arm: Auto-vectorization for MVE: add pack/unpack patterns

This patch adds vec_unpack<US>_hi_<mode>, vec_unpack<US>_lo_<mode>,
vec_pack_trunc_<mode> patterns for MVE.

It does so by moving the unpack patterns from neon.md to
vec-common.md, while adding them support for MVE. The pack expander is
derived from the Neon one (which in turn is renamed into
neon_quad_vec_pack_trunc_<mode>).

The patch introduces mve_vec_unpack<US>_lo_<mode> and
mve_vec_unpack<US>_hi_<mode> which are similar to their Neon
counterparts, except for the assembly syntax.

The patch introduces mve_vec_pack_trunc_lo_<mode> to avoid the need for a
zero-initialized temporary, which is needed if the
vec_pack_trunc_<mode> expander calls @mve_vmovn[bt]q_<supf><mode>
instead.

With this patch, we can now vectorize the 16 and 8-bit versions of
vclz and vshl, although the generated code could still be improved.
For test_clz_s16, we now generate
        vldrh.16        q3, [r1]
        vmovlb.s16   q2, q3
        vmovlt.s16   q3, q3
        vclz.i32  q2, q2
        vclz.i32  q3, q3
        vmovnb.i32      q1, q2
        vmovnt.i32      q1, q3
        vstrh.16        q1, [r0]
which could be improved to
        vldrh.16        q3, [r1]
	vclz.i16	q1, q3
        vstrh.16        q1, [r0]
if we could avoid the need for unpack/pack steps.

For reference, clang-12 generates:
	vldrh.s32       q0, [r1]
	vldrh.s32       q1, [r1, #8]
	vclz.i32        q0, q0
	vstrh.32        q0, [r0]
	vclz.i32        q0, q1
	vstrh.32        q0, [r0, #8]

2021-06-11  Christophe Lyon  <christophe.lyon@linaro.org>

	gcc/
	* config/arm/mve.md (mve_vec_unpack<US>_lo_<mode>): New pattern.
	(mve_vec_unpack<US>_hi_<mode>): New pattern.
	(@mve_vec_pack_trunc_lo_<mode>): New pattern.
	(mve_vmovntq_<supf><mode>): Prefix with '@'.
	* config/arm/neon.md (vec_unpack<US>_hi_<mode>): Move to
	vec-common.md.
	(vec_unpack<US>_lo_<mode>): Likewise.
	(vec_pack_trunc_<mode>): Rename to
	neon_quad_vec_pack_trunc_<mode>.
	* config/arm/vec-common.md (vec_unpack<US>_hi_<mode>): New
	pattern.
	(vec_unpack<US>_lo_<mode>): New.
	(vec_pack_trunc_<mode>): New.

	gcc/testsuite/
	* gcc.target/arm/simd/mve-vclz.c: Update expected results.
	* gcc.target/arm/simd/mve-vshl.c: Likewise.
	* gcc.target/arm/simd/mve-vec-pack.c: New test.
	* gcc.target/arm/simd/mve-vec-unpack.c: New test.

0 files changed, 0 insertions, 0 deletions