Segfault when running pippenger_bench

[jon-chuang]

make run_pippenger_bench -j16

initializing
executing normal fft
make[3]: *** [src/aztec/plonk/pippenger_bench/CMakeFiles/run_pippenger_bench.dir/build.make:57: src/aztec/plonk/pippenger_bench/CMakeFiles/run_pippenger_bench] Segmentation fault (core dumped)
make[2]: *** [CMakeFiles/Makefile2:3153: src/aztec/plonk/pippenger_bench/CMakeFiles/run_pippenger_bench.dir/all] Error 2
make[1]: *** [CMakeFiles/Makefile2:3160: src/aztec/plonk/pippenger_bench/CMakeFiles/run_pippenger_bench.dir/rule] Error 2

cmake version 3.16.3
make version 4.2.1

[jon-chuang]

polynomials_bench worked as an alternative.

Results for 2^20 pippenger on 8-core Ryzen 4800H w/ 8MB L3 and 4.1GHz sustained all core boost, with SMT on:
barretenberg: 520ms (490ms CPU)
Zexe: 551ms.

The fat to be trimmed is not a Zexe issue, but lies in the radix sort used from an external library, which takes around 90ms compared to barretenberg's < 48ms. Once we trim the radix sort to be more cache friendly for multicore, hopefully we can close the gap.