Fix Float16 sincos intrinsic

[christiangnrd]

Also same treatment to the Float32 intrinsics since I don't know why it used to work.

Closes #530

[github-actions]

Your PR requires formatting changes to meet the project's style guidelines.
Please consider running Runic (git runic main) to apply these changes.

Click here to view the suggested changes.

diff --git a/test/device/intrinsics.jl b/test/device/intrinsics.jl
index b71f62b1..4c88fb67 100644
--- a/test/device/intrinsics.jl
+++ b/test/device/intrinsics.jl
@@ -257,9 +257,9 @@ end
             return nothing
         end
 
-        Metal.@sync @metal threads = N intr_test3(bufferA, bufferB)
-        @test Array(bufferA) ≈ sin.(arr)
-        @test Array(bufferB) ≈ cos.(arr)
+            Metal.@sync @metal threads = N intr_test3(bufferA, bufferB)
+            @test Array(bufferA) ≈ sin.(arr)
+            @test Array(bufferB) ≈ cos.(arr)
     end
 
     let # clamp

[github-actions]

Metal Benchmarks

Benchmark suite	Current: 1981c5f	Previous: 924a130	Ratio
private array/construct	24989.583333333336 ns	25257 ns	0.99
private array/broadcast	463084 ns	463958 ns	1.00
private array/random/randn/Float32	814208.5 ns	760208 ns	1.07
private array/random/randn!/Float32	618500 ns	624687.5 ns	0.99
private array/random/rand!/Int64	558250 ns	563395.5 ns	0.99
private array/random/rand!/Float32	585541 ns	587854 ns	1.00
private array/random/rand/Int64	785292 ns	755083 ns	1.04
private array/random/rand/Float32	628667 ns	611333 ns	1.03
private array/copyto!/gpu_to_gpu	644250 ns	650916.5 ns	0.99
private array/copyto!/cpu_to_gpu	612062 ns	820417 ns	0.75
private array/copyto!/gpu_to_cpu	813583.5 ns	686687.5 ns	1.18
private array/accumulate/1d	1355104 ns	1362854.5 ns	0.99
private array/accumulate/2d	1391229 ns	1404375 ns	0.99
private array/iteration/findall/int	2101854.5 ns	2111541.5 ns	1.00
private array/iteration/findall/bool	1808708 ns	1849958 ns	0.98
private array/iteration/findfirst/int	1698167 ns	1715916 ns	0.99
private array/iteration/findfirst/bool	1664708 ns	1673542 ns	0.99
private array/iteration/scalar	3897542 ns	3935958 ns	0.99
private array/iteration/logical	3179583.5 ns	3227791 ns	0.99
private array/iteration/findmin/1d	1755396 ns	1778916.5 ns	0.99
private array/iteration/findmin/2d	1347875 ns	1352041 ns	1.00
private array/reductions/reduce/1d	1037291.5 ns	1043458 ns	0.99
private array/reductions/reduce/2d	663541 ns	669709 ns	0.99
private array/reductions/mapreduce/1d	1036208 ns	1052375 ns	0.98
private array/reductions/mapreduce/2d	663292 ns	675854.5 ns	0.98
private array/permutedims/4d	2510521 ns	2557249.5 ns	0.98
private array/permutedims/2d	1026000 ns	1026042 ns	1.00
private array/permutedims/3d	1577041.5 ns	1594916 ns	0.99
private array/copy	563875 ns	569146 ns	0.99
latency/precompile	8863922500 ns	8851374292 ns	1.00
latency/ttfp	3613381417 ns	3620472000 ns	1.00
latency/import	1231948875 ns	1235528208 ns	1.00
integration/metaldevrt	717292 ns	717959 ns	1.00
integration/byval/slices=1	1643125 ns	1539333 ns	1.07
integration/byval/slices=3	10184375 ns	9117125 ns	1.12
integration/byval/reference	1535499.5 ns	1560292 ns	0.98
integration/byval/slices=2	2646437 ns	2731458.5 ns	0.97
kernel/indexing	483792 ns	457750 ns	1.06
kernel/indexing_checked	474125 ns	455833 ns	1.04
kernel/launch	8125 ns	8166 ns	0.99
metal/synchronization/stream	14666.5 ns	14708 ns	1.00
metal/synchronization/context	15166 ns	14791 ns	1.03
shared array/construct	24305.5 ns	25149.333333333332 ns	0.97
shared array/broadcast	459167 ns	464959 ns	0.99
shared array/random/randn/Float32	812125.5 ns	835792 ns	0.97
shared array/random/randn!/Float32	617583 ns	635500 ns	0.97
shared array/random/rand!/Int64	559458 ns	560542 ns	1.00
shared array/random/rand!/Float32	580709 ns	598750 ns	0.97
shared array/random/rand/Int64	761250 ns	765583 ns	0.99
shared array/random/rand/Float32	635708 ns	604979 ns	1.05
shared array/copyto!/gpu_to_gpu	84458 ns	79167 ns	1.07
shared array/copyto!/cpu_to_gpu	84042 ns	82250 ns	1.02
shared array/copyto!/gpu_to_cpu	83875 ns	83000 ns	1.01
shared array/accumulate/1d	1344584 ns	1366937.5 ns	0.98
shared array/accumulate/2d	1391083.5 ns	1400458 ns	0.99
shared array/iteration/findall/int	1781042 ns	1854249.5 ns	0.96
shared array/iteration/findall/bool	1584417 ns	1619625 ns	0.98
shared array/iteration/findfirst/int	1393417 ns	1384771 ns	1.01
shared array/iteration/findfirst/bool	1365979 ns	1375770.5 ns	0.99
shared array/iteration/scalar	157583 ns	155000 ns	1.02
shared array/iteration/logical	2972542 ns	3012021 ns	0.99
shared array/iteration/findmin/1d	1471542 ns	1470417 ns	1.00
shared array/iteration/findmin/2d	1372791 ns	1355875 ns	1.01
shared array/reductions/reduce/1d	726000 ns	729437.5 ns	1.00
shared array/reductions/reduce/2d	668083 ns	674104.5 ns	0.99
shared array/reductions/mapreduce/1d	739584 ns	743833 ns	0.99
shared array/reductions/mapreduce/2d	685542 ns	658375 ns	1.04
shared array/permutedims/4d	2523750 ns	2575708 ns	0.98
shared array/permutedims/2d	1019708.5 ns	1011875 ns	1.01
shared array/permutedims/3d	1579416 ns	1573854 ns	1.00
shared array/copy	244334 ns	246417 ns	0.99

This comment was automatically generated by workflow using github-action-benchmark.