Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Jump to bottom

SIMD groundwork #1199

Merged
merged 25 commits into from
Sep 18, 2015
Merged

SIMD groundwork #1199

merged 25 commits into from
Sep 18, 2015

Commits on Jul 7, 2015

  1. Lay groundwork for SIMD.

    @huonw
    huonw committed Jul 7, 2015
    Configuration menu
    Copy the full SHA
    c7b72a0 View commit details
    Browse the repository at this point in the history

Commits on Jul 8, 2015

  1. First round of changes: remove extraneous traits etc.

    @huonw
    huonw committed Jul 8, 2015
    Configuration menu
    Copy the full SHA
    272ce9b View commit details
    Browse the repository at this point in the history

  2. Second round of changes: minor tweaks.

    @huonw
    huonw committed Jul 8, 2015
    Configuration menu
    Copy the full SHA
    5893b16 View commit details
    Browse the repository at this point in the history

Commits on Jul 9, 2015

  1. Clarify/fix typos.

    @huonw
    huonw committed Jul 9, 2015
    Configuration menu
    Copy the full SHA
    f9e48d1 View commit details
    Browse the repository at this point in the history

  2. Note that fixed-length arrays could be repr(simd)'d.

    @huonw
    huonw committed Jul 9, 2015
    Configuration menu
    Copy the full SHA
    e2fc223 View commit details
    Browse the repository at this point in the history

Commits on Jul 10, 2015

  1. Remove the simd_primitive_trait attribute. 

    Not really necessary: the type safety it offers can be provided by libraries.
    @huonw
    huonw committed Jul 10, 2015
    Configuration menu
    Copy the full SHA
    efeafdb View commit details
    Browse the repository at this point in the history

  2. Mention alignment changes due to repr(simd).

    @huonw
    huonw committed Jul 10, 2015
    Configuration menu
    Copy the full SHA
    a7c409b View commit details
    Browse the repository at this point in the history

  3. Note pre-RFC discussion.

    @huonw
    huonw committed Jul 10, 2015
    Configuration menu
    Copy the full SHA
    f4e2ecf View commit details
    Browse the repository at this point in the history

Commits on Jul 13, 2015

  1. Clarify how the intrinsics' structural typing works.

    @huonw
    huonw committed Jul 13, 2015
    Configuration menu
    Copy the full SHA
    1132ede View commit details
    Browse the repository at this point in the history

  2. Add arithmetic intrinsics alternative.

    @huonw
    huonw committed Jul 13, 2015
    Configuration menu
    Copy the full SHA
    8317ea4 View commit details
    Browse the repository at this point in the history

  3. Write down an answer to "why not asm!?".

    @huonw
    huonw committed Jul 13, 2015
    Configuration menu
    Copy the full SHA
    c6ed18a View commit details
    Browse the repository at this point in the history

Commits on Jul 14, 2015

  1. point to cfg-if.

    @huonw
    huonw committed Jul 14, 2015
    Configuration menu
    Copy the full SHA
    67f78ec View commit details
    Browse the repository at this point in the history

Commits on Aug 3, 2015

  1. Use intrinsics for arithmetic instead of built-in operators.

    @huonw
    huonw committed Aug 3, 2015
    Configuration menu
    Copy the full SHA
    f71c4b3 View commit details
    Browse the repository at this point in the history

  2. Accidentally: 

    - an extra word.
- a subject-verb agreement.
- an ly.
- a plural.
    @huonw
    huonw committed Aug 3, 2015
    Configuration menu
    Copy the full SHA
    8b2ec8c View commit details
    Browse the repository at this point in the history

Commits on Aug 6, 2015

  1. Use the platform-intrinsic ABI instead of rust-intrinsic.

    @huonw
    huonw committed Aug 6, 2015
    Configuration menu
    Copy the full SHA
    47f6ae9 View commit details
    Browse the repository at this point in the history

  2. feature(simd_basics) -> feature(repr_simd) 

    This feature gate now only applies to the attribute, so it might as well
be more specific.
    @huonw
    huonw committed Aug 6, 2015
    Configuration menu
    Copy the full SHA
    4a4e6ae View commit details
    Browse the repository at this point in the history

Commits on Aug 12, 2015

  1. Remove struct flattening. 

    This is non-trivial (for me) to implement, and ended up not being that
useful, i.e. it wasn't needed to make useful things.
    @huonw
    huonw committed Aug 12, 2015
    Configuration menu
    Copy the full SHA
    c4bf5e1 View commit details
    Browse the repository at this point in the history

  2. Change shuffles to use arrays of indices. 

    This is *far* more scalable than having an argument for each value.

Thanks to @pnkfelix for the suggestion.
    @huonw
    huonw committed Aug 12, 2015
    Configuration menu
    Copy the full SHA
    6532670 View commit details
    Browse the repository at this point in the history

  3. shuffles don't rely on generic types for return values. 

    This has less type safety, but doesn't require generic simd types to
exist:

    #[repr(simd)] struct Simd2<T>(T, T);
    @huonw
    huonw committed Aug 12, 2015
    Configuration menu
    Copy the full SHA
    8e3a0de View commit details
    Browse the repository at this point in the history

  4. Intrinsics-for-operations is now the RFC, not an alternative. 

    Also, the comparison comment no longer makes sense.
    @huonw
    huonw committed Aug 12, 2015
    Configuration menu
    Copy the full SHA
    54b0927 View commit details
    Browse the repository at this point in the history

  5. Out of bounds indices are errors (backwards compat to relax).

    @huonw
    huonw committed Aug 12, 2015
    Configuration menu
    Copy the full SHA
    9e31ad3 View commit details
    Browse the repository at this point in the history

  6. Only invalid to *call* intrinsics on bad platforms. 

    It's valid to `extern` them, though.
    @huonw
    huonw committed Aug 12, 2015
    Configuration menu
    Copy the full SHA
    91a2b36 View commit details
    Browse the repository at this point in the history

  7. There can be more shuffles.

    @huonw
    huonw committed Aug 12, 2015
    Configuration menu
    Copy the full SHA
    60931df View commit details
    Browse the repository at this point in the history

Commits on Aug 14, 2015

  1. Internal references are legal. 

    Automatic crazy boolean bit-packing is crazy.
    @huonw
    huonw committed Aug 14, 2015
    Configuration menu
    Copy the full SHA
    135ba7d View commit details
    Browse the repository at this point in the history

  2. Type-level integer/values alternatives for shuffles.

    @huonw
    huonw committed Aug 14, 2015
    Configuration menu
    Copy the full SHA
    67fea6e View commit details
    Browse the repository at this point in the history