Skip to main content

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index] [List Home]
Re: [cdt-dev] future of LRParser, XLC and UPC specific support in CDT

I think it would be good to remove these. If there is something left to be used by PTP, maybe it can be moved there. I don’t recall the outcome of the previous conversations about this but perhaps it was the same conclusion but no one had the time to look into moving things.


On Jan 8, 2020, at 1:13 PM, Jonah Graham <jonah@xxxxxxxxxxxxxxxx> wrote:

Hello folks,

This is a continuation of my ongoing review of what is in CDT. 

The LRParser, XLC and UPC components are currently part of our release and the SimRel, but it is not part of C/C++ IDE EPP project. UPC/LRParser is part of the Eclipse IDE for Scientific Computing and PTP has a project[0] that requires UPC  (which is why ptp-dev is included in this email).

The current state of these components is all similar. There has not been any substantive bugs fix in 7+ years in them, nor any bug reports. Virtually all the commits to these components have been cleanup and releng for years.
They each have a testsuite, but those suites have not been run as part of the build for a few years now. They can be run manually, but there are lots of failures. The failures are a combination of things, but include not handling new token types that have been added to CDT over the years[1] and that the tests cause assertions[2] in the core CDT code when run.

From this preliminary examination it looks improbable that anyone could be using this code. Does anyone know of any uses? 

Thank you,

[0] /tools/pldt/org.eclipse.ptp.pldt.upc
[1] tUSER_DEFINED_STRING_LITERAL added in Bug 379684 Support for User Defined Literals is not handled in case statements. See AST2CPPTests.testUserDefinedLiteralConcatenation2c
[2] CPPASTQualifiedName.setLastName(ICPPASTName) has an assertion that is hit when running the tests

Jonah Graham
Kichwa Coders
cdt-dev mailing list
To change your delivery options, retrieve your password, or unsubscribe from this list, visit

Back to the top