ruby-core

Mailing list archive

[#35446] [Ruby 1.9 - Bug #4477][Open] Kernel:exec and backtick (`) don't work for certain system commands — Joachim Wuttke <j.wuttke@...>

10 messages 2011/03/07

[#35476] [Ruby 1.9 - Bug #4489][Open] [PATCH] Encodings with /-(unix|dos|mac)\Z/ — "James M. Lawrence" <quixoticsycophant@...>

20 messages 2011/03/10

[#35552] [Ruby 1.9 - Feature #4523][Open] Kernel#require to return the path of the loaded file — Alex Young <alex@...>

14 messages 2011/03/24

[#35565] [Ruby 1.9 - Feature #4531][Open] [PATCH 0/7] use poll() instead of select() in certain cases — Eric Wong <normalperson@...>

33 messages 2011/03/28

[#35566] [Ruby 1.9 - Feature #4532][Open] [PATCH] add IO#pread and IO#pwrite methods — Eric Wong <normalperson@...>

12 messages 2011/03/28

[#35586] [Ruby 1.9 - Feature #4538][Open] [PATCH (cleanup)] avoid unnecessary select() calls before doing I/O — Eric Wong <normalperson@...>

9 messages 2011/03/29

[ruby-core:35425] Re: [Ruby 1.9 - Bug #4463][Open] [PATCH] release GVL for fcntl() for operations that may block

From: Eric Wong <normalperson@...>
Date: 2011-03-04 17:16:59 UTC
List: ruby-core #35425
KOSAKI Motohiro <kosaki.motohiro@gmail.com> wrote:
> Hi
> 
> I've commited slightly modified version today (r31025).
> The difference is,
> 
> 1) All IO.fcntl() and IO.iocntl() relese GVL instead only SETLCKW. because,
>     A) if a user are using network filesystem, almost all fcntl need network
>         communication. iow, they can be blocked.
>     B) We are sure ioctl() has similar issue. But, we don't have any knowledge
>         which ioctl can be blocked. It is strongly dependend a
> platform and a device.

Agreed on both points.

> 2) Added small test. It is based on your Fcntl::Flock patch.

Any chance of that patch making it into trunk?  I'd be happy to make
any changes/improvements necessary (+docs, too).  Thanks again.

-- 
Eric Wong

In This Thread