diagnostic_show_locus_one_liner): Call it.
* diagnostic.c (diagnostic_initialize): Initialize
context->show_labels_p.
* diagnostic.h (struct diagnostic_context): Add field
"show_labels_p".
* doc/invoke.texi (Diagnostic Message Formatting Options): Add
-fno-diagnostics-show-labels.
* dwarf2out.c (gen_producer_string): Add
OPT_fdiagnostics_show_labels to the ignored options.
* gcc-rich-location.c (gcc_rich_location::add_expr): Add "label"
param.
(gcc_rich_location::maybe_add_expr): Likewise.
* gcc-rich-location.h (gcc_rich_location::gcc_rich_location): Add
label" param, defaulting to NULL.
(gcc_rich_location::add_expr): Add "label" param.
(gcc_rich_location::maybe_add_expr): Likewise.
(class text_range_label): New class.
(class range_label_for_type_mismatch): New class.
* gimple-ssa-sprintf.c (fmtwarn): Pass NULL for new label params
of format_warning_va.
(fmtwarn_n): Likewise for new params of format_warning_n_va.
* lto-wrapper.c (merge_and_complain): Add
OPT_fdiagnostics_show_labels to the "pick one setting" options.
(append_compiler_options): Likewise to the dropped options.
(append_diag_options): Likewise to the passed-on options.
* opts.c (common_handle_option): Handle the new option.
* selftest-diagnostic.c
(test_diagnostic_context::test_diagnostic_context): Enable
show_labels_p.
* substring-locations.c: Include "gcc-rich-location.h".
(format_warning_n_va): Add "fmt_label" and "param_label" params
and use them as appropriate.
(format_warning_va): Add "fmt_label" and "param_label" params,
passing them on to format_warning_n_va.
(format_warning_at_substring): Likewise.
(format_warning_at_substring_n): Likewise.
* substring-locations.h (format_warning_va): Add "fmt_label" and
"param_label" params.
(format_warning_n_va): Likewise.
(format_warning_at_substring): Likewise.
(format_warning_at_substring_n): Likewise.
* toplev.c (general_init): Initialize global_dc->show_labels_p.
gcc/testsuite/ChangeLog:
* g++.dg/diagnostic/aka3.C: New test.
* g++.dg/diagnostic/param-type-mismatch-2.C: Update expected
output to show range labels.
* g++.dg/diagnostic/param-type-mismatch.C: Likewise.
* g++.dg/plugin/plugin.exp (plugin_test_list): Add...
* g++.dg/plugin/show-template-tree-color-labels.C: New test.
* gcc.dg/bad-binary-ops.c: Update expected output to show range
labels. Add an "aka" example.
* gcc.dg/cpp/pr66415-1.c: Update expected output to show range
labels.
* gcc.dg/format/diagnostic-ranges.c: Likewise.
* gcc.dg/format/pr72858.c: Likewise.
* gcc.dg/format/pr78498.c: Likewise.
* gcc.dg/param-type-mismatch.c: Add "-Wpointer-sign" to options.
Update expected output to show range labels. Add examples of
-Wincompatible-pointer-types and -Wpointer-sign for parameters.
* gcc.dg/plugin/diagnostic-test-show-locus-bw-line-numbers.c:
Update expected output to show range labels.
* gcc.dg/plugin/diagnostic-test-show-locus-bw.c: Likewise.
(test_very_wide_line): Adjust so that label is at left-clipping
boundary.
(test_very_wide_line_2): New test.
* gcc.dg/plugin/diagnostic-test-show-locus-color-line-numbers.c:
Update expected output to show range labels.
* gcc.dg/plugin/diagnostic-test-show-locus-color.c: Likewise.
* gcc.dg/plugin/diagnostic-test-show-locus-no-labels.c: New test.
* gcc.dg/plugin/diagnostic_plugin_show_trees.c (show_tree): Update
for new param to gcc_rich_location::add_expr.
* gcc.dg/plugin/diagnostic_plugin_test_show_locus.c (add_range):
Add "label" param.
(test_show_locus): Add examples of labels to various tests. Tweak
the "very wide_line" test case and duplicate it, to cover the
boundary values for clipping of labels against the left-margin.
* gcc.dg/plugin/plugin.exp (plugin_test_list): Add
diagnostic-test-show-locus-no-labels.c.
* gcc.dg/pr69554-1.c: Update expected output to show range labels.
Update line numbers of dg-locus directives.
* gcc.dg/pr69627.c: Update expected output to show range labels.
* lib/multiline.exp (proc _build_multiline_regex): Remove
special-case handling of lines with trailing '|'.
libcpp/ChangeLog:
* include/line-map.h (struct location_range): Add "m_label" field.
(class rich_location): Add description of labels to leading
comment.
(rich_location::rich_location): Add "label" param, defaulting to
NULL.
(rich_location::add_range): Likewise.
(struct label_text): New struct.
(class range_label): New abstract base class.
* line-map.c (rich_location::rich_location): Add "label" param;
use it.
(rich_location::add_range): Likewise.
From-SVN: r263564
2018-08-15 Prathamesh Kulkarni <prathamesh.kulkarni@linaro.org>
c/
* c-decl.c (start_decl): Do not warn if variables is named as main
and is a local variable.
testsuite/
* gcc.dg/wmain.c: New test.
From-SVN: r263562
Since there is no branch splitting anymore, the loop is no longer
necessary: pool chunkification can be done in one step.
gcc/ChangeLog:
2018-08-15 Ilya Leoshkevich <iii@linux.ibm.com>
* config/s390/s390.c (s390_reorg): Remove loop.
From-SVN: r263557
2018-08-15 Iain Sandoe <iain@sandoe.co.uk>
gcc/c:
PR c/19315
* c-decl.c (finish_decl): Don't add the 'extern' storage class to
objects of unknown size.
gcc/testsuite:
PR c/19315
gcc.dg/graphite/pr82451.c: Make array 'a' an extern.
gcc.dg/redecl-10.c: Expect warnings for the static vars with unknown
size.
From-SVN: r263556
2018-08-15 Martin Liska <mliska@suse.cz>
PR tree-optimization/86925
* predict.c (expr_expected_value_1): When taking
later predictor, assign also probability.
Use fold_build2_initializer_loc in order to fold
the expression in -frounding-math.
2018-08-15 Martin Liska <mliska@suse.cz>
PR tree-optimization/86925
* gcc.dg/predict-20.c: New test.
* gcc.dg/predict-21.c: New test.
From-SVN: r263552
/cp
2018-08-15 Paolo Carlini <paolo.carlini@oracle.com>
* decl.c (check_previous_goto_1): When decl_jump_unsafe returns 2
emit an error instead of a permerror.
/testsuite
2018-08-15 Paolo Carlini <paolo.carlini@oracle.com>
* g++.dg/init/goto3.C: Adjust for error instead of permerror.
From-SVN: r263551
Backport of RISC-V support for libffi go closures
* src/riscv/ffi.c (ffi_call_go, ffi_prep_go_closure): New
functions.
(ffi_call_int): Renamed from ffi_call.
(ffi_call_asm, ffi_closure_inner): Adjust interface.
* src/riscv/ffitarget.h (FFI_GO_CLOSURES): Define.
* src/riscv/sysv.S (ffi_go_closure_asm): New function.
(ffi_closure_asm, ffi_call_asm): Update for adjusted interfaces.
From-SVN: r263550
* config/i386/i386.c (expand_vec_perm_movs): New method matching movs
patterns.
(expand_vec_perm_1): Try the new method.
* gcc.target/i386/sse2-movs.c: New test.
From-SVN: r263549
As explained in the PR, there's no reason to call the nothrow delete,
we can just use the normal one.
PR libstdc++/86954
* include/bits/stl_tempbuf.h (return_temporary_buffer): Use
non-placement delete.
From-SVN: r263542
2018-08-14 Janus Weil <janus@gcc.gnu.org>
PR fortran/86116
* interface.c (compare_type): Remove a CLASS/TYPE check.
(compare_type_characteristics): New function that behaves like the old
'compare_type'.
(gfc_check_dummy_characteristics, gfc_check_result_characteristics):
Call 'compare_type_characteristics' instead of 'compare_type'.
2018-08-14 Janus Weil <janus@gcc.gnu.org>
PR fortran/86116
* gfortran.dg/generic_34.f90: New test case.
From-SVN: r263540
Define a class using std::mutex for when std::atomic<memory_resource*>
cannot be used to implement the default memory resource.
When std::mutex constructor is not constexpr the constant_init trick
won't work, so just define a global and use init_priority for it. The
compiler warns about using reserved priority, so put the definition in a
header file using #pragma GCC system_header to suppress the warning.
PR libstdc++/86846
* src/c++17/default_resource.h: New file, defining default_res.
* src/c++17/memory_resource.cc [ATOMIC_POINTER_LOCK_FREE != 2]
(atomic_mem_res): Define alternative for atomic<memory_resource*>
using a mutex instead of atomics.
From-SVN: r263536
[ios::failure] p2: "When throwing ios_base::failure exceptions,
implementations should provide values of ec that identify the specific
reason for the failure."
This adds a new overload of __throw_ios_failure that can be passed
errno, to store error_code(errno, system_category()) in the exception
object.
PR libstdc++/85343
* acinclude.m4 (libtool_VERSION): Bump version.
* config/abi/pre/gnu.ver (GLIBCXX_3.4.26): Add new symbol version.
Export new symbol.
* configure: Regenerate.
* doc/xml/manual/abi.xml: Document new versions.
* include/bits/fstream.tcc (basic_filebuf<C, T>::underflow)
(basic_filebuf<C, T>::xsgetn): Pass errno to __throw_ios_failure.
* include/bits/functexcept.h (__throw_ios_failure(const char*, int)):
Declare new overload.
* src/c++11/cxx11-ios_failure.cc (__ios_failure): Add new constructor
and static member function.
(__throw_ios_failure(const char*, int)): Define.
* src/c++98/ios_failure.cc [!_GLIBCXX_USE_DUAL_ABI]
(__throw_ios_failure(const char*, int)): Define.
* testsuite/util/testsuite_abi.cc: Update known and latest versions.
From-SVN: r263535
Rope iterators sometimes contain pointers to an internal buffer
inside the iterator itself. When such an iterator is copied, the
copy incorrectly retains pointers to the original.
This patch takes the simple approach of not copying the cached
information when the internal buffer is being used, instead
requiring it to be recomputed when the copied iterator is
dereferenced. An alternative would be to adjust the pointers so
they refer to the buffer in the copy.
2018-08-14 Jeremy Sawicki <jeremy-gcc@sawicki.us>
* include/ext/rope (_Rope_iterator_base(const _Rope_iterator_base&))
(_Rope_const_iterator::operator=(const _Rope_const_iterator&))
(_Rope_iterator::operator=(const _Rope_iterator&)): Ensure
copied/assigned rope iterators don't retain pointers to the iterator
they were copied/assigned from.
* testsuite/ext/rope/7.cc: New.
From-SVN: r263534
* gcc/pretty-print.c (eat_esc_sequence): Swap the foreground and background
colors if the COMMON_LVB_REVERSE_VIDEO flag is set, and clear it thereafter,
as it only works for DBCS.
From-SVN: r263531
The getentropy function, found on Linux, OpenBSD, and recently also
FreeBSD, can be used to get random bytes to initialize the PRNG. It
is similar to the traditional way of reading from /dev/urandom, but
being a system call rather than a special file, it doesn't suffer from
problems like running out of file descriptors, or failure when running
in a container where /dev/urandom may not be available.
Regtested on x86_64-pc-linux-gnu, Ok for trunk?
2018-08-13 Janne Blomqvist <jb@gcc.gnu.org>
* configure.ac: Check for getentropy.
* intrinsics/random.c (getosrandom): Use getentropy if available.
* config.h.in: Regenerated.
* configure: Regenerated.
From-SVN: r263522
After the combine 2-2 changes, this testcase does not have a ZERO_EXTEND
in the intermediate code, but an AND instead.
gcc/testsuite/
* gcc.target/powerpc/pr56605.c: The generated code can have an AND
instead of a ZERO_EXTEND.
From-SVN: r263521
[gcc]
2018-07-06 Will Schmidt <will_schmidt@vnet.ibm.com>
* gcc/config/rs6000/rs6000.c (rs6000_gimple_fold_builtin): Add support
for folding vec_perm.
From-SVN: r263520
[gcc]
2018-08-13 Will Schmidt <will_schmidt@vnet.ibm.com>
* config/rs6000/rs6000.c (rs6000_gimple_fold_builtin):
Add support for gimple-folding of vec_pack() and vec_unpack()
intrinsics.
From-SVN: r263519
* libsupc++/new_opa.cc (operator new(size_t, align_val_t)): Use
__is_pow2 to check for valid alignment. Avoid branching when rounding
size to multiple of alignment.
From-SVN: r263515
