[cfe-dev] bugzilla change

Daniel Dunbar daniel at zuster.org
Mon Mar 8 14:13:29 PST 2010


On Mon, Mar 8, 2010 at 1:26 PM, Chris Lattner <clattner at apple.com> wrote:
>
> On Mar 8, 2010, at 8:56 AM, Daniel Dunbar wrote:
>
>> Nice!
>>
>> Can we get rid of the unnecessarily refined categories? From my
>> perspective, we rarely use the large list of components and products
>> that we have, and they just complicate things for users.
>>
>> For Clang, for example, I think we could get away with
>> {Sema,C++,Codegen,[C++0x],Documentation} without any loss of
>> functionality.
>
> Yes, I plan to clean up the llvm components as well, adding a top level code generation component etc.
>
> If you tell me how you want the clang components, I'm happy to fix them.

I talked this over with Doug and Ted, lets remove: AST, Basic, Lex,
Parse, Sema, Driver, Headers, Preprocessor and replace them with a
single "Frontend" component. That leaves us with:

-New Bugs: New bugs can be filed here, no need to categorize.
Frontend: Issues in the preprocessor, parser, and semantic analyzer.
LLVM Codegen: Issues with the translation to LLVM intermediate
representation (IR).
Static Analyzer: Automatic Bug Finding, scan-build, etc
C++: C++ Related Issues
C++0x: Issues related to the implementation of the upcoming C++0x standard.

And we can kill C++0x if we get a C++0x keyword.

 - Daniel

>
> -Chris
>
>>
>> - Daniel
>>
>> On Sat, Mar 6, 2010 at 12:10 PM, Chris Lattner <clattner at apple.com> wrote:
>>> FYI, I changed bugzilla to simplify it a bit.  I removed all of the different priority levels (P1-5) so that we now just have one priority (I couldn't figure out how to remove the field) and simplified the severity field to only include "enhancement" "normal" and "release blocker".
>>>
>>> There was some confusion about why bugs were "just a p5" etc, hopefully that is defined away now.
>>>
>>> -Chris
>>> _______________________________________________
>>> cfe-dev mailing list
>>> cfe-dev at cs.uiuc.edu
>>> http://lists.cs.uiuc.edu/mailman/listinfo/cfe-dev
>>>
>
>




More information about the cfe-dev mailing list