PR fortran/66549
gcc/fortran/
* resolve.c (resolve_global_procedure): Don't save and restore
OpenMP state around the call to gfc_resolve.
(gfc_resolve): Save OpenMP state on entry and restore it on return.
gcc/testsuite/
* gfortran.dg/gomp/omp_parallel_1.f90: New file.
From-SVN: r224648
* config/i386/i386.md (*movsicc_noc_zext): New insn.
(zero-extended cmove with mem peephole2): New pattern.
(cmove with mem peephole2): Merge patterns.
testsuite/ChangeLog:
* gcc.target/i386/cmov9.c: New test.
From-SVN: r224631
PR middle-end/66253
* tree-vect-stmts.c (vectorizable_store): Implement non-SLP
grouped strided stores.
(vectorizable_load): Don't use the DR from first_stmt in
the non-SLP grouped strided case.
testsuite/
* gcc.dg/vect/pr66253.c: New testcase.
From-SVN: r224605
PR c++/66571
* pt.c (tsubst_omp_clause_decl): New function.
(tsubst_omp_clauses): Use it or tsubst_copy instead of
tsubst_expr on OMP_CLAUSE_DECL.
* g++.dg/gomp/pr66571-1.C: New test.
From-SVN: r224599
2015-06-18 Richard Biener <rguenther@suse.de>
PR tree-optimization/66510
* tree-vect-stmts.c (vectorizable_load): Properly compute the
number of vector loads for SLP permuted loads.
* tree-vect-data-refs.c (vect_compute_data_ref_alignment): Also
check the stride for loop vectorization.
(vect_enhance_data_refs_alignment): Deal with SLP adjusted
vectorization factor.
(vect_analyze_group_access): If the group size is not a power
of two require a epilogue loop.
* tree-vect-loop.c (vect_analyze_loop_2): Move alignment
compute and optimizing and alias test pruning after final
vectorization factor computation.
* tree-vect-slp.c (vect_build_slp_tree_1): Remove check on
vector alignment.
(vect_transform_slp_perm_load): Properly compute the original
number of vector load stmts.
* gcc.dg/vect/slp-perm-12.c: New testcase.
From-SVN: r224598
PR middle-end/66429
* omp-low.c (expand_omp_taskreg, expand_omp_target): Use child_cfun
instead of DECL_STRUCT_FUNCTION (child_fn). Or in has_simduid_loops
and has_force_vectorize_loops flags from cfun into
child_cfun.
(expand_omp_simd): For broken loop, set cfun->has_simduid_loops
if simduid is non-NULL.
* tree-pass.h (make_pass_simduid_cleanup): New prototype.
* passes.def (pass_simduid_cleanup): Add new pass after loop
passes.
* tree-vectorizer.c (adjust_simduid_builtins): Remove one unnecessary
indirection from htab argument's type.
(shrink_simd_arrays): New function.
(vectorize_loops): Use it. Adjust adjust_simduid_builtins caller.
Don't call adjust_simduid_builtins if there are no loops.
(pass_data_simduid_cleanup, pass_simduid_cleanup): New variables.
(pass_simduid_cleanup::execute): New method.
(make_pass_simduid_cleanup): New function.
From-SVN: r224571
PR c++/66571
* pt.c (tsubst_omp_clause_decl): New function.
(tsubst_omp_clauses): Use it or tsubst_copy instead of
tsubst_expr on OMP_CLAUSE_DECL.
* g++.dg/gomp/pr66571-1.C: New test.
From-SVN: r224569
PR middle-end/66429
* omp-low.c (expand_omp_taskreg): Use child_cfun instead of
DECL_STRUCT_FUNCTION (child_fn). Or in has_simduid_loops
and has_force_vectorize_loops flags from cfun into
child_cfun.
(expand_omp_simd): For broken loop, set cfun->has_simduid_loops
if simduid is non-NULL.
* tree-pass.h (make_pass_simduid_cleanup): New prototype.
* passes.def (pass_simduid_cleanup): Add new pass after loop
passes.
* tree-vectorizer.c (adjust_simduid_builtins): Remove one unnecessary
indirection from htab argument's type.
(shrink_simd_arrays): New function.
(vectorize_loops): Use it. Adjust adjust_simduid_builtins caller.
Don't call adjust_simduid_builtins if there are no loops.
(pass_data_simduid_cleanup, pass_simduid_cleanup): New variables.
(pass_simduid_cleanup::execute): New method.
(make_pass_simduid_cleanup): New function.
* c-c++-common/gomp/pr66429.c: New test.
From-SVN: r224568
2017-06-17 Andrew MacLeod <amacleod@redhat.com>
* tree-core.h (tree_target_option): Make opts field a pointer to a
cl_target_option instead of an instance of the struct.
* tree.h (TREE_TARGET_OPTION): Return the pointer, not an address of
the structure.
* tree.c (make_node_stat ): Allocate a cl_target_option struct for
TARGET_OPTION_NODE.
(copy_node_stat): Allocate and copy struct cl_target_option.
From-SVN: r224567
* include/bits/forward_list.h
(_Fwd_list_base(const _Node_alloc_type&)): Change parameter to
rvalue-reference.
(_Fwd_list_base(_Fwd_list_base&&, const _Node_alloc_type&)): Likewise.
(forward_list(const _Alloc&)): Split default constructor out to
separate function.
(forward_list(forward_list&&, const _Alloc&)): Move elements if base
class didn't do so.
(forward_list::_M_move_assign(forward_list&&, true_type)): Replace
swap call with two assignments.
* include/bits/forward_list.tcc
(_Fwd_list_base(_Fwd_list_base&&, const _Node_alloc_type&)): Don't
move elements when allocators are not equal.
* include/debug/forward_list (forward_list(const allocator_type&)):
Split default constructor out to separate function.
* include/profile/forward_list (forward_list(const _Alloc&)):
Likewise.
From-SVN: r224566
gcc/jit/ChangeLog:
* libgccjit.c (gcc_jit_lvalue_access_field): Verify that the field
is for the correct struct.
(gcc_jit_rvalue_access_field): Likewise.
gcc/testsuite/ChangeLog:
* jit.dg/test-error-accessing-field-in-other-struct.c: Rename to...
* jit.dg/test-error-gcc_jit_rvalue_dereference_field-wrong-struct.c:
...this.
* jit.dg/test-error-gcc_jit_lvalue_access_field-wrong-struct.c:
New testcase.
* jit.dg/test-error-gcc_jit_rvalue_access_field-wrong-struct.c:
New testcase.
From-SVN: r224565