2012-09-10 Martin Jambor <mjambor@suse.cz>
* params.def (PARAM_IPA_MAX_AGG_ITEMS): New parameter.
* ipa-prop.c: Include params.h.
(IPA_MAX_AFF_JF_ITEMS): Removed.
(determine_known_aggregate_parts): Use param value of
PARAM_IPA_MAX_AGG_ITEMS instead of IPA_MAX_AFF_JF_ITEMS.
* Makefile.in (ipa-prop.o): Add PARAMS_H dependency.
From-SVN: r191150
PR c++/54506
* decl.c (move_signature_fn_p): Split out from move_fn_p.
* method.c (process_subob_fn): Use it.
* cp-tree.h: Declare it.
From-SVN: r191140
* lto-cgraph.c (compute_ltrans_boundary): Do not care about aliases.
* lto-partition.c (partition_symbol_p): Forward declare.
(add_references_to_partition): Reimplement using partition_symbol_p.
(add_aliases_to_partition): Break out from add_references_to_partition;
reimplement using partition_symbol_p.
(add_cgraph_node_to_partition_1): Handle callees using partition_symbol_p;
add sanity checks.
(add_varpool_node_to_partition): Use add_aliases_to_partition.
(partition_varpool_node_p): Do not special case aliases.
From-SVN: r191129
* passes.c (ipa_write_summaries_1): Set state;
do not call compute_ltrans_boundary.
(ipa_write_optimization_summaries): Likewise.
(ipa_write_summaries): compute_ltrans_boundary here.
* lto-streamer.h (lto_symtab_encoder_d): NODES are allocated in heap.
(compute_ltrans_boundary): Update prototype.
* lto.c (lto_wpa_write_files): Do not delete partition encoder;
it is deleted after streaming.
* lto-partition.c (partition_symbol_p): New function.
(promote_var, promote_fn): Remove.
(promote_symbol): New function.
(lto_promote_cross_file_statics): First compute boundaries; rewrite
to lookup the actual boundaries instead of computing them ad-hoc.
From-SVN: r191113
gcc/
* config/i386/cpuid.h: Define signature_*_e[bcd]x macros for
matching results of level 0 calls to __cpuid to processor
manufacturers.
* config/i386/driver-i386.c (vendor_signatures): Removed.
(processor_signatures): Removed.
(host_detect_local_cpu): Replace uses of now-removed SIG_*
constants with the new signature_*_ebx constants.
libstdc++-v3/
* src/c++11/random.cc (random_device::_M_init): Use new macro
signature_INTEL_ebx to check for Intel processors.
From-SVN: r191109
For parallel LTO builds setting -fmem-report does not work very well
because all the LTRANS phases dump it in parallel and typically interleave
it to unreadability.
Since usually the memory bottleneck is WPA add a flag to only dump
the memory report for that.
gcc/:
2012-09-08 Andi Kleen <ak@linux.intel.com>
* gcc/common.opt (-fmem-report-wpa): Add
* gcc/doc/invoke.texi (-fmem-report-wpa): Document.
* gcc/lto/lto.c (do_whole_program_analysis): Run mem_report
when mem_report_wpa is set.
From-SVN: r191095
2012-09-07 Paolo Carlini <paolo.carlini@oracle.com>
* pt.c (num_template_headers_for_class): Rework per the code
inline in cp_parser_check_declarator_template_parameters.
* parser.c (cp_parser_check_declarator_template_parameters):
Use num_template_headers_for_class.
From-SVN: r191082
* config/v850/v850.h (DBX_DEBUGGING_INFO): Define.
(ASM_GENERATE_INTERNAL_LABEL): Define if not already provided.
* config/v850/v850.c (compute_register_save_size): Always include
the link pointer.
(increment_stack): New function - emits insns to increment or
decrement the stack pointer.
(expand_prologue, expand_epilogue): Use it.
(expand_prologue): Set the function stack size, if requested.
(v850_debug_unwind_info): New function.
(TARGET_DEBUG_UNWIND_INFO): Define.
From-SVN: r191076
This fixes part of the issue described in PR gcov-profile/54487 where
there were warnings about mismatches due to slight differences in the
merged histograms in different object files. This can happen due to
the truncating integer division in the merge routine, which could result
in slightly different histograms when summaries are merged in different
orders.
2012-09-07 Teresa Johnson <tejohnson@google.com>
PR gcov-profile/54487
* libgcc/libgcov.c (gcov_exit): Avoid warning on histogram
differences.
From-SVN: r191074
PR tree-ssa/54295
* tree-ssa-math-opts.c (widening_mult_conversion_strippable_p):
Sign-extension of a zero-extended value can be simplified to
just zero-extension.
testsuite:
* gcc.target/arm/pr50318-1.c: Scan for smlal.
* gcc.target/arm/smlaltb-1.c: XFAIL test.
* gcc.target/arm/smlaltt-1.c: Likewise.
From-SVN: r191066
* bitmap.c (bitmap_last_set_bit): Rewrite to return the correct bit.
* graphite.c (print_global_statistics): Use EDGE_COUNT instead
of VEC_length.
(print_graphite_scop_statistics): Likewise.
* graphite-scop-detection.c (get_bb_type): Use single_succ_p.
(print_graphite_scop_statistics): Use EDGE_COUNT, not VEC_length.
(canonicalize_loop_closed_ssa): Use single_pred_p.
* alias.c (reg_seen): Make this an sbitmap.
(record_set, init_alias_analysis): Update.
* tree-ssa-coalesce.c (ssa_conflicts_dump): Fix dumping.
From-SVN: r191063
2012-09-07 Tom de Vries <tom@codesourcery.com>
PR tree-optimization/53986
* tree-vrp.c (extract_range_from_multiplicative_op_1): Allow
LSHIFT_EXPR.
(extract_range_from_binary_expr_1): Handle LSHIFT with constant range as
shift amount.
From-SVN: r191057
With a LTO build of a large project (>11k subfiles incrementially linked)
storing the LTO resolutions took over 0.5GB memory:
lto/lto.c:1087 (lto_resolution_read) 0: 0.0% 540398500 15903: 0.0%
The reason is that the declaration indexes are quite sparse, but every subfile
got a full continuous vector for them. Since there are so many of them the
many vectors add up.
This patch instead stores the resolutions initially in a compact (index, resolution)
format. This is only expanded into a sparse vector for fast lookup when
the subfile is actually read, but then immediately freed. This means only one
vector is allocated at a time.
This brings the overhead for this down to less than 3MB for the test case:
lto/lto.c:1087 (lto_resolution_read) 0: 0.0% 2821456 42186: 0.0%
gcc/:
2012-09-06 Andi Kleen <ak@linux.intel.com>
* gcc/lto-streamer.h (res_pair): Add.
(lto_file_decl_data): Replace resolutions with respairs.
Add max_index.
* gcc/lto/lto.c (lto_resolution_read): Remove max_index. Add rp.
Initialize respairs.
(lto_file_finalize): Set up resolutions vector lazily from respairs.
From-SVN: r191051
2012-09-06 Thiago Macieira <thiago.macieira@intel.com>
PR libstdc++/54172
* libsupc++/guard.cc (__cxa_guard_acquire): Exit the loop earlier if
we detect that another thread has had success. Don't compare_exchange
from a finished state back to a waiting state. Comment.
From-SVN: r191042
PR c++/54341
PR c++/54253
* semantics.c (sort_constexpr_mem_initializers): New.
(build_constexpr_constructor_member_initializers): Use it.
(cx_check_missing_mem_inits): Skip artificial fields.
* init.c (expand_aggr_init_1): Don't zero out a class
with no data.
From-SVN: r191033
2012-09-06 Richard Guenther <rguenther@suse.de>
PR tree-optimization/54498
* tree-ssa-alias.h (get_continuation_for_phi): Add flag to
abort when reaching an already visited region.
* tree-ssa-alias.c (maybe_skip_until): Likewise. And do it.
(get_continuation_for_phi_1): Likewise.
(walk_non_aliased_vuses): When we translated the reference,
abort when we re-visit a region.
* tree-ssa-pre.c (translate_vuse_through_block): Adjust.
From-SVN: r191030
