Feature #11072: [PATCH] connect_nonblock(..., exception: false) does not raise EISCONN

• Status: Closed
• Assignee: nobu (Nobuyoshi Nakada)
• Target version: -

[normalperson (Eric Wong)]

This is to reduce exceptions for code which issues a (IMHO,
unnecessary) second connect() syscall.

I'll update ext/socket/lib/socket.rb to reduce exceptions on
:connect_timeout if this is accepted.

Personally, I never bother issuing a second connect() syscall
anymore and just catch the failures when calling
write/read/sendmsg/recvmsg/etc, but maybe we need to keep the
second connect() syscall in socket.rb for some odd systems...

[akr (Akira Tanaka)]

Eric Wong wrote:

This is to reduce exceptions for code which issues a (IMHO,
unnecessary) second connect() syscall.

I'll update ext/socket/lib/socket.rb to reduce exceptions on
:connect_timeout if this is accepted.

It is a good idea.

Personally, I never bother issuing a second connect() syscall
anymore and just catch the failures when calling
write/read/sendmsg/recvmsg/etc, but maybe we need to keep the
second connect() syscall in socket.rb for some odd systems...

I think it will affect applications.

[normalperson (Eric Wong)]

• Assignee changed from akr (Akira Tanaka) to nobu (Nobuyoshi Nakada)

nobu/matz: any comment on just this patch? thanks.

akr: thanks for the comment, I will leave the second connect_nonblock call in socket.rb

[nobu (Nobuyoshi Nakada)]

Seems fine to me.

• Status changed from Open to Closed
• % Done changed from 0 to 100

Applied in changeset r50347.

---

connect_nonblock(..., exception: false) does not raise EISCONN

• 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.