Re: [net] 4890b686f4: netperf.Throughput_Mbps -69.4% regression

[Feng Tang]

Hi Shakeel,

On Fri, Jul 01, 2022 at 08:47:29AM -0700, Shakeel Butt wrote:
> On Mon, Jun 27, 2022 at 8:49 PM Feng Tang <feng.tang@xxxxxxxxx> wrote:
> > I just tested it, it does perform better (the 4th is with your patch),
> > some perf-profile data is also listed.
> >
> >  7c80b038d23e1f4c 4890b686f4088c90432149bd6de 332b589c49656a45881bca4ecc0 e719635902654380b23ffce908d
> > ---------------- --------------------------- --------------------------- ---------------------------
> >      15722           -69.5%       4792           -40.8%       9300           -27.9%      11341        netperf.Throughput_Mbps
> >
> >       0.00            +0.3        0.26 ±  5%      +0.5        0.51            +1.3        1.27 ±  2%pp.self.__sk_mem_raise_allocated
> >       0.00            +0.3        0.32 ± 15%      +1.7        1.74 ±  2%      +0.4        0.40 ±  2%  pp.self.propagate_protected_usage
> >       0.00            +0.8        0.82 ±  7%      +0.9        0.90            +0.8        0.84        pp.self.__mod_memcg_state
> >       0.00            +1.2        1.24 ±  4%      +1.0        1.01            +1.4        1.44        pp.self.try_charge_memcg
> >       0.00            +2.1        2.06            +2.1        2.13            +2.1        2.11        pp.self.page_counter_uncharge
> >       0.00            +2.1        2.14 ±  4%      +2.7        2.71            +2.6        2.60 ±  2%  pp.self.page_counter_try_charge
> >       1.12 ±  4%      +3.1        4.24            +1.1        2.22            +1.4        2.51        pp.self.native_queued_spin_lock_slowpath
> >       0.28 ±  9%      +3.8        4.06 ±  4%      +0.2        0.48            +0.4        0.68        pp.self.sctp_eat_data
> >       0.00            +8.2        8.23            +0.8        0.83            +1.3        1.26        pp.self.__sk_mem_reduce_allocated
> >
> > And the size of 'mem_cgroup' is increased from 4224 Bytes to 4608.
> 
> Hi Feng, can you please try two more configurations? Take Eric's patch
> of adding ____cacheline_aligned_in_smp in page_counter and for first
> increase MEMCG_CHARGE_BATCH to 64 and for second increase it to 128.
> Basically batch increases combined with Eric's patch.

With increasing batch to 128, the regression could be reduced to -12.4%.

Some more details with perf-profile data below:

7c80b038d23e1f4c 4890b686f4088c90432149bd6de  Eric's patch                 Eric's patch + batch-64   Eric's patch + batch-128
---------------- --------------------------- --------------------------- --------------------------- --------------------------- 
     15722           -69.5%       4792           -27.9%      11341           -14.0%      13521           -12.4%      13772        netperf.Throughput_Mbps
 
      0.05            +0.2        0.27 ± 18%      +0.0        0.08 ±  6%      -0.1        0.00            -0.0        0.03 ±100%  pp.self.timekeeping_max_deferment
      0.00            +0.3        0.26 ±  5%      +1.3        1.27 ±  2%      +1.8        1.82 ± 10%      +2.0        1.96 ±  9%  pp.self.__sk_mem_raise_allocated
      0.00            +0.3        0.32 ± 15%      +0.4        0.40 ±  2%      +0.1        0.10 ±  5%      +0.0        0.00        pp.self.propagate_protected_usage
      0.00            +0.8        0.82 ±  7%      +0.8        0.84            +0.5        0.48            +0.4        0.36 ±  2%  pp.self.__mod_memcg_state
      0.00            +1.2        1.24 ±  4%      +1.4        1.44            +0.4        0.40 ±  3%      +0.2        0.24 ±  6%  pp.self.try_charge_memcg
      0.00            +2.1        2.06            +2.1        2.11            +0.5        0.50            +0.2        0.18 ±  8%  pp.self.page_counter_uncharge
      0.00            +2.1        2.14 ±  4%      +2.6        2.60 ±  2%      +0.6        0.58            +0.2        0.20        pp.self.page_counter_try_charge
      1.12 ±  4%      +3.1        4.24            +1.4        2.51            +1.0        2.10 ±  2%      +1.0        2.10 ±  9%  pp.self.native_queued_spin_lock_slowpath
      0.28 ±  9%      +3.8        4.06 ±  4%      +0.4        0.68            +0.6        0.90 ±  9%      +0.7        1.00 ± 11%  pp.self.sctp_eat_data
      0.00            +8.2        8.23            +1.3        1.26            +1.7        1.72 ±  6%      +2.0        1.95 ± 10%  pp.self.__sk_mem_reduce_allocated
 
 Thanks,
 Feng