|Subject:||Re: fix for int-> long overflows|
|From:||Amy Kang (Amy....@Sun.COM)|
|Date:||Oct 14, 2009 12:30:11 pm|
Your issue tracker #31 has been accepted and I'll look into incorporating the fix as early as possible.
On 10/06/09 16:18, Dies Koper wrote:
Thanks Amy for the quick response!
Could you also evaluate the other patch I submitted? https://mq.dev.java.net/issues/show_bug.cgi?id=31
Amy Kang wrote:
Your issue #32 has been accepted and processed. Thanks. Others may anwser your remaining question.
I have filed an issue for this in your issue tracker.
You can always file an IssueTracker and enter the changes you'd like
to see in the product. We would then take that and factor it in, as we do with every defect or enhancement report
we receive. In general all issues submitted to mq.dev.java.net IssueTracker are assigned to issues@mq. They are tracked and are considered for each release cycle based on
internal priority. When and how are they tracked?
Of the 32 issues in your issue tracker, many have never been assigned or accepted. I felt mixed signals about whether you really accept and welcome feedback.
I understand the ability and extend of complying with user requests depend on internal priorities and release schedules. But if you could just accept the issues and write a short comment, maybe set a target milestone or anything to show when suggestions will be looked at, that would be highly appreciated.
About my contributions, as OpenMQ doesn't have a source repository system I can connect to I used a diff tool to create the "patch" files. I don't know if you can easily "apply" them to your work space. If you have any suggestions for creating patch files that are easier to review/apply (different tool, particular options, etc.?), please do let me know.
Thanks. I have previously signed the SCA.
In April I found and fixed an issue in GlassFish's JTA where the multiplication of two integers and a subsequent cast to a long led to IllegalArgumentException because the value had become negative.
For example: timer.schedule(this,timeout * 1000);
schedule() takes a long, but if 'timeout' is bigger than MAX_INT/1000, the value will become negative.
Searching through the V3 source code I found many similar cases in MQ code. Some of the 'timeout' variables might never reach a big enough value, or code might not be used any more, but as it's better safe than sorry I've addressed them in the attached diff (added 'L' to integer 1000).
Last month I tried contributing a patch using your issue tracker, but it's left unassigned. https://mq.dev.java.net/issues/show_bug.cgi?id=31
Please let me know if you are interested in contributions like this, and if so, how would you like me to make them.
I found these by grepping "1000;" and "1000);", so I might have missed some. Thanks, Dies Koper GF committer