drop imul rewriting
commite8fa27bcdb2872588e39b398f7857157fb249628
authorQuentin Carbonneaux <quentin@c9x.me>
Wed, 19 Jun 2024 10:39:21 +0000 (19 12:39 +0200)
committerQuentin Carbonneaux <quentin@c9x.me>
Wed, 19 Jun 2024 10:39:26 +0000 (19 12:39 +0200)
treec10d441a877d92deae16e5fc23378dc5d718561c
parent7de34d3454e0a6a4e66b414f03441723d340dec5
drop imul rewriting

This was cute to do, but it is
largely inconsequential, as shown
by the rough timings below:

    benchmarking mul8_lea
      3.9 ticks ± 0.88 (min: 3)
    benchmarking mul8_imul
      3.3 ticks ± 0.27 (min: 3)
    benchmarking div8_udiv
      6.5 ticks ± 0.52 (min: 6)
    benchmarking div8_shr
      3.3 ticks ± 0.34 (min: 3)
amd64/emit.c