This is the mail archive of the gdb@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]

Re: RFA: Testsuite patches...


[suggest following up to gdb@sourceware - not patch specific]

Andrew Cagney wrote:
> 
> Um, careful here.

Double Um, I should be more careful, sorry :-(

> I don't think the question to ask is ``who will fix the failures''.
> Rather, I think people should be asking  ``does the test check correct
> functionality''?
> 
> If the test is correct and applicable to more than just HP then I think
> it should be enabled.  If it adds to the number of failures for certain
> targets then ``oops'' :-)  I'd assume the relevant maintainers would
> simply append this to their list of known problems.
> 
> Perhaphs post something like a transcript (assuming they are not to
> large) of each test so people can see them in action.
> 
> As a generalization, GDB desperatly needs more tests.

Its been pointed out I've got the current policy wrong.  There was a
debate (internal to Cygnus though :-() but the policy didn't get changed
- in general a testsuite shouldn't go in unless it is accompanied by a
fix.

From the above you can tell that I think this is too stringent.  I think
GDB should be accepting tests (provided that they are rigiously
examined) even when they add failures - just as long as the failures
examine real bugs.  I think this also better reflects what really goes
on.

BTW, as twist too all this, HP and Cygnus were quietly contributing
tests (call-*-st.exp in particular) which, at least initially, only
passed on HP platforms.

thoughs?

	enjoy,
		Andrew

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