2018-09-06 Will Schmidt <will_schmidt@vnet.ibm.com>
* config/rs6000/rs6000.c (rs6000_gimple_fold_builtin): Add support for
early gimple folding of vec_splat().
* tree-vect-generic.c: Remove static from tree_vec_extract() definition.
* gimple-fold.h: Add an extern define for tree_vec_extract().
Inhibit constant propagation inlining SYMBOL_REF loads into
UNSPECV_CAS. Even though reload can later undo it, the resulting
code will be less efficient.
gcc/ChangeLog:
2018-09-06 Ilya Leoshkevich <iii@linux.ibm.com>
PR target/80080
* config/s390/predicates.md: Add nonsym_memory_operand.
* config/s390/s390.c (s390_legitimize_cs_operand): If operand
contains a SYMBOL_REF, load it into an intermediate pseudo.
(s390_emit_compare_and_swap): Legitimize operand.
* config/s390/s390.md: Use the new nonsym_memory_operand
with UNSPECV_CAS patterns.
gcc/testsuite/ChangeLog:
2018-09-06 Ilya Leoshkevich <iii@linux.ibm.com>
PR target/80080
* gcc.target/s390/pr80080-3.c: New test.
* gcc.target/s390/s390.exp: Make sure the new test passes
on all optimization levels.
The dump file used to come at the end of the sorted dump file list,
because the pass was registered dynamically. This did not reflect the
order in which passes are executed. Static registration fixes this:
PR c++/86982, -Wreturn-local-addr and std::move and std::forward.
* typeck.c (maybe_warn_about_returning_address_of_local): Handle calls
to std::move or std::forward.
(is_std_forward_p): New function.
cesar [Wed, 5 Sep 2018 22:27:31 +0000 (22:27 +0000)]
[nvptx] Basic -misa support for nvptx
gcc/
* config/nvptx/nvptx-opts.h: New file.
* config/nvptx/nvptx.c (nvptx_file_start): Print the correct .target.
* config/nvptx/nvptx.h: Include "nvptx-opts.h".
(ASM_SPEC): Define.
(TARGET_SM35): New macro.
* config/nvptx/nvptx.md (atomic_fetch_<logic><mode>): Enable with the
correct predicate.
* config/nvptx/nvptx.opt (ptx_isa, sm_30, sm_35): New enum and its
values.
(misa=): New option.
* doc/invoke.texi (Nvidia PTX Options): Document -misa.
gcc/testsuite/
* gcc.target/nvptx/atomic_fetch-1.c: New test.
* gcc.target/nvptx/atomic_fetch-1.c: New test.
PR c++/87109, wrong overload with ref-qualifiers.
* call.c (build_user_type_conversion_1): Use NULL instead of 0. Bail
out if performing the maybe-rvalue overload resolution and a conversion
function is getting called.
PR52665 do not let .ident confuse assembler scan tests
A branch with a name matching scan-assembler pattern triggers
inappropriate FAIL.
E.g. branch fixups-testsuite and
- gcc.target/i386/pr65871-?.c (scan-assembler-not "test")
- gcc.target/i386/pr41442.c (scan-assembler-times "test|cmp" 2)
etc.
This is a recurring problem as can be seen by some -fno-ident additions
by commits from e.g. Michael Meissner over the years: builtins-58.c,
powerpc/pr46728-?.c
The patch below adds -fno-ident if a testcase contains one of
scan-assembler, scan-assembler-not or scan-assembler-times.
Regression tested on x86_64-unknown-linux on a fixups-testsuite branch
where it fixes several false FAILs without regressions.
gcc/testsuite/ChangeLog
2016-06-18 Bernhard Reutner-Fischer <aldot@gcc.gnu.org>
PR testsuite/52665
* lib/gcc-dg.exp (gcc-dg-test-1): Iterate over _required_options.
* lib/target-supports.exp (scan-assembler_required_options,
scan-assembler-not_required_options,
scan-assembler-times_required_options): Add -fno-ident.
* lib/scanasm.exp (scan-assembler-times): Fix error message.
* c-c++-common/ident-0a.c: New test.
* c-c++-common/ident-0b.c: New test.
* c-c++-common/ident-1a.c: New test.
* c-c++-common/ident-1b.c: New test.
* c-c++-common/ident-2a.c: New test.
* c-c++-common/ident-2b.c: New test.
This patch aims to optimise sequences involving uses of 1.0 / sqrt (a) under -freciprocal-math and -funsafe-math-optimizations.
In particular consider:
x = 1.0 / sqrt (a);
r1 = x * x; // same as 1.0 / a
r2 = a * x; // same as sqrt (a)
If x, r1 and r2 are all used further on in the code, this can be transformed into:
tmp1 = 1.0 / a
tmp2 = sqrt (a)
tmp3 = tmp1 * tmp2
x = tmp3
r1 = tmp1
r2 = tmp2
A bit convoluted, but this saves us one multiplication and, more importantly, the sqrt and division are now independent.
This also allows optimisation of a subset of these expressions.
For example:
x = 1.0 / sqrt (a)
r1 = x * x
can be transformed to r1 = 1.0 / a, eliminating the sqrt if x is not used anywhere else.
And similarly:
x = 1.0 / sqrt (a)
r1 = a * x
can be transformed to sqrt (a) eliminating the division.
For the testcase:
double res, res2, tmp;
void
foo (double a, double b)
{
tmp = 1.0 / __builtin_sqrt (a);
res = tmp * tmp;
res2 = a * tmp;
}
We now generate for aarch64 with -Ofast:
foo:
fmov d2, 1.0e+0
adrp x2, res2
fsqrt d1, d0
adrp x1, res
fdiv d0, d2, d0
adrp x0, tmp
str d1, [x2, #:lo12:res2]
fmul d1, d1, d0
str d0, [x1, #:lo12:res]
str d1, [x0, #:lo12:tmp]
ret
where before it generated:
foo:
fsqrt d2, d0
fmov d1, 1.0e+0
adrp x1, res2
adrp x2, tmp
adrp x0, res
fdiv d1, d1, d2
fmul d0, d1, d0
fmul d2, d1, d1
str d1, [x2, #:lo12:tmp]
str d0, [x1, #:lo12:res2]
str d2, [x0, #:lo12:res]
ret
As you can see, the new sequence has one fewer multiply and the fsqrt and fdiv are independent.
* tree-ssa-math-opts.c (is_mult_by): New function.
(is_square_of): Use the above.
(optimize_recip_sqrt): New function.
(pass_cse_reciprocals::execute): Use the above.
Group switch cases in switch lowering (PR tree-optimization/87205).
2018-09-05 Martin Liska <mliska@suse.cz>
PR tree-optimization/87205
* tree-switch-conversion.c (pass_lower_switch::execute):
Group cases for switch statements.
2018-09-05 Martin Liska <mliska@suse.cz>
PR tree-optimization/87205
* gcc.dg/tree-ssa/pr87205-2.c: New test.
* gcc.dg/tree-ssa/pr87205.c: New test.
https://gcc.gnu.org/ml/gcc-patches/2018-08/msg01966.html
PR c++/87137
* stor-layout.c (place_field): Scan forwards to check last
bitfield when ms_bitfield_placement is in effect.
gcc/testsuite/
* g++.dg/abi/pr87137.C: New.
This is a rewrite of the tag collision avoidance patch that Kugan had
written as a machine reorg pass back in February.
The falkor hardware prefetching system uses a combination of the
source, destination and offset to decide which prefetcher unit to
train with the load. This is great when loads in a loop are
sequential but sub-optimal if there are unrelated loads in a loop that
tag to the same prefetcher unit.
This pass attempts to rename the desination register of such colliding
loads using routines available in regrename.c so that their tags do
not collide. This shows some performance gains with mcf and xalancbmk
(~5% each) and will be tweaked further. The pass is placed near the
fag end of the pass list so that subsequent passes don't inadvertantly
end up undoing the renames.
GCOV: Print one decimal place in human readable mode.
2018-09-05 Martin Liska <mliska@suse.cz>
* doc/gcov.texi: Update documentation of humar
readable mode.
* gcov.c (format_count): Print one decimal place, it provides
more fine number of situations like '1G' vs. '1.4G'.
2018-09-05 Martin Liska <mliska@suse.cz>
* g++.dg/gcov/loop.C: Update test to support new format.
PR target/87164
* config/rs6000/rs6000.opt: Mark the option as Deprecated.
* optc-gen.awk: Allow 'Var' for Deprecated options in order
to generate a MASK value.
i386: Don't use frame pointer without stack access
When there is no stack access, there is no need to use frame pointer
even if -fno-omit-frame-pointer is used and caller's frame pointer is
unchanged.
frame pointer may not be available even if -fno-omit-frame-pointer is
used. When this happened, arg pointer may be eliminated by hard frame
pointer. Since hard frame pointer is encoded with DW_OP_fbreg which
uses the DW_AT_frame_base attribute, not hard frame pointer directly,
we should allow hard frame pointer when generating DWARF info even if
frame pointer isn't used.
gcc/
PR debug/86593
* dwarf2out.c (based_loc_descr): Allow hard frame pointer even
if frame pointer isn't used.
(compute_frame_pointer_to_fb_displacement): Likewise.
* wide-int-range.cc (wide_int_range_convert): New.
* wide-int-range.h (wide_int_range_convert): New.
* tree-vrp.c (extract_range_from_unary_expr): Abstract wide int
code into wide_int_range_convert.
(extract_range_into_wide_ints): Do not munge anti range constants
into the entire domain. Just return the range back.
PR tree-optimization/87211
* tree-ssa-sccvn.c (visit_phi): When value-numbering to a
backedge value we're supposed to treat as VARYING also number
the PHI to VARYING in case it got a different value-number already.
PR tree-optimization/87176
* tree-ssa-sccvn.c (visit_phi): Remove redundant allsame
variable. When value-numbering a virtual PHI node make sure
to not value-number to the backedge value.
* gcc.dg/torture/pr87176.c: New testcase.
* gcc.dg/torture/ssa-fre-1.c: Likewise.
* sort.cc (struct sort_ctx): New field 'nlim'. Use it...
(mergesort): ... here as maximum count for using netsort.
(gcc_qsort): Set nlim to 3 if stable sort is requested.
(gcc_stablesort): New.
* system.h (gcc_stablesort): Declare.
Our md files refer to {l,st}xsd%U<n>x, but no {l,st}xsdux insns exist.
This patch removes the update forms. All these use constraint "Z"
which does not allow update form, so there is no practical difference.
* config/rs6000/rs6000.md (*mov<mode>_hardfloat32): Remove %U from the
lxsdx and stxsdx alternatives.
(*mov<mode>_hardfloat64): Ditto.
* config/rs6000/vsx.md (*vsx_extract_<mode>_store): Ditto.
PR libstdc++/78179 run long double tests separately
Split the long double testing into a separate file, so that we can XFAIL
targets where the long double precision doesn't meet the expected
tolerances. The float and double tests are still expefted to PASS for
all targets.
PR libstdc++/78179
* testsuite/26_numerics/headers/cmath/hypot-long-double.cc: New test
that runs the long double part of hypot.cc.
* testsuite/26_numerics/headers/cmath/hypot.cc: Disable long double
tests unless TEST_HYPOT_LONG_DOUBLE is defined.
The pointer argument to allocator_traits::construct and
allocator_traits::destroy should be a raw pointer, not the allocator's
pointer type. _Temporary_value::_M_ptr was returning the wrong type.
* include/bits/stl_vector.h (vector::_Temporary_value::_M_ptr):
Return raw pointer not allocator's pointer type.
(vector::_Temporary_value::_M_val): Use _M_ptr.
PR libstdc++/87194 fix range insertion into maps and sets
Since C++11 range insertion and construction of maps and sets from a
pair of iterators only requires that the iterator's value_type is
convertible to the container's value_type (previously it had to be the
same).
This fixes the implementation to meet that relaxed requirement, by
defining a pair of overloads that either insert or emplace, depending on
the iterator's value_type. Instead of adding yet another overload of
_M_insert_unique and _M_insert_equal, the overloads taking iterators are
renamed to _M_insert_range_unique and _M_insert_range_equal.
PR libstdc++/87194
* include/bits/stl_map.h
(map::map(initializer_list<value_type>, const Compare&, const Alloc&))
(map::map(initializer_list<value_type>, const Alloc&))
(map::map(InputIterator, InputIterator, const Alloc&))
(map::map(InputIterator, InputIterator))
(map::map(InputIterator, InputIterator, const Compare&, const Alloc&))
(map::insert(InputIterator, InputIterator)):
Call _M_insert_range_unique instead of _M_insert_unique.
* include/bits/stl_multimap.h
(multimap::multimap(initializer_list<value_type>, const C&, const A&))
(multimap::multimap(initializer_list<value_type>, const A&))
(multimap::multimap(InputIterator, InputIterator, const A&))
(multimap::multimap(InputIterator, InputIterator))
(multimap::multimap(InputIterator, InputIterator, const C&, const A&))
(multimap::insert(InputIterator, InputIterator)): Call
_M_insert_range_equal instead of _M_insert_equal.
* include/bits/stl_multiset.h
(multiset::multiset(InputIterator, InputIterator))
(multiset::multiset(InputIterator, InputIterator, const C&, const A&))
(multiset::multiset(initializer_list<value_type>, const C&, const A&))
(multiset::multiset(initializer_list<value_type>, const A&))
(multiset::multiset(InputIterator, InputIterator, const A&))
(multiset::insert(InputIterator, InputIterator)): Call
_M_insert_range_equal instead of _M_insert_equal.
* include/bits/stl_set.h
(set::set(InputIterator, InputIterator))
(set::set(InputIterator, InputIterator, const Compare&, const Alloc&))
(set::set(initializer_list<value_type>, const Compare&, const Alloc&))
(set::set(initializer_list<value_type>, const Alloc&))
(set::set(InputIterator, InputIterator, const Alloc&))
(set::insert(InputIterator, InputIterator)):
Call _M_insert_range_unique instead of _M_insert_unique.
* include/bits/stl_tree.h
[__cplusplus >= 201103L] (_Rb_tree::__same_value_type): New alias
template for SFINAE constraints.
[__cplusplus >= 201103L] (_Rb_tree::_M_insert_range_unique): Pair of
constrained overloads that either insert or emplace, depending on
iterator's value_type.
[__cplusplus >= 201103L] (_Rb_tree::_M_insert_range_equal): Likewise.
[__cplusplus < 201103L] (_Rb_tree::_M_insert_range_unique)
(_Rb_tree::_M_insert_range_equal): New functions replacing range
versions of _M_insert_unique and _M_insert_equal.
(_Rb_tree::_M_insert_unique(_InputIterator, _InputIterator))
(_Rb_tree::_M_insert_equal(_InputIterator, _InputIterator)): Remove.
* testsuite/23_containers/map/modifiers/insert/87194.cc: New test.
* testsuite/23_containers/multimap/modifiers/insert/87194.cc: New test.
* testsuite/23_containers/multiset/modifiers/insert/87194.cc: New test.
* testsuite/23_containers/set/modifiers/insert/87194.cc: New test.
PR libstdc++/78595 implement insertion into maps in terms of emplace
C++14 simplified the specification of the generic insert function
templates to be equivalent to calling emplace (or emplace_hint).
Defining them in terms of emplace takes care of the problems described
in PR 78595, ensuring a single conversion to value_type is done at the
right time.
PR libstdc++/78595
* include/bits/stl_map.h (map::insert(_Pair&&))
(map::insert(const_iterator, _Pair&&)): Do emplace instead of insert.
* include/bits/stl_multimap.h (multimap::insert(_Pair&&))
(multimap::insert(const_iterator, _Pair&&)): Likewise.
* include/bits/unordered_map.h (unordered_map::insert(_Pair&&))
(unordered_map::insert(const_iterator, _Pair&&))
(unordered_multimap::insert(_Pair&&))
(unordered_multimap::insert(const_iterator, _Pair&&)): Likewise.
* testsuite/23_containers/map/modifiers/insert/78595.cc: New test.
* testsuite/23_containers/multimap/modifiers/insert/78595.cc: New test.
* testsuite/23_containers/unordered_map/modifiers/78595.cc: New test.
* testsuite/23_containers/unordered_multimap/modifiers/78595.cc: New
test.
PR tree-optimization/87197
* tree-ssa-sccvn.c (vn_nary_build_or_lookup_1): Mark the new def
visited. CSE the VN_INFO hashtable lookup.
* gcc.dg/torture/pr87197.c: New testcase.
PR tree-optimization/87169
* tree-ssa-sccvn.c (do_rpo_vn): When marking loops for not
iterating make sure there's no extra backedges from irreducible
regions feeding the header. Mark the destination block
executable.
fixincludes: vxworks: remove unnecessary parentheses in ioctl wrapper macro
The rationale for the fixinclude ioctl macro wrapper is, as far as I can
tell (https://gcc.gnu.org/ml/gcc-patches/2012-09/msg01619.html)
Fix 2: Add hack for ioctl() on VxWorks.
ioctl() is supposed to be variadic, but VxWorks only has a three
argument version with the third argument of type int. This messes up
when the third argument is not implicitly convertible to int. This
adds a macro which wraps around ioctl() and explicitly casts the third
argument to an int. This way, the most common use case of ioctl (with
a const char * for the third argument) will compile in C++, where
pointers must be explicitly casted to int.
However, we have existing C++ code that calls the ioctl function via
::ioctl(foo, bar, baz)
and obviously this breaks when it gets expanded to
::(ioctl)(foo, bar, (int)(baz))
Since the GNU C preprocessor already prevents recursive expansion of
function-like macros, the parentheses around ioctl are unnecessary.
Incidentally, there is also a macro sioIoctl() in the vxworks sioLib.h
header that expands to
This patch partly fixes that issue as well, but the third argument to
the pDrvFuncs->ioctl method should be void*, so the cast to (int) is
slightly annoying. Internally, we've simply patched the sioIoctl macro:
Come up with TARGET_GET_VALID_OPTION_VALUES option hook (PR driver/83193).
2018-09-03 Martin Liska <mliska@suse.cz>
PR driver/83193
* common/common-target.def: Add TARGET_GET_VALID_OPTION_VALUES.
* common/common-targhooks.c (default_get_valid_option_values):
New function.
* common/common-targhooks.h (default_get_valid_option_values):
Likewise.
* common/config/i386/i386-common.c: Move processor_target_table
from i386.c.
(ix86_get_valid_option_values): New function.
(TARGET_GET_VALID_OPTION_VALUES): New macro.
* config/i386/i386.c (struct ptt): Move to i386-common.c.
(PTA_*): Move all defined masks into i386-common.c.
(ix86_function_specific_restore): Use new processor_cost_table.
* config/i386/i386.h (struct ptt): Moved from i386.c.
(struct pta): Likewise.
* doc/tm.texi: Document new TARGET_GET_VALID_OPTION_VALUES.
* doc/tm.texi.in: Likewise.
* opt-suggestions.c (option_proposer::suggest_option):
Pass prefix to build_option_suggestions.
(option_proposer::get_completions): Likewise.
(option_proposer::build_option_suggestions): Use the new target
hook.
* opts.c (struct option_help_tuple): New struct.
(print_filtered_help): Use the new target hook.
2018-09-03 Martin Liska <mliska@suse.cz>
PR driver/83193
* gcc.dg/completion-4.c: New test.
Make __builtin_expect effective in switch statements (PR middle-end/PR59521).
2018-09-03 Martin Liska <mliska@suse.cz>
PR middle-end/59521
* predict.c (set_even_probabilities): Add likely_edges
argument and handle cases where we have precisely one
likely edge.
(combine_predictions_for_bb): Catch also likely_edges.
(tree_predict_by_opcode): Handle gswitch statements.
* tree-cfg.h (find_case_label_for_value): New declaration.
(find_taken_edge_switch_expr): Likewise.
* tree-switch-conversion.c (switch_decision_tree::balance_case_nodes):
Find pivot in decision tree based on probabily, not by number of
nodes.
2018-09-03 Martin Liska <mliska@suse.cz>
PR middle-end/59521
* c-c++-common/pr59521-1.c: New test.
* c-c++-common/pr59521-2.c: New test.
* gcc.dg/tree-prof/pr59521-3.c: New test.
* io/io.h: Change declaration of vlist type to
gfc_full_array_i4 to eliminate warning for mismatched type.
* io/format.c ((parse_format_list): Use gfc_full_array_i4.
* io/io.h: Use gfc_full_array_i4.
law [Sun, 2 Sep 2018 15:18:30 +0000 (15:18 +0000)]
* c-common.c (braced_list_to_string): Remove eval parameter.
Add some more checks. Always create zero-terminated STRING_CST.
* c-common.h (braced_list_to_string): Adjust prototype.
* c-decl.c (finish_decl): Call braced_list_to_string here ...
* c-parser.c (c_parser_declaration_or_fndef): ... instead of here.
* decl.c (eval_check_narrowing): Remove.
(check_initializer): Move call to braced_list_to_string from here ...
* typeck2.c (store_init_value): ... to here.
(digest_init_r): Remove handing of signed/unsigned char strings.
* c-c++-common/array-init.c: New test.
* g++.dg/init/string2.C: Remove xfail.
rguenth [Fri, 31 Aug 2018 16:50:13 +0000 (16:50 +0000)]
2018-08-31 Richard Biener <rguenther@suse.de>
PR tree-optimization/87168
* tree-ssa-sccvn.c (SSA_VAL): Add visited output parameter.
(rpo_elim::eliminate_avail): When OP was not visited it must
be available.
segher [Fri, 31 Aug 2018 08:52:13 +0000 (08:52 +0000)]
rs6000: Make lround<mode>di2 depend on TARGET_FPRND (PR86684)
TARGET_FPRND should be on for everything ISA 2.04 and later, and
TARGET_VSX implies ISA 2.06 or later; but it is possible to disable
TARGET_FPRND (separately via -mno-fprnd, but also implicitly)
currently, and then things fall down. This patch makes things not
fall down.
PR target/86684
PR target/87149
* config/rs6000/rs6000.md (lround<mode>di2): Gate on TARGET_FPRND.
pault [Fri, 31 Aug 2018 06:51:31 +0000 (06:51 +0000)]
2018-08-31 Paul Thomas <pault@gcc.gnu.org>
PR fortran/86328
PR fortran/86760
* trans-array.c (gfc_conv_scalarized_array_ref): Do not fix
info->descriptor but pass it directly to gfc_build_array_ref.
(gfc_conv_array_ref): Likewise for se->expr.
* trans.c (gfc_build_array_ref): If 'decl' is a COMPONENT_REF
obtain the span field directly from it.
2018-08-31 Paul Thomas <pault@gcc.gnu.org>
PR fortran/86328
PR fortran/86760
* gfortran.dg/pr86328.f90 : New test.
in comment 12 of the PR.
* gfortran.dg/pr86760.f90 : New test.
gcc/
* target.def (custom_function_descriptors): Improve documentation.
* doc/tm.texi.in (Trampolines): Expand discussion of function
descriptors and move TARGET_CUSTOM_FUNCTION_DESCRIPTORS to the
beginning of the section.
* doc/tm.texi: Regenerated.
jemarch [Thu, 30 Aug 2018 20:28:31 +0000 (20:28 +0000)]
gcc:
2018-08-30 Jose E. Marchesi <jose.marchesi@oracle.com>
* cfg.h (class auto_edge_flag): Spell out the template-id of the
base class in the initializer list. This is a workaround for
building with older GCC.
(class auto_bb_flag): Likewise.
hjl [Thu, 30 Aug 2018 15:59:41 +0000 (15:59 +0000)]
x86: Move AESNI generation to Skylake and Goldmont
The instruction set first appeared with Westmere, but not all processors
in that and the next few generations have the instructions. According to
Wikipedia[1], the first generation in which all SKUs have AES
instructions are Skylake and Goldmont. I can't find any Skylake,
Kabylake, Kabylake-R or Cannon Lake currently listed at
https://ark.intel.com that says "IntelĀ® AES New Instructions" "No".
redi [Thu, 30 Aug 2018 15:07:35 +0000 (15:07 +0000)]
Avoid undefined shifts in ceil2 operations
For values where the result cannot be represented the shift width would
be equal to the width of the type, which is undefined. Perform two
well-defined shifts instead of one possible undefined shift.
* include/bits/hashtable_policy.h (__clp2): Fix calculation for LLP64
targets where sizeof(size_t) > sizeof(long). Avoid undefined shifts
of the number of bits in the type.
* include/std/bit (__ceil2): Avoid undefined shifts.
* testsuite/26_numerics/bit/bit.pow.two/ceil2.cc: Test values with
the most signifiant bit set.
hubicka [Thu, 30 Aug 2018 15:05:38 +0000 (15:05 +0000)]
* lto-streamer-out.c (DFS::DFS_write_tree_body): Do not follow
TYPE_STUB_DECL.
(hash_tree): Do not visit TYPE_STUB_DECL.
* tree-streamer-out.c (write_ts_type_common_tree_pointers): Do not
stream TYPE_STUB_DECL.
* tree-streamer-in.c (lto_input_ts_type_common_tree_pointers): Likewise.
* ipa-utils.h (type_with_linkage_p): Do not rely on TYPE_STUB_DECL
after free_lang_data.
(type_in_anonymous_namespace_p): Likewise.
hubicka [Thu, 30 Aug 2018 12:58:42 +0000 (12:58 +0000)]
* sreal.h (SREAL_PART_BITS): Change to 31; remove seemingly unnecessary
comment that it has to be even number.
(class sreal): Change m_sig type to int32_t.
* sreal.c (sreal::dump, sreal::to_int, opreator+, operator-): Use
int64_t for temporary calculations.
(sreal_verify_basics): Drop one bit from minimum and maximum.
rguenth [Thu, 30 Aug 2018 12:37:10 +0000 (12:37 +0000)]
2018-08-30 Richard Biener <rguenther@suse.de>
PR tree-optimization/87147
* tree-ssa-sccvn.c (SSA_VISITED): New function.
(visit_phi): When the degenerate result is from the backedge and
we didn't visit its definition yet drop to VARYING.
(do_rpo_vn): Properly mark blocks with incoming backedges as executable.