Re: [SLUB] nfs_page cmpxchg_double_fail and perf lock perturbation on dual-socket NFS/RDMA

[Harry Yoo <harry@kernel.org>]

On Fri, Sep 18, 2026 at 09:08:37AM +0200, Vlastimil Babka (SUSE) wrote:
> On 9/17/26 18:25, Tim Menninger wrote:
> > It reproduces with slab_nomerge.
> > 
> > I booted the same kernel with:
> > 
> >     $ cat /proc/cmdline
> >     ... slab_nomerge ...
> > 
> > Uninstrumented throughput remains in the high-45 to low-46 GB/s
> > range, as before. During perf lock it was around 30 GB/s in these
> > runs, versus roughly 25 GB/s previously.
> > 
> > The dominant lock measurements are essentially unchanged:
> > 
> >                          default             slab_nomerge
> >     unpinned/node0
> >       contentions        4,782,839             4,858,759
> >       total wait          14.39 min             14.26 min
> >       average wait       180.47 us             176.13 us
> > 
> >     node0/node0
> >       contentions        4,582,012             4,686,729
> >       total wait          14.13 min             14.34 min
> >       average wait       185.05 us             183.53 us
> > 
> >     unpinned/balanced
> >       contentions        4,572,196             4,544,677
> >       total wait          12.81 min             12.56 min
> >       average wait       168.14 us             165.84 us
> > 
> > The uninstrumented cmpxchg_double_fail deltas were:
> > 
> >                          default     slab_nomerge
> >     unpinned/node0          7,471          8,892
> >     node0/node0               713            457
> >     unpinned/balanced       1,503          1,504
> > 
> > The CPU-utilization change while perf lock is active also remains.
> > System-wide %idle for uninstrumented/instrumented windows was:
> > 
> >                            default                  slab_nomerge
> >                        uninstr.   instr.         uninstr.   instr.
> >     unpinned/node0      31.46%    3.60%           23.02%    2.15%
> >     node0/node0         84.26%   12.07%           84.62%   11.85%
> >     unpinned/balanced   66.86%   14.83%           66.58%   14.78%
> 
> I believe the weirdest part in the original report was how slub stat
> counters such as free_fastpath etc got order of magnitude higher with
> instrumented vs non-instrumented.
> Did any of that change with slab_nomerge?

If it didn't change, that'd be very surprising.
... unless perf lock was writing data to an NFS filesystem?

-- 
Cheers,
Harry / Hyeonggon