Anyone managed to try this? Should just be a case of opening the proj and
running the test. Would be good to know I'm not going crazy.

Thanks :)

James

On 4 May 2010 21:24, James Crowley <[email protected]> wrote:

> Hi guys,
>
> I've posted a ZIP with a unit test up here:
> http://216.121.112.228/browse/NH-2192<http://216.121.112.228/browse/NH-2192?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel>
>
> Please let me know if you can reproduce or not!!
>
> Thank you
>
> James
>
> On 23 April 2010 16:41, Ayende Rahien <[email protected]> wrote:
>
>> There _was_ a problem there, IIRC.
>> I am pretty sure that we fix that.
>> Are you using the latest? Can you provide a failing test?
>>
>>
>> On Fri, Apr 23, 2010 at 3:09 PM, James Crowley 
>> <[email protected]>wrote:
>>
>>> Hi guys,
>>>
>>> Are you aware of a threading issue
>>> within QueryParameters.PrepareParameterTypes? I've got two threads
>>> (operating over different sessions) that, when run starting from the same
>>> execution point at the start of the PrepareParameterTypes method. All works
>>> fine when the other thread is frozen, but when both are run at the same
>>> time, there are issues with the parameter locations.
>>>
>>> It's something to do with here:
>>>
>>> while ((location < sqlParameters.Count) &&
>>> (sqlParameters[location].ParameterPosition != null))
>>> location++;
>>>
>>> in that this can actually result in location being equal to
>>> sqlParameters.Count (and later having an IndexOutOfRange exception being
>>> thrown). However, it's only happening because ParameterPosition is
>>> unexpectedly set. Still trying to figure out how, as all I can see are local
>>> variables right now.
>>>
>>> I'm still trying to pin this down exactly, but wanted to check if you
>>> were already aware of not?
>>>
>>> Thanks,
>>>
>>> James
>>>
>>>

Reply via email to