On Thu, 2020-08-27 at 11:04 -0700, Linus Torvalds wrote: > On Wed, Aug 26, 2020 at 6:36 PM Allen Pais <allen.l...@gmail.com> > wrote: > > > > Introduce container_from() as a generic helper instead of > > sub-systems defining a private from_ > > NAK. > > This seems completely broken. > > The documentation comment doesn't even match the macro, and claims > that "container" is a type. > > Which it isn't. That's what container_of()" already takes. > > And if the argument is that a broken commit introduced a broken > macro, > then that's not a great argument. Yes, we have that broken > "from_tasklet()" macro, but it's not even *USED* anywhere. > > So instead of adding a broken new concept that adds absolutely no > value, let's just remove the broken macro that isn't even used.
The argument is more over the tasklet rework adding from_tasklet that was effectively completely generic. https://lore.kernel.org/dri-devel/20200817091617.28119-1-allen.cryp...@gmail.com/ The tasklet rework people don't want to use container_of, which was our first suggestion, because it produces lines which are "too long". So container_from is a compromise that takes the actual structure pointer as a second argument instead of the structure type, thus being completely generic and suitable for line length reduction in the tasklet rework code. James