Remove dead macOS version branches from support queries

[KaanKesginLW]

Summary

The __init__ function in initialization.jl blocks execution on macOS < 13:

if macos_version() < v"13"
    @error "Metal.jl requires macOS 13 or later"
    return
end

This makes all branches for macOS 10.13, 10.14, 10.15, 11, and 12 in metallib_support(), air_support(), and metal_support() unreachable dead code.

This PR removes ~36 lines of code that can never execute.

Changes

• metallib_support(): Removed branches for macOS 10.13-12, kept macOS 13+ (minimum returns v"1.2.7")
• air_support(): Removed branches for macOS 10.13-12, kept macOS 13+ (minimum returns v"2.5")
• metal_support(): Removed branches for macOS 10.13-12, kept macOS 13+ (minimum returns v"3.0")

[codecov]

Codecov Report

✅ All modified and coverable lines are covered by tests.
✅ Project coverage is 81.97%. Comparing base (239fa4d) to head (7a1228c).
⚠️ Report is 1 commits behind head on main.

Additional details and impacted files

@@            Coverage Diff             @@
##             main     #720      +/-   ##
==========================================
+ Coverage   80.96%   81.97%   +1.00%     
==========================================
  Files          62       62              
  Lines        2837     2801      -36     
==========================================
- Hits         2297     2296       -1     
+ Misses        540      505      -35     

☔ View full report in Codecov by Sentry.
📢 Have feedback on the report? Share it here.

🚀 New features to boost your workflow:

• ❄️ Test Analytics: Detect flaky tests, report on failures, and find test suite problems.

[github-actions]

Metal Benchmarks

Benchmark suite	Current: 7a1228c	Previous: 239fa4d	Ratio
latency/precompile	24104876875 ns	24383716541 ns	0.99
latency/ttfp	2288838792 ns	2324081375 ns	0.98
latency/import	1408815458 ns	1427504083 ns	0.99
integration/metaldevrt	837334 ns	837292 ns	1.00
integration/byval/slices=1	1562895.5 ns	1598354 ns	0.98
integration/byval/slices=3	8819708.5 ns	19021791.5 ns	0.46
integration/byval/reference	1531604 ns	1590708.5 ns	0.96
integration/byval/slices=2	2595375 ns	2727250 ns	0.95
kernel/indexing	605250 ns	459062.5 ns	1.32
kernel/indexing_checked	604292 ns	463104.5 ns	1.30
kernel/launch	12250 ns	11625 ns	1.05
kernel/rand	568583 ns	526667 ns	1.08
array/construct	6417 ns	5958 ns	1.08
array/broadcast	603167 ns	545375 ns	1.11
array/random/randn/Float32	804312.5 ns	886167 ns	0.91
array/random/randn!/Float32	603542 ns	578875 ns	1.04
array/random/rand!/Int64	542417 ns	539083 ns	1.01
array/random/rand!/Float32	579125 ns	533229.5 ns	1.09
array/random/rand/Int64	738041.5 ns	887000 ns	0.83
array/random/rand/Float32	600375 ns	840959 ns	0.71
array/accumulate/Int64/1d	1253417 ns	1292146 ns	0.97
array/accumulate/Int64/dims=1	1825187.5 ns	1865375 ns	0.98
array/accumulate/Int64/dims=2	2125104 ns	2215437 ns	0.96
array/accumulate/Int64/dims=1L	11670417 ns	12096125 ns	0.96
array/accumulate/Int64/dims=2L	9743625 ns	10003417 ns	0.97
array/accumulate/Float32/1d	1115083 ns	1086042 ns	1.03
array/accumulate/Float32/dims=1	1545125 ns	1581542 ns	0.98
array/accumulate/Float32/dims=2	1853437.5 ns	1998167 ns	0.93
array/accumulate/Float32/dims=1L	9769062.5 ns	10248396 ns	0.95
array/accumulate/Float32/dims=2L	7230604 ns	7422792 ns	0.97
array/reductions/reduce/Int64/1d	1530479 ns	1312917 ns	1.17
array/reductions/reduce/Int64/dims=1	1075166 ns	1120125 ns	0.96
array/reductions/reduce/Int64/dims=2	1303062.5 ns	1153917 ns	1.13
array/reductions/reduce/Int64/dims=1L	1994792 ns	2041417 ns	0.98
array/reductions/reduce/Int64/dims=2L	4209750 ns	3778125 ns	1.11
array/reductions/reduce/Float32/1d	1020791.5 ns	796167 ns	1.28
array/reductions/reduce/Float32/dims=1	816416.5 ns	794000 ns	1.03
array/reductions/reduce/Float32/dims=2	844250 ns	818562.5 ns	1.03
array/reductions/reduce/Float32/dims=1L	1315833 ns	1329000 ns	0.99
array/reductions/reduce/Float32/dims=2L	1785708.5 ns	1796708.5 ns	0.99
array/reductions/mapreduce/Int64/1d	1552167 ns	1298666 ns	1.20
array/reductions/mapreduce/Int64/dims=1	1081208 ns	1086313 ns	1.00
array/reductions/mapreduce/Int64/dims=2	1129041 ns	1122666 ns	1.01
array/reductions/mapreduce/Int64/dims=1L	2003041 ns	2025395.5 ns	0.99
array/reductions/mapreduce/Int64/dims=2L	3610625 ns	3647583 ns	0.99
array/reductions/mapreduce/Float32/1d	1062375 ns	774083.5 ns	1.37
array/reductions/mapreduce/Float32/dims=1	814291 ns	791417 ns	1.03
array/reductions/mapreduce/Float32/dims=2	842625 ns	826542 ns	1.02
array/reductions/mapreduce/Float32/dims=1L	1308875 ns	1322667 ns	0.99
array/reductions/mapreduce/Float32/dims=2L	1806625 ns	1817916.5 ns	0.99
array/private/copyto!/gpu_to_gpu	641917 ns	533917 ns	1.20
array/private/copyto!/cpu_to_gpu	799458 ns	690271 ns	1.16
array/private/copyto!/gpu_to_cpu	408541.5 ns	668542 ns	0.61
array/private/iteration/findall/int	1583896 ns	1565687.5 ns	1.01
array/private/iteration/findall/bool	1436125 ns	1465333.5 ns	0.98
array/private/iteration/findfirst/int	2051125 ns	2079042 ns	0.99
array/private/iteration/findfirst/bool	2033833 ns	2020083 ns	1.01
array/private/iteration/scalar	5309166.5 ns	2787125 ns	1.90
array/private/iteration/logical	2610541.5 ns	2599208 ns	1.00
array/private/iteration/findmin/1d	2233270.5 ns	2265458 ns	0.99
array/private/iteration/findmin/2d	1516666 ns	1528791 ns	0.99
array/private/copy	557708 ns	847041.5 ns	0.66
array/shared/copyto!/gpu_to_gpu	84375 ns	84333 ns	1.00
array/shared/copyto!/cpu_to_gpu	92917 ns	83042 ns	1.12
array/shared/copyto!/gpu_to_cpu	84667 ns	83479.5 ns	1.01
array/shared/iteration/findall/int	1573417 ns	1558208 ns	1.01
array/shared/iteration/findall/bool	1438520.5 ns	1470708 ns	0.98
array/shared/iteration/findfirst/int	1637542 ns	1682792 ns	0.97
array/shared/iteration/findfirst/bool	1619417 ns	1644334 ns	0.98
array/shared/iteration/scalar	209958 ns	202000 ns	1.04
array/shared/iteration/logical	2273042 ns	2368458 ns	0.96
array/shared/iteration/findmin/1d	1824959 ns	1845542 ns	0.99
array/shared/iteration/findmin/2d	1504833.5 ns	1521583 ns	0.99
array/shared/copy	242834 ns	210959 ns	1.15
array/permutedims/4d	2363062.5 ns	2473375 ns	0.96
array/permutedims/2d	1154896 ns	1178666.5 ns	0.98
array/permutedims/3d	1660000 ns	1780750 ns	0.93
metal/synchronization/stream	19292 ns	19334 ns	1.00
metal/synchronization/context	19250 ns	20000 ns	0.96

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