* Paul E. McKenney <paulmck@xxxxxxxxxxxxxxxxxx> wrote: > > The stop_machine_run()-alike thing is for brlocks - for which > > Sasha sent patches already, see these patches on the > > kvm@xxxxxxxxxxxxxxx list: > > > > [PATCH 3/4] kvm tools: Add a brlock > > [PATCH 4/4] kvm tools: Use brlock in MMIO and IOPORT > > > > Wrt. brlocks, we'll keep them as simple as possible and indeed no > > involved tricks are needed AFAICS. read_lock() will be a compiler > > barrier(), that's as fast as it gets :-) > > Makes sense! > > The other debugging use for the read-side primitives is to execute > the read-side ready-do-respond-to-kvm-pause code. This can help > catch bugs where the developer put the br_read_lock() and the > br_read_unlock() in the wrong place. That's a very good suggestion - it might in fact be simpler to implement than the 'replace by rw_lock' debugging variant. Thanks, Ingo -- To unsubscribe from this list: send the line "unsubscribe kvm" in the body of a message to majordomo@xxxxxxxxxxxxxxx More majordomo info at http://vger.kernel.org/majordomo-info.html
