[llvm] [RISCV] Remove vmv.s.x and vmv.x.s lmul pseudo variants (PR #71501)

[Michael Maitland via llvm-commits]

================
@@ -8,7 +8,7 @@
 define signext i8 @extractelt_nxv1i8_0(<vscale x 1 x i8> %v) {
 ; CHECK-LABEL: extractelt_nxv1i8_0:
 ; CHECK:       # %bb.0:
-; CHECK-NEXT:    vsetivli zero, 1, e8, mf8, ta, ma
+; CHECK-NEXT:    vsetivli zero, 1, e8, m1, ta, ma
----------------
michaelmaitland wrote:

Is there agreement that m1 compared to mf8 is not a regression? Some microarchitectures have same behavior for M1 and fractional LMUL. I don't know whether there is any microarchitectures that can optimize on the fractional LMUL cases.

https://github.com/llvm/llvm-project/pull/71501