Add evp_kdf perftool

[andrewkdinh]

Summary

Adds a CLI tool evp_kdf that derives keys from random data with SHA-256 digest.
Runs for 5 seconds and prints the average execution time per computation.

Fixes openssl/project#1839

Features

• Four modes of operation:
  • evp_shared (default): Use EVP API and allow shared data between computations
  • evp_isolated: Use EVP API and don't allow shared data between computations
  • deprecated_shared: Use legacy API and allow shared data between computations
  • deprecated_isolated: Use legacy API and don't allow shared data between computations
• Configurable number of times to update
• Terse output for easier CI automation (-t)
• Configurable thread count

TODO: Since freeze functionality hasn't been properly added yet (see openssl/project#1834 & openssl/openssl#29433), support for freezing the context store has not been added yet.

Usage

$ ./evp_kdf -h
Usage: evp_kdf [-h] [-t] [-o operation] [-V] thread-count
-h - print this help output
-t - terse output
-o operation - mode of operation. One of [evp_isolated, evp_shared, deprecated_isolated, deprecated_shared] (default: evp_shared)
-V - print version information and exit
thread-count - number of threads

$ ./evp_kdf -o evp_shared 10 # evp_shared operation mode, 10 threads
Average time per computation: 2393.604289us

$ ./evp_kdf -o evp_isolated 10 # now using evp_isolated mode
Average time per computation: 2476.350849us

Findings

Initial results showed that all modes take approximately the same amount of time

[Sashan]

looks good. the nit I'm pointing out can be ignored.

I think we should also sort out print_help() vs. usage() I would like perftools to consistently use one or the other. but this is yet another PR.

[Sashan]

NIT: I think counts and run_err can be also static here.

[Sashan]

changes were merged in, thanks