[DTrace-devel] [PATCH 1/5] Implement setopt()

[Eugene Loh]

On 9/7/22 15:50, Kris Van Hees wrote:

> On Wed, Sep 07, 2022 at 03:32:33PM -0400, Eugene Loh via DTrace-devel wrote:
>> On 9/6/22 21:36, Kris Van Hees via DTrace-devel wrote:
>>
>>> The basic mechanics of the setopt() action are implemented in this
>>> patch.  Each dynamic runtime options that can be set using setopt()
>>> must provide support for it.
>> There ought to be at least one test, and if rudimentary setopt() support was
>> the only thing only keeping tst.keysort.d from passing then that could be
>> the test.
> I re-ordered the patches so that setopt is implemented before we add the
> xfail marker to tst.keysort.d (and deleted that addition) to avoid the whole
> situation.

You mean, ignore the posted patches?  Will you be posting the intended 
patch series?

Anyhow, whichever way tst.keysort.d is handled, 
test/unittest/multiaggs/tst.sortpos.d should probably be handled the 
same.  E.g., if you do end up putting setopt() in first, then just add 
@@nosort to tst.sortpos.d later.  I can post a patch, but maybe that 
would only confuse things more.

In any case, there ought to be some test that indicates that basic 
setopt() support was added.  Another candidate is 
test/unittest/misc/tst.badopt.d.  With the current patch, you can remove 
the xfail in tst.badopt.d.  Just be sure to fix the EPIDs in tst.badopt.r.