On Wed, 2019-02-20 at 07:33 -0500, John Ferlan wrote:
On 2/20/19 2:37 AM, Michal Privoznik wrote:
> On 2/19/19 9:19 PM, John Ferlan wrote:
> > Assuming extraction (sigh) of the VIR_AUTOFREE,
>
> While I'd definitely want this to be split into two patches if it was
> fixing something under src/, but this is under tests/ and therefore I
> did not bother. The reason for splitting a patch into smaller
> semanticaly divided patches is to help distro maintainers to ease
> backports.
And reviewers, and people digging through history possibly years
down the line.
Whether the change is in src/, tests/ or whatever else shouldn't
make a difference, it's all code and we routinely have to fix bugs
both in the library and in the corresponding test suite.
There is barely ever a reason *not* to split changes into smaller,
independent units; having to write "at the same time" in the commit
message should be your hint that you're doing it wrong ;)
> I don't think they will need to backport this patch, nor
will
> they want only a part of it.
Just going with precedent I've seen for other patches regardless of
where they're found in the tree. Personally, I'm fine with doing it all
at once. But I think perhaps some one should take the time to write
down what the "house rules" are on the hacking page. Makes it easier
that way.
We *kinda* have that already:
Split large changes into a series of smaller patches,
self-contained if possible [...]
Perhaps we could reword that passage so it's clearer, and extend
it by going into more detail.
--
Andrea Bolognani / Red Hat / Virtualization