We already use "static inline" heavily and there should be no
penalty for modern compilers; this adds type-checking, too.
This will make future changes easier-to-review.
git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@65775 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
* vm_core.h: remove `rb_execution_context_t::passed_bmethod_me`
and fix functions to pass the `me` directly.
`passed_bmethod_me` was used to make bmethod (methods defined by
`defined_method`). `rb_vm_invoke_bmethod` invoke `Proc` with `me`
information as method frame (`lambda` frame, actually).
If the proc call is not bmethod call, `passed_bmethod_me` should
be NULL. However, there is a bug which passes wrong `me` for
normal block call.
http://ci.rvm.jp/results/trunk-asserts@silicon-docker/1449470
This is because wrong `me` was remained in `passed_bmethod_me`
(and used incorrectly it after collected by GC).
We need to clear `passed_bmethod_me` just after bmethod call,
but clearing is not enough.
To solve this issue, I removed `passed_bmethod_me` and pass `me`
information as a function parameter of `rb_vm_invoke_bmethod`,
`invoke_block_from_c_proc` and `invoke_iseq_block_from_c` in vm.c.
* vm.c (invoke_iseq_block_from_c): the number of parameters is too
long so that I try to specify `ALWAYS_INLINE`.
* vm.c (invoke_block_from_c_proc): ditto.
* vm_insnhelper.c (vm_yield_with_cfunc): now there are no pathes
to use bmethod here.
git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@65636 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
* vm_core.h (rb_thread_struct): introduce new fields `invoke_type`
and `invoke_arg`.
There are two types threads: invoking proc (normal Ruby thread
created by `Thread.new do ... end`) and invoking func, created
by C-API. `invoke_type` shows the types.
* thread.c (thread_do_start): copy `invoke_arg.proc.args` contents
from Array to ALLOCA stack memory if args length is enough small (<8).
We don't need to keep Array and don't need to cancel using transient heap.
* vm.c (thread_mark): For func invoking threads, they can pass (void *)
parameter (rb_thread_t::invoke_arg::func::arg). However, a rubyspec test
(thread_spec.c) passes an Array object and it expect to mark it.
Clealy it is out of scope (misuse of `rb_thread_create` C-API). However,
I'm not sure someone else has such kind of misunderstanding.
So now we mark conservatively this (void *) arg with rb_gc_mark_maybe.
This misuse is found by this error log.
http://ci.rvm.jp/results/trunk-theap-asserts@silicon-docker/1448164
git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@65622 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
because it's not supported by this file. Also, shared `def_iseq_ptr`
instead of copying the main definition of it.
vm_core.h: moved `def_iseq_ptr` to this place. added `inline` to avoid
compiler warnings since it's not used in some files including vm_core.h.
vm_insnhelper.c: moved `def_iseq_ptr` to vm_core.h.
git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@65440 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
This reverts commit 5a1dfb04bc (r63451)
And mark the functions as async-signal-safe while we're at it to
prevent future developers from making the same mistake as I did :x
git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@65316 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
The comment didn't make sense. As it's allocated with
`ZALLOC_N(struct rb_call_cache, body->ci_size + body->ci_kw_size)`,
it's very likely to be forgotten to press shift key on US keyboard.
git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@65277 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
The instructions were used only for branch coverage.
Instead, it now uses a trace framework [Feature #14104].
git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@65225 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
We have several options to ensure there's no race condition between main
thread and MJIT thead about IC reference:
1) Give up caching ivar for multiple classes (or multiple versions of the
same class) in the same getinstancevariable (This commit's approach)
2) Allocate new inline cache every time
Other ideas we could think of couldn't eliminate possibilities of race
condition.
In 2, it's memory allocation would be slow and it may trigger JIT
cancellation frequently. So 1 would be fast for both VM and JIT
situations.
git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@65213 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
* configure.ac: introduce new configure option `--enable-mjit` and
`--disable-mjit`. Default is "enable".
`--disable-mjit` disables all of MJIT features so that `ruby --jit`
can't enable MJIT.
This option affect a macro `USE_MJIT`.
This change remove `--enable/disable-install-mjit-header` option.
* Makefile.in: introduce the `ENABLE_MJIT` variable.
* common.mk: use `ENABLE_MJIT` option.
* internal.h: respect `USE_MJIT`. Same as other *.c, *.h.
* test/ruby/test_jit.rb: check `ENABLE_MJIT` key of rbconfg.rb.
git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@65204 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
This patch introduces "oneshot_lines" mode for `Coverage.start`, which
checks "whether each line was executed at least once or not", instead of
"how many times each line was executed". A hook for each line is fired
at most once, and after it is fired, the hook flag was removed; it runs
with zero overhead.
See [Feature #15022] in detail.
git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@65195 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
Surprisingly, this constant (been there since around 1983) has
never been a part of any standards until now. We have to find
out the appropriate value.
NSIG_MAX is expected to become a part of forthcoming POSIX.
See: http://austingroupbugs.net/view.php?id=741
_SIG_MAXSIG is here because that is greater than NSIG. See
Python's relevant discussion: https://bugs.python.org/issue20584
git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@65161 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
like "error: static declaration of 'xxx' follows non-static declaration".
r64940 is successfully built on mswin but not built on almost all other environments.
internal.h: ditto
include/ruby/intern.h: MJIT_STATIC is moved to this file since this file
also needs to use this.
mjit.h: MJIT_STATIC is moved from this.
git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@64941 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
because r64849 seems to fix issues which we were confused about.
git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@64850 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
This reverts commit r64829. I'll prepare another temporary fix, but I'll
separately commit that to make it easier to revert that later.
git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@64838 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
not optimizing Array#& and Array#| because vm_insnhelper.c can't easily
inline it (large amount of array.c code would be needed in vm_insnhelper.c)
and the method body is a little complicated compared to Integer's ones.
So I thought only Integer#& and Integer#| have a significant impact,
and eliminating unnecessary branches would contribute to JIT's performance.
vm_insnhelper.c: ditto
tool/transform_mjit_header.rb: make sure these instructions are inlined
on JIT.
compile.c: compile vm_opt_and and vm_opt_or.
id.def: define id for them to be used in compile.c and vm*.c
vm.c: track redefinition of Integer#& and Integer#|
vm_core.h: allow detecting redefinition of & and |
test/ruby/test_jit.rb: test new insns
test/ruby/test_optimization.rb: ditto
* Optcarrot benchmark
This is a kind of experimental thing but I'm committing this since the
performance impact is significant especially on Optcarrot with JIT.
$ benchmark-driver benchmark.yml --rbenv 'before::before --disable-gems;before+JIT::before --disable-gems --jit;after::after --disable-gems;after+JIT::after --disable-gems --jit' -v --repeat-count 24
before: ruby 2.6.0dev (2018-09-24 trunk 64821) [x86_64-linux]
before+JIT: ruby 2.6.0dev (2018-09-24 trunk 64821) +JIT [x86_64-linux]
after: ruby 2.6.0dev (2018-09-24 opt_and 64821) [x86_64-linux]
last_commit=opt_or
after+JIT: ruby 2.6.0dev (2018-09-24 opt_and 64821) +JIT [x86_64-linux]
last_commit=opt_or
Calculating -------------------------------------
before before+JIT after after+JIT
Optcarrot Lan_Master.nes 51.460 66.315 53.023 71.173 fps
Comparison:
Optcarrot Lan_Master.nes
after+JIT: 71.2 fps
before+JIT: 66.3 fps - 1.07x slower
after: 53.0 fps - 1.34x slower
before: 51.5 fps - 1.38x slower
[close https://github.com/ruby/ruby/pull/1963]
git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@64824 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
This enhances rb_vm_insn_addr2insn which retrieves a decoded insn number
from encoded insn.
The insn data table include not only decoded insn number, but also its
len, trace and non-trace version of encoded insn.
This table can be used to simplify trace instrumentation.
git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@64518 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
Spurious interrupts from SIGCHLD cause Mutex#sleep (via
ConditionVariable#wait) to return early and breaks some use
cases. Since these are outside the programs's control with
MJIT, we will only consider pending interrupts (e.g. those
from Thread#run) and signals which cause a Ruby-level Signal.trap
handler to fire as "spurious" wakeups.
[ruby-core:88537] [Feature #15002]
git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@64444 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
On a 64-bit system, this reduces rb_thread_t from 536 to 520 bytes.
Depending on the allocation, this can reduce cacheline access
for checking the abort_on_exception, report_on_exception and
pending_interrupt_queue_checked flags.
git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@64376 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
This reverts commit 194a6a2c68 (r64203).
Race conditions which caused the original reversion will be fixed
in the subsequent commit.
[ruby-core:88360] [Misc #14937]
git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@64352 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
This is just a refactoring.
The receiver of "invokesuper" was a boolean to represent if it is ZSUPER
or not. This was used in vm_search_super_method to prohibit ZSUPER call
in define_method. (It is currently prohibited because of the limitation
of the implementation.)
This change removes the hack by introducing an explicit flag,
VM_CALL_SUPER, to signal the information. Now, the implementation of
"invokesuper" is consistent with "send" instruction.
git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@64268 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
By this commit's changes in other files, now MJIT started to work on VC++.
Unfortunately some features are still broken and they'll be fixed later.
This also suppresses cl.exe's default output to stdout because there
seems to be no option to do it. Tweaking some log messages as well.
vm_core.h: declare `__declspec(dllimport)` to export them correctly on mswin.
vm_insnhelper.h: ditto
mjit.h: ditto
test_jit.rb: skipped some pending tests.
git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@64221 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
To reduce resource use and reduce CI failure; remove
timer-thread. Single-threaded Ruby processes (including forked
children) will never see extra thread overhead. This prevents
glibc and jemalloc from going into multi-threaded mode and
initializing locks or causing fragmentation via arena explosion.
The GVL is implements its own wait-queue as a ccan/list to
permit controlling wakeup order. Timeslice under contention is
handled by a designated timer thread (similar to choosing a
"patrol_thread" for current deadlock checking).
There is only one self-pipe, now, as wakeups for timeslice are
done independently using condition variables. This reduces FD
pressure slightly.
Signal handling is handled directly by a Ruby Thread (instead
of timer-thread) by exposing signal self-pipe to callers of
rb_thread_fd_select, native_sleep, rb_wait_for_single_fd, etc...
Acquiring, using, and releasing the self-pipe is exposed via 4
new internal functions:
1) rb_sigwait_fd_get - exclusively acquire timer_thread_pipe.normal[0]
2) rb_sigwait_fd_sleep - sleep and wait for signal (and no other FDs)
3) rb_sigwait_fd_put - release acquired result from rb_sigwait_fd_get
4) rb_sigwait_fd_migrate - migrate signal handling to another thread
after calling rb_sigwait_fd_put.
rb_sigwait_fd_migrate is necessary for waitpid callers because
only one thread can wait on self-pipe at a time, otherwise a
deadlock will occur if threads fight over the self-pipe.
TRAP_INTERRUPT_MASK is now set for the main thread directly in
signal handler via rb_thread_wakeup_timer_thread.
Originally, I wanted to use POSIX timers
(timer_create/timer_settime) for this. Unfortunately, this
proved unfeasible as Mutex#sleep resumes on spurious wakeups and
test/thread/test_cv.rb::test_condvar_timed_wait failed. Using
pthread_sigmask to mask out SIGVTALRM fixed that test, but
test/fiddle/test_function.rb::test_nogvl_poll proved there'd be
some unavoidable (and frequent) incompatibilities from that
approach.
Finally, this allows us to drop thread_destruct_lock and
interrupt current ec directly.
We don't need to rely on vm->thread_destruct_lock or a coherent
vm->running_thread on any platform. Separate timer-thread for
time slice and signal handling is relegated to thread_win32.c,
now.
[ruby-core:88088] [Misc #14937]
git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@64107 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
Otherwise, an altstack may live past ObjectSpace destruction
and xfree-ing the altstack will segfault.
[ruby-core:85621] [Feature #14487]
git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@64102 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
Repack rb_thread_struct, rb_execution_context_struct, args_info and
iseq_compile_data to save 1 word per struct.
re_pattern_buffer remains unpacked due to the possible binary
compatibility.
[Fix GH-1907]
Based on the patch
From: Lourens Naudé <lourens@bearmetal.eu>
git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@64096 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
Revert r63968 and cast at caller side to prevent unintentional casting.
git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@63972 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
* variable.c (rb_const_search): call #const_missing method on
private constants, as well as uninitialized constants.
[Feature #14328]
git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@63871 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
Reading win32/win32.c waitpid implementation, maybe waitpid(-1, ...)
on that platform will never conflict with mjit use of waitpid.
In any case, I've added WAITPID_USE_SIGCHLD macro to vm_core.h
so it can be easy for Linux/BSD users to test (hopefully!)
win32-compatible code.
git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@63855 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
The change is unstable on Windows. Please re-commit it when it correctly
supports Windows.
git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@63852 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
Some systems lack SIGCHLD or have incomplete SIGCHLD
implementations. So enable polling mode for them.
[ruby-core:87705] [Bug #14867]
git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@63795 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
Use a global SIGCHLD handler to guard all callers of rb_waitpid.
To work safely with multi-threaded programs, we introduce a
VM-wide waitpid_lock to be acquired BEFORE fork/vfork spawns the
process. This is to be combined with the new ruby_waitpid_locked
function used by mjit.c in a non-Ruby thread.
Ruby-level SIGCHLD handlers registered with Signal.trap(:CHLD)
continues to work as before and there should be no regressions
in any existing use cases.
Splitting the wait queues for PID > 0 and groups (PID <= 0)
ensures we favor PID > 0 callers.
The disabling of SIGCHLD in rb_f_system is longer necessary,
as we use deferred signal handling and no longer make ANY
blocking waitpid syscalls in other threads which could "beat"
the waitpid call made by rb_f_system.
We prevent SIGCHLD from firing in normal Ruby Threads and only
enable it in the timer-thread, to prevent spurious wakeups
from in test/-ext-/gvl/test_last_thread.rb with MJIT enabled.
I've tried to guard as much of the code for RUBY_SIGCHLD==0
using C "if" statements rather than CPP "#if" so to reduce
the likelyhood of portability problems as the compiler will
see more code.
We also work to suppress false-positives from
Process.wait(-1, Process::WNOHANG) to quiets warnings from
spec/ruby/core/process/wait2_spec.rb with MJIT enabled.
Lastly, we must implement rb_grantpt for ext/pty. We need a
MJIT-compatible way of supporting grantpt(3) which may spawn
the `pt_chown' binary and call waitpid(2) on it.
[ruby-core:87605] [Ruby trunk Bug#14867]
git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@63758 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
vm->sleeper is never modified in signal handlers or without GVL,
so there's no need for volatile hocus-pocus.
git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@63713 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
The current VM_INSTRUCTION_SIZE is 198, so the linear search
painful during a major GC phase.
I noticed rb_vm_insn_addr2insn2 showing up at the top of some
profiles while working on some malloc-related stuff, so I
decided to attack it.
Most notably, the benchmark/bm_vm3_gc.rb improves by over 40%:
https://80x24.org/spew/20180602220554.GA9991@whir/raw
[ruby-core:87361] [Feature #14814]
git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@63594 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
* vm_core.h (VM_ENV_DATA_INDEX_ENV_PROC): ep[VM_ENV_DATA_INDEX_ENV_PROC] is
allocated to mark a Proc which is created from iseq block.
However, `lep[0]` keeps Proc object itself as a block handler (Proc).
So we don't need to keep it.
* vm_core.h (VM_ENV_PROCVAL): ditto.
* vm.c (vm_make_env_each): do not need to keep blockprocval as special value.
* vm.c (vm_block_handler_escape): simply return Proc value.
* proc.c (proc_new): we don't need to check Env because a Proc type block
handler is a Proc object itself.
[Bug #14782]
* test/ruby/test_proc.rb: add a test for [Bug #14782]
git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@63494 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
st_table allows the use of st_shift to act as an order-preserving
queue while allowing fast lookups to prevent duplicate jobs.
In typical Ruby apps, this table will only have one entry
for gc_finalize_deferred_register.
git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@63451 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
rb_atomic_t is 32-bit on 64-bit platforms (including the popular
x86-64 Linux), so save 4 bytes on this structure. This doesn't
result in any final size reduction due to padding, yet, but
future changes are possible to shrink rb_execution_context_t
git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@63435 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
This ec->machine.regs is marked by GC. However jmp_buf is
not defined by us. There are chances of unaligned access.
We should force it VALUE-aligned.
git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@63375 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
Don't abuse struct RString to hold arbitrary memory region.
Raw pointer should just suffice.
git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@63368 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
Instead of allocating and registering the altstack in different
places, do it together to reduce code and improve readability.
When thread cache is enabled, storing altstack in rb_thread_t
is wasteful and we may reuse altstack in the same pthread.
This also lets us clearly allow use of xmalloc to allow GC to
recover from ENOMEM.
[ruby-core:85621] [Feature #14487]
git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@63213 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
This is fairly non-intrusive bugfix to prevent children
from trying to reach into thread stacks of the parent.
I will probably reuse this idea and redo r62934, too
(same bug).
* vm_core.h (typedef struct rb_vm_struct): add fork_gen counter
* thread.c (rb_thread_atfork_internal): increment fork_gen
* variable.c (struct autoload_data_i): store fork_gen
* variable.c (check_autoload_data): remove (replaced with get_...)
* variable.c (get_autoload_data): check fork_gen when retrieving
* variable.c (check_autoload_required): use get_autoload_data
* variable.c (rb_autoloading_value): ditto
* variable.c (rb_autoload_p): ditto
* variable.c (current_autoload_data): ditto
* variable.c (autoload_reset): reset fork_gen, adjust indent
* variable.c (rb_autoload_load): set fork_gen when setting state
* test/ruby/test_autoload.rb (test_autoload_fork): new test
[ruby-core:86410] [Bug #14634]
git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@63210 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
We need to mark default values for kwarg methods. This also fixes
Bootsnap. IBF iseq loading needed to mark iseqs as "having markable
objects".
git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@62851 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
Directly marking iseq operands allows us to eliminate the "mark array"
stored on ISEQ objects, which will reduce the amount of memory ISEQ
objects consume. This patch changes the iseq mark function to:
* Directly marks ISEQ operands
* Iterate over and mark child ISEQs
It also introduces two flags on the ISEQ object. In order to mark
instruction operands, we have to disassemble the instructions and find
the instruction parameters and types. Instructions may also be
translated to jump addresses. Instruction sequences may get marked by
the GC *while* they're mid flight (being compiled). The
`ISEQ_TRANSLATED` flag is used to indicate whether or not the
instructions have been translated to jump addresses so that when we
decode the instructions we know whether or not we need to go from jump
location back to original instruction or not.
Not all ISEQ objects have any markable objects embedded in their
instructions. We can detect whether or not an ISEQ has markable objects
in the instructions at compile time. If the instructions contain
markable objects, we set a flag `ISEQ_MARKABLE_ISEQ` on the ISEQ object.
This means that during the mark phase, we can skip decompilation if the
flag is *not* set. In other words, we can avoid decompilation of we
know in advance there is nothing to mark.
`once` instructions have an operand that contains the result of a
one-time compilation of a regex. Before this patch, that operand was
called an "inline cache", even though the struct was actually an "inline
storage". This patch changes the operand to be an "inline storage" so
that we can differentiate between caches that need marking (the inline
storage) and caches that don't need marking (inline cache).
[ruby-core:84909]
git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@62706 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
to be used for MJIT's optimization. It's not used for optimization
in this commit yet.
vm_core.h: added catch_except_p field.
iseq.c: show the flag in ISeq disasm for debugging.
git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@62654 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
* vm_insnhelper.c (vm_get_ev_const): add flag argument of
`rb_autoloading_value`.
* constant.h (rb_autoloading_value): moved the declaration from
vm_core.h for `rb_const_flag_t`. [ruby-core:85516] [Bug #14469]
git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@62394 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
No implicit cast is defined between these types. Should be explicit.
Also, NULL is defined to be ((void*)0) so not usable as a function
pointer value.
git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@62221 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
which has been developed by Takashi Kokubun <takashikkbn@gmail> as
YARV-MJIT. Many of its bugs are fixed by wanabe <s.wanabe@gmail.com>.
This JIT compiler is designed to be a safe migration path to introduce
JIT compiler to MRI. So this commit does not include any bytecode
changes or dynamic instruction modifications, which are done in original
MJIT.
This commit even strips off some aggressive optimizations from
YARV-MJIT, and thus it's slower than YARV-MJIT too. But it's still
fairly faster than Ruby 2.5 in some benchmarks (attached below).
Note that this JIT compiler passes `make test`, `make test-all`, `make
test-spec` without JIT, and even with JIT. Not only it's perfectly safe
with JIT disabled because it does not replace VM instructions unlike
MJIT, but also with JIT enabled it stably runs Ruby applications
including Rails applications.
I'm expecting this version as just "initial" JIT compiler. I have many
optimization ideas which are skipped for initial merging, and you may
easily replace this JIT compiler with a faster one by just replacing
mjit_compile.c. `mjit_compile` interface is designed for the purpose.
common.mk: update dependencies for mjit_compile.c.
internal.h: declare `rb_vm_insn_addr2insn` for MJIT.
vm.c: exclude some definitions if `-DMJIT_HEADER` is provided to
compiler. This avoids to include some functions which take a long time
to compile, e.g. vm_exec_core. Some of the purpose is achieved in
transform_mjit_header.rb (see `IGNORED_FUNCTIONS`) but others are
manually resolved for now. Load mjit_helper.h for MJIT header.
mjit_helper.h: New. This is a file used only by JIT-ed code. I'll
refactor `mjit_call_cfunc` later.
vm_eval.c: add some #ifdef switches to skip compiling some functions
like Init_vm_eval.
win32/mkexports.rb: export thread/ec functions, which are used by MJIT.
include/ruby/defines.h: add MJIT_FUNC_EXPORTED macro alis to clarify
that a function is exported only for MJIT.
array.c: export a function used by MJIT.
bignum.c: ditto.
class.c: ditto.
compile.c: ditto.
error.c: ditto.
gc.c: ditto.
hash.c: ditto.
iseq.c: ditto.
numeric.c: ditto.
object.c: ditto.
proc.c: ditto.
re.c: ditto.
st.c: ditto.
string.c: ditto.
thread.c: ditto.
variable.c: ditto.
vm_backtrace.c: ditto.
vm_insnhelper.c: ditto.
vm_method.c: ditto.
I would like to improve maintainability of function exports, but I
believe this way is acceptable as initial merging if we clarify the
new exports are for MJIT (so that we can use them as TODO list to fix)
and add unit tests to detect unresolved symbols.
I'll add unit tests of JIT compilations in succeeding commits.
Author: Takashi Kokubun <takashikkbn@gmail.com>
Contributor: wanabe <s.wanabe@gmail.com>
Part of [Feature #14235]
---
* Known issues
* Code generated by gcc is faster than clang. The benchmark may be worse
in macOS. Following benchmark result is provided by gcc w/ Linux.
* Performance is decreased when Google Chrome is running
* JIT can work on MinGW, but it doesn't improve performance at least
in short running benchmark.
* Currently it doesn't perform well with Rails. We'll try to fix this
before release.
---
* Benchmark reslts
Benchmarked with:
Intel 4.0GHz i7-4790K with 16GB memory under x86-64 Ubuntu 8 Cores
- 2.0.0-p0: Ruby 2.0.0-p0
- r62186: Ruby trunk (early 2.6.0), before MJIT changes
- JIT off: On this commit, but without `--jit` option
- JIT on: On this commit, and with `--jit` option
** Optcarrot fps
Benchmark: https://github.com/mame/optcarrot
| |2.0.0-p0 |r62186 |JIT off |JIT on |
|:--------|:--------|:--------|:--------|:--------|
|fps |37.32 |51.46 |51.31 |58.88 |
|vs 2.0.0 |1.00x |1.38x |1.37x |1.58x |
** MJIT benchmarks
Benchmark: https://github.com/benchmark-driver/mjit-benchmarks
(Original: https://github.com/vnmakarov/ruby/tree/rtl_mjit_branch/MJIT-benchmarks)
| |2.0.0-p0 |r62186 |JIT off |JIT on |
|:----------|:--------|:--------|:--------|:--------|
|aread |1.00 |1.09 |1.07 |2.19 |
|aref |1.00 |1.13 |1.11 |2.22 |
|aset |1.00 |1.50 |1.45 |2.64 |
|awrite |1.00 |1.17 |1.13 |2.20 |
|call |1.00 |1.29 |1.26 |2.02 |
|const2 |1.00 |1.10 |1.10 |2.19 |
|const |1.00 |1.11 |1.10 |2.19 |
|fannk |1.00 |1.04 |1.02 |1.00 |
|fib |1.00 |1.32 |1.31 |1.84 |
|ivread |1.00 |1.13 |1.12 |2.43 |
|ivwrite |1.00 |1.23 |1.21 |2.40 |
|mandelbrot |1.00 |1.13 |1.16 |1.28 |
|meteor |1.00 |2.97 |2.92 |3.17 |
|nbody |1.00 |1.17 |1.15 |1.49 |
|nest-ntimes|1.00 |1.22 |1.20 |1.39 |
|nest-while |1.00 |1.10 |1.10 |1.37 |
|norm |1.00 |1.18 |1.16 |1.24 |
|nsvb |1.00 |1.16 |1.16 |1.17 |
|red-black |1.00 |1.02 |0.99 |1.12 |
|sieve |1.00 |1.30 |1.28 |1.62 |
|trees |1.00 |1.14 |1.13 |1.19 |
|while |1.00 |1.12 |1.11 |2.41 |
** Discourse's script/bench.rb
Benchmark: https://github.com/discourse/discourse/blob/v1.8.7/script/bench.rb
NOTE: Rails performance was somehow a little degraded with JIT for now.
We should fix this.
(At least I know opt_aref is performing badly in JIT and I have an idea
to fix it. Please wait for the fix.)
*** JIT off
Your Results: (note for timings- percentile is first, duration is second in millisecs)
categories_admin:
50: 17
75: 18
90: 22
99: 29
home_admin:
50: 21
75: 21
90: 27
99: 40
topic_admin:
50: 17
75: 18
90: 22
99: 32
categories:
50: 35
75: 41
90: 43
99: 77
home:
50: 39
75: 46
90: 49
99: 95
topic:
50: 46
75: 52
90: 56
99: 101
*** JIT on
Your Results: (note for timings- percentile is first, duration is second in millisecs)
categories_admin:
50: 19
75: 21
90: 25
99: 33
home_admin:
50: 24
75: 26
90: 30
99: 35
topic_admin:
50: 19
75: 20
90: 25
99: 30
categories:
50: 40
75: 44
90: 48
99: 76
home:
50: 42
75: 48
90: 51
99: 89
topic:
50: 49
75: 55
90: 58
99: 99
git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@62197 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
that allows to JIT-compile Ruby methods by generating C code and
using C compiler. See the first comment of mjit.c to know what this
file does.
mjit.c is authored by Vladimir Makarov <vmakarov@redhat.com>.
After he invented great method JIT infrastructure for MRI as MJIT,
Lars Kanis <lars@greiz-reinsdorf.de> sent the patch to support MinGW
in MJIT. In addition to merging it, I ported pthread to Windows native
threads. Now this MJIT infrastructure can be compiled on Visual Studio.
This commit simplifies mjit.c to decrease code at initial merge. For
example, this commit does not provide multiple JIT threads support.
We can resurrect them later if we really want them, but I wanted to minimize
diff to make it easier to review this patch.
`/tmp/_mjitXXX` file is renamed to `/tmp/_ruby_mjitXXX` because non-Ruby
developers may not know the name "mjit" and the file name should make
sure it's from Ruby and not from some harmful programs. TODO: it may be
better to store this to some temporary directory which Ruby is already using
by Tempfile, if it's not bad for performance.
mjit.h: New. It has `mjit_exec` interface similar to `vm_exec`, which is
for triggering MJIT. This drops interface for AOT compared to the original
MJIT.
Makefile.in: define macros to let MJIT know the path of MJIT header.
Probably we can refactor this to reduce the number of macros (TODO).
win32/Makefile.sub: ditto.
common.mk: compile mjit.o and mjit_compile.o. Unlike original MJIT, this
commit separates MJIT infrastructure and JIT compiler code as independent
object files. As initial patch is NOT going to have ultra-fast JIT compiler,
it's likely to replace JIT compiler, e.g. original MJIT's compiler or some
future JIT impelementations which are not public now.
inits.c: define MJIT module. This is added because `MJIT.enabled?` was
necessary for testing.
test/lib/zombie_hunter.rb: skip if `MJIT.enabled?`. Obviously this
wouldn't work with current code when JIT is enabled.
test/ruby/test_io.rb: skip this too. This would make no sense with MJIT.
ruby.c: define MJIT CLI options. As major difference from original MJIT,
"-j:l"/"--jit:llvm" are renamed to "--jit-cc" because I want to support
not only gcc/clang but also cl.exe (Visual Studio) in the future. But it
takes only "--jit-cc=gcc", "--jit-cc=clang" for now. And only long "--jit"
options are allowed since some Ruby committers preferred it at Ruby
developers Meeting on January, and some of options are renamed.
This file also triggers to initialize MJIT thread and variables.
eval.c: finalize MJIT worker thread and variables.
test/ruby/test_rubyoptions.rb: fix number of CLI options for --jit.
thread_pthread.c: change for pthread abstraction in MJIT. Prefix rb_ for
functions which are used by other files.
thread_win32.c: ditto, for Windows. Those pthread porting is one of major
works that YARV-MJIT created, which is my fork of MJIT, in Feature 14235.
thread.c: follow rb_ prefix changes
vm.c: trigger MJIT call on VM invocation. Also trigger `mjit_mark` to avoid
SEGV by race between JIT and GC of ISeq. The improvement was provided by
wanabe <s.wanabe@gmail.com>.
In JIT compiler I created and am going to add in my next commit, I found
that having `mjit_exec` after `vm_loop_start:` is harmful because the
JIT-ed function doesn't proceed other ISeqs on RESTORE_REGS of leave insn.
Executing non-FINISH frame is unexpected for my JIT compiler and
`exception_handler` triggers executions of such ISeqs. So `mjit_exec`
here should be executed only when it directly comes from `vm_exec` call.
`RubyVM::MJIT` module and `.enabled?` method is added so that we can skip
some tests which don't expect JIT threads or compiler file descriptors.
vm_insnhelper.h: trigger MJIT on method calls during VM execution.
vm_core.h: add fields required for mjit.c. `bp` must be `cfp[6]` because
rb_control_frame_struct is likely to be casted to another struct. The
last position is the safest place to add the new field.
vm_insnhelper.c: save initial value of cfp->ep as cfp->bp. This is an
optimization which are done in both MJIT and YARV-MJIT. So this change
is added in this commit. Calculating bp from ep is a little heavy work,
so bp is kind of cache for it.
iseq.c: notify ISeq GC to MJIT. We should know which iseq in MJIT queue
is GCed to avoid SEGV. TODO: unload some GCed units in some safe way.
gc.c: add hooks so that MJIT can wait GC, and vice versa. Simultaneous
JIT and GC executions may cause SEGV and so we should synchronize them.
cont.c: save continuation information in MJIT worker. As MJIT shouldn't
unload JIT-ed code which is being used, MJIT wants to know full list of
saved execution contexts for continuation and detect ISeqs in use.
mjit_compile.c: added empty JIT compiler so that you can reuse this commit
to build your own JIT compiler. This commit tries to compile ISeqs but
all of them are considered as not supported in this commit. So you can't
use JIT compiler in this commit yet while we added --jit option now.
Patch author: Vladimir Makarov <vmakarov@redhat.com>.
Contributors:
Takashi Kokubun <takashikkbn@gmail.com>.
wanabe <s.wanabe@gmail.com>.
Lars Kanis <lars@greiz-reinsdorf.de>.
Part of Feature 12589 and 14235.
git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@62189 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
We treat this as "int" through the vm_living_thread_num API
anyways, and "pid_t" is still 32-bits with glibc on 64-bit
platforms. I expect it'll be a long time before anybody needs
more than 2 billion native threads. For now, let's save one
cacheline on x86-64 (as reported by pahole(1)):
before: size: 1288, cachelines: 21, members: 45
after: size: 1280, cachelines: 20, members: 45
git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@62075 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
* vm_core.h (VM_INSN_INFO_TABLE_IMPL): use 1 (binary search)
because 2 (succinct bitvector) doesn't work on Rasbian (x86_64).
git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@61748 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
Because the name "code_range" is ambiguous with encoding's.
Abbreviations ("crange", and "cr") are also renamed to "loc".
The traditional "code_location" (a pair of lineno and column) is
renamed to "code_position". Abbreviations are also renamed
(first_loc to beg_pos, and last_loc to end_pos).
git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@61721 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
* insns.def (getblockparamproxy): introduce new instruction to return
the `rb_block_param_proxy` object if possible. This object responds
to `call` method and invoke given block (completely similar to `yield`).
* method.h (OPTIMIZED_METHOD_TYPE_BLOCK_CALL): add new optimized call type
which is for `rb_block_param_proxy.cal`.
* vm_insnhelper.c (vm_call_method_each_type): ditto.
* vm_insnhelper.c (vm_call_opt_block_call): ditto.
* vm_core.h (BOP_CALL, PROC_REDEFINED_OP_FLAG): add check for `Proc#call`
redefinition.
* compile.c (iseq_compile_each0): compile to use new insn
`getblockparamproxy` for method call.
git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@61659 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
This commit removes ISEQ_TYPE_DEFINED_GUARD because it is no longer
needed. And this introduces ISEQ_TYPE_PLAIN which means that the iseq
does nothing special but just wrap an expression. Currently, this is
used for once execution: `/foo#{ bar }baz/o`.
git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@61601 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
It is too error-prone to pass IMEMO_IFUNC object as NODE*.
git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@61592 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
* gc.c (rb_raw_obj_info): check block before using it.
* vm_core.h (vm_block_iseq): r61565 introduced NULL check but this
check is only needed by `rb_raw_obj_info()` and it is called at GC
debug mode. Above fix for `rb_raw_obj_info()` solves this problem and
NULL check should not be needed any more.
git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@61571 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
Currently, VM_INSN_INFO_TABLE_IMPL == 0 means linear search, and
VM_INSN_INFO_TABLE_IMPL == 1 means binary search. I plan to add
succinct bitvector algorithm later.
git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@61537 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
This makes TracePoint a bit fast by reducing cache misses of
`get_insn_info_binary_search`.
Also, I plan to use succinct bitvector algorithm for `get_insn_info`
instead of binary search. This change will make it easy.
git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@61536 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
This factors rb_iseq_constant_body#insns_info and #insns_info_size to
struct iseq_insn_info.
git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@61534 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
* vm_core.h (rb_vm_t): move `rb_execution_context_t::safe_level` to
`rb_vm_t::safe_level_` because `$SAFE` is a process (VM) global state.
* vm_core.h (rb_proc_t): remove `rb_proc_t::safe_level` because `Proc`
objects don't need to keep `$SAFE` at the creation.
Also make `is_from_method` and `is_lambda` as 1 bit fields.
* cont.c (cont_restore_thread): no need to keep `$SAFE` for Continuation.
* eval.c (ruby_cleanup): use `rb_set_safe_level_force()` instead of access
`vm->safe_level_` directly.
* eval_jump.c: End procs `END{}` doesn't keep `$SAFE`.
* proc.c (proc_dup): removed and introduce `rb_proc_dup` in vm.c.
* safe.c (rb_set_safe_level): don't check `$SAFE` 1 -> 0 changes.
* safe.c (safe_setter): use `rb_set_safe_level()`.
* thread.c (rb_thread_safe_level): `Thread#safe_level` returns `$SAFE`.
It should be obsolete.
* transcode.c (load_transcoder_entry): `rb_safe_level()` only returns
0 or 1 so that this check is not needed.
* vm.c (vm_proc_create_from_captured): don't need to keep `$SAFE` for Proc.
* vm.c (rb_proc_create): renamed to `proc_create`.
* vm.c (rb_proc_dup): moved from proc.c.
* vm.c (vm_invoke_proc): do not need to set and restore `$SAFE`
for `Proc#call`.
* vm_eval.c (rb_eval_cmd): rename a local variable to represent clearer
meaning.
* lib/drb/drb.rb: restore `$SAFE`.
* lib/erb.rb: restore `$SAFE`, too.
* test/lib/leakchecker.rb: check `$SAFE == 0` at the end of tests.
* test/rubygems/test_gem.rb: do not set `$SAFE = 1`.
* bootstraptest/test_proc.rb: catch up this change.
* spec/ruby/optional/capi/string_spec.rb: ditto.
* test/bigdecimal/test_bigdecimal.rb: ditto.
* test/fiddle/test_func.rb: ditto.
* test/fiddle/test_handle.rb: ditto.
* test/net/imap/test_imap_response_parser.rb: ditto.
* test/pathname/test_pathname.rb: ditto.
* test/readline/test_readline.rb: ditto.
* test/ruby/test_file.rb: ditto.
* test/ruby/test_optimization.rb: ditto.
* test/ruby/test_proc.rb: ditto.
* test/ruby/test_require.rb: ditto.
* test/ruby/test_thread.rb: ditto.
* test/rubygems/test_gem_specification.rb: ditto.
* test/test_tempfile.rb: ditto.
* test/test_tmpdir.rb: ditto.
* test/win32ole/test_win32ole.rb: ditto.
* test/win32ole/test_win32ole_event.rb: ditto.
git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@61510 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
2.5's line coverage measurement was about two times slower than 2.4
because of two reasons; (1) vm_trace uses rb_iseq_event_flags (which
takes O(n) currently where n is the length of iseq) to get an event
type, and (2) RUBY_EVENT_LINE uses setjmp to call an event hook.
This change adds a special event for line coverage,
RUBY_EVENT_COVERAGE_LINE, and adds `tracecoverage` instructions where
the event occurs in iseq.
`tracecoverage` instruction calls an event hook without vm_trace.
And, RUBY_EVENT_COVERAGE_LINE is an internal event which does not
use setjmp.
This change also cancells lineno change due to the deletion of trace
instructions [Feature #14104]. So fixes [Bug #14191].
git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@61350 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
* vm.c: introduce `ruby_vm_event_enabled_flags` which represents which
event flags are enabled before.
* vm_trace.c: do not turn off `trace_` prefix instructions because turn on
overhead is a matter if a program repeats turn on and turn off frequently.
* iseq.c (finish_iseq_build): respect `ruby_vm_event_enabled_flags`.
* vm_insnhelper.c (vm_trace): check `ruby_vm_event_flags` and disable
lazy trace-off technique (do not disable traces).
git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@61122 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
This change moves RUBY_EVENT_COVERAGE from include/ruby/ruby.h to
vm_core.h and renames it to RUBY_EVENT_COVERAGE_BRANCH.
git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@61049 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
This change makes each ISeq keep NODE's code range. This information is
needed for method coverage.
git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@61025 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
Traditionally, method coverage measurement was implemented by inserting
`trace2` instruction to the head of method iseq. So, it just measured
methods defined by `def` keyword.
This commit drastically changes the measuring mechanism of method
coverage; at `RUBY_EVENT_CALL`, it keeps a hash from rb_method_entry_t*
to runs (i.e., it counts the runs per method entry), and at
`Coverage.result`, it creates the result hash by enumerating all
`rb_method_entry_t*` objects (by `ObjectSpace.each_object`).
git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@61023 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
Currently the Fiber stack size is small for powerpc64 and it causes
test/ruby/test_backtrace.rb test to break, since it is using a 8kb stack
size.
It breaks on powerpc64 due to the fact that a frame in the stack is
usually 50% bigger on powerpc64 compared to Intel, due to some
considerations:
* The powerpc64 minimum frame is 2x bigger than on Intel
* Powerpc has more registers that might be saved in the frame compared
to Intel.
I ran the same ruby test that is failing on both Intel and Powerpc, and
each Fiber frame is ~50% bigger on powerpc64 for every single lambda
function, thus, we need to increase the stack size on powerpc64 to
accomodate the same tests/applications.
This fixes bug#13757.
Signed-off-by: Breno Leitao <leitao@debian.org>
git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@61020 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
* vm_insnhelper.c (vm_trace): use EXEC_EVENT_HOOK() instead of
EXEC_EVENT_HOOK_VM_TRACE(). The latter macro assumes
`ruby_vm_event_flags` is constant in `vm_trace()` function,
but it can be changed in hook functions.
* vm_core.h (EXEC_EVENT_HOOK_VM_TRACE): removed.
git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@60923 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
* vm_core.h (rb_iseq_t::aux): add `trace_events` which represents
which events are enabled on this iseq. With this information,
we can skip useless trace-on changes for ISeqs.
* vm_trace.c (RUBY_EVENTS_TRACE_BY_ISEQ): moved to iseq.h and rename it
with ISEQ_TRACE_EVENTS.
* iseq.h: introduce ISEQ_USE_COMPILE_DATA iseq (imemo) flag to represent
COMPILE_DATA is available. In other words, iseq->aux.trace_events is not
available when this flag is set.
* ISEQ_COMPILE_DATA() is changed from a macro.
* ISEQ_COMPILE_DATA_ALLOC() is added.
* ISEQ_COMPILE_DATA_CLEAR() is added.
* iseq.c: use them.
git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@60838 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
* vm_core.h (EXEC_EVENT_HOOK_VM_TRACE): added to pass vm_event_flags
(== ruby_vm_event_flags) as a macro parameter.
* vm_insnhelper.c (vm_trace): use an added macro.
git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@60819 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
* vm_core.h (rb_thread_t): remove rb_thread_t::event_hooks.
* vm_trace.c: all hooks are connected to vm->event_hooks and
add rb_event_hook_t::filter::th to filter invoke thread.
It will simplify invoking hooks code.
* thread.c (thread_start_func_2): clear thread specific trace_func.
* test/ruby/test_settracefunc.rb: add a test for Thread#add_trace_func.
git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@60776 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
* iseq.h (iseq_line_info_entry): rename to iseq_insn_info_entry.
* vm_core.h (rb_iseq_constant_body): rename field name line_info_table
to insns_info and also from line_info_size to insns_info_size.
* compile.c (INSN): add struct insn_info to contain per insn information.
* compile.c (add_insn_info): added to add new insn_info entry.
git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@60726 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
* vm_core.h (rb_execution_context_t): renmae ec::fiber to
ec::fiber_ptr make consistent with ec::thread_ptr.
git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@60670 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
* eval_error.c (rb_threadptr_error_print): renamed to
rb_ec_error_print() and it accepts `ec`.
git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@60545 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
* vm.c (rb_thread_method_id_and_class): rename to
rb_ec_frame_method_id_and_class() and accepts `ec` instead of `th`.
git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@60513 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
* vm_core.h: move rb_thread_t::passed_block_handler to
rb_execution_context_t::passed_block_handler.
Also move rb_thread_t::passed_bmethod_me to
rb_execution_context_t::passed_bmethod_me.
git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@60503 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
to represent execution context [Feature #14038]
* vm_core.h (rb_thread_t): rb_thread_t::ec is now a pointer.
There are many code using `th` to represent execution context
(such as cfp, VM stack and so on). To access `ec`, they need to
use `th->ec->...` (adding one indirection) so that we need to
replace them by passing `ec` instead of `th`.
* vm_core.h (GET_EC()): introduced to access current ec. Also
remove `ruby_current_thread` global variable.
* cont.c (rb_context_t): introduce rb_context_t::thread_ptr instead of
rb_context_t::thread_value.
* cont.c (ec_set_vm_stack): added to update vm_stack explicitly.
* cont.c (ec_switch): added to switch ec explicitly.
* cont.c (rb_fiber_close): added to terminate fibers explicitly.
git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@60440 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
[Feature #14045]
* insns.def (getblockparam, setblockparam): add special access
instructions for block parameters.
getblockparam checks VM_FRAME_FLAG_MODIFIED_BLOCK_PARAM and
if it is not set this instruction creates a Proc object from
a given blcok and set VM_FRAME_FLAG_MODIFIED_BLOCK_PARAM.
setblockparam is similar to setlocal, but set
VM_FRAME_FLAG_MODIFIED_BLOCK_PARAM.
* compile.c: use get/setblockparm instead get/setlocal instructions.
Note that they are used for method local block parameters (def m(&b)),
not for block local method parameters (iter{|&b|).
* proc.c (get_local_variable_ptr): creates Proc object for
Binding#local_variable_get/set.
* safe.c (safe_setter): we need to create Proc objects for postponed
block parameters when $SAFE is changed.
* vm_args.c (args_setup_block_parameter): used only for block local blcok
parameters.
* vm_args.c (vm_caller_setup_arg_block): if called with
VM_CALL_ARGS_BLOCKARG_BLOCKPARAM flag then passed block values should be
a block handler.
* test/ruby/test_optimization.rb: add tests.
* benchmark/bm_vm1_blockparam*: added.
git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@60397 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
Because NaCl and PNaCl are already sunset status.
see https://bugs.chromium.org/p/chromium/issues/detail?id=239656#c160
configure.ac: Patch for this file was provided by @nobu.
[Feature #14041][ruby-core:83497][fix GH-1726]
git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@60374 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
vm_insnhelper.h: ditto.
All changes are for reducing changes required to introduce JIT compiler.
Unlike functions that can be inlined by header, those variables should
be shared with JIT-ed code. This will help reducing cost of rebase
against upstream.
[close GH-1720]
git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@60231 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
* cont.c (rb_context_t): introduce saved_ec instaad of saved_thread.
We only need to transfer ec data (not all of thread data).
Introduce `thread_value` field to point creation thread.
To acccess this field, `cont_thread_value()` is introduced.
* vm.c (rb_execution_context_mark): remove `static` and use it
from cont.c (use this function instead of `rb_thread_mark`).
* vm_insnhelper.c (rb_vm_push_frame): accept ec instead of th.
git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@59831 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
* vm_insnhelper.c (rb_threadptr_stack_overflow): rb_fatal is not
available during GC. raise the preallocated fatal error.
git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@59618 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
* vm_core.h: Ruby processes run with two stacks, a machine stack and a
VM stack. To make it clear, this fix renames
rb_execution_context_t::stack(_size) to vm_stack(_size).
git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@59563 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
* vm_core.h (rb_vm_struct): separate trap_list cmd and safe to
each arrays, to shrink the size.
git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@59434 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
These caused numerous CI failures I haven't been able to
reproduce [ruby-core:82102]
git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@59364 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
The same hash keys may be loaded from tainted data sources
frequently (e.g. parsing headers from socket or loading
YAML data from a file). If a non-tainted fstring already
exists (because the application expects the hash key),
cache and deduplicate the tainted version in the new
tainted_frozen_strings table.
For non-embedded strings, this also allows sharing with the
underlying malloc-ed data.
* vm_core.h (rb_vm_struct): add tainted_frozen_strings
* vm.c (ruby_vm_destruct): free tainted_frozen_strings
(Init_vm_objects): initialize tainted_frozen_strings
(rb_vm_tfstring_table): accessor for tainted_frozen_strings
* internal.h: declare rb_fstring_existing, rb_vm_tfstring_table
* hash.c (fstring_existing_str): remove (moved to string.c)
(hash_aset_str): use rb_fstring_existing
* string.c (rb_fstring_existing): new, based on fstring_existing_str
(tainted_fstr_update): new
(rb_fstring_existing0): new, based on fstring_existing_str
(rb_tainted_fstring_existing): new, special case for tainted strings
(rb_str_free): delete from tainted_frozen_strings table
* test/ruby/test_optimization.rb (test_hash_reuse_fstring): new test
[ruby-core:82012] [Bug #13737]
git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@59354 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
* vm_core.h (rb_thread_ptr): added to replace GetThreadPtr() macro.
* thread.c (in some functions: use "target_th" instead of "th" to make clear
that it is not a current thread.
git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@59192 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
* vm_core.h (rb_thread_t): move several fields which are copied at cont.c
to rb_execution_context_t.
git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@59177 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
* vm_core.h (rb_vm_tag): move jmpbuf which should be stored at the
last in TH_PUSH_TAG, between tag and state so ensure to be
accessible. reapplied r40806.
git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@59160 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
* vm_core.h (rb_thread_t::tag_state): move to "rb_vm_tag::state".
Lifetime of "state" should be same as current tag.
git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@59159 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
* vm_core.h (rb_thread_t): rename rb_thread_t::state to tag_state
to make it clear.
git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@59158 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
Return value of EXEC_TAG() is saved by "int state".
Instead of "int", use "enum ruby_tag_type". First EXEC_TAG()
value should be 0, so that define TAG_NONE (= 0) and use it.
Some code used "status" instead of "state". To make them clear,
rename them to state.
We can change variable name from "state" to "tag_state", but this
ticket doesn't contain it.
git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@59155 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
* vm_core.h (OPT_CALL_CFUNC_WITHOUT_FRAME): removed because nobody use it.
This optimization tries to call C-methods without pushing VM frames,
however no big improvements compare with this complexity.
git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@59076 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
* thread.c (rbuy_kill): removed. This function is used
with SIGSEGV, SIGBUS, SIGKILL, SIGILL, SIGFPE and SIGSTOP
and these signals are affect immediately. So that `kill(2)'
is enough for them.
* signal.c (rb_f_kill): ditto.
* vm_core.h (rb_thread_t::interrupt_cond): removed because
only `ruby_kill()' uses this field.
* test/ruby/test_signal.rb: Without this patch sending SIGSTOP to own
process wait another interrupt even if another process sends SIGCONT.
git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@59066 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
* vm_core.h (vm_block_handler_verify): this function only checks
the given block handler with VM_ASSERT(). No need to return any value.
git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@59045 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
VM_FRAME_FLAG_LAMBDA. So that lambda block frame is
VM_FRAME_MAGIC_BLOCK | VM_FRAME_FLAG_LAMBDA.
* vm_core.h: remvoe VM_FRAME_MAGIC_PROC because nobody uses it.
git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@59006 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
* proc.c (ruby_binding_data_type): set RUBY_TYPED_WB_PROTECTED.
Insert write barriers (WBs) to all of writes.
git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@58988 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
Today, it increases IO#close performance with many threads:
Execution time (sec)
name trunk after
vm_thread_close 4.276 3.018
Speedup ratio: compare with the result of `trunk' (greater is better)
name after
vm_thread_close 1.417
This speedup comes because rb_notify_fd_close only scans threads
inside rb_thread_io_blocking_region, not all threads in the VM.
In the future, this type data structure may allow us to notify
waiters of multiple FDs on a single thread (when using
Fibers).
* thread.c (struct waiting_fd): declare
(rb_thread_io_blocking_region): use on-stack list waiter
(rb_notify_fd_close): walk vm->waiting_fds instead
(call_without_gvl): remove old field setting
(th_init): ditto
* vm_core.h (typedef struct rb_vm_struct): add waiting_fds list
* (typedef struct rb_thread_struct): remove waiting_fd field
(rb_vm_living_threads_init): initialize waiting_fds list
I am now kicking myself for not thinking about this 3 years ago
when I introduced ccan/list in [Feature #9632] to optimize this
same function :<
git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@58812 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
The goal is to reduce rb_context_t and rb_fiber_t size
by removing the need to store the entire rb_thread_t in
there.
[ruby-core:81045] Work-in-progress: soon, we will move more fields here.
git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@58614 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
The current range based current values of:
TIME_QUANTUM_USEC=100000
RUBY_THREAD_PRIORITY_MAX=3
RUBY_THREAD_PRIORITY_MIN=-3
Gives a range of 12500..800000, plenty enough for a 32-bit
integer. Clamping this also reduces potential implementation
bugs between 32-bit and 64-bit platforms.
I may consider a further reduction to uint16_t in the future
for M:N threading, but some users may want slightly larger
time quantums.
* vm_core.h (rb_thread_t): use 32-bit running_time_us
git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@58591 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
* vm_core.h (rb_vm_register_special_exception): make the contents
of special exception messages static.
git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@58285 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
* vm_core.h (ruby_special_exceptions): renamed
ruby_error_closed_stream as ruby_error_stream_closed, like the
message.
git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@58284 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
* internal.h: introduce imemo_type_p() which checks the given value is
T_IMEMO and imemo_type() == given imemo_type.
git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@58268 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
Unfortunately this enlarges insns.def by yet another
instruction. However, it is much prettier than opt_str_freeze
in use, and maybe we can avoid having so many instructions in
the future.
[ruby-core:80368]
* insns.def (DEFINE_INSN): new instruction: opt_str_uminus (maybe temporary)
* compile.c (iseq_compile_each0): split instructions
* test/ruby/test_optimization.rb (test_string_uminus): new test
* vm.c (vm_init_redefined_flag): set redefinintion flag for uminus
* vm_core.h (enum ruby_basic_operators): add BOP_UMINUS
git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@58144 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
* vm_core.h (vm_block_handler_verify): some compilers warn about comparison
with enum value and 0. Real assertion is in vm_block_handler_type() so
we only need to call vm_block_handler_type() and the value should be TRUE.
git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@57557 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
Use volatile instead of optnone to avoid optimization which causes
segmentation faults.
Patch by Dimitry Andric. [ruby-core:78531] [Bug #13014]
git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@57020 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
* vm_core.h (rb_vm_struct): trace_running should be a counter but
not a bit flag. [ruby-core:78514] [Bug #13011]
Author: David Rodríguez <deivid.rodriguez@gmail.com>
git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@57002 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
* vm_core.h (VM_ASSERT): stringify expr here before expansion in
RUBY_ASSERT_WHEN.
git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@56378 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
and VM_FRAME_CFRAME_P().
Most of case, RUBY_VM_NORMAL_ISEQ_P() is no
longer needed.
* vm_core.h: introduce rb_obj_is_iseq().
* cont.c, vm.c: VM_FRAME_MAGIC_DUMMY with
VM_FRAME_FLAG_CFRAME.
git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@55804 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
are imemo objects (imemo_env).
* NEWS: describe this change. I believe nobody touch these objects
because there are no method defined.
* vm_core.h: remove the following definitions.
* rb_cEnv decl.
* GetEnvPtr() because Env is no longer T_DATA object.
* vm_core.h (rb_env_t): fix layout for imemo values.
* vm_core.h (vm_assert_env): added.
* vm_core.h (vm_env_new): added.
git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@55768 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
[Bug #12628]
This patch introduce many changes.
* Introduce concept of "Block Handler (BH)" to represent
passed blocks.
* move rb_control_frame_t::flag to ep[0] (as a special local
variable). This flags represents not only frame type, but also
env flags such as escaped.
* rename `rb_block_t` to `struct rb_block`.
* Make Proc, Binding and RubyVM::Env objects wb-protected.
Check [Bug #12628] for more details.
git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@55766 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
instead of setting rb_thread_t::cfp directly.
* vm_insnhelper.c (vm_pop_frame): return the result of
finish frame or not.
git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@55755 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
* [Feature #12005] Unify Fixnum and Bignum into Integer
* include/ruby/ruby.h (rb_class_of): Return rb_cInteger for fixnums.
* insns.def (INTEGER_REDEFINED_OP_FLAG): Unified from
FIXNUM_REDEFINED_OP_FLAG and BIGNUM_REDEFINED_OP_FLAG.
* vm_core.h: Ditto.
* vm_insnhelper.c (opt_eq_func): Use INTEGER_REDEFINED_OP_FLAG instead
of FIXNUM_REDEFINED_OP_FLAG.
* vm.c (vm_redefinition_check_flag): Use rb_cInteger instead of
rb_cFixnum and rb_cBignum.
(C): Use Integer instead of Fixnum and Bignum.
* numeric.c (fix_succ): Removed.
(Init_Numeric): Define Fixnum as Integer.
* bignum.c (bignew): Use rb_cInteger instead of Rb_cBignum.
(rb_int_coerce): replaced from rb_big_coerce and return fixnums
as-is.
(Init_Bignum): Define Bignum as Integer.
Don't define ===.
* error.c (builtin_class_name): Return "Integer" for fixnums.
* sprintf.c (ruby__sfvextra): Use rb_cInteger instead of rb_cFixnum.
* ext/-test-/testutil: New directory to test.
Currently it provides utilities for fixnum and bignum.
* ext/json/generator/generator.c: Define mInteger_to_json.
* lib/mathn.rb (Fixnum#/): Redefinition removed.
git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@55024 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
* vm_core.h (rb_vm_struct): make at_exit a single linked list but
not RArray, not to mark the registered functions by the write
barrier. based on the patches by Evan Phoenix.
[ruby-core:73908] [Bug #12095]
git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@54484 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
* error.c (rb_compile_err_append): rb_thread_t::base_block is no
longer used.
* iseq.c (rb_iseq_compile_with_option): ditto, no protection is
needed.
git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@54345 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
normal
nobu
ko1
k0kubun
yui-knk
svn
mame
shyouhei
kazu
ktsj
naruse
tenderlove
hsbt
shugo
akr