aboutsummaryrefslogtreecommitdiffstats
path: root/process.c
Commit message (Collapse)AuthorAgeFilesLines
* move fields to ec.ko12017-06-281-1/+1
| | | | | | | | | | * vm_core.h (rb_thread.h): move errinfo and trace_arg to rb_execution_context_t. * cont.c (fiber_switch, rb_cont_call): do not restore "trace_arg" here. git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@59199 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
* move several fields from rb_thread_t to rb_execution_context_t.ko12017-06-261-1/+1
| | | | | | | | * 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
* removed ruby_error_printnobu2017-06-161-2/+4
| | | | git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@59097 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
* Improve performance of implicit type conversionwatson19782017-05-311-2/+2
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | 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
* process.c: Use getrusage(2) in Process.timesk0kubun2017-05-281-1/+12
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | if getrusage(2) is available, to improve precision of Process.times and its user like lib/benchmark.rb. On macOS, since getrusage(2) has better precision than times(3), they are much improved like: * Before Process.times => #<struct Process::Tms utime=0.56, stime=0.35, cutime=0.04, cstime=0.03> puts Benchmark.measure { "a" * 1_000_000_000 } 0.340000 0.310000 0.650000 ( 0.674025) * After Process.times => #<struct Process::Tms utime=0.561899, stime=0.35076, cutime=0.046483, cstime=0.038929> puts Benchmark.measure { "a" * 1_000_000_000 } 0.343223 0.310037 0.653260 ( 0.674025) On Linux, since struct rusage from getrusage(2) is used instead of struct tms from times(2), they are slightly improved like: * Before Process.times => #<struct Process::Tms utime=0.43, stime=0.11, cutime=0.0, cstime=0.0> puts Benchmark.measure { "a" * 1_000_000_000 } 0.120000 0.040000 0.170000 ( 0.171621) * After Process.times => #<struct Process::Tms utime=0.432, stime=0.116, cutime=0.0, cstime=0.0> puts Benchmark.measure { "a" * 1_000_000_000 } 0.124000 0.048000 0.172000 ( 0.171621) [ruby-dev:49471] [Feature #11952] git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@58935 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
* process.c: fix error checknobu2017-05-111-20/+8
| | | | | | | * process.c (obj2uid, obj2gid): check the error number returned by `getpwnam_r()` and `getgrnam_r()`, instead of `errno`. git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@58662 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
* .nobu2017-05-101-2/+12
| | | | git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@58659 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
* process.c: temporary string for buffernobu2017-05-101-14/+12
| | | | | | | | * process.c (obj2uid, obj2gid): use temporary string as the buffer instead of `rb_alloc_tmp_buffer`, which is `NODE_ALLOCA` since r51492. [ruby-core:81084] [Bug #13554] git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@58658 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
* Add documents on Process::CLOCK_* constantssorah2017-03-291-0/+26
| | | | | | | | * process.c: Add documents on Process::CLOCK_* constants. Patch by Sunao Komuro <sunao-komuro@cookpad.com>. Closes [GH-1567] [Bug #13386] git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@58208 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
* use HAVE_BUILTIN___BUILTIN_MUL_OVERFLOWshyouhei2017-03-061-2/+3
| | | | | | | | | | | | | | | | | | | We already check for __builtin_mul_overflow in configure but never actually referred it before. Why not call it if available, because that should render supposedly-optimial assembly outputs. Optionally if __builtin_mul_overflow_p is available, which is the case for recent GCC, use that to detect fixnum overflow. This is much faster than the previous. On my machine generated assembly of numeric.c:int_pow reduces from 480 to 448 bytes, according to nm(1). Also on my machine, following script boosts from 7.819 to 6.929 sec. time ./miniruby -e 'i=0; while i < 30_000_000 do i += 1; 7 ** 23; end' Signed-off-by: Urabe, Shyouhei <shyouhei@ruby-lang.org> git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@57784 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
* process.c: PATH env in spawnnobu2016-11-061-8/+24
| | | | | | | * process.c (rb_exec_fillarg): honor the given path environment variable. [ruby-core:53103] [Bug #8004] git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@56618 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
* [DOC] replace Fixnum with Integer [ci skip]nobu2016-10-261-1/+1
| | | | | | * numeric.c: [DOC] update document for Integer class. git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@56492 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
* process.c: use uaspawnnobu2016-10-141-0/+6
| | | | | | | * process.c (proc_exec_cmd): use UTF-8 version aspawn. [ruby-dev:49838] [Bug #12841] git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@56424 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
* fid typos [ci skip]nobu2016-09-241-1/+1
| | | | | | * fix typos, "a" before "Integer" to "an". [Fix GH-1438] git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@56225 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
* * process.c (InitVM_process): Support CLOCK_MONOTONIC_RAW_APPROX,naruse2016-09-211-5/+17
| | | | | | | CLOCK_UPTIME_RAW, and CLOCK_UPTIME_RAW_APPROX which are introduced by macOS 10.12. git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@56200 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
* replace fixnum by integer in documents.akr2016-09-081-45/+45
| | | | git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@56102 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
* * vm_core.h: revisit the structure of frame, block and env.ko12016-07-281-1/+0
| | | | | | | | | | | | | | | | | | | | | | [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
* fix typonaruse2016-07-041-1/+1
| | | | git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@55572 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
* * process.c: define sig_t if not exist.naruse2016-07-041-0/+3
| | | | | | at least Solaris 10 and 11 doesn't have sig_t. git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@55571 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
* process.c (disable_child_handler_fork_child): simplifynormal2016-07-021-52/+9
| | | | | | | | | | | | | | | | | signal(2) is portable for SIG_DFL and SIG_IGN, so we do not need the extra code for sigaction(2). Also, execve will reset all signal handlers to default anyways, so there is little sense in preserving old signal handler besides SIG_IGN. Hopefully this makes the code easier-to-understand and maintain. * process.c (disable_child_handler_fork_child): simplify [ruby-core:75781] [Misc #12439] git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@55566 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
* process.c: [DOC] Process.abort [ci skip]nobu2016-05-271-1/+1
| | | | | | | * process.c (rb_f_abort): [DOC] Process.abort is a singleton method of Process, but not an instance method. git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@55183 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
* fix build on no-fork-spawnv platformsnobu2016-05-201-6/+27
| | | | | | | | * process.c (rb_execarg_commandline): build command line string from argument vector in rb_execarg. [ruby-core:75611] [Bug #12398] git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@55091 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
* * configure.in: check function attirbute const and pure,naruse2016-05-081-0/+1
| | | | | | | | | | | | and define CONSTFUNC and PUREFUNC if available. Note that I don't add those options as default because it still shows many false-positive (it seems not to consider longjmp). * vm_eval.c (stack_check): get rb_thread_t* as an argument to avoid duplicate call of GET_THREAD(). git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@54952 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
* * configure.in: add -Wsuggest-attribute=noreturn and suppress warnings.naruse2016-05-081-1/+1
| | | | git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@54943 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
* process.c: argument types over conversionnobu2016-05-061-2/+14
| | | | | | | | * process.c (rb_exec_getargs): honor the expected argument types over the conversion method. the basic language functionality should be robust. [ruby-core:75388] [Bug #12355] git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@54934 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
* process.c (disable_child_handler_fork_child): initialize handler for SIGPIPEnormal2016-05-021-1/+5
| | | | | | | | | | | | | | | | | | This only happens on the !POSIX_SIGNAL case; but one may test with the following: --- a/process.c +++ b/process.c @@ -3479,6 +3479,7 @@ disable_child_handler_fork_parent(struct child_handler_disabler_state *old) #endif } +#undef POSIX_SIGNAL /* This function should be async-signal-safe. Actually it is. */ static int disable_child_handler_fork_child(struct child_handler_disabler_state *old, char *errmsg, size_t errmsg_buflen) git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@54890 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
* process.c: prefer rb_check_aritynobu2016-04-141-28/+26
| | | | | | | * process.c: use rb_check_arity instead of rb_scan_args for simple optional arguments. git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@54573 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
* * process.c (rb_execarg_parent_start1): need to convert the encoding tousa2015-12-271-0/+1
| | | | | | | ospath's one. git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@53348 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
* * process.c (rb_execarg_addopt): need to convert to ospath.usa2015-12-271-0/+1
| | | | git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@53347 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
* * process.c: use rb_w32_uchdir() instead of plain chdir() on Windows.usa2015-12-271-0/+5
| | | | | | | reported by naruse via twitter. git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@53346 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
* prefer rb_syserr_failnobu2015-12-231-42/+27
| | | | | | | | * file.c, io.c, util.c: prefer rb_syserr_fail with saved errno over setting errno then call rb_sys_fail, not to be clobbered potentially and to reduce thread local errno accesses. git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@53264 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
* * enum.c: fix a typo in documentation.hsbt2015-12-141-1/+1
| | | | | | | | | | | | [ci skip][fix GH-1140] Patch by @jutaz * io.c: ditto. * iseq.c: ditto. * numeric.c: ditto. * process.c: ditto. * string.c: ditto. * vm_trace.c: ditto. git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@53105 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
* process.c: fix rdoc of Process.getsid [ci skip]nobu2015-11-041-1/+1
| | | | | | | * process.c (proc_getsid): [DOC] Fix double word 'for' and typo. [Fix GH-1080] git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@52441 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
* * ChangeLog: Good-bye OS/2.kosaki2015-10-181-42/+1
| | | | | | | | | | | | | | | | | | | | * common.mk: ditto. * configure.in: ditto. * dln_find.c: ditto. * ext/Setup.emx: ditto. * ext/extmk.rb: ditto. * ext/socket/extconf.rb: ditto. * ext/zlib/extconf.rb: ditto. * file.c: ditto. * include/ruby/defines.h: ditto. * io.c: ditto. * lib/mkmf.rb: ditto. * missing/os2.c: ditto. * process.c: ditto. * ruby.c: ditto. * NEWS: announce OS/2 is no longer supported. git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@52161 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
* process.c: do not inherit saved fdsnobu2015-09-121-4/+75
| | | | | | | | * process.c (save_redirect_fd): make saved FDs close-on-exec not to be inherited. * process.c (run_exec_dup2): restore close-on-exec flags too. git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@51832 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
* Clarify spawn pgroup documentation [ci skip]nobu2015-09-101-5/+5
| | | | | | | | | | * process.c (rb_f_spawn): Be more specific regarding "other values" by having "non-zero positive integers" Add nil, the default value, as a possible value and what it means. Try to use more consistent language. [Fix GH-1008] git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@51819 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
* process.c: keep GVL if WNOHANGnobu2015-09-081-11/+19
| | | | | | | * process.c (rb_waitpid): do not release GVL when WNOHANG is set, it should return immediately without blocking. git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@51793 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
* process.c: retry loopnobu2015-09-081-14/+17
| | | | | | * process.c (rb_waitpid): refactor retry loop by interrupt. git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@51792 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
* process.c: open exception messagenobu2015-09-071-1/+1
| | | | | | | * process.c (rb_execarg_parent_start1): raise with the target path name when open() failed. git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@51787 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
* process.c: chdir exception messagenobu2015-09-071-9/+31
| | | | | | | * process.c (rb_exec_fail): raise with the target directory name when chdir() failed. pointed out by sorah. git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@51786 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
* Fix typo in Process.detatch doc [ci skip]nobu2015-09-021-1/+1
| | | | | | | * process.c (proc_detach): [DOC] fix typo "intent" as "intend" in rdoc. [Fix GH-1011] git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@51739 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
* process.c: [DOC] env values in spawn [ci skip]nobu2015-08-191-1/+1
| | | | | | | * process.c (rb_f_spawn): [DOC] elaborate environment variable values. [ruby-core:70456] [Bug #11463] git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@51641 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
* improve handling of timer thread shutdownnormal2015-08-141-37/+11
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | Shutting down the timer thread now always closes pipes to free FDs. In fact, we close the write ends of the pipes is done in the main RubyVM to signal the timer thread shutdown. To effectively close pipes, we implement userspace locks via atomics to force the pipe closing thread to wait on any signal handlers which may be waking up. While we're at it, improve robustness during resource exhaustion and allow it to limp along non-fatally if restarting a timer thread fails. This reverts r51268 Note: this change is tested with VM_CHECK_MODE 1 in vm_core.h * process.c (close_unless_reserved): add extra check (dup2_with_divert): remove (redirect_dup2): use dup2 without divert (before_exec_non_async_signal_safe): adjust call + comment (rb_f_exec): stop timer thread for all OSes (rb_exec_without_timer_thread): remove * eval.c (ruby_cleanup): adjust call * thread.c (rb_thread_stop_timer_thread): always close pipes * thread_pthread.c (struct timer_thread_pipe): add writing field, mark owner_process volatile for signal handlers (rb_thread_wakeup_timer_thread_fd): check valid FD (rb_thread_wakeup_timer_thread): set writing flag to prevent close (rb_thread_wakeup_timer_thread_low): ditto (CLOSE_INVALIDATE): new macro (close_invalidate): new function (close_communication_pipe): removed (setup_communication_pipe_internal): make errors non-fatal (setup_communication_pipe): ditto (thread_timer): close reading ends inside timer thread (rb_thread_create_timer_thread): make errors non-fatal (native_stop_timer_thread): close write ends only, always, wait for signal handlers to finish (rb_divert_reserved_fd): remove * thread_win32.c (native_stop_timer_thread): adjust (untested) (rb_divert_reserved_fd): remove * vm_core.h: adjust prototype git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@51576 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
* revert r31760 and r31761nobu2015-07-261-23/+0
| | | | | | | seems that rb_bug_errno() is called in sigpipe() intentionally. https://gist.github.com/sorah/831169 git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@51390 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
* avoid redundant GET_THREAD() callsnormal2015-07-161-1/+2
| | | | | | | | | | | | | | | | | | This reduces binary size slightly on my 32-bit system: text data bss dec hex filename 2847705 12360 30632 2890697 2c1bc9 ruby.orig 2847641 12360 30632 2890633 2c1b89 ruby * iseq.c (rb_iseq_compile_with_option): reuse result of previous GET_THREAD() call * thread.c (thread_create_core): ditto (rb_mutex_trylock): ditto (rb_mutex_lock): ditto * process.c (rb_waitpid): avoid multiple eval from RUBY_VM_CHECK_INTS * thread.c (rb_thread_check_ints): ditto git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@51273 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
* * process.c (redirect_dup2): when the new FD of dup2() coflictsngoto2015-07-161-2/+12
| | | | | | | | | | | | | | | | | | | | | | with one of the timer thread FDs, the internal FD is diverted. [Bug #11336] [ruby-core:69886] [Bug #11350] [ruby-core:69961] * process.c (dup2_with_divert): new function for the above purpose. * thread_pthread.c (rb_divert_reserved_fd): new function for diverting reserved FD. If the given FD is the same as one of the reserved FDs, the reserved FD number is internally changed. It returns -1 when error. Otherwise, returns 0. It also returns 0 if there is no need to change reserved FD number. * thread_win32.c (rb_divert_reserved_fd): always returns 0 because of no reserved FDs. * internal.h (rb_divert_reserved_fd): prototype declaration. It is Ruby internal use only. git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@51268 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
* Revert r51209 because signals after stopping timer thread beforengoto2015-07-161-0/+10
| | | | | | | entering exec(2) system call may be lost. git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@51265 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
* process.c (close_unless_reserved): declare type of `fd' argnormal2015-07-161-1/+2
| | | | git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@51262 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
* * process.c (redirect_close, parent_redirect_close): should not closengoto2015-07-151-4/+14
| | | | | | | | | | | | | | | reserved FD. It should be closed in the exec system call due to the O_CLOEXEC or FD_CLOEXEC flag. [Bug #11353] [ruby-core:69977] * process.c (close_unless_reserved): new function to close FD unless it is reserved for internal communication. * thread_pthread.c (rb_reserved_fd_p): should check owner_process pid to avoid false positive in forked child process. git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@51255 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
* * process.c (rb_f_exec): rb_exec_without_timer_thread should bengoto2015-07-101-10/+0
| | | | | | | | | | used on every OS, not only on Mac OS X or Haiku, to prevent timer thread communications using file descriptor 3 after setting redirection of the fd 3 before calling exec. [Bug #11336] [ruby-core:69886] git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@51209 b2dd03c8-39d4-4d8f-98ff-823fe69b080e