Re: [cdt-dev] LRParser test failures

[Chris Recoskie <recoskie@xxxxxxxxxx>]

John will have a look. Thanks for letting us know.

The fact that the LR parser tests inherit from the GNU parser tests is both a blessing and a curse. When new tests are added for cases that the existing parser logic ought to cover, we get those tests automatically. Unfortunately, when new features are added such as C++0x support, we get those tests automatically too, and they fail because we don't have the new features in the LR parser.

This is compounded by the fact that hudson only emails build results to people who recently made checkins. Since the changes that started the failures were made elsewhere, none of us got an email. So, it's not a matter of the parsers not getting any love... it's that we just didn't know there was a problem.

===========================
Chris Recoskie
Team Lead, IBM CDT and RDT
IBM Toronto

James Blackburn ---01/22/2011 06:50:12 AM---Hi All, We've had LRParser tests failing for more than 100 builds:

From:	James Blackburn <jamesblackburn@xxxxxxxxx>
To:	"CDT General developers list." <cdt-dev@xxxxxxxxxxx>
Date:	01/22/2011 06:50 AM
Subject:	[cdt-dev] LRParser test failures
Sent by:	cdt-dev-bounces@xxxxxxxxxxx

---

Hi All,

We've had LRParser tests failing for more than 100 builds:
https://hudson.eclipse.org/hudson/job/cdt-nightly/lastCompletedBuild/testReport/

There are 59 failures, and the volume drowns out any real sporadic
failures that might be occurring.  It looks like the lrparser isn't
get much love at the moment, is anyone planning on fixing the tests
before the summer?  If not we should disable the tests...

Cheers,
James
_______________________________________________
cdt-dev mailing list
cdt-dev@xxxxxxxxxxx
https://dev.eclipse.org/mailman/listinfo/cdt-dev