Hi Alvaro,
That's a very interesting case and might indeed be fixed or at least improved by this patch. I tried to reproduce this, but at least when running a simple, serial query with increasing numbers of functions, the time spent per function is linear or even slightly sub-linear (same as Tom observed in [1]).
I also couldn't reproduce the JIT runtimes you shared, when running the attached catalog query. The catalog query ran serially for me with the following JIT stats:
JIT:
Functions: 169
Options: Inlining true, Optimization true, Expressions true, Deforming true
Timing: Generation 12.223 ms, Inlining 17.323 ms, Optimization 388.491 ms, Emission 283.464 ms, Total 701.501 ms
Is it possible that the query ran in parallel for you? For parallel queries, every worker JITs all of the functions it uses. Even though the workers might JIT the functions in parallel, the time reported in the EXPLAIN ANALYZE output is the sum of the time spent by all workers. With this patch applied, the JIT time drops significantly, as many of the generated functions remain unused.
JIT:
Modules: 15
Functions: 26
Options: Inlining true, Optimization true, Expressions true, Deforming true
Timing: Generation 1.931 ms, Inlining 0.722 ms, Optimization 67.195 ms, Emission 70.347 ms, Total 140.195 ms
Of course, this does not prove that the nonlinearity that you observed went away. Could you share with me how you ran the query so that I can reproduce your numbers on master to then compare them with the patched version? Also, which LLVM version did you run with? I'm currently running with LLVM 13.
Thanks!
--
David Geier
(ServiceNow)
