On Fri, 24 Jul 2020 23:14:17 +1000, Nicholas Piggin wrote:
> Updated with everybody's feedback (thanks all), and more performance
> results.
>
> What I've found is I might have been measuring the worst load point for
> the paravirt case, and by looking at a range of loads it's clear that
> queued spinlocks are overall better even on PV, doubly so when you look
> at the generally much improved worst case latencies.
>
> [...]
Applied to powerpc/next.
[1/6] powerpc/pseries: Move some PAPR paravirt functions to their own file
https://git.kernel.org/powerpc/c/20d444d06f97504d165b08558678b4737dcefb02
[2/6] powerpc: Move spinlock implementation to simple_spinlock
https://git.kernel.org/powerpc/c/12d0b9d6c843e7dbe739ebefcf16c7e4a45e4e78
[3/6] powerpc/64s: Implement queued spinlocks and rwlocks
https://git.kernel.org/powerpc/c/aa65ff6b18e0366db1790609956a4ac7308c5668
[4/6] powerpc/pseries: Implement paravirt qspinlocks for SPLPAR
https://git.kernel.org/powerpc/c/20c0e8269e9d515e677670902c7e1cc0209d6ad9
[5/6] powerpc/qspinlock: Optimised atomic_try_cmpxchg_lock() that adds the lock
hint
https://git.kernel.org/powerpc/c/2f6560e652dfdbdb59df28b45a3458bf36d3c580
[6/6] powerpc: Implement smp_cond_load_relaxed()
https://git.kernel.org/powerpc/c/49a7d46a06c30c7beabbf9d1a8ea1de0f9e4fdfe
cheers
