Re: [PATCH tip/core/rcu 01/21] doc: READ_ONCE() now implies smp_barrier_depends()

[Peter Zijlstra]

On Mon, Dec 04, 2017 at 10:39:05PM +0000, David Howells wrote:
> Peter Zijlstra <peterz@xxxxxxxxxxxxx> wrote:
> 
> > > Good point!  How about as shown in the updated patch below?
> > 
> > Humm, I thought the idea was to completely remove read_barrier_depends
> > from the lkmm and memory-barriers.txt, making it an Alpha implementation
> > detail.
> 
> memory-barriers.txt explains how the barriers used by the kernel work, amongst
> other things.

The thing is, read-barrier-depends will no longer be used. read-read
dependencies will henceforth be assumed to 'just work'.

> Don't forget, btw, that Blackfin uses it also.

We should delete that blackfin code, its beyond insane.