On Mon, 28 Mar 2022 00:09:38 +0200 Albert Astals Cid <aa...@kde.org> wrote: [...] > Results of running clang-tidy with some of my favorite options > attached.
Thanks! > The first group [bugprone-integer-division] seems an actual bug since > double RKGraphicsDeviceFrontendTransmitter::lwdscale = 72/96; > means lwdscale value is 0 Fixed. > The second group [bugprone-parent-virtual-call] is worth looking at, > may be what you want or may not. None of these caused real misbehavior, but in none of the cases there was a terribly good reason for writing them, this way, either. Changed. > The third group [modernize-use-bool-literals] is totally just for our > monkey brains, the compiler doesn't care, so if you don't care either > it's fine to not change it Fixed. One was actually a (minor / optimization) bug, where the bool was meant to be an int. > The fourth group [performance-unnecessary-value-param] is just about > adding a few const & to copy less things, mostly it's Qt things that > are cheap to copy, so you're not going to get much performance, but > the const & is faster anyway, so why not do it?. Will do, but deferred for a quiet hour: https://invent.kde.org/education/rkward/-/issues/9 Regards Thomas
pgp9ORnxDk8LR.pgp
Description: OpenPGP digital signature
