Stream: git-wasmtime

Topic: wasmtime / Issue #2146 Enhance machinst x64 instruction s...

Wasmtime GitHub notifications bot (Aug 19 2020 at 14:32):

github-actions[bot] commented on Issue #2146:

Subscribe to Label Action

cc @bnjbvr

<details>
This issue or pull request has been labeled: "cranelift", "cranelift:area:machinst", "cranelift:area:x64"

Thus the following users have been cc'd because of the following labels:

• bnjbvr: cranelift

To subscribe or unsubscribe from this label, edit the <code>.github/subscribe-to-label.json</code> configuration file.

Learn more.
</details>

Wasmtime GitHub notifications bot (Aug 20 2020 at 19:08):

bjorn3 commented on Issue #2146:

For reference this is the speedup of simple-raytracer compiled with cg_clif:

Benchmark #1: raytracer_cg_clif_x64_post
  Time (mean ± σ):      8.365 s ±  0.047 s    [User: 8.357 s, System: 0.008 s]
  Range (min … max):    8.297 s …  8.470 s    10 runs

Benchmark #2: ./raytracer_cg_clif_x64_pre
  Time (mean ± σ):      8.492 s ±  0.052 s    [User: 8.485 s, System: 0.007 s]
  Range (min … max):    8.425 s …  8.600 s    10 runs

Summary
  'raytracer_cg_clif_x64_post' ran
    1.02 ± 0.01 times faster than './raytracer_cg_clif_x64_pre'

Wasmtime GitHub notifications bot (Aug 20 2020 at 19:09):

bjorn3 edited a comment on Issue #2146:

For reference this is the speedup of simple-raytracer compiled with cg_clif:

Benchmark #1: ./raytracer_cg_clif_x64_post
  Time (mean ± σ):      8.365 s ±  0.047 s    [User: 8.357 s, System: 0.008 s]
  Range (min … max):    8.297 s …  8.470 s    10 runs

Benchmark #2: ./raytracer_cg_clif_x64_pre
  Time (mean ± σ):      8.492 s ±  0.052 s    [User: 8.485 s, System: 0.007 s]
  Range (min … max):    8.425 s …  8.600 s    10 runs

Summary
  './raytracer_cg_clif_x64_post' ran
    1.02 ± 0.01 times faster than './raytracer_cg_clif_x64_pre'

Wasmtime GitHub notifications bot (Aug 24 2020 at 07:56):

julian-seward1 commented on Issue #2146:

TODO relating to PR2149: change the lowering for Opcode::AtomicStore so as to use lower_to_amode rather than forcing the amode to be of the form 0(reg).

Wasmtime GitHub notifications bot (Aug 24 2020 at 07:58):

julian-seward1 edited a comment on Issue #2146:

TODO relating to PR2149: change the lowering for Opcode::AtomicStore so as to use lower_to_amode rather than forcing the amode to be of the form 0(reg). Also, potentially the same for AtomicCas and AtomicStore.

Last updated: Dec 23 2024 at 12:05 UTC