2017-03-09 Martin Liska <mliska@suse.cz>
PR ipa/79764
(chkp_narrow_size_and_offset): New function.
(chkp_parse_array_and_component_ref): Support BIT_FIELD_REF.
(void chkp_parse_bit_field_ref): New function.
(chkp_make_addressed_object_bounds): Add case for BIT_FIELD_REF.
(chkp_process_stmt): Use chkp_parse_bit_field_ref.
2017-03-09 Martin Liska <mliska@suse.cz>
PR ipa/79764
* g++.dg/pr79764.C: New test.
From-SVN: r245994
2017-03-09 Martin Liska <mliska@suse.cz>
PR ipa/79761
* tree-chkp.c (chkp_get_bound_for_parm): Get bounds for a param.
(chkp_find_bounds_1): Remove gcc_unreachable.
2017-03-09 Martin Liska <mliska@suse.cz>
PR ipa/79761
* g++.dg/pr79761.C: New test.
From-SVN: r245993
PR sanitizer/79944
* asan.c (get_mem_refs_of_builtin_call): For BUILT_IN_ATOMIC* and
BUILT_IN_SYNC*, determine the access type from the size suffix and
always build a MEM_REF with that type. Handle forgotten
BUILT_IN_SYNC_FETCH_AND_NAND_16 and BUILT_IN_SYNC_NAND_AND_FETCH_16.
* c-c++-common/asan/pr79944.c: New test.
From-SVN: r245991
PR c/79940
* gimplify.c (gimplify_omp_for): Replace index var in outer
taskloop statement with an artificial variable and add
OMP_CLAUSE_PRIVATE clause for it.
* testsuite/libgomp.c/pr79940.c: New test.
From-SVN: r245980
2017-03-08 Richard Biener <rguenther@suse.de>
PR tree-optimization/79955
* tree-ssa-uninit.c (warn_uninitialized_vars): Do not warn
for accesses that are completely outside of the variable.
* gcc.dg/uninit-24.c: New testcase.
From-SVN: r245976
2017-03-08 Andrew Haley <aph@redhat.com>
PR tree-optimization/79943
* tree-ssa-loop-split.c (compute_new_first_bound): When
calculating the new upper bound, (END-BEG) should be added, not
subtracted.
From-SVN: r245974
2017-03-08 Richard Biener <rguenther@suse.de>
PR tree-optimization/79920
* tree-vect-slp.c (vect_create_mask_and_perm): Remove and inline
with ncopies == 1 to ...
(vect_transform_slp_perm_load): ... here. Properly compute
all element loads by iterating VF times over the group. Do
not handle ncopies (computed in a broken way) in
vect_create_mask_and_perm.
* gcc.dg/vect/pr79920.c: New testcase.
From-SVN: r245968
PR sanitizer/79904
* internal-fn.c (expand_vector_ubsan_overflow): If arg0 or arg1
is a uniform vector, use uniform_vector_p return value instead of
building ARRAY_REF on folded VIEW_CONVERT_EXPR to array type.
* gcc.dg/ubsan/pr79904.c: New test.
From-SVN: r245967
2017-03-07 Martin Liska <mliska@suse.cz>
PR middle-end/68270
* tree-chkp.c (chkp_may_narrow_to_field): Add new argument ref.
Use array_at_struct_end_p instead of DECL_CHAIN (field).
(chkp_narrow_bounds_for_field): Likewise.
(chkp_parse_array_and_component_ref): Pass one more argument to
call.
2017-03-07 Martin Liska <mliska@suse.cz>
PR middle-end/68270
* g++.dg/pr68270.C: New test.
From-SVN: r245951
As Shmuel reported in <https://gcc.gnu.org/ml/gcc-help/2017-03/msg00009.html>,
on x86-64 small structures in automatic storage are aligned to 16 bytes.
This seems to be because of a mix-up between bits and bytes in the i386
target code.
* config/i386/i386.c (ix86_local_alignment): Align most aggregates
of 16 bytes and more to 16 bytes, not those of 16 bits and more.
From-SVN: r245949
PR c/79855
* params.def (PARAM_STORE_MERGING_ALLOW_UNALIGNED): Add full stop
to end of description.
(PARAM_MAX_STORES_TO_MERGE): Likewise.
From-SVN: r245948
PR rtl-optimization/79901
* expr.c (expand_expr_real_2): For vector MIN/MAX, if there is no
min/max expander, expand it using expand_vec_cond_expr.
From-SVN: r245946
2017-03-06 Vladimir Makarov <vmakarov@redhat.com>
PR rtl-optimization/79571
* lra-constraints.c (process_alt_operands): Claculate static
reject and subtract it from overal when there will be only address
reloads.
2017-03-06 Vladimir Makarov <vmakarov@redhat.com>
PR rtl-optimization/79571
* gcc.target/i386/pr79571.c: New.
From-SVN: r245928
For 64-bit targets, the incoming stack of interrupt handler is aligned
to 16 bytes. Update ix86_minimum_incoming_stack_boundary to set the
incoming stack boundary of interrupt handler to 128 for 64-bit targets.
gcc/
2017-03-06 Julia Koval <julia.koval@intel.com>
PR target/79793
* config/i386/i386.c (ix86_minimum_incoming_stack_boundary): Set
incoming stack boundary to 128 for 64-bit targets.
gcc/testsuite/
2017-03-06 Julia Koval <julia.koval@intel.com>
PR target/79793
* gcc.target/i386/interrupt-12.c: Update scan-assembler-times
directives.
* gcc.target/i386/interrupt-13.c: Ditto.
* gcc.target/i386/interrupt-14.c: Ditto.
* gcc.target/i386/interrupt-15.c: Ditto.
From-SVN: r245926
2017-03-06 Richard Biener <rguenther@suse.de>
PR tree-optimization/79894
* tree-vectorizer.c (vectorize_loops): Set loop_vectorized_call
to NULL after folding it.
* gcc.dg/vect/pr79887.c: New testcase.
From-SVN: r245923
The following patch defines the PARAM_MIN_VECT_LOOP_BOUND parameter in
the s390 backend. It helps with the vectorization epilogue problem
described here [1].
I see an overall performance increase of > 1% in SPECfp2006, yet some
cases like cactusADM regress. This seems to be caused by the vectorizer
creating an epilogue guard for one more iteration than before, which, in
turn, causes e.g. predcom to run on the epilogue that it used to ignore
before ("Loop iterates only 1 time, nothing to do."). Subsequent,
minor, effects cause an eventual slowdown.
Until the reason for the bad epilogue code is understood, this patch
mitigates the problem. When investigating the issue, I stumbled across
an attempt to vectorize the epilogue itself as well as combine it with
the vectorized loop in addition to vector masking [2]. A similar
approach might also help here. My original observation of high register
pressure within the epilogue still stands. In this specific case, it
would most likely suffice to save all registers once, run the epilogue
and restore the registers. I'm pretty sure this would be faster than
the "spill fest" that's currently happening.
Regards
Robin
[1] https://gcc.gnu.org/ml/gcc/2017-01/msg00234.html
[2] https://gcc.gnu.org/ml/gcc-patches/2016-05/msg01562.html
gcc/ChangeLog:
2017-03-06 Robin Dapp <rdapp@linux.vnet.ibm.com>
* config/s390/s390.c (s390_option_override_internal): Set
PARAM_MIN_VECT_LOOP_BOUND
From-SVN: r245918
2017-03-06 Martin Liska <mliska@suse.cz>
PR sanitize/79783
* asan.c (asan_expand_poison_ifn): Do not expand ASAN_POISON
when having a SSA NAME w/o VAR_DECL assigned to it.
2017-03-06 Martin Liska <mliska@suse.cz>
PR sanitize/79783
* g++.dg/asan/pr79783.C: New test.
From-SVN: r245912
gcc/
* config/mips/mips.c (mips_gen_const_int_vector): Change type of last
argument.
* config/mips/mips-protos.h (mips_gen_const_int_vector): Likewise.
gcc/testsuite/
* gcc.target/mips/msa-bclri.c: New test.
From-SVN: r245910
* config/i386/sse.md (sse_storehps, sse_storelps,
avx_<castmode><avxsizesuffix>_<castmode>,
avx512f_<castmode><avxsizesuffix>_<castmode>,
avx512f_<castmode><avxsizesuffix>_256<castmode>): Require
in condition that at least one operand is not a MEM.
From-SVN: r245904
PR middle-end/79805
* internal-fn.def (ATOMIC_BIT_TEST_AND_SET, ATOMIC_BIT_TEST_AND_RESET,
ATOMIC_BIT_TEST_AND_COMPLEMENT, ATOMIC_COMPARE_EXCHANGE): Remove
ECF_NOTHROW.
* gimple-fold.c (fold_builtin_atomic_compare_exchange): Set
gimple_call_nothrow_p flag based on whether original builtin can throw.
If it can, emit following stmts on the fallthrough edge.
* tree-ssa-ccp.c (optimize_atomic_bit_test_and): Similarly, except
don't create new bb if inserting just debug stmts on the edge, try to
insert them on the fallthru bb or just reset debug stmts.
* g++.dg/opt/pr79805.C: New test.
From-SVN: r245882
If using -mwarn-cell-microcode, rs6000_final_prescan_insn calls
get_insn_template to get the name of the machine instruction. But,
get_insn_template calls the output template if that is code, and that
then can modify recog_data (it is normal to change the operands, for
example).
This patch saves and restores recog_data around the call to
get_insn_template to fix the problems this causes.
PR target/43763
* config/rs6000/rs6000.c (rs6000_final_prescan_insn): Save and
restore recog_data (including the operand rtxes inside it) around
the call to get_insn_template.
From-SVN: r245880