lima/ppir: enable vectorize optimization

pp has vector units and some operations can be optimized when bundled
together.
Benchmarking this with piglit shaders shows that the instruction count
can be greatly reduced on many examples with vectorize.

Signed-off-by: Erico Nunes <nunes.erico@gmail.com>
Reviewed-by: Qiang Yu <yuq825@gmail.com>
22 jobs for !1639 with lima-ppir-vectorize in 24 minutes and 11 seconds (queued for 10 seconds)
latest detached
Status Name Job ID Coverage
  Containers Build
passed debian #532691

00:00:21

 
  Build+Test
passed meson-arm64 #532700

00:02:22

passed meson-armhf #532699

00:02:25

passed meson-clang #532693

00:10:03

passed meson-clover #532697

00:11:09

passed meson-i386 #532701

00:02:48

passed meson-main #532696

00:09:21

passed meson-swr-glvnd #532692

00:06:07

passed meson-vulkan #532698

00:03:56

passed scons-llvm #532703

00:04:51

passed scons-nollvm #532702

00:03:43

passed scons-swr #532694

00:08:19

passed scons-win64 #532695

00:08:10

 
  Test
passed test-llvmpipe-gles2 1/4 #532704

00:04:13

passed test-llvmpipe-gles2 2/4 #532705

00:04:37

passed test-llvmpipe-gles2 3/4 #532706

00:04:03

passed test-llvmpipe-gles2 4/4 #532707

00:02:29

passed test-softpipe-gles2 1/4 #532708

00:01:28

passed test-softpipe-gles2 2/4 #532709

00:02:22

passed test-softpipe-gles2 3/4 #532710

00:02:10

passed test-softpipe-gles2 4/4 #532711

00:01:33

passed test-softpipe-gles3-limited #532712

00:04:01