This is the mail archive of the gdb-patches@sources.redhat.com mailing list for the GDB project.


Index Nav: [Date Index] [Subject Index] [Author Index] [Thread Index]
Message Nav: [Date Prev] [Date Next] [Thread Prev] [Thread Next]
Other format: [Raw text]

Re: [rfa:testsuite} Overhaul sizeof.exp


Michael Elizabeth Chastain wrote:
> 
> Hmmm, let's start by looking at the alternatives.  Suppose someone writes
> a new test that revels an unexpected bug in gdb (hi Andrew).
> 
> [0] Stqtus quo: we reject the test.

I agree that this is not good.


> [1] FAIL it:  we accept the test and let it FAIL.

Makes it harder for engineers to check if a toolchain is healthy.


> [2] XFAIL it: we accept the test and mark it with XFAIL.

It will be forgotten in the middle of so many other XFAILs.


> [3] KFAIL it: we accept the test and mark it with KFAIL.
> 

With this one you can have a script that produces a "Release Notes"
section with the known bugs by retrieving the bug description from
the database (based on the bug id mentioned in the KFAIL).

The list of KFAILs corresponds to the bugs we all should be trying to 
fix (and that we can fix in GDB independently from any fixes to
compilers,
operation systems, etc.)



> [3] has a problem because KFAIL does not exist.  If the senior maintainers
> decide to approve [3] then I can add it to my analysis tables right away,
> but someone will have to add it to dejagnu.
> 

It is just copy and paste xfail code and add a bit to force the bug id
to be
mandatory.



> I agree with Daniel Jacobowitz; I think [2] is the right thing to do.

This same discussion keeps coming back.  We discovered by trial and
error that XFAILing things only serves to hide them.

But if everybody wants to try again, lets enforce some very stringent
police using the Gnats/Bugzilla database so that the XFAILs are tracked
and removed when the fix is in.



> It's also practical because we can start today.
> 

I don't think this is a good argument.  This has a long time impact 
and we should not do it the "easy" way.  We once xfailed things in
mass and it did not go very well. If we are going to do it lets do
it very cautiously, after discussing how effective our enforcement
police can be.





-- 
Fernando Nasser
Red Hat Canada Ltd.                     E-Mail:  fnasser@redhat.com
2323 Yonge Street, Suite #300
Toronto, Ontario   M4P 2C9


Index Nav: [Date Index] [Subject Index] [Author Index] [Thread Index]
Message Nav: [Date Prev] [Date Next] [Thread Prev] [Thread Next]