Re: [PATCH 17/19] srcu: Optimize SRCU-fast-updown for arm64

["Paul E. McKenney" <paulmck@kernel.org>]

On Mon, Nov 03, 2025 at 08:34:10AM -0500, Mathieu Desnoyers wrote:
> On 2025-11-02 16:44, Paul E. McKenney wrote:
> > Some arm64 platforms have slow per-CPU atomic operations, for example,
> > the Neoverse V2.  This commit therefore moves SRCU-fast from per-CPU
> > atomic operations to interrupt-disabled non-read-modify-write-atomic
> > atomic_read()/atomic_set() operations.  This works because
> > SRCU-fast-updown is not invoked from read-side primitives, which
> > means that if srcu_read_unlock_fast() NMI handlers.  This means that
> > srcu_read_lock_fast_updown() and srcu_read_unlock_fast_updown() can
> > exclude themselves and each other
> > 
> > This reduces the overhead of calls to srcu_read_lock_fast_updown() and
> > srcu_read_unlock_fast_updown() from about 100ns to about 12ns on an ARM
> > Neoverse V2.  Although this is not excellent compared to about 2ns on x86,
> > it sure beats 100ns.
> > 
> > This command was used to measure the overhead:
> > 
> > tools/testing/selftests/rcutorture/bin/kvm.sh --torture refscale --allcpus --duration 5 --configs NOPREEMPT --kconfig "CONFIG_NR_CPUS=64 CONFIG_TASKS_TRACE_RCU=y" --bootargs "refscale.loops=100000 refscale.guest_os_delay=5 refscale.nreaders=64 refscale.holdoff=30 torture.disable_onoff_at_boot refscale.scale_type=srcu-fast-updown refscale.verbose_batched=8 torture.verbose_sleep_frequency=8 torture.verbose_sleep_duration=8 refscale.nruns=100" --trust-make
> > 
> Hi Paul,
> 
> At a high level, what are you trying to achieve with this ?

I am working around the high single-CPU cost of arm64 LSE instructions,
as in about 50ns per compared non-LSE of about 5ns per.  The 50ns rules
them out for uretprobes, for example.

But Catalin's later patch is in all ways better than mine, so I will be
keeping this one only until Catalin's hits mainline.  Once that happens,
I will revert this one the following merge window.  (It might be awhile
because of the testing required on a wide range of platforms.)

> AFAIU, you are trying to remove the cost of atomics on per-cpu
> data from srcu-fast read lock/unlock for frequent calls for
> CONFIG_NEED_SRCU_NMI_SAFE=y, am I on the right track ?
> 
> [disclaimer: I've looked only briefly at your proposed patch.]
> Then there are various other less specific approaches to consider
> before introducing such architecture and use-case specific work-around.
> 
> One example is the libside (user level) rcu implementation which uses
> two counters per cpu [1]. One counter is the rseq fast path, and the
> second counter is for atomics (as fallback).
> 
> If the typical scenario we want to optimize for is thread context, we
> can probably remove the atomic from the fast path with just preempt off
> by partitioning the per-cpu counters further, one possibility being:
> 
> struct percpu_srcu_fast_pair {
> 	unsigned long lock, unlock;
> };
> 
> struct percpu_srcu_fast {
> 	struct percpu_srcu_fast_pair thread;
> 	struct percpu_srcu_fast_pair irq;
> };
> 
> And the grace period sums both thread and irq counters.
> 
> Thoughts ?

One complication here is that we need srcu_down_read() at task level
and the matching srcu_up_read() at softirq and/or hardirq level.

Or am I missing a trick in your proposed implementation?

							Thanx, Paul

> Thanks,
> 
> Mathieu
> 
> [1] https://github.com/compudj/libside/blob/master/src/rcu.h#L71
> 
> -- 
> Mathieu Desnoyers
> EfficiOS Inc.
> https://www.efficios.com