When registering relations in the oracle, search for other relations which
imply new transitive relations.
gcc/
* value-relation.cc (rr_transitive_table): New.
(relation_transitive): New.
(value_relation::swap): Remove.
(value_relation::apply_transitive): New.
(relation_oracle::relation_oracle): Allocate a new tmp bitmap.
(relation_oracle::register_relation): Call register_transitives.
(relation_oracle::register_transitives): New.
* value-relation.h (relation_oracle): Add new temporary bitmap and
methods.
gcc/testsuite/
* gcc.dg/predict-1.c: Disable evrp.
* gcc.dg/tree-ssa/evrp-trans.c: New.
Clang warns about this, but GCC doesn't (see PR c++/102036).
Signed-off-by: Jonathan Wakely <jwakely@redhat.com>
libstdc++-v3/ChangeLog:
* src/c++11/cxx11-shim_facets.cc: Fix mismatched class-key in
explicit instantiation definitions.
Broadcast from integer to a pseudo vector register instead of a hard
vector register to allow LRA to remove redundant move instruction after
broadcast.
gcc/
PR target/102021
* config/i386/i386-expand.c (ix86_expand_vector_move): Broadcast
from integer to a pseudo vector register.
gcc/testsuite/
PR target/102021
* gcc.target/i386/pr100865-10b.c: Expect vzeroupper.
* gcc.target/i386/pr100865-4b.c: Likewise.
* gcc.target/i386/pr100865-6b.c: Expect vmovdqu and vzeroupper.
* gcc.target/i386/pr100865-7b.c: Likewise.
* gcc.target/i386/pr102021.c: New test.
This avoids leaving scalar if-converted code around for the case
of BB vectorizing an if-converted loop body when using the very-cheap
cost model. In this case we scan not vectorized scalar stmts in
the basic-block vectorized for COND_EXPRs and force the vectorization
to be marked as not profitable.
The patch also makes sure to always consider all BB vectorization
subgraphs together for costing purposes when vectorizing an
if-converted loop body.
2021-08-24 Richard Biener <rguenther@suse.de>
PR tree-optimization/100089
* tree-vectorizer.h (vect_slp_bb): Rename to ...
(vect_slp_if_converted_bb): ... this and get the original
loop as new argument.
* tree-vectorizer.c (try_vectorize_loop_1): Revert previous fix,
pass original loop to vect_slp_if_converted_bb.
* tree-vect-slp.c (vect_bb_vectorization_profitable_p):
If orig_loop was passed scan the not vectorized stmts
for COND_EXPRs and force not profitable if found.
(vect_slp_region): Pass down all SLP instances to costing
if orig_loop was specified.
(vect_slp_bbs): Pass through orig_loop.
(vect_slp_bb): Rename to ...
(vect_slp_if_converted_bb): ... this and get the original
loop as new argument.
(vect_slp_function): Adjust.
For Armv8.1-m we generate code that emits VLLDM directly and do not
rely on support code in the library, so emit the mitigation directly
as well, when required. In this case, we can use the compiler options
to determine when to apply the fix and when it is safe to omit it.
gcc:
PR target/102035
* config/arm/arm.md (attribute arch): Add fix_vlldm.
(arch_enabled): Use it.
* config/arm/vfp.md (lazy_store_multiple_insn): Add alternative to
use when erratum mitigation is needed.
Add the recommended erratum mitigation sequence to
__gnu_cmse_nonsecure_call for use on Armv8-m.main devices. Since this
is in the library code we cannot know in advance whether the core we
are running on will be affected by this, so always enable it.
libgcc:
PR target/102035
* config/arm/cmse_nonsecure_call.S (__gnu_cmse_nonsecure_call):
Add vlldm erratum work-around.
Add a new option, -mfix-cmse-cve-2021-35465 and document it. Enable it
automatically for cortex-m33, cortex-m35p and cortex-m55.
gcc:
PR target/102035
* config/arm/arm.opt (mfix-cmse-cve-2021-35465): New option.
* doc/invoke.texi (Arm Options): Document it.
* config/arm/arm-cpus.in (quirk_vlldm): New feature bit.
(ALL_QUIRKS): Add quirk_vlldm.
(cortex-m33): Add quirk_vlldm.
(cortex-m35p, cortex-m55): Likewise.
* config/arm/arm.c (arm_option_override): Enable fix_vlldm if
targetting an affected CPU and not explicitly controlled on
the command line.
The test for CMSE support being available in hardware currently
relies on the compiler not optimizing away a secure gateway operation.
But even that is suspect, because the SG instruction is just a NOP
on armv8-m implementations that do not support the security extension.
Replace the existing test with a new one that reads and checks
the appropriate hardware feature register (memory mapped). This has
to be run from secure mode, but that shouldn't matter, because if we
can't do that we can't really test the CMSE extensions anyway. We
retain the SG instruction to ensure the test can't pass accidentally
if run on pre-armv8-m devices.
gcc/testsuite:
* lib/target-supports.exp (check_effective_target_arm_cmse_hw):
Check the CMSE feature register, rather than relying on the
SG operation causing an execution fault.
Both lazy_store_multiple_insn and lazy_load_multiple_insn contain
invalid RTL (eg they contain a post_inc statement outside of a mem).
What's more, the instructions concerned do not modify their input
address register. We probably got away with this because they are
generated so late in the compilation that no subsequent pass needed to
understand them. Nevertheless, this could cause problems someday, so
fixed to use a simple legal unspec.
gcc:
* config/arm/vfp.md (lazy_store_multiple_insn): Rewrite as valid RTL.
(lazy_load_multiple_insn): Likewise.
Also optimize below 3 forms to vpternlog, op1, op2, op3 are
register_operand or unary_p as (not reg)
A: (any_logic (any_logic op1 op2) op3)
B: (any_logic (any_logic op1 op2) (any_logic op3 op4)) op3/op4 should
be equal to op1/op2
C: (any_logic (any_logic (any_logic:op1 op2) op3) op4) op3/op4 should
be equal to op1/op2
gcc/ChangeLog:
PR target/101989
* config/i386/i386.c (ix86_rtx_costs): Define cost for
UNSPEC_VTERNLOG.
* config/i386/i386.h (STRIP_UNARY): New macro.
* config/i386/predicates.md (reg_or_notreg_operand): New
predicate.
* config/i386/sse.md (*<avx512>_vternlog<mode>_all): New define_insn.
(*<avx512>_vternlog<mode>_1): New pre_reload
define_insn_and_split.
(*<avx512>_vternlog<mode>_2): Ditto.
(*<avx512>_vternlog<mode>_3): Ditto.
(any_logic1,any_logic2): New code iterator.
(logic_op): New code attribute.
(ternlogsuffix): Extend to VNxDF and VNxSF.
gcc/testsuite/ChangeLog:
PR target/101989
* gcc.target/i386/pr101989-1.c: New test.
* gcc.target/i386/pr101989-2.c: New test.
* gcc.target/i386/avx512bw-shiftqihi-constant-1.c: Adjust testcase.
This makes use of the estimated number of iterations of the inner loop
to limit --param vect-inner-loop-cost-factor scaling. It also reduces
the maximum value of vect-inner-loop-cost-factor to 10000 making it
less likely to cause overflow of costs.
2021-08-23 Richard Biener <rguenther@suse.de>
* doc/invoke.texi (vect-inner-loop-cost-factor): Adjust.
* params.opt (--param vect-inner-loop-cost-factor): Adjust
maximum value.
* tree-vect-loop.c (vect_analyze_loop_form): Initialize
inner_loop_cost_factor to the minimum of the estimated number
of iterations of the inner loop and vect-inner-loop-cost-factor.
There are a few problems with download_prerequisites are
described in PR 82704. The first is on busy-box version of
shasum and md5sum the extended option --check don't exist
so just use -c. The second issue is the code for which
shasum program to use is included twice and is different.
So move which program to use for the checksum after argument
parsing. The last issue is --md5 option has been broken for
sometime now as the program is named md5sum and not just md5.
Nobody updated switch table to be correct.
contrib/ChangeLog:
PR other/82704
* download_prerequisites: Fix issues with --md5 and
--sha512 options.
Back in June I briefly mentioned in one of my gcc-patches posts that
a change that should have always reduced code size, would mysteriously
occasionally result in slightly larger code (according to CSiBE):
https://gcc.gnu.org/pipermail/gcc-patches/2021-June/573233.html
Investigating further, the cause turns out to be that x86_64's
scalar-to-vector (stv) pass is relying on poor estimates of the size
costs/benefits. This patch tweaks the backend's compute_convert_gain
method to provide slightly more accurate values when compiling with
-Os. Compilation without -Os is (should be) unaffected. And for
completeness, I'll mention that the stv pass is a net win for code
size so it's much better to improve its heuristics than simply gate
the pass on !optimize_for_size.
The net effect of this change is to save 1399 bytes on the CSiBE
code size benchmark when compiling with -Os.
2021-08-24 Roger Sayle <roger@nextmovesoftware.com>
Richard Biener <rguenther@suse.de>
gcc/ChangeLog
* config/i386/i386-features.c (compute_convert_gain): Provide
more accurate values for CONST_INT, when optimizing for size.
* config/i386/i386.c (COSTS_N_BYTES): Move definition from here...
* config/i386/i386.h (COSTS_N_BYTES): to here.
My sincere apologies to everyone (again). As diagnosed by
Jakub Jelinek, my recent patch to fold the signedness of LSHIFT_EXPR
needs to be careful not to attempt transforming a left shift in an
integer type into an invalid left shift of a pointer type.
2021-08-24 Roger Sayle <roger@nextmovesoftware.com>
Jakub Jelinek <jakub@redhat.com>
gcc/ChangeLog
PR middle-end/102029
* match.pd (shift transformations): Add an additional check for
!POINTER_TYPE_P in the recently added left shift transformation.
gcc/testsuite/ChangeLog
PR middle-end/102029
* gcc.dg/fold-convlshift-3.c: New test case.
When investigating false positives on the Linux kernel from
-Wanalyzer-use-of-uninitialized-value, I noticed that the existing
implementation of switch statements in the analyzer is broken.
Specifically, the existing implementation assumes a 1:1 association
between CFG out-edges from the basic block and case labels in the
gimple switch statement. This happened to be the case in the
examples I had tested, but there is no such association in general.
In particular, in the motivating example:
arch/x86/kernel/cpu/mtrr/if.c: mtrr_ioctl
the switch statement has 3 blocks, each covering multiple ranges of
ioctl command IDs for which different local variables are initialized,
which the existing implementation gets badly wrong. [1]
This patch reimplements switch handling in the analyzer to eliminate
this false assumption - instead, for each out-edge we gather the set
of case labels for that out-edge, and use that to determine the
set of value ranges for the edge. Avoiding false positives for the
above example requires that we accurately track value ranges for
symbolic values, so the patch extends constraint_manager with a new
bounded_ranges_constraint, adding just enough information to capture the
ranges for switch statements whilst retaining combatility with the
existing constraint-handling (ultimately I'd prefer to simply throw
all of this into a SAT solver and let it track things).
Doing so fixes the false positives seen on the Linux kernel and an
existing xfail in the test suite.
The patch also fixes a long-standing bug in
constraint_manager::add_unknown_constraint when updating constraints
due to combining equivalence classes, spotted when debugging the
same logic for the new kind of constraints.
[1] a reduced version of this code is captured in this patch, in
gcc.dg/analyzer/torture/switch-3.c
gcc/analyzer/ChangeLog:
* analyzer.h (struct rejected_constraint): Convert to...
(class rejected_constraint): ...this.
(class bounded_ranges): New forward decl.
(class bounded_ranges_manager): New forward decl.
* constraint-manager.cc: Include "analyzer/analyzer-logging.h" and
"tree-pretty-print.h".
(can_plus_one_p): New.
(plus_one): New.
(can_minus_one_p): New.
(minus_one): New.
(bounded_range::bounded_range): New.
(dump_cst): New.
(bounded_range::dump_to_pp): New.
(bounded_range::dump): New.
(bounded_range::to_json): New.
(bounded_range::set_json_attr): New.
(bounded_range::contains_p): New.
(bounded_range::intersects_p): New.
(bounded_range::operator==): New.
(bounded_range::cmp): New.
(bounded_ranges::bounded_ranges): New.
(bounded_ranges::bounded_ranges): New.
(bounded_ranges::bounded_ranges): New.
(bounded_ranges::canonicalize): New.
(bounded_ranges::validate): New.
(bounded_ranges::operator==): New.
(bounded_ranges::dump_to_pp): New.
(bounded_ranges::dump): New.
(bounded_ranges::to_json): New.
(bounded_ranges::eval_condition): New.
(bounded_ranges::contain_p): New.
(bounded_ranges::cmp): New.
(bounded_ranges_manager::~bounded_ranges_manager): New.
(bounded_ranges_manager::get_or_create_empty): New.
(bounded_ranges_manager::get_or_create_point): New.
(bounded_ranges_manager::get_or_create_range): New.
(bounded_ranges_manager::get_or_create_union): New.
(bounded_ranges_manager::get_or_create_intersection): New.
(bounded_ranges_manager::get_or_create_inverse): New.
(bounded_ranges_manager::consolidate): New.
(bounded_ranges_manager::get_or_create_ranges_for_switch): New.
(bounded_ranges_manager::create_ranges_for_switch): New.
(bounded_ranges_manager::make_case_label_ranges): New.
(bounded_ranges_manager::log_stats): New.
(bounded_ranges_constraint::print): New.
(bounded_ranges_constraint::to_json): New.
(bounded_ranges_constraint::operator==): New.
(bounded_ranges_constraint::add_to_hash): New.
(constraint_manager::constraint_manager): Update for new field
m_bounded_ranges_constraints.
(constraint_manager::operator=): Likewise.
(constraint_manager::hash): Likewise.
(constraint_manager::operator==): Likewise.
(constraint_manager::print): Likewise.
(constraint_manager::dump_to_pp): Likewise.
(constraint_manager::to_json): Likewise.
(constraint_manager::add_unknown_constraint): Update the lhs_ec_id
if necessary in existing constraints when combining equivalence
classes. Add similar code for handling
m_bounded_ranges_constraints.
(constraint_manager::add_constraint_internal): Add comment.
(constraint_manager::add_bounded_ranges): New.
(constraint_manager::eval_condition): Use new field
m_bounded_ranges_constraints.
(constraint_manager::purge): Update bounded_ranges_constraint
instances.
(constraint_manager::canonicalize): Update for new field.
(merger_fact_visitor::on_ranges): New.
(constraint_manager::for_each_fact): Use new field
m_bounded_ranges_constraints.
(constraint_manager::validate): Fix off-by-one error needed due
to bug fixed above in add_unknown_constraint. Validate the EC IDs
in m_bounded_ranges_constraints.
(constraint_manager::get_range_manager): New.
(selftest::assert_dump_bounded_range_eq): New.
(ASSERT_DUMP_BOUNDED_RANGE_EQ): New.
(selftest::test_bounded_range): New.
(selftest::assert_dump_bounded_ranges_eq): New.
(ASSERT_DUMP_BOUNDED_RANGES_EQ): New.
(selftest::test_bounded_ranges): New.
(selftest::run_constraint_manager_tests): Call the new selftests.
* constraint-manager.h (struct bounded_range): New.
(struct bounded_ranges): New.
(template <> struct default_hash_traits<bounded_ranges::key_t>): New.
(class bounded_ranges_manager): New.
(fact_visitor::on_ranges): New pure virtual function.
(class bounded_ranges_constraint): New.
(constraint_manager::add_bounded_ranges): New decl.
(constraint_manager::get_range_manager): New decl.
(constraint_manager::m_bounded_ranges_constraints): New field.
* diagnostic-manager.cc (epath_finder::process_worklist_item):
Transfer ownership of rc to add_feasibility_problem.
* engine.cc (feasibility_problem::dump_to_pp): Use get_model.
* feasible-graph.cc (infeasible_node::dump_dot): Update for
conversion of m_rc to a pointer.
(feasible_graph::add_feasibility_problem): Pass RC by pointer and
take ownership.
* feasible-graph.h (infeasible_node::infeasible_node): Pass RC by
pointer and take ownership.
(infeasible_node::~infeasible_node): New.
(infeasible_node::m_rc): Convert to a pointer.
(feasible_graph::add_feasibility_problem): Pass RC by pointer and
take ownership.
* region-model-manager.cc: Include
"analyzer/constraint-manager.h".
(region_model_manager::region_model_manager): Initializer new
field m_range_mgr.
(region_model_manager::~region_model_manager): Delete it.
(region_model_manager::log_stats): Call log_stats on it.
* region-model.cc (region_model::add_constraint): Use new subclass
rejected_op_constraint.
(region_model::apply_constraints_for_gswitch): Reimplement using
bounded_ranges_manager.
(rejected_constraint::dump_to_pp): Convert to...
(rejected_op_constraint::dump_to_pp): ...this.
(rejected_ranges_constraint::dump_to_pp): New.
* region-model.h (struct purge_stats): Add field
m_num_bounded_ranges_constraints.
(region_model_manager::get_range_manager): New.
(region_model_manager::m_range_mgr): New.
(region_model::get_range_manager): New.
(struct rejected_constraint): Split into...
(class rejected_constraint):...this new abstract base class,
and...
(class rejected_op_constraint): ...this new concrete subclass.
(class rejected_ranges_constraint): New.
* supergraph.cc: Include "tree-cfg.h".
(supergraph::supergraph): Drop idx param from add_cfg_edge.
(supergraph::add_cfg_edge): Drop idx param.
(switch_cfg_superedge::switch_cfg_superedge): Move here from
header. Populate m_case_labels with all cases which go to DST.
(switch_cfg_superedge::dump_label_to_pp): Reimplement to use
m_case_labels.
(switch_cfg_superedge::get_case_label): Delete.
* supergraph.h (supergraphadd_cfg_edge): Drop "idx" param.
(switch_cfg_superedge::switch_cfg_superedge): Drop idx param and
move implementation to supergraph.cc.
(switch_cfg_superedge::get_case_label): Delete.
(switch_cfg_superedge::get_case_labels): New.
(switch_cfg_superedge::m_idx): Delete.
(switch_cfg_superedge::m_case_labels): New field.
gcc/testsuite/ChangeLog:
* gcc.dg/analyzer/switch.c: Remove xfail. Add various tests.
* gcc.dg/analyzer/torture/switch-2.c: New test.
* gcc.dg/analyzer/torture/switch-3.c: New test.
* gcc.dg/analyzer/torture/switch-4.c: New test.
* gcc.dg/analyzer/torture/switch-5.c: New test.
2021-08-23 Bill Schmidt <wschmidt@linux.ibm.com>
gcc/
* config/rs6000/rs6000-gen-builtins.c (parse_bif_entry): Don't call
asprintf, which is not available on AIX.
gcc/analyzer/ChangeLog:
PR analyzer/101837
* analyzer.cc (maybe_reconstruct_from_def_stmt): Bail if fn is
NULL, and assert that it's non-NULL before passing it to
build_call_array_loc.
gcc/testsuite/ChangeLog:
PR analyzer/101837
* gcc.dg/analyzer/pr101837.c: New test.
gcc/analyzer/ChangeLog:
PR analyzer/101962
* region-model.cc (region_model::eval_condition_without_cm):
Refactor comparison against zero, adding a check for
POINTER_PLUS_EXPR of non-NULL.
gcc/testsuite/ChangeLog:
PR analyzer/101962
* gcc.dg/analyzer/data-model-23.c: New test.
* gcc.dg/analyzer/pr101962.c: New test.
gcc/analyzer/ChangeLog:
* store.cc (bit_range::intersects_p): New overload.
(bit_range::operator-): New.
(binding_cluster::maybe_get_compound_binding): Handle the partial
overlap case.
(selftest::test_bit_range_intersects_p): Add test coverage for
new overload of bit_range::intersects_p.
* store.h (bit_range::intersects_p): New overload.
(bit_range::operator-): New.
gcc/testsuite/ChangeLog:
* gcc.dg/analyzer/data-model-22.c: New test.
* gcc.dg/analyzer/uninit-6.c: New test.
* gcc.dg/analyzer/uninit-6b.c: New test.
r12-3005-g220c410162ebece4f missed a cast for the set_32 call.
Fixed thus.
Signed-off-by: Iain Sandoe <iain@sandoe.co.uk>
libiberty/ChangeLog:
* simple-object-mach-o.c (simple_object_mach_o_write_segment):
Cast the first argument to set_32 as needed.
In PR101296 Richard noticed that modref is giving up on analysis in milc by
hitting --param=modref-max-accesses limit. While cleaning up original modref
patch I removed code that tried to do smart things while merging accesses
because it had bugs and wanted to reimplement it later which I later forgot.
This patch adds logic that avoids adding access and its subaccess to the list
which is just waste of memory and compile time. Incrementally I will add logic
merging the ranges.
gcc/ChangeLog:
2021-08-23 Jan Hubicka <hubicka@ucw.cz>
* ipa-modref-tree.h (modref_access_node::range_info_useful_p):
Improve range compare.
(modref_access_node::contains): New member function.
(modref_access_node::search): Remove.
(modref_access_node::insert): Be smarter about subaccesses.
gcc/testsuite/ChangeLog:
2021-08-23 Jan Hubicka <hubicka@ucw.cz>
* gcc.dg/tree-ssa/modref-7.c: New test.
Intel MIC (emulated) offloading execution failure remains to be analyzed.
libgomp/
* testsuite/libgomp.c/address-space-1.c: New file.
Co-authored-by: Jakub Jelinek <jakub@redhat.com>
..., and use that to improve XFAILing for Intel MIC offloading execution
instead of compilation in 'libgomp.c-c++-common/target-45.c',
'libgomp.fortran/target10.f90'.
gcc/
* config/i386/i386-options.c (ix86_omp_device_kind_arch_isa)
<omp_device_arch> [ACCEL_COMPILER]: Match "intel_mic".
* config/i386/t-omp-device (omp-device-properties-i386) <arch>:
Add "intel_mic".
libgomp/
* testsuite/lib/libgomp.exp
(check_effective_target_offload_target_intelmic): Remove 'proc'.
(check_effective_target_offload_device_intel_mic): New 'proc'.
* testsuite/libgomp.c-c++-common/on_device_arch.h
(device_arch_intel_mic, on_device_arch_intel_mic): New.
* testsuite/libgomp.c-c++-common/target-45.c: Use that for
'dg-xfail-run-if'.
* testsuite/libgomp.fortran/target10.f90: Likewise.
The standard shows this default template argument in the <ranges>
synopsis, but it was missing in libstdc++.
libstdc++-v3/ChangeLog:
* include/std/ranges (basic_istream_view): Add default template
argument.
* testsuite/std/ranges/istream_view.cc: Check it.
My sincere apologies to everyone, but especially Andrew Pinski
who warned me in advance that TRULY_NOOP_TRUNCATION results in
different code paths/optimizations on some targets. This restores
the build on nvptx-none (and presumably others) where mysteriously
(truncate:QI (reg:QI)) fails to be simplified to (reg:QI), which
is expected (everywhere) in my recently added self-tests.
2021-08-23 Roger Sayle <roger@nextmovesoftware.com>
gcc/ChangeLog
* simplify-rtx.c (simplify_unary_operation_1): [TRUNCATE]:
Handle case where the operand is already the desired mode.
Looks like the existing check using has_gimple_body_p isn't enough
at LTRANS time but I need to check in_other_partition as well.
2021-08-23 Richard Biener <rguenther@suse.de>
PR ipa/97565
* tree-ssa-structalias.c (ipa_pta_execute): Check in_other_partition
in addition to has_gimple_body.
* g++.dg/lto/pr97565_0.C: New testcase.
* g++.dg/lto/pr97565_1.C: Likewise.
The null pointer check is never needed for correct code, only to
gracefully handle undefined cases. Add __builtin_expect to be sure that
we don't pessimize the valid uses.
libstdc++-v3/ChangeLog:
* libsupc++/dyncast.cc (__dynamic_cast): Add __builtin_expect to
precondition check.
This function should be inline, so that's it's not emitted in tests that
don't use it, to avoid undefined references to geteuid().
libstdc++-v3/ChangeLog:
PR libstdc++/90787
* testsuite/util/testsuite_fs.h (permissions_are_testable):
Define as inline.
This patch adds support for the 'strict' modifier on grainsize/num_tasks
clauses, an OpenMP 5.1 feature supported in C/C++ since commit
r12-3066-g3bc75533d1f87f0617be6c1af98804f9127ec637
gcc/fortran/ChangeLog:
* dump-parse-tree.c (show_omp_clauses): Handle 'strict' modifier
on grainsize/num_tasks
* gfortran.h (gfc_omp_clauses): Add grainsize_strict
and num_tasks_strict.
* trans-openmp.c (gfc_trans_omp_clauses, gfc_split_omp_clauses):
Handle 'strict' modifier on grainsize/num_tasks.
* openmp.c (gfc_match_omp_clauses): Likewise.
libgomp/ChangeLog:
* testsuite/libgomp.fortran/taskloop-4-a.f90: New test.
* testsuite/libgomp.fortran/taskloop-4.f90: New test.
* testsuite/libgomp.fortran/taskloop-5-a.f90: New test.
* testsuite/libgomp.fortran/taskloop-5.f90: New test.
This fixes double-scaling of the inner loop scalar cost caused
by routing the scalar costs through the add_stmt_cost hook and
using vect_body as the location. The issue makes almost every
outer loop vectorization profitable.
2021-08-23 Richard Biener <rguenther@suse.de>
* tree-vect-loop.c (vect_compute_single_scalar_iteration_cost):
Properly scale the inner loop cost only once.
This patch implements support for TRUNC_MOD_EXPR and TRUNC_DIV_EXPR
in tree-ssa's bit CCP pass. This is mostly for completeness, as the
VRP pass already provides better bounds for these operations, but
seeing mask values of all_ones in my debugging/instrumentation logs
seemed overly pessimistic. With this patch, the expression X%10
has a nonzero bits of 0x0f (for unsigned X), likewise (X&1)/3 has
a known value of zero, and (X&3)/3 has a nonzero bits mask of 0x1.
2021-08-23 Roger Sayle <roger@nextmovesoftware.com>
gcc/ChangeLog
* tree-ssa-ccp.c (bit_value_binop) [TRUNC_MOD_EXPR, TRUNC_DIV_EXPR]:
Provide bounds for unsigned (and signed with non-negative operands)
division and modulus.
Whilst working on a backend patch, I noticed that the middle-end's
RTL optimizers weren't simplifying a truncation of a paradoxical
subreg extension, though it does transform closely related (more
complex) expressions. The main (first) part of this patch
implements this simplification, reusing much of the logic already
in place.
I briefly considered suggesting that it's difficult to provide a new
testcase for this change, but then realized the reviewer's response
would be that this type of transformation should be self-tested
in simplify-rtx, so this patch adds a bunch of tests that integer
extensions and truncations are simplified as expected. No good
deed goes unpunished and I was equally surprised to see that we
don't currently simplify/check/defend (zero_extend:SI (reg:SI)),
i.e. useless no-op extensions to the same mode. So I've added
some logic to simplify (or more accurately prevent us generating
dubious RTL for) those.
2021-08-23 Roger Sayle <roger@nextmovesoftware.com>
gcc/ChangeLog
* simplify-rtx.c (simplify_truncation): Generalize simplification
of (truncate:A (subreg:B X)).
(simplify_unary_operation_1) [FLOAT_TRUNCATE, FLOAT_EXTEND,
SIGN_EXTEND, ZERO_EXTEND]: Handle cases where the operand
already has the desired machine mode.
(test_scalar_int_ops): Add tests that useless extensions and
truncations are optimized away.
(test_scalar_int_ext_ops): New self-test function to confirm
that truncations of extensions are correctly simplified.
(test_scalar_int_ext_ops2): New self-test function to check
truncations of truncations, extensions of extensions, and
truncations of extensions.
(test_scalar_ops): Call the above two functions with a
representative sampling of integer machine modes.
This short patch teaches fold that it is "safe" to change the sign
of a left shift, to reduce the number of type conversions in gimple.
As an example:
unsigned int foo(unsigned int i) {
return (int)i << 8;
}
is currently optimized to:
unsigned int foo (unsigned int i)
{
int i.0_1;
int _2;
unsigned int _4;
<bb 2> [local count: 1073741824]:
i.0_1 = (int) i_3(D);
_2 = i.0_1 << 8;
_4 = (unsigned int) _2;
return _4;
}
with this patch, this now becomes:
unsigned int foo (unsigned int i)
{
unsigned int _2;
<bb 2> [local count: 1073741824]:
_2 = i_1(D) << 8;
return _2;
}
which generates exactly the same assembly language. Aside from the
reduced memory usage, the real benefit is that no-op conversions tend
to interfere with many folding optimizations. For example,
unsigned int bar(unsigned char i) {
return (i ^ (i<<16)) | (i<<8);
}
currently gets (tangled in conversions and) optimized to:
unsigned int bar (unsigned char i)
{
unsigned int _1;
unsigned int _2;
int _3;
int _4;
unsigned int _6;
unsigned int _8;
<bb 2> [local count: 1073741824]:
_1 = (unsigned int) i_5(D);
_2 = _1 * 65537;
_3 = (int) i_5(D);
_4 = _3 << 8;
_8 = (unsigned int) _4;
_6 = _2 | _8;
return _6;
}
but with this patch, bar now optimizes down to:
unsigned int bar(unsigned char i)
{
unsigned int _1;
unsigned int _4;
<bb 2> [local count: 1073741824]:
_1 = (unsigned int) i_3(D);
_4 = _1 * 65793;
return _4;
}
2021-08-23 Roger Sayle <roger@nextmovesoftware.com>
gcc/ChangeLog
* match.pd (shift transformations): Change the sign of an
LSHIFT_EXPR if it reduces the number of explicit conversions.
gcc/testsuite/ChangeLog
* gcc.dg/fold-convlshift-1.c: New test case.
* gcc.dg/fold-convlshift-2.c: New test case.
The following patch recognizes in the bswap pass (only there for now,
haven't done it for store merging pass yet) code sequences that can
be handled by (int32) __builtin_bswap64 (arg), i.e. where we have
0x05060708 n->n with 64-bit non-memory argument (if it is memory, we
can just load the 32-bit at 4 bytes into the address and n->n would
be 0x01020304; and only 64 -> 32 bit, because 64 -> 16 bit or 32 -> 16 bit
would mean only two bytes in the result and probably not worth it),
and furthermore the case where we have in the 0x0102030405060708 etc.
numbers some bytes 0 (i.e. known to contain zeros rather than source bytes),
as long as we have at least two original bytes in the right
positions (and no unknown bytes). This can be handled by
__builtin_bswap64 (arg) & 0xff0000ffffff00ffULL etc.
The latter change is the reason why counting the bswap messages doesn't work
too well in optimize-bswap* tests anymore, while the pass iterates from end
of basic block towards start, it will often match both the bswap at the end
and some of the earlier bswaps with some masks (not a problem generally,
we'll just DCE it away whenever possible). The pass right now doesn't
handle __builtin_bswap* calls in the pattern matching (which is the reason
why it operates backwards), but it uses FOR_EACH_BB_FN (bb, fun) order
of handling blocks and matched sequences can span multiple blocks, so I was
worried about cases like:
void bar (unsigned long long);
unsigned long long
foo (unsigned long long value, int x)
{
unsigned long long tmp = (((value & 0x00000000000000ffull) << 56)
| ((value & 0x000000000000ff00ull) << 40)
| ((value & 0x00000000ff000000ull) << 8));
if (x)
bar (tmp);
return (tmp
| ((value & 0x000000ff00000000ull) >> 8)
| ((value & 0x0000ff0000000000ull) >> 24)
| ((value & 0x0000000000ff0000ull) << 24)
| ((value & 0x00ff000000000000ull) >> 40)
| ((value & 0xff00000000000000ull) >> 56));
}
but it seems we handle even that fine, while bb2 ending in GIMPLE_COND
is processed first, we recognize there a __builtin_bswap64 (value) & mask1,
in the last bb we recognize tmp | (__builtin_bswap64 (value) & mask2) and
PRE optimizes that into t = __builtin_bswap64 (value); tmp = t & mask1;
in the first bb and return t; in the last one.
2021-08-23 Jakub Jelinek <jakub@redhat.com>
PR tree-optimization/86723
* gimple-ssa-store-merging.c (find_bswap_or_nop_finalize): Add
cast64_to_32 argument, set *cast64_to_32 to false, unless n is
non-memory permutation of 64-bit src which only has bytes of
0 or [5..8] and n->range is 4.
(find_bswap_or_nop): Add cast64_to_32 and mask arguments, adjust
find_bswap_or_nop_finalize caller, support bswap with some bytes
zeroed, as long as at least two bytes are not zeroed.
(bswap_replace): Add mask argument and handle masking of bswap
result.
(maybe_optimize_vector_constructor): Adjust find_bswap_or_nop
caller, punt if cast64_to_32 or mask is not all ones.
(pass_optimize_bswap::execute): Adjust find_bswap_or_nop_finalize
caller, for now punt if cast64_to_32.
* gcc.dg/pr86723.c: New test.
* gcc.target/i386/pr86723.c: New test.
* gcc.dg/optimize-bswapdi-1.c: Use -fdump-tree-optimized instead of
-fdump-tree-bswap and scan for number of __builtin_bswap64 calls.
* gcc.dg/optimize-bswapdi-2.c: Likewise.
* gcc.dg/optimize-bswapsi-1.c: Use -fdump-tree-optimized instead of
-fdump-tree-bswap and scan for number of __builtin_bswap32 calls.
* gcc.dg/optimize-bswapsi-5.c: Likewise.
* gcc.dg/optimize-bswapsi-3.c: Likewise. Expect one __builtin_bswap32
call instead of zero.
This replicates tree-eh.c in_array_bound_p into VNs
vn_reference_may_trap to fix hoisting of a possibly trapping
ARRAY_REF across a call that might not return.
2021-08-23 Richard Biener <rguenther@suse.de>
PR tree-optimization/79334
* tree-ssa-sccvn.c (copy_reference_ops_from_ref): Record
a type also for COMPONENT_REFs.
(vn_reference_may_trap): Check ARRAY_REF with constant index
against the array domain.
* gcc.dg/torture/pr79334-0.c: New testcase.
* gcc.dg/torture/pr79334-1.c: Likewise.
