LSUThorns/Vectors: Simplify API for partial vector stores

Issue #682 closed
Erik Schnetter created an issue

Implement vec_store_nta_partial, which offers a simpler interface, similar to the one used in OpenCL.

Keyword:

Comments (4)

  1. Barry Wardell
    • removed comment

    It looks like you have two "else if (v8stp_lo)" in the SSE4A part of the patch. Should the second of these be "else if (v8stp_hi)"? If so,would the tests have picked this up on an SSE4A machine? If so, and the AVX code also pases the test, I think this should be applied. I'd also recommend adding a pseudo-code comment explaining what the AVX version does.

  2. Log in to comment