instead of st_table.
* iseq.c (prepare_iseq_build): don't allocate ivar_cache_table
until it has at least one element.
* iseq.c (compile_data_free): free ivar_cache_table only if it
is allocated.
* compile.c (get_ivar_ic_value): allocate if the table is not
allocated yet.
git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@55002 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
instance variable accesses. Reducing memory consumption,
rasing cache hit rate and rasing branch prediction hit rate
are expected. A part of [Bug #12274].
* iseq.h (struct iseq_compile_data): introduce instance
variable IC table for sharing.
* iseq.c (prepare_iseq_build, compile_data_free):
construct/destruct above table.
git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@54976 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
* compile.c (iseq_peephole_optimize): remove successive line trace
instructions except for the last.
git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@54733 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
* compile.c (append_compile_error): make SyntaxError instance by
rb_syntax_error_append on demand after formatting the message.
git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@54652 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
* compile.c (append_compile_error): use rb_syntax_error_append.
* error.c (rb_syntax_error_append): append messages into a
SyntaxError exception instance.
* parse.y (yycompile0): make new SyntaxError instance in main
mode, otherwize error_buffer should be a SyntaxError if error
has occurred.
git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@54650 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
* compile.c (append_compile_error, compile_bug): pass iseq and get
error info and file from it, not by the thread error info.
* error.c (rb_report_bug_valist): take va_list instead of variadic
arguments, and just report the bug but not abort.
git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@54648 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
* compile.c (iseq_peephole_optimize): should not replace the
current target INSN, not to follow the replaced dangling link in
the caller. [ruby-core:74993] [Bug #11816]
git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@54628 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
* compile.c (new_label_body): initialize bit fields, since
compile_data_alloc does not clear the memory. [Bug #12082]
git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@54548 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
* parse.y (struct parser_params): move parse_in_eval flag from
rb_thread_t.
* parse.y (rb_parser_set_context): set parsing context, not only
mild error flag.
* iseq.c (rb_iseq_compile_with_option): the parser now refers no
thread local states to be restored.
* vm_eval.c (eval_string_with_cref): ditto.
git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@54343 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
* compile.c (prepare_compile_error): add & to explicit address of
functions to suppress warnings by old VC.
git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@54294 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
* compile.c (iseq_specialized_instruction): specialize only
concatenated newarray and send, no labels and no adjusts.
git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@54157 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
Now `[x, y].max` is optimized so that a temporal array object is not
created in some condition.
* insns.def (opt_newarray_max, opt_newarray_min): added.
git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@54153 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
All of the strings created here eventually get converted to
fstrings when they are frozen into the iseq. Prepare the
fstring early so we may reduce a one or two objects.
This is a very minor change, mainly for the '<main>' dedupe.
* compile.c (caller_location): use rb_fstring_cstr for "<compiled>"
(it is converted to fstring anyways inside rb_iseq_new_with_opt)
* iseq.c (iseqw_s_compile): ditto
* iseq.c (rb_iseq_new_main): use rb_fstring_cstr for "<main>"
* vm.c (Init_VM): ditto, share with with above
* iseq.c (iseqw_s_compile_file): rb_fstring before rb_io_t->pathv
share "<main>" with above
* vm.c (rb_binding_add_dynavars): fstring "<temp>" immediately
git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@53659 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
* compile.c (iseq_peephole_optimize): don't apply tailcall
optimization to send/invokesuper instructions with blockiseq.
This is a follow-up to the changes in r51903; blockiseq is now
the third operand of send/invokesuper instructions.
[ruby-core:73413] [Bug #12018]
git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@53658 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
* compile.c (compile_massign_lhs): when index ends with splat,
append rhs value to it like POSTARG, since VM_CALL_ARGS_SPLAT
splats the last argument only. [ruby-core:72777] [Bug #11970]
git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@53495 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
* compile.c (new_adjust_body): labels referred by adjuststack
shoud not be optimized away.
git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@53367 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
There is no need to dup the fstring unless we want to set the
debug ivar for it.
[ruby-core:72018] <5668DB6E.8000101@ruby-lang.org>
git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@53021 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
without --debug or --debug=frozen-string-literal option
because String#dup slows down with debug information.
[Feature #11725]
* NEWS: apply about it.
* test/ruby/test_rubyoptions.rb: catch up this fix with refactoring.
git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@53009 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
NULL checking is finished Before call of memsize functions.
See r52979.
git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@52986 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
* compile.c (ibf_dump_overwrite): cast to unsigned long to
suppress sign-compare warning.
git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@52971 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
* compile.c (iseq_ibf_load_extra_data): fix offset and
length of extra_str, which is not NUL-terminated.
git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@52968 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
* compile.c (iseq_ibf_dump): rb_check_string_type() returns nil if
the object does not have to_str method.
git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@52958 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
and a pre-compilation/runtime loader sample.
[Feature #11788]
* iseq.c: add new methods:
* RubyVM::InstructionSequence#to_binary_format(extra_data = nil)
* RubyVM::InstructionSequence.from_binary_format(binary)
* RubyVM::InstructionSequence.from_binary_format_extra_data(binary)
* compile.c: implement body of this new feature.
* load.c (rb_load_internal0), iseq.c (rb_iseq_load_iseq):
call RubyVM::InstructionSequence.load_iseq(fname) with
loading script name if this method is defined.
We can return any ISeq object as a result value.
Otherwise loading will be continue as usual.
This interface is not matured and is not extensible.
So that we don't guarantee the future compatibility of this method.
Basically, you should'nt use this method.
* iseq.h: move ISEQ_MAJOR/MINOR_VERSION (and some definitions)
from iseq.c.
* encoding.c (rb_data_is_encoding), internal.h: added.
* vm_core.h: add several supports for lazy load.
* add USE_LAZY_LOAD macro to specify enable or disable of
this feature.
* add several fields to rb_iseq_t.
* introduce new macro rb_iseq_check().
* insns.def: some check for lazy loading feature.
* vm_insnhelper.c: ditto.
* proc.c: ditto.
* vm.c: ditto.
* test/lib/iseq_loader_checker.rb: enabled iff suitable
environment variables are provided.
* test/runner.rb: enable lib/iseq_loader_checker.rb.
* sample/iseq_loader.rb: add sample compiler and loader.
$ ruby sample/iseq_loader.rb [dir]
will compile all ruby scripts in [dir].
With default setting, this compile creates *.rb.yarb files
in same directory of target .rb scripts.
$ ruby -r sample/iseq_loader.rb [app]
will run with enable to load compiled binary data.
git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@52949 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
This reverts r16587
("compile.c (iseq_build_body): remove side effect from
VM::InstructionSequence.load."
as that change was obsoleted by r48705 ("mostly fix rb_iseq_load")
git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@52938 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
nil/true/false are special literals just like floats, integers,
literal strings, and symbols. Optimize when statements with
them by using a jump table, too.
target 0: a (ruby 2.3.0dev (2015-12-08 trunk 52928) [x86_64-linux]) at "/home/ew/rrrr/b/ruby"
target 1: b (ruby 2.3.0dev (2015-12-08 master 52928) [x86_64-linux]) at "/home/ew/ruby/b/ruby"
benchmark results:
minimum results in each 5 measurements.
Execution time (sec)
name a b
loop_whileloop2 0.102 0.103
vm2_case_lit* 1.657 0.549
Speedup ratio: compare with the result of `a' (greater is better)
name b
loop_whileloop2 0.988
vm2_case_lit* 3.017
* benchmark/bm_vm2_case_lit.rb: new benchmark
* compile.c (case_when_optimizable_literal): add nil/true/false
* insns.def (opt_case_dispatch): ditto
* vm.c (vm_redefinition_check_flag): ditto
* vm.c (vm_init_redefined_flag): ditto
* vm_core.h: ditto
* object.c (InitVM_Object): define === explicitly for nil/true/false
* test/ruby/test_case.rb (test_deoptimize_nil): new test
* test/ruby/test_optimization.rb (test_opt_case_dispatch): update
(test_eqq): new test
[ruby-core:71923] [Feature #11769]
Original patch by Aaron Patterson <tenderlove@ruby-lang.org>
git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@52931 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
naruse
tarui
nobu
mame
normal
ko1
hsbt