Hi all,
please review the following patch to API in
jdk.incubator.httpclient. The patch is more like a clean-up:
* With current RequestProcessor implementation, constructing multiple
HTTP Exchange may be blocked if the first request doesn't response
by server side. This is introduced
Thanks to Christoph, Vyom and Mark for the reviews.
I've improved the testcase as per feedback. Hope this meets all requests :
http://cr.openjdk.java.net/~coffeys/webrev.8182672.jdk10.v2/webrev/
Regards,
Sean.
On 22/06/17 22:36, Langer, Christoph wrote:
Hi Sean,
I played with it a bit more a
Hi Seán,
this looks very good.
I've got a few minor points:
1. In line 65 where you catch a possible unexpected exception, you could also
mention the thread name in the output.
2. I would check for failure condition at the end of each test loop(line 91,
for (NetworkInterface ni : toTest)... )
Hi Sean,
latest test code looks good, one minor bit, in MT program
"ex.printStackTrace();" will not help as output will be screwed. I will
suggested to pass the original exception to line 106.
throw new RuntimeException(originalException);
Thanks,
Vyom
On Friday 23 June 2017 03:26 PM, Seá
Hi Felix,
Looks good to me.
+1 for removing unnecessary calls that need synchronization and
cause deadlocks.
If you still have the full thread-dump where the issue could
be observed it might be good to record it in JDK-8178699
(if you no longer have it then don't bother).
best regards,
-- dan
Hi Daniel，
thread dump can be found in JBS attachment .
Thanks
-Felix
Sent from my iphone
> 在 2017年6月23日，18:40，Daniel Fuchs 写道：
>
> Hi Felix,
>
> Looks good to me.
>
> +1 for removing unnecessary calls that need synchronization and
> cause deadlocks.
>
> If you still have the full thr
On 23/06/17 10:56, Seán Coffey wrote:
Thanks to Christoph, Vyom and Mark for the reviews.
I've improved the testcase as per feedback. Hope this meets all requests :
http://cr.openjdk.java.net/~coffeys/webrev.8182672.jdk10.v2/webrev/
The change looks good.
Sean and I did a live coding sessi
Hi,
looks like a great piece of modern concurrent Java coding :) Well done! +1
Best regards
Christoph
> -Original Message-
> From: Chris Hegarty [mailto:chris.hega...@oracle.com]
> Sent: Freitag, 23. Juni 2017 15:28
> To: Seán Coffey ; Langer, Christoph
>
> Cc: net-dev
> Subject: Re: R
Thanks all. There were a few shouts from Chris in the office when he saw
my Enumeration code ;)
I'm running this through our test system again and will push if I get
green results.
Regards,
Sean.
On 23/06/17 14:40, Langer, Christoph wrote:
Hi,
looks like a great piece of modern concurrent
Hi Chris,
test looks good, one minor comment can be ignored, in failing case it
print stack and "Exception in thread "main" java.lang.RuntimeException:
Failed" both. I can see main is declared to throw Exception. Is is
possible to throw exception instead of printing trace or pass this
except
a minor observation:
perhaps a slight modification to the test to allow adaptation to jdk8
(the genesis of the reported problem)
replacing the NetworkInterface.networkInterfaces() with static method
which encapsulates the Stream
creation
public static Stream
getNetworkInterfacesAsStream(
Nice suggestion Mark - I've taken that on board with latest webrev.
http://cr.openjdk.java.net/~coffeys/webrev.8182672.jdk10.v3/webrev/
There was an issue with the hasHardwareAddress method also. A 'return
false' was missing in the exception block.
Vyom - I discussed your suggestion with Chri
12 matches
Mail list logo
