x64: Add non-SSE4.1 lowerings of pmov{s,z}x* (#6279)

* x64: Add non-SSE4.1 lowerings of `pmov{s,z}x*`

This commit adds lowerings for a suite of sign/zero extension
instructions which don't require SSE4.1. Like before these lowerings are
based on LLVM's output.

This commit also deletes special casees for `i16x8.extmul_{low,high}_*`
since the output of the special case is the same as the default lowering
of all the component instructions used within as well.

* Remove SSE4.1 specialization of `uwiden_high`

LLVM prefers the `punpckh*`-based lowerings and at least according to
`llvm-mca` these are slightly better cycle-wise too.

cranelift/codegen/src/isa/x64/inst/args.rs

@@ -1116,6 +1116,7 @@ pub enum SseOpcode {
     Ucomiss,
     Ucomisd,
     Unpcklps,
+    Unpckhps,
     Xorps,
     Xorpd,
     Phaddw,
@@ -1168,6 +1169,7 @@ impl SseOpcode {
             | SseOpcode::Subss
             | SseOpcode::Ucomiss
             | SseOpcode::Unpcklps
+            | SseOpcode::Unpckhps
             | SseOpcode::Xorps => SSE,
 
             SseOpcode::Addpd
@@ -1516,6 +1518,7 @@ impl fmt::Debug for SseOpcode {
             SseOpcode::Ucomiss => "ucomiss",
             SseOpcode::Ucomisd => "ucomisd",
             SseOpcode::Unpcklps => "unpcklps",
+            SseOpcode::Unpckhps => "unpckhps",
             SseOpcode::Xorps => "xorps",
             SseOpcode::Xorpd => "xorpd",
             SseOpcode::Phaddw => "phaddw",
@@ -1611,6 +1614,7 @@ impl AvxOpcode {
             | AvxOpcode::Vpunpckhwd
             | AvxOpcode::Vpunpcklwd
             | AvxOpcode::Vunpcklps
+            | AvxOpcode::Vunpckhps
             | AvxOpcode::Vaddps
             | AvxOpcode::Vaddpd
             | AvxOpcode::Vsubps

cranelift/codegen/src/isa/x64/inst/emit.rs

@@ -2060,6 +2060,7 @@ pub(crate) fn emit(
                 SseOpcode::Subss => (LegacyPrefixes::_F3, 0x0F5C, 2),
                 SseOpcode::Subsd => (LegacyPrefixes::_F2, 0x0F5C, 2),
                 SseOpcode::Unpcklps => (LegacyPrefixes::None, 0x0F14, 2),
+                SseOpcode::Unpckhps => (LegacyPrefixes::None, 0x0F15, 2),
                 SseOpcode::Xorps => (LegacyPrefixes::None, 0x0F57, 2),
                 SseOpcode::Xorpd => (LegacyPrefixes::_66, 0x0F57, 2),
                 SseOpcode::Phaddw => (LegacyPrefixes::_66, 0x0F3801, 3),
@@ -2206,6 +2207,7 @@ pub(crate) fn emit(
                 AvxOpcode::Vpunpckhwd => (LP::_66, OM::_0F, 0x69),
                 AvxOpcode::Vpunpcklwd => (LP::_66, OM::_0F, 0x61),
                 AvxOpcode::Vunpcklps => (LP::None, OM::_0F, 0x14),
+                AvxOpcode::Vunpckhps => (LP::None, OM::_0F, 0x15),
                 AvxOpcode::Vaddps => (LP::None, OM::_0F, 0x58),
                 AvxOpcode::Vaddpd => (LP::_66, OM::_0F, 0x58),
                 AvxOpcode::Vsubps => (LP::None, OM::_0F, 0x5C),