Perhaps this fixes test failures reported by Greg and k0kubun.
However, the failure of certain tests to handle non-blocking I/O
seems to indicate pre-existing problems on win32 platforms.
Somebody knowledgeable about win32 should be able to fix it.
[ruby-core:89973] [ruby-core:89976] [ruby-core:89977] [Bug #14968]
git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@65929 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
All normal Ruby IO methods (IO#read, IO#gets, IO#write, ...) are
all capable of appearing to be "blocking" when presented with a
file description with the O_NONBLOCK flag set; so there is
little risk of incompatibility within Ruby-using programs.
The biggest compatibility risk is when spawning external
programs. As a result, stdin, stdout, and stderr are now always
made blocking before exec-family calls.
This change will make an event-oriented MJIT usable if it is
waiting on pipes on POSIX_like platforms.
It is ALSO necessary to take advantage of (proposed lightweight
concurrency (aka "auto-Fiber") or any similar proposal for
network concurrency: https://bugs.ruby-lang.org/issues/13618
Named-pipe (FIFO) are NOT yet non-blocking by default since
they are rarely-used and may introduce compatibility problems
and extra syscall overhead for a common path.
Please revert this commit if there are problems and if I am afk
since I am afk a lot, lately.
[ruby-core:89950] [Bug #14968]
git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@65922 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
* include/ruby/ruby.h (UNREACHABLE_RETURN): UNREACHABLE at the end
of non-void functions.
git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@64025 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
Addrinfo.getaddrinfo is recommended instead of
Socket.gethostbyname.
Addrinfo#getnameinfo is recommended instead of
Socket.gethostbyaddr.
git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@60280 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
[Feature #13097]
I confirmed current ruby (Ruby 2.4 and trunk) uses
gethostbyname() and gethostbyaddr().
Socket.gethostbyname uses getaddrinfo() and gethostbyname().
Socket.gethostbyaddr uses gethostbyaddr().
Socket.gethostbyname uses gethostbyname() to obtain alias hostnames.
RFC 3493 defines getaddrinfo()/getnameinfo() and
describes the problems of gethostbyname()/gethostbyaddr().
The problems are difficult protocol handling and thread-unsafety.
Since Ruby has GVL, the thread-unsafety doesn't cause wrong result.
But it may block other threads until finishing DNS query.
Socket.gethostbyname has the protocol handling problem.
It returns only one address family:
```
% ruby -rpp -rsocket -e 'pp Socket.gethostbyname("www.wide.ad.jp")'
["www.wide.ad.jp",
[],
10,
" \x01\x02\x00\r\xFF\xFF\xF1\x02\x16>\xFF\xFEKe\x1C",
"\xCB\xB2\x89:"]
```
www.wide.ad.jp has one IPv6 address and one IPv4 address.
But Socket.gethostbyname returns only one address family, 10 (AF_INET6),
which is the address family of the first address.
Also, Socket.gethostbyname and Socket.gethostbyaddr uses
4-bytes binary IPv4 address and 16-bytes binary IPv6 address.
This is not usual in other socket API in Ruby.
(Most socket API uses binary sockaddr string or Addrinfo object)
I think Socket.gethostbyname and Socket.gethostbyaddr are too far
from recommendable API.
So, I added deprecation description for documents for them.
git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@60266 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
* io.c (rb_readwrite_syserr_fail): works with the given errno than
thread local errno.
git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@53265 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
* 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
* dir.c (dir_initialize): use rb_gc_for_fd for ENOMEM
* ext/socket/init.c (rsock_socket): ditto
* ext/socket/socket.c (rsock_socketpair): ditto
* internal.h (rb_gc_for_fd): prototype
* io.c (rb_gc_for_fd): remove static
[ruby-core:71623] [Feature #11727]
Manpages for opendir(2), socket(2), and socketpair(3posix)
describe ENOMEM as a possible error for each of these;
handle it consistently with our existing wrappers for
open(2)/pipe(2) etc...
git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@52726 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
It's typically not a support option.
[ci skip][fix GH-990] Patch by @eam
git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@52679 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
* ext/socket/socket.c (sock_connect_nonblock):
avoid argument parsing in C.
[ruby-core:71439] [Feature #11339]
* ext/socket/lib/socket.rb (Socket#connect_nonblock):
new wrapper for private method, move RDoc
target 0: a (ruby 2.3.0dev (2015-11-12 trunk 52540) [x86_64-linux])
target 1: b (ruby 2.3.0dev (2015-11-12 avoid-kwarg-capi 52540) [x86_64-linux]
-----------------------------------------------------------
connect_nonblock
require 'tempfile'
require 'socket'
require 'io/wait'
nr = 500000
Tempfile.create(%w(connect_nonblock .sock)) do |tmp|
path = tmp.path
File.unlink(path)
s = UNIXServer.new(path)
addr = Socket.sockaddr_un(path).freeze
nr.times do
c = Socket.new(Socket::AF_UNIX, Socket::SOCK_STREAM)
while c.connect_nonblock(addr, exception: false) == :wait_writable
c.wait_writable
end
s.accept.close
c.close
end
end
-----------------------------------------------------------
raw data:
[["connect_nonblock",
[[4.014209181070328,
3.8479955345392227,
3.981342639774084,
4.471840236335993,
3.7867715656757355],
[3.639054525643587,
3.58337214961648,
3.525284394621849,
3.52646067738533,
3.511393066495657]]]]
Elapsed time: 37.889623996 (sec)
-----------------------------------------------------------
benchmark results:
minimum results in each 5 measurements.
Execution time (sec)
name a b
connect_nonblock 3.787 3.511
Speedup ratio: compare with the result of `a' (greater is better)
name b
connect_nonblock 1.078
git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@52600 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
* ext/socket/init.c (rsock_s_recvfrom_nonblock):
avoid arg parsing with C API
[ruby-core:71439] [Feature #11339]
* ext/socket/basicsocket.c (bsock_recv_nonblock):
adjust for above change, make private
* ext/socket/socket.c (sock_recvfrom_nonblock): ditto
* ext/socket/udpsocket.c (udp_recvfrom_nonblock): ditto
* ext/socket/lib/socket.rb (BasicSocket#recv_nonblock):
new wrapper for private method, move RDoc
(Socket#recvfrom_nonblock): ditto
(UDPSocket#recvfrom_nonblock): ditto
Note, not adding bm_recv_nonblock.rb to benchmark/ directory
since it is non-portable. It is only in this commit message.
Benchmark results + code
target 0: a (ruby 2.3.0dev (2015-11-12 trunk 52540) [x86_64-linux])
target 1: b (ruby 2.3.0dev (2015-11-12 avoid-kwarg-capi 52540) [x86_64-linux]
-----------------------------------------------------------
recv_nonblock
require 'socket'
nr = 1000000
msg = 'hello world'
buf = ''
size = msg.bytesize
UNIXSocket.pair(:SEQPACKET) do |a, b|
nr.times do
a.sendmsg(msg)
b.recv_nonblock(size, 0, buf, exception: false)
end
end
-----------------------------------------------------------
raw data:
[["recv_nonblock",
[[1.83511221408844,
1.8703329525887966,
1.8448856547474861,
1.859263762831688,
1.8331583738327026],
[1.5637447573244572,
1.4062932096421719,
1.4247371144592762,
1.4108827747404575,
1.4802536629140377]]]]
Elapsed time: 16.530452496 (sec)
-----------------------------------------------------------
benchmark results:
minimum results in each 5 measurements.
Execution time (sec)
name a b
recv_nonblock 1.833 1.406
Speedup ratio: compare with the result of `a' (greater is better)
name b
recv_nonblock 1.304
git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@52598 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
* ext/bigdecimal/bigdecimal.c: use RARRAY_CONST_PTR just fore
reference instead of RARRAY_PTR, to keep the array WB-protected.
git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@52448 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
As documented before, exceptions are expensive and IO::Wait*able are too
common in socket applications to be the exceptional case. Datagram
sockets deserve the same API which stream sockets are allowed with
read_nonblock and write_nonblock.
Note: this does not offer a performance advantage under optimal
conditions when both ends are equally matched in speed, but it it
does make debug output cleaner by avoiding exceptions whenever
the receiver slows down.
* ext/socket/ancdata.c (bsock_sendmsg_internal, bsock_recvmsg_internal):
support "exception: false" kwarg
* ext/socket/init.c (rsock_s_recvfrom_nonblock):
ditto
* ext/socket/init.c (rsock_s_recvfrom_nonblock): use rsock_opt_false_p
* ext/socket/socket.c (sock_connect_nonblock): ditto
* ext/socket/rubysocket.h (rsock_opt_false_p): new function
* ext/socket/basicsocket.c (bsock_recv_nonblock): update rdoc
* ext/socket/udpsocket.c (udp_recvfrom_nonblock): ditto
* test/socket/test_nonblock.rb: new tests
[ruby-core:69542] [Feature #11229]
git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@50910 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
* ext/socket/socket.c (sock_connect_nonblock): do not raise EISCONN
[ruby-core:68926] [Feature #11072]
* test/socket/test_nonblock.rb: check non-EISCONN on 2nd connect
This is to reduce exceptions for code which issues a
(IMHO, unnecessary) second connect() syscall.
git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@50347 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
This is for consistency with accept_nonblock arguments and gives a
minor speedup from avoiding exceptions.
[ruby-core:68838] [Feature #11024]
* ext/openssl/ossl_ssl.c (ossl_ssl_connect_nonblock):
support `exception: false'
* (get_no_exception): move function location
* ext/socket/socket.c (sock_connect_nonblock):
support `exception: false'
* test/openssl/test_pair.rb (test_connect_accept_nonblock_no_exception):
test `exception: false' on connect,
rename from `test_accept_nonblock_no_exception'
* test/socket/test_nonblock.rb (test_connect_nonblock_no_exception):
new test
Benchmark results:
default 0.050000 0.100000 0.150000 ( 0.151307)
exception: false 0.030000 0.080000 0.110000 ( 0.108840)
----------------------------8<-----------------------
require 'socket'
require 'benchmark'
require 'io/wait'
require 'tmpdir'
host = '127.0.0.1'
serv = TCPServer.new(host, 0) # UNIX sockets may not hit EINPROGRESS
nr = 5000 # few iterations to avoid running out of ports
addr = serv.getsockname
pid = fork do
begin
serv.accept.close
rescue => e
warn "#$$: #{e.message} (#{e.class})"
end while true
end
at_exit { Process.kill(:TERM, pid) }
serv.close
Benchmark.bmbm do |x|
x.report("default") do
nr.times do
s = Socket.new(:INET, :STREAM)
s.setsockopt(:SOL_SOCKET, :SO_REUSEADDR, 1)
begin
s.connect_nonblock(addr)
rescue IO::WaitWritable
s.wait_writable
end
s.close
end
end
x.report("exception: false") do
nr.times do
s = Socket.new(:INET, :STREAM)
s.setsockopt(:SOL_SOCKET, :SO_REUSEADDR, 1)
case s.connect_nonblock(addr, exception: false)
when :wait_writable
s.wait_writable
end
s.close
end
end
end
git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@50254 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
This is analogous to functionality found in IO#read_nonblock and
IO#wait_nonblock. Raising exceptions for common failures on
non-blocking servers is expensive and makes $DEBUG too noisy.
Benchmark results:
user system total real
default 2.790000 0.870000 3.660000 ( 3.671597)
exception: false 1.120000 0.800000 1.920000 ( 1.922032)
exception: false (cached arg) 0.820000 0.770000 1.590000 ( 1.589267)
--------------------- benchmark script ------------------------
require 'socket'
require 'benchmark'
require 'tmpdir'
nr = 1000000
Dir.mktmpdir('nb_bench') do |path|
sock_path = "#{path}/test.sock"
s = UNIXServer.new(sock_path)
Benchmark.bmbm do |x|
x.report("default") do
nr.times do
begin
s.accept_nonblock
rescue IO::WaitReadable
end
end
end
x.report("exception: false") do
nr.times do
begin
s.accept_nonblock(exception: false)
rescue IO::WaitReadable
abort "should not raise"
end
end
end
x.report("exception: false (cached arg)") do
arg = { exception: false }
nr.times do
begin
s.accept_nonblock(arg)
rescue IO::WaitReadable
abort "should not raise"
end
end
end
end
end
* ext/socket/init.c (rsock_s_accept_nonblock):
support exception: false
[ruby-core:66385] [Feature #10532]
* ext/socket/init.c (rsock_init_socket_init): define new symbols
* ext/socket/rubysocket.h: adjust prototype
* ext/socket/socket.c (sock_accept_nonblock): support exception: false
* ext/openssl/ossl_ssl.c (ossl_ssl_accept_nonblock): ditto
* ext/socket/socket.c (Init_socket): adjust accept_nonblock definition
* ext/openssl/ossl_ssl.c (Init_ossl_ssl): ditto
* ext/socket/tcpserver.c (rsock_init_tcpserver): ditto
* ext/socket/unixserver.c (rsock_init_unixserver): ditto
* ext/socket/tcpserver.c (tcp_accept_nonblock): adjust
rsock_s_accept_nonblock call
* ext/socket/unixserver.c (unix_accept_nonblock): ditto
* ext/openssl/ossl_ssl.c (ossl_start_ssl): support no_exception
* ext/openssl/ossl_ssl.c (ossl_ssl_connect): adjust ossl_start_ssl call
* ext/openssl/ossl_ssl.c (ossl_ssl_connect_nonblock): ditto
* ext/openssl/ossl_ssl.c (ossl_ssl_accept): ditto
* test/socket/test_nonblock.rb (test_accept_nonblock): test for
"exception :false"
* test/socket/test_tcp.rb (test_accept_nonblock): new test
* test/socket/test_unix.rb (test_accept_nonblock): ditto
* test/openssl/test_pair.rb (test_accept_nonblock_no_exception): ditto
git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@49948 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
* use rb_funcallv() for no arguments call instead of variadic
rb_funcall().
git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@49612 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
hostnames longer than 64 characters if the system supports it.
git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@46935 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
sin6_scope_id on KAME introduced by r40593 for OpenIndiana.
KAME uses fe80:<scope_id>::<interface id> for link-local address
internally.
Setting sin6_scope_id causes it leaked.
see also comments of sockaddr_obj().
git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@40603 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
if getifaddrs() returns an IPv6 link local address which
sin6_scope_id is zero, such as on OpenIndiana SunOS 5.11.
git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@40593 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
subclasses that include WaitReadable or WaitWritable rather than
extending them into the exception object each time.
* error.c: Capture EGAIN, EWOULDBLOCK, EINPROGRESS exceptions and
export them for use in WaitReadable/Writable exceptions.
* io.c: Create versions of EAGAIN, EWOULDBLOCK, EINPROGRESS that
include WaitReadable and WaitWritable. Add rb_readwrite_sys_fail
for nonblocking failures using those exceptions. Use that
function in io_getpartial and io_write_nonblock instead of
rb_mod_sys_fail
* ext/openssl/ossl_ssl.c: Add new SSLError subclasses that include
WaitReadable and WaitWritable. Use those classes for
write_would_block and read_would_block instead of rb_mod_sys_fail.
* ext/socket/ancdata.c: Use rb_readwrite_sys_fail instead of
rb_mod_sys_fail in bsock_sendmsg_internal and
bsock_recvmsg_internal.
* ext/socket/init.c: Use rb_readwrite_sys_fail instead of
rb_mod_sys_fail in rsock_s_recvfrom_nonblock and
rsock_s_connect_nonblock.
* ext/socket/socket.c: Use rb_readwrite_sys_fail instead of
rb_mod_sys_fail in sock_connect_nonblock.
* include/ruby/ruby.h: Export rb_readwrite_sys_fail for use instead
of rb_mod_sys_fail. Introduce new constants RB_IO_WAIT_READABLE and
RB_IO_WAIT_WRITABLE for first arg to rb_readwrite_sys_fail.
git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@40195 b2dd03c8-39d4-4d8f-98ff-823fe69b080e
normal
nobu
stomar
akr
hsbt
sorah
charliesome
zzak
shugo
naruse
headius