Review request for OPENJDK6-38: java.net.Socket getTcpNoDelay/setTcpNoDelay work incorrectly on Windows

[Vladislav Karnaukhov]

Hello Omair,

thank you for review. I agree with your proposal to backport 8010213. Please note that the bug record for 8010213 isn't accessible outside Oracle, I suppose due to security restrictions. I won't be able to create backport record for 8010213 and update main record.

Could you please point me at backport process in OpenJDK6? Are there any docs?

Regards,
- Vlad

-----Original Message-----
From: Omair Majid [mailto:omajid at redhat.com] 
Sent: Friday, July 18, 2014 19:34
To: Vladislav Karnaukhov
Cc: jdk6-dev at openjdk.java.net
Subject: Re: Review request for OPENJDK6-38: java.net.Socket getTcpNoDelay/setTcpNoDelay work incorrectly on Windows

Hi,

* Vladislav Karnaukhov <vladislav at azulsystems.com> [2014-07-18 04:09]:
> please review the fix for OPENJDK6-38: java.net.Socket 
> getTcpNoDelay/setTcpNoDelay work incorrectly on Windows
> 
> webrev: http://cr.openjdk.java.net/~vkarnauk/OPENJDK6-38/webrev.00/
>
> I fixed the bug independently from the fix 8010213, though the fixes 
> are identical. I don't mind at all if my fix will be marked as a 
> backport of 8010213 for consistency.

I am not familiar with windows networking, but since this is identical to the JDK 7 fix I have no objections.

I would strongly prefer however, that 8010213 was backported directly (through hg export/hg import), rather than through an OpenJDK6-specific bug. This keeps track of the bug id (which makes tracking of backported fixes easier in the future) and also keeps the code as close as possible (which makes applying future fixes easier).

Thanks,
Omair

--
PGP Key: 66484681 (http://pgp.mit.edu/)
Fingerprint = F072 555B 0A17 3957 4E95  0056 F286 F14F 6648 4681