[DTrace-devel] [PATCH] Record CPU that BEGIN and END probes run on in the 'state' BPF map

Kris Van Hees kris.van.hees at oracle.com
Tue Sep 8 09:16:14 PDT 2020


On Tue, Sep 08, 2020 at 08:45:20AM -0700, Eugene Loh wrote:
> On 09/04/2020 11:41 PM, Kris Van Hees wrote:
> 
> > On Fri, Sep 04, 2020 at 10:12:25PM -0700, Eugene Loh wrote:
> >> There should be some test here, but that also means pairing this patch
> >> with something that makes the change observable and testable. So, hold
> >> off on this patch until it's usable?
> > It is harmless so no test would be required for this because it does not
> > affect any functionality.  Once the recording of beganon and endedon gets
> > used, that will get tests.  Or rather, we already have tests for that (and
> > they are failing) so we'll see tests passing.
> 
> If a patch is not tested, it should not go in.  "Harmless" is too low a 
> standard... in general, it feels like a different standard is being used 
> for patches depending on who is submitting the patch. Just hold off 
> until a complete patch(set) is ready.

No one said that this patch is going in until the rest of the patchset is
completed.  Every patch posted to the list has been tested, and sometimes
they do cause regressions in some areas because we occasionally have to
break how something is done because we need to put it back together the way
is really ought to be done (i.e. improve the implementation).

	Kris



More information about the DTrace-devel mailing list