Re: [PATCH v2] bpf/verifier: optimize ID mapping reset in states_equal

[Eduard Zingerman]

On Fri, 2026-01-16 at 17:08 +0800, Qiliang Yuan wrote:

[...]

> Metric          | Baseline      | Patched       | Delta
> ----------------|---------------|---------------|----------
> Iterations      | 5710          | 5731          | +0.37%
> Instructions    | 1.714 T       | 1.555 T       | -9.28%
> Inst/Iter       | 300.2 M       | 271.3 M       | -9.63%
> Cycles          | 1.436 T       | 1.335 T       | -7.03%
> Branches        | 350.4 B       | 311.9 B       | -10.99%
> Migrations      | 25,977        | 23,524        | -9.44%
> 
> Test Command:
>   seq 1 2000000 | sudo perf stat -a -- \
>     timeout 60s xargs -P $(nproc) -I {} ./veristat access_map_in_map.bpf.o

[...]

As discussed in a separate thread, I don't think that system-wide
profiling is a good fit here. When using perf stat for collecting
pyperf180.bpf.o processing stats I don't see a big difference.
Regardless of specific statistics, I think this change should be landed.
 
> Suggested-by: Eduard Zingerman <eddyz87@xxxxxxxxx>

Please remove this 'Suggested-by', you found this change yourself,
Andrii suggested improving it further (this usually goes to a changelog).

> Suggested-by: Andrii Nakryiko <andrii@xxxxxxxxxx>
> Signed-off-by: Qiliang Yuan <realwujing@xxxxxxxxx>
> ---

Acked-by: Eduard Zingerman <eddyz87@xxxxxxxxx>

[...]