On Thu, 10 Sep 2015, Vlastimil Babka wrote: > > For a partial cacheline it would have to read the rest of the cacheline > > before updating. And I would expect the processor to have exclusive access > > to the cacheline that is held in a store buffer. If not then there is > > trouble afoot. > > IIRC that (or something similar with same guarantees) basically happens on x86 > when you use the LOCK prefix, i.e. for atomic inc etc. Doing that always would > destroy performance. Well yes but it also happens anytime you try to write to a cacheline. -- To unsubscribe, send a message with 'unsubscribe linux-mm' in the body to majordomo@xxxxxxxxx. For more info on Linux MM, see: http://www.linux-mm.org/ . Don't email: <a href=mailto:"dont@xxxxxxxxx";> email@xxxxxxxxx </a>