Lokathor/safe_arch

Issues

• Is `safe_arch` actually any safer then `core::arch`?

  #112 opened 9 months ago by CeleritasCelery
  4

• ARM Support?

  #98 opened 3 years ago by BillyDM
  4

• source contains backup files

  #110 opened a year ago by jonassmedegaard
  0

• please tag release v0.7.0

  #109 opened a year ago by jonassmedegaard
  3

• Publish version of new crate

  #108 opened a year ago by mcroomp
  1

• Functions should say in their docs what intrinsics / assembly they implement

  #49 opened 4 years ago by Lokathor
  6

• How would you feel about macrofying more of the implementation?

  #106 opened a year ago by HadrienG2
  4

• Missing 256-bit vpermps ?

  #104 opened a year ago by HadrienG2
  2

• Naming convention cleanup

  #105 opened a year ago by HadrienG2
  1

• License questions

  #103 opened a year ago by werdahias
  2

• [avx512] General AVX-512 tracking issue

  #77 opened 3 years ago by Lokathor
  4

• check all integer multiplies

  #102 opened 3 years ago by Lokathor
  0

• Compîlation failed in amd64 on copr

  #99 opened 3 years ago
  1

• convert most macro_rules to min-const-generics

  #96 opened 3 years ago by Lokathor
  1

• Full double check of all mainline functions (`sse` through `avx2`)

  #44 opened 3 years ago by Lokathor
  0

• Convert all doc tests to integration tests

  #38 opened 4 years ago by Lokathor
  4

• [avx2] gather operations

  #62 opened 4 years ago by Lokathor
  2

• [fma] add all the fma operations

  #47 opened 4 years ago by Lokathor
  1

• Try to make macro names terser without losing clarity

  #52 opened 4 years ago by Lokathor
  1

• Ensure all macros have absolute `use` statements.

  #55 opened 4 years ago by Lokathor
  3

• [avx][avx2][fma] Gate high-power-draw instructions behind additional cargo features

  #42 opened 4 years ago by Lokathor
  2

• Full double check of all sideline functions (The ones not `sse` through `avx2`)

  #45 opened 3 years ago by Lokathor
  1

• [neon] General ARM/Neon tracking issue.

  #86 opened 3 years ago by Lokathor
  0

• `cmp_ordinary` should be named `cmp_ordered`

  #93 opened 3 years ago by Lokathor
  0

• macro blend_i32_m128i should have name with `imm` in it

  #92 opened 3 years ago by Lokathor
  0

• move_mask_m256i should be move_mask_i8_m256i

  #94 opened 3 years ago by Lokathor
  0

• Some macros continue to not prefix the value constructor used with `$crate::`

  #89 opened 4 years ago by Lokathor
  0

• Improve CI

  #90 opened 4 years ago by Lokathor
  0

• Put an entry for `blend` into the naming conventions document

  #88 opened 4 years ago by Lokathor
  0

• fix shuffle/permute naming

  #79 opened 4 years ago by Lokathor
  4

• [avx] the ops that chop 256-bit lanes to 128-bit lanes.

  #41 opened 4 years ago by Lokathor
  0

• [avx] testc, textnzc, testz

  #40 opened 4 years ago by Lokathor
  1

• Sort out all the "convert" operations

  #76 opened 4 years ago by Lokathor
  1

• rename: andnot to bitandnot

  #85 opened 4 years ago by Lokathor
  0

• widen mul

  #80 opened 4 years ago by Lokathor
  2

• change and/or/xor to be bitand/bitor/bitxor

  #82 opened 4 years ago by Lokathor
  0

• Put more verbs in the glossary

  #51 opened 4 years ago by Lokathor
  3

• Sort out all bit shifting conventions

  #75 opened 4 years ago by Lokathor
  1

• Remove super::* imports

  #64 opened 4 years ago by Soveu
  5

• Delete underscores from submodule filenames

  #63 opened 4 years ago by Soveu
  3

• Fix all floating point docs about "ord" and "unord"

  #37 opened 4 years ago by Lokathor
  3

• [avx2] add all the avx2 operations

  #43 opened 4 years ago by Lokathor
  0

• Remove [must_use] from functions that don't return anything

  #72 opened 4 years ago by Soveu
  0

• _mm256_insertf128_si256(avx) vs _mm256_inserti128_si256 (avx2)

  #67 opened 4 years ago by Lokathor
  8

• splat / set_splat inconsistency

  #66 opened 4 years ago by Lokathor
  0

• [avx][avx2] Operator overloads for 256-bit types

  #46 opened 4 years ago by Lokathor
  0

• inside of `sse4_2.rs` toggles on the `sse4.1` feature (thankfully `lib.rs` is still correct)

  #60 opened 4 years ago by Lokathor
  0

• test_mixed_ones_and_zeroes_m128i

  #48 opened 4 years ago by Lokathor
  0

• 256-bit registers aren't "sse" registers they're "avx" registers.

  #50 opened 4 years ago by Lokathor
  0

• Shuffle / Permute cleanup

  #39 opened 4 years ago by Lokathor
  1