| blob | c70fb93be346d2c3e38c8b79c0985c23665aa616 |
1 2017-01-27 David Malcolm <dmalcolm@redhat.com>
3 * doc/sourcebuild.texi (Testsuites): Add "GIMPLE Tests" and
4 "RTL Tests" to menu.
5 (GIMPLE Tests): New node.
6 (RTL Tests): New node.
8 2017-01-27 Richard Biener <rguenther@suse.de>
10 PR tree-optimization/79245
11 * tree-loop-distribution.c (distribute_loop): Apply cost
12 modeling also to detected patterns.
14 2017-01-27 Richard Biener <rguenther@suse.de>
16 PR tree-optimization/71433
17 * tree-vrp.c (register_new_assert_for): Revert earlier changes.
18 (compare_assert_loc): New function.
19 (process_assert_insertions): Sort and optimize assert locations
20 to remove duplicates and push down identical assertions on
21 edges to their destination block.
23 2017-01-27 Richard Biener <rguenther@suse.de>
25 PR tree-optimization/79244
26 * tree-vrp.c (remove_range_assertions): Forcefully propagate
27 out SSA names even if abnormal.
29 2017-01-27 Jakub Jelinek <jakub@redhat.com>
31 * realmpfr.h: Poison MPFR_RND{N,Z,U,D}.
32 * gimple-ssa-sprintf.c (format_floating_max): Use GMP_RNDN
33 instead of MPFR_RNDN.
35 2017-01-27 Richard Earnshaw <rearnsha@arm.com>
37 PR target/79239
38 * arm.c (arm_option_override): Don't call build_target_option_node
39 until after doing all option overrides.
40 (arm_valid_target_attribute_tree): Likewise.
42 2017-01-27 Martin Liska <mliska@suse.cz>
44 * doc/invoke.texi (-fprofile-arcs): Document profiling support
45 for {cd}tors and C++ {cd}tors.
47 2017-01-27 Dominik Vogt <vogt@linux.vnet.ibm.com>
49 * config/s390/s390.md ("*setmem_long_and")
50 ("*setmem_long_and_31z"): Use zero_extend instead of and.
52 2017-01-26 Martin Sebor <msebor@redhat.com>
54 * gimple-ssa-sprintf.c (format_floating): Simplify the computation
55 of precision.
57 2017-01-26 Martin Sebor <msebor@redhat.com>
59 * gimple-ssa-sprintf.c (format_floating): Test HAVE_XFmode and
60 HAVE_DFmode before using XFmode or DFmode.
61 (parse_directive): Avoid using the z length modifier to avoid
62 the ISO C++98 does not support the ‘z’ gnu_printf length modifier.
64 PR middle-end/78703
65 * gimple-ssa-sprintf.c (adjust_for_width_or_precision): Change
66 to accept adjustment as an array.
67 (get_int_range): New function.
68 (struct directive): Make width and prec arrays.
69 (directive::set_width, directive::set_precision): Call get_int_range.
70 (format_integer, format_floating): Handle width and precision ranges.
71 (format_string, parse_directive): Same.
73 2017-01-26 Jakub Jelinek <jakub@redhat.com>
75 PR debug/79129
76 * dwarf2out.c (generate_skeleton_bottom_up): For children with
77 comdat_type_p set, just clone them, but keep the children in the
78 original DIE.
80 PR debug/78835
81 * dwarf2out.c (prune_unused_types): Mark all functions with DIEs
82 which have direct callers with -fvar-tracking-assignments enabled
83 in the current TU.
84 (resolve_addr): Avoid adding skeleton DIEs for DW_AT_call_origin
85 inside of type units.
87 2017-01-26 Martin Sebor <msebor@redhat.com>
89 PR middle-end/78703
90 * gimple-ssa-sprintf.c (struct result_range): Add likely and
91 unlikely counters.
92 (struct format_result): Replace number_chars, number_chars_min,
93 and number_chars_max with a single member of struct result_range.
94 Remove bounded.
95 (format_result::operator+=): Adjust.
96 (struct fmtresult): Remove bounded. Handle likely and unlikely
97 counters.
98 (fmtresult::adjust_for_width_or_precision): New function.
99 (fmtresult:type_max_digits): New function.
100 (bytes_remaining): Handle likely and unlikely counters.
101 (min_bytes_remaining): Remove.
102 (format_percent): Simplify.
103 (format_integer, format_floating): Set likely and unlikely counters.
104 (get_string_length, format_character, format_string): Same.
105 (format_plain, should_warn_p): New function.
106 (maybe_warn): Call should_warn_p. Update diagnostic messages
107 and handle those for all directives, including plain strings.
108 (format_directive): Handle likely and unlikely counters.
109 Remove unnecessary quoting from diagnostics. Add an informational
110 note.
111 (add_bytes): Remove.
112 (pass_sprintf_length::compute_format_length): Simplify.
113 (try_substitute_return_value): Handle likely and unlikely counters.
115 2017-01-26 Carl Love <cel@us.ibm.com>
117 * config/rs6000/rs6000-c (altivec_overloaded_builtins): Remove
118 bogus entries for the P8V_BUILTIN_VEC_VGBBD built-ins
120 2017-01-26 Vladimir Makarov <vmakarov@redhat.com>
122 PR target/79131
123 * lra-assigns.c (setup_live_pseudos_and_spill_after_risky): Take
124 endianess for subregs into account.
125 * lra-constraints.c (lra_constraints): Do risky transformations
126 always on the first iteration.
127 * lra-lives.c (check_pseudos_live_through_calls): Add arg
128 last_call_used_reg_set.
129 (process_bb_lives): Define and use last_call_used_reg_set.
130 * lra.c (lra): Always continue after lra_constraints on the first
131 iteration.
133 2017-01-26 Kirill Yukhin <kirill.yukhin@gmail.com>
135 * gcc.target/i386/avx512bw-kshiftlq-2.c: Use unsigned long long
136 constant.
137 * gcc.target/i386/avx512bw-kshiftrq-2.c: Ditto.
139 2017-01-26 Jakub Jelinek <jakub@redhat.com>
141 * config/i386/avx512fintrin.h (_ktest_mask16_u8,
142 _ktestz_mask16_u8, _ktestc_mask16_u8, _kadd_mask16): Move to ...
143 * config/i386/avx512dqintrin.h (_ktest_mask16_u8,
144 _ktestz_mask16_u8, _ktestc_mask16_u8, _kadd_mask16): ... here.
145 * config/i386/i386-builtin.def (__builtin_ia32_ktestchi,
146 __builtin_ia32_ktestzhi, __builtin_ia32_kaddhi): Use
147 OPTION_MASK_ISA_AVX512DQ instead of OPTION_MASK_ISA_AVX512F.
148 * config/i386/sse.md (SWI1248_AVX512BWDQ2): New mode iterator.
149 (kadd<mode>, ktest<mode>): Use it instead of SWI1248_AVX512BWDQ.
151 2017-01-26 Marek Polacek <polacek@redhat.com>
153 PR c/79199
154 * fold-const.c (operand_equal_p) [COND_EXPR]: Use OP_SAME_WITH_NULL
155 for the third operand.
157 2017-01-26 Jakub Jelinek <jakub@redhat.com>
159 PR middle-end/79236
160 * omp-low.c (struct omp_context): Add simt_stmt field.
161 (scan_omp_for): Return omp_context *.
162 (scan_omp_simd): Set simt_stmt on the non-_simt_ SIMD
163 context to the _simt_ SIMD stmt.
164 (lower_omp_for): For combined SIMD with sibling _simt_
165 SIMD, make sure to use the same decls in _looptemp_
166 clauses as in the sibling.
168 2017-01-26 David Sherwood <david.sherwood@arm.com>
170 PR middle-end/79212
171 * gimplify.c (omp_notice_variable): Add GOVD_SEEN flag to variables in
172 all contexts.
174 2017-01-26 Jakub Jelinek <jakub@redhat.com>
176 PR target/70465
177 * reg-stack.c (emit_swap_insn): Instead of fld a; fld b; fxchg %st(1);
178 emit fld b; fld a; if possible.
180 * brig-builtins.def: Update copyright years.
181 * config/arm/arm_acle_builtins.def: Update copyright years.
183 2017-01-25 Michael Meissner <meissner@linux.vnet.ibm.com>
185 PR target/79179
186 * config/rs6000/vsx.md (vsx_extract_<mode>_store): Use wY
187 constraint instead of o for the stxsd instruction.
189 2017-01-25 Carl Love <cel@us.ibm.com>
191 * config/rs6000/rs6000-c (altivec_overloaded_builtins): Fix order
192 of entries for ALTIVEC_BUILTIN_VEC_PACKS and P8V_BUILTIN_VEC_VGBBD.
194 2017-01-25 Jonathan Wakely <jwakely@redhat.com>
196 * doc/invoke.texi (C++ Dialect Options): Fix typo.
198 2017-01-25 Richard Biener <rguenther@suse.de>
200 PR tree-optimization/69264
201 * target.def (vector_alignment_reachable): Improve documentation.
202 * doc/tm.texi: Regenerate.
203 * targhooks.c (default_builtin_vector_alignment_reachable): Simplify
204 and add a comment.
205 * tree-vect-data-refs.c (vect_supportable_dr_alignment): Revert
206 earlier changes with respect to TYPE_USER_ALIGN.
207 (vector_alignment_reachable_p): Likewise. Improve dumping.
209 2016-01-25 Kyrylo Tkachov <kyrylo.tkachov@arm.com>
211 PR target/79145
212 * config/arm/arm.md (xordi3): Force constant operand into a register
213 for TARGET_IWMMXT.
215 2016-01-25 Kyrylo Tkachov <kyrylo.tkachov@arm.com>
217 * doc/invoke.texi (-fstore-merging): Correct default optimization
218 levels at which it is enabled.
219 (-O): Move -fstore-merging from list to...
220 (-O2): ... Here.
222 2017-01-25 Richard Biener <rguenther@suse.de>
224 PR debug/78363
225 * omp-expand.c: Include debug.h.
226 (expand_omp_taskreg): Make sure to generate early debug before
227 outlining anything from a function.
228 (expand_omp_target): Likewise.
229 (grid_expand_target_grid_body): Likewise.
231 2017-01-25 Maxim Ostapenko <m.ostapenko@samsung.com>
233 PR lto/79061
234 * asan.c (get_translation_unit_decl): New function.
235 (asan_add_global): Extract modules file name from globals
236 TRANSLATION_UNIT_DECL name.
238 2017-01-24 Eric Botcazou <ebotcazou@adacore.com>
240 PR target/77439
241 * config/arm/arm.c (arm_function_ok_for_sibcall): Add back restriction
242 for long calls with APCS frame and VFP.
244 2017-01-24 David Malcolm <dmalcolm@redhat.com>
246 * cfg.c (original_copy_tables_initialized_p): New function.
247 * cfg.h (original_copy_tables_initialized_p): New decl.
248 * cfgrtl.c (relink_block_chain): Guard the call to
249 free_original_copy_tables with a call to
250 original_copy_tables_initialized_p.
251 * cgraph.h (symtab_node::native_rtl_p): New decl.
252 * cgraphunit.c (symtab_node::native_rtl_p): New function.
253 (symtab_node::needed_p): Don't assert for early assembly output
254 for __RTL functions.
255 (cgraph_node::finalize_function): Set "force_output" for __RTL
256 functions.
257 (cgraph_node::analyze): Bail out early for __RTL functions.
258 (analyze_functions): Update assertion to support __RTL functions.
259 (cgraph_node::expand): Bail out early for __RTL functions.
260 * final.c (rest_of_clean_state): Don't call delete_tree_ssa for
261 __RTL functions.
262 * function.h (struct function): Update comment for field
263 "pass_startwith".
264 * gimple-expr.c: Include "tree-pass.h".
265 (gimple_has_body_p): Return false for __RTL functions.
266 * Makefile.in (OBJS): Add run-rtl-passes.o.
267 * pass_manager.h (gcc::pass_manager::get_rest_of_compilation): New
268 accessor.
269 (gcc::pass_manager::get_clean_slate): New accessor.
270 * passes.c: Include "insn-addr.h".
271 (should_skip_pass_p): Add logging. Update logic for running
272 "expand" to be compatible with both __GIMPLE and __RTL. Guard
273 property-provider override so it is only done for gimple passes.
274 Don't skip dfinit.
275 (skip_pass): New function.
276 (execute_one_pass): Call skip_pass when skipping passes.
277 * read-md.c (md_reader::read_char): Support filtering
278 the input to a subset of line numbers.
279 (md_reader::md_reader): Initialize fields
280 m_first_line and m_last_line.
281 (md_reader::read_file_fragment): New function.
282 * read-md.h (md_reader::read_file_fragment): New decl.
283 (md_reader::m_first_line): New field.
284 (md_reader::m_last_line): New field.
285 * read-rtl-function.c (function_reader::create_function): Only
286 create cfun if it doesn't already exist. Set PROP_rtl on cfun's
287 curr_properties. Set DECL_INITIAL to a dummy block.
288 (read_rtl_function_body_from_file_range): New function.
289 * read-rtl-function.h (read_rtl_function_body_from_file_range):
290 New decl.
291 * run-rtl-passes.c: New file.
292 * run-rtl-passes.h: New file.
294 2017-01-24 Jeff Law <law@redhat.com>
296 * config/microblaze/microblaze.h (ASM_FORMAT_PRIVATE_NAME): Increase
297 buffer size.
299 2017-01-24 Bin Cheng <bin.cheng@arm.com>
301 PR tree-optimization/79159
302 * tree-ssa-loop-niter.c (get_cst_init_from_scev): New function.
303 (record_nonwrapping_iv): Improve boundary using above function if no
304 value range information.
306 2017-01-24 Pekka Jääskeläinen <pekka@parmance.com>
307 Martin Jambor <mjambor@suse.cz>
309 * brig-builtins.def: New file.
310 * builtins.def (DEF_HSAIL_BUILTIN): New macro.
311 (DEF_HSAIL_ATOMIC_BUILTIN): Likewise.
312 (DEF_HSAIL_SAT_BUILTIN): Likewise.
313 (DEF_HSAIL_INTR_BUILTIN): Likewise.
314 (DEF_HSAIL_CVT_ZEROI_SAT_BUILTIN): Likewise.
315 * builtin-types.def (BT_INT8): New.
316 (BT_INT16): Likewise.
317 (BT_UINT8): Likewise.
318 (BT_UINT16): Likewise.
319 (BT_FN_ULONG): Likewise.
320 (BT_FN_UINT_INT): Likewise.
321 (BT_FN_UINT_ULONG): Likewise.
322 (BT_FN_UINT_LONG): Likewise.
323 (BT_FN_UINT_PTR): Likewise.
324 (BT_FN_ULONG_PTR): Likewise.
325 (BT_FN_INT8_FLOAT): Likewise.
326 (BT_FN_INT16_FLOAT): Likewise.
327 (BT_FN_UINT32_FLOAT): Likewise.
328 (BT_FN_UINT16_FLOAT): Likewise.
329 (BT_FN_UINT8_FLOAT): Likewise.
330 (BT_FN_UINT64_FLOAT): Likewise.
331 (BT_FN_UINT16_UINT32): Likewise.
332 (BT_FN_UINT32_UINT16): Likewise.
333 (BT_FN_UINT16_UINT16_UINT16): Likewise.
334 (BT_FN_INT_PTR_INT): Likewise.
335 (BT_FN_UINT_PTR_UINT): Likewise.
336 (BT_FN_LONG_PTR_LONG): Likewise.
337 (BT_FN_ULONG_PTR_ULONG): Likewise.
338 (BT_FN_VOID_UINT64_UINT64): Likewise.
339 (BT_FN_UINT8_UINT8_UINT8): Likewise.
340 (BT_FN_INT8_INT8_INT8): Likewise.
341 (BT_FN_INT16_INT16_INT16): Likewise.
342 (BT_FN_INT_INT_INT): Likewise.
343 (BT_FN_UINT_FLOAT_UINT): Likewise.
344 (BT_FN_FLOAT_UINT_UINT): Likewise.
345 (BT_FN_ULONG_UINT_UINT): Likewise.
346 (BT_FN_ULONG_UINT_PTR): Likewise.
347 (BT_FN_ULONG_ULONG_ULONG): Likewise.
348 (BT_FN_UINT_UINT_UINT): Likewise.
349 (BT_FN_VOID_UINT_PTR): Likewise.
350 (BT_FN_UINT_UINT_PTR: Likewise.
351 (BT_FN_UINT32_UINT64_PTR): Likewise.
352 (BT_FN_INT_INT_UINT_UINT): Likewise.
353 (BT_FN_UINT_UINT_UINT_UINT): Likewise.
354 (BT_FN_UINT_UINT_UINT_PTR): Likewise.
355 (BT_FN_UINT_ULONG_ULONG_UINT): Likewise.
356 (BT_FN_ULONG_ULONG_ULONG_ULONG): Likewise.
357 (BT_FN_LONG_LONG_UINT_UINT): Likewise.
358 (BT_FN_ULONG_ULONG_UINT_UINT): Likewise.
359 (BT_FN_VOID_UINT32_UINT64_PTR): Likewise.
360 (BT_FN_VOID_UINT32_UINT32_PTR): Likewise.
361 (BT_FN_UINT_UINT_UINT_UINT_UINT): Likewise.
362 (BT_FN_UINT_FLOAT_FLOAT_FLOAT_FLOAT): Likewise.
363 (BT_FN_ULONG_ULONG_ULONG_UINT_UINT): Likewise.
364 * doc/frontends.texi: List BRIG FE.
365 * doc/install.texi (Testing): Add BRIG tesring requirements.
366 * doc/invoke.texi (Overall Options): Mention BRIG.
367 * doc/standards.texi (Standards): Doucment BRIG HSA version.
369 2017-01-24 Richard Biener <rguenther@suse.de>
371 PR translation/79208
372 * ipa-devirt.c (odr_types_equivalent_p): Fix typo in diagnostic.
374 2017-01-24 Martin Jambor <mjambor@suse.cz>
376 PR bootstrap/79198
377 * ipa-prop.c (ipa_free_all_node_params): Call summary destructor.
378 * ipa-prop.c (ipa_node_params_t::insert): Initialize fields known_csts
379 and known_contexts.
381 2017-01-24 Aldy Hernandez <aldyh@redhat.com>
383 PR middle-end/79123
384 * gimple-ssa-warn-alloca.c (alloca_call_type): Make sure
385 casts from signed to unsigned really don't have a range.
387 2017-01-24 Markus Trippelsdorf <markus@trippelsdorf.de>
389 * gimple-ssa-sprintf.c (format_floating): Change MPFR_RNDx to
390 GMP_RNDx for compatiblity.
392 2017-01-24 Martin Liska <mliska@suse.cz>
394 PR bootstrap/79132
395 * tree-ssa-reassoc.c (rewrite_expr_tree_parallel): Insert assert
396 that would prevent us to call alloca with -1 as argument.
398 2017-01-24 Jakub Jelinek <jakub@redhat.com>
400 * dwarf2out.c (output_compilation_unit_header, output_file_names):
401 Avoid -Wformat-security warning.
403 2017-01-23 Andrew Pinski <apinski@cavium.com>
405 * config/aarch64/aarch64.c (thunderx2t99_addrcost_table): Improve
406 cost table.
408 2017-01-23 Martin Sebor <msebor@redhat.com>
410 PR middle-end/78703
411 * gimple-ssa-sprintf.c (warn_level): New global.
412 (format_integer): Use it here and throughout the rest of the file.
413 Use the same switch to compute sign as base.
414 (maybe_warn): New function.
415 (format_directive): Factor out warnings into maybe_warn.
416 Add debugging output. Use warn_level.
417 (add_bytes): Use warn_level.
418 (pass_sprintf_length::compute_format_length): Add debugging output.
419 (try_substitute_return_value): Same.
420 (pass_sprintf_length::handle_gimple_call): Set and use warn_level.
422 PR middle-end/78703
423 * gimple-ssa-sprintf.c (struct format_result): Remove constant member.
424 (struct fmtresult, format_integer, format_floating): Adjust.
425 (fmtresult::fmtresult): Set max correctly in two argument ctor.
426 (get_string_length, format_string,format_directive): Same.
427 (pass_sprintf_length::compute_format_length): Same.
428 (try_substitute_return_value): Simplify slightly.
430 PR middle-end/78703
431 * gimple-ssa-sprintf.c (pass_sprintf_length::gate): Adjust formatting.
432 (fmtresult::operator+=): Outlined.
433 (struct fmtresult): Add ctors.
434 (struct conversion_spec): Rename...
435 (struct directive): ...to this. Add and remove data members.
436 (directive::set_width, directive::set_precision): New functions.
437 (format_percent): Use fmtresult ctor.
438 (get_width_and_precision): Remove.
439 (format_integer): Make naming changes. Avoid computing width and
440 precision.
441 (format_floating): Same. Adjust indentation.
442 (format_character, format_none): New functions.
443 (format_string): Moved character handling to format_character.
444 (format_directive): Remove arguments, change return type.
445 (parse_directive): New function.
446 (pass_sprintf_length::compute_format_length): Move directive
447 parsing to parse_directive.
449 2017-01-23 Jakub Jelinek <jakub@redhat.com>
451 * tree.h (assign_assembler_name_if_neeeded): Rename to ...
452 (assign_assembler_name_if_needed): ... this.
453 * tree.c (assign_assembler_name_if_neeeded): Rename to ...
454 (assign_assembler_name_if_needed): ... this.
455 (free_lang_data_in_cgraph): Adjust callers.
456 * cgraphunit.c (cgraph_node::analyze): Likewise.
457 * omp-expand.c (expand_omp_taskreg, expand_omp_target):
458 Likewise.
460 2017-01-23 Richard Biener <rguenther@suse.de>
462 PR tree-optimization/79088
463 PR tree-optimization/79188
464 * tree-ssa-threadupdate.c (mark_threaded_blocks): Move code
465 resetting loop bounds after last path deletion. Reset loop
466 bounds of the target loop, make code match the comments.
467 * tree-ssa-threadbackwards.c (pass_early_thread_jumps::execute):
468 Make sure loops need no fixups.
470 2017-01-23 Kelvin Nilsen <kelvin@gcc.gnu.org>
472 * config/rs6000/rs6000-builtin.def (VSIEDPF): Add scalar insert
473 exponent support with double type for first argument.
474 * config/rs6000/rs6000-c.c (altivec_overloaded_builtins): Changed
475 type returned by __builtin_vec_extract_sig,
476 __builtin_vec_extract_sig_sp, and __builtin_vec_extract_sig_dp
477 functions from "vector int" to "vector unsigned int" or from
478 "vector long long int" to "vector unsigned long long int".
479 Changed type returned by __builtin_vec_extract_exp,
480 __builtin_vec_extract_exp_sp, and __builtin_vec_extract_exp_dp
481 functions from "vector int" to "vector unsigned int" or from
482 "vector long long int" to "vector unsigned long long int".
483 Changed return type of __builtin_vec_test_data_class,
484 __builtin_vec_test_data_class_sp, and
485 __builtin_vec_test_data_class_dp from "vector int" to
486 "vector bool int" or from "vector long long int" to "vector bool
487 long long int" and changed second argument type from "unsigned
488 int" to "int". Added new overloaded function forms "vector float
489 __builtin_vec_insert_exp (vector float, vector unsigned int)" and
490 "vector float __builtin_vec_insert_exp_sp (vector float, vector
491 unsigned int)" and "vector double __builtin_vec_insert_exp (vector
492 double, vector unsigned long long int)" and "vector double
493 __builtin_vec_insert_exp_dp (vector double, vector unsigned long
494 long int)". Changed return type of
495 __builtin_scalar_test_data_class and
496 __builtin_scalar_test_data_class_sp and
497 __builtin_scalar_test_data_class_dp from "unsigned int" to "bool
498 int" and changed second argument from "unsigned int" to "int".
499 Changed type returned by __builtin_scalar_test_neg,
500 __builtin_scalar_test_neg_sp, and __builtin_scalar_test_neg_dp
501 from "int" to "bool int". Added new overloaded function form
502 "double __builtin_scalar_insert_exp (double, unsigned long long int)".
503 * config/rs6000/vsx.md (xsiexpdpf): New insn for scalar insert
504 exponent double-precision with floating point first argument.
505 * doc/extend.texi (PowerPC AltiVec Built-in Functions): Adjust
506 documentation of scalar_test_data_class, scalar_test_neg,
507 scalar_extract_sig, scalar_extract_exp, scalar_insert_exp,
508 vector_extract_exp, vec_extract_sig, vec_insert_exp, and
509 vec_test_data_class built-in functions to reflect refinements in
510 their type signatures.
512 2017-01-23 Andreas Tobler <andreast@gcc.gnu.org>
514 * config/aarch64/aarch64.c (aarch64_elf_asm_constructor): Increase
515 size of buf.
516 (aarch64_elf_asm_destructor): Likewise.
518 2017-01-23 Bernd Schmidt <bschmidt@redhat.com>
520 PR rtl-optimization/78634
521 * config/i386/i386.c (ix86_max_noce_ifcvt_seq_cost): New function.
522 (TARGET_MAX_NOCE_IFCVT_SEQ_COST): Define.
523 * ifcvt.c (noce_try_cmove): Add missing cost check.
525 PR rtl-optimization/71724
526 * combine.c (if_then_else_cond): Look for situations where it is
527 beneficial to undo the work of one of the recursive calls.
529 2017-01-23 Bin Cheng <bin.cheng@arm.com>
531 PR tree-optimization/70754
532 * tree-predcom.c (stmt_combining_refs): New parameter INSERT_BEFORE.
533 (reassociate_to_the_same_stmt): New parameter INSERT_BEFORE. Insert
534 combined stmt before it if not NULL.
535 (combine_chains): Process refs reversely and compute dominance point
536 for root ref.
538 2017-01-23 Martin Liska <mliska@suse.cz>
540 PR tree-optimization/79196
541 * tree-ssa-strlen.c (fold_strstr_to_memcmp): Rename to ...
542 (fold_strstr_to_strncmp): ... this. Fold the pattern to strncmp
543 instead of memcmp.
544 (strlen_optimize_stmt): Call the renamed function.
546 2017-01-23 Michael Matz <matz@suse.de>
548 PR tree-optimization/78384
549 * tree-ssa-loop-split.c (patch_loop_exit): Use correct edge.
551 2017-01-23 Richard Biener <rguenther@suse.de>
553 PR tree-optimization/79186
554 * tree-vrp.c (register_new_assert_for): Make sure we've seen
555 both incoming edges before moving an assert.
557 2017-01-23 Martin Jambor <mjambor@suse.cz>
559 * ipa-prop.c (load_from_param_1): Removed.
560 (load_from_unmodified_param): Bits from load_from_param_1 put back
561 here.
562 (load_from_param): Removed.
563 (compute_complex_assign_jump_func): Removed stmt2 and just replaced it
564 with stmt. Reverted back to use of load_from_unmodified_param.
566 2017-01-23 Martin Jambor <mjambor@suse.cz>
568 PR ipa/79108
569 * ipa-prop.h (ipa_param_descriptor): Anotate with with GTY(()).
570 (ipa_node_params): Annotate with GTY((for_user)). Make descriptors
571 field a pointer to garbage collected vector, mark lattices and
572 ipcp_orig_node with GTY((skip)).
573 (ipa_get_param_count): Adjust to descriptors being a pointer.
574 (ipa_get_param): Likewise.
575 (ipa_get_type): Likewise.
576 (ipa_get_param_move_cost): Likewise.
577 (ipa_set_param_used): Likewise.
578 (ipa_get_controlled_uses): Likewise.
579 (ipa_set_controlled_uses): Likewise.
580 (ipa_is_param_used): Likewise.
581 (ipa_node_params_t): Move into garbage collector. New methods insert
582 and remove.
583 (ipa_node_params_sum): Annotate wth GTY(()).
584 (ipa_check_create_node_params): Adjust to ipa_node_params_sum being
585 garbage collected.
586 (ipa_load_from_parm_agg): Adjust declaration.
587 * ipa-icf.c (param_used_p): Adjust to descriptors being a pointer.
588 * ipa-profile.c (ipa_profile): Likewise.
589 * ipa-prop.c (ipa_get_param_decl_index_1): Likewise.
590 (ipa_populate_param_decls): Make descriptors parameter garbage
591 collected.
592 (ipa_dump_param): Adjust to descriptors being a pointer.
593 (ipa_alloc_node_params): Likewise.
594 (ipa_initialize_node_params): Likewise.
595 (load_from_param_1): Make descriptors parameter garbage collected.
596 (load_from_unmodified_param): Likewise.
597 (load_from_param): Likewise.
598 (ipa_load_from_parm_agg): Likewise.
599 (ipa_node_params::~ipa_node_params): Removed.
600 (ipa_free_all_node_params): Remove call to delete operator.
601 (ipa_node_params_t::insert): New.
602 (ipa_node_params_t::remove): Likewise.
603 (ipa_node_params_t::duplicate): Adjust to descriptors being a pointer,
604 copy known_csts and known_contexts vectors.
605 (ipa_read_node_info): Adjust to descriptors being a pointer.
606 (ipcp_modif_dom_walker): Make m_descriptors field garbage
607 collected.
608 (ipcp_transform_function): Make descriptors variable garbage
609 collected.
611 2017-01-23 Andrew Senkevich <andrew.senkevich@intel.com>
613 * config/i386/avx512bwintrin.h: Add k-mask test, kortest intrinsics.
614 * config/i386/avx512dqintrin.h: Ditto.
615 * config/i386/avx512fintrin.h: Ditto.
616 * gcc/config/i386/i386.c: Handle new builtins.
617 * config/i386/i386-builtin.def: Add new builtins.
618 * config/i386/sse.md (ktest<mode>, kortest<mode>): New.
619 (UNSPEC_KORTEST, UNSPEC_KTEST): New.
621 2017-01-23 Jakub Jelinek <jakub@redhat.com>
622 Martin Liska <mliska@suse.cz>
624 * asan.h: Define ASAN_USE_AFTER_SCOPE_ATTRIBUTE.
625 * asan.c (asan_expand_poison_ifn): Support stores and use
626 appropriate ASAN report function.
627 * internal-fn.c (expand_ASAN_POISON_USE): New function.
628 * internal-fn.def (ASAN_POISON_USE): Declare.
629 * tree-into-ssa.c (maybe_add_asan_poison_write): New function.
630 (maybe_register_def): Create ASAN_POISON_USE when sanitizing.
631 * tree-ssa-dce.c (eliminate_unnecessary_stmts): Remove
632 ASAN_POISON calls w/o LHS.
633 * tree-ssa.c (execute_update_addresses_taken): Create clobber
634 for ASAN_MARK (UNPOISON, &x, ...) in order to prevent usage of a LHS
635 from ASAN_MARK (POISON, &x, ...) coming to a PHI node.
636 * gimplify.c (asan_poison_variables): Add attribute
637 use_after_scope_memory to variables that really needs to live
638 in memory.
639 * tree-ssa.c (is_asan_mark_p): Do not rewrite into SSA when
640 having the attribute.
642 2017-01-23 Martin Liska <mliska@suse.cz>
644 * asan.c (create_asan_shadow_var): New function.
645 (asan_expand_poison_ifn): Likewise.
646 * asan.h (asan_expand_poison_ifn): New declaration.
647 * internal-fn.c (expand_ASAN_POISON): Likewise.
648 * internal-fn.def (ASAN_POISON): New builtin.
649 * sanopt.c (pass_sanopt::execute): Expand
650 asan_expand_poison_ifn.
651 * tree-inline.c (copy_decl_for_dup_finish): Make function
652 external.
653 * tree-inline.h (copy_decl_for_dup_finish): Likewise.
654 * tree-ssa.c (is_asan_mark_p): New function.
655 (execute_update_addresses_taken): Rewrite local variables
656 (identified just by use-after-scope as addressable) into SSA.
658 2017-01-22 Gerald Pfeifer <gerald@pfeifer.com>
660 * doc/install.texi (Specific): opensource.apple.com uses https
661 now. Remove trailing slash.
663 2017-01-22 Gerald Pfeifer <gerald@pfeifer.com>
665 * README.Portability: Remove note on an Irix compatibility issue.
667 2017-01-22 Dimitry Andric <dim@FreeBSD.org>
669 * gcov.c (INCLUDE_ALGORITHM): Define.
670 (INCLUDE_VECTOR): Define.
671 No longer include <vector> and <algorithm> directly.
673 2017-01-21 Gerald Pfeifer <gerald@pfeifer.com>
675 * doc/extend.texi (Thread-Local): Change www.akkadia.org reference
676 to https.
677 * doc/invoke.texi (Code Gen Options): Ditto.
679 2017-01-21 Jan Hubicka <hubicka@ucw.cz>
681 PR lto/78407
682 * cfg.c (update_bb_profile_for_threading): Fix updating of probablity.
684 2017-01-21 Bernd Schmidt <bschmidt@redhat.com>
686 rtl-optimization/79125
687 * cprop.c (local_cprop_pass): Handle cases where we make an
688 unconditional trap.
690 2017-01-20 Segher Boessenkool <segher@kernel.crashing.org>
692 PR target/61729
693 PR target/77850
694 * config/rs6000/rs6000.c (rs6000_gimplify_va_arg): Adjust address to
695 read from, for big endian.
697 2017-01-20 Jiong Wang <jiong.wang@arm.com>
699 * config/aarch64/aarch64-builtins.c (aarch64_init_builtins): Register
700 register pauth builtins for LP64 only.
702 2017-01-20 Marek Polacek <polacek@redhat.com>
704 PR c/79152
705 * gimplify.c (should_warn_for_implicit_fallthrough): Handle consecutive
706 non-case labels.
708 2017-01-20 Alexander Monakov <amonakov@ispras.ru>
710 * omp-expand.c (expand_omp_simd): Clear PROP_gimple_lomp_dev regardless
711 of safelen status.
712 * omp-offload.c (pass_omp_device_lower::gate): Use PROP_gimple_lomp_dev.
713 * passes.c (dump_properties): Handle PROP_gimple_lomp_dev.
714 * tree-inline.c (expand_call_inline): Propagate PROP_gimple_lomp_dev.
716 2017-01-20 Kyrylo Tkachov <kyrylo.tkachov@arm.com>
718 PR target/71270
719 * config/arm/arm.c (neon_valid_immediate): Reject vector constants
720 in big-endian mode when they are not a single duplicated value.
722 2017-01-20 Richard Biener <rguenther@suse.de>
724 * BASE-VER: Bump to 7.0.1.
726 2017-01-20 Alexander Monakov <amonakov@ispras.ru>
728 * omp-low.c (omplow_simd_context): New struct. Use it...
729 (lower_rec_simd_input_clauses): ...here and...
730 (lower_rec_input_clauses): ...here to hold common data. Adjust all
731 references to idx, lane, max_vf, is_simt.
733 2017-01-20 Graham Markall <graham.markall@embecosm.com>
735 * config/arc/arc.h (LINK_SPEC): Use arclinux_nps emulation when
736 mcpu=nps400.
738 2017-01-20 Martin Jambor <mjambor@suse.cz>
740 * hsa.h: Renaed to hsa-common.h. Adjusted a comment.
741 * hsa.c: Renaed to hsa-common.c. Change include of gt-hsa.h to
742 gt-hsa-common.h.
743 * Makefile.in (OBJS): Rename hsa.o to hsa-common.o.
744 (GTFILES): Rename hsa.c to hsa-common.c.
745 * hsa-brig.c: Change include of hsa.h to hsa-common.h.
746 * hsa-dump.c: Likewise.
747 * hsa-gen.c: Likewise.
748 * hsa-regalloc.c: Likewise.
749 * ipa-hsa.c: Likewise.
750 * omp-expand.c: Likewise.
751 * omp-low.c: Likewise.
752 * toplev.c: Likewise.
754 2017-01-20 Marek Polacek <polacek@redhat.com>
756 PR c/64279
757 * doc/invoke.texi: Document -Wduplicated-branches.
758 * fold-const.c (operand_equal_p): Handle MODIFY_EXPR, INIT_EXPR,
759 COMPOUND_EXPR, PREDECREMENT_EXPR, PREINCREMENT_EXPR,
760 POSTDECREMENT_EXPR, POSTINCREMENT_EXPR, CLEANUP_POINT_EXPR, EXPR_STMT,
761 STATEMENT_LIST, and RETURN_EXPR. For non-pure non-const functions
762 return 0 only when not OEP_LEXICOGRAPHIC.
763 (fold_build_cleanup_point_expr): Use the expression
764 location when building CLEANUP_POINT_EXPR.
765 * tree-core.h (enum operand_equal_flag): Add OEP_LEXICOGRAPHIC.
766 * tree.c (add_expr): Handle error_mark_node.
768 2017-01-20 Martin Liska <mliska@suse.cz>
770 PR lto/69188
771 * tree-profile.c (init_ic_make_global_vars): Do not call
772 finalize_decl.
773 (gimple_init_gcov_profiler): Likewise.
775 2017-01-20 Martin Liska <mliska@suse.cz>
777 PR ipa/71190
778 * cgraph.h (maybe_create_reference): Remove argument and
779 update comment.
780 * cgraphclones.c (cgraph_node::create_virtual_clone): Remove one
781 argument.
782 * ipa-cp.c (create_specialized_node): Likewise.
783 * symtab.c (symtab_node::maybe_create_reference): Handle
784 VAR_DECLs and ADDR_EXPRs and select ipa_ref_use type.
786 2017-01-20 Martin Liska <mliska@suse.cz>
788 * read-rtl-function.c (function_reader::create_function): Use
789 build_decl instread of build_decl_stat.
791 2017-01-20 Andrew Senkevich <andrew.senkevich@intel.com>
793 * config/i386/avx512bwintrin.h: Add k-mask registers shift intrinsics.
794 * config/i386/avx512dqintrin.h: Ditto.
795 * config/i386/avx512fintrin.h: Ditto.
796 * config/i386/i386-builtin-types.def: Add new types.
797 * gcc/config/i386/i386.c: Handle new types.
798 * config/i386/i386-builtin.def (__builtin_ia32_kshiftliqi)
799 (__builtin_ia32_kshiftlihi, __builtin_ia32_kshiftlisi)
800 (__builtin_ia32_kshiftlidi, __builtin_ia32_kshiftriqi)
801 (__builtin_ia32_kshiftrihi, __builtin_ia32_kshiftrisi)
802 (__builtin_ia32_kshiftridi): New.
803 * config/i386/sse.md (k<code><mode>): Rename *k<code><mode>.
805 2017-01-19 Segher Boessenkool <segher@kernel.crashing.org>
807 PR target/78875
808 PR target/79140
809 * config/rs6000/rs6000.c (TARGET_STACK_PROTECT_GUARD): Unconditionally
810 define to rs6000_init_stack_protect_guard.
811 (rs6000_init_stack_protect_guard): New function.
813 2017-01-19 Matthew Fortune <matthew.fortune@imgtec.com>
814 Yunqiang Su <yunqiang.su@imgtec.com>
816 * config.gcc (supported_defaults): Add madd4.
817 (with_madd4): Add validation.
818 (all_defaults): Add madd4.
819 * config/mips/mips.opt (mmadd4): New option.
820 * gcc/config/mips/mips.h (OPTION_DEFAULT_SPECS): Add a default for
821 mmadd4.
822 (TARGET_CPU_CPP_BUILTINS): Add builtin_define for
823 __mips_no_madd4.
824 (ISA_HAS_UNFUSED_MADD4): Gate with mips_madd4.
825 (ISA_HAS_FUSED_MADD4): Likewise.
826 * gcc/doc/invoke.texi (-mmadd4): Document the new option.
827 * gcc/doc/install.texi (--with-madd4): Document the new option.
829 2017-01-19 Jiong Wang <jiong.wang@arm.com>
831 * config/aarch64/aarch64-builtins.c (enum aarch64_builtins): New
832 entries for AARCH64_PAUTH_BUILTIN_XPACLRI,
833 AARCH64_PAUTH_BUILTIN_PACIA1716, AARCH64_PAUTH_BUILTIN_AUTIA1716.
834 (aarch64_init_pauth_hint_builtins): New.
835 (aarch64_init_builtins): Call aarch64_init_pauth_hint_builtins.
836 (aarch64_expand_builtin): Expand new builtins.
838 2017-01-19 Jiong Wang <jiong.wang@arm.com>
840 * reg-notes.def (CFA_TOGGLE_RA_MANGLE): New reg-note.
841 * combine-stack-adj.c (no_unhandled_cfa): Handle
842 REG_CFA_TOGGLE_RA_MANGLE.
843 * dwarf2cfi.c (dwarf2out_frame_debug): Handle REG_CFA_TOGGLE_RA_MANGLE.
844 * config/aarch64/aarch64.c (aarch64_expand_prologue): Generates DWARF
845 info for return address signing.
846 (aarch64_expand_epilogue): Likewise.
848 2017-01-19 Jiong Wang <jiong.wang@arm.com>
850 * config/aarch64/aarch64-opts.h (aarch64_function_type): New enum.
851 * config/aarch64/aarch64-protos.h
852 (aarch64_return_address_signing_enabled): New declaration.
853 * config/aarch64/aarch64.c (aarch64_return_address_signing_enabled):
854 New function.
855 (aarch64_expand_prologue): Sign return address before it's pushed onto
856 stack.
857 (aarch64_expand_epilogue): Authenticate return address fetched from
858 stack.
859 (aarch64_override_options): Sanity check for ILP32 and ISA level.
860 (aarch64_attributes): New function attributes for "sign-return-address".
861 * config/aarch64/aarch64.md (UNSPEC_AUTI1716, UNSPEC_AUTISP,
862 UNSPEC_PACI1716, UNSPEC_PACISP, UNSPEC_XPACLRI): New unspecs.
863 ("*do_return"): Generate combined instructions according to key index.
864 ("<pauth_mnem_prefix>sp", "<pauth_mnem_prefix1716", "xpaclri"): New.
865 * config/aarch64/iterators.md (PAUTH_LR_SP, PAUTH_17_16): New integer
866 iterators.
867 (pauth_mnem_prefix, pauth_hint_num_a): New integer attributes.
868 * config/aarch64/aarch64.opt (msign-return-address=): New.
869 * doc/extend.texi (AArch64 Function Attributes): Documents
870 "sign-return-address=".
871 * doc/invoke.texi (AArch64 Options): Documents "-msign-return-address=".
873 2017-01-19 Matthew Fortune <matthew.fortune@imgtec.com>
875 * doc/invoke.texi: Add missing -mlxc1-sxc1 options to
876 overall option summary.
878 2017-01-19 Jiong Wang <jiong.wang@arm.com>
880 * config/aarch64/aarch64-arches.def: New entry for "armv8.3-a".
881 * config/aarch64/aarch64.h (AARCH64_FL_V8_3, AARCH64_FL_FOR_ARCH8_3,
882 AARCH64_ISA_V8_3, TARGET_ARMV8_3): New.
883 * doc/invoke.texi (AArch64 Options): Document "armv8.3-a".
885 2017-01-19 Michael Meissner <meissner@linux.vnet.ibm.com>
887 * config/rs6000/rs6000-cpus.def (ISA_3_0_MASKS_SERVER): Enable
888 -mpower9-minmax by default for -mcpu=power9.
889 (ISA_3_MASKS_IEEE): Require -mvsx-small-integer to enable IEEE
890 128-bit floating point.
892 2017-01-20 Alan Modra <amodra@gmail.com>
894 * config/rs6000/rs6000.md (cmpstrnsi, cmpstrsi): Fail if
895 optimizing for size.
897 2017-01-20 Alan Modra <amodra@gmail.com>
899 PR target/79144
900 * config/rs6000/rs6000.c (expand_strn_compare): Get the asm name
901 for strcmp and strncmp from corresponding builtin decl.
903 2017-01-19 Uros Bizjak <ubizjak@gmail.com>
905 * config.gcc (x86_64-*-rtems*): Use i386/rtemself.h
906 instead of i386/rtems-64.h.
907 * config/i386/rtems-64.h: Remove.
909 2017-01-19 Uros Bizjak <ubizjak@gmail.com>
911 PR target/78478
912 Revert:
913 2013-11-05 Uros Bizjak <ubizjak@gmail.com>
915 * config/i386/rtemself.h (LONG_DOUBLE_TYPE_SIZE): New define.
917 2017-01-19 Tamar Christina <tamar.christina@arm.com>
919 * config/aarch64/aarch64.c (aarch64_simd_gen_const_vector_dup):
920 Change int to HOST_WIDE_INT.
921 * config/aarch64/aarch64-protos.h
922 (aarch64_simd_gen_const_vector_dup): Likewise.
923 * config/aarch64/aarch64-simd.md: Add copysign<mode>3.
925 2017-01-19 David Malcolm <dmalcolm@redhat.com>
927 * langhooks-def.h (lhd_type_for_size): New decl.
928 (LANG_HOOKS_TYPE_FOR_SIZE): Define as lhd_type_for_size.
929 * langhooks.c (lhd_type_for_size): New function, taken from
930 lto_type_for_size.
932 2017-01-19 Pat Haugen <pthaugen@us.ibm.com>
934 * config/rs6000/power9.md (power9-alu): Remove 'cmp' type and add
935 define_bypass for CR latency.
936 (power9-cracked-alu): Update bypass latency and remove power9-branch.
937 (power9-alu2): Add define_bypass for CR latency.
938 (power9-cmp): New.
939 (power9-mul): Update insn latency.
940 (power9-mul-compare): Update insn latency, bypass latency and remove
941 power9-branch.
943 2016-01-19 Kyrylo Tkachov <kyrylo.tkachov@arm.com>
945 * config/aarch64/aarch64-protos.h (aarch64_nopcrelative_literal_loads):
946 Delete.
947 * config/aarch64/aarch64.md
948 (aarch64_reload_movcp<GPF_TF:mode><P:mode>): Delete reference to
949 aarch64_nopcrelative_literal_loads.
950 (aarch64_reload_movcp<VALL:mode><P:mode>): Likewise.
952 2017-01-19 Chenghua Xu <paul.hua.gm@gmail.com>
954 * config/mips/mips.h (ISA_HAS_FUSED_MADD4): Enable for
955 TARGET_LOONGSON_3A.
956 (ISA_HAS_UNFUSED_MADD4): Exclude TARGET_LOONGSON_3A.
958 2017-01-19 Doug Gilmore <doug.gilmore@imgtec.com>
960 PR target/78176
961 * config.gcc (supported_defaults): Add lxc1-sxc1.
962 (with_lxc1_sxc1): Add validation.
963 (all_defaults): Add lxc1-sxc1.
964 * config/mips/mips.opt (mlxc1-sxc1): New option.
965 * gcc/config/mips/mips.h (OPTION_DEFAULT_SPECS): Add a default for
966 mlxc1-sxc1.
967 (TARGET_CPU_CPP_BUILTINS): Add builtin_define for
968 __mips_no_lxc1_sxc1.
969 (ISA_HAS_LXC1_SXC1): Gate with mips_lxc1_sxc1.
970 * gcc/doc/invoke.texi (-mlxc1-sxc1): Document the new option.
971 * doc/install.texi (--with-lxc1-sxc1): Document the new option.
973 2017-01-19 Richard Biener <rguenther@suse.de>
975 PR tree-optimization/72488
976 * tree-ssa-sccvn.c (run_scc_vn): When we abort the VN make
977 sure to restore SSA info.
978 * tree-ssa.c (verify_ssa): Verify SSA info is not shared.
980 2017-01-19 Richard Earnshaw <rearnsha@arm.com>
982 PR rtl-optimization/79121
983 * expr.c (expand_expr_real_2, case LSHIFT_EXPR): Look at the signedness
984 of the inner type when shifting an extended value.
986 2017-01-17 Jan Hubicka <hubicka@ucw.cz>
988 PR lto/78407
989 * symtab.c (symtab_node::equal_address_to): Fix comparing of
990 interposable aliases.
992 2017-01-18 Peter Bergner <bergner@vnet.ibm.com>
994 PR target/78516
995 * config/rs6000/spe.md (mov_si<mode>_e500_subreg0): Fix constraints.
996 Use the evmergelohi instruction.
997 (mov_si<mode>_e500_subreg4_2_le): Likewise.
998 (mov_sitf_e500_subreg8_2_be): Likewise.
999 (mov_sitf_e500_subreg12_2_le): Likewise.
1000 (mov_si<mode>_e500_subreg0_2_le): Fix constraints.
1001 (mov_si<mode>_e500_subreg4_2_be): Likewise.
1002 (mov_sitf_e500_subreg8_2_le): Likewise.
1003 (mov_sitf_e500_subreg12_2_be): Likewise.
1005 2017-01-18 Bill Schmidt <wschmidt@linux.vnet.ibm.com>
1007 * config/rs6000/altivec.md (altivec_vbpermq): Change "type"
1008 attribute from vecsimple to vecperm.
1009 (altivec_vbpermq2): Likewise.
1011 2017-01-18 Bill Schmidt <wschmidt@linux.vnet.ibm.com>
1013 PR target/79040
1014 * config/rs6000/altivec.h: Fix typo of vec_cntlz to vec_cnttz.
1016 2017-01-18 Aaron Sawdey <acsawdey@linux.vnet.ibm.com>
1017 * config/rs6000/rs6000-protos.h (expand_strn_compare): Add arg.
1018 * config/rs6000/rs6000.c (expand_strn_compare): Add ability to expand
1019 strcmp. Fix bug where comparison didn't stop with zero byte. Fix
1020 case where N arg is SIZE_MAX.
1021 * config/rs6000/rs6000.md (cmpstrnsi): Args to expand_strn_compare.
1022 (cmpstrsi): Add pattern.
1024 2017-01-18 Michael Meissner <meissner@linux.vnet.ibm.com>
1026 * config/rs6000/rs6000-c.c (altivec_overloaded_builtins): Add
1027 __builtin_vec_revb builtins.
1028 * config/rs6000/rs6000-builtins.def (P9V_BUILTIN_XXBRQ_V16QI): Add
1029 built-in functions to support generation of the ISA 3.0 XXBR<x>
1030 vector byte reverse instructions.
1031 (P9V_BUILTIN_XXBRQ_V1TI): Likewise.
1032 (P9V_BUILTIN_XXBRD_V2DI): Likewise.
1033 (P9V_BUILTIN_XXBRD_V2DF): Likewise.
1034 (P9V_BUILTIN_XXBGW_V4SI): Likewise.
1035 (P9V_BUILTIN_XXBGW_V4SF): Likewise.
1036 (P9V_BUILTIN_XXBGH_V8HI): Likewise.
1037 (P9V_BUILTIN_VEC_REVB): Likewise.
1038 * config/rs6000/vsx.md (p9_xxbrq_v1ti): New insns/expanders to
1039 generate the ISA 3.0 XXBR<x> vector byte reverse instructions.
1040 (p9_xxbrq_v16qi): Likewise.
1041 (p9_xxbrd_<mode>, VSX_D iterator): Likewise.
1042 (p9_xxbrw_<mode>, VSX_W iterator): Likewise.
1043 (p9_xxbrh_v8hi): Likewise.
1044 * config/rs6000/altivec.h (vec_revb): Define if ISA 3.0.
1045 * doc/extend.texi (RS/6000 Altivec Built-ins): Document the
1046 vec_revb built-in functions.
1048 2017-01-18 Uros Bizjak <ubizjak@gmail.com>
1050 PR rtl-optimization/78952
1051 * config/i386/i386.md (any_extract): New code iterator.
1052 (*insvqi_2): Use any_extract for source operand.
1053 (*insvqi_3): Use any_shiftrt for source operand.
1055 2017-01-18 Wilco Dijkstra <wdijkstr@arm.com>
1057 * config/aarch64/aarch64.c (aarch64_sched_adjust_priority)
1058 New function.
1059 (TARGET_SCHED_ADJUST_PRIORITY): Define target hook.
1061 2017-01-18 Matthias Klose <doko@ubuntu.com>
1063 * doc/install.texi: Allow default for --with-target-bdw-gc-include.
1065 2016-01-18 Bill Schmidt <wschmidt@linux.vnet.ibm.com>
1067 * config/rs6000/altivec.h (vec_bperm): Change #define.
1068 * config/rs6000/altivec.md (UNSPEC_VBPERMD): New enum constant.
1069 (altivec_vbpermq2): New define_insn.
1070 (altivec_vbpermd): Likewise.
1071 * config/rs6000/rs6000-builtin.def (VBPERMQ2): New monomorphic
1072 function interface.
1073 (VBPERMD): Likewise.
1074 (VBPERM): New polymorphic function interface.
1075 * config/rs6000/r6000-c.c (altivec_overloaded_builtins_table):
1076 Add entries for P9V_BUILTIN_VEC_VBPERM.
1077 * doc/extend.texi: Add interfaces for vec_bperm.
1079 2017-01-18 Andreas Krebbel <krebbel@linux.vnet.ibm.com>
1081 * config/s390/s390-c.c (s390_expand_overloaded_builtin): Downcase
1082 first letter of error messages.
1083 (s390_resolve_overloaded_builtin): Likewise.
1084 * config/s390/s390.c (s390_expand_builtin): Likewise.
1085 (s390_invalid_arg_for_unprototyped_fn): Likewise.
1086 (s390_valid_target_attribute_inner_p): Likewise.
1087 * config/s390/s390.md ("tabort"): Likewise.
1089 2017-01-18 Toma Tabacu <toma.tabacu@imgtec.com>
1091 * config/mips/mips.h (ISA_HAS_DIV3): Remove unused macro.
1092 (ISA_AVOID_DIV_HILO): New macro.
1093 (ISA_HAS_DIV): Use new ISA_AVOID_DIV_HILO macro.
1094 (ISA_HAS_DDIV): Likewise.
1096 2017-01-18 Markus Trippelsdorf <markus@trippelsdorf.de>
1098 * doc/invoke.texi (fabi-version): Correct number of occurrences.
1100 2017-01-18 Markus Trippelsdorf <markus@trippelsdorf.de>
1102 * doc/invoke.texi (fabi-version): Spelling fix.
1104 2017-01-18 Markus Trippelsdorf <markus@trippelsdorf.de>
1106 PR c++/70182
1107 * doc/invoke.texi (fabi-version): Mention mangling fix for
1108 operator names.
1110 2017-01-18 Markus Trippelsdorf <markus@trippelsdorf.de>
1112 PR c++/77489
1113 * doc/invoke.texi (fabi-version): Document discriminator mangling.
1115 2017-01-17 Segher Boessenkool <segher@kernel.crashing.org>
1117 PR target/78875
1118 * config/rs6000/rs6000-opts.h (stack_protector_guard): New enum.
1119 * config/rs6000/rs6000.c (rs6000_option_override_internal): Handle
1120 the new options.
1121 * config/rs6000/rs6000.md (stack_protect_set): Handle the new more
1122 flexible settings.
1123 (stack_protect_test): Ditto.
1124 * config/rs6000/rs6000.opt (mstack-protector-guard=,
1125 mstack-protector-guard-reg=, mstack-protector-guard-offset=): New
1126 options.
1127 * doc/invoke.texi (Option Summary) [RS/6000 and PowerPC Options]:
1128 Add -mstack-protector-guard=, -mstack-protector-guard-reg=, and
1129 -mstack-protector-guard-offset=.
1130 (RS/6000 and PowerPC Options): Ditto.
1132 2017-01-17 Uros Bizjak <ubizjak@gmail.com>
1134 * config/i386/i386.h (MASK_CLASS_P): New define.
1135 * config/i386/i386.c (inline_secondary_memory_needed): Ensure that
1136 there are no registers from different register sets also when
1137 mask registers are used. Update function comment.
1138 * config/i386/i386.md (*movsi_internal): Split (*k/*krm) alternative
1139 to (*k/*r) and (*k/*km) alternatives.
1141 2017-01-17 Wilco Dijkstra <wdijkstr@arm.com>
1143 * config/aarch64/aarch64.md (eh_return): Remove pattern and splitter.
1144 * config/aarch64/aarch64.h (AARCH64_EH_STACKADJ_REGNUM): Remove.
1145 (EH_RETURN_HANDLER_RTX): New define.
1146 * config/aarch64/aarch64.c (aarch64_frame_pointer_required):
1147 Force frame pointer in EH return functions.
1148 (aarch64_expand_epilogue): Add barrier for eh_return.
1149 (aarch64_final_eh_return_addr): Remove.
1150 (aarch64_eh_return_handler_rtx): New function.
1151 * config/aarch64/aarch64-protos.h (aarch64_final_eh_return_addr):
1152 Remove.
1153 (aarch64_eh_return_handler_rtx): New prototype.
1155 2017-01-17 Bill Schmidt <wschmidt@linux.vnet.ibm.com>
1157 * config/rs6000/altivec.h (vec_rlmi): New #define.
1158 (vec_vrlnm): Likewise.
1159 (vec_rlnm): Likewise.
1160 * config/rs6000/altivec.md (UNSPEC_VRLMI): New UNSPEC enum value.
1161 (UNSPEC_VRLNM): Likewise.
1162 (VIlong): New mode iterator.
1163 (altivec_vrl<VI_char>mi): New define_insn.
1164 (altivec_vrl<VI_char>nm): Likewise.
1165 * config/rs6000/rs6000-builtin.def (VRLWNM): New monomorphic
1166 function entry.
1167 (VRLDNM): Likewise.
1168 (RLNM): New polymorphic function entry.
1169 (VRLWMI): New monomorphic function entry.
1170 (VRLDMI): Likewise.
1171 (RLMI): New polymorphic function entry.
1172 * config/rs6000/r6000-c.c (altivec_overloaded_builtin_table): Add
1173 new entries for P9V_BUILTIN_VEC_RLMI and P9V_BUILTIN_VEC_RLNM.
1174 * doc/extend.texi: Add description of vec_rlmi, vec_rlnm, and
1175 vec_vrlnm.
1177 2017-01-17 Jakub Jelinek <jakub@redhat.com>
1179 PR debug/78839
1180 * dwarf2out.c (field_byte_offset): Restore the
1181 PCC_BITFIELD_TYPE_MATTERS behavior for INTEGER_CST DECL_FIELD_OFFSET
1182 and DECL_FIELD_BIT_OFFSET. Use fold_build2 instead of build2 + fold.
1183 (analyze_variants_discr, gen_variant_part): Use fold_build2 instead
1184 of build2 + fold.
1186 2017-01-17 Eric Botcazou <ebotcazou@adacore.com>
1188 PR ada/67205
1189 * config/aarch64/aarch64.c (TARGET_CUSTOM_FUNCTION_DESCRIPTORS): Define
1191 2017-01-17 Jakub Jelinek <jakub@redhat.com>
1193 PR debug/71669
1194 * dwarf2out.c (add_data_member_location_attribute): For constant
1195 offset bitfield emit for -gdwarf-5 DW_AT_data_bit_offset attribute
1196 instead of DW_AT_data_member_location, DW_AT_bit_offset and
1197 DW_AT_byte_size attributes.
1199 2017-01-17 Eric Botcazou <ebotcazou@adacore.com>
1201 * config/rs6000/rs6000.c (rs6000_emit_move): Also use a TOC reference
1202 after forcing to constant memory when the code model is medium.
1204 2017-01-17 Julia Koval <julia.koval@intel.com>
1206 PR target/76731
1207 * config/i386/avx512fintrin.h
1208 (_mm512_i32gather_ps): Change __addr type to void const*.
1209 (_mm512_mask_i32gather_ps): Ditto.
1210 (_mm512_i32gather_pd): Ditto.
1211 (_mm512_mask_i32gather_pd): Ditto.
1212 (_mm512_i64gather_ps): Ditto.
1213 (_mm512_mask_i64gather_ps): Ditto.
1214 (_mm512_i64gather_pd): Ditto.
1215 (_mm512_mask_i64gather_pd): Ditto.
1216 (_mm512_i32gather_epi32): Ditto.
1217 (_mm512_mask_i32gather_epi32): Ditto.
1218 (_mm512_i32gather_epi64): Ditto.
1219 (_mm512_mask_i32gather_epi64): Ditto.
1220 (_mm512_i64gather_epi32): Ditto.
1221 (_mm512_mask_i64gather_epi32): Ditto.
1222 (_mm512_i64gather_epi64): Ditto.
1223 (_mm512_mask_i64gather_epi64): Ditto.
1224 (_mm512_i32scatter_ps): Change __addr type to void*.
1225 (_mm512_mask_i32scatter_ps): Ditto.
1226 (_mm512_i32scatter_pd): Ditto.
1227 (_mm512_mask_i32scatter_pd): Ditto.
1228 (_mm512_i64scatter_ps): Ditto.
1229 (_mm512_mask_i64scatter_ps): Ditto.
1230 (_mm512_i64scatter_pd): Ditto.
1231 (_mm512_mask_i64scatter_pd): Ditto.
1232 (_mm512_i32scatter_epi32): Ditto.
1233 (_mm512_mask_i32scatter_epi32): Ditto.
1234 (_mm512_i32scatter_epi64): Ditto.
1235 (_mm512_mask_i32scatter_epi64): Ditto.
1236 (_mm512_i64scatter_epi32): Ditto.
1237 (_mm512_mask_i64scatter_epi32): Ditto.
1238 (_mm512_i64scatter_epi64): Ditto.
1239 (_mm512_mask_i64scatter_epi64): Ditto.
1240 * config/i386/avx512pfintrin.h
1241 (_mm512_mask_prefetch_i32gather_pd): Change __addr type to void const*.
1242 (_mm512_mask_prefetch_i32gather_ps): Ditto.
1243 (_mm512_mask_prefetch_i64gather_pd): Ditto.
1244 (_mm512_mask_prefetch_i64gather_ps): Ditto.
1245 (_mm512_prefetch_i32scatter_pd): Change __addr type to void*.
1246 (_mm512_prefetch_i32scatter_ps): Ditto.
1247 (_mm512_mask_prefetch_i32scatter_pd): Ditto.
1248 (_mm512_mask_prefetch_i32scatter_ps): Ditto.
1249 (_mm512_prefetch_i64scatter_pd): Ditto.
1250 (_mm512_prefetch_i64scatter_ps): Ditto.
1251 (_mm512_mask_prefetch_i64scatter_pd): Ditto.
1252 (_mm512_mask_prefetch_i64scatter_ps): Ditto.
1253 * config/i386/avx512vlintrin.h
1254 (_mm256_mmask_i32gather_ps): Change __addr type to void const*.
1255 (_mm_mmask_i32gather_ps): Ditto.
1256 (_mm256_mmask_i32gather_pd): Ditto.
1257 (_mm_mmask_i32gather_pd): Ditto.
1258 (_mm256_mmask_i64gather_ps): Ditto.
1259 (_mm_mmask_i64gather_ps): Ditto.
1260 (_mm256_mmask_i64gather_pd): Ditto.
1261 (_mm_mmask_i64gather_pd): Ditto.
1262 (_mm256_mmask_i32gather_epi32): Ditto.
1263 (_mm_mmask_i32gather_epi32): Ditto.
1264 (_mm256_mmask_i32gather_epi64): Ditto.
1265 (_mm_mmask_i32gather_epi64): Ditto.
1266 (_mm256_mmask_i64gather_epi32): Ditto.
1267 (_mm_mmask_i64gather_epi32): Ditto.
1268 (_mm256_mmask_i64gather_epi64): Ditto.
1269 (_mm_mmask_i64gather_epi64): Ditto.
1270 (_mm256_i32scatter_ps): Change __addr type to void*.
1271 (_mm256_mask_i32scatter_ps): Ditto.
1272 (_mm_i32scatter_ps): Ditto.
1273 (_mm_mask_i32scatter_ps): Ditto.
1274 (_mm256_i32scatter_pd): Ditto.
1275 (_mm256_mask_i32scatter_pd): Ditto.
1276 (_mm_i32scatter_pd): Ditto.
1277 (_mm_mask_i32scatter_pd): Ditto.
1278 (_mm256_i64scatter_ps): Ditto.
1279 (_mm256_mask_i64scatter_ps): Ditto.
1280 (_mm_i64scatter_ps): Ditto.
1281 (_mm_mask_i64scatter_ps): Ditto.
1282 (_mm256_i64scatter_pd): Ditto.
1283 (_mm256_mask_i64scatter_pd): Ditto.
1284 (_mm_i64scatter_pd): Ditto.
1285 (_mm_mask_i64scatter_pd): Ditto.
1286 (_mm256_i32scatter_epi32): Ditto.
1287 (_mm256_mask_i32scatter_epi32): Ditto.
1288 (_mm_i32scatter_epi32): Ditto.
1289 (_mm_mask_i32scatter_epi32): Ditto.
1290 (_mm256_i32scatter_epi64): Ditto.
1291 (_mm256_mask_i32scatter_epi64): Ditto.
1292 (_mm_i32scatter_epi64): Ditto.
1293 (_mm_mask_i32scatter_epi64): Ditto.
1294 (_mm256_i64scatter_epi32): Ditto.
1295 (_mm256_mask_i64scatter_epi32): Ditto.
1296 (_mm_i64scatter_epi32): Ditto.
1297 (_mm_mask_i64scatter_epi32): Ditto.
1298 (_mm256_i64scatter_epi64): Ditto.
1299 (_mm256_mask_i64scatter_epi64): Ditto.
1300 (_mm_i64scatter_epi64): Ditto.
1301 (_mm_mask_i64scatter_epi64): Ditto.
1302 * config/i386/i386-builtin-types.def (V16SF_V16SF_PCFLOAT_V16SI_HI_INT)
1303 (V8DF_V8DF_PCDOUBLE_V8SI_QI_INT, V8SF_V8SF_PCFLOAT_V8DI_QI_INT)
1304 (V8DF_V8DF_PCDOUBLE_V8DI_QI_INT, V16SI_V16SI_PCINT_V16SI_HI_INT)
1305 (V8DI_V8DI_PCINT64_V8SI_QI_INT, V8SI_V8SI_PCINT_V8DI_QI_INT)
1306 (V8DI_V8DI_PCINT64_V8DI_QI_INT, V2DF_V2DF_PCDOUBLE_V4SI_QI_INT)
1307 (V4DF_V4DF_PCDOUBLE_V4SI_QI_INT, V2DF_V2DF_PCDOUBLE_V2DI_QI_INT)
1308 (V4DF_V4DF_PCDOUBLE_V4DI_QI_INT, V4SF_V4SF_PCFLOAT_V4SI_QI_INT)
1309 (V8SF_V8SF_PCFLOAT_V8SI_QI_INT, V4SF_V4SF_PCFLOAT_V2DI_QI_INT)
1310 (V4SF_V4SF_PCFLOAT_V4DI_QI_INT, V2DI_V2DI_PCINT64_V4SI_QI_INT)
1311 (V4DI_V4DI_PCINT64_V4SI_QI_INT, V2DI_V2DI_PCINT64_V2DI_QI_INT)
1312 (V4DI_V4DI_PCINT64_V4DI_QI_INT, V4SI_V4SI_PCINT_V4SI_QI_INT)
1313 (V8SI_V8SI_PCINT_V8SI_QI_INT, V4SI_V4SI_PCINT_V2DI_QI_INT)
1314 (V4SI_V4SI_PCINT_V4DI_QI_INT, VOID_PFLOAT_HI_V16SI_V16SF_INT)
1315 (VOID_PFLOAT_QI_V8SI_V8SF_INT, VOID_PFLOAT_QI_V4SI_V4SF_INT)
1316 (VOID_PDOUBLE_QI_V8SI_V8DF_INT, VOID_PDOUBLE_QI_V4SI_V4DF_INT)
1317 (VOID_PDOUBLE_QI_V4SI_V2DF_INT, VOID_PFLOAT_QI_V8DI_V8SF_INT)
1318 (VOID_PFLOAT_QI_V4DI_V4SF_INT, VOID_PFLOAT_QI_V2DI_V4SF_INT)
1319 (VOID_PDOUBLE_QI_V8DI_V8DF_INT, VOID_PDOUBLE_QI_V4DI_V4DF_INT)
1320 (VOID_PDOUBLE_QI_V2DI_V2DF_INT, VOID_PINT_HI_V16SI_V16SI_INT)
1321 (VOID_PINT_QI_V8SI_V8SI_INT, VOID_PINT_QI_V4SI_V4SI_INT)
1322 (VOID_PLONGLONG_QI_V8SI_V8DI_INT, VOID_PLONGLONG_QI_V4SI_V4DI_INT)
1323 (VOID_PLONGLONG_QI_V4SI_V2DI_INT, VOID_PINT_QI_V8DI_V8SI_INT)
1324 (VOID_PINT_QI_V4DI_V4SI_INT, VOID_PINT_QI_V2DI_V4SI_INT)
1325 (VOID_PLONGLONG_QI_V8DI_V8DI_INT, VOID_QI_V8SI_PCINT64_INT_INT)
1326 (VOID_PLONGLONG_QI_V4DI_V4DI_INT, VOID_PLONGLONG_QI_V2DI_V2DI_INT)
1327 (VOID_HI_V16SI_PCINT_INT_INT, VOID_QI_V8DI_PCINT64_INT_INT)
1328 (VOID_QI_V8DI_PCINT_INT_INT): Remove.
1329 (V16SF_V16SF_PCVOID_V16SI_HI_INT, V8DF_V8DF_PCVOID_V8SI_QI_INT)
1330 (V8SF_V8SF_PCVOID_V8DI_QI_INT, V8DF_V8DF_PCVOID_V8DI_QI_INT)
1331 (V16SI_V16SI_PCVOID_V16SI_HI_INT, V8DI_V8DI_PCVOID_V8SI_QI_INT)
1332 (V8SI_V8SI_PCVOID_V8DI_QI_INT, V8DI_V8DI_PCVOID_V8DI_QI_INT)
1333 (VOID_PVOID_HI_V16SI_V16SF_INT, VOID_PVOID_QI_V8SI_V8DF_INT)
1334 (VOID_PVOID_QI_V8DI_V8SF_INT, VOID_PVOID_QI_V8DI_V8DF_INT)
1335 (VOID_PVOID_HI_V16SI_V16SI_INT, VOID_PVOID_QI_V8SI_V8DI_INT)
1336 (VOID_PVOID_QI_V8DI_V8SI_INT, VOID_PVOID_QI_V8DI_V8DI_INT)
1337 (V2DF_V2DF_PCVOID_V4SI_QI_INT, V4DF_V4DF_PCVOID_V4SI_QI_INT)
1338 (V2DF_V2DF_PCVOID_V2DI_QI_INT, V4DF_V4DF_PCVOID_V4DI_QI_INT
1339 (V4SF_V4SF_PCVOID_V4SI_QI_INT, V8SF_V8SF_PCVOID_V8SI_QI_INT)
1340 (V4SF_V4SF_PCVOID_V2DI_QI_INT, V4SF_V4SF_PCVOID_V4DI_QI_INT)
1341 (V2DI_V2DI_PCVOID_V4SI_QI_INT, V4DI_V4DI_PCVOID_V4SI_QI_INT)
1342 (V2DI_V2DI_PCVOID_V2DI_QI_INT, V4DI_V4DI_PCVOID_V4DI_QI_INT)
1343 (V4SI_V4SI_PCVOID_V4SI_QI_INT, V8SI_V8SI_PCVOID_V8SI_QI_INT)
1344 (V4SI_V4SI_PCVOID_V2DI_QI_INT, V4SI_V4SI_PCVOID_V4DI_QI_INT)
1345 (VOID_PVOID_QI_V8SI_V8SF_INT, VOID_PVOID_QI_V4SI_V4SF_INT)
1346 (VOID_PVOID_QI_V4SI_V4DF_INT, VOID_PVOID_QI_V4SI_V2DF_INT)
1347 (VOID_PVOID_QI_V4DI_V4SF_INT, VOID_PVOID_QI_V2DI_V4SF_INT)
1348 (VOID_PVOID_QI_V4DI_V4DF_INT, VOID_PVOID_QI_V2DI_V2DF_INT)
1349 (VOID_PVOID_QI_V8SI_V8SI_INT, VOID_PVOID_QI_V4SI_V4SI_INT)
1350 (VOID_PVOID_QI_V4SI_V4DI_INT, VOID_PVOID_QI_V4SI_V2DI_INT)
1351 (VOID_PVOID_QI_V4DI_V4SI_INT, VOID_PVOID_QI_V2DI_V4SI_INT)
1352 (VOID_PVOID_QI_V4DI_V4DI_INT, VOID_PVOID_QI_V2DI_V2DI_INT)
1353 (VOID_QI_V8SI_PCVOID_INT_INT, VOID_HI_V16SI_PCVOID_INT_INT)
1354 (VOID_QI_V8DI_PCVOID_INT_INT): Add.
1355 * config/i386/i386.c (ix86_init_mmx_sse_builtins): Adjust builtin
1356 definitions accordingly.
1358 2017-01-17 Kito Cheng <kito.cheng@gmail.com>
1359 Kuan-Lin Chen <kuanlinchentw@gmail.com>
1361 PR target/79079
1362 * internal-fn.c (expand_mul_overflow): Use convert_modes instead of
1363 gen_lowpart.
1365 2017-01-17 Vladimir Makarov <vmakarov@redhat.com>
1367 PR target/79058
1368 * ira-conflicts.c (ira_build_conflicts): Update total conflict
1369 hard regs for inner regno.
1371 2017-01-17 Martin Liska <mliska@suse.cz>
1373 PR ipa/71207
1374 * ipa-polymorphic-call.c (contains_type_p): Fix wrong
1375 assumption and add comment.
1377 2017-01-17 Nathan Sidwell <nathan@acm.org>
1379 * ipa-visibility.c (localize_node): New function, broken out of ...
1380 (function_and_variable_visibility): ... here. Call it.
1382 2017-01-17 Jan Hubicka <hubicka@ucw.cz>
1384 PR middle-end/77445
1385 * tree-ssa-threadupdate.c (remove_ctrl_stmt_and_useless_edges):
1386 correctly set frequency of oudgoing edge.
1387 (duplicate_thread_path): Fix profile updating.
1389 2017-01-17 Jakub Jelinek <jakub@redhat.com>
1391 PR other/79046
1392 * configure.ac: Add GCC_BASE_VER.
1393 * Makefile.in (version): Use @get_gcc_base_ver@ instead of cat to get
1394 version from BASE-VER file.
1395 (CFLAGS-gcc.o): Add -DBASEVER=$(BASEVER_s).
1396 (gcc.o): Depend on $(BASEVER).
1397 * common.opt (dumpfullversion): New option.
1398 * gcc.c (driver_handle_option): Handle OPT_dumpfullversion.
1399 * doc/invoke.texi: Document -dumpfullversion.
1400 * doc/install.texi: Document --with-gcc-major-version-only.
1401 * configure: Regenerated.
1403 2017-01-17 Richard Biener <rguenther@suse.de>
1405 PR tree-optimization/71433
1406 * tree-vrp.c (register_new_assert_for): Merge same asserts
1407 on all incoming edges.
1408 (process_assert_insertions_for): Handle insertions at the
1409 beginning of BBs.
1411 2017-01-17 Gerald Pfeifer <gerald@pfeifer.com>
1413 * config/i386/cygwin.h (LIBGCJ_SONAME): Remove.
1414 * config/i386/mingw32.h (LIBGCJ_SONAME): Remove.
1416 2017-01-17 Kaz Kojima <kkojima@gcc.gnu.org>
1418 PR target/78633
1419 * config/sh/sh.md (cmpeqsi_t+1): Call copy_rtx to avoid invalid
1420 RTL sharing.
1422 2017-01-17 Alan Modra <amodra@gmail.com>
1424 PR target/79066
1425 * config/rs6000/rs6000.md (elf_high, elf_low): Disable when pic.
1426 * config/rs6000/rs6000.c (rs6000_emit_allocate_stack): Don't allow
1427 symbolic stack limit when pic.
1429 2017-01-16 Martin Sebor <msebor@redhat.com>
1431 PR tree-optimization/78608
1432 * gimple-ssa-sprintf.c (tree_digits): Avoid negating TYPE_MIN.
1434 2017-01-16 Jeff Law <law@redhat.com>
1436 Revert:
1437 2016-12-02 Tadek Kijkowski <tkijkowski@gmail.com>
1438 * Makefile.in (PREPROCESSOR_DEFINES): Add a level of indirection
1439 for several include directories that may be relative to sysroot.
1440 * config/i386/x-mingw32 (gplus_includedir): Define.
1441 (gplus_tool_includedir, gplus_backward_include_dir): Likewise.
1442 (native_system_includedir): Likewise.
1443 * config/i386/mingw32.h (STANDARD_STARTFILE_PREFIX_1): Do not
1444 override if TARGET_SYSTEM_ROOT is defined.
1445 (NATIVE_SYSTEM_HEADER_DIR): Likewise.
1447 PR tree-optimization/79090
1448 PR tree-optimization/33562
1449 PR tree-optimization/61912
1450 PR tree-optimization/77485
1451 * tree-ssa-dse.c (compute_trims): Accept STMT argument. Dump STMT
1452 and computed trims into the dump file.
1454 2017-01-17 Uros Bizjak <ubizjak@gmail.com>
1456 * config/i386/i386.h (LIMIT_RELOAD_CLASS): Remove.
1458 2017-01-16 Jakub Jelinek <jakub@redhat.com>
1460 PR c/79089
1461 * gimplify.c (gimplify_init_constructor): If want_value and
1462 object == lhs, unshare lhs to avoid invalid tree sharing. Formatting
1463 fix.
1465 PR target/79080
1466 * loop-doloop.c (doloop_modify): Call unshare_all_rtl_in_chain on
1467 sequence. Formatting fixes.
1468 (doloop_optimize): Formatting fixes.
1470 PR driver/49726
1471 * gcc.c (debug_level_greater_than_spec_func): New function.
1472 (static_spec_functions): Add debug-level-gt spec function.
1473 (ASM_DEBUG_SPEC, cpp_options): Use %:debug-level-gt(0) instead of
1474 !g0.
1475 * config/darwin.h (DSYMUTIL_SPEC, ASM_DEBUG_SPEC): Likewise.
1476 * config/darwin9.h (DSYMUTIL_SPEC, ASM_DEBUG_SPEC): Likewise.
1477 * common.opt (g, gcoff, gdwarf, gdwarf-, ggdb, gno-pubnames,
1478 gpubnames, ggnu-pubnames, gno-record-gcc-switches,
1479 grecord-gcc-switches, gno-strict-dwarf, gstrict-dwarf, gstabs,
1480 gstabs+, gtoggle, gvms, gxcoff, gxcoff+): Add Driver flag.
1482 2017-01-16 Uros Bizjak <ubizjak@gmail.com>
1484 * config/i386/i386.h (HARD_REGNO_CALLER_SAVE_MODE): Apply HImode and
1485 QImode fixups to general and mask registers only.
1487 2017-01-16 Carl Love <cel@us.ibm.com>
1489 * config/rs6000/rs6000-c (altivec_overloaded_builtins): Add support
1490 for built-in functions
1491 vector signed char vec_nabs (vector signed char)
1492 vector signed short vec_nabs (vector signed short)
1493 vector signed int vec_nabs (vector signed int)
1494 vector signed long long vec_nabs (vector signed long long)
1495 vector float vec_nabs (vector float)
1496 vector double vec_nabs (vector double)
1497 * config/rs6000/rs6000-builtin.def: Add definitions for NABS functions
1498 and NABS overload.
1499 * config/rs6000/altivec.md: New define_expand nabs<mode>2 types
1500 * config/rs6000/altivec.h: New define for vec_nabs built-in function.
1501 * doc/extend.texi: Update the documentation file for the new built-in
1502 functions.
1504 2017-01-16 Martin Sebor <msebor@redhat.com>
1506 * gimple-ssa-sprintf.c (format_directive): Correct a typo in a warning
1507 message.
1509 2017-01-16 Bill Schmidt <wschmidt@linux.vnet.ibm.com>
1511 * config/rs6000/rs6000.c (rtx_is_swappable_p): Change
1512 UNSPEC_VSX__XXSPLTD to require special splat handling.
1514 2017-01-16 David Malcolm <dmalcolm@redhat.com>
1516 PR bootstrap/78616
1517 * system.h: Poison strndup.
1519 2017-01-16 Alan Modra <amodra@gmail.com>
1521 PR target/79098
1522 * config/rs6000/rs6000.c (rs6000_legitimate_combined_insn): Don't
1523 use a switch.
1525 2017-01-16 Georg-Johann Lay <avr@gjlay.de>
1527 * config/avr/avr.h (BRANCH_COST) [reload_completed]: Increase by 4.
1529 2017-01-15 Uros Bizjak <ubizjak@gmail.com>
1531 * config/i386/i386.c (ix86_legitimate_combined_insn): Do not
1532 call recog here. Assert that INSN_CODE (insn) is non-negative.
1534 2017-01-15 Segher Boessenkool <segher@kernel.crashing.org>
1536 PR target/72749
1537 * cfgrtl.c (rtl_split_edge): Also patch jump insns that jump to the
1538 fallthrough.
1539 * haifa-sched.c (dump_insn_stream): Don't crash if there is a label
1540 in the currently scheduled RTL fragment.
1542 2017-01-15 Segher Boessenkool <segher@kernel.crashing.org>
1544 PR rtl-optimization/78751
1545 * ifcvt.c (find_cond_trap): If we generated a non-existing insn,
1546 give up.
1548 2017-01-14 Jeff Law <law@redhat.com>
1550 PR tree-optimization/79090
1551 * tree-ssa-dse.c (valid_ao_ref_for_dse): Reject zero length and
1552 variable length stores.
1553 (compute_trims): Delete dead assignment to *trim_tail.
1554 (dse_dom_walker::dse_optimize_stmt): Optimize mem* calls with
1555 zero length.
1557 2017-01-14 Bernd Schmidt <bschmidt@redhat.com>
1559 PR rtl-optimization/78626
1560 PR rtl-optimization/78727
1561 * cprop.c (one_cprop_pass): Collect unconditional traps in the middle
1562 of a block, and split such blocks after everything else is finished.
1564 2017-01-14 Alan Modra <amodra@gmail.com>
1566 PR target/72749
1567 * combine.c (recog_for_combine_1): Set INSN_CODE before calling
1568 target legitimate_combined_insn.
1569 * config/rs6000/rs6000.c (TARGET_LEGITIMATE_COMBINED_INSN): Define.
1570 (rs6000_legitimate_combined_insn): New function.
1571 * config/rs6000/rs6000.md (UNSPEC_DOLOOP): Delete, and remove
1572 all uses.
1573 (ctr<mode>_internal3): Rename from *ctr<mode>_internal5.
1574 (ctr<mode>_internal4): Rename from *ctr<mode>_internal6.
1575 (ctr<mode>_internal1, ctr<mode>_internal2): Remove '*' from name.
1577 2017-01-14 Gerald Pfeifer <gerald@pfeifer.com>
1579 * doc/frontends.texi (G++ and GCC): Remove references to Java.
1581 2017-01-13 Jeff Law <law@redhat.com>
1583 PR tree-optimization/33562
1584 PR tree-optimization/61912
1585 PR tree-optimization/77485
1586 * tree-ssa-dse.c (delete_dead_call): Accept gsi rather than
1587 a statement.
1588 (delete_dead_assignment): Likewise.
1589 (dse_dom_walker::dse_optimize_stmt): Pass in the gsi rather than
1590 statement to delete_dead_call and delete_dead_assignment.
1592 2017-01-13 David Malcolm <dmalcolm@redhat.com>
1594 PR c/78304
1595 * substring-locations.c (format_warning_va): Strengthen case 1 so
1596 that both endpoints of the substring must be within the format
1597 range for just the substring to be printed.
1599 2017-01-13 Uros Bizjak <ubizjak@gmail.com>
1601 * config/i386/i386.opt (msgx): Use ix86_isa_flags2 variable.
1602 * config/i386/i386.c (ix86_target_string): Add missing options
1603 to isa_opts and reorder options by implied ISAs. Rename isa_opts2 to
1604 isa2_opts, ix86_flag_opts to flag2_opts, ix86_target_other to
1605 flags_other and ix86_target_other to flags2_other. Display unknown
1606 isa2 options.
1607 (ix86_valid_target_attribute_inner_p): Add missing options and
1608 reorder options by implied ISAs, as in ix86_target_string.
1610 2017-01-13 Richard Sandiford <richard.sandiford@arm.com>
1612 * hash-table.h (hash_table::too_empty_p): New function.
1613 (hash_table::expand): Use it.
1614 (hash_table::traverse): Likewise.
1615 (hash_table::empty_slot): Use sizeof (value_type) instead of
1616 sizeof (PTR) to convert bytes to elements. Shrink the table
1617 if the current size is excessive for the current number of
1618 elements.
1620 2017-01-13 Richard Sandiford <richard.sandiford@arm.com>
1622 * ira-costs.c (record_reg_classes): Break from the inner loop
1623 early once alt_fail is known to be true. Update outer loop
1624 handling accordingly.
1626 2017-01-13 Jeff Law <law@redhat.com>
1628 * tree-ssa-dse.c (decrement_count): New function.
1629 (increment_start_addr, maybe_trim_memstar_call): Likewise.
1630 (dse_dom_walker::optimize_stmt): Call maybe_trim_memstar_call directly
1631 when we know the partially dead statement is a mem* function.
1633 PR tree-optimization/61912
1634 PR tree-optimization/77485
1635 * tree-ssa-dse.c: Include expr.h.
1636 (maybe_trim_constructor_store): New function.
1637 (maybe_trim_partially_dead_store): Call maybe_trim_constructor_store.
1639 PR tree-optimization/33562
1640 PR tree-optimization/61912
1641 PR tree-optimization/77485
1642 * doc/invoke.texi: Document new dse-max-object-size param.
1643 * params.def (PARM_DSE_MAX_OBJECT_SIZE): New PARAM.
1644 * tree-ssa-dse.c: Include params.h.
1645 (dse_store_status): New enum.
1646 (initialize_ao_ref_for_dse): New, partially extracted from
1647 dse_optimize_stmt.
1648 (valid_ao_ref_for_dse, normalize_ref): New.
1649 (setup_live_bytes_from_ref, compute_trims): Likewise.
1650 (clear_bytes_written_by, maybe_trim_complex_store): Likewise.
1651 (maybe_trim_partially_dead_store): Likewise.
1652 (maybe_trim_complex_store): Likewise.
1653 (dse_classify_store): Renamed from dse_possibly_dead_store_p.
1654 Track what bytes live from the original store. Return tri-state
1655 for dead, partially dead or live.
1656 (dse_dom_walker): Add constructor, destructor and new private members.
1657 (delete_dead_call, delete_dead_assignment): New extracted from
1658 dse_optimize_stmt.
1659 (dse_optimize_stmt): Make a member of dse_dom_walker.
1660 Use initialize_ao_ref_for_dse.
1662 PR tree-optimization/33562
1663 PR tree-optimization/61912
1664 PR tree-optimization/77485
1665 * sbitmap.h (bitmap_count_bits): Prototype.
1666 (bitmap_clear_range, bitmap_set_range): Likewise.
1667 * sbitmap.c (bitmap_clear_range): New function.
1668 (bitmap_set_range, sbitmap_popcount, bitmap_count_bits): Likewise.
1670 2017-01-13 Martin Liska <mliska@suse.cz>
1672 PR ipa/79043
1673 * function.c (set_cfun): Add new argument force.
1674 * function.h (set_cfun): Likewise.
1675 * ipa-inline-transform.c (inline_call): Use the function when
1676 strict alising from is dropped for function we inline to.
1678 2017-01-13 Richard Biener <rguenther@suse.de>
1680 * tree-pretty-print.c (dump_generic_node): Fix inverted condition
1681 for dumping GIMPLE INTEGER_CSTs.
1683 2017-01-13 Rainer Orth <ro@CeBiTec.Uni-Bielefeld.DE>
1685 * config/sol2.h (TARGET_OS_CPP_BUILTINS): Define __STDC_VERSION__
1686 to 201112L since C++17.
1688 2017-01-13 Maxim Ostapenko <m.ostapenko@samsung.com>
1690 PR sanitizer/78887
1691 * asan.c (asan_needs_odr_indicator_p): Don't emit ODR indicators
1692 if -fsanitize=kernel-address is present.
1694 2017-01-13 Richard Biener <rguenther@suse.de>
1696 * tree-pretty-print.c (dump_generic_node): Dump INTEGER_CSTs
1697 as _Literal ( type ) number in case usual suffixes do not
1698 preserve all information.
1700 2017-01-13 Richard Biener <rguenther@suse.de>
1702 PR tree-optimization/77283
1703 * gimple-ssa-split-paths.c: Include gimple-ssa.h, tree-phinodes.h
1704 and ssa-iterators.h.
1705 (is_feasible_trace): Implement a cost model based on joiner
1706 PHI node uses.
1708 2017-01-12 Michael Meissner <meissner@linux.vnet.ibm.com>
1710 PR target/79004
1711 * config/rs6000/rs6000.md (FP_ISA3): Do not optimize converting
1712 char or short to __float128/_Float128 directly.
1714 2017-01-12 Martin Sebor <msebor@redhat.com>
1716 to -Wformat-overflow.
1717 * gimple-ssa-sprintf.c (pass_sprintf_length::gate): Adjust.
1718 (min_bytes_remaining): Same.
1719 (get_string_length): Same.
1720 (format_string): Same.
1721 (format_directive): Same.
1722 (add_bytes): Same.
1723 (pass_sprintf_length::handle_gimple_call): Same.
1725 2017-01-12 Jakub Jelinek <jakub@redhat.com>
1727 * gimple-ssa-sprintf.c (try_substitute_return_value): Remove
1728 info.nowrite calls with no lhs that can't throw. Return bool
1729 whether gsi_remove has been called or not.
1730 (pass_sprintf_length::handle_gimple_call): Return bool whether
1731 try_substitute_return_value called gsi_remove. Formatting fix.
1732 (pass_sprintf_length::execute): Don't use gsi_remove if
1733 handle_gimple_call returned true.
1735 PR bootstrap/79069
1736 * cfgrtl.c (rtl_tidy_fallthru_edge): For any_uncondjump_p that can't
1737 be removed due to side-effects, don't remove following barrier nor
1738 turn the successor edge into fallthru edge.
1740 2017-01-12 Bill Schmidt <wschmidt@linux.vnet.ibm.com>
1742 PR target/79044
1743 * config/rs6000/rs6000.c (insn_is_swappable_p): Mark
1744 element-reversing loads and stores as not swappable.
1746 2017-01-12 Nathan Sidwell <nathan@acm.org>
1747 Nicolai Stange <nicstange@gmail.com>
1749 * combine.c (try_combine): Don't ignore result of overlap checking
1750 loop. Combine overlap & asm check into single loop.
1752 2017-01-12 Richard Biener <rguenther@suse.de>
1754 * tree-pretty-print.c (dump_generic_node): Provide -gimple
1755 variant for MEM_REF. Sanitize INTEGER_CST for -gimple.
1757 2017-01-12 Richard Biener <rguenther@suse.de>
1759 * tree.c (initialize_tree_contains_struct): Make TS_OPTIMIZATION
1760 and TS_TARGET_OPTION directly derive from TS_BASE.
1761 * tree-core.h (tree_optimization_option): Derive from tree_base.
1762 (tree_target_option): Likewise.
1764 2017-01-11 Uros Bizjak <ubizjak@gmail.com>
1766 * config/i386/i386.c (memory_address_length): Increase len
1767 only when rip_relative_addr_p returns false.
1769 2017-01-11 Julia Koval <julia.koval@intel.com>
1771 * common/config/i386/i386-common.c (OPTION_MASK_ISA_SGX_UNSET): New.
1772 (OPTION_MASK_ISA_SGX_SET): New.
1773 (ix86_handle_option): Handle OPT_msgx.
1774 * config.gcc: Added sgxintrin.h.
1775 * config/i386/driver-i386.c (host_detect_local_cpu): Detect sgx.
1776 * config/i386/i386-c.c (ix86_target_macros_internal): Define __SGX__.
1777 * config/i386/i386.c (ix86_target_string): Add -msgx.
1778 (PTA_SGX): New.
1779 (ix86_option_override_internal): Handle new options.
1780 (ix86_valid_target_attribute_inner_p): Add sgx.
1781 * config/i386/i386.h (TARGET_SGX, TARGET_SGX_P): New.
1782 * config/i386/i386.opt: Add msgx.
1783 * config/i386/sgxintrin.h: New file.
1784 * config/i386/x86intrin.h: Add sgxintrin.h.
1786 2017-01-11 Jakub Jelinek <jakub@redhat.com>
1788 PR c++/71537
1789 * fold-const.c (maybe_nonzero_address): Return 1 for function
1790 local objects.
1791 (tree_single_nonzero_warnv_p): Don't handle function local objects
1792 here.
1794 PR c++/72813
1795 * gcc.c (default_compilers): Don't add -o %g.s for -S -save-temps
1796 of c-header.
1798 2017-01-11 David Malcolm <dmalcolm@redhat.com>
1800 PR driver/78877
1801 * opts.c: Include "spellcheck.h"
1802 (struct string_fragment): New struct.
1803 (struct edit_distance_traits<const string_fragment &>): New
1804 struct.
1805 (get_closest_sanitizer_option): New function.
1806 (parse_sanitizer_options): Offer suggestions for unrecognized arguments.
1808 2017-01-11 Jakub Jelinek <jakub@redhat.com>
1810 * dwarf2out.c (DWARF_COMPILE_UNIT_HEADER_SIZE): For DWARF5 decrease
1811 by 12.
1812 (DWARF_COMDAT_TYPE_UNIT_HEADER_SIZE): Always
1813 DWARF_COMPILE_UNIT_HEADER_SIZE plus 12.
1814 (DWARF_COMPILE_UNIT_SKELETON_HEADER_SIZE): Define.
1815 (calc_base_type_die_sizes): Use DWARF_COMPILE_UNIT_SKELETON_HEADER_SIZE
1816 for initial die_offset if dwarf_split_debug_info.
1817 (output_comp_unit): Use DWARF_COMPILE_UNIT_SKELETON_HEADER_SIZE for
1818 initial next_die_offset if dwo_id is non-NULL. Don't emit padding
1819 fields.
1820 (output_skeleton_debug_sections): Formatting fix. Use
1821 DWARF_COMPILE_UNIT_SKELETON_HEADER_SIZE instead of
1822 DWARF_COMPILE_UNIT_HEADER_SIZE. Don't emit padding.
1824 2017-01-11 Wilco Dijkstra <wdijkstr@arm.com>
1826 * config/arm/cortex-a53.md: Add bypasses for
1827 cortex_a53_r2f_cvt.
1828 (cortex_a53_r2f): Only use for transfers.
1829 (cortex_a53_f2r): Likewise.
1830 (cortex_a53_r2f_cvt): Add reservation for conversions.
1831 (cortex_a53_f2r_cvt): Likewise.
1833 2017-01-11 Tamar Christina <tamar.christina@arm.com>
1835 * config/arm/arm_neon.h: Add __artificial__ and gnu_inline
1836 to all inlined functions, change static to extern.
1838 2017-01-11 Christophe Lyon <christophe.lyon@linaro.org>
1840 PR target/78253
1841 * config/arm/arm.c (legitimize_pic_address): Handle reference to
1842 weak symbol.
1843 (arm_assemble_integer): Likewise.
1845 2017-01-11 Richard Earnshaw <rearnsha@arm.com>
1847 * config.gcc: Use new awk script to check CPU, FPU and architecture
1848 parameters for --with-... options.
1849 * config/arm/parsecpu.awk: New file
1850 * config/arm/arm-cpus.in: New file.
1851 * config/arm/arm-opts.h: Include arm-cpu.h instead of processing .def
1852 files.
1853 * config/arm/arm.c: Include arm-cpu-data.h instead of processing .def
1854 files.
1855 * config/arm/t-arm: Update dependency rules.
1856 * common/config/arm/arm-common.c: Include arm-cpu-cdata.h instead
1857 of processing .def files.
1858 * config/arm/genopt.sh: Deleted.
1859 * config/arm/gentune.sh: Deleted.
1860 * config/arm/arm-cores.def: Deleted.
1861 * config/arm/arm-arches.def: Deleted.
1862 * config/arm/arm-fpus.def: Deleted.
1863 * config/arm/arm-tune.md: Regenerated.
1864 * config/arm/arm-tables.opt: Regenerated.
1865 * config/arm/arm-cpu.h: New generated file.
1866 * config/arm/arm-cpu-data.h: New generated file.
1867 * config/arm/arm-cpu-cdata.h: New generated file.
1869 2017-01-11 Maxim Ostapenko <m.ostapenko@samsung.com>
1871 PR lto/79042
1872 * lto-cgraph.c (lto_output_varpool_node): Pack dynamically_initialized
1873 bit.
1874 (input_varpool_node): Unpack dynamically_initialized bit.
1876 2017-01-11 Eric Botcazou <ebotcazou@adacore.com>
1878 PR rtl-optimization/79032
1879 * lra-constraints.c (simplify_operand_subreg): In the MEM case, test
1880 the alignment of the adjusted memory reference against that of MODE,
1881 instead of the alignment of the original memory reference.
1883 2017-01-11 Martin Jambor <mjambor@suse.cz>
1885 * hsa.c (hsa_callable_function_p): Revert addition of DECL_ARTIFICIAL
1886 test.
1887 * ipa-hsa.c (process_hsa_functions): Only duplicate non-artificial
1888 decorated functions.
1890 2017-01-11 Richard Biener <rguenther@suse.de>
1892 * tree-vrp.c (evrp_dom_walker::before_dom_children): Also
1893 set range/nonnull info for PHI results. Do not set it on
1894 stmts marked for removal.
1896 2017-01-10 Eric Botcazou <ebotcazou@adacore.com>
1898 * expr.c (store_field): In the bitfield case, fetch the return value
1899 from the registers before applying a single big-endian adjustment.
1900 Always do a final load for a BLKmode value not larger than a word.
1902 2017-01-10 David Malcolm <dmalcolm@redhat.com>
1904 PR c++/77949
1905 * input.c (selftest::test_accessing_ordinary_linemaps): Verify
1906 that we correctly handle column numbers greater than
1907 LINE_MAP_MAX_COLUMN_NUMBER.
1909 2017-01-10 Martin Sebor <msebor@redhat.com>
1911 PR middle-end/78245
1912 * gimple-ssa-sprintf.c (get_destination_size): Call
1913 {init,fini}object_sizes.
1914 * tree-object-size.c (addr_object_size): Adjust.
1915 (pass_through_call): Adjust.
1916 (pass_object_sizes::execute): Adjust.
1917 * tree-object-size.h (fini_object_sizes): Declare.
1919 2017-01-10 Martin Sebor <msebor@redhat.com>
1921 PR tree-optimization/78775
1922 * builtins.c (get_size_range): Move...
1923 * calls.c: ...to here.
1924 (alloc_max_size): Accept zero argument.
1925 (operand_signed_p): Remove.
1926 (maybe_warn_alloc_args_overflow): Call get_size_range.
1927 * calls.h (get_size_range): Declare.
1929 2017-01-10 Joe Seymour <joe.s@somniumtech.com>
1931 * config/msp430/driver-msp430.c (msp430_mcu_data): Sync with data
1932 from TI's devices.csv file as of September 2016.
1933 * config/msp430/msp430.c (msp430_mcu_data): Likewise.
1935 2017-01-10 Sandra Loosemore <sandra@codesourcery.com>
1937 * doc/extend.texi: Tweak formatting to fix overfull hbox warnings.
1938 * doc/invoke.texi: Likewise.
1939 * doc/md.texi: Likewise.
1940 * doc/objc.texi: Likewise.
1942 2017-01-10 Joshua Conner <joshconner@google.com>
1944 * config/arm/fuchsia-elf.h: New file.
1945 * config/fuchsia.h: New file.
1946 * config.gcc (*-*-fuchsia*): Set native_system_header_dir.
1947 (aarch64*-*-fuchsia*, arm*-*-fuchsia*, x86_64-*-fuchsia*): Add to
1948 targets.
1949 * config.host: (aarch64*-*-fuchsia*, arm*-*-fuchsia*): Add to hosts.
1951 2016-01-10 Richard Biener <rguenther@suse.de>
1953 PR tree-optimization/79034
1954 * tree-call-cdce.c (shrink_wrap_one_built_in_call_with_conds):
1955 Propagate out degenerate PHIs in the joiner.
1957 2017-01-10 Martin Liska <mliska@suse.cz>
1959 * ipa-icf.c (sort_sem_items_by_decl_uid): New function.
1960 (sort_congruence_classes_by_decl_uid): Likewise.
1961 (sort_congruence_class_groups_by_decl_uid): Likewise.
1962 (sem_item_optimizer::merge_classes): Sort class, groups in these
1963 classes and members in the groups by DECL_UID of declarations.
1964 This would make merge operations stable.
1966 2017-01-10 Martin Liska <mliska@suse.cz>
1968 * ipa-icf.c (sem_item_optimizer::sem_item_optimizer): Remove
1969 usage of m_classes_vec.
1970 (sem_item_optimizer::~sem_item_optimizer): Likewise.
1971 (sem_item_optimizer::get_group_by_hash): Likewise.
1972 (sem_item_optimizer::subdivide_classes_by_equality): Likewise.
1973 (sem_item_optimizer::subdivide_classes_by_sensitive_refs): Likewise.
1974 (sem_item_optimizer::verify_classes): Likewise.
1975 (sem_item_optimizer::process_cong_reduction): Likewise.
1976 (sem_item_optimizer::dump_cong_classes): Likewise.
1977 (sem_item_optimizer::merge_classes): Likewise.
1978 * ipa-icf.h (congruence_class_hash): Rename from
1979 congruence_class_group_hash. Remove declaration of m_classes_vec.
1981 2017-01-10 Andrew Senkevich <andrew.senkevich@intel.com>
1983 * common/config/i386/i386-common.c (OPTION_MASK_ISA_AVX512VPOPCNTDQ_SET,
1984 OPTION_MASK_ISA_AVX512VPOPCNTDQ_UNSET): New.
1985 * config.gcc: Add avx512vpopcntdqintrin.h.
1986 * config/i386/avx512vpopcntdqintrin.h: New.
1987 * config/i386/cpuid.h (bit_AVX512VPOPCNTDQ): New.
1988 * config/i386/i386-builtin-types.def: Add new types.
1989 * config/i386/i386-builtin.def (__builtin_ia32_vpopcountd_v16si,
1990 __builtin_ia32_vpopcountd_v16si_mask, __builtin_ia32_vpopcountq_v8di,
1991 __builtin_ia32_vpopcountq_v8di_mask): New.
1992 * config/i386/i386-c.c (ix86_target_macros_internal): Define
1993 __AVX512VPOPCNTDQ__.
1994 * config/i386/i386.c (ix86_target_string): Add -mavx512vpopcntdq.
1995 (PTA_AVX512VPOPCNTDQ): Define.
1996 * config/i386/i386.h (TARGET_AVX512VPOPCNTDQ,
1997 TARGET_AVX512VPOPCNTDQ_P): Define.
1998 * config/i386/i386.opt: Add mavx512vpopcntdq.
1999 * config/i386/immintrin.h: Include avx512vpopcntdqintrin.h.
2000 * config/i386/sse.md (define_insn "vpopcount<mode><mask_name>"): New.
2002 2017-01-01 Jan Hubicka <hubicka@ucw.cz>
2004 PR middle-end/77484
2005 * predict.def (PRED_CALL): Set to 67.
2007 2017-01-09 Eric Botcazou <ebotcazou@adacore.com>
2009 * expr.c (store_field): In the bitfield case, if the value comes from
2010 a function call and is of an aggregate type returned in registers, do
2011 not modify the field mode; extract the value in all cases if the mode
2012 is BLKmode and the size is not larger than a word.
2014 2017-01-09 Dominique d'Humieres <dominiq@lps.ens.fr>
2016 PR target/71017
2017 * config/i386/cpuid.h: Fix undefined behavior.
2019 2017-01-04 Jeff Law <law@redhat.com>
2021 PR tree-optimization/79007
2022 PR tree-optimization/67955
2023 * tree-ssa-alias.c (same_addr_size_stores_p): Only need to be
2024 conservative for pt.null when flag_non_call_exceptions is on.
2026 2017-01-09 Jakub Jelinek <jakub@redhat.com>
2028 PR translation/79019
2029 PR translation/79020
2030 * params.def (PARAM_INLINE_MIN_SPEEDUP,
2031 PARAM_IPA_CP_SINGLE_CALL_PENALTY,
2032 PARAM_USE_AFTER_SCOPE_DIRECT_EMISSION_THRESHOLD): Fix typos
2033 in descriptions.
2034 * config/avr/avr.opt (maccumulate-args): Likewise.
2035 * config/msp430/msp430.opt (mwarn-mcu): Likewise.
2036 * common.opt (freport-bug): Likewise.
2037 * cif-code.def (CIF_FINAL_ERROR): Likewise.
2038 * doc/invoke.texi (ipa-cp-single-call-penalty): Likewise.
2039 * config/s390/s390.c (s390_invalid_binary_op): Fix spelling in
2040 translatable string.
2041 * config/i386/i386.c (function_value_32): Likewise.
2042 * config/nios2/nios2.c (nios2_valid_target_attribute_rec): Likewise.
2043 * config/msp430/msp430.c (msp430_option_override, msp430_attr):
2044 Likewise.
2045 * config/msp430/driver-msp430.c (msp430_select_hwmult_lib): Likewise.
2046 * common/config/msp430/msp430-common.c (msp430_handle_option):
2047 Likewise.
2048 * symtab.c (symtab_node::verify_base): Likewise.
2049 * opts.c (set_debug_level): Likewise.
2050 * tree.c (verify_type_variant): Likewise. Fix typo in comment.
2051 * config/rs6000/rs6000-c.c (altivec_resolve_overloaded_builtin): Add
2052 missing whitespace to translatable strings.
2053 * config/avr/avr.md (bswapsi2): Fix typo in comment.
2054 * config/sh/superh.h: Likewise.
2055 * config/i386/xopintrin.h: Likewise.
2056 * config/i386/znver1.md: Likewise.
2057 * config/rs6000/rs6000.c (struct rs6000_opt_mask): Likewise.
2058 * ipa-inline-analysis.c (compute_inline_parameters): Likewise.
2059 * double-int.h (struct double_int): Likewise.
2060 * double-int.c (div_and_round_double): Likewise.
2061 * wide-int.cc: Likewise.
2062 * tree-ssa.c (non_rewritable_mem_ref_base): Likewise.
2063 * tree-ssa-sccvn.c (vn_reference_lookup_3): Likewise.
2064 * cfgcleanup.c (crossjumps_occured): Renamed to ...
2065 (crossjumps_occurred): ... this.
2066 (try_crossjump_bb, try_head_merge_bb, try_optimize_cfg, cleanup_cfg):
2067 Adjust all uses.
2069 PR tree-optimization/78899
2070 * tree-if-conv.c (version_loop_for_if_conversion): Instead of
2071 returning bool return struct loop *, NULL for failure and the new
2072 loop on success.
2073 (versionable_outer_loop_p): Don't version outer loop if it has
2074 dont_vectorized bit set.
2075 (tree_if_conversion): When versioning outer loop, ensure
2076 tree_if_conversion is performed also on the inner loop of the
2077 non-vectorizable outer loop copy.
2078 * tree-vectorizer.c (set_uid_loop_bbs): Formatting fix. Fold
2079 LOOP_VECTORIZED in inner loop of the scalar outer loop and
2080 prevent vectorization of it.
2081 (vectorize_loops): For outer + inner LOOP_VECTORIZED, ensure
2082 the outer loop vectorization of the non-scalar version is attempted
2083 before vectorization of the inner loop in scalar version. If
2084 outer LOOP_VECTORIZED guarded loop is not vectorized, prevent
2085 vectorization of its inner loop.
2086 * tree-vect-loop-manip.c (rename_variables_in_bb): If outer_loop
2087 has 2 inner loops, rename also on edges from bb whose single pred
2088 is outer_loop->header. Fix typo in function comment.
2090 2017-01-09 Martin Sebor <msebor@redhat.com>
2092 PR bootstrap/79033
2093 * asan.c (asan_emit_stack_protection): Increase local buffer size
2094 to avoid snprintf truncation warning.
2096 2017-01-09 Andrew Pinski <apinski@cavium.com>
2098 * config/aarch64/aarch64-cores.def: Add thunderx2t99. Change vulcan
2099 to reference thunderx2t99 for the tuning structure
2100 * config/aarch64/aarch64-cost-tables.h (vulcan_extra_costs):
2101 Rename to ...
2102 (thunderx2t99_extra_costs): This.
2103 * config/aarch64/aarch64-tune.md: Regenerate.
2104 * config/aarch64/aarch64.c (vulcan_addrcost_table): Rename to ...
2105 (vulcan_addrcost_table): This.
2106 (vulcan_regmove_cost): Rename to ...
2107 (thunderx2t99_regmove_cost): This.
2108 (vulcan_vector_cost): Rename to ...
2109 (thunderx2t99_vector_cost): this.
2110 (vulcan_branch_cost): Rename to ...
2111 (thunderx2t99_branch_cost): This.
2112 (vulcan_tunings): Rename to ...
2113 (thunderx2t99_tunings): This and s/vulcan/thunderx2t99 .
2114 * doc/invoke.texi (AARCH64/mtune): Add thunderx2t99.
2116 2017-01-09 Martin Jambor <mjambor@suse.cz>
2118 PR ipa/78365
2119 PR ipa/78599
2120 * ipa-prop.h (ipa_jump_func): Swap positions of vr_known and m_vr.
2121 * ipa-cp.c (ipa_vr_operation_and_type_effects): New function.
2122 (propagate_vr_accross_jump_function): Use the above function for all
2123 value range computations for pass-through jump functions and type
2124 converasion from explicit value range values.
2125 (ipcp_propagate_stage): Do not attempt to deduce types of formal
2126 parameters from TYPE_ARG_TYPES.
2127 * ipa-prop.c (ipa_write_jump_function): Remove trailing whitespace.
2128 (ipa_write_node_info): Stream type of the actual argument.
2129 (ipa_read_node_info): Likewise. Also remove trailing whitespace.
2131 2017-01-09 Martin Liska <mliska@suse.cz>
2133 PR pch/78970
2134 * gcc.c (driver_handle_option): Handle OPT_E and set have_E.
2135 (lookup_compiler): Do not show error message with have_E.
2137 2017-01-09 Jakub Jelinek <jakub@redhat.com>
2139 PR tree-optimization/78938
2140 * tree-vect-stmts.c (vectorizable_condition): For non-masked COND_EXPR
2141 where comp_vectype is VECTOR_BOOLEAN_TYPE_P, use
2142 BIT_{NOT,XOR,AND,IOR}_EXPR on the comparison operands instead of
2143 {EQ,NE,GE,GT,LE,LT}_EXPR directly inside of VEC_COND_EXPR. Formatting
2144 fixes.
2146 2017-01-09 Kyrylo Tkachov <kyrylo.tkachov@arm.com>
2148 * tree-ssa-address.c (gen_addr_rtx): Don't handle index if it
2149 is const0_rtx.
2151 2017-01-09 Richard Biener <rguenther@suse.de>
2153 PR tree-optimization/78997
2154 * tree-vect-slp.c (vect_mask_constant_operand_p): Handle SSA
2155 name condition properly.
2157 2017-01-09 Richard Biener <rguenther@suse.de>
2159 PR debug/79000
2160 * dwarf2out.c (is_cxx): New overload with context.
2161 (is_naming_typedef_decl): Use it.
2163 2017-01-08 Sandra Loosemore <sandra@codesourcery.com>
2165 * invoke.texi (Option Summary): Correct spacing in option lists
2166 and add line breaks to fix over-long lines.
2168 2017-01-08 Sandra Loosemore <sandra@codesourcery.com>
2170 PR middle-end/17660
2172 * extend.texi (Common Variable Attributes): Add xref to GCC
2173 Internals manual to explain mode attribute keywords.
2175 2017-01-08 Sandra Loosemore <sandra@codesourcery.com>
2177 PR other/16519
2178 * doc/invoke.texi (Option Summary): Move -pthread to Linker Options
2179 and Preprocessor Options.
2180 (Options for Linking): Document -pthread here....
2181 (RS/6000 and PowerPC Options): ...not here.
2182 (Solaris 2 Options): ...or here.
2183 * doc/cppopts.texi: Document -pthread.
2185 2017-01-08 Martin Sebor <msebor@redhat.com>
2187 PR middle-end/77708
2188 * doc/invoke.texi (Warning Options): Document -Wformat-truncation.
2189 * gimple-ssa-sprintf.c (call_info::reval_used, call_info::warnopt):
2190 New member functions.
2191 (format_directive): Used them.
2192 (add_bytes): Same.
2193 (pass_sprintf_length::handle_gimple_call): Same.
2194 * graphite-sese-to-poly.c (tree_int_to_gmp): Increase buffer size
2195 to avoid truncation for any argument.
2196 (extract_affine_mul): Same.
2197 * tree.c (get_file_function_name): Same.
2199 2017-01-01 Jan Hubicka <hubicka@ucw.cz>
2201 PR middle-end/77484
2202 * predict.def (PRED_INDIR_CALL): Set to 86.
2204 2017-01-07 Sandra Loosemore <sandra@codesourcery.com>
2206 PR preprocessor/54124
2207 * doc/cppopts.texi: Reformat -d subtable to list the full name
2208 of the options. Add cross-reference to the docs for the general
2209 compiler -d options.
2210 * doc/invoke.texi (Developer Options): Add cross-reference to the
2211 preprocessor-specific -d option documentation.
2213 2017-01-07 Sandra Loosemore <sandra@codesourcery.com>
2215 PR preprocessor/13498
2216 * doc/cpp.texi (Search Path): Rewrite to remove obsolete and
2217 redudant material, and reflect new command-line options.
2218 (System Headers): Likewise.
2220 2017-01-07 Sandra Loosemore <sandra@codesourcery.com>
2222 * doc/cppdiropts.texi: Merge documentation of -I, -iquote,
2223 -isystem, and -idirafter. Copy-edit.
2224 * doc/cppopts.texi: Copy-edit. Remove contradiction about
2225 default for -ftrack-macro-expansion. Delete obsolete and
2226 badly-formatted implementation details about -fdebug-cpp output.
2227 * doc/cppwarnopts.texi: Copy-edit.
2229 2017-01-07 David Malcolm <dmalcolm@redhat.com>
2231 PR c++/72803
2232 * input.c (selftest::test_accessing_ordinary_linemaps): Verify
2233 that the transition from a max line width >= 1<<10 to narrower
2234 lines works correctly.
2236 2017-01-07 Alexandre Oliva <aoliva@redhat.com>
2238 * doc/options.texi (PerFunction): New.
2239 * opt-functions.awk (switch_flags): Map both Optimization and
2240 PerFunction to CL_OPTIMIZATION.
2241 * opth-gen.awk: Test for PerFunction flag along with
2242 Optimization.
2243 * optc-save-gen.awk: Likewise. Introduce var_opt_hash and set
2244 it only when the latter is present. Skip those that don't in
2245 the hash function generator.
2246 * common.opt (fvar-tracking): Mark as PerFunction instead of
2247 Optimization.
2248 (fvar-tracking-assignments): Likewise.
2249 (fvar-tracking-assignments-toggle): Likewise.
2250 (fvar-tracking-uninit): Likewise.
2252 2017-01-07 Jakub Jelinek <jakub@redhat.com>
2254 PR translation/79018
2255 * params.def (PARAM_MAX_STORES_TO_MERGE): Add missing space between
2256 the and store.
2258 2017-01-06 Mikael Pettersson <mikpelinux@gmail.com>
2260 PR target/57583
2261 * config/m68k/m68k.opt (LONG_JUMP_TABLE_OFFSETS): New option.
2262 * config/m68k/linux.h (ASM_RETURN_CASE_JUMP): Handle
2263 TARGET_LONG_JUMP_TABLE_OFFSETS.
2264 * config/m68k/m68kelf.h (ASM_RETURN_CASE_JUMP): Likewise.
2265 * config/m68k/netbsd-elf.h (ASM_RETURN_CASE_JUMP): Likewise.
2266 * config/m68k/m68k.h (CASE_VECTOR_MODE): Likewise.
2267 (ASM_OUTPUT_ADDR_DIFF_ELF): Likewise.
2268 * config/m68k/m68k.md (tablejump expander): Likewise.
2269 (*tablejump_pcrel_hi): Renamed from unnamed insn, reject
2270 TARGET_LONG_JUMP_TABLE_OFFSETS.
2271 (*tablejump_pcrel_si): New insn, handle TARGET_LONG_JUMP_TABLE_OFFSETS.
2272 * doc/invoke.texi (M68K options): Add -mlong-jump-table-offsets.
2274 2017-01-06 Edgar E. Iglesias <edgar.iglesias@xilinx.com>
2275 David Holsgrove <david.holsgrove@xilinx.com>
2277 * common/config/microblaze/microblaze-common.c
2278 (TARGET_EXCEPT_UNWIND_INFO): Remove.
2279 * config/microblaze/microblaze-protos.h (microblaze_eh_return):
2280 New prototype.
2281 * config/microblaze/microblaze.c (microblaze_must_save_register)
2282 (microblaze_expand_epilogue, microblaze_return_addr): Handle
2283 calls_eh_return.
2284 (microblaze_eh_return): New function.
2285 * config/microblaze/microblaze.h (RETURN_ADDR_OFFSET)
2286 (EH_RETURN_DATA_REGNO, MB_EH_STACKADJ_REGNUM)
2287 (EH_RETURN_STACKADJ_RTX, ASM_PREFERRED_EH_DATA_FORMAT): New macros.
2288 * config/microblaze/microblaze.md (eh_return): New pattern.
2290 2017-01-06 Jakub Jelinek <jakub@redhat.com>
2292 * system.h (GCC_DIAGNOSTIC_PUSH_IGNORED, GCC_DIAGNOSTIC_POP,
2293 GCC_DIAGNOSTIC_STRINGIFY): Define.
2295 * read-rtl.c (rtx_reader::read_rtx_code): Avoid -Wsign-compare warning.
2297 2017-01-06 Andre Vieira <andre.simoesdiasvieira@arm.com>
2299 * config/arm/arm.md (<mcrr>): New.
2300 (<mrrc>): New.
2301 * config/arm/arm.c (arm_arch5te): New.
2302 (arm_option_override): Set arm_arch5te.
2303 (arm_coproc_builtin_available): Add support for mcrr, mcrr2, mrrc
2304 and mrrc2.
2305 * config/arm/arm-builtins.c (MCRR_QUALIFIERS): Define to...
2306 (arm_mcrr_qualifiers): ... this. New.
2307 (MRRC_QUALIFIERS): Define to...
2308 (arm_mrrc_qualifiers): ... this. New.
2309 * config/arm/arm_acle.h (__arm_mcrr, __arm_mcrr2, __arm_mrrc,
2310 __arm_mrrc2): New.
2311 * config/arm/arm_acle_builtins.def (mcrr, mcrr2, mrrc, mrrc2): New.
2312 * config/arm/iterators.md (MCRRI, mcrr, MCRR): New.
2313 (MRRCI, mrrc, MRRC): New.
2314 * config/arm/unspecs.md (VUNSPEC_MCRR, VUNSPEC_MCRR2, VUNSPEC_MRRC,
2315 VUNSPEC_MRRC2): New.
2317 2017-01-06 Andre Vieira <andre.simoesdiasvieira@arm.com>
2319 * config/arm/arm.md (<mcr>): New.
2320 (<mrc>): New.
2321 * config/arm/arm.c (arm_coproc_builtin_available): Add
2322 support for mcr, mrc, mcr2 and mrc2.
2323 * config/arm/arm-builtins.c (MCR_QUALIFIERS): Define to...
2324 (arm_mcr_qualifiers): ... this. New.
2325 (MRC_QUALIFIERS): Define to ...
2326 (arm_mrc_qualifiers): ... this. New.
2327 (MCR_QUALIFIERS): Define to ...
2328 (arm_mcr_qualifiers): ... this. New.
2329 * config/arm/arm_acle.h (__arm_mcr, __arm_mrc, __arm_mcr2,
2330 __arm_mrc2): New.
2331 * config/arm/arm_acle_builtins.def (mcr, mcr2, mrc, mrc2): New.
2332 * config/arm/iterators.md (MCRI, mcr, MCR, MRCI, mrc, MRC): New.
2333 * config/arm/unspecs.md (VUNSPEC_MCR, VUNSPEC_MCR2, VUNSPEC_MRC,
2334 VUNSPEC_MRC2): New.
2336 2017-01-06 Andre Vieira <andre.simoesdiasvieira@arm.com>
2338 * config/arm/arm.md (*ldc): New.
2339 (*stc): New.
2340 (<ldc>): New.
2341 (<stc>): New.
2342 * config/arm/arm.c (arm_coproc_builtin_available): Add
2343 support for ldc,ldcl,stc,stcl,ldc2,ldc2l,stc2 and stc2l.
2344 (arm_coproc_ldc_stc_legitimate_address): New.
2345 * config/arm/arm-builtins.c (arm_type_qualifiers): Add
2346 'qualifier_const_pointer'.
2347 (LDC_QUALIFIERS): Define to...
2348 (arm_ldc_qualifiers): ... this. New.
2349 (STC_QUALIFIERS): Define to...
2350 (arm_stc_qualifiers): ... this. New.
2351 * config/arm/arm-protos.h
2352 (arm_coproc_ldc_stc_legitimate_address): New.
2353 * config/arm/arm_acle.h (__arm_ldc, __arm_ldcl, __arm_stc,
2354 __arm_stcl, __arm_ldc2, __arm_ldc2l, __arm_stc2, __arm_stc2l): New.
2355 * config/arm/arm_acle_builtins.def (ldc, ldc2, ldcl, ldc2l, stc,
2356 stc2, stcl, stc2l): New.
2357 * config/arm/constraints.md (Uz): New.
2358 * config/arm/iterators.md (LDCI, STCI, ldc, stc, LDC STC): New.
2359 * config/arm/unspecs.md (VUNSPEC_LDC, VUNSPEC_LDC2, VUNSPEC_LDCL,
2360 VUNSPEC_LDC2L, VUNSPEC_STC, VUNSPEC_STC2, VUNSPEC_STCL,
2361 VUNSPEC_STC2L): New.
2363 2017-01-06 Andre Vieira <andre.simoesdiasvieira@arm.com>
2365 * config/arm/arm.md (<cdp>): New.
2366 * config/arm/arm.c (neon_const_bounds): Rename this ...
2367 (arm_const_bounds): ... this.
2368 (arm_coproc_builtin_available): New.
2369 * config/arm/arm-builtins.c (SIMD_MAX_BUILTIN_ARGS): Increase.
2370 (arm_type_qualifiers): Add 'qualifier_unsigned_immediate'.
2371 (CDP_QUALIFIERS): Define to...
2372 (arm_cdp_qualifiers): ... this. New.
2373 (void_UP): Define.
2374 (arm_expand_builtin_args): Add case for 6 arguments.
2375 * config/arm/arm-protos.h (neon_const_bounds): Rename this ...
2376 (arm_const_bounds): ... this.
2377 (arm_coproc_builtin_available): New.
2378 * config/arm/arm_acle.h (__arm_cdp): New.
2379 (__arm_cdp2): New.
2380 * config/arm/arm_acle_builtins.def (cdp): New.
2381 (cdp2): New.
2382 * config/arm/iterators.md (CDPI,CDP,cdp): New.
2383 * config/arm/neon.md: Rename all 'neon_const_bounds' to
2384 'arm_const_bounds'.
2385 * config/arm/types.md (coproc): New.
2386 * config/arm/unspecs.md (VUNSPEC_CDP, VUNSPEC_CDP2): New.
2387 * gcc/doc/extend.texi (ACLE): Add a mention of Coprocessor intrinsics.
2388 * gcc/doc/sourcebuild.texi (arm_coproc1_ok, arm_coproc2_ok,
2389 arm_coproc3_ok, arm_coproc4_ok): Document new effective targets.
2391 2017-01-06 Andre Vieira <andre.simoesdiasvieira@arm.com>
2393 * config/arm/arm-builtins.c (arm_unsigned_binop_qualifiers): New.
2394 (UBINOP_QUALIFIERS): New.
2395 (si_UP): Define.
2396 (acle_builtin_data): New. Change comment.
2397 (arm_builtins): Remove ARM_BUILTIN_CRC32B, ARM_BUILTIN_CRC32H,
2398 ARM_BUILTIN_CRC32W, ARM_BUILTIN_CRC32CB, ARM_BUILTIN_CRC32CH,
2399 ARM_BUILTIN_CRC32CW. Add ARM_BUILTIN_ACLE_BASE and include
2400 arm_acle_builtins.def.
2401 (ARM_BUILTIN_ACLE_PATTERN_START): Define.
2402 (arm_init_acle_builtins): New.
2403 (CRC32_BUILTIN): Remove.
2404 (bdesc_2arg): Remove entries for crc32b, crc32h, crc32w,
2405 crc32cb, crc32ch and crc32cw.
2406 (arm_init_crc32_builtins): Remove.
2407 (arm_init_builtins): Use arm_init_acle_builtins rather
2408 than arm_init_crc32_builtins.
2409 (arm_expand_acle_builtin): New.
2410 (arm_expand_builtin): Use 'arm_expand_acle_builtin'.
2411 * config/arm/arm_acle_builtins.def: New.
2413 2017-01-06 Andre Vieira <andre.simoesdiasvieira@arm.com>
2415 * config/arm/arm-builtins.c (neon_builtin_datum): Rename to ..
2416 (arm_builtin_datum): ... this.
2417 (arm_init_neon_builtin): Rename to ...
2418 (arm_init_builtin): ... this. Add a new parameters PREFIX
2419 and USE_SIG_IN_NAME.
2420 (arm_init_neon_builtins): Replace 'arm_init_neon_builtin' with
2421 'arm_init_builtin'. Replace type 'neon_builtin_datum' with
2422 'arm_builtin_datum'.
2423 (arm_init_vfp_builtins): Likewise.
2424 (builtin_arg): Rename enum's replacing 'NEON_ARG' with
2425 'ARG_BUILTIN' and add a 'ARG_BUILTIN_NEON_MEMORY.
2426 (arm_expand_neon_args): Rename to ...
2427 (arm_expand_builtin_args): ... this. Rename builtin_arg
2428 enum values and differentiate between ARG_BUILTIN_MEMORY
2429 and ARG_BUILTIN_NEON_MEMORY.
2430 (arm_expand_neon_builtin_1): Rename to ...
2431 (arm_expand_builtin_1): ... this. Rename builtin_arg enum
2432 values, arm_expand_builtin_args and add bool parameter NEON.
2433 (arm_expand_neon_builtin): Use arm_expand_builtin_1.
2434 (arm_expand_vfp_builtin): Likewise.
2435 (NEON_MAX_BUILTIN_ARGS): Remove, it was unused.
2437 2017-01-01 Jan Hubicka <hubicka@ucw.cz>
2439 PR middle-end/77484
2440 * predict.def (PRED_POLYMORPHIC_CALL): Set to 59.
2441 * predict.c (tree_estimate_probability_bb): Reverse direction of
2442 polymorphic call predictor.
2444 2017-01-06 David Malcolm <dmalcolm@redhat.com>
2446 * passes.c (execute_one_pass): Split out pass-skipping logic into...
2447 (determine_pass_name_match): ...this new function and...
2448 (should_skip_pass_p): ...this new function.
2450 2017-01-06 Nathan Sidwell <nathan@acm.org>
2452 * ipa-visibility.c (function_and_variable_visibility): Reformat
2453 comments and long lines. Remove extrneous if.
2454 * symtab.c (symtab_node::make_decl_local): Fix code format.
2455 (symtab_node::set_section_for_node): Fix comment typo.
2457 2017-01-06 Martin Liska <mliska@suse.cz>
2459 PR bootstrap/79003
2460 * lra-constraints.c: Rename invariant to lra_invariant.
2461 * predict.c (set_even_probabilities): Initialize e to NULL.
2463 2017-01-05 Martin Sebor <msebor@redhat.com>
2465 PR tree-optimization/78910
2466 * gimple-ssa-sprintf.c (tree_digits): Add an argument.
2467 (format_integer): Correct off-by-one error in the handling
2468 of precision with negative numbers in signed conversions..
2470 2017-01-05 Eric Botcazou <ebotcazou@adacore.com>
2472 * doc/invoke.texi (C Dialect Options): Adjust -fsso-struct entry.
2474 2017-01-05 Jakub Jelinek <jakub@redhat.com>
2476 PR tree-optimization/71016
2477 * tree-ssa-phiopt.c (tree_ssa_phiopt_worker): Pass cond_stmt to
2478 factor_out_conditional_conversion. Formatting fix.
2479 (factor_out_conditional_conversion): Add cond_stmt argument.
2480 If arg1 is INTEGER_CST, punt if new_arg0 is not any operand of
2481 cond_stmt and if arg0_def_stmt is not the only stmt in its bb.
2482 Formatting fix.
2484 2017-01-05 David Malcolm <dmalcolm@redhat.com>
2486 * Makefile.in (OBJS): Add read-md.o, read-rtl.o,
2487 read-rtl-function.o, and selftest-rtl.o.
2488 * config/aarch64/aarch64.c: Include selftest.h and selftest-rtl.h.
2489 (selftest::aarch64_test_loading_full_dump): New function.
2490 (selftest::aarch64_run_selftests): New function.
2491 (TARGET_RUN_TARGET_SELFTESTS): Wire it up to
2492 selftest::aarch64_run_selftests.
2493 * config/i386/i386.c
2494 (selftest::ix86_test_loading_dump_fragment_1): New function.
2495 (selftest::ix86_test_loading_call_insn): New function.
2496 (selftest::ix86_test_loading_full_dump): New function.
2497 (selftest::ix86_test_loading_unspec): New function.
2498 (selftest::ix86_run_selftests): Call the new functions.
2499 * emit-rtl.c (maybe_set_max_label_num): New function.
2500 * emit-rtl.h (maybe_set_max_label_num): New decl.
2501 * function.c (instantiate_decls): Guard call to
2502 instantiate_decls_1 with if (DECL_INITIAL (fndecl)).
2503 * function-tests.c (selftest::verify_three_block_rtl_cfg): Remove
2504 "static".
2505 * gensupport.c (gen_reader::gen_reader): Pass "false"
2506 for new "compact" param of rtx_reader.
2507 * print-rtl.c (rtx_writer::print_rtx_operand): Print "(nil)"
2508 rather than an empty string for NULL strings.
2509 * read-md.c: Potentially include config.h rather than bconfig.h.
2510 Wrap include of errors.h with #ifdef GENERATOR_FILE.
2511 (have_error): New global, copied from errors.c.
2512 (md_reader::read_name): Rename to...
2513 (md_reader::read_name_1): ...this, adding "out_loc" param,
2514 and converting "missing name or number" to returning false, rather
2515 than failing.
2516 (md_reader::read_name): Reimplement in terms of read_name_1.
2517 (md_reader::read_name_or_nil): New function.
2518 (md_reader::read_string): Handle "(nil)" by returning NULL.
2519 (md_reader::md_reader): Add new param "compact".
2520 (md_reader::read_md_files): Wrap with #ifdef GENERATOR_FILE.
2521 (md_reader::read_file): New method.
2522 * read-md.h (md_reader::md_reader): Add new param "compact".
2523 (md_reader::read_file): New method.
2524 (md_reader::is_compact): New accessor.
2525 (md_reader::read_name): Convert return type from void to file_location.
2526 (md_reader::read_name_or_nil): New decl.
2527 (md_reader::read_name_1): New decl.
2528 (md_reader::m_compact): New field.
2529 (noop_reader::noop_reader): Pass "false" for new "compact" param
2530 of rtx_reader.
2531 (rtx_reader::rtx_reader): Add new "compact" param.
2532 (rtx_reader::read_rtx_operand): Make virtual and convert return
2533 type from void to rtx.
2534 (rtx_reader::read_until): New decl.
2535 (rtx_reader::handle_any_trailing_information): New virtual function.
2536 (rtx_reader::postprocess): New virtual function.
2537 (rtx_reader::finalize_string): New virtual function.
2538 (rtx_reader::m_in_call_function_usage): New field.
2539 (rtx_reader::m_reuse_rtx_by_id): New field.
2540 * read-rtl-function.c: New file.
2541 * selftest-rtl.c (selftest::assert_rtx_ptr_eq_at): New function.
2542 * selftest-rtl.h (ASSERT_RTX_PTR_EQ): New macro.
2543 (selftest::verify_three_block_rtl_cfg): New decl.
2544 * read-rtl-function.h: New file.
2545 * read-rtl.c: Potentially include config.h rather than bconfig.h.
2546 For host, include function.h, memmodel.h, and emit-rtl.h.
2547 (one_time_initialization): New function.
2548 (struct compact_insn_name): New struct.
2549 (compact_insn_names): New array.
2550 (find_code): Handle insn codes in compact dumps.
2551 (apply_subst_iterator): Wrap with #ifdef GENERATOR_FILE.
2552 (bind_subst_iter_and_attr): Likewise.
2553 (add_condition_to_string): Likewise.
2554 (add_condition_to_rtx): Likewise.
2555 (apply_attribute_uses): Likewise.
2556 (add_current_iterators): Likewise.
2557 (apply_iterators): Likewise.
2558 (initialize_iterators): Guard usage of apply_subst_iterator with
2559 #ifdef GENERATOR_FILE.
2560 (read_conditions): Wrap with #ifdef GENERATOR_FILE.
2561 (md_reader::read_mapping): Likewise.
2562 (add_define_attr_for_define_subst): Likewise.
2563 (add_define_subst_attr): Likewise.
2564 (read_subst_mapping): Likewise.
2565 (check_code_iterator): Likewise.
2566 (rtx_reader::read_rtx): Likewise. Move one-time initialization
2567 logic to...
2568 (one_time_initialization): New function.
2569 (rtx_reader::read_until): New method.
2570 (read_flags): New function.
2571 (parse_reg_note_name): New function.
2572 (rtx_reader::read_rtx_code): Initialize "iterator" to NULL.
2573 Handle reuse_rtx ids.
2574 Wrap iterator lookup within #ifdef GENERATOR_FILE.
2575 Add parsing support for RTL dumps, mirroring the special-cases in
2576 print_rtx, by calling read_flags, reading REG_NOTE names, INSN_UID
2577 values, and calling handle_any_trailing_information.
2578 (rtx_reader::read_rtx_operand): Convert return type from void
2579 to rtx, returning return_rtx. Handle case 'e'. Call
2580 finalize_string on XSTR and XTMPL fields.
2581 (rtx_reader::read_nested_rtx): Handle dumps in which trailing
2582 "(nil)" values were omitted. Call the postprocess vfunc on the
2583 return_rtx.
2584 (rtx_reader::rtx_reader): Add new "compact" param and pass to base
2585 class ctor. Initialize m_in_call_function_usage. Call
2586 one_time_initialization.
2587 * rtl-tests.c (selftest::test_uncond_jump): Call
2588 set_new_first_and_last_insn.
2589 * rtl.h (read_rtx): Wrap decl with #ifdef GENERATOR_FILE.
2590 * selftest-rtl.c: New file.
2591 * selftest-rtl.h (class selftest::rtl_dump_test): New class.
2592 (selftest::get_insn_by_uid): New decl.
2593 * selftest-run-tests.c (selftest::run_tests): Call
2594 read_rtl_function_c_tests.
2595 * selftest.h (selftest::read_rtl_function_c_tests): New decl.
2596 * tree-dfa.c (ssa_default_def): Return NULL_TREE for rtl function
2597 dumps.
2599 2017-01-05 Uros Bizjak <ubizjak@gmail.com>
2601 * config/i386/i386.md (*testqi_ext_3): No need to handle memory
2602 operands in a special way. Assert that pos+len <= mode precision.
2604 2017-01-05 Jakub Jelinek <jakub@redhat.com>
2606 * common.opt (fvect-cost-model): Remove RejectNegative flag, use
2607 3 argument Alias with unlimited for the negative form.
2608 (fno-vect-cost-model): Removed.
2610 2017-01-05 Martin Liska <mliska@suse.cz>
2612 * hsa-gen.c (gen_hsa_divmod): New function.
2613 (gen_hsa_insn_for_internal_fn_call): Use the function for IFN_DIVMOD.
2615 2017-01-05 Martin Liska <mliska@suse.cz>
2617 PR pch/78970
2618 * gcc.c (lookup_compiler): Reject '-' filename for a precompiled
2619 header.
2621 2017-01-05 Andreas Krebbel <krebbel@linux.vnet.ibm.com>
2623 * config/s390/s390.c (s390_expand_setmem): Unroll the loop for
2624 small constant length operands.
2626 2017-01-05 Andreas Krebbel <krebbel@linux.vnet.ibm.com>
2628 * config/s390/s390.c (s390_expand_setmem): Avoid overlapping bytes
2629 between loop iterations.
2631 2017-01-05 Martin Liska <mliska@suse.cz>
2633 PR sanitizer/78815
2634 * gimplify.c (gimplify_decl_expr): Compare to
2635 asan_poisoned_variables instread of checking flags.
2636 (gimplify_target_expr): Likewise.
2637 (gimplify_expr): Likewise.
2638 (gimplify_function_tree): Conditionally initialize
2639 asan_poisoned_variables.
2641 2017-01-04 Jeff Law <law@redhat.com>
2643 PR tree-optimizatin/78812
2644 * rtl.h (contains_mem_rtx_p): Prototype.
2645 * ifcvt.c (containts_mem_rtx_p): Move from here to...
2646 * rtlanal.c (contains_mem_rtx_p): Here and remove static linkage.
2647 * gcse.c (prune_expressions): Use contains_mem_rtx_p to discover
2648 and prune MEMs that are not at the toplevel of a SET_SRC rtx. Look
2649 through ZERO_EXTEND and SIGN_EXTEND when trying to avoid pruning MEMs.
2651 2017-01-04 Alexandre Oliva <aoliva@redhat.com>
2653 * input.c (assert_char_at_range): Default-initialize actual_range.
2655 2017-01-04 Alexandre Oliva <aoliva@redhat.com>
2657 * df-scan.c (df_ref_create_structure): Make regno unsigned,
2658 to match the caller.
2660 2017-01-04 Alexandre Oliva <aoliva@redhat.com>
2662 * cfgexpand.c (expand_gimple_basic_block): Disregard debug
2663 insns after final jump in test to emit dummy move.
2665 2017-01-04 Alexandre Oliva <aoliva@redhat.com>
2667 * gimple-iterator.h (gsi_one_nondebug_before_end_p): New.
2668 * tree-eh.c (cleanup_empty_eh): Skip more debug stmts.
2670 2017-01-04 Alexandre Oliva <aoliva@redhat.com>
2672 * multiple_target.c (create_dispatcher_calls): Init e_next.
2673 * tree-ssa-loop-split.c (split_loop): Init border.
2674 * tree-vect-loop.c (vect_determine_vectorization_factor): Init
2675 scalar_type.
2677 2017-01-04 Michael Meissner <meissner@linux.vnet.ibm.com>
2679 PR target/71977
2680 PR target/70568
2681 PR target/78823
2682 * config/rs6000/predicates.md (sf_subreg_operand): New predicate.
2683 (altivec_register_operand): Do not return true if the operand
2684 contains a SUBREG mixing SImode and SFmode.
2685 (vsx_register_operand): Likewise.
2686 (vsx_reg_sfsubreg_ok): New predicate.
2687 (vfloat_operand): Do not return true if the operand contains a
2688 SUBREG mixing SImode and SFmode.
2689 (vint_operand): Likewise.
2690 (vlogical_operand): Likewise.
2691 (gpc_reg_operand): Likewise.
2692 (int_reg_operand): Likewise.
2693 * config/rs6000/rs6000-protos.h (valid_sf_si_move): Add declaration.
2694 * config/rs6000/rs6000.c (valid_sf_si_move): New function to
2695 determine if a MOVSI or MOVSF operation contains SUBREGs that mix
2696 SImode and SFmode.
2697 (rs6000_emit_move_si_sf_subreg): New helper function.
2698 (rs6000_emit_move): Call rs6000_emit_move_si_sf_subreg to possbily
2699 fixup SUBREGs involving SImode and SFmode.
2700 * config/rs6000/vsx.md (SFBOOL_*): New constants that are operand
2701 numbers for the new peephole2 optimization.
2702 (peephole2 for SFmode unions): New peephole2 to optimize cases in
2703 the GLIBC math library that do AND/IOR/XOR operations on single
2704 precision floating point.
2705 * config/rs6000/rs6000.h (TARGET_NO_SF_SUBREG): New internal
2706 target macros to say whether we need to avoid SUBREGs mixing
2707 SImode and SFmode.
2708 (TARGET_ALLOW_SF_SUBREG): Likewise.
2709 * config/rs6000/rs6000.md (UNSPEC_SF_FROM_SI): New unspecs.
2710 (UNSPEC_SI_FROM_SF): Likewise.
2711 (iorxor): Change spacing.
2712 (and_ior_xor): New iterator for AND, IOR, and XOR.
2713 (movsi_from_sf): New insns for SImode/SFmode SUBREG support.
2714 (movdi_from_sf_zero_ext): Likewise.
2715 (mov<mode>_hardfloat, FMOVE32 iterator): Use register_operand
2716 instead of gpc_reg_operand. Add SImode/SFmode SUBREG support.
2717 (movsf_from_si): New insn for SImode/SFmode SUBREG support.
2718 (fma<mode>4): Use gpc_reg_operand instead of register_operand.
2719 (fms<mode>4): Likewise.
2720 (fnma<mode>4): Likewise.
2721 (fnms<mode>4): Likewise.
2722 (nfma<mode>4): Likewise.
2723 (nfms<mode>4): Likewise.
2725 2017-01-04 Marek Polacek <polacek@redhat.com>
2727 PR c++/64767
2728 * doc/invoke.texi: Document -Wpointer-compare.
2730 2017-01-04 Jakub Jelinek <jakub@redhat.com>
2732 * optc-gen.awk: Emit #error for -W*/-f*/-m* Enum without
2733 RejectNegative.
2735 * dwarf2out.c (output_loc_list): Don't throw away 64K+ location
2736 descriptions for -gdwarf-5 and emit them as uleb128 instead of
2737 2-byte data.
2739 2017-01-04 Kelvin Nilsen <kelvin@gcc.gnu.org>
2741 PR target/78056
2742 * doc/sourcebuild.texi (PowerPC-specific attributes): Add
2743 documentation of the powerpc_popcntb_ok attribute.
2744 * config/rs6000/rs6000.c (rs6000_option_override_internal): Add
2745 code to issue warning messages if a requested CPU configuration is
2746 not supported by the binary (assembler and loader) toolchain.
2747 (spe_init_builtins): Add two assertions to prevent ICE if attempt is
2748 made to define a built-in function that has been disabled.
2749 (paired_init_builtins): Add assertion to prevent ICE if attempt is
2750 made to define a built-in function that has been disabled.
2751 (altivec_init_builtins): Add comment explaining why definition
2752 of the DST built-in functions is not preceded by an assertion
2753 check. Add assertions to prevent ICE if attempts are made to
2754 define an altivec predicate or an abs* built-in function that has
2755 been disabled.
2756 (htm_init_builtins): Add comment explaining why definition of the
2757 htm built-in functions is not preceded by an assertion check.
2759 2017-01-04 Jeff Law <law@redhat.com>
2761 PR tree-optimizatin/67955
2762 * tree-ssa-alias.c (same_addr_size_stores_p): Check offsets first.
2763 Allow any SSA_VAR_P as the base objects. Use integer_zerop. Verify
2764 the points-to solution does not include pt_null. Use DECL_PT_UID
2765 unconditionally.
2767 2017-01-04 Uros Bizjak <ubizjak@gmail.com>
2769 * config/i386/i386.md (HI/SImode test with imm to QImode splitters):
2770 Use gen_int_mode instead of gen_lopwart for const_int operands.
2772 2017-01-04 Jakub Jelinek <jakub@redhat.com>
2774 PR tree-optimization/71563
2775 * match.pd: Simplify X << Y into X if Y is known to be 0 or
2776 out of range value - has low bits known to be zero.
2778 2017-01-04 Alan Modra <amodra@gmail.com>
2780 * Makefile.in (aclocal_deps): Update and order as per aclocal.m4.
2781 * configure: Regenerate.
2782 * config.in: Regenerate.
2784 2017-01-04 Jakub Jelinek <jakub@redhat.com>
2786 PR bootstrap/77569
2787 * input.c (ebcdic_execution_charset::on_error): Don't use strstr for
2788 a substring of the message, but strcmp with the whole message. Ifdef
2789 ENABLE_NLS, translate the message first using dgettext.
2791 2017-01-03 Jeff Law <law@redhat.com>
2793 PR tree-optimizatin/78856
2794 * tree-ssa-threadupdate.c: Include tree-vectorizer.h.
2795 (mark_threaded_blocks): Remove code to truncate thread paths that
2796 cross multiple loop headers. Instead invalidate the cached loop
2797 iteration information and handle case of a thread path walking
2798 into an irreducible region.
2800 2017-01-03 Michael Meissner <meissner@linux.vnet.ibm.com>
2802 PR target/78900
2803 * config/rs6000/rs6000.c (rs6000_split_signbit): Change some
2804 assertions. Add support for doing the signbit if the IEEE 128-bit
2805 floating point value is in a GPR.
2806 * config/rs6000/rs6000.md (Fsignbit): Delete.
2807 (signbit<mode>2_dm): Delete using <Fsignbit> and just use "wa".
2808 Update the length attribute if the value is in a GPR.
2809 (signbit<mode>2_dm_<su>ext): Add combiner pattern to eliminate
2810 the sign or zero extension instruction, since the value is always 0/1.
2811 (signbit<mode>2_dm2): Delete using <Fsignbit>.
2813 PR target/78953
2814 * config/rs6000/vsx.md (vsx_extract_<mode>_store_p9): If we are
2815 extracting SImode to a GPR register so that we can generate a
2816 store, limit the vector to be in a traditional Altivec register
2817 for the vextuwrx instruction.
2819 2017-01-03 Ian Lance Taylor <iant@google.com>
2821 * godump.c (go_format_type): Treat ENUMERAL_TYPE like INTEGER_TYPE.
2823 2017-01-03 Martin Sebor <msebor@redhat.com>
2825 PR tree-optimization/78696
2826 * gimple-ssa-sprintf.c (format_floating): Correct handling of
2827 precision. Use MPFR for %f for greater fidelity. Correct handling
2828 of %g.
2829 (pass_sprintf_length::compute_format_length): Set width and precision
2830 specified by asrerisk to void_node for vararg functions.
2831 (try_substitute_return_value): Adjust dump output.
2833 2017-01-03 David Edelsohn <dje.gcc@gmail.com>
2835 * doc/invoke.texi (RS6000 options): LRA is enabled by default.
2837 2017-01-03 Eric Botcazou <ebotcazou@adacore.com>
2839 * doc/invoke.texi (SPARC options): Document -mlra as the default.
2840 * config/sparc/sparc.c (sparc_option_override): Force LRA unless
2841 -mlra/-mno-lra was passed to the compiler.
2843 2017-01-03 James Cowgill <James.Cowgill@imgtec.com>
2845 PR rtl-optimization/65618
2846 * emit-rtl.c (try_split): Move initialization of "before" and
2847 "after" to just before the call to emit_insn_after_setloc.
2849 2017-01-03 Gerald Pfeifer <gerald@pfeifer.com>
2851 * doc/md.texi (Standard Names): Remove reference to Java frontend.
2853 2017-01-03 Pierre-Marie de Rodat <derodat@adacore.com>
2855 * dwarf2out.c (gen_enumeration_type_die): When
2856 -gno-strict-dwarf, add a DW_AT_encoding attribute.
2858 2017-01-03 Jakub Jelinek <jakub@redhat.com>
2860 PR tree-optimization/78965
2861 * gimple-ssa-sprintf.c (pass_sprintf_length::compute_format_length):
2862 Change first argument from const call_info & to call_info &. For %n
2863 set info.nowrite to false.
2865 PR middle-end/78901
2866 * gimple-ssa-sprintf.c (try_substitute_return_value): Don't change
2867 possibly throwing calls.
2869 * genmatch.c (dt_node::gen_kids_1): If generic_exprs include SSA_NAME
2870 and exprs_len || fns_len, emit the code for SSA_NAME next to the exprs
2871 and fns handling, rather than in a separate case SSA_NAME.
2873 2017-01-02 Jeff Law <law@redhat.com>
2875 * config/darwin-driver.c (darwin_driver_init): Const-correctness
2876 fixes for first_period and second_period variables.
2878 2017-01-02 Uros Bizjak <ubizjak@gmail.com>
2880 PR target/78967
2881 * config/i386/i386.md (UNSPEC_NOREX_MEM): New unspec.
2882 (*insvqi_1): New insn pattern.
2883 (*insvqi_1_mem_rex64): Ditto.
2884 (*insvqi_2): Ditto.
2885 (*insvqi_3): Rename from *insvqi.
2887 (*extzvqi_mem_rex64): Add UNSPEC_NOREX_MEM tag.
2889 2017-01-02 Gerald Pfeifer <gerald@pfeifer.com>
2891 * doc/cfg.texi (Edges): Remove reference to Java.
2892 (Maintaining the CFG): Ditto.
2894 2017-01-01 Jan Hubicka <hubicka@ucw.cz>
2896 PR middle-end/77674
2897 * symtab.c (symtab_node::binds_to_current_def_p): Fix handling of
2898 transparent aliases.
2900 2017-01-01 Jan Hubicka <hubicka@ucw.cz>
2902 PR middle-end/77484
2903 * predict.def (PRED_CALL): Update hitrate.
2904 (PRED_INDIR_CALL, PRED_POLYMORPHIC_CALL): New predictors.
2905 * predict.c (tree_estimate_probability_bb): Split CALL predictor
2906 into direct/indirect/polymorphic variants.
2908 2017-01-01 Jakub Jelinek <jakub@redhat.com>
2910 Update copyright years.
2912 * gcc.c (process_command): Update copyright notice dates.
2913 * gcov-dump.c (print_version): Ditto.
2914 * gcov.c (print_version): Ditto.
2915 * gcov-tool.c (print_version): Ditto.
2916 * gengtype.c (create_file): Ditto.
2917 * doc/cpp.texi: Bump @copying's copyright year.
2918 * doc/cppinternals.texi: Ditto.
2919 * doc/gcc.texi: Ditto.
2920 * doc/gccint.texi: Ditto.
2921 * doc/gcov.texi: Ditto.
2922 * doc/install.texi: Ditto.
2923 * doc/invoke.texi: Ditto.
2924 \f
2925 Copyright (C) 2017 Free Software Foundation, Inc.
2927 Copying and distribution of this file, with or without modification,
2928 are permitted in any medium without royalty provided the copyright
2929 notice and this notice are preserved.