| blob | b70acef6953e1a8d867a5a949f7bc0bb8a6ac6e4 |
1 2017-01-27 Bin Cheng <bin.cheng@arm.com>
3 PR rtl-optimization/78559
4 * combine.c (try_combine): Discard REG_EQUAL and REG_EQUIV for
5 other_insn in combine.
7 2017-01-27 Pekka Jääskeläinen <pekka.jaaskelainen@parmance.com>
9 * builtin-types.def: Use unsigned_char_type_node for BT_UINT8. Use
10 uint16_type_node for BT_UINT16.
12 2017-01-27 David Malcolm <dmalcolm@redhat.com>
14 * doc/sourcebuild.texi (Testsuites): Add "GIMPLE Tests" and
15 "RTL Tests" to menu.
16 (GIMPLE Tests): New node.
17 (RTL Tests): New node.
19 2017-01-27 Richard Biener <rguenther@suse.de>
21 PR tree-optimization/79245
22 * tree-loop-distribution.c (distribute_loop): Apply cost
23 modeling also to detected patterns.
25 2017-01-27 Richard Biener <rguenther@suse.de>
27 PR tree-optimization/71433
28 * tree-vrp.c (register_new_assert_for): Revert earlier changes.
29 (compare_assert_loc): New function.
30 (process_assert_insertions): Sort and optimize assert locations
31 to remove duplicates and push down identical assertions on
32 edges to their destination block.
34 2017-01-27 Richard Biener <rguenther@suse.de>
36 PR tree-optimization/79244
37 * tree-vrp.c (remove_range_assertions): Forcefully propagate
38 out SSA names even if abnormal.
40 2017-01-27 Jakub Jelinek <jakub@redhat.com>
42 * realmpfr.h: Poison MPFR_RND{N,Z,U,D}.
43 * gimple-ssa-sprintf.c (format_floating_max): Use GMP_RNDN
44 instead of MPFR_RNDN.
46 2017-01-27 Richard Earnshaw <rearnsha@arm.com>
48 PR target/79239
49 * arm.c (arm_option_override): Don't call build_target_option_node
50 until after doing all option overrides.
51 (arm_valid_target_attribute_tree): Likewise.
53 2017-01-27 Martin Liska <mliska@suse.cz>
55 * doc/invoke.texi (-fprofile-arcs): Document profiling support
56 for {cd}tors and C++ {cd}tors.
58 2017-01-27 Dominik Vogt <vogt@linux.vnet.ibm.com>
60 * config/s390/s390.md ("*setmem_long_and")
61 ("*setmem_long_and_31z"): Use zero_extend instead of and.
63 2017-01-26 Martin Sebor <msebor@redhat.com>
65 * gimple-ssa-sprintf.c (format_floating): Simplify the computation
66 of precision.
68 2017-01-26 Martin Sebor <msebor@redhat.com>
70 * gimple-ssa-sprintf.c (format_floating): Test HAVE_XFmode and
71 HAVE_DFmode before using XFmode or DFmode.
72 (parse_directive): Avoid using the z length modifier to avoid
73 the ISO C++98 does not support the ‘z’ gnu_printf length modifier.
75 PR middle-end/78703
76 * gimple-ssa-sprintf.c (adjust_for_width_or_precision): Change
77 to accept adjustment as an array.
78 (get_int_range): New function.
79 (struct directive): Make width and prec arrays.
80 (directive::set_width, directive::set_precision): Call get_int_range.
81 (format_integer, format_floating): Handle width and precision ranges.
82 (format_string, parse_directive): Same.
84 2017-01-26 Jakub Jelinek <jakub@redhat.com>
86 PR debug/79129
87 * dwarf2out.c (generate_skeleton_bottom_up): For children with
88 comdat_type_p set, just clone them, but keep the children in the
89 original DIE.
91 PR debug/78835
92 * dwarf2out.c (prune_unused_types): Mark all functions with DIEs
93 which have direct callers with -fvar-tracking-assignments enabled
94 in the current TU.
95 (resolve_addr): Avoid adding skeleton DIEs for DW_AT_call_origin
96 inside of type units.
98 2017-01-26 Martin Sebor <msebor@redhat.com>
100 PR middle-end/78703
101 * gimple-ssa-sprintf.c (struct result_range): Add likely and
102 unlikely counters.
103 (struct format_result): Replace number_chars, number_chars_min,
104 and number_chars_max with a single member of struct result_range.
105 Remove bounded.
106 (format_result::operator+=): Adjust.
107 (struct fmtresult): Remove bounded. Handle likely and unlikely
108 counters.
109 (fmtresult::adjust_for_width_or_precision): New function.
110 (fmtresult:type_max_digits): New function.
111 (bytes_remaining): Handle likely and unlikely counters.
112 (min_bytes_remaining): Remove.
113 (format_percent): Simplify.
114 (format_integer, format_floating): Set likely and unlikely counters.
115 (get_string_length, format_character, format_string): Same.
116 (format_plain, should_warn_p): New function.
117 (maybe_warn): Call should_warn_p. Update diagnostic messages
118 and handle those for all directives, including plain strings.
119 (format_directive): Handle likely and unlikely counters.
120 Remove unnecessary quoting from diagnostics. Add an informational
121 note.
122 (add_bytes): Remove.
123 (pass_sprintf_length::compute_format_length): Simplify.
124 (try_substitute_return_value): Handle likely and unlikely counters.
126 2017-01-26 Carl Love <cel@us.ibm.com>
128 * config/rs6000/rs6000-c (altivec_overloaded_builtins): Remove
129 bogus entries for the P8V_BUILTIN_VEC_VGBBD built-ins
131 2017-01-26 Vladimir Makarov <vmakarov@redhat.com>
133 PR target/79131
134 * lra-assigns.c (setup_live_pseudos_and_spill_after_risky): Take
135 endianess for subregs into account.
136 * lra-constraints.c (lra_constraints): Do risky transformations
137 always on the first iteration.
138 * lra-lives.c (check_pseudos_live_through_calls): Add arg
139 last_call_used_reg_set.
140 (process_bb_lives): Define and use last_call_used_reg_set.
141 * lra.c (lra): Always continue after lra_constraints on the first
142 iteration.
144 2017-01-26 Kirill Yukhin <kirill.yukhin@gmail.com>
146 * gcc.target/i386/avx512bw-kshiftlq-2.c: Use unsigned long long
147 constant.
148 * gcc.target/i386/avx512bw-kshiftrq-2.c: Ditto.
150 2017-01-26 Jakub Jelinek <jakub@redhat.com>
152 * config/i386/avx512fintrin.h (_ktest_mask16_u8,
153 _ktestz_mask16_u8, _ktestc_mask16_u8, _kadd_mask16): Move to ...
154 * config/i386/avx512dqintrin.h (_ktest_mask16_u8,
155 _ktestz_mask16_u8, _ktestc_mask16_u8, _kadd_mask16): ... here.
156 * config/i386/i386-builtin.def (__builtin_ia32_ktestchi,
157 __builtin_ia32_ktestzhi, __builtin_ia32_kaddhi): Use
158 OPTION_MASK_ISA_AVX512DQ instead of OPTION_MASK_ISA_AVX512F.
159 * config/i386/sse.md (SWI1248_AVX512BWDQ2): New mode iterator.
160 (kadd<mode>, ktest<mode>): Use it instead of SWI1248_AVX512BWDQ.
162 2017-01-26 Marek Polacek <polacek@redhat.com>
164 PR c/79199
165 * fold-const.c (operand_equal_p) [COND_EXPR]: Use OP_SAME_WITH_NULL
166 for the third operand.
168 2017-01-26 Jakub Jelinek <jakub@redhat.com>
170 PR middle-end/79236
171 * omp-low.c (struct omp_context): Add simt_stmt field.
172 (scan_omp_for): Return omp_context *.
173 (scan_omp_simd): Set simt_stmt on the non-_simt_ SIMD
174 context to the _simt_ SIMD stmt.
175 (lower_omp_for): For combined SIMD with sibling _simt_
176 SIMD, make sure to use the same decls in _looptemp_
177 clauses as in the sibling.
179 2017-01-26 David Sherwood <david.sherwood@arm.com>
181 PR middle-end/79212
182 * gimplify.c (omp_notice_variable): Add GOVD_SEEN flag to variables in
183 all contexts.
185 2017-01-26 Jakub Jelinek <jakub@redhat.com>
187 PR target/70465
188 * reg-stack.c (emit_swap_insn): Instead of fld a; fld b; fxchg %st(1);
189 emit fld b; fld a; if possible.
191 * brig-builtins.def: Update copyright years.
192 * config/arm/arm_acle_builtins.def: Update copyright years.
194 2017-01-25 Michael Meissner <meissner@linux.vnet.ibm.com>
196 PR target/79179
197 * config/rs6000/vsx.md (vsx_extract_<mode>_store): Use wY
198 constraint instead of o for the stxsd instruction.
200 2017-01-25 Carl Love <cel@us.ibm.com>
202 * config/rs6000/rs6000-c (altivec_overloaded_builtins): Fix order
203 of entries for ALTIVEC_BUILTIN_VEC_PACKS and P8V_BUILTIN_VEC_VGBBD.
205 2017-01-25 Jonathan Wakely <jwakely@redhat.com>
207 * doc/invoke.texi (C++ Dialect Options): Fix typo.
209 2017-01-25 Richard Biener <rguenther@suse.de>
211 PR tree-optimization/69264
212 * target.def (vector_alignment_reachable): Improve documentation.
213 * doc/tm.texi: Regenerate.
214 * targhooks.c (default_builtin_vector_alignment_reachable): Simplify
215 and add a comment.
216 * tree-vect-data-refs.c (vect_supportable_dr_alignment): Revert
217 earlier changes with respect to TYPE_USER_ALIGN.
218 (vector_alignment_reachable_p): Likewise. Improve dumping.
220 2016-01-25 Kyrylo Tkachov <kyrylo.tkachov@arm.com>
222 PR target/79145
223 * config/arm/arm.md (xordi3): Force constant operand into a register
224 for TARGET_IWMMXT.
226 2016-01-25 Kyrylo Tkachov <kyrylo.tkachov@arm.com>
228 * doc/invoke.texi (-fstore-merging): Correct default optimization
229 levels at which it is enabled.
230 (-O): Move -fstore-merging from list to...
231 (-O2): ... Here.
233 2017-01-25 Richard Biener <rguenther@suse.de>
235 PR debug/78363
236 * omp-expand.c: Include debug.h.
237 (expand_omp_taskreg): Make sure to generate early debug before
238 outlining anything from a function.
239 (expand_omp_target): Likewise.
240 (grid_expand_target_grid_body): Likewise.
242 2017-01-25 Maxim Ostapenko <m.ostapenko@samsung.com>
244 PR lto/79061
245 * asan.c (get_translation_unit_decl): New function.
246 (asan_add_global): Extract modules file name from globals
247 TRANSLATION_UNIT_DECL name.
249 2017-01-24 Eric Botcazou <ebotcazou@adacore.com>
251 PR target/77439
252 * config/arm/arm.c (arm_function_ok_for_sibcall): Add back restriction
253 for long calls with APCS frame and VFP.
255 2017-01-24 David Malcolm <dmalcolm@redhat.com>
257 * cfg.c (original_copy_tables_initialized_p): New function.
258 * cfg.h (original_copy_tables_initialized_p): New decl.
259 * cfgrtl.c (relink_block_chain): Guard the call to
260 free_original_copy_tables with a call to
261 original_copy_tables_initialized_p.
262 * cgraph.h (symtab_node::native_rtl_p): New decl.
263 * cgraphunit.c (symtab_node::native_rtl_p): New function.
264 (symtab_node::needed_p): Don't assert for early assembly output
265 for __RTL functions.
266 (cgraph_node::finalize_function): Set "force_output" for __RTL
267 functions.
268 (cgraph_node::analyze): Bail out early for __RTL functions.
269 (analyze_functions): Update assertion to support __RTL functions.
270 (cgraph_node::expand): Bail out early for __RTL functions.
271 * final.c (rest_of_clean_state): Don't call delete_tree_ssa for
272 __RTL functions.
273 * function.h (struct function): Update comment for field
274 "pass_startwith".
275 * gimple-expr.c: Include "tree-pass.h".
276 (gimple_has_body_p): Return false for __RTL functions.
277 * Makefile.in (OBJS): Add run-rtl-passes.o.
278 * pass_manager.h (gcc::pass_manager::get_rest_of_compilation): New
279 accessor.
280 (gcc::pass_manager::get_clean_slate): New accessor.
281 * passes.c: Include "insn-addr.h".
282 (should_skip_pass_p): Add logging. Update logic for running
283 "expand" to be compatible with both __GIMPLE and __RTL. Guard
284 property-provider override so it is only done for gimple passes.
285 Don't skip dfinit.
286 (skip_pass): New function.
287 (execute_one_pass): Call skip_pass when skipping passes.
288 * read-md.c (md_reader::read_char): Support filtering
289 the input to a subset of line numbers.
290 (md_reader::md_reader): Initialize fields
291 m_first_line and m_last_line.
292 (md_reader::read_file_fragment): New function.
293 * read-md.h (md_reader::read_file_fragment): New decl.
294 (md_reader::m_first_line): New field.
295 (md_reader::m_last_line): New field.
296 * read-rtl-function.c (function_reader::create_function): Only
297 create cfun if it doesn't already exist. Set PROP_rtl on cfun's
298 curr_properties. Set DECL_INITIAL to a dummy block.
299 (read_rtl_function_body_from_file_range): New function.
300 * read-rtl-function.h (read_rtl_function_body_from_file_range):
301 New decl.
302 * run-rtl-passes.c: New file.
303 * run-rtl-passes.h: New file.
305 2017-01-24 Jeff Law <law@redhat.com>
307 * config/microblaze/microblaze.h (ASM_FORMAT_PRIVATE_NAME): Increase
308 buffer size.
310 2017-01-24 Bin Cheng <bin.cheng@arm.com>
312 PR tree-optimization/79159
313 * tree-ssa-loop-niter.c (get_cst_init_from_scev): New function.
314 (record_nonwrapping_iv): Improve boundary using above function if no
315 value range information.
317 2017-01-24 Pekka Jääskeläinen <pekka@parmance.com>
318 Martin Jambor <mjambor@suse.cz>
320 * brig-builtins.def: New file.
321 * builtins.def (DEF_HSAIL_BUILTIN): New macro.
322 (DEF_HSAIL_ATOMIC_BUILTIN): Likewise.
323 (DEF_HSAIL_SAT_BUILTIN): Likewise.
324 (DEF_HSAIL_INTR_BUILTIN): Likewise.
325 (DEF_HSAIL_CVT_ZEROI_SAT_BUILTIN): Likewise.
326 * builtin-types.def (BT_INT8): New.
327 (BT_INT16): Likewise.
328 (BT_UINT8): Likewise.
329 (BT_UINT16): Likewise.
330 (BT_FN_ULONG): Likewise.
331 (BT_FN_UINT_INT): Likewise.
332 (BT_FN_UINT_ULONG): Likewise.
333 (BT_FN_UINT_LONG): Likewise.
334 (BT_FN_UINT_PTR): Likewise.
335 (BT_FN_ULONG_PTR): Likewise.
336 (BT_FN_INT8_FLOAT): Likewise.
337 (BT_FN_INT16_FLOAT): Likewise.
338 (BT_FN_UINT32_FLOAT): Likewise.
339 (BT_FN_UINT16_FLOAT): Likewise.
340 (BT_FN_UINT8_FLOAT): Likewise.
341 (BT_FN_UINT64_FLOAT): Likewise.
342 (BT_FN_UINT16_UINT32): Likewise.
343 (BT_FN_UINT32_UINT16): Likewise.
344 (BT_FN_UINT16_UINT16_UINT16): Likewise.
345 (BT_FN_INT_PTR_INT): Likewise.
346 (BT_FN_UINT_PTR_UINT): Likewise.
347 (BT_FN_LONG_PTR_LONG): Likewise.
348 (BT_FN_ULONG_PTR_ULONG): Likewise.
349 (BT_FN_VOID_UINT64_UINT64): Likewise.
350 (BT_FN_UINT8_UINT8_UINT8): Likewise.
351 (BT_FN_INT8_INT8_INT8): Likewise.
352 (BT_FN_INT16_INT16_INT16): Likewise.
353 (BT_FN_INT_INT_INT): Likewise.
354 (BT_FN_UINT_FLOAT_UINT): Likewise.
355 (BT_FN_FLOAT_UINT_UINT): Likewise.
356 (BT_FN_ULONG_UINT_UINT): Likewise.
357 (BT_FN_ULONG_UINT_PTR): Likewise.
358 (BT_FN_ULONG_ULONG_ULONG): Likewise.
359 (BT_FN_UINT_UINT_UINT): Likewise.
360 (BT_FN_VOID_UINT_PTR): Likewise.
361 (BT_FN_UINT_UINT_PTR: Likewise.
362 (BT_FN_UINT32_UINT64_PTR): Likewise.
363 (BT_FN_INT_INT_UINT_UINT): Likewise.
364 (BT_FN_UINT_UINT_UINT_UINT): Likewise.
365 (BT_FN_UINT_UINT_UINT_PTR): Likewise.
366 (BT_FN_UINT_ULONG_ULONG_UINT): Likewise.
367 (BT_FN_ULONG_ULONG_ULONG_ULONG): Likewise.
368 (BT_FN_LONG_LONG_UINT_UINT): Likewise.
369 (BT_FN_ULONG_ULONG_UINT_UINT): Likewise.
370 (BT_FN_VOID_UINT32_UINT64_PTR): Likewise.
371 (BT_FN_VOID_UINT32_UINT32_PTR): Likewise.
372 (BT_FN_UINT_UINT_UINT_UINT_UINT): Likewise.
373 (BT_FN_UINT_FLOAT_FLOAT_FLOAT_FLOAT): Likewise.
374 (BT_FN_ULONG_ULONG_ULONG_UINT_UINT): Likewise.
375 * doc/frontends.texi: List BRIG FE.
376 * doc/install.texi (Testing): Add BRIG tesring requirements.
377 * doc/invoke.texi (Overall Options): Mention BRIG.
378 * doc/standards.texi (Standards): Doucment BRIG HSA version.
380 2017-01-24 Richard Biener <rguenther@suse.de>
382 PR translation/79208
383 * ipa-devirt.c (odr_types_equivalent_p): Fix typo in diagnostic.
385 2017-01-24 Martin Jambor <mjambor@suse.cz>
387 PR bootstrap/79198
388 * ipa-prop.c (ipa_free_all_node_params): Call summary destructor.
389 * ipa-prop.c (ipa_node_params_t::insert): Initialize fields known_csts
390 and known_contexts.
392 2017-01-24 Aldy Hernandez <aldyh@redhat.com>
394 PR middle-end/79123
395 * gimple-ssa-warn-alloca.c (alloca_call_type): Make sure
396 casts from signed to unsigned really don't have a range.
398 2017-01-24 Markus Trippelsdorf <markus@trippelsdorf.de>
400 * gimple-ssa-sprintf.c (format_floating): Change MPFR_RNDx to
401 GMP_RNDx for compatiblity.
403 2017-01-24 Martin Liska <mliska@suse.cz>
405 PR bootstrap/79132
406 * tree-ssa-reassoc.c (rewrite_expr_tree_parallel): Insert assert
407 that would prevent us to call alloca with -1 as argument.
409 2017-01-24 Jakub Jelinek <jakub@redhat.com>
411 * dwarf2out.c (output_compilation_unit_header, output_file_names):
412 Avoid -Wformat-security warning.
414 2017-01-23 Andrew Pinski <apinski@cavium.com>
416 * config/aarch64/aarch64.c (thunderx2t99_addrcost_table): Improve
417 cost table.
419 2017-01-23 Martin Sebor <msebor@redhat.com>
421 PR middle-end/78703
422 * gimple-ssa-sprintf.c (warn_level): New global.
423 (format_integer): Use it here and throughout the rest of the file.
424 Use the same switch to compute sign as base.
425 (maybe_warn): New function.
426 (format_directive): Factor out warnings into maybe_warn.
427 Add debugging output. Use warn_level.
428 (add_bytes): Use warn_level.
429 (pass_sprintf_length::compute_format_length): Add debugging output.
430 (try_substitute_return_value): Same.
431 (pass_sprintf_length::handle_gimple_call): Set and use warn_level.
433 PR middle-end/78703
434 * gimple-ssa-sprintf.c (struct format_result): Remove constant member.
435 (struct fmtresult, format_integer, format_floating): Adjust.
436 (fmtresult::fmtresult): Set max correctly in two argument ctor.
437 (get_string_length, format_string,format_directive): Same.
438 (pass_sprintf_length::compute_format_length): Same.
439 (try_substitute_return_value): Simplify slightly.
441 PR middle-end/78703
442 * gimple-ssa-sprintf.c (pass_sprintf_length::gate): Adjust formatting.
443 (fmtresult::operator+=): Outlined.
444 (struct fmtresult): Add ctors.
445 (struct conversion_spec): Rename...
446 (struct directive): ...to this. Add and remove data members.
447 (directive::set_width, directive::set_precision): New functions.
448 (format_percent): Use fmtresult ctor.
449 (get_width_and_precision): Remove.
450 (format_integer): Make naming changes. Avoid computing width and
451 precision.
452 (format_floating): Same. Adjust indentation.
453 (format_character, format_none): New functions.
454 (format_string): Moved character handling to format_character.
455 (format_directive): Remove arguments, change return type.
456 (parse_directive): New function.
457 (pass_sprintf_length::compute_format_length): Move directive
458 parsing to parse_directive.
460 2017-01-23 Jakub Jelinek <jakub@redhat.com>
462 * tree.h (assign_assembler_name_if_neeeded): Rename to ...
463 (assign_assembler_name_if_needed): ... this.
464 * tree.c (assign_assembler_name_if_neeeded): Rename to ...
465 (assign_assembler_name_if_needed): ... this.
466 (free_lang_data_in_cgraph): Adjust callers.
467 * cgraphunit.c (cgraph_node::analyze): Likewise.
468 * omp-expand.c (expand_omp_taskreg, expand_omp_target):
469 Likewise.
471 2017-01-23 Richard Biener <rguenther@suse.de>
473 PR tree-optimization/79088
474 PR tree-optimization/79188
475 * tree-ssa-threadupdate.c (mark_threaded_blocks): Move code
476 resetting loop bounds after last path deletion. Reset loop
477 bounds of the target loop, make code match the comments.
478 * tree-ssa-threadbackwards.c (pass_early_thread_jumps::execute):
479 Make sure loops need no fixups.
481 2017-01-23 Kelvin Nilsen <kelvin@gcc.gnu.org>
483 * config/rs6000/rs6000-builtin.def (VSIEDPF): Add scalar insert
484 exponent support with double type for first argument.
485 * config/rs6000/rs6000-c.c (altivec_overloaded_builtins): Changed
486 type returned by __builtin_vec_extract_sig,
487 __builtin_vec_extract_sig_sp, and __builtin_vec_extract_sig_dp
488 functions from "vector int" to "vector unsigned int" or from
489 "vector long long int" to "vector unsigned long long int".
490 Changed type returned by __builtin_vec_extract_exp,
491 __builtin_vec_extract_exp_sp, and __builtin_vec_extract_exp_dp
492 functions from "vector int" to "vector unsigned int" or from
493 "vector long long int" to "vector unsigned long long int".
494 Changed return type of __builtin_vec_test_data_class,
495 __builtin_vec_test_data_class_sp, and
496 __builtin_vec_test_data_class_dp from "vector int" to
497 "vector bool int" or from "vector long long int" to "vector bool
498 long long int" and changed second argument type from "unsigned
499 int" to "int". Added new overloaded function forms "vector float
500 __builtin_vec_insert_exp (vector float, vector unsigned int)" and
501 "vector float __builtin_vec_insert_exp_sp (vector float, vector
502 unsigned int)" and "vector double __builtin_vec_insert_exp (vector
503 double, vector unsigned long long int)" and "vector double
504 __builtin_vec_insert_exp_dp (vector double, vector unsigned long
505 long int)". Changed return type of
506 __builtin_scalar_test_data_class and
507 __builtin_scalar_test_data_class_sp and
508 __builtin_scalar_test_data_class_dp from "unsigned int" to "bool
509 int" and changed second argument from "unsigned int" to "int".
510 Changed type returned by __builtin_scalar_test_neg,
511 __builtin_scalar_test_neg_sp, and __builtin_scalar_test_neg_dp
512 from "int" to "bool int". Added new overloaded function form
513 "double __builtin_scalar_insert_exp (double, unsigned long long int)".
514 * config/rs6000/vsx.md (xsiexpdpf): New insn for scalar insert
515 exponent double-precision with floating point first argument.
516 * doc/extend.texi (PowerPC AltiVec Built-in Functions): Adjust
517 documentation of scalar_test_data_class, scalar_test_neg,
518 scalar_extract_sig, scalar_extract_exp, scalar_insert_exp,
519 vector_extract_exp, vec_extract_sig, vec_insert_exp, and
520 vec_test_data_class built-in functions to reflect refinements in
521 their type signatures.
523 2017-01-23 Andreas Tobler <andreast@gcc.gnu.org>
525 * config/aarch64/aarch64.c (aarch64_elf_asm_constructor): Increase
526 size of buf.
527 (aarch64_elf_asm_destructor): Likewise.
529 2017-01-23 Bernd Schmidt <bschmidt@redhat.com>
531 PR rtl-optimization/78634
532 * config/i386/i386.c (ix86_max_noce_ifcvt_seq_cost): New function.
533 (TARGET_MAX_NOCE_IFCVT_SEQ_COST): Define.
534 * ifcvt.c (noce_try_cmove): Add missing cost check.
536 PR rtl-optimization/71724
537 * combine.c (if_then_else_cond): Look for situations where it is
538 beneficial to undo the work of one of the recursive calls.
540 2017-01-23 Bin Cheng <bin.cheng@arm.com>
542 PR tree-optimization/70754
543 * tree-predcom.c (stmt_combining_refs): New parameter INSERT_BEFORE.
544 (reassociate_to_the_same_stmt): New parameter INSERT_BEFORE. Insert
545 combined stmt before it if not NULL.
546 (combine_chains): Process refs reversely and compute dominance point
547 for root ref.
549 2017-01-23 Martin Liska <mliska@suse.cz>
551 PR tree-optimization/79196
552 * tree-ssa-strlen.c (fold_strstr_to_memcmp): Rename to ...
553 (fold_strstr_to_strncmp): ... this. Fold the pattern to strncmp
554 instead of memcmp.
555 (strlen_optimize_stmt): Call the renamed function.
557 2017-01-23 Michael Matz <matz@suse.de>
559 PR tree-optimization/78384
560 * tree-ssa-loop-split.c (patch_loop_exit): Use correct edge.
562 2017-01-23 Richard Biener <rguenther@suse.de>
564 PR tree-optimization/79186
565 * tree-vrp.c (register_new_assert_for): Make sure we've seen
566 both incoming edges before moving an assert.
568 2017-01-23 Martin Jambor <mjambor@suse.cz>
570 * ipa-prop.c (load_from_param_1): Removed.
571 (load_from_unmodified_param): Bits from load_from_param_1 put back
572 here.
573 (load_from_param): Removed.
574 (compute_complex_assign_jump_func): Removed stmt2 and just replaced it
575 with stmt. Reverted back to use of load_from_unmodified_param.
577 2017-01-23 Martin Jambor <mjambor@suse.cz>
579 PR ipa/79108
580 * ipa-prop.h (ipa_param_descriptor): Anotate with with GTY(()).
581 (ipa_node_params): Annotate with GTY((for_user)). Make descriptors
582 field a pointer to garbage collected vector, mark lattices and
583 ipcp_orig_node with GTY((skip)).
584 (ipa_get_param_count): Adjust to descriptors being a pointer.
585 (ipa_get_param): Likewise.
586 (ipa_get_type): Likewise.
587 (ipa_get_param_move_cost): Likewise.
588 (ipa_set_param_used): Likewise.
589 (ipa_get_controlled_uses): Likewise.
590 (ipa_set_controlled_uses): Likewise.
591 (ipa_is_param_used): Likewise.
592 (ipa_node_params_t): Move into garbage collector. New methods insert
593 and remove.
594 (ipa_node_params_sum): Annotate wth GTY(()).
595 (ipa_check_create_node_params): Adjust to ipa_node_params_sum being
596 garbage collected.
597 (ipa_load_from_parm_agg): Adjust declaration.
598 * ipa-icf.c (param_used_p): Adjust to descriptors being a pointer.
599 * ipa-profile.c (ipa_profile): Likewise.
600 * ipa-prop.c (ipa_get_param_decl_index_1): Likewise.
601 (ipa_populate_param_decls): Make descriptors parameter garbage
602 collected.
603 (ipa_dump_param): Adjust to descriptors being a pointer.
604 (ipa_alloc_node_params): Likewise.
605 (ipa_initialize_node_params): Likewise.
606 (load_from_param_1): Make descriptors parameter garbage collected.
607 (load_from_unmodified_param): Likewise.
608 (load_from_param): Likewise.
609 (ipa_load_from_parm_agg): Likewise.
610 (ipa_node_params::~ipa_node_params): Removed.
611 (ipa_free_all_node_params): Remove call to delete operator.
612 (ipa_node_params_t::insert): New.
613 (ipa_node_params_t::remove): Likewise.
614 (ipa_node_params_t::duplicate): Adjust to descriptors being a pointer,
615 copy known_csts and known_contexts vectors.
616 (ipa_read_node_info): Adjust to descriptors being a pointer.
617 (ipcp_modif_dom_walker): Make m_descriptors field garbage
618 collected.
619 (ipcp_transform_function): Make descriptors variable garbage
620 collected.
622 2017-01-23 Andrew Senkevich <andrew.senkevich@intel.com>
624 * config/i386/avx512bwintrin.h: Add k-mask test, kortest intrinsics.
625 * config/i386/avx512dqintrin.h: Ditto.
626 * config/i386/avx512fintrin.h: Ditto.
627 * gcc/config/i386/i386.c: Handle new builtins.
628 * config/i386/i386-builtin.def: Add new builtins.
629 * config/i386/sse.md (ktest<mode>, kortest<mode>): New.
630 (UNSPEC_KORTEST, UNSPEC_KTEST): New.
632 2017-01-23 Jakub Jelinek <jakub@redhat.com>
633 Martin Liska <mliska@suse.cz>
635 * asan.h: Define ASAN_USE_AFTER_SCOPE_ATTRIBUTE.
636 * asan.c (asan_expand_poison_ifn): Support stores and use
637 appropriate ASAN report function.
638 * internal-fn.c (expand_ASAN_POISON_USE): New function.
639 * internal-fn.def (ASAN_POISON_USE): Declare.
640 * tree-into-ssa.c (maybe_add_asan_poison_write): New function.
641 (maybe_register_def): Create ASAN_POISON_USE when sanitizing.
642 * tree-ssa-dce.c (eliminate_unnecessary_stmts): Remove
643 ASAN_POISON calls w/o LHS.
644 * tree-ssa.c (execute_update_addresses_taken): Create clobber
645 for ASAN_MARK (UNPOISON, &x, ...) in order to prevent usage of a LHS
646 from ASAN_MARK (POISON, &x, ...) coming to a PHI node.
647 * gimplify.c (asan_poison_variables): Add attribute
648 use_after_scope_memory to variables that really needs to live
649 in memory.
650 * tree-ssa.c (is_asan_mark_p): Do not rewrite into SSA when
651 having the attribute.
653 2017-01-23 Martin Liska <mliska@suse.cz>
655 * asan.c (create_asan_shadow_var): New function.
656 (asan_expand_poison_ifn): Likewise.
657 * asan.h (asan_expand_poison_ifn): New declaration.
658 * internal-fn.c (expand_ASAN_POISON): Likewise.
659 * internal-fn.def (ASAN_POISON): New builtin.
660 * sanopt.c (pass_sanopt::execute): Expand
661 asan_expand_poison_ifn.
662 * tree-inline.c (copy_decl_for_dup_finish): Make function
663 external.
664 * tree-inline.h (copy_decl_for_dup_finish): Likewise.
665 * tree-ssa.c (is_asan_mark_p): New function.
666 (execute_update_addresses_taken): Rewrite local variables
667 (identified just by use-after-scope as addressable) into SSA.
669 2017-01-22 Gerald Pfeifer <gerald@pfeifer.com>
671 * doc/install.texi (Specific): opensource.apple.com uses https
672 now. Remove trailing slash.
674 2017-01-22 Gerald Pfeifer <gerald@pfeifer.com>
676 * README.Portability: Remove note on an Irix compatibility issue.
678 2017-01-22 Dimitry Andric <dim@FreeBSD.org>
680 * gcov.c (INCLUDE_ALGORITHM): Define.
681 (INCLUDE_VECTOR): Define.
682 No longer include <vector> and <algorithm> directly.
684 2017-01-21 Gerald Pfeifer <gerald@pfeifer.com>
686 * doc/extend.texi (Thread-Local): Change www.akkadia.org reference
687 to https.
688 * doc/invoke.texi (Code Gen Options): Ditto.
690 2017-01-21 Jan Hubicka <hubicka@ucw.cz>
692 PR lto/78407
693 * cfg.c (update_bb_profile_for_threading): Fix updating of probablity.
695 2017-01-21 Bernd Schmidt <bschmidt@redhat.com>
697 rtl-optimization/79125
698 * cprop.c (local_cprop_pass): Handle cases where we make an
699 unconditional trap.
701 2017-01-20 Segher Boessenkool <segher@kernel.crashing.org>
703 PR target/61729
704 PR target/77850
705 * config/rs6000/rs6000.c (rs6000_gimplify_va_arg): Adjust address to
706 read from, for big endian.
708 2017-01-20 Jiong Wang <jiong.wang@arm.com>
710 * config/aarch64/aarch64-builtins.c (aarch64_init_builtins): Register
711 register pauth builtins for LP64 only.
713 2017-01-20 Marek Polacek <polacek@redhat.com>
715 PR c/79152
716 * gimplify.c (should_warn_for_implicit_fallthrough): Handle consecutive
717 non-case labels.
719 2017-01-20 Alexander Monakov <amonakov@ispras.ru>
721 * omp-expand.c (expand_omp_simd): Clear PROP_gimple_lomp_dev regardless
722 of safelen status.
723 * omp-offload.c (pass_omp_device_lower::gate): Use PROP_gimple_lomp_dev.
724 * passes.c (dump_properties): Handle PROP_gimple_lomp_dev.
725 * tree-inline.c (expand_call_inline): Propagate PROP_gimple_lomp_dev.
727 2017-01-20 Kyrylo Tkachov <kyrylo.tkachov@arm.com>
729 PR target/71270
730 * config/arm/arm.c (neon_valid_immediate): Reject vector constants
731 in big-endian mode when they are not a single duplicated value.
733 2017-01-20 Richard Biener <rguenther@suse.de>
735 * BASE-VER: Bump to 7.0.1.
737 2017-01-20 Alexander Monakov <amonakov@ispras.ru>
739 * omp-low.c (omplow_simd_context): New struct. Use it...
740 (lower_rec_simd_input_clauses): ...here and...
741 (lower_rec_input_clauses): ...here to hold common data. Adjust all
742 references to idx, lane, max_vf, is_simt.
744 2017-01-20 Graham Markall <graham.markall@embecosm.com>
746 * config/arc/arc.h (LINK_SPEC): Use arclinux_nps emulation when
747 mcpu=nps400.
749 2017-01-20 Martin Jambor <mjambor@suse.cz>
751 * hsa.h: Renaed to hsa-common.h. Adjusted a comment.
752 * hsa.c: Renaed to hsa-common.c. Change include of gt-hsa.h to
753 gt-hsa-common.h.
754 * Makefile.in (OBJS): Rename hsa.o to hsa-common.o.
755 (GTFILES): Rename hsa.c to hsa-common.c.
756 * hsa-brig.c: Change include of hsa.h to hsa-common.h.
757 * hsa-dump.c: Likewise.
758 * hsa-gen.c: Likewise.
759 * hsa-regalloc.c: Likewise.
760 * ipa-hsa.c: Likewise.
761 * omp-expand.c: Likewise.
762 * omp-low.c: Likewise.
763 * toplev.c: Likewise.
765 2017-01-20 Marek Polacek <polacek@redhat.com>
767 PR c/64279
768 * doc/invoke.texi: Document -Wduplicated-branches.
769 * fold-const.c (operand_equal_p): Handle MODIFY_EXPR, INIT_EXPR,
770 COMPOUND_EXPR, PREDECREMENT_EXPR, PREINCREMENT_EXPR,
771 POSTDECREMENT_EXPR, POSTINCREMENT_EXPR, CLEANUP_POINT_EXPR, EXPR_STMT,
772 STATEMENT_LIST, and RETURN_EXPR. For non-pure non-const functions
773 return 0 only when not OEP_LEXICOGRAPHIC.
774 (fold_build_cleanup_point_expr): Use the expression
775 location when building CLEANUP_POINT_EXPR.
776 * tree-core.h (enum operand_equal_flag): Add OEP_LEXICOGRAPHIC.
777 * tree.c (add_expr): Handle error_mark_node.
779 2017-01-20 Martin Liska <mliska@suse.cz>
781 PR lto/69188
782 * tree-profile.c (init_ic_make_global_vars): Do not call
783 finalize_decl.
784 (gimple_init_gcov_profiler): Likewise.
786 2017-01-20 Martin Liska <mliska@suse.cz>
788 PR ipa/71190
789 * cgraph.h (maybe_create_reference): Remove argument and
790 update comment.
791 * cgraphclones.c (cgraph_node::create_virtual_clone): Remove one
792 argument.
793 * ipa-cp.c (create_specialized_node): Likewise.
794 * symtab.c (symtab_node::maybe_create_reference): Handle
795 VAR_DECLs and ADDR_EXPRs and select ipa_ref_use type.
797 2017-01-20 Martin Liska <mliska@suse.cz>
799 * read-rtl-function.c (function_reader::create_function): Use
800 build_decl instread of build_decl_stat.
802 2017-01-20 Andrew Senkevich <andrew.senkevich@intel.com>
804 * config/i386/avx512bwintrin.h: Add k-mask registers shift intrinsics.
805 * config/i386/avx512dqintrin.h: Ditto.
806 * config/i386/avx512fintrin.h: Ditto.
807 * config/i386/i386-builtin-types.def: Add new types.
808 * gcc/config/i386/i386.c: Handle new types.
809 * config/i386/i386-builtin.def (__builtin_ia32_kshiftliqi)
810 (__builtin_ia32_kshiftlihi, __builtin_ia32_kshiftlisi)
811 (__builtin_ia32_kshiftlidi, __builtin_ia32_kshiftriqi)
812 (__builtin_ia32_kshiftrihi, __builtin_ia32_kshiftrisi)
813 (__builtin_ia32_kshiftridi): New.
814 * config/i386/sse.md (k<code><mode>): Rename *k<code><mode>.
816 2017-01-19 Segher Boessenkool <segher@kernel.crashing.org>
818 PR target/78875
819 PR target/79140
820 * config/rs6000/rs6000.c (TARGET_STACK_PROTECT_GUARD): Unconditionally
821 define to rs6000_init_stack_protect_guard.
822 (rs6000_init_stack_protect_guard): New function.
824 2017-01-19 Matthew Fortune <matthew.fortune@imgtec.com>
825 Yunqiang Su <yunqiang.su@imgtec.com>
827 * config.gcc (supported_defaults): Add madd4.
828 (with_madd4): Add validation.
829 (all_defaults): Add madd4.
830 * config/mips/mips.opt (mmadd4): New option.
831 * gcc/config/mips/mips.h (OPTION_DEFAULT_SPECS): Add a default for
832 mmadd4.
833 (TARGET_CPU_CPP_BUILTINS): Add builtin_define for
834 __mips_no_madd4.
835 (ISA_HAS_UNFUSED_MADD4): Gate with mips_madd4.
836 (ISA_HAS_FUSED_MADD4): Likewise.
837 * gcc/doc/invoke.texi (-mmadd4): Document the new option.
838 * gcc/doc/install.texi (--with-madd4): Document the new option.
840 2017-01-19 Jiong Wang <jiong.wang@arm.com>
842 * config/aarch64/aarch64-builtins.c (enum aarch64_builtins): New
843 entries for AARCH64_PAUTH_BUILTIN_XPACLRI,
844 AARCH64_PAUTH_BUILTIN_PACIA1716, AARCH64_PAUTH_BUILTIN_AUTIA1716.
845 (aarch64_init_pauth_hint_builtins): New.
846 (aarch64_init_builtins): Call aarch64_init_pauth_hint_builtins.
847 (aarch64_expand_builtin): Expand new builtins.
849 2017-01-19 Jiong Wang <jiong.wang@arm.com>
851 * reg-notes.def (CFA_TOGGLE_RA_MANGLE): New reg-note.
852 * combine-stack-adj.c (no_unhandled_cfa): Handle
853 REG_CFA_TOGGLE_RA_MANGLE.
854 * dwarf2cfi.c (dwarf2out_frame_debug): Handle REG_CFA_TOGGLE_RA_MANGLE.
855 * config/aarch64/aarch64.c (aarch64_expand_prologue): Generates DWARF
856 info for return address signing.
857 (aarch64_expand_epilogue): Likewise.
859 2017-01-19 Jiong Wang <jiong.wang@arm.com>
861 * config/aarch64/aarch64-opts.h (aarch64_function_type): New enum.
862 * config/aarch64/aarch64-protos.h
863 (aarch64_return_address_signing_enabled): New declaration.
864 * config/aarch64/aarch64.c (aarch64_return_address_signing_enabled):
865 New function.
866 (aarch64_expand_prologue): Sign return address before it's pushed onto
867 stack.
868 (aarch64_expand_epilogue): Authenticate return address fetched from
869 stack.
870 (aarch64_override_options): Sanity check for ILP32 and ISA level.
871 (aarch64_attributes): New function attributes for "sign-return-address".
872 * config/aarch64/aarch64.md (UNSPEC_AUTI1716, UNSPEC_AUTISP,
873 UNSPEC_PACI1716, UNSPEC_PACISP, UNSPEC_XPACLRI): New unspecs.
874 ("*do_return"): Generate combined instructions according to key index.
875 ("<pauth_mnem_prefix>sp", "<pauth_mnem_prefix1716", "xpaclri"): New.
876 * config/aarch64/iterators.md (PAUTH_LR_SP, PAUTH_17_16): New integer
877 iterators.
878 (pauth_mnem_prefix, pauth_hint_num_a): New integer attributes.
879 * config/aarch64/aarch64.opt (msign-return-address=): New.
880 * doc/extend.texi (AArch64 Function Attributes): Documents
881 "sign-return-address=".
882 * doc/invoke.texi (AArch64 Options): Documents "-msign-return-address=".
884 2017-01-19 Matthew Fortune <matthew.fortune@imgtec.com>
886 * doc/invoke.texi: Add missing -mlxc1-sxc1 options to
887 overall option summary.
889 2017-01-19 Jiong Wang <jiong.wang@arm.com>
891 * config/aarch64/aarch64-arches.def: New entry for "armv8.3-a".
892 * config/aarch64/aarch64.h (AARCH64_FL_V8_3, AARCH64_FL_FOR_ARCH8_3,
893 AARCH64_ISA_V8_3, TARGET_ARMV8_3): New.
894 * doc/invoke.texi (AArch64 Options): Document "armv8.3-a".
896 2017-01-19 Michael Meissner <meissner@linux.vnet.ibm.com>
898 * config/rs6000/rs6000-cpus.def (ISA_3_0_MASKS_SERVER): Enable
899 -mpower9-minmax by default for -mcpu=power9.
900 (ISA_3_MASKS_IEEE): Require -mvsx-small-integer to enable IEEE
901 128-bit floating point.
903 2017-01-20 Alan Modra <amodra@gmail.com>
905 * config/rs6000/rs6000.md (cmpstrnsi, cmpstrsi): Fail if
906 optimizing for size.
908 2017-01-20 Alan Modra <amodra@gmail.com>
910 PR target/79144
911 * config/rs6000/rs6000.c (expand_strn_compare): Get the asm name
912 for strcmp and strncmp from corresponding builtin decl.
914 2017-01-19 Uros Bizjak <ubizjak@gmail.com>
916 * config.gcc (x86_64-*-rtems*): Use i386/rtemself.h
917 instead of i386/rtems-64.h.
918 * config/i386/rtems-64.h: Remove.
920 2017-01-19 Uros Bizjak <ubizjak@gmail.com>
922 PR target/78478
923 Revert:
924 2013-11-05 Uros Bizjak <ubizjak@gmail.com>
926 * config/i386/rtemself.h (LONG_DOUBLE_TYPE_SIZE): New define.
928 2017-01-19 Tamar Christina <tamar.christina@arm.com>
930 * config/aarch64/aarch64.c (aarch64_simd_gen_const_vector_dup):
931 Change int to HOST_WIDE_INT.
932 * config/aarch64/aarch64-protos.h
933 (aarch64_simd_gen_const_vector_dup): Likewise.
934 * config/aarch64/aarch64-simd.md: Add copysign<mode>3.
936 2017-01-19 David Malcolm <dmalcolm@redhat.com>
938 * langhooks-def.h (lhd_type_for_size): New decl.
939 (LANG_HOOKS_TYPE_FOR_SIZE): Define as lhd_type_for_size.
940 * langhooks.c (lhd_type_for_size): New function, taken from
941 lto_type_for_size.
943 2017-01-19 Pat Haugen <pthaugen@us.ibm.com>
945 * config/rs6000/power9.md (power9-alu): Remove 'cmp' type and add
946 define_bypass for CR latency.
947 (power9-cracked-alu): Update bypass latency and remove power9-branch.
948 (power9-alu2): Add define_bypass for CR latency.
949 (power9-cmp): New.
950 (power9-mul): Update insn latency.
951 (power9-mul-compare): Update insn latency, bypass latency and remove
952 power9-branch.
954 2016-01-19 Kyrylo Tkachov <kyrylo.tkachov@arm.com>
956 * config/aarch64/aarch64-protos.h (aarch64_nopcrelative_literal_loads):
957 Delete.
958 * config/aarch64/aarch64.md
959 (aarch64_reload_movcp<GPF_TF:mode><P:mode>): Delete reference to
960 aarch64_nopcrelative_literal_loads.
961 (aarch64_reload_movcp<VALL:mode><P:mode>): Likewise.
963 2017-01-19 Chenghua Xu <paul.hua.gm@gmail.com>
965 * config/mips/mips.h (ISA_HAS_FUSED_MADD4): Enable for
966 TARGET_LOONGSON_3A.
967 (ISA_HAS_UNFUSED_MADD4): Exclude TARGET_LOONGSON_3A.
969 2017-01-19 Doug Gilmore <doug.gilmore@imgtec.com>
971 PR target/78176
972 * config.gcc (supported_defaults): Add lxc1-sxc1.
973 (with_lxc1_sxc1): Add validation.
974 (all_defaults): Add lxc1-sxc1.
975 * config/mips/mips.opt (mlxc1-sxc1): New option.
976 * gcc/config/mips/mips.h (OPTION_DEFAULT_SPECS): Add a default for
977 mlxc1-sxc1.
978 (TARGET_CPU_CPP_BUILTINS): Add builtin_define for
979 __mips_no_lxc1_sxc1.
980 (ISA_HAS_LXC1_SXC1): Gate with mips_lxc1_sxc1.
981 * gcc/doc/invoke.texi (-mlxc1-sxc1): Document the new option.
982 * doc/install.texi (--with-lxc1-sxc1): Document the new option.
984 2017-01-19 Richard Biener <rguenther@suse.de>
986 PR tree-optimization/72488
987 * tree-ssa-sccvn.c (run_scc_vn): When we abort the VN make
988 sure to restore SSA info.
989 * tree-ssa.c (verify_ssa): Verify SSA info is not shared.
991 2017-01-19 Richard Earnshaw <rearnsha@arm.com>
993 PR rtl-optimization/79121
994 * expr.c (expand_expr_real_2, case LSHIFT_EXPR): Look at the signedness
995 of the inner type when shifting an extended value.
997 2017-01-17 Jan Hubicka <hubicka@ucw.cz>
999 PR lto/78407
1000 * symtab.c (symtab_node::equal_address_to): Fix comparing of
1001 interposable aliases.
1003 2017-01-18 Peter Bergner <bergner@vnet.ibm.com>
1005 PR target/78516
1006 * config/rs6000/spe.md (mov_si<mode>_e500_subreg0): Fix constraints.
1007 Use the evmergelohi instruction.
1008 (mov_si<mode>_e500_subreg4_2_le): Likewise.
1009 (mov_sitf_e500_subreg8_2_be): Likewise.
1010 (mov_sitf_e500_subreg12_2_le): Likewise.
1011 (mov_si<mode>_e500_subreg0_2_le): Fix constraints.
1012 (mov_si<mode>_e500_subreg4_2_be): Likewise.
1013 (mov_sitf_e500_subreg8_2_le): Likewise.
1014 (mov_sitf_e500_subreg12_2_be): Likewise.
1016 2017-01-18 Bill Schmidt <wschmidt@linux.vnet.ibm.com>
1018 * config/rs6000/altivec.md (altivec_vbpermq): Change "type"
1019 attribute from vecsimple to vecperm.
1020 (altivec_vbpermq2): Likewise.
1022 2017-01-18 Bill Schmidt <wschmidt@linux.vnet.ibm.com>
1024 PR target/79040
1025 * config/rs6000/altivec.h: Fix typo of vec_cntlz to vec_cnttz.
1027 2017-01-18 Aaron Sawdey <acsawdey@linux.vnet.ibm.com>
1028 * config/rs6000/rs6000-protos.h (expand_strn_compare): Add arg.
1029 * config/rs6000/rs6000.c (expand_strn_compare): Add ability to expand
1030 strcmp. Fix bug where comparison didn't stop with zero byte. Fix
1031 case where N arg is SIZE_MAX.
1032 * config/rs6000/rs6000.md (cmpstrnsi): Args to expand_strn_compare.
1033 (cmpstrsi): Add pattern.
1035 2017-01-18 Michael Meissner <meissner@linux.vnet.ibm.com>
1037 * config/rs6000/rs6000-c.c (altivec_overloaded_builtins): Add
1038 __builtin_vec_revb builtins.
1039 * config/rs6000/rs6000-builtins.def (P9V_BUILTIN_XXBRQ_V16QI): Add
1040 built-in functions to support generation of the ISA 3.0 XXBR<x>
1041 vector byte reverse instructions.
1042 (P9V_BUILTIN_XXBRQ_V1TI): Likewise.
1043 (P9V_BUILTIN_XXBRD_V2DI): Likewise.
1044 (P9V_BUILTIN_XXBRD_V2DF): Likewise.
1045 (P9V_BUILTIN_XXBGW_V4SI): Likewise.
1046 (P9V_BUILTIN_XXBGW_V4SF): Likewise.
1047 (P9V_BUILTIN_XXBGH_V8HI): Likewise.
1048 (P9V_BUILTIN_VEC_REVB): Likewise.
1049 * config/rs6000/vsx.md (p9_xxbrq_v1ti): New insns/expanders to
1050 generate the ISA 3.0 XXBR<x> vector byte reverse instructions.
1051 (p9_xxbrq_v16qi): Likewise.
1052 (p9_xxbrd_<mode>, VSX_D iterator): Likewise.
1053 (p9_xxbrw_<mode>, VSX_W iterator): Likewise.
1054 (p9_xxbrh_v8hi): Likewise.
1055 * config/rs6000/altivec.h (vec_revb): Define if ISA 3.0.
1056 * doc/extend.texi (RS/6000 Altivec Built-ins): Document the
1057 vec_revb built-in functions.
1059 2017-01-18 Uros Bizjak <ubizjak@gmail.com>
1061 PR rtl-optimization/78952
1062 * config/i386/i386.md (any_extract): New code iterator.
1063 (*insvqi_2): Use any_extract for source operand.
1064 (*insvqi_3): Use any_shiftrt for source operand.
1066 2017-01-18 Wilco Dijkstra <wdijkstr@arm.com>
1068 * config/aarch64/aarch64.c (aarch64_sched_adjust_priority)
1069 New function.
1070 (TARGET_SCHED_ADJUST_PRIORITY): Define target hook.
1072 2017-01-18 Matthias Klose <doko@ubuntu.com>
1074 * doc/install.texi: Allow default for --with-target-bdw-gc-include.
1076 2016-01-18 Bill Schmidt <wschmidt@linux.vnet.ibm.com>
1078 * config/rs6000/altivec.h (vec_bperm): Change #define.
1079 * config/rs6000/altivec.md (UNSPEC_VBPERMD): New enum constant.
1080 (altivec_vbpermq2): New define_insn.
1081 (altivec_vbpermd): Likewise.
1082 * config/rs6000/rs6000-builtin.def (VBPERMQ2): New monomorphic
1083 function interface.
1084 (VBPERMD): Likewise.
1085 (VBPERM): New polymorphic function interface.
1086 * config/rs6000/r6000-c.c (altivec_overloaded_builtins_table):
1087 Add entries for P9V_BUILTIN_VEC_VBPERM.
1088 * doc/extend.texi: Add interfaces for vec_bperm.
1090 2017-01-18 Andreas Krebbel <krebbel@linux.vnet.ibm.com>
1092 * config/s390/s390-c.c (s390_expand_overloaded_builtin): Downcase
1093 first letter of error messages.
1094 (s390_resolve_overloaded_builtin): Likewise.
1095 * config/s390/s390.c (s390_expand_builtin): Likewise.
1096 (s390_invalid_arg_for_unprototyped_fn): Likewise.
1097 (s390_valid_target_attribute_inner_p): Likewise.
1098 * config/s390/s390.md ("tabort"): Likewise.
1100 2017-01-18 Toma Tabacu <toma.tabacu@imgtec.com>
1102 * config/mips/mips.h (ISA_HAS_DIV3): Remove unused macro.
1103 (ISA_AVOID_DIV_HILO): New macro.
1104 (ISA_HAS_DIV): Use new ISA_AVOID_DIV_HILO macro.
1105 (ISA_HAS_DDIV): Likewise.
1107 2017-01-18 Markus Trippelsdorf <markus@trippelsdorf.de>
1109 * doc/invoke.texi (fabi-version): Correct number of occurrences.
1111 2017-01-18 Markus Trippelsdorf <markus@trippelsdorf.de>
1113 * doc/invoke.texi (fabi-version): Spelling fix.
1115 2017-01-18 Markus Trippelsdorf <markus@trippelsdorf.de>
1117 PR c++/70182
1118 * doc/invoke.texi (fabi-version): Mention mangling fix for
1119 operator names.
1121 2017-01-18 Markus Trippelsdorf <markus@trippelsdorf.de>
1123 PR c++/77489
1124 * doc/invoke.texi (fabi-version): Document discriminator mangling.
1126 2017-01-17 Segher Boessenkool <segher@kernel.crashing.org>
1128 PR target/78875
1129 * config/rs6000/rs6000-opts.h (stack_protector_guard): New enum.
1130 * config/rs6000/rs6000.c (rs6000_option_override_internal): Handle
1131 the new options.
1132 * config/rs6000/rs6000.md (stack_protect_set): Handle the new more
1133 flexible settings.
1134 (stack_protect_test): Ditto.
1135 * config/rs6000/rs6000.opt (mstack-protector-guard=,
1136 mstack-protector-guard-reg=, mstack-protector-guard-offset=): New
1137 options.
1138 * doc/invoke.texi (Option Summary) [RS/6000 and PowerPC Options]:
1139 Add -mstack-protector-guard=, -mstack-protector-guard-reg=, and
1140 -mstack-protector-guard-offset=.
1141 (RS/6000 and PowerPC Options): Ditto.
1143 2017-01-17 Uros Bizjak <ubizjak@gmail.com>
1145 * config/i386/i386.h (MASK_CLASS_P): New define.
1146 * config/i386/i386.c (inline_secondary_memory_needed): Ensure that
1147 there are no registers from different register sets also when
1148 mask registers are used. Update function comment.
1149 * config/i386/i386.md (*movsi_internal): Split (*k/*krm) alternative
1150 to (*k/*r) and (*k/*km) alternatives.
1152 2017-01-17 Wilco Dijkstra <wdijkstr@arm.com>
1154 * config/aarch64/aarch64.md (eh_return): Remove pattern and splitter.
1155 * config/aarch64/aarch64.h (AARCH64_EH_STACKADJ_REGNUM): Remove.
1156 (EH_RETURN_HANDLER_RTX): New define.
1157 * config/aarch64/aarch64.c (aarch64_frame_pointer_required):
1158 Force frame pointer in EH return functions.
1159 (aarch64_expand_epilogue): Add barrier for eh_return.
1160 (aarch64_final_eh_return_addr): Remove.
1161 (aarch64_eh_return_handler_rtx): New function.
1162 * config/aarch64/aarch64-protos.h (aarch64_final_eh_return_addr):
1163 Remove.
1164 (aarch64_eh_return_handler_rtx): New prototype.
1166 2017-01-17 Bill Schmidt <wschmidt@linux.vnet.ibm.com>
1168 * config/rs6000/altivec.h (vec_rlmi): New #define.
1169 (vec_vrlnm): Likewise.
1170 (vec_rlnm): Likewise.
1171 * config/rs6000/altivec.md (UNSPEC_VRLMI): New UNSPEC enum value.
1172 (UNSPEC_VRLNM): Likewise.
1173 (VIlong): New mode iterator.
1174 (altivec_vrl<VI_char>mi): New define_insn.
1175 (altivec_vrl<VI_char>nm): Likewise.
1176 * config/rs6000/rs6000-builtin.def (VRLWNM): New monomorphic
1177 function entry.
1178 (VRLDNM): Likewise.
1179 (RLNM): New polymorphic function entry.
1180 (VRLWMI): New monomorphic function entry.
1181 (VRLDMI): Likewise.
1182 (RLMI): New polymorphic function entry.
1183 * config/rs6000/r6000-c.c (altivec_overloaded_builtin_table): Add
1184 new entries for P9V_BUILTIN_VEC_RLMI and P9V_BUILTIN_VEC_RLNM.
1185 * doc/extend.texi: Add description of vec_rlmi, vec_rlnm, and
1186 vec_vrlnm.
1188 2017-01-17 Jakub Jelinek <jakub@redhat.com>
1190 PR debug/78839
1191 * dwarf2out.c (field_byte_offset): Restore the
1192 PCC_BITFIELD_TYPE_MATTERS behavior for INTEGER_CST DECL_FIELD_OFFSET
1193 and DECL_FIELD_BIT_OFFSET. Use fold_build2 instead of build2 + fold.
1194 (analyze_variants_discr, gen_variant_part): Use fold_build2 instead
1195 of build2 + fold.
1197 2017-01-17 Eric Botcazou <ebotcazou@adacore.com>
1199 PR ada/67205
1200 * config/aarch64/aarch64.c (TARGET_CUSTOM_FUNCTION_DESCRIPTORS): Define
1202 2017-01-17 Jakub Jelinek <jakub@redhat.com>
1204 PR debug/71669
1205 * dwarf2out.c (add_data_member_location_attribute): For constant
1206 offset bitfield emit for -gdwarf-5 DW_AT_data_bit_offset attribute
1207 instead of DW_AT_data_member_location, DW_AT_bit_offset and
1208 DW_AT_byte_size attributes.
1210 2017-01-17 Eric Botcazou <ebotcazou@adacore.com>
1212 * config/rs6000/rs6000.c (rs6000_emit_move): Also use a TOC reference
1213 after forcing to constant memory when the code model is medium.
1215 2017-01-17 Julia Koval <julia.koval@intel.com>
1217 PR target/76731
1218 * config/i386/avx512fintrin.h
1219 (_mm512_i32gather_ps): Change __addr type to void const*.
1220 (_mm512_mask_i32gather_ps): Ditto.
1221 (_mm512_i32gather_pd): Ditto.
1222 (_mm512_mask_i32gather_pd): Ditto.
1223 (_mm512_i64gather_ps): Ditto.
1224 (_mm512_mask_i64gather_ps): Ditto.
1225 (_mm512_i64gather_pd): Ditto.
1226 (_mm512_mask_i64gather_pd): Ditto.
1227 (_mm512_i32gather_epi32): Ditto.
1228 (_mm512_mask_i32gather_epi32): Ditto.
1229 (_mm512_i32gather_epi64): Ditto.
1230 (_mm512_mask_i32gather_epi64): Ditto.
1231 (_mm512_i64gather_epi32): Ditto.
1232 (_mm512_mask_i64gather_epi32): Ditto.
1233 (_mm512_i64gather_epi64): Ditto.
1234 (_mm512_mask_i64gather_epi64): Ditto.
1235 (_mm512_i32scatter_ps): Change __addr type to void*.
1236 (_mm512_mask_i32scatter_ps): Ditto.
1237 (_mm512_i32scatter_pd): Ditto.
1238 (_mm512_mask_i32scatter_pd): Ditto.
1239 (_mm512_i64scatter_ps): Ditto.
1240 (_mm512_mask_i64scatter_ps): Ditto.
1241 (_mm512_i64scatter_pd): Ditto.
1242 (_mm512_mask_i64scatter_pd): Ditto.
1243 (_mm512_i32scatter_epi32): Ditto.
1244 (_mm512_mask_i32scatter_epi32): Ditto.
1245 (_mm512_i32scatter_epi64): Ditto.
1246 (_mm512_mask_i32scatter_epi64): Ditto.
1247 (_mm512_i64scatter_epi32): Ditto.
1248 (_mm512_mask_i64scatter_epi32): Ditto.
1249 (_mm512_i64scatter_epi64): Ditto.
1250 (_mm512_mask_i64scatter_epi64): Ditto.
1251 * config/i386/avx512pfintrin.h
1252 (_mm512_mask_prefetch_i32gather_pd): Change __addr type to void const*.
1253 (_mm512_mask_prefetch_i32gather_ps): Ditto.
1254 (_mm512_mask_prefetch_i64gather_pd): Ditto.
1255 (_mm512_mask_prefetch_i64gather_ps): Ditto.
1256 (_mm512_prefetch_i32scatter_pd): Change __addr type to void*.
1257 (_mm512_prefetch_i32scatter_ps): Ditto.
1258 (_mm512_mask_prefetch_i32scatter_pd): Ditto.
1259 (_mm512_mask_prefetch_i32scatter_ps): Ditto.
1260 (_mm512_prefetch_i64scatter_pd): Ditto.
1261 (_mm512_prefetch_i64scatter_ps): Ditto.
1262 (_mm512_mask_prefetch_i64scatter_pd): Ditto.
1263 (_mm512_mask_prefetch_i64scatter_ps): Ditto.
1264 * config/i386/avx512vlintrin.h
1265 (_mm256_mmask_i32gather_ps): Change __addr type to void const*.
1266 (_mm_mmask_i32gather_ps): Ditto.
1267 (_mm256_mmask_i32gather_pd): Ditto.
1268 (_mm_mmask_i32gather_pd): Ditto.
1269 (_mm256_mmask_i64gather_ps): Ditto.
1270 (_mm_mmask_i64gather_ps): Ditto.
1271 (_mm256_mmask_i64gather_pd): Ditto.
1272 (_mm_mmask_i64gather_pd): Ditto.
1273 (_mm256_mmask_i32gather_epi32): Ditto.
1274 (_mm_mmask_i32gather_epi32): Ditto.
1275 (_mm256_mmask_i32gather_epi64): Ditto.
1276 (_mm_mmask_i32gather_epi64): Ditto.
1277 (_mm256_mmask_i64gather_epi32): Ditto.
1278 (_mm_mmask_i64gather_epi32): Ditto.
1279 (_mm256_mmask_i64gather_epi64): Ditto.
1280 (_mm_mmask_i64gather_epi64): Ditto.
1281 (_mm256_i32scatter_ps): Change __addr type to void*.
1282 (_mm256_mask_i32scatter_ps): Ditto.
1283 (_mm_i32scatter_ps): Ditto.
1284 (_mm_mask_i32scatter_ps): Ditto.
1285 (_mm256_i32scatter_pd): Ditto.
1286 (_mm256_mask_i32scatter_pd): Ditto.
1287 (_mm_i32scatter_pd): Ditto.
1288 (_mm_mask_i32scatter_pd): Ditto.
1289 (_mm256_i64scatter_ps): Ditto.
1290 (_mm256_mask_i64scatter_ps): Ditto.
1291 (_mm_i64scatter_ps): Ditto.
1292 (_mm_mask_i64scatter_ps): Ditto.
1293 (_mm256_i64scatter_pd): Ditto.
1294 (_mm256_mask_i64scatter_pd): Ditto.
1295 (_mm_i64scatter_pd): Ditto.
1296 (_mm_mask_i64scatter_pd): Ditto.
1297 (_mm256_i32scatter_epi32): Ditto.
1298 (_mm256_mask_i32scatter_epi32): Ditto.
1299 (_mm_i32scatter_epi32): Ditto.
1300 (_mm_mask_i32scatter_epi32): Ditto.
1301 (_mm256_i32scatter_epi64): Ditto.
1302 (_mm256_mask_i32scatter_epi64): Ditto.
1303 (_mm_i32scatter_epi64): Ditto.
1304 (_mm_mask_i32scatter_epi64): Ditto.
1305 (_mm256_i64scatter_epi32): Ditto.
1306 (_mm256_mask_i64scatter_epi32): Ditto.
1307 (_mm_i64scatter_epi32): Ditto.
1308 (_mm_mask_i64scatter_epi32): Ditto.
1309 (_mm256_i64scatter_epi64): Ditto.
1310 (_mm256_mask_i64scatter_epi64): Ditto.
1311 (_mm_i64scatter_epi64): Ditto.
1312 (_mm_mask_i64scatter_epi64): Ditto.
1313 * config/i386/i386-builtin-types.def (V16SF_V16SF_PCFLOAT_V16SI_HI_INT)
1314 (V8DF_V8DF_PCDOUBLE_V8SI_QI_INT, V8SF_V8SF_PCFLOAT_V8DI_QI_INT)
1315 (V8DF_V8DF_PCDOUBLE_V8DI_QI_INT, V16SI_V16SI_PCINT_V16SI_HI_INT)
1316 (V8DI_V8DI_PCINT64_V8SI_QI_INT, V8SI_V8SI_PCINT_V8DI_QI_INT)
1317 (V8DI_V8DI_PCINT64_V8DI_QI_INT, V2DF_V2DF_PCDOUBLE_V4SI_QI_INT)
1318 (V4DF_V4DF_PCDOUBLE_V4SI_QI_INT, V2DF_V2DF_PCDOUBLE_V2DI_QI_INT)
1319 (V4DF_V4DF_PCDOUBLE_V4DI_QI_INT, V4SF_V4SF_PCFLOAT_V4SI_QI_INT)
1320 (V8SF_V8SF_PCFLOAT_V8SI_QI_INT, V4SF_V4SF_PCFLOAT_V2DI_QI_INT)
1321 (V4SF_V4SF_PCFLOAT_V4DI_QI_INT, V2DI_V2DI_PCINT64_V4SI_QI_INT)
1322 (V4DI_V4DI_PCINT64_V4SI_QI_INT, V2DI_V2DI_PCINT64_V2DI_QI_INT)
1323 (V4DI_V4DI_PCINT64_V4DI_QI_INT, V4SI_V4SI_PCINT_V4SI_QI_INT)
1324 (V8SI_V8SI_PCINT_V8SI_QI_INT, V4SI_V4SI_PCINT_V2DI_QI_INT)
1325 (V4SI_V4SI_PCINT_V4DI_QI_INT, VOID_PFLOAT_HI_V16SI_V16SF_INT)
1326 (VOID_PFLOAT_QI_V8SI_V8SF_INT, VOID_PFLOAT_QI_V4SI_V4SF_INT)
1327 (VOID_PDOUBLE_QI_V8SI_V8DF_INT, VOID_PDOUBLE_QI_V4SI_V4DF_INT)
1328 (VOID_PDOUBLE_QI_V4SI_V2DF_INT, VOID_PFLOAT_QI_V8DI_V8SF_INT)
1329 (VOID_PFLOAT_QI_V4DI_V4SF_INT, VOID_PFLOAT_QI_V2DI_V4SF_INT)
1330 (VOID_PDOUBLE_QI_V8DI_V8DF_INT, VOID_PDOUBLE_QI_V4DI_V4DF_INT)
1331 (VOID_PDOUBLE_QI_V2DI_V2DF_INT, VOID_PINT_HI_V16SI_V16SI_INT)
1332 (VOID_PINT_QI_V8SI_V8SI_INT, VOID_PINT_QI_V4SI_V4SI_INT)
1333 (VOID_PLONGLONG_QI_V8SI_V8DI_INT, VOID_PLONGLONG_QI_V4SI_V4DI_INT)
1334 (VOID_PLONGLONG_QI_V4SI_V2DI_INT, VOID_PINT_QI_V8DI_V8SI_INT)
1335 (VOID_PINT_QI_V4DI_V4SI_INT, VOID_PINT_QI_V2DI_V4SI_INT)
1336 (VOID_PLONGLONG_QI_V8DI_V8DI_INT, VOID_QI_V8SI_PCINT64_INT_INT)
1337 (VOID_PLONGLONG_QI_V4DI_V4DI_INT, VOID_PLONGLONG_QI_V2DI_V2DI_INT)
1338 (VOID_HI_V16SI_PCINT_INT_INT, VOID_QI_V8DI_PCINT64_INT_INT)
1339 (VOID_QI_V8DI_PCINT_INT_INT): Remove.
1340 (V16SF_V16SF_PCVOID_V16SI_HI_INT, V8DF_V8DF_PCVOID_V8SI_QI_INT)
1341 (V8SF_V8SF_PCVOID_V8DI_QI_INT, V8DF_V8DF_PCVOID_V8DI_QI_INT)
1342 (V16SI_V16SI_PCVOID_V16SI_HI_INT, V8DI_V8DI_PCVOID_V8SI_QI_INT)
1343 (V8SI_V8SI_PCVOID_V8DI_QI_INT, V8DI_V8DI_PCVOID_V8DI_QI_INT)
1344 (VOID_PVOID_HI_V16SI_V16SF_INT, VOID_PVOID_QI_V8SI_V8DF_INT)
1345 (VOID_PVOID_QI_V8DI_V8SF_INT, VOID_PVOID_QI_V8DI_V8DF_INT)
1346 (VOID_PVOID_HI_V16SI_V16SI_INT, VOID_PVOID_QI_V8SI_V8DI_INT)
1347 (VOID_PVOID_QI_V8DI_V8SI_INT, VOID_PVOID_QI_V8DI_V8DI_INT)
1348 (V2DF_V2DF_PCVOID_V4SI_QI_INT, V4DF_V4DF_PCVOID_V4SI_QI_INT)
1349 (V2DF_V2DF_PCVOID_V2DI_QI_INT, V4DF_V4DF_PCVOID_V4DI_QI_INT
1350 (V4SF_V4SF_PCVOID_V4SI_QI_INT, V8SF_V8SF_PCVOID_V8SI_QI_INT)
1351 (V4SF_V4SF_PCVOID_V2DI_QI_INT, V4SF_V4SF_PCVOID_V4DI_QI_INT)
1352 (V2DI_V2DI_PCVOID_V4SI_QI_INT, V4DI_V4DI_PCVOID_V4SI_QI_INT)
1353 (V2DI_V2DI_PCVOID_V2DI_QI_INT, V4DI_V4DI_PCVOID_V4DI_QI_INT)
1354 (V4SI_V4SI_PCVOID_V4SI_QI_INT, V8SI_V8SI_PCVOID_V8SI_QI_INT)
1355 (V4SI_V4SI_PCVOID_V2DI_QI_INT, V4SI_V4SI_PCVOID_V4DI_QI_INT)
1356 (VOID_PVOID_QI_V8SI_V8SF_INT, VOID_PVOID_QI_V4SI_V4SF_INT)
1357 (VOID_PVOID_QI_V4SI_V4DF_INT, VOID_PVOID_QI_V4SI_V2DF_INT)
1358 (VOID_PVOID_QI_V4DI_V4SF_INT, VOID_PVOID_QI_V2DI_V4SF_INT)
1359 (VOID_PVOID_QI_V4DI_V4DF_INT, VOID_PVOID_QI_V2DI_V2DF_INT)
1360 (VOID_PVOID_QI_V8SI_V8SI_INT, VOID_PVOID_QI_V4SI_V4SI_INT)
1361 (VOID_PVOID_QI_V4SI_V4DI_INT, VOID_PVOID_QI_V4SI_V2DI_INT)
1362 (VOID_PVOID_QI_V4DI_V4SI_INT, VOID_PVOID_QI_V2DI_V4SI_INT)
1363 (VOID_PVOID_QI_V4DI_V4DI_INT, VOID_PVOID_QI_V2DI_V2DI_INT)
1364 (VOID_QI_V8SI_PCVOID_INT_INT, VOID_HI_V16SI_PCVOID_INT_INT)
1365 (VOID_QI_V8DI_PCVOID_INT_INT): Add.
1366 * config/i386/i386.c (ix86_init_mmx_sse_builtins): Adjust builtin
1367 definitions accordingly.
1369 2017-01-17 Kito Cheng <kito.cheng@gmail.com>
1370 Kuan-Lin Chen <kuanlinchentw@gmail.com>
1372 PR target/79079
1373 * internal-fn.c (expand_mul_overflow): Use convert_modes instead of
1374 gen_lowpart.
1376 2017-01-17 Vladimir Makarov <vmakarov@redhat.com>
1378 PR target/79058
1379 * ira-conflicts.c (ira_build_conflicts): Update total conflict
1380 hard regs for inner regno.
1382 2017-01-17 Martin Liska <mliska@suse.cz>
1384 PR ipa/71207
1385 * ipa-polymorphic-call.c (contains_type_p): Fix wrong
1386 assumption and add comment.
1388 2017-01-17 Nathan Sidwell <nathan@acm.org>
1390 * ipa-visibility.c (localize_node): New function, broken out of ...
1391 (function_and_variable_visibility): ... here. Call it.
1393 2017-01-17 Jan Hubicka <hubicka@ucw.cz>
1395 PR middle-end/77445
1396 * tree-ssa-threadupdate.c (remove_ctrl_stmt_and_useless_edges):
1397 correctly set frequency of oudgoing edge.
1398 (duplicate_thread_path): Fix profile updating.
1400 2017-01-17 Jakub Jelinek <jakub@redhat.com>
1402 PR other/79046
1403 * configure.ac: Add GCC_BASE_VER.
1404 * Makefile.in (version): Use @get_gcc_base_ver@ instead of cat to get
1405 version from BASE-VER file.
1406 (CFLAGS-gcc.o): Add -DBASEVER=$(BASEVER_s).
1407 (gcc.o): Depend on $(BASEVER).
1408 * common.opt (dumpfullversion): New option.
1409 * gcc.c (driver_handle_option): Handle OPT_dumpfullversion.
1410 * doc/invoke.texi: Document -dumpfullversion.
1411 * doc/install.texi: Document --with-gcc-major-version-only.
1412 * configure: Regenerated.
1414 2017-01-17 Richard Biener <rguenther@suse.de>
1416 PR tree-optimization/71433
1417 * tree-vrp.c (register_new_assert_for): Merge same asserts
1418 on all incoming edges.
1419 (process_assert_insertions_for): Handle insertions at the
1420 beginning of BBs.
1422 2017-01-17 Gerald Pfeifer <gerald@pfeifer.com>
1424 * config/i386/cygwin.h (LIBGCJ_SONAME): Remove.
1425 * config/i386/mingw32.h (LIBGCJ_SONAME): Remove.
1427 2017-01-17 Kaz Kojima <kkojima@gcc.gnu.org>
1429 PR target/78633
1430 * config/sh/sh.md (cmpeqsi_t+1): Call copy_rtx to avoid invalid
1431 RTL sharing.
1433 2017-01-17 Alan Modra <amodra@gmail.com>
1435 PR target/79066
1436 * config/rs6000/rs6000.md (elf_high, elf_low): Disable when pic.
1437 * config/rs6000/rs6000.c (rs6000_emit_allocate_stack): Don't allow
1438 symbolic stack limit when pic.
1440 2017-01-16 Martin Sebor <msebor@redhat.com>
1442 PR tree-optimization/78608
1443 * gimple-ssa-sprintf.c (tree_digits): Avoid negating TYPE_MIN.
1445 2017-01-16 Jeff Law <law@redhat.com>
1447 Revert:
1448 2016-12-02 Tadek Kijkowski <tkijkowski@gmail.com>
1449 * Makefile.in (PREPROCESSOR_DEFINES): Add a level of indirection
1450 for several include directories that may be relative to sysroot.
1451 * config/i386/x-mingw32 (gplus_includedir): Define.
1452 (gplus_tool_includedir, gplus_backward_include_dir): Likewise.
1453 (native_system_includedir): Likewise.
1454 * config/i386/mingw32.h (STANDARD_STARTFILE_PREFIX_1): Do not
1455 override if TARGET_SYSTEM_ROOT is defined.
1456 (NATIVE_SYSTEM_HEADER_DIR): Likewise.
1458 PR tree-optimization/79090
1459 PR tree-optimization/33562
1460 PR tree-optimization/61912
1461 PR tree-optimization/77485
1462 * tree-ssa-dse.c (compute_trims): Accept STMT argument. Dump STMT
1463 and computed trims into the dump file.
1465 2017-01-17 Uros Bizjak <ubizjak@gmail.com>
1467 * config/i386/i386.h (LIMIT_RELOAD_CLASS): Remove.
1469 2017-01-16 Jakub Jelinek <jakub@redhat.com>
1471 PR c/79089
1472 * gimplify.c (gimplify_init_constructor): If want_value and
1473 object == lhs, unshare lhs to avoid invalid tree sharing. Formatting
1474 fix.
1476 PR target/79080
1477 * loop-doloop.c (doloop_modify): Call unshare_all_rtl_in_chain on
1478 sequence. Formatting fixes.
1479 (doloop_optimize): Formatting fixes.
1481 PR driver/49726
1482 * gcc.c (debug_level_greater_than_spec_func): New function.
1483 (static_spec_functions): Add debug-level-gt spec function.
1484 (ASM_DEBUG_SPEC, cpp_options): Use %:debug-level-gt(0) instead of
1485 !g0.
1486 * config/darwin.h (DSYMUTIL_SPEC, ASM_DEBUG_SPEC): Likewise.
1487 * config/darwin9.h (DSYMUTIL_SPEC, ASM_DEBUG_SPEC): Likewise.
1488 * common.opt (g, gcoff, gdwarf, gdwarf-, ggdb, gno-pubnames,
1489 gpubnames, ggnu-pubnames, gno-record-gcc-switches,
1490 grecord-gcc-switches, gno-strict-dwarf, gstrict-dwarf, gstabs,
1491 gstabs+, gtoggle, gvms, gxcoff, gxcoff+): Add Driver flag.
1493 2017-01-16 Uros Bizjak <ubizjak@gmail.com>
1495 * config/i386/i386.h (HARD_REGNO_CALLER_SAVE_MODE): Apply HImode and
1496 QImode fixups to general and mask registers only.
1498 2017-01-16 Carl Love <cel@us.ibm.com>
1500 * config/rs6000/rs6000-c (altivec_overloaded_builtins): Add support
1501 for built-in functions
1502 vector signed char vec_nabs (vector signed char)
1503 vector signed short vec_nabs (vector signed short)
1504 vector signed int vec_nabs (vector signed int)
1505 vector signed long long vec_nabs (vector signed long long)
1506 vector float vec_nabs (vector float)
1507 vector double vec_nabs (vector double)
1508 * config/rs6000/rs6000-builtin.def: Add definitions for NABS functions
1509 and NABS overload.
1510 * config/rs6000/altivec.md: New define_expand nabs<mode>2 types
1511 * config/rs6000/altivec.h: New define for vec_nabs built-in function.
1512 * doc/extend.texi: Update the documentation file for the new built-in
1513 functions.
1515 2017-01-16 Martin Sebor <msebor@redhat.com>
1517 * gimple-ssa-sprintf.c (format_directive): Correct a typo in a warning
1518 message.
1520 2017-01-16 Bill Schmidt <wschmidt@linux.vnet.ibm.com>
1522 * config/rs6000/rs6000.c (rtx_is_swappable_p): Change
1523 UNSPEC_VSX__XXSPLTD to require special splat handling.
1525 2017-01-16 David Malcolm <dmalcolm@redhat.com>
1527 PR bootstrap/78616
1528 * system.h: Poison strndup.
1530 2017-01-16 Alan Modra <amodra@gmail.com>
1532 PR target/79098
1533 * config/rs6000/rs6000.c (rs6000_legitimate_combined_insn): Don't
1534 use a switch.
1536 2017-01-16 Georg-Johann Lay <avr@gjlay.de>
1538 * config/avr/avr.h (BRANCH_COST) [reload_completed]: Increase by 4.
1540 2017-01-15 Uros Bizjak <ubizjak@gmail.com>
1542 * config/i386/i386.c (ix86_legitimate_combined_insn): Do not
1543 call recog here. Assert that INSN_CODE (insn) is non-negative.
1545 2017-01-15 Segher Boessenkool <segher@kernel.crashing.org>
1547 PR target/72749
1548 * cfgrtl.c (rtl_split_edge): Also patch jump insns that jump to the
1549 fallthrough.
1550 * haifa-sched.c (dump_insn_stream): Don't crash if there is a label
1551 in the currently scheduled RTL fragment.
1553 2017-01-15 Segher Boessenkool <segher@kernel.crashing.org>
1555 PR rtl-optimization/78751
1556 * ifcvt.c (find_cond_trap): If we generated a non-existing insn,
1557 give up.
1559 2017-01-14 Jeff Law <law@redhat.com>
1561 PR tree-optimization/79090
1562 * tree-ssa-dse.c (valid_ao_ref_for_dse): Reject zero length and
1563 variable length stores.
1564 (compute_trims): Delete dead assignment to *trim_tail.
1565 (dse_dom_walker::dse_optimize_stmt): Optimize mem* calls with
1566 zero length.
1568 2017-01-14 Bernd Schmidt <bschmidt@redhat.com>
1570 PR rtl-optimization/78626
1571 PR rtl-optimization/78727
1572 * cprop.c (one_cprop_pass): Collect unconditional traps in the middle
1573 of a block, and split such blocks after everything else is finished.
1575 2017-01-14 Alan Modra <amodra@gmail.com>
1577 PR target/72749
1578 * combine.c (recog_for_combine_1): Set INSN_CODE before calling
1579 target legitimate_combined_insn.
1580 * config/rs6000/rs6000.c (TARGET_LEGITIMATE_COMBINED_INSN): Define.
1581 (rs6000_legitimate_combined_insn): New function.
1582 * config/rs6000/rs6000.md (UNSPEC_DOLOOP): Delete, and remove
1583 all uses.
1584 (ctr<mode>_internal3): Rename from *ctr<mode>_internal5.
1585 (ctr<mode>_internal4): Rename from *ctr<mode>_internal6.
1586 (ctr<mode>_internal1, ctr<mode>_internal2): Remove '*' from name.
1588 2017-01-14 Gerald Pfeifer <gerald@pfeifer.com>
1590 * doc/frontends.texi (G++ and GCC): Remove references to Java.
1592 2017-01-13 Jeff Law <law@redhat.com>
1594 PR tree-optimization/33562
1595 PR tree-optimization/61912
1596 PR tree-optimization/77485
1597 * tree-ssa-dse.c (delete_dead_call): Accept gsi rather than
1598 a statement.
1599 (delete_dead_assignment): Likewise.
1600 (dse_dom_walker::dse_optimize_stmt): Pass in the gsi rather than
1601 statement to delete_dead_call and delete_dead_assignment.
1603 2017-01-13 David Malcolm <dmalcolm@redhat.com>
1605 PR c/78304
1606 * substring-locations.c (format_warning_va): Strengthen case 1 so
1607 that both endpoints of the substring must be within the format
1608 range for just the substring to be printed.
1610 2017-01-13 Uros Bizjak <ubizjak@gmail.com>
1612 * config/i386/i386.opt (msgx): Use ix86_isa_flags2 variable.
1613 * config/i386/i386.c (ix86_target_string): Add missing options
1614 to isa_opts and reorder options by implied ISAs. Rename isa_opts2 to
1615 isa2_opts, ix86_flag_opts to flag2_opts, ix86_target_other to
1616 flags_other and ix86_target_other to flags2_other. Display unknown
1617 isa2 options.
1618 (ix86_valid_target_attribute_inner_p): Add missing options and
1619 reorder options by implied ISAs, as in ix86_target_string.
1621 2017-01-13 Richard Sandiford <richard.sandiford@arm.com>
1623 * hash-table.h (hash_table::too_empty_p): New function.
1624 (hash_table::expand): Use it.
1625 (hash_table::traverse): Likewise.
1626 (hash_table::empty_slot): Use sizeof (value_type) instead of
1627 sizeof (PTR) to convert bytes to elements. Shrink the table
1628 if the current size is excessive for the current number of
1629 elements.
1631 2017-01-13 Richard Sandiford <richard.sandiford@arm.com>
1633 * ira-costs.c (record_reg_classes): Break from the inner loop
1634 early once alt_fail is known to be true. Update outer loop
1635 handling accordingly.
1637 2017-01-13 Jeff Law <law@redhat.com>
1639 * tree-ssa-dse.c (decrement_count): New function.
1640 (increment_start_addr, maybe_trim_memstar_call): Likewise.
1641 (dse_dom_walker::optimize_stmt): Call maybe_trim_memstar_call directly
1642 when we know the partially dead statement is a mem* function.
1644 PR tree-optimization/61912
1645 PR tree-optimization/77485
1646 * tree-ssa-dse.c: Include expr.h.
1647 (maybe_trim_constructor_store): New function.
1648 (maybe_trim_partially_dead_store): Call maybe_trim_constructor_store.
1650 PR tree-optimization/33562
1651 PR tree-optimization/61912
1652 PR tree-optimization/77485
1653 * doc/invoke.texi: Document new dse-max-object-size param.
1654 * params.def (PARM_DSE_MAX_OBJECT_SIZE): New PARAM.
1655 * tree-ssa-dse.c: Include params.h.
1656 (dse_store_status): New enum.
1657 (initialize_ao_ref_for_dse): New, partially extracted from
1658 dse_optimize_stmt.
1659 (valid_ao_ref_for_dse, normalize_ref): New.
1660 (setup_live_bytes_from_ref, compute_trims): Likewise.
1661 (clear_bytes_written_by, maybe_trim_complex_store): Likewise.
1662 (maybe_trim_partially_dead_store): Likewise.
1663 (maybe_trim_complex_store): Likewise.
1664 (dse_classify_store): Renamed from dse_possibly_dead_store_p.
1665 Track what bytes live from the original store. Return tri-state
1666 for dead, partially dead or live.
1667 (dse_dom_walker): Add constructor, destructor and new private members.
1668 (delete_dead_call, delete_dead_assignment): New extracted from
1669 dse_optimize_stmt.
1670 (dse_optimize_stmt): Make a member of dse_dom_walker.
1671 Use initialize_ao_ref_for_dse.
1673 PR tree-optimization/33562
1674 PR tree-optimization/61912
1675 PR tree-optimization/77485
1676 * sbitmap.h (bitmap_count_bits): Prototype.
1677 (bitmap_clear_range, bitmap_set_range): Likewise.
1678 * sbitmap.c (bitmap_clear_range): New function.
1679 (bitmap_set_range, sbitmap_popcount, bitmap_count_bits): Likewise.
1681 2017-01-13 Martin Liska <mliska@suse.cz>
1683 PR ipa/79043
1684 * function.c (set_cfun): Add new argument force.
1685 * function.h (set_cfun): Likewise.
1686 * ipa-inline-transform.c (inline_call): Use the function when
1687 strict alising from is dropped for function we inline to.
1689 2017-01-13 Richard Biener <rguenther@suse.de>
1691 * tree-pretty-print.c (dump_generic_node): Fix inverted condition
1692 for dumping GIMPLE INTEGER_CSTs.
1694 2017-01-13 Rainer Orth <ro@CeBiTec.Uni-Bielefeld.DE>
1696 * config/sol2.h (TARGET_OS_CPP_BUILTINS): Define __STDC_VERSION__
1697 to 201112L since C++17.
1699 2017-01-13 Maxim Ostapenko <m.ostapenko@samsung.com>
1701 PR sanitizer/78887
1702 * asan.c (asan_needs_odr_indicator_p): Don't emit ODR indicators
1703 if -fsanitize=kernel-address is present.
1705 2017-01-13 Richard Biener <rguenther@suse.de>
1707 * tree-pretty-print.c (dump_generic_node): Dump INTEGER_CSTs
1708 as _Literal ( type ) number in case usual suffixes do not
1709 preserve all information.
1711 2017-01-13 Richard Biener <rguenther@suse.de>
1713 PR tree-optimization/77283
1714 * gimple-ssa-split-paths.c: Include gimple-ssa.h, tree-phinodes.h
1715 and ssa-iterators.h.
1716 (is_feasible_trace): Implement a cost model based on joiner
1717 PHI node uses.
1719 2017-01-12 Michael Meissner <meissner@linux.vnet.ibm.com>
1721 PR target/79004
1722 * config/rs6000/rs6000.md (FP_ISA3): Do not optimize converting
1723 char or short to __float128/_Float128 directly.
1725 2017-01-12 Martin Sebor <msebor@redhat.com>
1727 to -Wformat-overflow.
1728 * gimple-ssa-sprintf.c (pass_sprintf_length::gate): Adjust.
1729 (min_bytes_remaining): Same.
1730 (get_string_length): Same.
1731 (format_string): Same.
1732 (format_directive): Same.
1733 (add_bytes): Same.
1734 (pass_sprintf_length::handle_gimple_call): Same.
1736 2017-01-12 Jakub Jelinek <jakub@redhat.com>
1738 * gimple-ssa-sprintf.c (try_substitute_return_value): Remove
1739 info.nowrite calls with no lhs that can't throw. Return bool
1740 whether gsi_remove has been called or not.
1741 (pass_sprintf_length::handle_gimple_call): Return bool whether
1742 try_substitute_return_value called gsi_remove. Formatting fix.
1743 (pass_sprintf_length::execute): Don't use gsi_remove if
1744 handle_gimple_call returned true.
1746 PR bootstrap/79069
1747 * cfgrtl.c (rtl_tidy_fallthru_edge): For any_uncondjump_p that can't
1748 be removed due to side-effects, don't remove following barrier nor
1749 turn the successor edge into fallthru edge.
1751 2017-01-12 Bill Schmidt <wschmidt@linux.vnet.ibm.com>
1753 PR target/79044
1754 * config/rs6000/rs6000.c (insn_is_swappable_p): Mark
1755 element-reversing loads and stores as not swappable.
1757 2017-01-12 Nathan Sidwell <nathan@acm.org>
1758 Nicolai Stange <nicstange@gmail.com>
1760 * combine.c (try_combine): Don't ignore result of overlap checking
1761 loop. Combine overlap & asm check into single loop.
1763 2017-01-12 Richard Biener <rguenther@suse.de>
1765 * tree-pretty-print.c (dump_generic_node): Provide -gimple
1766 variant for MEM_REF. Sanitize INTEGER_CST for -gimple.
1768 2017-01-12 Richard Biener <rguenther@suse.de>
1770 * tree.c (initialize_tree_contains_struct): Make TS_OPTIMIZATION
1771 and TS_TARGET_OPTION directly derive from TS_BASE.
1772 * tree-core.h (tree_optimization_option): Derive from tree_base.
1773 (tree_target_option): Likewise.
1775 2017-01-11 Uros Bizjak <ubizjak@gmail.com>
1777 * config/i386/i386.c (memory_address_length): Increase len
1778 only when rip_relative_addr_p returns false.
1780 2017-01-11 Julia Koval <julia.koval@intel.com>
1782 * common/config/i386/i386-common.c (OPTION_MASK_ISA_SGX_UNSET): New.
1783 (OPTION_MASK_ISA_SGX_SET): New.
1784 (ix86_handle_option): Handle OPT_msgx.
1785 * config.gcc: Added sgxintrin.h.
1786 * config/i386/driver-i386.c (host_detect_local_cpu): Detect sgx.
1787 * config/i386/i386-c.c (ix86_target_macros_internal): Define __SGX__.
1788 * config/i386/i386.c (ix86_target_string): Add -msgx.
1789 (PTA_SGX): New.
1790 (ix86_option_override_internal): Handle new options.
1791 (ix86_valid_target_attribute_inner_p): Add sgx.
1792 * config/i386/i386.h (TARGET_SGX, TARGET_SGX_P): New.
1793 * config/i386/i386.opt: Add msgx.
1794 * config/i386/sgxintrin.h: New file.
1795 * config/i386/x86intrin.h: Add sgxintrin.h.
1797 2017-01-11 Jakub Jelinek <jakub@redhat.com>
1799 PR c++/71537
1800 * fold-const.c (maybe_nonzero_address): Return 1 for function
1801 local objects.
1802 (tree_single_nonzero_warnv_p): Don't handle function local objects
1803 here.
1805 PR c++/72813
1806 * gcc.c (default_compilers): Don't add -o %g.s for -S -save-temps
1807 of c-header.
1809 2017-01-11 David Malcolm <dmalcolm@redhat.com>
1811 PR driver/78877
1812 * opts.c: Include "spellcheck.h"
1813 (struct string_fragment): New struct.
1814 (struct edit_distance_traits<const string_fragment &>): New
1815 struct.
1816 (get_closest_sanitizer_option): New function.
1817 (parse_sanitizer_options): Offer suggestions for unrecognized arguments.
1819 2017-01-11 Jakub Jelinek <jakub@redhat.com>
1821 * dwarf2out.c (DWARF_COMPILE_UNIT_HEADER_SIZE): For DWARF5 decrease
1822 by 12.
1823 (DWARF_COMDAT_TYPE_UNIT_HEADER_SIZE): Always
1824 DWARF_COMPILE_UNIT_HEADER_SIZE plus 12.
1825 (DWARF_COMPILE_UNIT_SKELETON_HEADER_SIZE): Define.
1826 (calc_base_type_die_sizes): Use DWARF_COMPILE_UNIT_SKELETON_HEADER_SIZE
1827 for initial die_offset if dwarf_split_debug_info.
1828 (output_comp_unit): Use DWARF_COMPILE_UNIT_SKELETON_HEADER_SIZE for
1829 initial next_die_offset if dwo_id is non-NULL. Don't emit padding
1830 fields.
1831 (output_skeleton_debug_sections): Formatting fix. Use
1832 DWARF_COMPILE_UNIT_SKELETON_HEADER_SIZE instead of
1833 DWARF_COMPILE_UNIT_HEADER_SIZE. Don't emit padding.
1835 2017-01-11 Wilco Dijkstra <wdijkstr@arm.com>
1837 * config/arm/cortex-a53.md: Add bypasses for
1838 cortex_a53_r2f_cvt.
1839 (cortex_a53_r2f): Only use for transfers.
1840 (cortex_a53_f2r): Likewise.
1841 (cortex_a53_r2f_cvt): Add reservation for conversions.
1842 (cortex_a53_f2r_cvt): Likewise.
1844 2017-01-11 Tamar Christina <tamar.christina@arm.com>
1846 * config/arm/arm_neon.h: Add __artificial__ and gnu_inline
1847 to all inlined functions, change static to extern.
1849 2017-01-11 Christophe Lyon <christophe.lyon@linaro.org>
1851 PR target/78253
1852 * config/arm/arm.c (legitimize_pic_address): Handle reference to
1853 weak symbol.
1854 (arm_assemble_integer): Likewise.
1856 2017-01-11 Richard Earnshaw <rearnsha@arm.com>
1858 * config.gcc: Use new awk script to check CPU, FPU and architecture
1859 parameters for --with-... options.
1860 * config/arm/parsecpu.awk: New file
1861 * config/arm/arm-cpus.in: New file.
1862 * config/arm/arm-opts.h: Include arm-cpu.h instead of processing .def
1863 files.
1864 * config/arm/arm.c: Include arm-cpu-data.h instead of processing .def
1865 files.
1866 * config/arm/t-arm: Update dependency rules.
1867 * common/config/arm/arm-common.c: Include arm-cpu-cdata.h instead
1868 of processing .def files.
1869 * config/arm/genopt.sh: Deleted.
1870 * config/arm/gentune.sh: Deleted.
1871 * config/arm/arm-cores.def: Deleted.
1872 * config/arm/arm-arches.def: Deleted.
1873 * config/arm/arm-fpus.def: Deleted.
1874 * config/arm/arm-tune.md: Regenerated.
1875 * config/arm/arm-tables.opt: Regenerated.
1876 * config/arm/arm-cpu.h: New generated file.
1877 * config/arm/arm-cpu-data.h: New generated file.
1878 * config/arm/arm-cpu-cdata.h: New generated file.
1880 2017-01-11 Maxim Ostapenko <m.ostapenko@samsung.com>
1882 PR lto/79042
1883 * lto-cgraph.c (lto_output_varpool_node): Pack dynamically_initialized
1884 bit.
1885 (input_varpool_node): Unpack dynamically_initialized bit.
1887 2017-01-11 Eric Botcazou <ebotcazou@adacore.com>
1889 PR rtl-optimization/79032
1890 * lra-constraints.c (simplify_operand_subreg): In the MEM case, test
1891 the alignment of the adjusted memory reference against that of MODE,
1892 instead of the alignment of the original memory reference.
1894 2017-01-11 Martin Jambor <mjambor@suse.cz>
1896 * hsa.c (hsa_callable_function_p): Revert addition of DECL_ARTIFICIAL
1897 test.
1898 * ipa-hsa.c (process_hsa_functions): Only duplicate non-artificial
1899 decorated functions.
1901 2017-01-11 Richard Biener <rguenther@suse.de>
1903 * tree-vrp.c (evrp_dom_walker::before_dom_children): Also
1904 set range/nonnull info for PHI results. Do not set it on
1905 stmts marked for removal.
1907 2017-01-10 Eric Botcazou <ebotcazou@adacore.com>
1909 * expr.c (store_field): In the bitfield case, fetch the return value
1910 from the registers before applying a single big-endian adjustment.
1911 Always do a final load for a BLKmode value not larger than a word.
1913 2017-01-10 David Malcolm <dmalcolm@redhat.com>
1915 PR c++/77949
1916 * input.c (selftest::test_accessing_ordinary_linemaps): Verify
1917 that we correctly handle column numbers greater than
1918 LINE_MAP_MAX_COLUMN_NUMBER.
1920 2017-01-10 Martin Sebor <msebor@redhat.com>
1922 PR middle-end/78245
1923 * gimple-ssa-sprintf.c (get_destination_size): Call
1924 {init,fini}object_sizes.
1925 * tree-object-size.c (addr_object_size): Adjust.
1926 (pass_through_call): Adjust.
1927 (pass_object_sizes::execute): Adjust.
1928 * tree-object-size.h (fini_object_sizes): Declare.
1930 2017-01-10 Martin Sebor <msebor@redhat.com>
1932 PR tree-optimization/78775
1933 * builtins.c (get_size_range): Move...
1934 * calls.c: ...to here.
1935 (alloc_max_size): Accept zero argument.
1936 (operand_signed_p): Remove.
1937 (maybe_warn_alloc_args_overflow): Call get_size_range.
1938 * calls.h (get_size_range): Declare.
1940 2017-01-10 Joe Seymour <joe.s@somniumtech.com>
1942 * config/msp430/driver-msp430.c (msp430_mcu_data): Sync with data
1943 from TI's devices.csv file as of September 2016.
1944 * config/msp430/msp430.c (msp430_mcu_data): Likewise.
1946 2017-01-10 Sandra Loosemore <sandra@codesourcery.com>
1948 * doc/extend.texi: Tweak formatting to fix overfull hbox warnings.
1949 * doc/invoke.texi: Likewise.
1950 * doc/md.texi: Likewise.
1951 * doc/objc.texi: Likewise.
1953 2017-01-10 Joshua Conner <joshconner@google.com>
1955 * config/arm/fuchsia-elf.h: New file.
1956 * config/fuchsia.h: New file.
1957 * config.gcc (*-*-fuchsia*): Set native_system_header_dir.
1958 (aarch64*-*-fuchsia*, arm*-*-fuchsia*, x86_64-*-fuchsia*): Add to
1959 targets.
1960 * config.host: (aarch64*-*-fuchsia*, arm*-*-fuchsia*): Add to hosts.
1962 2016-01-10 Richard Biener <rguenther@suse.de>
1964 PR tree-optimization/79034
1965 * tree-call-cdce.c (shrink_wrap_one_built_in_call_with_conds):
1966 Propagate out degenerate PHIs in the joiner.
1968 2017-01-10 Martin Liska <mliska@suse.cz>
1970 * ipa-icf.c (sort_sem_items_by_decl_uid): New function.
1971 (sort_congruence_classes_by_decl_uid): Likewise.
1972 (sort_congruence_class_groups_by_decl_uid): Likewise.
1973 (sem_item_optimizer::merge_classes): Sort class, groups in these
1974 classes and members in the groups by DECL_UID of declarations.
1975 This would make merge operations stable.
1977 2017-01-10 Martin Liska <mliska@suse.cz>
1979 * ipa-icf.c (sem_item_optimizer::sem_item_optimizer): Remove
1980 usage of m_classes_vec.
1981 (sem_item_optimizer::~sem_item_optimizer): Likewise.
1982 (sem_item_optimizer::get_group_by_hash): Likewise.
1983 (sem_item_optimizer::subdivide_classes_by_equality): Likewise.
1984 (sem_item_optimizer::subdivide_classes_by_sensitive_refs): Likewise.
1985 (sem_item_optimizer::verify_classes): Likewise.
1986 (sem_item_optimizer::process_cong_reduction): Likewise.
1987 (sem_item_optimizer::dump_cong_classes): Likewise.
1988 (sem_item_optimizer::merge_classes): Likewise.
1989 * ipa-icf.h (congruence_class_hash): Rename from
1990 congruence_class_group_hash. Remove declaration of m_classes_vec.
1992 2017-01-10 Andrew Senkevich <andrew.senkevich@intel.com>
1994 * common/config/i386/i386-common.c (OPTION_MASK_ISA_AVX512VPOPCNTDQ_SET,
1995 OPTION_MASK_ISA_AVX512VPOPCNTDQ_UNSET): New.
1996 * config.gcc: Add avx512vpopcntdqintrin.h.
1997 * config/i386/avx512vpopcntdqintrin.h: New.
1998 * config/i386/cpuid.h (bit_AVX512VPOPCNTDQ): New.
1999 * config/i386/i386-builtin-types.def: Add new types.
2000 * config/i386/i386-builtin.def (__builtin_ia32_vpopcountd_v16si,
2001 __builtin_ia32_vpopcountd_v16si_mask, __builtin_ia32_vpopcountq_v8di,
2002 __builtin_ia32_vpopcountq_v8di_mask): New.
2003 * config/i386/i386-c.c (ix86_target_macros_internal): Define
2004 __AVX512VPOPCNTDQ__.
2005 * config/i386/i386.c (ix86_target_string): Add -mavx512vpopcntdq.
2006 (PTA_AVX512VPOPCNTDQ): Define.
2007 * config/i386/i386.h (TARGET_AVX512VPOPCNTDQ,
2008 TARGET_AVX512VPOPCNTDQ_P): Define.
2009 * config/i386/i386.opt: Add mavx512vpopcntdq.
2010 * config/i386/immintrin.h: Include avx512vpopcntdqintrin.h.
2011 * config/i386/sse.md (define_insn "vpopcount<mode><mask_name>"): New.
2013 2017-01-01 Jan Hubicka <hubicka@ucw.cz>
2015 PR middle-end/77484
2016 * predict.def (PRED_CALL): Set to 67.
2018 2017-01-09 Eric Botcazou <ebotcazou@adacore.com>
2020 * expr.c (store_field): In the bitfield case, if the value comes from
2021 a function call and is of an aggregate type returned in registers, do
2022 not modify the field mode; extract the value in all cases if the mode
2023 is BLKmode and the size is not larger than a word.
2025 2017-01-09 Dominique d'Humieres <dominiq@lps.ens.fr>
2027 PR target/71017
2028 * config/i386/cpuid.h: Fix undefined behavior.
2030 2017-01-04 Jeff Law <law@redhat.com>
2032 PR tree-optimization/79007
2033 PR tree-optimization/67955
2034 * tree-ssa-alias.c (same_addr_size_stores_p): Only need to be
2035 conservative for pt.null when flag_non_call_exceptions is on.
2037 2017-01-09 Jakub Jelinek <jakub@redhat.com>
2039 PR translation/79019
2040 PR translation/79020
2041 * params.def (PARAM_INLINE_MIN_SPEEDUP,
2042 PARAM_IPA_CP_SINGLE_CALL_PENALTY,
2043 PARAM_USE_AFTER_SCOPE_DIRECT_EMISSION_THRESHOLD): Fix typos
2044 in descriptions.
2045 * config/avr/avr.opt (maccumulate-args): Likewise.
2046 * config/msp430/msp430.opt (mwarn-mcu): Likewise.
2047 * common.opt (freport-bug): Likewise.
2048 * cif-code.def (CIF_FINAL_ERROR): Likewise.
2049 * doc/invoke.texi (ipa-cp-single-call-penalty): Likewise.
2050 * config/s390/s390.c (s390_invalid_binary_op): Fix spelling in
2051 translatable string.
2052 * config/i386/i386.c (function_value_32): Likewise.
2053 * config/nios2/nios2.c (nios2_valid_target_attribute_rec): Likewise.
2054 * config/msp430/msp430.c (msp430_option_override, msp430_attr):
2055 Likewise.
2056 * config/msp430/driver-msp430.c (msp430_select_hwmult_lib): Likewise.
2057 * common/config/msp430/msp430-common.c (msp430_handle_option):
2058 Likewise.
2059 * symtab.c (symtab_node::verify_base): Likewise.
2060 * opts.c (set_debug_level): Likewise.
2061 * tree.c (verify_type_variant): Likewise. Fix typo in comment.
2062 * config/rs6000/rs6000-c.c (altivec_resolve_overloaded_builtin): Add
2063 missing whitespace to translatable strings.
2064 * config/avr/avr.md (bswapsi2): Fix typo in comment.
2065 * config/sh/superh.h: Likewise.
2066 * config/i386/xopintrin.h: Likewise.
2067 * config/i386/znver1.md: Likewise.
2068 * config/rs6000/rs6000.c (struct rs6000_opt_mask): Likewise.
2069 * ipa-inline-analysis.c (compute_inline_parameters): Likewise.
2070 * double-int.h (struct double_int): Likewise.
2071 * double-int.c (div_and_round_double): Likewise.
2072 * wide-int.cc: Likewise.
2073 * tree-ssa.c (non_rewritable_mem_ref_base): Likewise.
2074 * tree-ssa-sccvn.c (vn_reference_lookup_3): Likewise.
2075 * cfgcleanup.c (crossjumps_occured): Renamed to ...
2076 (crossjumps_occurred): ... this.
2077 (try_crossjump_bb, try_head_merge_bb, try_optimize_cfg, cleanup_cfg):
2078 Adjust all uses.
2080 PR tree-optimization/78899
2081 * tree-if-conv.c (version_loop_for_if_conversion): Instead of
2082 returning bool return struct loop *, NULL for failure and the new
2083 loop on success.
2084 (versionable_outer_loop_p): Don't version outer loop if it has
2085 dont_vectorized bit set.
2086 (tree_if_conversion): When versioning outer loop, ensure
2087 tree_if_conversion is performed also on the inner loop of the
2088 non-vectorizable outer loop copy.
2089 * tree-vectorizer.c (set_uid_loop_bbs): Formatting fix. Fold
2090 LOOP_VECTORIZED in inner loop of the scalar outer loop and
2091 prevent vectorization of it.
2092 (vectorize_loops): For outer + inner LOOP_VECTORIZED, ensure
2093 the outer loop vectorization of the non-scalar version is attempted
2094 before vectorization of the inner loop in scalar version. If
2095 outer LOOP_VECTORIZED guarded loop is not vectorized, prevent
2096 vectorization of its inner loop.
2097 * tree-vect-loop-manip.c (rename_variables_in_bb): If outer_loop
2098 has 2 inner loops, rename also on edges from bb whose single pred
2099 is outer_loop->header. Fix typo in function comment.
2101 2017-01-09 Martin Sebor <msebor@redhat.com>
2103 PR bootstrap/79033
2104 * asan.c (asan_emit_stack_protection): Increase local buffer size
2105 to avoid snprintf truncation warning.
2107 2017-01-09 Andrew Pinski <apinski@cavium.com>
2109 * config/aarch64/aarch64-cores.def: Add thunderx2t99. Change vulcan
2110 to reference thunderx2t99 for the tuning structure
2111 * config/aarch64/aarch64-cost-tables.h (vulcan_extra_costs):
2112 Rename to ...
2113 (thunderx2t99_extra_costs): This.
2114 * config/aarch64/aarch64-tune.md: Regenerate.
2115 * config/aarch64/aarch64.c (vulcan_addrcost_table): Rename to ...
2116 (vulcan_addrcost_table): This.
2117 (vulcan_regmove_cost): Rename to ...
2118 (thunderx2t99_regmove_cost): This.
2119 (vulcan_vector_cost): Rename to ...
2120 (thunderx2t99_vector_cost): this.
2121 (vulcan_branch_cost): Rename to ...
2122 (thunderx2t99_branch_cost): This.
2123 (vulcan_tunings): Rename to ...
2124 (thunderx2t99_tunings): This and s/vulcan/thunderx2t99 .
2125 * doc/invoke.texi (AARCH64/mtune): Add thunderx2t99.
2127 2017-01-09 Martin Jambor <mjambor@suse.cz>
2129 PR ipa/78365
2130 PR ipa/78599
2131 * ipa-prop.h (ipa_jump_func): Swap positions of vr_known and m_vr.
2132 * ipa-cp.c (ipa_vr_operation_and_type_effects): New function.
2133 (propagate_vr_accross_jump_function): Use the above function for all
2134 value range computations for pass-through jump functions and type
2135 converasion from explicit value range values.
2136 (ipcp_propagate_stage): Do not attempt to deduce types of formal
2137 parameters from TYPE_ARG_TYPES.
2138 * ipa-prop.c (ipa_write_jump_function): Remove trailing whitespace.
2139 (ipa_write_node_info): Stream type of the actual argument.
2140 (ipa_read_node_info): Likewise. Also remove trailing whitespace.
2142 2017-01-09 Martin Liska <mliska@suse.cz>
2144 PR pch/78970
2145 * gcc.c (driver_handle_option): Handle OPT_E and set have_E.
2146 (lookup_compiler): Do not show error message with have_E.
2148 2017-01-09 Jakub Jelinek <jakub@redhat.com>
2150 PR tree-optimization/78938
2151 * tree-vect-stmts.c (vectorizable_condition): For non-masked COND_EXPR
2152 where comp_vectype is VECTOR_BOOLEAN_TYPE_P, use
2153 BIT_{NOT,XOR,AND,IOR}_EXPR on the comparison operands instead of
2154 {EQ,NE,GE,GT,LE,LT}_EXPR directly inside of VEC_COND_EXPR. Formatting
2155 fixes.
2157 2017-01-09 Kyrylo Tkachov <kyrylo.tkachov@arm.com>
2159 * tree-ssa-address.c (gen_addr_rtx): Don't handle index if it
2160 is const0_rtx.
2162 2017-01-09 Richard Biener <rguenther@suse.de>
2164 PR tree-optimization/78997
2165 * tree-vect-slp.c (vect_mask_constant_operand_p): Handle SSA
2166 name condition properly.
2168 2017-01-09 Richard Biener <rguenther@suse.de>
2170 PR debug/79000
2171 * dwarf2out.c (is_cxx): New overload with context.
2172 (is_naming_typedef_decl): Use it.
2174 2017-01-08 Sandra Loosemore <sandra@codesourcery.com>
2176 * invoke.texi (Option Summary): Correct spacing in option lists
2177 and add line breaks to fix over-long lines.
2179 2017-01-08 Sandra Loosemore <sandra@codesourcery.com>
2181 PR middle-end/17660
2183 * extend.texi (Common Variable Attributes): Add xref to GCC
2184 Internals manual to explain mode attribute keywords.
2186 2017-01-08 Sandra Loosemore <sandra@codesourcery.com>
2188 PR other/16519
2189 * doc/invoke.texi (Option Summary): Move -pthread to Linker Options
2190 and Preprocessor Options.
2191 (Options for Linking): Document -pthread here....
2192 (RS/6000 and PowerPC Options): ...not here.
2193 (Solaris 2 Options): ...or here.
2194 * doc/cppopts.texi: Document -pthread.
2196 2017-01-08 Martin Sebor <msebor@redhat.com>
2198 PR middle-end/77708
2199 * doc/invoke.texi (Warning Options): Document -Wformat-truncation.
2200 * gimple-ssa-sprintf.c (call_info::reval_used, call_info::warnopt):
2201 New member functions.
2202 (format_directive): Used them.
2203 (add_bytes): Same.
2204 (pass_sprintf_length::handle_gimple_call): Same.
2205 * graphite-sese-to-poly.c (tree_int_to_gmp): Increase buffer size
2206 to avoid truncation for any argument.
2207 (extract_affine_mul): Same.
2208 * tree.c (get_file_function_name): Same.
2210 2017-01-01 Jan Hubicka <hubicka@ucw.cz>
2212 PR middle-end/77484
2213 * predict.def (PRED_INDIR_CALL): Set to 86.
2215 2017-01-07 Sandra Loosemore <sandra@codesourcery.com>
2217 PR preprocessor/54124
2218 * doc/cppopts.texi: Reformat -d subtable to list the full name
2219 of the options. Add cross-reference to the docs for the general
2220 compiler -d options.
2221 * doc/invoke.texi (Developer Options): Add cross-reference to the
2222 preprocessor-specific -d option documentation.
2224 2017-01-07 Sandra Loosemore <sandra@codesourcery.com>
2226 PR preprocessor/13498
2227 * doc/cpp.texi (Search Path): Rewrite to remove obsolete and
2228 redudant material, and reflect new command-line options.
2229 (System Headers): Likewise.
2231 2017-01-07 Sandra Loosemore <sandra@codesourcery.com>
2233 * doc/cppdiropts.texi: Merge documentation of -I, -iquote,
2234 -isystem, and -idirafter. Copy-edit.
2235 * doc/cppopts.texi: Copy-edit. Remove contradiction about
2236 default for -ftrack-macro-expansion. Delete obsolete and
2237 badly-formatted implementation details about -fdebug-cpp output.
2238 * doc/cppwarnopts.texi: Copy-edit.
2240 2017-01-07 David Malcolm <dmalcolm@redhat.com>
2242 PR c++/72803
2243 * input.c (selftest::test_accessing_ordinary_linemaps): Verify
2244 that the transition from a max line width >= 1<<10 to narrower
2245 lines works correctly.
2247 2017-01-07 Alexandre Oliva <aoliva@redhat.com>
2249 * doc/options.texi (PerFunction): New.
2250 * opt-functions.awk (switch_flags): Map both Optimization and
2251 PerFunction to CL_OPTIMIZATION.
2252 * opth-gen.awk: Test for PerFunction flag along with
2253 Optimization.
2254 * optc-save-gen.awk: Likewise. Introduce var_opt_hash and set
2255 it only when the latter is present. Skip those that don't in
2256 the hash function generator.
2257 * common.opt (fvar-tracking): Mark as PerFunction instead of
2258 Optimization.
2259 (fvar-tracking-assignments): Likewise.
2260 (fvar-tracking-assignments-toggle): Likewise.
2261 (fvar-tracking-uninit): Likewise.
2263 2017-01-07 Jakub Jelinek <jakub@redhat.com>
2265 PR translation/79018
2266 * params.def (PARAM_MAX_STORES_TO_MERGE): Add missing space between
2267 the and store.
2269 2017-01-06 Mikael Pettersson <mikpelinux@gmail.com>
2271 PR target/57583
2272 * config/m68k/m68k.opt (LONG_JUMP_TABLE_OFFSETS): New option.
2273 * config/m68k/linux.h (ASM_RETURN_CASE_JUMP): Handle
2274 TARGET_LONG_JUMP_TABLE_OFFSETS.
2275 * config/m68k/m68kelf.h (ASM_RETURN_CASE_JUMP): Likewise.
2276 * config/m68k/netbsd-elf.h (ASM_RETURN_CASE_JUMP): Likewise.
2277 * config/m68k/m68k.h (CASE_VECTOR_MODE): Likewise.
2278 (ASM_OUTPUT_ADDR_DIFF_ELF): Likewise.
2279 * config/m68k/m68k.md (tablejump expander): Likewise.
2280 (*tablejump_pcrel_hi): Renamed from unnamed insn, reject
2281 TARGET_LONG_JUMP_TABLE_OFFSETS.
2282 (*tablejump_pcrel_si): New insn, handle TARGET_LONG_JUMP_TABLE_OFFSETS.
2283 * doc/invoke.texi (M68K options): Add -mlong-jump-table-offsets.
2285 2017-01-06 Edgar E. Iglesias <edgar.iglesias@xilinx.com>
2286 David Holsgrove <david.holsgrove@xilinx.com>
2288 * common/config/microblaze/microblaze-common.c
2289 (TARGET_EXCEPT_UNWIND_INFO): Remove.
2290 * config/microblaze/microblaze-protos.h (microblaze_eh_return):
2291 New prototype.
2292 * config/microblaze/microblaze.c (microblaze_must_save_register)
2293 (microblaze_expand_epilogue, microblaze_return_addr): Handle
2294 calls_eh_return.
2295 (microblaze_eh_return): New function.
2296 * config/microblaze/microblaze.h (RETURN_ADDR_OFFSET)
2297 (EH_RETURN_DATA_REGNO, MB_EH_STACKADJ_REGNUM)
2298 (EH_RETURN_STACKADJ_RTX, ASM_PREFERRED_EH_DATA_FORMAT): New macros.
2299 * config/microblaze/microblaze.md (eh_return): New pattern.
2301 2017-01-06 Jakub Jelinek <jakub@redhat.com>
2303 * system.h (GCC_DIAGNOSTIC_PUSH_IGNORED, GCC_DIAGNOSTIC_POP,
2304 GCC_DIAGNOSTIC_STRINGIFY): Define.
2306 * read-rtl.c (rtx_reader::read_rtx_code): Avoid -Wsign-compare warning.
2308 2017-01-06 Andre Vieira <andre.simoesdiasvieira@arm.com>
2310 * config/arm/arm.md (<mcrr>): New.
2311 (<mrrc>): New.
2312 * config/arm/arm.c (arm_arch5te): New.
2313 (arm_option_override): Set arm_arch5te.
2314 (arm_coproc_builtin_available): Add support for mcrr, mcrr2, mrrc
2315 and mrrc2.
2316 * config/arm/arm-builtins.c (MCRR_QUALIFIERS): Define to...
2317 (arm_mcrr_qualifiers): ... this. New.
2318 (MRRC_QUALIFIERS): Define to...
2319 (arm_mrrc_qualifiers): ... this. New.
2320 * config/arm/arm_acle.h (__arm_mcrr, __arm_mcrr2, __arm_mrrc,
2321 __arm_mrrc2): New.
2322 * config/arm/arm_acle_builtins.def (mcrr, mcrr2, mrrc, mrrc2): New.
2323 * config/arm/iterators.md (MCRRI, mcrr, MCRR): New.
2324 (MRRCI, mrrc, MRRC): New.
2325 * config/arm/unspecs.md (VUNSPEC_MCRR, VUNSPEC_MCRR2, VUNSPEC_MRRC,
2326 VUNSPEC_MRRC2): New.
2328 2017-01-06 Andre Vieira <andre.simoesdiasvieira@arm.com>
2330 * config/arm/arm.md (<mcr>): New.
2331 (<mrc>): New.
2332 * config/arm/arm.c (arm_coproc_builtin_available): Add
2333 support for mcr, mrc, mcr2 and mrc2.
2334 * config/arm/arm-builtins.c (MCR_QUALIFIERS): Define to...
2335 (arm_mcr_qualifiers): ... this. New.
2336 (MRC_QUALIFIERS): Define to ...
2337 (arm_mrc_qualifiers): ... this. New.
2338 (MCR_QUALIFIERS): Define to ...
2339 (arm_mcr_qualifiers): ... this. New.
2340 * config/arm/arm_acle.h (__arm_mcr, __arm_mrc, __arm_mcr2,
2341 __arm_mrc2): New.
2342 * config/arm/arm_acle_builtins.def (mcr, mcr2, mrc, mrc2): New.
2343 * config/arm/iterators.md (MCRI, mcr, MCR, MRCI, mrc, MRC): New.
2344 * config/arm/unspecs.md (VUNSPEC_MCR, VUNSPEC_MCR2, VUNSPEC_MRC,
2345 VUNSPEC_MRC2): New.
2347 2017-01-06 Andre Vieira <andre.simoesdiasvieira@arm.com>
2349 * config/arm/arm.md (*ldc): New.
2350 (*stc): New.
2351 (<ldc>): New.
2352 (<stc>): New.
2353 * config/arm/arm.c (arm_coproc_builtin_available): Add
2354 support for ldc,ldcl,stc,stcl,ldc2,ldc2l,stc2 and stc2l.
2355 (arm_coproc_ldc_stc_legitimate_address): New.
2356 * config/arm/arm-builtins.c (arm_type_qualifiers): Add
2357 'qualifier_const_pointer'.
2358 (LDC_QUALIFIERS): Define to...
2359 (arm_ldc_qualifiers): ... this. New.
2360 (STC_QUALIFIERS): Define to...
2361 (arm_stc_qualifiers): ... this. New.
2362 * config/arm/arm-protos.h
2363 (arm_coproc_ldc_stc_legitimate_address): New.
2364 * config/arm/arm_acle.h (__arm_ldc, __arm_ldcl, __arm_stc,
2365 __arm_stcl, __arm_ldc2, __arm_ldc2l, __arm_stc2, __arm_stc2l): New.
2366 * config/arm/arm_acle_builtins.def (ldc, ldc2, ldcl, ldc2l, stc,
2367 stc2, stcl, stc2l): New.
2368 * config/arm/constraints.md (Uz): New.
2369 * config/arm/iterators.md (LDCI, STCI, ldc, stc, LDC STC): New.
2370 * config/arm/unspecs.md (VUNSPEC_LDC, VUNSPEC_LDC2, VUNSPEC_LDCL,
2371 VUNSPEC_LDC2L, VUNSPEC_STC, VUNSPEC_STC2, VUNSPEC_STCL,
2372 VUNSPEC_STC2L): New.
2374 2017-01-06 Andre Vieira <andre.simoesdiasvieira@arm.com>
2376 * config/arm/arm.md (<cdp>): New.
2377 * config/arm/arm.c (neon_const_bounds): Rename this ...
2378 (arm_const_bounds): ... this.
2379 (arm_coproc_builtin_available): New.
2380 * config/arm/arm-builtins.c (SIMD_MAX_BUILTIN_ARGS): Increase.
2381 (arm_type_qualifiers): Add 'qualifier_unsigned_immediate'.
2382 (CDP_QUALIFIERS): Define to...
2383 (arm_cdp_qualifiers): ... this. New.
2384 (void_UP): Define.
2385 (arm_expand_builtin_args): Add case for 6 arguments.
2386 * config/arm/arm-protos.h (neon_const_bounds): Rename this ...
2387 (arm_const_bounds): ... this.
2388 (arm_coproc_builtin_available): New.
2389 * config/arm/arm_acle.h (__arm_cdp): New.
2390 (__arm_cdp2): New.
2391 * config/arm/arm_acle_builtins.def (cdp): New.
2392 (cdp2): New.
2393 * config/arm/iterators.md (CDPI,CDP,cdp): New.
2394 * config/arm/neon.md: Rename all 'neon_const_bounds' to
2395 'arm_const_bounds'.
2396 * config/arm/types.md (coproc): New.
2397 * config/arm/unspecs.md (VUNSPEC_CDP, VUNSPEC_CDP2): New.
2398 * gcc/doc/extend.texi (ACLE): Add a mention of Coprocessor intrinsics.
2399 * gcc/doc/sourcebuild.texi (arm_coproc1_ok, arm_coproc2_ok,
2400 arm_coproc3_ok, arm_coproc4_ok): Document new effective targets.
2402 2017-01-06 Andre Vieira <andre.simoesdiasvieira@arm.com>
2404 * config/arm/arm-builtins.c (arm_unsigned_binop_qualifiers): New.
2405 (UBINOP_QUALIFIERS): New.
2406 (si_UP): Define.
2407 (acle_builtin_data): New. Change comment.
2408 (arm_builtins): Remove ARM_BUILTIN_CRC32B, ARM_BUILTIN_CRC32H,
2409 ARM_BUILTIN_CRC32W, ARM_BUILTIN_CRC32CB, ARM_BUILTIN_CRC32CH,
2410 ARM_BUILTIN_CRC32CW. Add ARM_BUILTIN_ACLE_BASE and include
2411 arm_acle_builtins.def.
2412 (ARM_BUILTIN_ACLE_PATTERN_START): Define.
2413 (arm_init_acle_builtins): New.
2414 (CRC32_BUILTIN): Remove.
2415 (bdesc_2arg): Remove entries for crc32b, crc32h, crc32w,
2416 crc32cb, crc32ch and crc32cw.
2417 (arm_init_crc32_builtins): Remove.
2418 (arm_init_builtins): Use arm_init_acle_builtins rather
2419 than arm_init_crc32_builtins.
2420 (arm_expand_acle_builtin): New.
2421 (arm_expand_builtin): Use 'arm_expand_acle_builtin'.
2422 * config/arm/arm_acle_builtins.def: New.
2424 2017-01-06 Andre Vieira <andre.simoesdiasvieira@arm.com>
2426 * config/arm/arm-builtins.c (neon_builtin_datum): Rename to ..
2427 (arm_builtin_datum): ... this.
2428 (arm_init_neon_builtin): Rename to ...
2429 (arm_init_builtin): ... this. Add a new parameters PREFIX
2430 and USE_SIG_IN_NAME.
2431 (arm_init_neon_builtins): Replace 'arm_init_neon_builtin' with
2432 'arm_init_builtin'. Replace type 'neon_builtin_datum' with
2433 'arm_builtin_datum'.
2434 (arm_init_vfp_builtins): Likewise.
2435 (builtin_arg): Rename enum's replacing 'NEON_ARG' with
2436 'ARG_BUILTIN' and add a 'ARG_BUILTIN_NEON_MEMORY.
2437 (arm_expand_neon_args): Rename to ...
2438 (arm_expand_builtin_args): ... this. Rename builtin_arg
2439 enum values and differentiate between ARG_BUILTIN_MEMORY
2440 and ARG_BUILTIN_NEON_MEMORY.
2441 (arm_expand_neon_builtin_1): Rename to ...
2442 (arm_expand_builtin_1): ... this. Rename builtin_arg enum
2443 values, arm_expand_builtin_args and add bool parameter NEON.
2444 (arm_expand_neon_builtin): Use arm_expand_builtin_1.
2445 (arm_expand_vfp_builtin): Likewise.
2446 (NEON_MAX_BUILTIN_ARGS): Remove, it was unused.
2448 2017-01-01 Jan Hubicka <hubicka@ucw.cz>
2450 PR middle-end/77484
2451 * predict.def (PRED_POLYMORPHIC_CALL): Set to 59.
2452 * predict.c (tree_estimate_probability_bb): Reverse direction of
2453 polymorphic call predictor.
2455 2017-01-06 David Malcolm <dmalcolm@redhat.com>
2457 * passes.c (execute_one_pass): Split out pass-skipping logic into...
2458 (determine_pass_name_match): ...this new function and...
2459 (should_skip_pass_p): ...this new function.
2461 2017-01-06 Nathan Sidwell <nathan@acm.org>
2463 * ipa-visibility.c (function_and_variable_visibility): Reformat
2464 comments and long lines. Remove extrneous if.
2465 * symtab.c (symtab_node::make_decl_local): Fix code format.
2466 (symtab_node::set_section_for_node): Fix comment typo.
2468 2017-01-06 Martin Liska <mliska@suse.cz>
2470 PR bootstrap/79003
2471 * lra-constraints.c: Rename invariant to lra_invariant.
2472 * predict.c (set_even_probabilities): Initialize e to NULL.
2474 2017-01-05 Martin Sebor <msebor@redhat.com>
2476 PR tree-optimization/78910
2477 * gimple-ssa-sprintf.c (tree_digits): Add an argument.
2478 (format_integer): Correct off-by-one error in the handling
2479 of precision with negative numbers in signed conversions..
2481 2017-01-05 Eric Botcazou <ebotcazou@adacore.com>
2483 * doc/invoke.texi (C Dialect Options): Adjust -fsso-struct entry.
2485 2017-01-05 Jakub Jelinek <jakub@redhat.com>
2487 PR tree-optimization/71016
2488 * tree-ssa-phiopt.c (tree_ssa_phiopt_worker): Pass cond_stmt to
2489 factor_out_conditional_conversion. Formatting fix.
2490 (factor_out_conditional_conversion): Add cond_stmt argument.
2491 If arg1 is INTEGER_CST, punt if new_arg0 is not any operand of
2492 cond_stmt and if arg0_def_stmt is not the only stmt in its bb.
2493 Formatting fix.
2495 2017-01-05 David Malcolm <dmalcolm@redhat.com>
2497 * Makefile.in (OBJS): Add read-md.o, read-rtl.o,
2498 read-rtl-function.o, and selftest-rtl.o.
2499 * config/aarch64/aarch64.c: Include selftest.h and selftest-rtl.h.
2500 (selftest::aarch64_test_loading_full_dump): New function.
2501 (selftest::aarch64_run_selftests): New function.
2502 (TARGET_RUN_TARGET_SELFTESTS): Wire it up to
2503 selftest::aarch64_run_selftests.
2504 * config/i386/i386.c
2505 (selftest::ix86_test_loading_dump_fragment_1): New function.
2506 (selftest::ix86_test_loading_call_insn): New function.
2507 (selftest::ix86_test_loading_full_dump): New function.
2508 (selftest::ix86_test_loading_unspec): New function.
2509 (selftest::ix86_run_selftests): Call the new functions.
2510 * emit-rtl.c (maybe_set_max_label_num): New function.
2511 * emit-rtl.h (maybe_set_max_label_num): New decl.
2512 * function.c (instantiate_decls): Guard call to
2513 instantiate_decls_1 with if (DECL_INITIAL (fndecl)).
2514 * function-tests.c (selftest::verify_three_block_rtl_cfg): Remove
2515 "static".
2516 * gensupport.c (gen_reader::gen_reader): Pass "false"
2517 for new "compact" param of rtx_reader.
2518 * print-rtl.c (rtx_writer::print_rtx_operand): Print "(nil)"
2519 rather than an empty string for NULL strings.
2520 * read-md.c: Potentially include config.h rather than bconfig.h.
2521 Wrap include of errors.h with #ifdef GENERATOR_FILE.
2522 (have_error): New global, copied from errors.c.
2523 (md_reader::read_name): Rename to...
2524 (md_reader::read_name_1): ...this, adding "out_loc" param,
2525 and converting "missing name or number" to returning false, rather
2526 than failing.
2527 (md_reader::read_name): Reimplement in terms of read_name_1.
2528 (md_reader::read_name_or_nil): New function.
2529 (md_reader::read_string): Handle "(nil)" by returning NULL.
2530 (md_reader::md_reader): Add new param "compact".
2531 (md_reader::read_md_files): Wrap with #ifdef GENERATOR_FILE.
2532 (md_reader::read_file): New method.
2533 * read-md.h (md_reader::md_reader): Add new param "compact".
2534 (md_reader::read_file): New method.
2535 (md_reader::is_compact): New accessor.
2536 (md_reader::read_name): Convert return type from void to file_location.
2537 (md_reader::read_name_or_nil): New decl.
2538 (md_reader::read_name_1): New decl.
2539 (md_reader::m_compact): New field.
2540 (noop_reader::noop_reader): Pass "false" for new "compact" param
2541 of rtx_reader.
2542 (rtx_reader::rtx_reader): Add new "compact" param.
2543 (rtx_reader::read_rtx_operand): Make virtual and convert return
2544 type from void to rtx.
2545 (rtx_reader::read_until): New decl.
2546 (rtx_reader::handle_any_trailing_information): New virtual function.
2547 (rtx_reader::postprocess): New virtual function.
2548 (rtx_reader::finalize_string): New virtual function.
2549 (rtx_reader::m_in_call_function_usage): New field.
2550 (rtx_reader::m_reuse_rtx_by_id): New field.
2551 * read-rtl-function.c: New file.
2552 * selftest-rtl.c (selftest::assert_rtx_ptr_eq_at): New function.
2553 * selftest-rtl.h (ASSERT_RTX_PTR_EQ): New macro.
2554 (selftest::verify_three_block_rtl_cfg): New decl.
2555 * read-rtl-function.h: New file.
2556 * read-rtl.c: Potentially include config.h rather than bconfig.h.
2557 For host, include function.h, memmodel.h, and emit-rtl.h.
2558 (one_time_initialization): New function.
2559 (struct compact_insn_name): New struct.
2560 (compact_insn_names): New array.
2561 (find_code): Handle insn codes in compact dumps.
2562 (apply_subst_iterator): Wrap with #ifdef GENERATOR_FILE.
2563 (bind_subst_iter_and_attr): Likewise.
2564 (add_condition_to_string): Likewise.
2565 (add_condition_to_rtx): Likewise.
2566 (apply_attribute_uses): Likewise.
2567 (add_current_iterators): Likewise.
2568 (apply_iterators): Likewise.
2569 (initialize_iterators): Guard usage of apply_subst_iterator with
2570 #ifdef GENERATOR_FILE.
2571 (read_conditions): Wrap with #ifdef GENERATOR_FILE.
2572 (md_reader::read_mapping): Likewise.
2573 (add_define_attr_for_define_subst): Likewise.
2574 (add_define_subst_attr): Likewise.
2575 (read_subst_mapping): Likewise.
2576 (check_code_iterator): Likewise.
2577 (rtx_reader::read_rtx): Likewise. Move one-time initialization
2578 logic to...
2579 (one_time_initialization): New function.
2580 (rtx_reader::read_until): New method.
2581 (read_flags): New function.
2582 (parse_reg_note_name): New function.
2583 (rtx_reader::read_rtx_code): Initialize "iterator" to NULL.
2584 Handle reuse_rtx ids.
2585 Wrap iterator lookup within #ifdef GENERATOR_FILE.
2586 Add parsing support for RTL dumps, mirroring the special-cases in
2587 print_rtx, by calling read_flags, reading REG_NOTE names, INSN_UID
2588 values, and calling handle_any_trailing_information.
2589 (rtx_reader::read_rtx_operand): Convert return type from void
2590 to rtx, returning return_rtx. Handle case 'e'. Call
2591 finalize_string on XSTR and XTMPL fields.
2592 (rtx_reader::read_nested_rtx): Handle dumps in which trailing
2593 "(nil)" values were omitted. Call the postprocess vfunc on the
2594 return_rtx.
2595 (rtx_reader::rtx_reader): Add new "compact" param and pass to base
2596 class ctor. Initialize m_in_call_function_usage. Call
2597 one_time_initialization.
2598 * rtl-tests.c (selftest::test_uncond_jump): Call
2599 set_new_first_and_last_insn.
2600 * rtl.h (read_rtx): Wrap decl with #ifdef GENERATOR_FILE.
2601 * selftest-rtl.c: New file.
2602 * selftest-rtl.h (class selftest::rtl_dump_test): New class.
2603 (selftest::get_insn_by_uid): New decl.
2604 * selftest-run-tests.c (selftest::run_tests): Call
2605 read_rtl_function_c_tests.
2606 * selftest.h (selftest::read_rtl_function_c_tests): New decl.
2607 * tree-dfa.c (ssa_default_def): Return NULL_TREE for rtl function
2608 dumps.
2610 2017-01-05 Uros Bizjak <ubizjak@gmail.com>
2612 * config/i386/i386.md (*testqi_ext_3): No need to handle memory
2613 operands in a special way. Assert that pos+len <= mode precision.
2615 2017-01-05 Jakub Jelinek <jakub@redhat.com>
2617 * common.opt (fvect-cost-model): Remove RejectNegative flag, use
2618 3 argument Alias with unlimited for the negative form.
2619 (fno-vect-cost-model): Removed.
2621 2017-01-05 Martin Liska <mliska@suse.cz>
2623 * hsa-gen.c (gen_hsa_divmod): New function.
2624 (gen_hsa_insn_for_internal_fn_call): Use the function for IFN_DIVMOD.
2626 2017-01-05 Martin Liska <mliska@suse.cz>
2628 PR pch/78970
2629 * gcc.c (lookup_compiler): Reject '-' filename for a precompiled
2630 header.
2632 2017-01-05 Andreas Krebbel <krebbel@linux.vnet.ibm.com>
2634 * config/s390/s390.c (s390_expand_setmem): Unroll the loop for
2635 small constant length operands.
2637 2017-01-05 Andreas Krebbel <krebbel@linux.vnet.ibm.com>
2639 * config/s390/s390.c (s390_expand_setmem): Avoid overlapping bytes
2640 between loop iterations.
2642 2017-01-05 Martin Liska <mliska@suse.cz>
2644 PR sanitizer/78815
2645 * gimplify.c (gimplify_decl_expr): Compare to
2646 asan_poisoned_variables instread of checking flags.
2647 (gimplify_target_expr): Likewise.
2648 (gimplify_expr): Likewise.
2649 (gimplify_function_tree): Conditionally initialize
2650 asan_poisoned_variables.
2652 2017-01-04 Jeff Law <law@redhat.com>
2654 PR tree-optimizatin/78812
2655 * rtl.h (contains_mem_rtx_p): Prototype.
2656 * ifcvt.c (containts_mem_rtx_p): Move from here to...
2657 * rtlanal.c (contains_mem_rtx_p): Here and remove static linkage.
2658 * gcse.c (prune_expressions): Use contains_mem_rtx_p to discover
2659 and prune MEMs that are not at the toplevel of a SET_SRC rtx. Look
2660 through ZERO_EXTEND and SIGN_EXTEND when trying to avoid pruning MEMs.
2662 2017-01-04 Alexandre Oliva <aoliva@redhat.com>
2664 * input.c (assert_char_at_range): Default-initialize actual_range.
2666 2017-01-04 Alexandre Oliva <aoliva@redhat.com>
2668 * df-scan.c (df_ref_create_structure): Make regno unsigned,
2669 to match the caller.
2671 2017-01-04 Alexandre Oliva <aoliva@redhat.com>
2673 * cfgexpand.c (expand_gimple_basic_block): Disregard debug
2674 insns after final jump in test to emit dummy move.
2676 2017-01-04 Alexandre Oliva <aoliva@redhat.com>
2678 * gimple-iterator.h (gsi_one_nondebug_before_end_p): New.
2679 * tree-eh.c (cleanup_empty_eh): Skip more debug stmts.
2681 2017-01-04 Alexandre Oliva <aoliva@redhat.com>
2683 * multiple_target.c (create_dispatcher_calls): Init e_next.
2684 * tree-ssa-loop-split.c (split_loop): Init border.
2685 * tree-vect-loop.c (vect_determine_vectorization_factor): Init
2686 scalar_type.
2688 2017-01-04 Michael Meissner <meissner@linux.vnet.ibm.com>
2690 PR target/71977
2691 PR target/70568
2692 PR target/78823
2693 * config/rs6000/predicates.md (sf_subreg_operand): New predicate.
2694 (altivec_register_operand): Do not return true if the operand
2695 contains a SUBREG mixing SImode and SFmode.
2696 (vsx_register_operand): Likewise.
2697 (vsx_reg_sfsubreg_ok): New predicate.
2698 (vfloat_operand): Do not return true if the operand contains a
2699 SUBREG mixing SImode and SFmode.
2700 (vint_operand): Likewise.
2701 (vlogical_operand): Likewise.
2702 (gpc_reg_operand): Likewise.
2703 (int_reg_operand): Likewise.
2704 * config/rs6000/rs6000-protos.h (valid_sf_si_move): Add declaration.
2705 * config/rs6000/rs6000.c (valid_sf_si_move): New function to
2706 determine if a MOVSI or MOVSF operation contains SUBREGs that mix
2707 SImode and SFmode.
2708 (rs6000_emit_move_si_sf_subreg): New helper function.
2709 (rs6000_emit_move): Call rs6000_emit_move_si_sf_subreg to possbily
2710 fixup SUBREGs involving SImode and SFmode.
2711 * config/rs6000/vsx.md (SFBOOL_*): New constants that are operand
2712 numbers for the new peephole2 optimization.
2713 (peephole2 for SFmode unions): New peephole2 to optimize cases in
2714 the GLIBC math library that do AND/IOR/XOR operations on single
2715 precision floating point.
2716 * config/rs6000/rs6000.h (TARGET_NO_SF_SUBREG): New internal
2717 target macros to say whether we need to avoid SUBREGs mixing
2718 SImode and SFmode.
2719 (TARGET_ALLOW_SF_SUBREG): Likewise.
2720 * config/rs6000/rs6000.md (UNSPEC_SF_FROM_SI): New unspecs.
2721 (UNSPEC_SI_FROM_SF): Likewise.
2722 (iorxor): Change spacing.
2723 (and_ior_xor): New iterator for AND, IOR, and XOR.
2724 (movsi_from_sf): New insns for SImode/SFmode SUBREG support.
2725 (movdi_from_sf_zero_ext): Likewise.
2726 (mov<mode>_hardfloat, FMOVE32 iterator): Use register_operand
2727 instead of gpc_reg_operand. Add SImode/SFmode SUBREG support.
2728 (movsf_from_si): New insn for SImode/SFmode SUBREG support.
2729 (fma<mode>4): Use gpc_reg_operand instead of register_operand.
2730 (fms<mode>4): Likewise.
2731 (fnma<mode>4): Likewise.
2732 (fnms<mode>4): Likewise.
2733 (nfma<mode>4): Likewise.
2734 (nfms<mode>4): Likewise.
2736 2017-01-04 Marek Polacek <polacek@redhat.com>
2738 PR c++/64767
2739 * doc/invoke.texi: Document -Wpointer-compare.
2741 2017-01-04 Jakub Jelinek <jakub@redhat.com>
2743 * optc-gen.awk: Emit #error for -W*/-f*/-m* Enum without
2744 RejectNegative.
2746 * dwarf2out.c (output_loc_list): Don't throw away 64K+ location
2747 descriptions for -gdwarf-5 and emit them as uleb128 instead of
2748 2-byte data.
2750 2017-01-04 Kelvin Nilsen <kelvin@gcc.gnu.org>
2752 PR target/78056
2753 * doc/sourcebuild.texi (PowerPC-specific attributes): Add
2754 documentation of the powerpc_popcntb_ok attribute.
2755 * config/rs6000/rs6000.c (rs6000_option_override_internal): Add
2756 code to issue warning messages if a requested CPU configuration is
2757 not supported by the binary (assembler and loader) toolchain.
2758 (spe_init_builtins): Add two assertions to prevent ICE if attempt is
2759 made to define a built-in function that has been disabled.
2760 (paired_init_builtins): Add assertion to prevent ICE if attempt is
2761 made to define a built-in function that has been disabled.
2762 (altivec_init_builtins): Add comment explaining why definition
2763 of the DST built-in functions is not preceded by an assertion
2764 check. Add assertions to prevent ICE if attempts are made to
2765 define an altivec predicate or an abs* built-in function that has
2766 been disabled.
2767 (htm_init_builtins): Add comment explaining why definition of the
2768 htm built-in functions is not preceded by an assertion check.
2770 2017-01-04 Jeff Law <law@redhat.com>
2772 PR tree-optimizatin/67955
2773 * tree-ssa-alias.c (same_addr_size_stores_p): Check offsets first.
2774 Allow any SSA_VAR_P as the base objects. Use integer_zerop. Verify
2775 the points-to solution does not include pt_null. Use DECL_PT_UID
2776 unconditionally.
2778 2017-01-04 Uros Bizjak <ubizjak@gmail.com>
2780 * config/i386/i386.md (HI/SImode test with imm to QImode splitters):
2781 Use gen_int_mode instead of gen_lopwart for const_int operands.
2783 2017-01-04 Jakub Jelinek <jakub@redhat.com>
2785 PR tree-optimization/71563
2786 * match.pd: Simplify X << Y into X if Y is known to be 0 or
2787 out of range value - has low bits known to be zero.
2789 2017-01-04 Alan Modra <amodra@gmail.com>
2791 * Makefile.in (aclocal_deps): Update and order as per aclocal.m4.
2792 * configure: Regenerate.
2793 * config.in: Regenerate.
2795 2017-01-04 Jakub Jelinek <jakub@redhat.com>
2797 PR bootstrap/77569
2798 * input.c (ebcdic_execution_charset::on_error): Don't use strstr for
2799 a substring of the message, but strcmp with the whole message. Ifdef
2800 ENABLE_NLS, translate the message first using dgettext.
2802 2017-01-03 Jeff Law <law@redhat.com>
2804 PR tree-optimizatin/78856
2805 * tree-ssa-threadupdate.c: Include tree-vectorizer.h.
2806 (mark_threaded_blocks): Remove code to truncate thread paths that
2807 cross multiple loop headers. Instead invalidate the cached loop
2808 iteration information and handle case of a thread path walking
2809 into an irreducible region.
2811 2017-01-03 Michael Meissner <meissner@linux.vnet.ibm.com>
2813 PR target/78900
2814 * config/rs6000/rs6000.c (rs6000_split_signbit): Change some
2815 assertions. Add support for doing the signbit if the IEEE 128-bit
2816 floating point value is in a GPR.
2817 * config/rs6000/rs6000.md (Fsignbit): Delete.
2818 (signbit<mode>2_dm): Delete using <Fsignbit> and just use "wa".
2819 Update the length attribute if the value is in a GPR.
2820 (signbit<mode>2_dm_<su>ext): Add combiner pattern to eliminate
2821 the sign or zero extension instruction, since the value is always 0/1.
2822 (signbit<mode>2_dm2): Delete using <Fsignbit>.
2824 PR target/78953
2825 * config/rs6000/vsx.md (vsx_extract_<mode>_store_p9): If we are
2826 extracting SImode to a GPR register so that we can generate a
2827 store, limit the vector to be in a traditional Altivec register
2828 for the vextuwrx instruction.
2830 2017-01-03 Ian Lance Taylor <iant@google.com>
2832 * godump.c (go_format_type): Treat ENUMERAL_TYPE like INTEGER_TYPE.
2834 2017-01-03 Martin Sebor <msebor@redhat.com>
2836 PR tree-optimization/78696
2837 * gimple-ssa-sprintf.c (format_floating): Correct handling of
2838 precision. Use MPFR for %f for greater fidelity. Correct handling
2839 of %g.
2840 (pass_sprintf_length::compute_format_length): Set width and precision
2841 specified by asrerisk to void_node for vararg functions.
2842 (try_substitute_return_value): Adjust dump output.
2844 2017-01-03 David Edelsohn <dje.gcc@gmail.com>
2846 * doc/invoke.texi (RS6000 options): LRA is enabled by default.
2848 2017-01-03 Eric Botcazou <ebotcazou@adacore.com>
2850 * doc/invoke.texi (SPARC options): Document -mlra as the default.
2851 * config/sparc/sparc.c (sparc_option_override): Force LRA unless
2852 -mlra/-mno-lra was passed to the compiler.
2854 2017-01-03 James Cowgill <James.Cowgill@imgtec.com>
2856 PR rtl-optimization/65618
2857 * emit-rtl.c (try_split): Move initialization of "before" and
2858 "after" to just before the call to emit_insn_after_setloc.
2860 2017-01-03 Gerald Pfeifer <gerald@pfeifer.com>
2862 * doc/md.texi (Standard Names): Remove reference to Java frontend.
2864 2017-01-03 Pierre-Marie de Rodat <derodat@adacore.com>
2866 * dwarf2out.c (gen_enumeration_type_die): When
2867 -gno-strict-dwarf, add a DW_AT_encoding attribute.
2869 2017-01-03 Jakub Jelinek <jakub@redhat.com>
2871 PR tree-optimization/78965
2872 * gimple-ssa-sprintf.c (pass_sprintf_length::compute_format_length):
2873 Change first argument from const call_info & to call_info &. For %n
2874 set info.nowrite to false.
2876 PR middle-end/78901
2877 * gimple-ssa-sprintf.c (try_substitute_return_value): Don't change
2878 possibly throwing calls.
2880 * genmatch.c (dt_node::gen_kids_1): If generic_exprs include SSA_NAME
2881 and exprs_len || fns_len, emit the code for SSA_NAME next to the exprs
2882 and fns handling, rather than in a separate case SSA_NAME.
2884 2017-01-02 Jeff Law <law@redhat.com>
2886 * config/darwin-driver.c (darwin_driver_init): Const-correctness
2887 fixes for first_period and second_period variables.
2889 2017-01-02 Uros Bizjak <ubizjak@gmail.com>
2891 PR target/78967
2892 * config/i386/i386.md (UNSPEC_NOREX_MEM): New unspec.
2893 (*insvqi_1): New insn pattern.
2894 (*insvqi_1_mem_rex64): Ditto.
2895 (*insvqi_2): Ditto.
2896 (*insvqi_3): Rename from *insvqi.
2898 (*extzvqi_mem_rex64): Add UNSPEC_NOREX_MEM tag.
2900 2017-01-02 Gerald Pfeifer <gerald@pfeifer.com>
2902 * doc/cfg.texi (Edges): Remove reference to Java.
2903 (Maintaining the CFG): Ditto.
2905 2017-01-01 Jan Hubicka <hubicka@ucw.cz>
2907 PR middle-end/77674
2908 * symtab.c (symtab_node::binds_to_current_def_p): Fix handling of
2909 transparent aliases.
2911 2017-01-01 Jan Hubicka <hubicka@ucw.cz>
2913 PR middle-end/77484
2914 * predict.def (PRED_CALL): Update hitrate.
2915 (PRED_INDIR_CALL, PRED_POLYMORPHIC_CALL): New predictors.
2916 * predict.c (tree_estimate_probability_bb): Split CALL predictor
2917 into direct/indirect/polymorphic variants.
2919 2017-01-01 Jakub Jelinek <jakub@redhat.com>
2921 Update copyright years.
2923 * gcc.c (process_command): Update copyright notice dates.
2924 * gcov-dump.c (print_version): Ditto.
2925 * gcov.c (print_version): Ditto.
2926 * gcov-tool.c (print_version): Ditto.
2927 * gengtype.c (create_file): Ditto.
2928 * doc/cpp.texi: Bump @copying's copyright year.
2929 * doc/cppinternals.texi: Ditto.
2930 * doc/gcc.texi: Ditto.
2931 * doc/gccint.texi: Ditto.
2932 * doc/gcov.texi: Ditto.
2933 * doc/install.texi: Ditto.
2934 * doc/invoke.texi: Ditto.
2935 \f
2936 Copyright (C) 2017 Free Software Foundation, Inc.
2938 Copying and distribution of this file, with or without modification,
2939 are permitted in any medium without royalty provided the copyright
2940 notice and this notice are preserved.