Remove the Cranelift vselect instruction (#5918)

* Remove the Cranelift `vselect` instruction

This instruction is documented as selecting lanes based on the "truthy"
value of the condition lane, but the current status of the
implementation of this instruction is:

* x64 - uses the high bit for `f32x4` and `f64x2` and otherwise uses the
  high bit of each byte doing a byte-wise lane select rather than
  whatever the controlling type is.

* AArch64 - this is the same as `bitselect` which is a bit-wise
  selection rather than a lane-wise selection.

* s390x - this is the same as AArch64, a bit-wise selection rather than
  lane-wise.

* interpreter - the interpreter implements the documented semantics of
  selecting based on "truthy" values.

Coupled with the status of the implementation is the fact that this
instruction is not used by WebAssembly SIMD today either. The only use
of this instruction in Cranelift is the nan-canonicalization pass. By
moving nan-canonicalization to `bitselect`, since that has the desired
semantics, there's no longer any need for `vselect`.

Given this situation this commit subsqeuently removes `vselect` and all
usage of it throughout Cranelift.

Closes #5917

* Review comments

* Bring back vselect opts as bitselect opts

* Clean up vselect usage in the interpreter

* Move bitcast in nan canonicalization

* Add a comment about float optimization

cranelift/filetests/filetests/runtests/simd-vselect.clif

@@ -1,82 +0,0 @@
-test interpret
-test run
-target s390x
-target aarch64
-set enable_simd
-target x86_64 has_sse3 has_ssse3 has_sse41
-
-function %vselect_i8x16() -> i8x16 {
-block0:
-    v1 = vconst.i8x16 [0 -1 0 -1 0 -1 -1 -1 -1 -1 0 0 0 0 0 0]
-    v2 = vconst.i8x16 [100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115]
-    v3 = vconst.i8x16 [200 201 202 203 204 205 206 207 208 209 210 211 212 213 214 215]
-    v4 = vselect v1, v2, v3
-    return v4
-}
-; run: %vselect_i8x16() == [200 101 202 103 204 105 106 107 108 109 210 211 212 213 214 215]
-
-function %vselect_i16x8() -> i16x8 {
-block0:
-    v1 = vconst.i16x8 [0 -1 0 -1 0 -1 -1 -1]
-    v2 = vconst.i16x8 [100 101 102 103 104 105 106 107]
-    v3 = vconst.i16x8 [200 201 202 203 204 205 206 207]
-    v4 = vselect v1, v2, v3
-    return v4
-}
-; run: %vselect_i16x8() == [200 101 202 103 204 105 106 107]
-
-function %vselect_i32x4_const() -> i32x4 {
-block0:
-    v1 = vconst.i32x4 [0 -1 0 -1]
-    v2 = vconst.i32x4 [100 101 102 103]
-    v3 = vconst.i32x4 [200 201 202 203]
-    v4 = vselect v1, v2, v3
-    return v4
-}
-; run: %vselect_i32x4_const() == [200 101 202 103]
-
-function %vselect_i32x4(i32x4, i32x4, i32x4) -> i32x4 {
-block0(v0: i32x4, v1: i32x4, v2: i32x4):
-    v3 = vselect v0, v1, v2
-    return v3
-}
-; Remember that vselect accepts: 1) the selector vector, 2) the "if true" vector, and 3) the "if false" vector.
-; run: %vselect_i32x4([-1 -1 0 0], [1 2 -1 -1], [-1 -1 3 4]) == [1 2 3 4]
-
-function %vselect_i64x2() -> i64x2 {
-block0:
-    v1 = vconst.i64x2 [0 -1]
-    v2 = vconst.i64x2 [100 101]
-    v3 = vconst.i64x2 [200 201]
-    v4 = vselect v1, v2, v3
-    return v4
-}
-; run: %vselect_i64x2() == [200 101]
-
-function %vselect_p_i8x16(i8x16, i8x16, i8x16) -> i8x16 {
-block0(v0: i8x16, v1: i8x16, v2: i8x16):
-    v3 = vselect v0, v1, v2
-    return v3
-}
-; run: %vselect_p_i8x16([-1 0 -1 -1 -1 0 0 0 -1 0 -1 -1 -1 0 0 0], [1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16], [17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32]) == [1 18 3 4 5 22 23 24 9 26 11 12 13 30 31 32]
-
-function %vselect_p_i16x8(i16x8, i16x8, i16x8) -> i16x8 {
-block0(v0: i16x8, v1: i16x8, v2: i16x8):
-    v3 = vselect v0, v1, v2
-    return v3
-}
-; run: %vselect_p_i16x8([-1 0 -1 -1 -1 0 0 0], [1 2 3 4 5 6 7 8], [17 18 19 20 21 22 23 24]) == [1 18 3 4 5 22 23 24]
-
-function %vselect_p_i32x4(i32x4, i32x4, i32x4) -> i32x4 {
-block0(v0: i32x4, v1: i32x4, v2: i32x4):
-    v3 = vselect v0, v1, v2
-    return v3
-}
-; run: %vselect_p_i32x4([-1 0 -1 -1], [1 2 3 4], [100000 200000 300000 400000]) == [1 200000 3 4]
-
-function %vselect_p_i64x2(i64x2, i64x2, i64x2) -> i64x2 {
-block0(v0: i64x2, v1: i64x2, v2: i64x2):
-    v3 = vselect v0, v1, v2
-    return v3
-}
-; run: %vselect_p_i64x2([-1 0], [1 2], [100000000000 200000000000]) == [1 200000000000]