* config/rs6000/rs6000.c (rs6000_declare_alias): Rename and globalize
both the symbol and the "dot" symbol for function descriptors. Fix
inversion for rename of symbols with dollar sign.
From-SVN: r230867
2015-11-25 Tom de Vries <tom@codesourcery.com>
* cfgloop.c (find_single_latch): New function, factored out of ...
(flow_loops_find): ... here.
(verify_loop_structure): Improve verification of loop->latch.
* cfgloop.h (find_single_latch): Declare.
* omp-low.c (expand_omp_for_generic): Initialize latch of orig_loop.
From-SVN: r230866
PR rtl-optimization/68435
* ifcvt.c (noce_try_cmove_arith): Skip final insn when checking
for conflicts between a, b and the set destinations.
* gcc.dg/pr68435.c: New test.
From-SVN: r230859
* ipa-inline-analysis.c: Include gimplify.h
(set_cond_stmt_execution_predicate,
set_switch_stmt_execution_predicate): Be sure to not leak locations
to function body.
From-SVN: r230857
2015-11-25 Richard Biener <rguenther@suse.de>
PR tree-optimization/68492
* tree-vect-slp.c (vect_build_slp_tree): Consistently build up
op from scalars after operand swapping.
From-SVN: r230855
2015-11-25 Richard Biener <rguenther@suse.de>
PR tree-optimization/68502
* tree-vect-data-refs.c (vect_analyze_group_access_1): Restore
check that the step is a multiple of the type size.
* gcc.dg/vect/pr68502-1.c: New testcase.
* gcc.dg/vect/pr68502-2.c: Likewise.
From-SVN: r230854
* config/aarch64/aarch64-simd.md (widen_ssum, widen_usum)
(aarch64_<ANY_EXTEND:su><ADDSUB:optab>w<mode>_internal): New patterns
* config/aarch64/iterators.md (Vhalf, VDBLW): New mode attributes.
* gcc.target/aarch64/saddw-1.c: New test.
* gcc.target/aarch64/saddw-2.c: New test.
* gcc.target/aarch64/uaddw-1.c: New test.
* gcc.target/aarch64/uaddw-2.c: New test.
* gcc.target/aarch64/uaddw-3.c: New test.
* lib/target-support.exp
(check_effective_target_vect_widen_sum_hi_to_si_pattern):
Add aarch64 to list of support targets.
From-SVN: r230853
2015-11-24 Steve Ellcey <sellcey@imgtec.com>
* frame-header-opt.c (gate): Check for optimize > 0.
(has_inlined_assembly): New function.
(needs_frame_header_p): Remove is_leaf_function check,
add argument type check.
(callees_functions_use_frame_header): Add is_leaf_function
and has_inlined_assembly calls..
(set_callers_may_not_allocate_frame): New function.
(frame_header_opt): Add is_leaf_function call, add
set_callers_may_not_allocate_frame call.
* config/mips/mips.c (mips_compute_frame_info): Add check
to see if callee saved regs can be put in frame header.
(mips_expand_prologue): Add check to see if step1 is zero,
fix cfa restores when using frame header to store regs.
(mips_can_use_return_insn): Check to see if registers are
stored in frame header.
* config/mips/mips.h (machine_function): Add
callers_may_not_allocate_frame and
use_frame_header_for_callee_saved_regs fields.
From-SVN: r230845
Part of the shrink-wrapping algorithm has this comment:
/* Now see if we can put the prologue at the start of PRO. Putting it
there might require duplicating a block that cannot be duplicated,
or in some cases we cannot insert the prologue there at all. If PRO
wont't do, try again with the immediate dominator of PRO, and so on.
The blocks that need duplicating are those reachable from PRO but
not dominated by it. We keep in BB_WITH a bitmap of the blocks
reachable from PRO that we already found, and in VEC a stack of
those we still need to consider (to find successors). */
Two of the cases that push to that stack do not actually check the
bitmap first. Either I thought those blocks could not be on the stack
already, or more likely I didn't think and it just didn't crash during
any testing. But of course those blocks *can* already be on the stack
(if you have a hideous loop structure), and then we end up with the
same block on the stack more than once. This is harmless, except that
(like in the PR) this can overflow the stack.
This fixes it, by doing the necessary bitmap checks before pushing.
PR rtl-optimization/68520
* shrink-wrap.c (try_shrink_wrapping): Don't push a block to VEC if
its bit was already set in BB_WITH.
From-SVN: r230843
* alias.c (get_alias_set): Before checking TYPE_ALIAS_SET_KNOWN_P
double check that type is main variant.
* tree.c (build_variant_type_copy): Clear TYPE_ALIAS_SET when producing
variant.
(verify_type_variant): Verify that variants have no
TYPE_ALIAS_SET_KNOWN_P set
* tree-streamer-out.c (pack_ts_type_common_value_fields): Reorder
streaming so constant fields come first; stream TYPE_ALIAS_SET==0
only for main variants; stream TYPE_ALIAS_SET as a bit.
* tree-streamer-in.c (unpack_ts_type_common_value_fields): Update
accordingly.
From-SVN: r230838
* tree.c (make_vector_type): Properly compute canonical type of the
main variant.
(verify_type): Verify that TYPE_CANONICAL of TYPE_MAIN_VARIANT is
a main variant.
From-SVN: r230837
* lto-streamer-in.c (lto_read_body_or_constructor): Set TYPE_CANONICAL
only for types where LTO sets them.
* tree.c (build_array_type_1): Do ont set TYPE_CANONICAL for LTO.
(make_vector_type): Likewise.
(gimple_canonical_types_compatible_p): Use canonical_type_used_p.
* tree.h (canonical_type_used_p): New inline.
* alias.c (get_alias_set): Handle structural equality for all
types that pass canonical_type_used_p.
(record_component_aliases): Look through all types with
record_component_aliases for possible pointers; sanity check that
the alias sets match.
* lto.c (iterative_hash_canonical_type): Recruse for all types
which pass !canonical_type_used_p.
(gimple_register_canonical_type_1): Sanity check we do not compute
canonical type of anything with !canonical_type_used_p.
(gimple_register_canonical_type): Skip all types that are
!canonical_type_used_p
From-SVN: r230835
2015-11-24 Steven G. Kargl <kargl@gcc.gnu.org>
PR fortran/68486
* simplify.c (gfc_simplify_cshift): Allow all EXPR_*.
2015-11-24 Steven G. Kargl <kargl@gcc.gnu.org>
PR fortran/68486
* gfortran.dg/simplify_cshift_3.f90
From-SVN: r230832
2015-11-24 David Edelsohn <dje.gcc@gmail.com>
Michael Meissner <meissner@linux.vnet.ibm.com>
* config/rs6000/rs6000.md (UNSPEC_XSRDPI): New unspec.
(Fv2): New mode attribute to be used when ISA 2.06 instructions
are used on SF/DF values.
(abs<mode>2_fpr): Use <Fv2> instead of <Fv>.
(nabs<mode>2_fpr): Likewise.
(neg<mode>2_fpr): Likewise.
(copysign<mode>3_fcpsgn): Likewise.
(smax<mode>3_vsx): Likewise.
(smin<mode>3_vsx): Likewise.
(floatsi<mode>2_lfiwax): Likewise.
(floatunssi<mode>2_lfiwz): Likewise.
(fctiwz_<mode>): Likewise.
(fctiwuz_<mode>): Likewise.
(btrunc<mode>2): Likewise.
(ceil<mode>2): Likewise.
(floor<mode>2): Likewise.
(xsrdpi<mode>): Add support for the lround function.
(lround<mode>2): Likewise.
Co-Authored-By: Michael Meissner <meissner@linux.vnet.ibm.com>
From-SVN: r230829
... so that debugger users can reference them instead of the renamed
subprograms. This is in accordance with the DWARF specification: the
section 3.2.3 (Imported (or Renamed) Declaration Entities) allows
DW_TAG_imported_declaration DIEs to be used "as a general means to
rename or provide an alias for an entity regardless of the context in
which the importing declaration or the imported entity occur." (wording
from the DWARFv3 and DWARFv4 specifications, allowed in DWARFv2).
gcc/ada/ChangeLog:
* gcc-interface/utils.c (gnat_write_global_declarations): Output
debugging information for top-level imported declarations.
gcc/ChangeLog:
* dwarf2out.c (gen_decl_die): Generate DW_TAG_imported_* instead
of DW_TAG_namespace for IMPORTED_DECL declarations. Call
dwarf2out_imported_module_or_decl_1 for all DWARF versions as
this function already takes care of checking what it can emit.
Bootstrapped, regtested and checked with GDB's testsuite on
x86_64-linux.
From-SVN: r230821
2015-11-24 Tom de Vries <tom@codesourcery.com>
* tree-ssa-loop-im.c (tree_ssa_lim): Make static.
(pass_lim::execute): Allow to run outside pass_tree_loop.
From-SVN: r230815
gcc/
* gcc/config/aarch64/aarch64.md (add<mode>3):
Block early expansion into 2 add instructions.
(add<mode>3_pluslong): New pattern to combine complex
immediates into 2 additions.
From-SVN: r230814
Calling rs6000_is_valid_and_mask on a reg instead of on a const_int is
not a good idea, as PR68332 and PR67677 as well as testing with
--enable-checking=yes,rtl show. Fix this.
PR target/66217
PR target/67677
PR target/68332
* config/rs6000/predicates.md (and_operand): Check that the operand
is a const_int before calling rs6000_is_valid_and_mask.
From-SVN: r230811
gcc/ChangeLog:
2015-11-24 Trevor Saunders <tbsaunde+gcc@tbsaunde.org>
* hash-map-traits.h (simple_hashmap_traits ::remove): call
destructors on values that are being removed.
* mem-stats.h (hash_map): Pass type of values to
simple_hashmap_traits.
* tree-sra.c (sra_deinitialize): Remove work around for hash
maps not destructing values.
* genmatch.c (sinfo_hashmap_traits): Adjust.
* tree-ssa-uncprop.c (val_ssa_equiv_hash_traits): Likewise.
From-SVN: r230801
PR target/68483
* tree-vect-generic.c (lower_vec_perm): If VEC_PERM_EXPR
is valid vec_shr pattern, don't lower it even if can_vec_perm_p
returns false.
* optabs.c (shift_amt_for_vec_perm_mask): Return NULL_RTX
whenever first is nelt or above. Don't mask expected with
2 * nelt - 1.
* gcc.target/i386/pr68483-1.c: New test.
* gcc.target/i386/pr68483-2.c: New test.
From-SVN: r230797
