On Fri, 2023-02-17 at 10:31 +0100, Greg KH wrote:
> On Fri, Feb 17, 2023 at 10:13:08AM +0100, Johannes Berg wrote:
> > On Fri, 2023-02-17 at 08:43 +0100, Greg KH wrote:
> > > On Fri, Feb 17, 2023 at 02:11:38PM +0900, Jaewan Kim wrote:
> > > > BTW, can I expect you to review my changes for further patchsets?
> > > > I sometimes get conflicting opinions (e.g. line limits)
> > >
> > > Sorry, I was the one that said "you can use 100 columns", if that's not
> > > ok in the networking subsystem yet, that was my fault as it's been that
> > > way in other parts of the kernel tree for a while.
> > >
> >
> > Hah. Maybe that's my mistake then, I was still at "use 80 columns where
> > it's simple, and more if it would look worse" ...
>
> It was changed back in 2020:
> bdc48fa11e46 ("checkpatch/coding-style: deprecate 80-column warning")
>
> seems to take a while to propagate out to all the subsystems :)
Ah no, I was aware of that, but I guess we interpret this bit
differently:
+Statements longer than 80 columns should be broken into sensible chunks,
+unless exceeding 80 columns significantly increases readability and does
+not hide information.
Here, I would've said something like:
+ if (request->request_lci && nla_put_flag(msg, NL80211_PMSR_FTM_REQ_ATTR_REQUEST_LCI))
+ return -ENOBUFS;
can indeed "be broken into sensible chunks, unless ..."
Just like this one already did:
+ if (request->request_civicloc &&
+ nla_put_flag(msg, NL80211_PMSR_FTM_REQ_ATTR_REQUEST_CIVICLOC))
+ return -ENOBUFS;
Personally I think the latter is easier to read because scanning the
long line for the logical break at "&&" is harder for me, but YMMV.
johannes
