Make vectorized store convert and perform multiple stores if required

[maleadt]

Similarly to #109, when doing a vectorized store, convert the elements so that they match the element type of the layout. In addition, if we want to store more values than what we can using a single vectorized store, perform multiple stores.

This matters when the element type of the global layout differs from the shared layout. For example, if global_layout is AlignedColMajor{Float16}, but shared_layout_a is AlignedColMajor{Float32}, we initially load 8 Float16's, but then only store 4 values into the Float32 shared-memory workspace, because the number of elements is currently always computed as 16 ÷ sizeof(T).

The above happens when doing Float16xFloat32, in which case I'm trying to use T=promote_type(Float16, Float32)=Float32, causing a type mismatch between global and shared memory. I want to take this approach (as opposed to keeping the shared layout Float16 too) because this opens up a path to using WMMA for incompatible inputs (say, for Float16xFloat32 we could then use WMMA on a GPU that has Float32xFloat32=... tensor cores).

[codecov]

Codecov Report

Patch coverage: 50.00% and project coverage change: +0.34 🎉

Comparison is base (2a6ad1d) 29.61% compared to head (8881635) 29.96%.

❗ Current head 8881635 differs from pull request most recent head 27903f6. Consider uploading reports for the commit 27903f6 to get more accurate results

Additional details and impacted files

@@            Coverage Diff             @@
##           master     #111      +/-   ##
==========================================
+ Coverage   29.61%   29.96%   +0.34%     
==========================================
  Files          11       11              
  Lines         763      761       -2     
==========================================
+ Hits          226      228       +2     
+ Misses        537      533       -4     

Impacted Files	Coverage Δ
src/layout.jl	21.60% <50.00%> (+1.91%)	⬆️

☔ View full report in Codecov by Sentry.
📢 Do you have feedback about the report comment? Let us know in this issue.

[maleadt]

Benchmark results for commit 2b25c99 (comparing to 48279bc):

ID	before	after	change

[maleadt]

The 1.6 CI failures is JuliaGPU/GPUCompiler.jl#481