Reinterpret between different LMUL under the same SEW

[riscv-non-isa]

Reinterpret between different LMUL under the same SEW

[zakk0610]

vint8m2_t and vint8m1_t occupy the different size, I think it should not be called reinterpret.
Could you share the user scenario?
Are you doing vector-length agnostic programming?
Or are you going to do the similar thing like #28 mention, implement combine operation for vcombine instruction in NEON?

[ChunyuLiao]

How to convert between data types and mask types use rvv intrinsics?

Eg: <vscale x 1 x i64> <-> <vscale x 1 x i1>

[zakk0610]

mask->data type:
vuint64m1_t data = vmerge_vxm_u64m1(mask, vundefined_u64m1(), 1);

data type->mask:
vbool64_t mask = vmseq_vx_u64m1(data, 1);

[eopXD]

We have vlmul_ext, vlmul_trunc now. Closing the issue.