thread_pthread.c: timer thread stack size
* thread_pthread.c (rb_thread_create_timer_thread): define the stack
size for timer thread at compile time.
* thread_pthread.c (rb_thread_create_timer_thread): expand timer
thread stack size to get rid of segfault on FreeBSD/powerpc64.
based on the patch by Steve Wills at [ruby-core:59923].
[ruby-core:56590] [Bug #8783]
* thread_pthread.c (rb_thread_create_timer_thread): fix for platforms
where PTHREAD_STACK_MIN is a dynamic value and not a compile-time
constant. [ruby-dev:47911] [Bug #9436]
git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/branches/ruby_1_9_3@44940 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
* configure.in: check struct timeval exist or not.
* include/ruby/missing.h (struct timeval): check HAVE_STRUCT_TIMEVAL
properly. and don't include sys/time.h if struct timeval exist.
* file.c: include sys/time.h explicitly.
* random.c: ditto.
* thread_pthread.c: ditto.
* time.c: ditto.
* ext/date/date_strftime.c: ditto.
* include/ruby/missing.h (struct timespec): include <sys/time.h>
* include/ruby/missing.h (__syscall): moved to...
* io.c: here. because __syscall() is only used from io.c.
* include/ruby/missing.h: move "#include <sys/type.h>" to ....
* include/ruby/intern.h: here. because it was introduced for
fixing NFDBITS issue. [ruby-core:05179].
* thread.c: disabled _FORTIFY_SOURCE for avoid to hit glibc bug.
[Bug #8080] [ruby-core:53349]
* test/ruby/test_io.rb (TestIO#test_io_select_with_many_files):
test for the above.
* include/ruby/missing.h: removed __linux__. it's unnecessary.
git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/branches/ruby_1_9_3@39985 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
after using it. Patch by Stanislav Sedov. Thank you.
[Bug #7041] [ruby-core:47619]
git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/branches/ruby_1_9_3@37002 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
* thread_pthread.c (ruby_init_stack): use stack info if possible.
* thread_pthread.c (ruby_init_stack): adjust stack size for offset of
addr from the bottom.
* thread_pthread.c (get_stack): seems stack size does not include
guard size on Mac OS X.
* gc.h (IS_STACK_DIR_UPPER): utility macro.
* thread_pthread.c (get_stack): Linux is the only OS which includes
the size of guard page into the stack size.
* thread_pthread.c (ruby_init_stack): STACK_GROW_DIR_DETECTION is
necessary on platforms with unknown stack direction. [Bug #6761]
git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/branches/ruby_1_9_3@36789 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
* thread_pthread.c (ubf_select): call rb_thread_wakeup_timer_thread()
only when it is not timer_thread. [Bug #5757] [ruby-dev:44985]
patched by Tomoyuki Chikanaga.
* thread_pthread.c (ping_signal_thread_list): remove return value.
* thread_pthread.c (check_signal_thread_list): add a new function to
check if signal thread list is empty.
* thread_pthread.c (thread_timer): check signal thread list after
timer_thread_function(). main thread might be added into signal thread
list during timer_thread_function().
git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/branches/ruby_1_9_3@34425 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
* thread_pthread.c (ubf_select): activate timer thread when interrupt
blocking thread.
A patch created by Koichi Sasada. [ruby-core:39634] [Bug #5343]
to cover race condition, timer thread periodically send SIGVTARLM to
threads in signal thread list. so you should activate timer thread
when interrupt a thread.
git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/branches/ruby_1_9_3@33310 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
* thread_pthread.c (native_cond_signal): retry to call pthread_cond_signal
and pthread_cond_broadcast if they return EAGAIN in
native_cond_signal and native_cond_broadcast, respectively.
It is for the pthread implementation of Mac OS X 10.7 (Lion).
fixes#5155. [ruby-dev:44342].
* thread_pthread.c (native_cond_broadcast): ditto.
* thread_pthread.c (struct cached_thread_entry): stop using
pthread_cond_t and its functions directly.
* thread_pthread.c (register_cached_thread_and_wait): ditto.
* thread_pthread.c (use_cached_thread): ditto.
git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/branches/ruby_1_9_3@32853 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
100,000,000 because Solaris cond_timedwait() return EINVAL if an
argument is greater than current_time + 100,000,000. This is
considered as a kind of spurious wakeup. The caller to native_sleep
should care about spurious wakeup.
git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@32409 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
pthread_attr_getstack() if possible. and, remove an assumption
of stack growing direction.
git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@32403 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
must be the highest address of the current thread's stack.
git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@32395 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
RB_RESERVED_FD_P): added. This C API is to limit to access
fds which are used by RubyVM internal. In this version of
CRuby, return 1 if fd is communication pipe.
If your application needs to close all file descriptors to
preent resource leak, skip internal fds using this C API.
We also define a macro RB_RESERVED_FD_P(fd). So you can write
#ifndef RB_RESERVED_FD_P
#define RB_RESERVED_FD_P(fd) 0
#endif
for Ruby 1.9.2 or previous version to write compatible extensions.
See [ruby-core:37727]
* thread_win32.c (rb_reserved_fd_p): added (return 0 for any fds).
* io.c (rb_io_initialize): raise ArgumentError if given fd is reserved by Ruby.
git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@32394 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
return stack start addres, but stack base address. Thus,
we need to add stack size for getting stack start address.
And, we don't have to decrease guard size twice.
* thread_pthread.c (thread_start_func_1): don't use inaccurate
stack start guess if native_thread_init_stack() can be used.
[Bug #1813] [ruby-core:24540]
git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@32371 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
(we observed that select(2) was canceled by errno=514 on
boron == Linux/Xen environment)
git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@32325 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
machine stack for the timer thread at least 12KB. FreeBSD 8.2
AMD64 causes machine stack overflow (SIGSEGV) only with
PTHREAD_STACK_MIN (maybe defined as 2KB).
git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@32319 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
communication pipe to avoid timing bug (process termination timing).
The communication pipe will closed by OS.
git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@32276 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
In timer thread, signal handler shoul use it.
The patch is contributed by Eric Wong <normalperson@yhbt.net>.
Refs: [ruby-core:37644] and [ruby-core:37647]
* thread_pthread.c: use rb_async_bug_errno().
And replace all fprintf() to write().
* internal.h (rb_async_bug_errno): add decl. of above func.
git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@32274 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
Make "buff" as static. (Maybe) "buff" can be shared between
any caller (any threads) because no one use the read values.
And remove const value "buff_size", and define CCP_READ_BUFF_SIZE
macro.
git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@32266 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
no waiting thread. If there are 2 or more runnable threads,
the timer thread does polling. Avoid polling makes power save
for several computers (0.2W per a Ruby process, when I measured).
If outside-event such as signal or Thread#kill was occuerred
when the timer thread does not do polling, then wake-up
the timer thread using communication-pipe (the timer thread
waits this communication-pipe with select(2)).
The discussion about this modification can be found from the post
[ruby-core:33456] and other related posts.
Note that Eric Wong and KOSAKI Motohiro give us the huge
contributions for this modification. Thanks.
* thread_pthread.c (rb_thread_wakeup_timer_thread): add a function.
This function wakes up the timer thread using communication-pipe.
* thread.c (rb_thread_stop_timer_thread): add a parameter which
specify closing communication-pipe or not.
* thread.c (rb_thread_terminate_all): do not stop timer thread here
(ruby_cleanup() terminate timer thread).
* signal.c: wake up timer thread using
rb_thread_wakeup_timer_thread() from signal handler.
* eval.c (ruby_cleanup): use rb_thread_stop_timer_thread(1).
* process.c: use rb_thread_stop_timer_thread(0)
(reuse communication-pipe).
* thread_win32.c (rb_thread_wakeup_timer_thread): add a dummy
function.
* vm_core.h: add and fix decl. of functions.
git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@32244 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
* thread_pthread.c: remove HAVE_GVL_YIELD macro.
* thread_win32.c (gvl_yield): new. this fallback logic was moved from
rb_thread_schedule_rec().
git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@32069 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
When native_cond_timedwait() return 0 by spurious wakeup, we
don't have to neither 1) call timer_thread_function and 2)
exit the timer thread.
git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@31926 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
this is a regression since r31457. [Bug #4808] [ruby-dev:43606]
git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@31893 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
