On Friday, 2018-03-09 15:48:51 +0000, Emil Velikov wrote: > On 9 March 2018 at 11:47, Eric Engestrom <eric.engestrom@xxxxxxxxxx> wrote: > > >> > + dep_atomic_ops = [] > >> > > >> The dummy dep_atomic_ops is needed for almost all cases in the if/else > >> ladder. The exception being the .found() case. > > > > Actually, because of the way the `if` chain is written, the dep_atomic_ops > > has to be created before the list; in the `if` cases before this one, > > that dep might exist, but we don't need it. In the branch that uses it, > > if we got in the branch then it exists, and we need it. In the branches > > after that, it doesn't exist and we don't need it either. > > > > That means that only up to that branch can it contain anything; > > afterwards, it's empty anyway. > > > > And like I said, it only has the effect of maybe on the branches before > > adding unused flags to the compiler & linker, which shouldn't have any > > effect anyway. > > > Let me see if I get this right: > > You're saying that when dep_atomic_ops.found() == false, the > dep_atomic_ops dependency listed in the respective will expand to > no-op. > I wasn't quite sure if that's the case, hence my pedantic suggestion ;-) Exactly :) > > Makes sense though, rb still stands. OK, pushing in a minute > > Thanks Eric! > > -Emil _______________________________________________ dri-devel mailing list dri-devel@xxxxxxxxxxxxxxxxxxxxx https://lists.freedesktop.org/mailman/listinfo/dri-devel
