On Fri, 2018-01-05 at 15:52 -0500, John Ferlan wrote:
Suggestion - follow cmdNodeinfo for virhostcputest.c and don't
print the
MHz if it's 0. It's the indication that the data wasn't obtainable. I
think printing 0 is perhaps just as bad as printing the erroneous 100.
I trust you can make the adjustment without another round of patches.
Good idea. However, I've implemented it as a separate patch for a
few reasons:
* I've converted the test to use virBuffer, so the changes are not
trivial enough that I feel comfortable pushing them without review;
* there are already a couple of files that contain 'MHz: 0' in the
test suite, so it feels cleaner to change all of them at once;
* it's better to avoid performing more than one change per commit.
The follow-up patch is already on the list, in case you feel like
taking a look at it :)
--
Andrea Bolognani / Red Hat / Virtualization