RISC-V/libgcc: Reduce the size of RV64 millicode by 6 bytes
Rewrite code sequences throughout the 64-bit RISC-V `__riscv_save_*'
routines replacing `li t1, -48', `li t1, -64', and `li t1, -80',
instructions, which do not have a compressed encoding, respectively with
`li t1, 3', `li t1, 4', and `li t1, 4', which do, and then adjusting the
remaining code accordingly observing that `sub sp, sp, t1' takes the
same amount of space as an `slli t1, t1, 4'/`add sp, sp, t1' instruction
pair does, again due to the use of compressed encodings, saving 6 bytes
total.
This change does increase code size by 4 bytes for RISC-V processors
lacking the compressed instruction set, however their users couldn't
care about the code size or they would have chosen an implementation
that does have the compressed instructions, wouldn't they?
libgcc/
* config/riscv/save-restore.S [__riscv_xlen == 64]
(__riscv_save_10, __riscv_save_8, __riscv_save_6, __riscv_save_4)
(__riscv_save_2): Replace negative immediates used for the final
stack pointer adjustment with positive ones, right-shifted by 4.
François Dumont [Tue, 21 Jan 2020 06:18:08 +0000 (07:18 +0100)]
libstdc++: Fix and improve std::vector<bool> implementation.
Do not consider allocator noexcept qualification for vector<bool> move
constructor.
Improve swap performance using TBAA like in main vector implementation. Bypass
_M_initialize_dispatch/_M_assign_dispatch in post-c++11 modes.
libstdc++-v3/ChangeLog:
* include/bits/stl_bvector.h
[_GLIBCXX_INLINE_VERSION](_Bvector_impl_data::_M_start): Define as
_Bit_type*.
(_Bvector_impl_data(const _Bvector_impl_data&)): Default.
(_Bvector_impl_data(_Bvector_impl_data&&)): Delegate to latter.
(_Bvector_impl_data::operator=(const _Bvector_impl_data&)): Default.
(_Bvector_impl_data::_M_move_data(_Bvector_impl_data&&)): Use latter.
(_Bvector_impl_data::_M_reset()): Likewise.
(_Bvector_impl_data::_M_swap_data): New.
(_Bvector_impl::_Bvector_impl(_Bvector_impl&&)): Implement explicitely.
(_Bvector_impl::_Bvector_impl(_Bit_alloc_type&&, _Bvector_impl&&)): New.
(_Bvector_base::_Bvector_base(_Bvector_base&&, const allocator_type&)):
New, use latter.
(vector::vector(vector&&, const allocator_type&, true_type)): New, use
latter.
(vector::vector(vector&&, const allocator_type&, false_type)): New.
(vector::vector(vector&&, const allocator_type&)): Use latters.
(vector::vector(const vector&, const allocator_type&)): Adapt.
[__cplusplus >= 201103](vector::vector(_InputIt, _InputIt,
const allocator_type&)): Use _M_initialize_range.
(vector::operator[](size_type)): Use iterator operator[].
(vector::operator[](size_type) const): Use const_iterator operator[].
(vector::swap(vector&)): Add assertions on allocators. Use _M_swap_data.
[__cplusplus >= 201103](vector::insert(const_iterator, _InputIt,
_InputIt)): Use _M_insert_range.
(vector::_M_initialize(size_type)): Adapt.
[__cplusplus >= 201103](vector::_M_initialize_dispatch): Remove.
[__cplusplus >= 201103](vector::_M_insert_dispatch): Remove.
* python/libstdcxx/v6/printers.py (StdVectorPrinter._iterator): Stop
using start _M_offset.
(StdVectorPrinter.to_string): Likewise.
* testsuite/23_containers/vector/bool/allocator/swap.cc: Adapt.
* testsuite/23_containers/vector/bool/cons/noexcept_move_construct.cc:
Add check.
Jakub Jelinek [Fri, 31 Jul 2020 21:08:00 +0000 (23:08 +0200)]
c++: Use error_at rather than warning_at for missing return in constexpr functions [PR96182]
For C++11 we already emit an error if a constexpr function doesn't contain
a return statement, because in C++11 that is the only thing it needs to
contain, but for C++14 we would normally issue a -Wreturn-type warning.
As mentioned by Jonathan, such constexpr functions are invalid, no
diagnostics required, because there doesn't exist any arguments for
which it would result in valid constant expression.
This raises it to an error in such cases. The !LAMBDA_TYPE_P case
is to avoid error on g++.dg/pr81194.C where the user didn't write
constexpr anywhere and the operator() is compiler generated.
2020-07-31 Jakub Jelinek <jakub@redhat.com>
PR c++/96182
* decl.c (finish_function): In constexpr functions use for C++14 and
later error instead of warning if no return statement is present and
diagnose it regardless of warn_return_type. Move the warn_return_type
diagnostics earlier in the function.
Jonathan Wakely [Fri, 31 Jul 2020 18:58:03 +0000 (19:58 +0100)]
libstdc++: Avoid using __float128 in strict modes
libstdc++-v3/ChangeLog:
* testsuite/26_numerics/numbers/float128.cc: Check
__STRICT_ANSI__ before using __float128.
* testsuite/std/concepts/concepts.lang/concept.arithmetic/floating_point.cc:
Likewise.
Jonathan Wakely [Fri, 31 Jul 2020 18:58:02 +0000 (19:58 +0100)]
libstdc++: Add dg-require-effective-target to std::span assert tests
The current dg directives say that the tests can run for any standard
mode, but should fail for C++20. What we want is that they only run for
C++20, and are always expected to fail.
Jonathan Wakely [Fri, 31 Jul 2020 18:58:02 +0000 (19:58 +0100)]
libstdc++: Use c++NN_only effective target to tests
Some tests really are only intended for a specific -std mode, so add a
target selector to make that explicit.
Also reorder the dg-do directives to come after the dg-options ones, so
that the target selector in the dg-do directive is applied after the
dg-options that sets the -std option.
libstdc++-v3/ChangeLog:
* testsuite/20_util/reference_wrapper/83427.cc: Adjust
effective-target to specific language mode only.
* testsuite/24_iterators/headers/iterator/range_access_c++11.cc:
Likewise.
* testsuite/24_iterators/headers/iterator/range_access_c++14.cc:
Likewise.
* testsuite/24_iterators/headers/iterator/synopsis_c++11.cc:
Likewise.
* testsuite/24_iterators/headers/iterator/synopsis_c++14.cc:
Likewise.
* testsuite/26_numerics/valarray/69116.cc:
Likewise.
* testsuite/30_threads/headers/condition_variable/std_c++0x_neg.cc:
Remove whitespace at end of file.
* testsuite/30_threads/headers/future/std_c++0x_neg.cc:
Likewise.
Roger Sayle [Thu, 30 Jul 2020 09:42:06 +0000 (11:42 +0200)]
nvptx: Define TARGET_TRULY_NOOP_TRUNCATION to false
Many thanks to Richard Biener for approving the midde-end
patch that cleared the way for this one. This nvptx patch
defines the target hook TARGET_TRULY_NOOP_TRUNCATION to
false, indicating that integer truncations require explicit
instructions. nvptx.c already defines TARGET_MODES_TIEABLE_P
and TARGET_CAN_CHANGE_MODE_CLASS to false, and as (previously)
documented that may require TARGET_TRULY_NOOP_TRUNCATION to
be defined likewise.
This patch decreases the number of unexpected failures in
the testsuite by 10, and increases the number of expected
passes by 4, including these previous FAILs/ICEs:
gcc.c-torture/compile/opout.c
gcc.dg/torture/pr79125.c
gcc.dg/tree-ssa/pr92085-1.c
Unfortunately there is one testsuite failure that used to
pass gcc.target/nvptx/v2si-cvt.c, but this isn't an ICE or
incorrect code. This regression has been filed as PR96403,
and the failing scan-assembler directives have been replaced
by a reference to the PR.
This patch has been tested on nvptx-none hosted on
x86_64-pc-linux-gnu with "make" and "make check" with
fewer ICEs and no wrong code regressions.
2020-07-31 Roger Sayle <roger@nextmovesoftware.com>
Tom de Vries <tdevries@suse.de>
Jonathan Wakely [Fri, 31 Jul 2020 16:51:00 +0000 (17:51 +0100)]
libstdc++: Adjust tests that give different results in C++20
libstdc++-v3/ChangeLog:
* testsuite/20_util/is_aggregate/value.cc: Adjust for changes to
definition of aggregates in C++20.
* testsuite/20_util/optional/requirements.cc: Adjust for
defaulted comparisons in C++20.
Jonathan Wakely [Fri, 31 Jul 2020 16:51:00 +0000 (17:51 +0100)]
libstdc++: Add -Wno-deprecated for tests that warn in C++20
libstdc++-v3/ChangeLog:
* testsuite/20_util/tuple/78939.cc: Suppress warnings about
deprecation of volatile-qualified structured bindings in C++20.
* testsuite/20_util/variable_templates_for_traits.cc: Likewise
for deprecation of is_pod in C++20
d: Split up the grouped compilable and runnable tests.
The majority of tests in runnable are really compilable/ICE tests, and
have have dg-do adjusted where necessary. Tests that had a dependency
on Phobos have also been reproduced and reduced with all imports
stripped from the test.
The end result is a collection of tests that only check the compiler bug
that was being fixed, rather than the library, and a reduction in time
spent running all tests.
gcc/testsuite/ChangeLog:
* gdc.dg/compilable.d: Removed.
* gdc.dg/gdc108.d: New test.
* gdc.dg/gdc115.d: New test.
* gdc.dg/gdc121.d: New test.
* gdc.dg/gdc122.d: New test.
* gdc.dg/gdc127.d: New test.
* gdc.dg/gdc131.d: New test.
* gdc.dg/gdc133.d: New test.
* gdc.dg/gdc141.d: New test.
* gdc.dg/gdc142.d: New test.
* gdc.dg/gdc15.d: New test.
* gdc.dg/gdc17.d: New test.
* gdc.dg/gdc170.d: New test.
* gdc.dg/gdc171.d: New test.
* gdc.dg/gdc179.d: New test.
* gdc.dg/gdc183.d: New test.
* gdc.dg/gdc186.d: New test.
* gdc.dg/gdc187.d: New test.
* gdc.dg/gdc19.d: New test.
* gdc.dg/gdc191.d: New test.
* gdc.dg/gdc194.d: New test.
* gdc.dg/gdc196.d: New test.
* gdc.dg/gdc198.d: New test.
* gdc.dg/gdc200.d: New test.
* gdc.dg/gdc204.d: New test.
* gdc.dg/gdc210.d: New test.
* gdc.dg/gdc212.d: New test.
* gdc.dg/gdc213.d: New test.
* gdc.dg/gdc218.d: New test.
* gdc.dg/gdc223.d: New test.
* gdc.dg/gdc231.d: New test.
* gdc.dg/gdc239.d: New test.
* gdc.dg/gdc24.d: New test.
* gdc.dg/gdc240.d: New test.
* gdc.dg/gdc241.d: New test.
* gdc.dg/gdc242a.d: New test.
* gdc.dg/gdc242b.d: New test.
* gdc.dg/gdc248.d: New test.
* gdc.dg/gdc250.d: New test.
* gdc.dg/gdc251.d: New test.
* gdc.dg/gdc253a.d: New test.
* gdc.dg/gdc253b.d: New test.
* gdc.dg/gdc255.d: New test.
* gdc.dg/gdc256.d: New test.
* gdc.dg/gdc261.d: New test.
* gdc.dg/gdc27.d: New test.
* gdc.dg/gdc273.d: New test.
* gdc.dg/gdc280.d: New test.
* gdc.dg/gdc284.d: New test.
* gdc.dg/gdc285.d: New test.
* gdc.dg/gdc286.d: New test.
* gdc.dg/gdc300.d: New test.
* gdc.dg/gdc309.d: New test.
* gdc.dg/gdc31.d: New test.
* gdc.dg/gdc35.d: New test.
* gdc.dg/gdc36.d: New test.
* gdc.dg/gdc37.d: New test.
* gdc.dg/gdc4.d: New test.
* gdc.dg/gdc43.d: New test.
* gdc.dg/gdc47.d: New test.
* gdc.dg/gdc51.d: New test.
* gdc.dg/gdc57.d: New test.
* gdc.dg/gdc66.d: New test.
* gdc.dg/gdc67.d: New test.
* gdc.dg/gdc71.d: New test.
* gdc.dg/gdc77.d: New test.
* gdc.dg/imports/gdc239.d: Remove phobos dependency.
* gdc.dg/imports/gdc241a.d: Updated imports.
* gdc.dg/imports/gdc241b.d: Likewise.
* gdc.dg/imports/gdc251a.d: Likewise.
* gdc.dg/imports/gdc253.d: Rename to...
* gdc.dg/imports/gdc253a.d: ...this.
* gdc.dg/imports/gdc253b.d: New.
* gdc.dg/imports/gdc36.d: New.
* gdc.dg/imports/runnable.d: Removed.
* gdc.dg/link.d: Removed.
* gdc.dg/runnable.d: Removed.
* gdc.dg/runnable2.d: Removed.
* gdc.dg/simd.d: Remove phobos dependency.
For 32-bit btr(), BIT_NOT_EXPR was being generated twice, something that
was not seen with the 64-bit variant. Removed the second call to fix
the generated code.
Richard Biener [Thu, 30 Jul 2020 09:46:43 +0000 (11:46 +0200)]
debug/96383 - emit debug info for used external functions
This makes sure to emit full declaration DIEs including
formal parameters for used external functions. This helps
debugging when debug information of the external entity is
not available and also helps external tools cross-checking
ABI compatibility which was the bug reporters use case.
For cc1 this affects debug information size as follows:
and DWARF compression via DWZ can only shave off minor bits
here.
Previously we emitted no DIEs for external functions at all
unless they were referenced via DW_TAG_GNU_call_site which
for some GCC revs caused a regular DIE to appear and since
GCC 4.9 only a stub without formal parameters. This means
at -O0 we did not emit any DIE for external functions
but with optimization we emitted stubs.
2020-07-30 Richard Biener <rguenther@suse.de>
PR debug/96383
* langhooks-def.h (lhd_finalize_early_debug): Declare.
(LANG_HOOKS_FINALIZE_EARLY_DEBUG): Define.
(LANG_HOOKS_INITIALIZER): Amend.
* langhooks.c: Include cgraph.h and debug.h.
(lhd_finalize_early_debug): Default implementation from
former code in finalize_compilation_unit.
* langhooks.h (lang_hooks::finalize_early_debug): Add.
* cgraphunit.c (symbol_table::finalize_compilation_unit):
Call the finalize_early_debug langhook.
gcc/c-family/
* c-common.h (c_common_finalize_early_debug): Declare.
* c-common.c: Include debug.h.
(c_common_finalize_early_debug): finalize_early_debug langhook
implementation generating debug for extern declarations.
gcc/c/
* c-objc-common.h (LANG_HOOKS_FINALIZE_EARLY_DEBUG):
Define to c_common_finalize_early_debug.
gcc/cp/
* cp-objcp-common.h (LANG_HOOKS_FINALIZE_EARLY_DEBUG):
Define to c_common_finalize_early_debug.
gcc/testsuite/
* gcc.dg/debug/dwarf2/pr96383-1.c: New testcase.
* gcc.dg/debug/dwarf2/pr96383-2.c: Likewise.
libstdc++-v3/
* testsuite/20_util/assume_aligned/3.cc: Use -g0.
to make the simplification only apply in case both plus operations
in the result end up simplified to a simple operand.
2020-07-31 Richard Biener <rguenther@suse.de>
* genmatch.c (expr::force_leaf): Add and initialize.
(expr::gen_transform): Honor force_leaf by passing
NULL as sequence argument to maybe_push_res_to_seq.
(parser::parse_expr): Allow ! marker on result expression
operations.
* doc/match-and-simplify.texi: Amend.
Kewen Lin [Fri, 31 Jul 2020 12:49:39 +0000 (07:49 -0500)]
vect: Don't consider branch costs if no peeled iterations
Currently vectorizer cost modeling counts branch taken costs for
prologue and epilogue if the number of iterations is unknown.
But it isn't sensible if there are no peeled iterations.
This patch is to guard them under peel_iters_prologue > 0 or
peel_iters_epilogue > 0.
Bootstrapped/regtested on powerpc64le-linux-gnu and aarch64-linux-gnu.
gcc/ChangeLog:
* tree-vect-loop.c (vect_get_known_peeling_cost): Don't consider branch
taken costs for prologue and epilogue if they don't exist.
(vect_estimate_min_profitable_iters): Likewise.
gcc/testsuite/ChangeLog:
* gcc.target/aarch64/sve/cost_model_2.c: Adjust due to cost model
change.
Richard Biener [Fri, 31 Jul 2020 06:41:56 +0000 (08:41 +0200)]
middle-end/96369 - fix missed short-circuiting during range folding
This makes the special case of constant evaluated LHS for a
short-circuiting or/and explicit rather than doing range
merging and eventually exposing a side-effect that shouldn't be
evaluated.
Richard Biener [Fri, 24 Jul 2020 11:44:09 +0000 (13:44 +0200)]
Improve var-tracking dataflow iteration order
This builds upon the rev_post_order_and_mark_dfs_back_seme improvements
and makes vt_find_locations iterate over the dataflow problems for
each toplevel SCC separately, improving memory locality and avoiding
to process nodes after the SCC before the SCC itself stabilized.
On the asan_interceptors.cc testcase this for example reduces the
number of visited blocks from 3751 to 2867. For stage3-gcc
this reduces the number of visited blocks by ~4%.
2020-07-28 Richard Biener <rguenther@suse.de>
PR debug/78288
* var-tracking.c (vt_find_locations): Use
rev_post_order_and_mark_dfs_back_seme and separately iterate
over toplevel SCCs.
Richard Biener [Mon, 20 Jul 2020 11:38:16 +0000 (13:38 +0200)]
Compute RPO with adjacent SCC members, expose toplevel SCC extents
This produces a more optimal RPO order for iteration processing
by making sure that SCC exits are processed before SCC members
themselves.. This avoids iterating blocks unrelated to the current
iteration for RPO VN and has the chance to improve code-generation
for the non-iterative mode of RPO VN. The patch also exposes toplevel
SCCs and gets rid of the ad-hoc max_rpo computation in RPO VN.
For simplicity it also removes the odd reverse ordering of the RPO
array returned from rev_post_order_and_mark_dfs_back_seme.
Overall reduction in the number of visited blocks isn't spectacular
for bootstrap (~2.5%) but single cases see up to a 10% reduction.
The same function can be used to optimize var-tracking iteration order
as seen in the followup.
2020-07-28 Richard Biener <rguenther@suse.de>
* cfganal.h (rev_post_order_and_mark_dfs_back_seme): Adjust
prototype.
* cfganal.c (rpoamdbs_bb_data): New struct with pre BB data.
(tag_header): New helper.
(cmp_edge_dest_pre): Likewise.
(rev_post_order_and_mark_dfs_back_seme): Compute SCCs,
find SCC exits and perform a DFS walk with extra edges to
compute a RPO with adjacent SCC members when requesting an
iteration optimized order and populate the toplevel SCC array.
* tree-ssa-sccvn.c (do_rpo_vn): Remove ad-hoc computation
of max_rpo and fill it in from SCC extent info instead.
Patrick Palka [Fri, 31 Jul 2020 02:21:41 +0000 (22:21 -0400)]
c++: decl_constant_value and unsharing [PR96197]
In the testcase from the PR we're seeing excessive memory use (> 5GB)
during constexpr evaluation, almost all of which is due to the call to
decl_constant_value in the VAR_DECL/CONST_DECL branch of
cxx_eval_constant_expression. We reach here every time we evaluate an
ARRAY_REF of a constexpr VAR_DECL, and from there decl_constant_value
makes an unshared copy of the VAR_DECL's initializer. But unsharing
here is unnecessary because callers of cxx_eval_constant_expression
already unshare its result when necessary.
To fix this excessive unsharing, this patch adds a new defaulted
parameter unshare_p to decl_really_constant_value and
decl_constant_value so that callers can control whether to unshare.
As a simplification, we can also move the call to unshare_expr in
constant_value_1 outside of the loop, since doing unshare_expr on a
DECL_P is a no-op.
Now that we no longer unshare the result of decl_constant_value and
decl_really_constant_value from cxx_eval_constant_expression, memory use
during constexpr evaluation for the testcase from the PR falls from ~5GB
to 15MB according to -ftime-report.
gcc/cp/ChangeLog:
PR c++/96197
* constexpr.c (cxx_eval_constant_expression) <case CONST_DECL>:
Pass false to decl_constant_value and decl_really_constant_value
so that they don't unshare their result.
* cp-tree.h (decl_constant_value): New declaration with an added
bool parameter.
(decl_really_constant_value): Add bool parameter defaulting to
true to existing declaration.
* init.c (constant_value_1): Add bool parameter which controls
whether to unshare the initializer before returning. Call
unshare_expr at most once.
(scalar_constant_value): Pass true to constant_value_1's new
bool parameter.
(decl_really_constant_value): Add bool parameter and forward it
to constant_value_1.
(decl_constant_value): Likewise, but instead define a new
overload with an added bool parameter.
gcc/testsuite/ChangeLog:
PR c++/96197
* g++.dg/cpp1y/constexpr-array8.C: New test.
d: Fix associative array literals that don't have alignment holes filled
Associative array literal keys with alignment holes are now filled using
memset() prior to usage, with LTR evaluation of side-effects enforced.
gcc/d/ChangeLog:
PR d/96152
* d-codegen.cc (build_array_from_exprs): New function.
* d-tree.h (build_array_from_exprs): Declare.
* expr.cc (ExprVisitor::visit (AssocArrayLiteralExp *)): Use
build_array_from_exprs to generate key and value arrays.
Both intrinsics did not handle the case where the va_list object comes
from a ref parameter.
gcc/d/ChangeLog:
PR d/96140
* intrinsics.cc (expand_intrinsic_vaarg): Handle ref parameters as
arguments to va_arg().
(expand_intrinsic_vastart): Handle ref parameters as arguments to
va_start().
Jonathan Wakely [Thu, 30 Jul 2020 19:58:09 +0000 (20:58 +0100)]
libstdc++: Make COW string use allocator_traits for nested types
When compiled as C++20 the COW std::string fails due to assuming that
the allocator always defines size_type and difference_type. That has
been incorrect since C++11, but we got away with it for specializations
using std::allocator until those members were removed in C++20.
libstdc++-v3/ChangeLog:
* include/bits/basic_string.h (size_type, difference_type):
Use allocator_traits to obtain the allocator's size_type and
difference_type.
Jonathan Wakely [Thu, 30 Jul 2020 19:55:56 +0000 (20:55 +0100)]
libstdc++: Check _GLIBCXX_USE_C99_STDLIB for strtof and strtold
On broken systems we only have strtod, not strtof and strtold. Just use
strtod for all types, even though that will produce incorrect results in
some cases.
Similarly, if _GLIBCXX_USE_C99_MATH is not defined then std::isinf won't
be declared. Just refer to it unqualified, which should find the C
library's isinf macro if that hasn't been #undef'd by <cmath>.
libstdc++-v3/ChangeLog:
* src/c++17/floating_from_chars.cc (from_chars_impl): Use
isinf unqualified.
[!_GLIBCXX_USE_C99_STDLIB]: Use strtod for float and long
double.
Jonathan Wakely [Thu, 30 Jul 2020 17:41:47 +0000 (18:41 +0100)]
libstdc++: Fix tests using wrong allocator type
libstdc++-v3/ChangeLog:
* testsuite/23_containers/unordered_multiset/cons/noexcept_default_construct.cc:
Use allocator with the correct value type.
* testsuite/23_containers/unordered_set/cons/noexcept_default_construct.cc:
Likewise.
d: Inline bounds checking for simple array assignments.
This optimizes the code generation of simple array assignments, inlining
the array bounds checking code so there is no reliance on the library
routine _d_arraycopy(), which also deals with postblit and copy
constructors for non-trivial arrays.
d: Refactor use of built-in memcmp/memcpy/memset into helper functions.
Generating calls to memset, memcpy, and memcmp is frequent enough that
it becomes beneficial to put them into their own routine. All parts of
the front-end have been updated to call the new helper functions instead
of doing it themselves.
d: Move private functions out of ExprVisitor into local statics
None of these functions need access to the context pointer of the
visitor class, so have been made free standing.
gcc/d/ChangeLog:
* expr.cc (needs_postblit): Move out of ExprVisitor as a static
function. Update all callers.
(needs_dtor): Likewise.
(lvalue_p): Likewise.
(binary_op): Likewise.
(binop_assignment): Likewise.
H.J. Lu [Wed, 15 Jul 2020 13:16:01 +0000 (06:16 -0700)]
Require CET support only for the final GCC build
With --enable-cet, require CET support only for the final GCC build.
Don't enable CET without CET support for non-bootstrap build, in stage1
nor for build support.
config/
PR bootstrap/96202
* cet.m4 (GCC_CET_HOST_FLAGS): Don't enable CET without CET
support in stage1 nor for build support.
Jonathan Wakely [Thu, 30 Jul 2020 11:23:55 +0000 (12:23 +0100)]
libstdc++: cv bool can't be an integer-like type (LWG 3467)
libstdc++-v3/ChangeLog:
* include/bits/iterator_concepts.h (__detail::__cv_bool): New
helper concept.
(__detail::__integral_nonbool): Likewise.
(__detail::__is_integer_like): Use __integral_nonbool.
* testsuite/std/ranges/access/lwg3467.cc: New test.
Jonathan Wakely [Thu, 30 Jul 2020 11:23:54 +0000 (12:23 +0100)]
libstdc++: Make testsuite usable with -fno-exceptions
Previously it was not possible to add -fno-exceptions to the testsuite
flags, because some files that are compiled by the v3-build_support
procedure failed with exceptions disabled.
This adjusts those files to still compile without exceptions (with
degraded functionality in some cases).
The sole testcase that explicitly checks for -fno-exceptions has also
been adjusted to use the more robust exceptions_enabled effective-target
keyword from gcc/testsuite/lib/target-supports.exp.
libstdc++-v3/ChangeLog:
* testsuite/23_containers/vector/bool/72847.cc: Use the
exceptions_enabled effective-target keyword instead of
checking for an explicit -fno-exceptions option.
* testsuite/util/testsuite_abi.cc (examine_symbol): Remove
redundant try-catch.
* testsuite/util/testsuite_allocator.h [!__cpp_exceptions]:
Do not define check_allocate_max_size and memory_resource.
* testsuite/util/testsuite_containers.h: Replace comment with
#error if wrong standard dialect used.
* testsuite/util/testsuite_shared.cc: Likewise.
d: Refactor matching and lowering of intrinsic functions.
Intrinsics are now matched explicitly, rather than through a common
alias where there are multiple overrides for a common intrinsic.
Where there is a corresponding DECL_FUNCTION_CODE, that is now stored in
the D intrinsic array. All run-time std.math intrinsics have been
removed, as the library implementation already forwards to core.math.
gcc/d/ChangeLog:
* d-tree.h (DEF_D_INTRINSIC): Rename second argument from A to B.
* intrinsics.cc (intrinsic_decl): Add built_in field.
(DEF_D_INTRINSIC): Rename second argument from ALIAS to BUILTIN.
(maybe_set_intrinsic): Handle new intrinsic codes.
(expand_intrinsic_bt): Likewise.
(expand_intrinsic_checkedint): Likewise.
(expand_intrinsic_bswap): Remove.
(expand_intrinsic_sqrt): Remove.
(maybe_expand_intrinsic): Group together intrinsic cases that map
directly to gcc built-ins.
* intrinsics.def (DEF_D_BUILTIN): Rename second argument from A to B.
Update all callers to pass equivalent DECL_FUNCTION_CODE.
(DEF_CTFE_BUILTIN): Likewise.
(STD_COS): Remove intrinsic.
(STD_FABS): Remove intrinsic.
(STD_LDEXP): Remove intrinsic.
(STD_RINT): Remove intrinsic.
(STD_RNDTOL): Remove intrinsic.
(STD_SIN): Remove intrinsic.
(STD_SQRTF): Remove intrinsic.
(STD_SQRT): Remove intrinsic.
(STD_SQRTL): Remove intrinsic.
Richard Biener [Thu, 30 Jul 2020 08:24:42 +0000 (10:24 +0200)]
tree-optimization/96370 - make reassoc expr rewrite more robust
In the face of the more complex tricks in reassoc with respect
to negate processing it can happen that the expression rewrite
is fooled to recurse on a leaf and pick up a bogus expression
code. The following patch makes the expression rewrite more
robust in providing the expression code to it directly since
it is the same for all operations in a chain.
2020-07-30 Richard Biener <rguenther@suse.de>
PR tree-optimization/96370
* tree-ssa-reassoc.c (rewrite_expr_tree): Add operation
code parameter and use it instead of picking it up from
the stmt that is being rewritten.
(reassociate_bb): Pass down the operation code.
Roger Sayle [Thu, 30 Jul 2020 08:23:38 +0000 (10:23 +0200)]
nvptx: Provide vec_set<mode> and vec_extract<vmode><mode> patterns
This patch provides standard vec_extract and vec_set patterns to the
nvptx backend, to extract an element from a PTX vector and set an
element of a PTX vector respectively. PTX vectors (I hesitate to
call them SIMD vectors) may contain up to four elements, so vector
modes up to size four are supported by this patch even though the
nvptx backend currently only allows V2SI and V2DI, i.e. two out
of the ten possible vector modes.
As an example of the improvement, the following C function:
typedef int __v2si __attribute__((__vector_size__(8)));
int foo (__v2si arg) { return arg[0]+arg[1]; }
I've implemented these getters and setters as their own instructions
instead of attempting the much more intrusive patch of changing the
backend's definition of register_operand. Given the limited utility
of PTX vectors, I'm not convinced that attempting to support them as
operands in every instruction would be worth the effort involved.
This patch has been tested on nvptx-none hosted on x86_64-pc-linux-gnu
with "make" and "make check" with no new regressions.
2020-07-15 Roger Sayle <roger@nextmovesoftware.com>
Tom de Vries <tdevries@suse.de>
gcc/ChangeLog:
* config/nvptx/nvptx.md (nvptx_vector_index_operand): New predicate.
(VECELEM): New mode attribute for a vector's uppercase element mode.
(Vecelem): New mode attribute for a vector's lowercase element mode.
(*vec_set<mode>_0, *vec_set<mode>_1, *vec_set<mode>_2)
(*vec_set<mode>_3): New instructions.
(vec_set<mode>): New expander to generate one of the above insns.
(vec_extract<mode><Vecelem>): New instruction.
gcc/testsuite/ChangeLog:
* gcc.target/nvptx/v2si-vec-set-extract.c: New test.
Patrick Palka [Thu, 30 Jul 2020 02:06:44 +0000 (22:06 -0400)]
c++: overload sets and placeholder return type [PR64194]
In the testcase below, template argument deduction for the call
g(id<int>) goes wrong because the functions in the overload set id<int>
each have a yet-undeduced auto return type, and this undeduced return
type makes try_one_overload fail to match up any of the overloads with
g's parameter type, leading to g's template argument going undeduced and
to the overload set going unresolved.
This patch fixes this issue by performing return type deduction via
instantiation before doing try_one_overload, in a manner similar to what
resolve_address_of_overloaded_function does.
gcc/cp/ChangeLog:
PR c++/64194
* pt.c (resolve_overloaded_unification): If the function
template specialization has a placeholder return type,
then instantiate it before attempting unification.
gcc/testsuite/ChangeLog:
PR c++/64194
* g++.dg/cpp1y/auto-fn60.C: New test.
Patrick Palka [Thu, 30 Jul 2020 02:06:41 +0000 (22:06 -0400)]
c++: alias_ctad_tweaks and constrained dguide [PR95486]
In the below testcase, we're ICEing from alias_ctad_tweaks ultimately
because the implied deduction guide for X's user-defined constructor
already has constraints associated with it. We then carry over these
constraints to 'fprime', the overlying deduction guide for the alias
template Y, via tsubst_decl from alias_ctad_tweaks. Later in
alias_ctad_tweaks we call get_constraints followed by set_constraints
without doing remove_constraints in between, which triggers the !found
assert in set_constraints.
This patch fixes this issue by adding an intervening call to
remove_constraints.
gcc/cp/ChangeLog:
PR c++/95486
* pt.c (alias_ctad_tweaks): Call remove_constraints before
calling set_constraints.
gcc/testsuite/ChangeLog:
PR c++/95486
* g++.dg/cpp2a/class-deduction-alias3.C: New test.
Patrick Palka [Thu, 30 Jul 2020 02:06:36 +0000 (22:06 -0400)]
c++: abbreviated function template friend matching [PR96106]
In the below testcase, duplicate_decls wasn't merging the tsubsted
friend declaration for 'void add(auto)' with its definition, because
reduce_template_parm_level (during tsubst_friend_function) lost the
DECL_VIRTUAL_P flag on the auto's invented template parameter, which
caused template_heads_equivalent_p to deem the two template heads as not
equivalent in C++20 mode.
This patch makes reduce_template_parm_level carry over the
DECL_VIRTUAL_P flag from the original TEMPLATE_PARM_DECL.
gcc/cp/ChangeLog:
PR c++/96106
* pt.c (reduce_template_parm_level): Propagate DECL_VIRTUAL_P
from the original TEMPLATE_PARM_DECL to the new lowered one.
gcc/testsuite/ChangeLog:
PR c++/96106
* g++.dg/concepts/abbrev7.C: New test.
Patrick Palka [Thu, 30 Jul 2020 02:06:33 +0000 (22:06 -0400)]
c++: constraints and explicit instantiation [PR96164]
When considering to instantiate a member of a class template as part of
an explicit instantiation of the class template, we need to first check
the member's constraints before proceeding with the instantiation of the
member.
gcc/cp/ChangeLog:
PR c++/96164
* constraint.cc (constraints_satisfied_p): Return true if
!flags_concepts.
* pt.c (do_type_instantiation): Update a paragraph taken from
[temp.explicit] to reflect the latest specification. Don't
instantiate a member with unsatisfied constraints.
gcc/testsuite/ChangeLog:
PR c++/96164
* g++.dg/cpp2a/concepts-explicit-inst5.C: New test.
Roger Sayle [Tue, 28 Jul 2020 13:55:47 +0000 (15:55 +0200)]
nvptx: Support floating point reciprocal instructions
The following patch addds support for PTX's rcp.rn.f32 and rcp.rn.f64
instructions. Note that the "rcp.rn" forms of this instruction
calculate the fully IEEE compliant result for the reciprocal, unlike
the rcp.approx variants that just provide fast approximations.
This patch has been tested on nvptx-none hosted on x86_64-pc-linux-gnu
with "make" and "make check" with no new regressions.
2020-07-12 Roger Sayle <roger@nextmovesoftware.com>
gcc/ChangeLog:
* config/nvptx/nvptx.md (recip<mode>2): New instruction.
_Local_iterator_base inherits _Node_iterator_base and so share the same
comparison operators. It avoids to expose special method _M_curr for debug
mode to compare such iterators.
P1825R0 extends the C++11 implicit move on return by removing the
constraints on the called constructor: previously, it needed to take an
rvalue reference to the type of the returned variable. The paper also
allows move on throw of parameters and implicit move of rvalue references.
Discussion on the CWG reflector about how to avoid breaking the PR91212 test
in the new model settled on the model of doing only a single overload
resolution, with the variable treated as an xvalue that can bind to
non-const lvalue references. So this patch implements that approach. The
implementation does not use the existing LOOKUP_PREFER_RVALUE flag, but
instead sets a flag on the representation of the static_cast turning the
variable into an xvalue.
For the time being I'm limiting the new semantics to C++20 mode; since it
was moved as a DR, we will probably want to apply the change to other
standard modes as well once we have a better sense of the impact on existing
code, probably in GCC 12.
Jason Merrill [Wed, 29 Jul 2020 04:57:40 +0000 (00:57 -0400)]
c++: Avoid calling const copy ctor on implicit move. [PR91212]
Our implementation of C++11 implicit move was wrong for return; we didn't
actually hit the check for the type of the first parameter of the selected
constructor, because we didn't see LOOKUP_PREFER_RVALUE set properly.
Fixing that to look at the right flags fixed the issue for this testcase,
but broke implicit move for a by-value converting constructor (PR58051). I
think this was not allowed in C++17, but it is allowed under the implicit
move changes from C++20, and those changes were voted to apply as a DR to
earlier standards as well, so I don't want to break it now.
So after fixing the flags check I changed the test to allow value
parameters.
gcc/cp/ChangeLog:
PR c++/91212
* call.c (build_over_call): Don't call a const ref
overload for implicit move.
gcc/testsuite/ChangeLog:
PR c++/91212
* g++.dg/cpp0x/move-return3.C: New test.
preprocessor: Teach traditional about has_include [PR95889]
Traditional cpp (used by fortran) didn;t know about the new
__has_include__ implementation. Hey, since when did traditional cpp
grow __has_include__? That wasn't in knr!
atomic_float/value_init.cc requires libatomic for some targets, i.e., when
it tries to perform an atomic operation with a 64 bit floating point
double type on a 32 bit target. This patch adds AIX and Darwin to the
list of targets that require the libatomic option and adds the option to
the atomic_float/value_init.cc testcase.
libstdc++-v3/ChangeLog:
2020-07-28 David Edelsohn <dje.gcc@gmail.com>
Jonathan Wakely <jwakely@redhat.com>
Rainer Orth <ro@CeBiTec.Uni-Bielefeld.DE>
* testsuite/lib/dg-options.exp (add_options_for_libatomic): Add
target powerpc-ibm-aix* and powerpc*-*-darwin*.
* testsuite/29_atomics/atomic_float/value_init.cc: Add options
for libatomic.
Joe Ramsay [Wed, 29 Jul 2020 13:04:28 +0000 (14:04 +0100)]
arm: Enable no-writeback vldr.16/vstr.16.
There was previously no way to specify that a register operand cannot
have any writeback modifiers, and as a result the argument to vldr.16
and vstr.16 could be erroneously output with post-increment. This
change adds a constraint which forbids all writeback, and
selects it in the relevant case for vldr.16 and vstr.16
Bootstrapped on arm-linux, gcc and CMSIS-DSP testsuites are clean.
Is this patch OK for trunk? If yes, please commit on my behalf as I
don't
have commit rights.
gcc/ChangeLog:
* config/arm/arm-protos.h (arm_coproc_mem_operand_no_writeback):
Declare prototype.
(arm_mve_mode_and_operands_type_check): Declare prototype.
* config/arm/arm.c (arm_coproc_mem_operand): Refactor to use
_arm_coproc_mem_operand.
(arm_coproc_mem_operand_wb): New function to cover full, limited
and no writeback.
(arm_coproc_mem_operand_no_writeback): New constraint for memory
operand with no writeback.
(arm_print_operand): Extend 'E' specifier for memory operand
that does not support writeback.
(arm_mve_mode_and_operands_type_check): New constraint check for
MVE memory operands.
* config/arm/constraints.md: Add Uj constraint for VFP vldr.16
and vstr.16.
* config/arm/vfp.md (*mov_load_vfp_hf16): New pattern for
vldr.16.
(*mov_store_vfp_hf16): New pattern for vstr.16.
(*mov<mode>_vfp_<mode>16): Remove MVE moves.
Richard Biener [Tue, 28 Jul 2020 07:45:52 +0000 (09:45 +0200)]
tree-optimization/96349 - avoid abnormal coalescing issues in loop split
This avoids splitting a loop when the entry value of a loop PHI is
involved with abnormal coalescing.
2020-07-28 Richard Biener <rguenther@suse.de>
PR tree-optimization/96349
* tree-ssa-loop-split.c (stmt_semi_invariant_p_1): When the
condition runs into a loop PHI with an abnormal entry value give up.
Richard Biener [Wed, 29 Jul 2020 10:30:59 +0000 (12:30 +0200)]
more SCEV cache clearing
This fixes two more places, in loop interchange and in the
vectorizer where the SCEV verifier sees stale entries.
2020-07-29 Richard Biener <rguenther@suse.de>
* tree-vectorizer.c (vectorize_loops): Reset the SCEV
cache if we removed any SIMD UID SSA defs.
* gimple-loop-interchange.cc (pass_linterchange::execute):
Reset the scev cache if we interchanged a loop.
Richard Biener [Wed, 29 Jul 2020 07:59:01 +0000 (09:59 +0200)]
tree-optimization/95679 - properly signal changes from propagate_into_phi_args
This restores a lost setting of something_changed with the
recent refactoring of the substitute and fold engine. The
reported ICE in the PR was meanwhile mitigated in other ways
but the issue can still result in missed optimizations via
failed runs of CFG cleanup.
Mark Eggleston [Fri, 17 Jul 2020 13:22:48 +0000 (14:22 +0100)]
Fortran : ICE in gfc_conv_scalarized_array_ref PR53298
When an array of characters is an argument to a subroutine and
is accessed using (:)(1:) an ICE occurs. The upper bound of the
substring does not have an expression and such should not have
a Scalarization State structure added to the Scalarization State
chain.
2020-07-29 Mark Eggleston <markeggleston@gcc.gnu.org>
gcc/fortran/
PR fortran/53298
* trans-array.c (gfc_walk_array_ref): If ref->ss.end is set
call gfc_get_scalar_ss.
2020-07-29 Mark Eggleston <markeggleston@gcc.gnu.org>
gcc/testsuite/
PR fortran/53298
* gfortran.dg/pr53298.f90: New test.
_Hashtable move constructor is wrongly qualified as noexcept(true) regardless of
_Equal and _H1 copy constructor qualifications.
_Hashtable allocator-aware move constructor is missing its noexcept
qualification like the depending unordered containers ones.
libstdc++-v3/ChangeLog:
* include/bits/hashtable.h
(_Hashtable(_Hashtable&& __ht, __node_alloc_type&& __a, true_type)):
Add noexcept qualification.
(_Hashtable(_Hashtable&&)): Fix noexcept qualification.
(_Hashtable(_Hashtable&&, const allocator_type&)): Add noexcept
qualification.
* include/bits/unordered_map.h
(unordered_map(unordered_map&&, const allocator_type&)): Add noexcept
qualification.
(unordered_multimap(unordered_multimap&&, const allocator_type&)):
Likewise.
* include/bits/unordered_set.h
(unordered_set(unordered_set&&, const allocator_type&)): Likewise.
(unordered_multiset(unordered_multiset&&, const allocator_type&)):
Likewise.
* include/debug/unordered_map
(unordered_map(unordered_map&&, const allocator_type&)): Likewise.
(unordered_multimap(unordered_multimap&&, const allocator_type&)):
Likewise.
* include/debug/unordered_set
(unordered_set(unordered_set&&, const allocator_type&)): Likewise.
(unordered_multiset(unordered_multiset&&, const allocator_type&)):
Likewise.
* testsuite/23_containers/unordered_map/allocator/default_init.cc:
New test.
* testsuite/23_containers/unordered_map/cons/noexcept_default_construct.cc:
New test.
* testsuite/23_containers/unordered_map/cons/noexcept_move_construct.cc:
New test.
* testsuite/23_containers/unordered_map/modifiers/move_assign.cc:
New test.
* testsuite/23_containers/unordered_multimap/cons/noexcept_default_construct.cc:
New test.
* testsuite/23_containers/unordered_multimap/cons/noexcept_move_construct.cc:
New test.
* testsuite/23_containers/unordered_multiset/cons/noexcept_default_construct.cc:
New test.
* testsuite/23_containers/unordered_multiset/cons/noexcept_move_construct.cc:
New test.
* testsuite/23_containers/unordered_set/allocator/default_init.cc:
New test.
* testsuite/23_containers/unordered_set/cons/noexcept_default_construct.cc:
New test.
* testsuite/23_containers/unordered_set/cons/noexcept_move_construct.cc:
New test.
vect: Fix infinite loop while determining peeling amount
This is a follow up to commit 5c9669a0e6c respectively discussion
https://gcc.gnu.org/pipermail/gcc-patches/2020-June/549132.html
In case a type has a lower alignment than its size, ensure that we
advance. For example, on s390x we have for a long double an alignment
constraint of 8 bytes whereas the size is 16 bytes. Increasing the loop
variable only by TARGET_ALIGN / DR_SIZE which equals zero in case of a
long double results in an infinite loop. This is fixed by ensuring that
we advance at least by one.
gcc/ChangeLog:
* tree-vect-data-refs.c (vect_enhance_data_refs_alignment):
Ensure that loop variable npeel_tmp advances in each iteration.
* gfortran.dg/gomp/requires-1.f90: New test.
* gfortran.dg/gomp/requires-2.f90: New test.
* gfortran.dg/gomp/requires-3.f90: New test.
* gfortran.dg/gomp/requires-4.f90: New test.
* gfortran.dg/gomp/requires-5.f90: New test.
* gfortran.dg/gomp/requires-6.f90: New test.
* gfortran.dg/gomp/requires-7.f90: New test.
* gfortran.dg/gomp/requires-8.f90: New test.
* gfortran.dg/gomp/requires-9.f90: New test.
config/mmix/mmix.h (NO_FUNCTION_CSE): Define to 1.
The tests gcc.dg/tree-ssa/loop-1.c and gcc.dg/weak/typeof-2.c
assume this setting and are as a consequence riddled with
exceptions for targets that actually do yield better code when
calling through a register rather than repeatedly the same
symbol. Nonetheless, defining it makes sense for MMIX.
(Even better IMHO, this macro shouldn't exist and instead rtx
costs be used or perhaps just a target-specific default of
flag_no_function_cse.)
gcc:
* config/mmix/mmix.h (NO_FUNCTION_CSE): Define to 1.
mmix.h (ASM_OUTPUT_EXTERNAL): Define to default_elf_asm_output_external.
Whoops. When un-disabling visibility support for mmix, I missed that
some of the newly enabled tests were FAILs, for not emitting .hidden
for references to external declarations. This takes care of
gcc.dg/visibility-14.c .. -19.c, and gcc.dg/visibility-23.c.
gcc:
* config/mmix/mmix.h (ASM_OUTPUT_EXTERNAL): Define to
default_elf_asm_output_external.
ipa/96291: don't crash on unoptimized lto functions
In PR ipa/96291 the test contained an SCC with one
unoptimized function. This tricked ipa-cp into NULL dereference.
has_undead_caller_from_outside_scc_p() did not take into account
that unoptimized funtions don't have IPA summary analysis. And
dereferenced NULL pointer causing an ICE.
Roger Sayle [Tue, 28 Jul 2020 21:55:12 +0000 (22:55 +0100)]
middle-end: Parity and popcount folding optimizations.
This patch implements several constant folding optimizations
for __builtin_parity and friends. We canonicalize popcount(x)&1
as parity(x) in gimple, and potentially convert back again when
we expand to RTL. parity(~x) is simplified to parity(x), which
is true for all integer modes with an even number of bits.
But probably most usefully, parity(x)^parity(y) can be simplified
to a parity(x^y), requiring only a single libcall or popcount.
This patch optimizes popcount and parity of an argument known to have
at most a single bit set, to be that single bit. Hence, popcount(x&8)
is simplified to (x>>3)&1. This generalizes the existing optimization
of popcount(x&1) being simplified to x&1, which is cleaned up with
this patch.
2020-07-28 Roger Sayle <roger@nextmovesoftware.com>
Richard Biener <rguenther@suse.de>
gcc/ChangeLog
* match.pd (popcount(x)&1 -> parity(x)): New simplification.
(parity(~x) -> parity(x)): New simplification.
(parity(x)^parity(y) -> parity(x^y)): New simplification.
(parity(x&1) -> x&1): New simplification.
(popcount(x) -> x>>C): New simplification.
Jakub Jelinek [Tue, 28 Jul 2020 13:41:30 +0000 (15:41 +0200)]
c++: Fix up cp_lexer_safe_previous_token [PR96328]
The following testcase ICEs, because cp_lexer_safe_previous_token calls
cp_lexer_previous_token and that ICEs, because all tokens in the lexer
buffer before the current one (CPP_EOF) have been purged.
cp_lexer_safe_previous_token is used in the context where it is ok if it
punts, so the patch changes the function so that it doesn't assert there is
some previous token, but instead returns NULL like in other cases where it
punts.
In addition to this, in the last hunk it does a micro-optimization, don't
call the potentially expensive function if it will not need the result,
instead check the least expensive condition first.
And the middle hunk is a similar change from Mark's version of the patch,
to use the safe variant in there because it is again just about a hint
and it is better not to provide the hint than to ICE, though we don't have a
testcase that would ICE.
2020-07-28 Jakub Jelinek <jakub@redhat.com>
Mark Wielaard <mark@klomp.org>
PR c++/96328
* parser.c (cp_lexer_safe_previous_token): Don't call
cp_lexer_previous_token, instead inline it by hand and return NULL
instead of failing assertion if all previous tokens until the first
one are purged.
(cp_parser_error_1): Optimize - only call cp_lexer_safe_previous_token
if token->type is CPP_NAME. Use cp_lexer_safe_previous_token instead
of cp_lexer_previous_token for the missing_token_desc != RT_NONE
case too.
Jakub Jelinek [Tue, 28 Jul 2020 13:40:15 +0000 (15:40 +0200)]
libcpp: Fix up raw string literal parsing error-recovery [PR96323]
For (invalid) newline inside of the raw string literal delimiter, doing
continue means we skip the needed processing of newlines. Instead of
duplicating that, this patch just doesn't continue for those.
2020-07-28 Jakub Jelinek <jakub@redhat.com>
PR preprocessor/96323
* lex.c (lex_raw_string): For c == '\n' don't continue after reporting
an prefix delimiter error.
François Dumont [Mon, 20 Jan 2020 18:01:18 +0000 (19:01 +0100)]
libstdc++: Do not over-size hashtable buckets on range insertion
We used to consider range size on insertion but on unique keys container
not all range values might be inserted resulting in over-sizing. In this
case we just consider user reservation and if none then the container will
adapt to actually inserted elements.
libstdc++-v3/ChangeLog:
* include/bits/hashtable.h
(_Hashtable<>(_InputIterator, _InputIterator, size_t, const _H1&,
const _H2&, const _Hash&, const _Equal&, const _ExtractKey&,
const allocator_type&, true_type)): New.
(_Hashtable<>(_InputIterator, _InputIterator, size_t, const _H1&,
const _H2&, const _Hash&, const _Equal&, const _ExtractKey&,
const allocator_type&, false_type)): New.
(_Hashtable<>(_InputIterator, _InputIterator, size_t, const _H1&,
const _H2&, const _Hash&, const _Equal&, const _ExtractKey&,
const allocator_type&)): Delegate to latters.
(operator=(initializer_list<value_type>)): Rehash if too small.
(_M_insert(_Arg&&, const _NodeGenerator&, true_type)): Remove
size_t len parameter.
* include/bits/hashtable_policy.h (_Insert_base<>::_M_insert_range):
Do not try to get input range distance.
* testsuite/23_containers/unordered_set/cons/bucket_hint.cc: New test.
* testsuite/23_containers/unordered_set/modifiers/insert.cc: New test.
Jakub Jelinek [Tue, 28 Jul 2020 09:08:29 +0000 (11:08 +0200)]
expander: Fix ICE in maybe_warn_rdwr_sizes [PR96335]
The following testcase ICEs in maybe_warn_rdwr_sizes. The problem is that
the caller uses its fndecl and fntype variables to fill up rdwr_map, and
the fntype in that case is a prototype with the access attribute and all
the checks needed for that performed. But the maybe_warn_rdwr_sizes
function tries to rediscover fndecl/fntype itself and does it differently
from how the caller did (for fndecl get_callee_fndecl and fntype from that
FUNCTION_DECL, otherwise sets fntype to CALL_EXPR_FN's type).
On the testcase, get_callee_fndecl does find a FUNCTION_DECL because
it does STRIP_NOPS in between.
Instead of trying to rediscover those, this patch just passes them down,
like is done in several other functions.
2020-07-28 Jakub Jelinek <jakub@redhat.com>
PR middle-end/96335
* calls.c (maybe_warn_rdwr_sizes): Add FNDECL and FNTYPE arguments,
instead of trying to rediscover them in the body.
(initialize_argument_information): Adjust caller.
This patch is to refactor the existing peel_iters_prologue and
peel_iters_epilogue cost model handlings, by following the structure
below suggested by Richard Sandiford:
- calculate peel_iters_prologue
- calculate peel_iters_epilogue
- add costs associated with peel_iters_prologue
- add costs associated with peel_iters_epilogue
- add costs related to branch taken/not_taken.
Bootstrapped/regtested on aarch64-linux-gnu.
gcc/ChangeLog:
* tree-vect-loop.c (vect_get_known_peeling_cost): Factor out some code
to determine peel_iters_epilogue to...
(vect_get_peel_iters_epilogue): ...this new function.
(vect_estimate_min_profitable_iters): Refactor cost calculation on
peel_iters_prologue and peel_iters_epilogue.
git://gcc.gnu.org / gcc.git / log