Merge branch 'main' into apply_simd

Quafadas · Nov 15, 2024 · 5324782 · 5324782
2 parents b5c2b34 + 70eead5
commit 5324782
Show file tree

Hide file tree

Showing 84 changed files with 4,157 additions and 865 deletions.
diff --git a/.github/FUNDING.yml b/.github/FUNDING.yml
@@ -0,0 +1,4 @@
+# These are supported funding model platforms
+
+github: [quafadas]
+
diff --git a/.github/workflows/benchmark.yml b/.github/workflows/benchmark.yml
@@ -15,10 +15,9 @@ jobs:
       - uses: coursier/setup-action@main
         with:
           jvm: temurin@17
-          apps: mill
       # Run benchmark with `go test -bench` and stores the output to a file
       - name: Run benchmark
-        run:  mill benchmark.runJmh -jvmArgs --add-modules=jdk.incubator.vector -rf json
+        run:  ./millw benchmark.runJmh -jvmArgs --add-modules=jdk.incubator.vector -rf json
 
       - name: Set variables
         run: |

diff --git a/.github/workflows/ci.yml b/.github/workflows/ci.yml
@@ -42,7 +42,7 @@ jobs:
       - uses: coursier/setup-action@main
         with:
           jvm: temurin@17
-          apps: mill scala-cli
+          apps: scala-cli
 
       - name: Setup NodeJS v18 LTS
         if: matrix.project == 'js'
@@ -51,7 +51,7 @@ jobs:
           node-version: 18
           cache: npm
       - name: formatCheck
-        run: mill mill.scalalib.scalafmt.ScalafmtModule/checkFormatAll __.sources
+        run: ./millw mill.scalalib.scalafmt.ScalafmtModule/checkFormatAll __.sources
 
       - if: matrix.project == 'js'
         run: npm install
@@ -61,14 +61,14 @@ jobs:
 
       - name: scalaJSLink
         if: matrix.project == 'js'
-        run: mill vecxt.${{ matrix.project }}.fastLinkJS
+        run: ./millw vecxt.${{ matrix.project }}.fastLinkJS
 
       - name: nativeLink
         if: matrix.project == 'native'
-        run: mill vecxt.${{ matrix.project }}.test.nativeLink
+        run: ./millw vecxt.${{ matrix.project }}.test.nativeLink
 
       - name: Test
-        run: mill vecxt.${{ matrix.project }}.test
+        run: ./millw vecxt.${{ matrix.project }}.test
 
   publish:
     if: github.repository == 'Quafadas/vecxt' && contains(github.ref, 'refs/tags/')
@@ -83,7 +83,7 @@ jobs:
       - uses: coursier/setup-action@main
         with:
           jvm: temurin@17
-          apps: mill scala-cli
+          apps: scala-cli
 
       - name: Setup GPG secrets
         run: |
@@ -92,7 +92,7 @@ jobs:
           gpg --list-secret-keys --keyid-format LONG
 
       - name: Publish to Maven Central
-        run: mill -i mill.scalalib.PublishModule/publishAll __.publishArtifacts --sonatypeUri https://s01.oss.sonatype.org/service/local --sonatypeCreds "${{ secrets.SONATYPE_USERNAME }}:${{ secrets.SONATYPE_PASSWORD }}" --gpgArgs "--passphrase=${{ secrets.PGP_PASSPHRASE}},--batch,--yes,-a,-b,--pinentry-mode,loopback" --readTimeout 1200000 --awaitTimeout 1200000 --release true --signed true
+        run: ./millw -i mill.scalalib.PublishModule/publishAll __.publishArtifacts --sonatypeUri https://s01.oss.sonatype.org/service/local --sonatypeCreds "${{ secrets.SONATYPE_USERNAME }}:${{ secrets.SONATYPE_PASSWORD }}" --gpgArgs "--passphrase=${{ secrets.PGP_PASSPHRASE}},--batch,--yes,-a,-b,--pinentry-mode,loopback" --readTimeout 1200000 --awaitTimeout 1200000 --release true --signed true
 
   site:
     if: github.event_name != 'pull_request' && github.ref == 'refs/heads/main'
@@ -102,7 +102,7 @@ jobs:
       - uses: coursier/setup-action@main
         with:
           jvm: temurin@17
-          apps: mill scala-cli
+          apps: scala-cli
       - uses: actions/checkout@v4
         with:
           fetch-depth: 0
@@ -115,18 +115,18 @@ jobs:
             ./aggregate.sh
 
       - name: Switch to main branch
-        run: | 
+        run: |
           git fetch origin main
           git checkout -f main
           git reset --hard origin/main
           ls
 
       - name: Copy benchmark results into docs generation
-        run: mkdir -p site/docs/_assets/benchmarks && cp benchmark_history.json site/docs/_assets/benchmarks/benchmark_history.json    
+        run: mkdir -p site/docs/_assets/benchmarks && cp benchmark_history.json site/docs/_assets/benchmarks/benchmark_history.json
 
       - name: Generate static site
-        run: mill site.publishDocs          
-        
+        run: ./millw site.publishDocs
+
       - name: Setup Pages
         uses: actions/configure-pages@v4
       - uses: actions/upload-artifact@v3

diff --git a/.mill-jvm-opts b/.mill-jvm-opts
@@ -0,0 +1 @@
+--add-modules=jdk.incubator.vector
diff --git a/.mill-version b/.mill-version
@@ -1 +1 @@
-0.11.12
+0.12.2
diff --git a/.scalafmt.conf b/.scalafmt.conf
@@ -11,4 +11,4 @@ rewrite.scala3.countEndMarkerLines = lastBlockOnly
 rewrite.scala3.insertEndMarkerMinLines = 1
 indent.main = 2
 maxColumn = 120
-project.excludeFilters = [ ".*/build\\.sc"]
+project.excludeFilters = [ ".*/build\\.mill"]
diff --git a/benchmark/src/Benchmark.scala b/benchmark/src/Benchmark.scala
@@ -11,7 +11,7 @@ import java.util.concurrent.TimeUnit;
 @OutputTimeUnit(TimeUnit.SECONDS)
 @State(Scope.Thread)
 @Fork(value = 1)
-@Warmup(iterations = 3)
+@Warmup(iterations = 1)
 @Measurement(iterations = 3)
 abstract class BLASBenchmark:
 

diff --git a/benchmark/src/and.scala b/benchmark/src/and.scala
@@ -30,7 +30,7 @@ import org.openjdk.jmh.infra.Blackhole
 // import vecxt.Matrix.*
 import vecxt.BoundsCheck
 import scala.compiletime.uninitialized
-import vecxt.*
+import vecxt.all.*
 import jdk.incubator.vector.VectorSpecies
 import jdk.incubator.vector.VectorOperators
 import jdk.incubator.vector.DoubleVector

diff --git a/benchmark/src/countTrue.scala b/benchmark/src/countTrue.scala
@@ -30,7 +30,7 @@ import org.openjdk.jmh.infra.Blackhole
 // import vecxt.Matrix.*
 import vecxt.BoundsCheck
 import scala.compiletime.uninitialized
-import vecxt.*
+import vecxt.all.*
 import jdk.incubator.vector.VectorSpecies
 import jdk.incubator.vector.VectorOperators
 import jdk.incubator.vector.DoubleVector

diff --git a/benchmark/src/increments.scala b/benchmark/src/increments.scala
@@ -30,7 +30,7 @@ import org.openjdk.jmh.infra.Blackhole
 // import vecxt.Matrix.*
 import vecxt.BoundsCheck
 import scala.compiletime.uninitialized
-import vecxt.*
+import vecxt.all.*
 import jdk.incubator.vector.VectorSpecies
 import jdk.incubator.vector.VectorOperators
 import jdk.incubator.vector.DoubleVector

diff --git a/benchmark/src/logical.scala b/benchmark/src/logical.scala
@@ -30,7 +30,7 @@ import org.openjdk.jmh.infra.Blackhole
 // import vecxt.Matrix.*
 import vecxt.BoundsCheck
 import scala.compiletime.uninitialized
-import vecxt.*
+import vecxt.all.*
 import jdk.incubator.vector.VectorSpecies
 import jdk.incubator.vector.VectorOperators
 import jdk.incubator.vector.DoubleVector
@@ -42,12 +42,12 @@ class LogicalBenchmark extends BLASBenchmark:
   var len: String = uninitialized;
 
   var arr: Array[Double] = uninitialized
-  
+
 
   // format: off
   @Setup(Level.Trial)
   def setup: Unit =
-    arr = randomDoubleArray(len.toInt);    
+    arr = randomDoubleArray(len.toInt);
     ()
   end setup
 
@@ -68,14 +68,14 @@ class LogicalBenchmark extends BLASBenchmark:
     val r = arr <= 4.0
     bh.consume(r);
   end lte_vec
-  
-  
+
+
 
   @Benchmark
   def lte_loop(bh: Blackhole) =
     val r = arr.lte2(4.0)
     bh.consume(r);
   end lte_loop
-  
+
 end LogicalBenchmark
-  
+
diff --git a/benchmark/src/matmul.scala b/benchmark/src/matmul.scala
@@ -27,10 +27,11 @@ package vecxt.benchmark
 
 import org.openjdk.jmh.annotations.*
 import org.openjdk.jmh.infra.Blackhole
-import vecxt.Matrix.*
+import vecxt.all.*
 import vecxt.BoundsCheck
 import scala.compiletime.uninitialized
 import BoundsCheck.DoBoundsCheck.no
+
 @State(Scope.Thread)
 class DgemmBenchmark extends BLASBenchmark:
 

diff --git a/benchmark/src/or.scala b/benchmark/src/or.scala
@@ -30,7 +30,7 @@ import org.openjdk.jmh.infra.Blackhole
 // import vecxt.Matrix.*
 import vecxt.BoundsCheck
 import scala.compiletime.uninitialized
-import vecxt.*
+import vecxt.all.*
 import jdk.incubator.vector.VectorSpecies
 import jdk.incubator.vector.VectorOperators
 import jdk.incubator.vector.DoubleVector

diff --git a/benchmark/src/sum.scala b/benchmark/src/sum.scala
@@ -30,7 +30,7 @@ import org.openjdk.jmh.infra.Blackhole
 // import vecxt.Matrix.*
 import vecxt.BoundsCheck
 import scala.compiletime.uninitialized
-import vecxt.*
+import vecxt.all.*
 import jdk.incubator.vector.VectorSpecies
 import jdk.incubator.vector.VectorOperators
 import jdk.incubator.vector.DoubleVector
@@ -52,17 +52,17 @@ class SumBenchmark extends BLASBenchmark:
 
   end setup
 
+  final val spd = DoubleVector.SPECIES_PREFERRED
   extension (vec: Array[Double])
 
 
     inline def sum2 =
       var sum: Double = 0.0
       var i: Int = 0
-      val sp = Matrix.Matrix.doubleSpecies
-      val l = sp.length()
+      val l = spd.length()
 
-      while i < sp.loopBound(vec.length) do
-        sum = sum + DoubleVector.fromArray(sp, vec, i).reduceLanes(VectorOperators.ADD)
+      while i < spd.loopBound(vec.length) do
+        sum = sum + DoubleVector.fromArray(spd, vec, i).reduceLanes(VectorOperators.ADD)
         i += l
       end while
       while i < vec.length do

diff --git a/benchmark/src/vectorAddition.scala b/benchmark/src/vectorAddition.scala
@@ -27,8 +27,8 @@ package vecxt.benchmark
 
 import org.openjdk.jmh.annotations.*
 import org.openjdk.jmh.infra.Blackhole
-import vecxt.Matrix.*
-import vecxt.*
+import jdk.incubator.vector.DoubleVector
+import vecxt.all.*
 import scala.compiletime.uninitialized
 
 // format: off
@@ -61,26 +61,34 @@ class AddScalarBenchmark extends BLASBenchmark:
   end setup
 
   extension (vec: Array[Double])
-    inline def scalarPlus(d: Double): Unit =
-      var i = 0
+    inline def scalarPlusVec(d: Double): Unit =
+      val species = DoubleVector.SPECIES_PREFERRED
+      var i: Int = 0
+      val l = species.length()
+
+      while i < species.loopBound(vec.length) do
+        DoubleVector.fromArray(species, vec, i).add(DoubleVector.broadcast(species, d)).intoArray(vec, i)
+        i += l
+      end while
+
       while i < vec.length do
         vec(i) += d
         i += 1
       end while
-    end scalarPlus
+    end scalarPlusVec
 
 
   end extension
 
   @Benchmark
   def vecxt_add(bh: Blackhole) =
-    vec.scalarPlus(4.5)
+    vec +:+= (4.5)
     bh.consume(vec);
   end vecxt_add
 
   @Benchmark
   def vecxt_add_vec(bh: Blackhole) =
-    vec2 +:+= 4.5
+    vec2.scalarPlusVec(4.5)
     bh.consume(vec2);
   end vecxt_add_vec
 end AddScalarBenchmark
Original file line number	Diff line number	Diff line change
		@@ -0,0 +1,4 @@
		# These are supported funding model platforms

		github: [quafadas]