Re: [PATCH 0/8] Introduce simple hazard pointers for lockdep

[Boqun Feng <boqun.feng@gmail.com>]

On Wed, Jun 25, 2025 at 08:25:52AM -0400, Mathieu Desnoyers wrote:
> On 2025-06-24 23:10, Boqun Feng wrote:
> > Hi,
> > 
> > This is the official first version of simple hazard pointers following
> > the RFC:
> > 
> > 	https://lore.kernel.org/lkml/20250414060055.341516-1-boqun.feng@gmail.com/
> > 
> > I rebase it onto v6.16-rc3 and hope to get more feedback this time.
> > 
> > Thanks a lot for Breno Leitao to try the RFC out and share the numbers.
> > 
> > I did an extra comparison this time, between the shazptr solution and
> > the synchronize_rcu_expedited() solution. In my test, during a 100 times
> > "tc qdisc replace" run:
> > 
> > * IPI rate with the shazptr solution: ~14 per second per core.
> > * IPI rate with synchronize_rcu_expedited(): ~140 per second per core.
> > 
> > (IPI results were from the 'CAL' line in /proc/interrupt)
> > 
> > This shows that while both solutions have the similar speedup, shazptr
> > solution avoids the introduce of high IPI rate compared to
> > synchronize_rcu_expedited().
> > 
> > Feedback is welcome and please let know if there is any concern or
> > suggestion. Thanks!
> 
> Hi Boqun,
> 
> What is unclear to me is what is the delta wrt:
> 
> https://lore.kernel.org/lkml/20241008135034.1982519-4-mathieu.desnoyers@efficios.com/
> 

shazptr is more close to the general hazptr I proposed earlier:

	https://lore.kernel.org/lkml/20240917143402.930114-1-boqun.feng@gmail.com/

, it's aimed as a global facility therefore no hazptr_domain is needed,
plus it supports non-busy waiting synchronize_shazptr() at the
beginning.

> and whether this helper against compiler optimizations would still be needed here:
> 
> https://lore.kernel.org/lkml/20241008135034.1982519-2-mathieu.desnoyers@efficios.com/
> 

For the current user, no, but eventually we are going to need it.

Regards,
Boqun

> Thanks,
> 
> Mathieu
> 
> > 
> > Regards,
> > Boqun
> > 
> > --------------------------------------
> > Please find the old performance below:
> > 
> > On my system (a 96-cpu VMs), the results of:
> > 
> > 	time /usr/sbin/tc qdisc replace dev eth0 root handle 0x1: mq
> > 
> > are (with lockdep enabled):
> > 
> > 	(without the patchset)
> > 	real    0m1.039s
> > 	user    0m0.001s
> > 	sys     0m0.069s
> > 
> > 	(with the patchset)
> > 	real    0m0.053s
> > 	user    0m0.000s
> > 	sys     0m0.051s
> > 
> > i.e. almost 20x speed-up.
> > 
> > Other comparisons between RCU and shazptr, the rcuscale results (using
> > default configuration from
> > tools/testing/selftests/rcutorture/bin/kvm.sh):
> > 
> > RCU:
> > 
> > 	Average grace-period duration: 7470.02 microseconds
> > 	Minimum grace-period duration: 3981.6
> > 	50th percentile grace-period duration: 6002.73
> > 	90th percentile grace-period duration: 7008.93
> > 	99th percentile grace-period duration: 10015
> > 	Maximum grace-period duration: 142228
> > 
> > shazptr:
> > 
> > 	Average grace-period duration: 0.845825 microseconds
> > 	Minimum grace-period duration: 0.199
> > 	50th percentile grace-period duration: 0.585
> > 	90th percentile grace-period duration: 1.656
> > 	99th percentile grace-period duration: 3.872
> > 	Maximum grace-period duration: 3049.05
> > 
> > shazptr (skip_synchronize_self_scan=1, i.e. always let scan kthread to
> > wakeup):
> > 
> > 	Average grace-period duration: 467.861 microseconds
> > 	Minimum grace-period duration: 92.913
> > 	50th percentile grace-period duration: 440.691
> > 	90th percentile grace-period duration: 460.623
> > 	99th percentile grace-period duration: 650.068
> > 	Maximum grace-period duration: 5775.46
> > 
> > shazptr_wildcard (i.e. readers always use SHAZPTR_WILDCARD):
> > 
> > 	Average grace-period duration: 599.569 microseconds
> > 	Minimum grace-period duration: 1.432
> > 	50th percentile grace-period duration: 582.631
> > 	90th percentile grace-period duration: 781.704
> > 	99th percentile grace-period duration: 1160.26
> > 	Maximum grace-period duration: 6727.53
> > 
> > shazptr_wildcard (skip_synchronize_self_scan=1):
> > 
> > 	Average grace-period duration: 460.466 microseconds
> > 	Minimum grace-period duration: 303.546
> > 	50th percentile grace-period duration: 424.334
> > 	90th percentile grace-period duration: 482.637
> > 	99th percentile grace-period duration: 600.214
> > 	Maximum grace-period duration: 4126.94
> > 
> > Boqun Feng (8):
> >    Introduce simple hazard pointers
> >    shazptr: Add refscale test
> >    shazptr: Add refscale test for wildcard
> >    shazptr: Avoid synchronize_shaptr() busy waiting
> >    shazptr: Allow skip self scan in synchronize_shaptr()
> >    rcuscale: Allow rcu_scale_ops::get_gp_seq to be NULL
> >    rcuscale: Add tests for simple hazard pointers
> >    locking/lockdep: Use shazptr to protect the key hashlist
> > 
> >   include/linux/shazptr.h  |  73 +++++++++
> >   kernel/locking/Makefile  |   2 +-
> >   kernel/locking/lockdep.c |  11 +-
> >   kernel/locking/shazptr.c | 318 +++++++++++++++++++++++++++++++++++++++
> >   kernel/rcu/rcuscale.c    |  60 +++++++-
> >   kernel/rcu/refscale.c    |  77 ++++++++++
> >   6 files changed, 534 insertions(+), 7 deletions(-)
> >   create mode 100644 include/linux/shazptr.h
> >   create mode 100644 kernel/locking/shazptr.c
> > 
> 
> 
> -- 
> Mathieu Desnoyers
> EfficiOS Inc.
> https://www.efficios.com