When no instruction is emitted in `iseq_compile_each0`
(i.e., when the line has no significant code), trace
instruction for `RUBY_EVENT_LINE` has been optimized out.
But trace for `RUBY_EVENT_COVERAGE` has not been removed.
Now, it is also removed.
`TestISeq#test_to_a_lines` has failed a long time under
coverage measurement (`make test-all COVERAGE=true`).
This change makes it pass.
git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@59769 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
To prepare new measuring targets: branch and method coverages.
So far, iseq->coverage was an array of counts executed for line coverage.
Now, it is a three-element array for each measuring target,
whose first element is an array for line coverage.
The second element is planned for branch coverage, and the third will be
for method coverage.
git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@59738 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
* compile.c (ibf_load_iseq_each): location.pathobj may not be a
mere string now.
git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@59709 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
* compile.c (iseq_peephole_optimize): remove unreachable chunk
which appeared by useless jump elimination.
git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@59656 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
* template/insns.inc.tmpl (ASSERT_VM_INSTRUCTION_SIZE): static
assertion for VM instruction info tables.
git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@59635 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
* compile.c (compile_array_keyword_arg): fix the condition of
KW_SPLAT flag. splat is value node only without key node,
simple assoc argument is not. [ruby-core:82291] [Bug #13793]
git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@59556 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
* compile.c (iseq_set_sequence): replace adjuststack with pop, or
remove if possible instead of two nops.
git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@59446 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
* compile.c (iseq_compile_each0): as compile_flip_flop always ends
with a jump instruction, successive instruction is never reached,
but caused stack consistency error without peephole optimization.
git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@59442 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
* compile.c: raise COMPILE_ERROR instead of compile_bug which is
very rarely (or never, actually) useful to debug instruction
sequence. COMPILE_ERROR is usually SyntaxError, or fatal error
if compile_debug is enabled,
git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@59388 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
* compile.c (iseq_compile_each0): restore the stack depth after
return to the previous depth, to fix the stack depth at
returning from rescue iseq. [ruby-core:82108] [Bug #13755]
git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@59374 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
* compile.c (iseq_compile_each0): turned dregx context in "once"
into "guarded" type from "block" type, to disallow `next`,
`break`, `redo` as well as outside "once".
[ruby-core:81805] [Bug #13690]
git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@59202 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
* compile.c (iseq_compile_each0): adjust stack after return in
toplevel ensure, when the value is used.
[ruby-core:81777] [Bug #13682]
git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@59184 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
* compile.c (iseq_compile_each0): debugs/ruby_debug_printf do not
support PRIsVALUE.
git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@59172 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
* compile.c (fix_sp_depth): separate fix-up of sp depth from code
generation.
git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@59171 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
* compile.c (iseq_compile_each0): adjust the stack before return
in an evstr/argument (reported by Balazs <balazs@kutilovi.cz>)
to fix stack consistency error. [ruby-core:81761] [Bug #13678]
git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@59169 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
* compile.c (iseq_set_sequence): separate instruction dump and
error, before allocation of sequence.
git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@59079 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
* compile.c (iseq_set_sequence): initialize with NULL for line_info_table
and generated_iseq and check NULL at BADINSN_ERROR.
git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@59074 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
* compile.c (insn_set_sc_state): dump the whole instructions and
mark the destination when label state mismatch.
git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@59073 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
* compile.c: expand ruby_sourcefile not to evaluate twice in
RSTRING_PTR.
git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@59060 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
* compile.c (iseq_compile_each0): save target child_iseq in the catch-table
for break. This iseq is not for continuation, but for search key at
vm_throw_start().
* vm_insnhelper.c (vm_throw_start): check saved iseq first.
* iseq.h: add comment for it.
* test/ruby/test_iterator.rb (test_ljump): add a test for the issue:
def call b; b.call; end
call(Proc.new{break}){} #=> (1) should raise LocalJumpError
call(Proc.new{break}) #=> (2) shoudd raies LocalJumpError, too.
but (1) doesn't raise LocalJumpError.
This issue is reported by Matz.
git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@59043 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
* compile.c: use NULL instead of 0 for ADD_CATCH_ENTRY()
to specify don't pass iseq.
git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@59042 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
* vm_core.h: rename absolute_path to realpath because it is expected name.
external APIs (#absolute_path methods) are remained.
* vm_core.h: remove rb_iseq_location_struct::path and
rb_iseq_location_struct::absolute_path and introduce pathobj.
if given path equals to given absolute_path (and most of case
it is true), pathobj is simply given path String. If it is not same,
pathobj is Array and pathobj[0] is path and pathobj[1] is realpath.
This size optimization reduce 8 bytes and
sizeof(struct rb_iseq_constant_body) is 200 bytes -> 192 bytes
on 64bit CPU.
To support this change, the following functions are introduced:
* pathobj_path() (defined in vm_core.h)
* pathobj_realpath() (ditto)
* rb_iseq_path() (decl. in vm_core.h)
* rb_iseq_realpath() (ditto)
* rb_iseq_pathobj_new() (ditto)
* rb_iseq_pathobj_set() (ditto)
* vm_core.h (rb_binding_t): use pathobj instead of path. If binding
is given at eval methods, realpath (absolute_path) was caller's
realpath. However, they should use binding's realpath.
git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@58979 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
To convert the object implicitly, it has had two parts in convert_type() which are
1. lookink up the method's id
2. calling the method
Seems that strncmp() and strcmp() in convert_type() are slightly heavy to look up
the method's id for type conversion.
This patch will add and use internal APIs (rb_convert_type_with_id, rb_check_convert_type_with_id)
to call the method without looking up the method's id when convert the object.
Array#flatten -> 19 % up
Array#+ -> 3 % up
[ruby-dev:50024] [Bug #13341] [Fix GH-1537]
### Before
Array#flatten 104.119k (± 1.1%) i/s - 525.690k in 5.049517s
Array#+ 1.993M (± 1.8%) i/s - 10.010M in 5.024258s
### After
Array#flatten 124.005k (± 1.0%) i/s - 624.240k in 5.034477s
Array#+ 2.058M (± 4.8%) i/s - 10.302M in 5.019328s
### Test Code
require 'benchmark/ips'
class Foo
def to_ary
[1,2,3]
end
end
Benchmark.ips do |x|
ary = []
100.times { |i| ary << i }
array = [ary]
x.report "Array#flatten" do |i|
i.times { array.flatten }
end
x.report "Array#+" do |i|
obj = Foo.new
i.times { array + obj }
end
end
git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@58978 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
LABEL::unremovable added by r58810 is not initialized by
new_label_body(), making the optimization unstable.
git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@58894 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
