Replace CraneStation by bytecodealliance everywhere; (#1221)

cranelift/codegen/meta/src/isa/x86/encodings.rs

@@ -1879,7 +1879,7 @@ pub(crate) fn define(
 
     // SIMD register movement: store, load, spill, fill, regmove. All of these use encodings of
     // MOVUPS and MOVAPS from SSE (TODO ideally all of these would either use MOVAPS when we have
-    // alignment or type-specific encodings, see https://github.com/CraneStation/cranelift/issues/1039).
+    // alignment or type-specific encodings, see https://github.com/bytecodealliance/cranelift/issues/1039).
     for ty in ValueType::all_lane_types().filter(allowed_simd_type) {
         // Store
         let bound_store = store.bind(vector(ty, sse_vector_size)).bind(Any);
@@ -2003,7 +2003,7 @@ pub(crate) fn define(
     // allows SIMD shifts to be legalized more easily. TODO ideally this would be typed as an
     // I128x1 but restrictions on the type builder prevent this; the general idea here is that
     // the upper bits are all zeroed and do not form parts of any separate lane. See
-    // https://github.com/CraneStation/cranelift/issues/1146.
+    // https://github.com/bytecodealliance/cranelift/issues/1146.
     e.enc_both(
         bitcast.bind(vector(I64, sse_vector_size)).bind(I32),
         rec_frurm.opcodes(&MOVD_LOAD_XMM),

cranelift/codegen/meta/src/isa/x86/legalize.rs

@@ -349,7 +349,7 @@ pub(crate) fn define(shared: &mut SharedDefinitions, x86_instructions: &Instruct
             def!(y = splat_any8x16(x)),
             vec![
                 def!(a = scalar_to_vector(x)), // move into the lowest 8 bits of an XMM register
-                // TODO replace the following two instructions with `vconst(0)` when this is possible; see https://github.com/CraneStation/cranelift/issues/1052
+                // TODO replace the following two instructions with `vconst(0)` when this is possible; see https://github.com/bytecodealliance/cranelift/issues/1052
                 def!(b = f64const(ieee64_zero)), // zero out a different XMM register; the shuffle mask for moving the lowest byte to all other byte lanes is 0x0
                 def!(c = bitcast_f64_to_any8x16(b)), // no instruction emitted; informs the SSA that the 0 in b can be used as a vector of this type
                 def!(y = x86_pshufb(a, c)), // PSHUFB takes two XMM operands, one of which is a shuffle mask (i.e. b)