Adds support to decode and encode the AVX-512 compressed displacement. 3 helper functions have been added that have been optimized to cover all cases of Intel's Vol.2A 2.6.5 "Compressed Displacement (disp8*N) Support in EVEX".
Adds tests for the above by duplicating relevant api.ir evex cases using a compressible displacement. The results have been verified manually and compared to binutils and llvm-mc, lacking/pending #3638.
Removes stale comment in ir_x86.c.
Issue: #1312