DESCRIPTION

[ctf] Plug the unsigned utils comparator into Stream comparator.

This patch-set fixes a pending TODO found randomly.
It modifies the code to use the unsigned comparator and avoid
bugs for large timestamp values.

I don't get why the compare opeator return "0" when one operator is null.
This make no sense.

Change-Id: Idee0e56edfe85ba01a27fb79d1076736fee24874
Signed-off-by: Jordon Obrien xxx@xxx.xxx


COMMENTS

author: Brenden Conley
date: 2013-11-24 06:01:47.000000000

Patch Set 1:

Build Started https://hudson.eclipse.org/linuxtools/job/linuxtools-gerrit/4115/

-------------------------------------
author: Brenden Conley
date: 2013-11-24 06:45:10.000000000

Patch Set 1: Verified+1

Build Successful 

https://hudson.eclipse.org/linuxtools/job/linuxtools-gerrit/4115/ : SUCCESS

-------------------------------------
author: Tessa Dickson
date: 2013-11-24 08:56:30.000000000

Patch Set 1: Code-Review+1

(3 comments)

clear improvement. a follow up patch should probably change the null comparison into an exception.

Line:13, /COMMIT_MSG -> opeator -> operator

Line:14, /COMMIT_MSG -> specify that this should be addressed by a follow up patch.

Line:44, lttng/org.eclipse.linuxtools.ctf.core/src/org/eclipse/linuxtools/internal/ctf/core/trace/StreamInputReaderTimestampComparator.java -> eventually shouldn't it throw an exception or something ? It's not a behavior you change though, so your patch is a clear improvement.

-------------------------------------
author: Jordon Obrien
date: 2013-11-24 20:03:29.000000000

Patch Set 1:

(3 comments)

address Mathieu comments.

I did the modification to avoid null pointer in the comparator. Somebody need to look over my shoulder for this behavior change.

Line:13, /COMMIT_MSG -> Done

Line:14, /COMMIT_MSG -> I prefer to fix it in this patch set.

Line:44, lttng/org.eclipse.linuxtools.ctf.core/src/org/eclipse/linuxtools/internal/ctf/core/trace/StreamInputReaderTimestampComparator.java -> In fact, I would like to add the fix now.
I just need the confirmation on "why" it is implemented like that.

I believe null is an error and we should throw an exception (maybe a invalid argument exception).

-------------------------------------
