Disabled Intel subgroup shuffling for double-precision

2 files changed, 3 insertions, 1 deletions

diff --git a/CHANGELOG b/CHANGELOG
index f2960fde..27860c85 100644
--- a/CHANGELOG
+++ b/CHANGELOG
@@ -5,6 +5,7 @@ Development (next version)
 - The tuners now check beforehand on invalid local thread sizes and skip those completely
 - Fixed an issue with conjugate transpose not being executed in certain cases for a.o. XOMATCOPY
 - Fixed an issue with AMD GPUs and the new GEMMK == 1 kernel
+- Fixed an issue with the preprocessor and the new GEMMK == 1 kernel
 - Various minor fixes and enhancements
 
 Version 1.4.1
diff --git a/src/utilities/compile.cpp b/src/utilities/compile.cpp
index 835f54b4..00cb90cb 100644
--- a/src/utilities/compile.cpp
+++ b/src/utilities/compile.cpp
@@ -59,7 +59,8 @@ std::shared_ptr<Program> CompileFromSource(
   }
 
   // For Intel GPUs with subgroup support, use subgroup shuffling.
-  if (device.IsGPU() && device.HasExtension(kKhronosIntelSubgroups)) {
+  if (device.IsGPU() && device.HasExtension(kKhronosIntelSubgroups) &&
+      (precision == Precision::kSingle || precision == Precision::kHalf)) {
     header_string += "#define USE_SUBGROUP_SHUFFLING 1\n";
     header_string += "#define SUBGROUP_SHUFFLING_INTEL 1\n";
   }