Skip to main content

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index] [List Home]
Re: [cdt-dev] Bug 327001

My interpretation would be:
We are in a case where fdata member is under initialization (cf CConfigurationDescriptionCache.loadData(CConfigurationDescriptionCache.java:95) in the stack trace)
For any reason, we need to check if current instance provides a Build Variable Contributor (cf CConfigurationDescriptionCache.getBuildVariablesContributor(CConfigurationDescriptionCache.java:450))
As fdata is not yet initialized at this time, our patch proposal is to temporarily consider that there is no Build Variable Contributor (until fdata will be non-null)
Since build variables seems not to be cached anywhere, the next time the variable is evaluated, fdata member is non-null and normal behavior is resumed.

Anyway, if you need additional JUnit test case to validate the correction, can you tell me how can I propose a new test for integration in your test campaign?

Thanks & Regards
David

> The concern is that we don't understand the ramifications of the
> change, and it looks to be a workaround for a real issue in the model.
> Ideally we need to commit a Junit test as well, showing the problem and the fix.


Back to the top