[#61171] Re: [ruby-changes:33145] normal:r45224 (trunk): gc.c: fix build for testing w/o RGenGC — SASADA Koichi <ko1@...>
(2014/03/01 16:15), normal wrote:
[#61243] [ruby-trunk - Feature #9425] [PATCH] st: use power-of-two sizes to avoid slow modulo ops — normalperson@...
Issue #9425 has been updated by Eric Wong.
[#61359] [ruby-trunk - Bug #9609] [Open] [PATCH] vm_eval.c: fix misplaced RB_GC_GUARDs — normalperson@...
Issue #9609 has been reported by Eric Wong.
(2014/03/07 19:09), normalperson@yhbt.net wrote:
SASADA Koichi <ko1@atdot.net> wrote:
[#61424] [REJECT?] xmalloc/xfree: reduce atomic ops w/ thread-locals — Eric Wong <normalperson@...>
I'm unsure about this. I _hate_ the extra branches this adds;
Hi Eric,
SASADA Koichi <ko1@atdot.net> wrote:
(2014/03/14 2:12), Eric Wong wrote:
SASADA Koichi <ko1@atdot.net> wrote:
[#61452] [ruby-trunk - Feature #9632] [Open] [PATCH 0/2] speedup IO#close with linked-list from ccan — normalperson@...
Issue #9632 has been reported by Eric Wong.
[#61496] [ruby-trunk - Feature #9638] [Open] [PATCH] limit IDs to 32-bits on 64-bit systems — normalperson@...
Issue #9638 has been reported by Eric Wong.
[#61568] hash function for global method cache — Eric Wong <normalperson@...>
I came upon this because I noticed existing st numtable worked poorly
(2014/03/18 8:03), Eric Wong wrote:
SASADA Koichi <ko1@atdot.net> wrote:
what's the profit from using binary tree in place of hash?
Юрий Соколов <funny.falcon@gmail.com> wrote:
[#61687] [ruby-trunk - Bug #9606] Ocassional SIGSEGV inTestException#test_machine_stackoverflow on OpenBSD — normalperson@...
Issue #9606 has been updated by Eric Wong.
[#61760] [ruby-trunk - Feature #9632] [PATCH 0/2] speedup IO#close with linked-list from ccan — normalperson@...
Issue #9632 has been updated by Eric Wong.
[ruby-core:61518] [ruby-trunk - Bug #9544] Ruby resolver not using autoport
Issue #9544 has been updated by Jakub Szafranski.
Just a little follow-up:
I still think that OS should take care of that, this is a wrong layer for me.
For instance: if a system is trying to choose a free port - either random or sequential - it will randomize from a pool of *free* ports. You do realize that at high traffic, the probability of ruby hitting into a free port is significantly lower? Also, if port_min/port_max has been altered (which isn't really any magic), that chance gets really dropped. Aren't you bothered by performance issues that your workaround introduced?
----------------------------------------
Bug #9544: Ruby resolver not using autoport
https://bugs.ruby-lang.org/issues/9544#change-45807
* Author: Jakub Szafranski
* Status: Closed
* Priority: Normal
* Assignee:
* Category: core
* Target version: current: 2.2.0
* ruby -v: ruby 2.1.0p0 (2013-12-25 revision 44422) [x86_64-freebsd9.1]
* Backport: 1.9.3: UNKNOWN, 2.0.0: UNKNOWN, 2.1: UNKNOWN
----------------------------------------
### Problem
On one of my production servers I've noticed that customers were failing to install anything using gem and the latest ruby. After a bit of debugging we've found out, that it's related to ruby resolve module:
<pre>
> p Resolv.getaddress "google.com"
Errno::EPERM: Operation not permitted - bind(2) for "0.0.0.0" port 62374
from /home/pudlobe/.rvm/rubies/ruby-2.1.0/lib/ruby/2.1.0/resolv.rb:654:in `bind'
from /home/pudlobe/.rvm/rubies/ruby-2.1.0/lib/ruby/2.1.0/resolv.rb:654:in `bind_random_port'
from /home/pudlobe/.rvm/rubies/ruby-2.1.0/lib/ruby/2.1.0/resolv.rb:747:in `block in initialize'
from /home/pudlobe/.rvm/rubies/ruby-2.1.0/lib/ruby/2.1.0/resolv.rb:735:in `each'
...
</pre>
The interesting part is _bind_random_port_ function. What for? The standard way of binding to a random port for udp connection is to use port 0. And on that particular machine it fails because it's using a mac_portacl module to filter which user can bind to what ports. **However, port 0 is excepted from this rule, because it's the AUTOPORT** - practically every system that allows such port filtering also allows to set an exception for the autoport.
### Docs
<pre>
Purpose:
Port 0 is officially a reserved port in TCP/IP networking, meaning that it should not be used for any TCP or UDP network communications. However, port 0 sometimes takes on a special meaning in network programming, particularly Unix socket programming. In that environment, port 0 is a programming technique for specifying system-allocated (dynamic) ports.
Description:
Configuring a new socket connection requires assigning a TCP or UDP port number. Instead of hard-coding a particular port number, or writing code that searches for an available port on the local system, network programmers can instead specify port 0 as a connection parameter. That triggers the operating system to automatically search for and return the next available port in the dynamic port number range.</pre>
### Impact
This bug affects every system that has a restricted port-binding policy, making ruby unavailable for security-freak admins ;)
### Suggested fix:
Either use port 0 to bind to the port, or at least make an option for the system admin/end user to specify the port by himself.
--
http://bugs.ruby-lang.org/