Added XgemvFastRot and Xgemm 16-bit tuning results: just defaults which are now automatically taken from 32-bit if there are no entries at all

4 files changed, 28 insertions, 4 deletions

diff --git a/src/database/database.cpp b/src/database/database.cpp
index 38974b95..34c44a29 100644
--- a/src/database/database.cpp
+++ b/src/database/database.cpp
@@ -35,9 +35,9 @@ const std::vector<Database::DatabaseEntry> Database::database = {
   XdotHalf, XdotSingle, XdotDouble, XdotComplexSingle, XdotComplexDouble,
   XgemvHalf, XgemvSingle, XgemvDouble, XgemvComplexSingle, XgemvComplexDouble,
   XgemvFastHalf, XgemvFastSingle, XgemvFastDouble, XgemvFastComplexSingle, XgemvFastComplexDouble,
-  /* XgemvFastRotHalf, */ XgemvFastRotSingle, XgemvFastRotDouble, XgemvFastRotComplexSingle, XgemvFastRotComplexDouble,
+  XgemvFastRotHalf, XgemvFastRotSingle, XgemvFastRotDouble, XgemvFastRotComplexSingle, XgemvFastRotComplexDouble,
   XgerHalf, XgerSingle, XgerDouble, XgerComplexSingle, XgerComplexDouble,
-  /* XgemmHalf, */ XgemmSingle, XgemmDouble, XgemmComplexSingle, XgemmComplexDouble,
+  XgemmHalf, XgemmSingle, XgemmDouble, XgemmComplexSingle, XgemmComplexDouble,
   CopyHalf, CopySingle, CopyDouble, CopyComplexSingle, CopyComplexDouble,
   PadHalf, PadSingle, PadDouble, PadComplexSingle, PadComplexDouble,
   TransposeHalf, TransposeSingle, TransposeDouble, TransposeComplexSingle, TransposeComplexDouble,
diff --git a/src/database/database.hpp b/src/database/database.hpp
index 8d6d3863..a6ab49c5 100644
--- a/src/database/database.hpp
+++ b/src/database/database.hpp
@@ -72,9 +72,9 @@ class Database {
   static const DatabaseEntry XdotHalf, XdotSingle, XdotDouble, XdotComplexSingle, XdotComplexDouble;
   static const DatabaseEntry XgemvHalf, XgemvSingle, XgemvDouble, XgemvComplexSingle, XgemvComplexDouble;
   static const DatabaseEntry XgemvFastHalf, XgemvFastSingle, XgemvFastDouble, XgemvFastComplexSingle, XgemvFastComplexDouble;
-  static const DatabaseEntry /* XgemvFastRotHalf, */ XgemvFastRotSingle, XgemvFastRotDouble, XgemvFastRotComplexSingle, XgemvFastRotComplexDouble;
+  static const DatabaseEntry XgemvFastRotHalf, XgemvFastRotSingle, XgemvFastRotDouble, XgemvFastRotComplexSingle, XgemvFastRotComplexDouble;
   static const DatabaseEntry XgerHalf, XgerSingle, XgerDouble, XgerComplexSingle, XgerComplexDouble;
-  static const DatabaseEntry /* XgemmHalf, */ XgemmSingle, XgemmDouble, XgemmComplexSingle, XgemmComplexDouble;
+  static const DatabaseEntry XgemmHalf, XgemmSingle, XgemmDouble, XgemmComplexSingle, XgemmComplexDouble;
   static const DatabaseEntry CopyHalf, CopySingle, CopyDouble, CopyComplexSingle, CopyComplexDouble;
   static const DatabaseEntry PadHalf, PadSingle, PadDouble, PadComplexSingle, PadComplexDouble;
   static const DatabaseEntry TransposeHalf, TransposeSingle, TransposeDouble, TransposeComplexSingle, TransposeComplexDouble;
diff --git a/src/database/kernels/xgemm.hpp b/src/database/kernels/xgemm.hpp
index 7e793076..d19c55b5 100644
--- a/src/database/kernels/xgemm.hpp
+++ b/src/database/kernels/xgemm.hpp
@@ -14,6 +14,18 @@
 namespace clblast {
 // =================================================================================================
 
+const Database::DatabaseEntry Database::XgemmHalf = {
+  "Xgemm", Precision::kHalf, {
+    { // Default
+      kDeviceTypeAll, "default", {
+        { "default",                                         { {"KWG",16}, {"KWI",2}, {"MDIMA",8}, {"MDIMC",8}, {"MWG",32}, {"NDIMB",8}, {"NDIMC",8}, {"NWG",16}, {"SA",0}, {"SB",0}, {"STRM",0}, {"STRN",0}, {"VWM",1}, {"VWN",1} } },
+      }
+    },
+  }
+};
+
+// =================================================================================================
+
 const Database::DatabaseEntry Database::XgemmSingle = {
   "Xgemm", Precision::kSingle, {
     { // AMD GPUs
diff --git a/src/database/kernels/xgemv_fast_rot.hpp b/src/database/kernels/xgemv_fast_rot.hpp
index 3d2e0d3a..328094e1 100644
--- a/src/database/kernels/xgemv_fast_rot.hpp
+++ b/src/database/kernels/xgemv_fast_rot.hpp
@@ -14,6 +14,18 @@
 namespace clblast {
 // =================================================================================================
 
+const Database::DatabaseEntry Database::XgemvFastRotHalf = {
+  "XgemvFastRot", Precision::kHalf, {
+    { // Default
+      kDeviceTypeAll, "default", {
+        { "default",                                         { {"VW3",8}, {"WGS3",32}, {"WPT3",32} } },
+      }
+    },
+  }
+};
+
+// =================================================================================================
+
 const Database::DatabaseEntry Database::XgemvFastRotSingle = {
   "XgemvFastRot", Precision::kSingle, {
     { // AMD GPUs