From 643dad66ee4b32a8850ca0b9e577d3af8012ad8f Mon Sep 17 00:00:00 2001
From: Gilbert Lee <gilbert.lee@amd.com>
Date: Sat, 17 Jan 2026 17:41:16 -0600
Subject: [PATCH] Fixing some Makefile issues, Null memory parsing

---
 CHANGELOG.md                 |  8 ++++++++
 CMakeLists.txt               |  2 +-
 Makefile                     | 35 ++++++++++++++++++++---------------
 src/client/EnvVars.hpp       |  2 +-
 src/header/TransferBench.hpp |  7 ++++++-
 5 files changed, 36 insertions(+), 18 deletions(-)

diff --git a/CHANGELOG.md b/CHANGELOG.md
index a1d2abe..124761b 100644
--- a/CHANGELOG.md
+++ b/CHANGELOG.md
@@ -3,6 +3,14 @@
 Documentation for TransferBench is available at
 [https://rocm.docs.amd.com/projects/TransferBench](https://rocm.docs.amd.com/projects/TransferBench).
 
+## v1.66.01
+## Fixed
+- Adding support for TheRock
+- Fixing parsing issue when using NULL memory type
+- Fixing CUAD compilation flags when enabling NIC/MPI
+## Modified
+- TransferBenchCuda must now be explicitly built with via 'make TransferBenchCuda'
+
 ## v1.66.00
 ### Added
 - Adding multi-node support
diff --git a/CMakeLists.txt b/CMakeLists.txt
index 7cf8ba3..db7da3c 100644
--- a/CMakeLists.txt
+++ b/CMakeLists.txt
@@ -9,7 +9,7 @@ if (NOT CMAKE_TOOLCHAIN_FILE)
   message(STATUS "CMAKE_TOOLCHAIN_FILE: ${CMAKE_TOOLCHAIN_FILE}")
 endif()
 
-set(VERSION_STRING "1.66.00")
+set(VERSION_STRING "1.66.01")
 project(TransferBench VERSION ${VERSION_STRING} LANGUAGES CXX)
 
 ## Load CMake modules
diff --git a/Makefile b/Makefile
index 8360d60..042ecc2 100644
--- a/Makefile
+++ b/Makefile
@@ -26,15 +26,21 @@ SINGLE_KERNEL ?= 0
 # Default is the native GPU target
 GPU_TARGETS ?= native
 
+EXE=TransferBench
 DEBUG ?= 0
 
+# Only perform this check if 'make clean' is not the target
 ifeq ($(filter clean,$(MAKECMDGOALS)),)
-  # Compile TransferBenchCuda if nvidia-smi returns successfully and nvcc detected
-  ifeq ("$(shell nvidia-smi > /dev/null 2>&1 && test -e $(NVCC) && echo found)", "found")
-    EXE=TransferBenchCuda
-    CXX=$(NVCC)
+  ifeq ($(MAKECMDGOALS),TransferBenchCuda)
+    # Check for nvcc
+    ifneq ($(shell test -e $(NVCC) && echo found), found)
+      $(error "Could not find $(NVCC).  Please set CUDA_PATH appropriately")
+    else
+      $(info Compiling TransferBenchCuda using $(NVCC))
+    endif
+    NVFLAGS = -x cu -lnuma -arch=native
   else
-    EXE=TransferBench
+    # Check for HIP compiler
     ifeq ("$(shell test -e $(HIPCC) && echo found)", "found")
       CXX=$(HIPCC)
     else ifeq ("$(shell test -e $(ROCM_PATH)/bin/hipcc && echo found)", "found")
@@ -44,18 +50,17 @@ ifeq ($(filter clean,$(MAKECMDGOALS)),)
       $(error "Could not find $(HIPCC) or $(ROCM_PATH)/bin/hipcc. Check if the path is correct if you want to build $(EXE)")
     endif
     GPU_TARGETS_FLAGS = $(foreach target,$(GPU_TARGETS),"--offload-arch=$(target)")
-  endif
 
-  CXXFLAGS = -I$(ROCM_PATH)/include -I$(ROCM_PATH)/include/hip -I$(ROCM_PATH)/include/hsa
-  HIPLDFLAGS= -lnuma -L$(ROCM_PATH)/lib -lhsa-runtime64 -lamdhip64
-  HIPFLAGS = -Wall -x hip -D__HIP_PLATFORM_AMD__ -D__HIPCC__ $(GPU_TARGETS_FLAGS)
-  ifneq ($(strip $(ROCM_DEVICE_LIB_PATH)),)
-    HIPFLAGS += --rocm-device-lib-path=$(ROCM_DEVICE_LIB_PATH)
+    CXXFLAGS = -I$(ROCM_PATH)/include -I$(ROCM_PATH)/include/hip -I$(ROCM_PATH)/include/hsa
+    HIPLDFLAGS= -lnuma -L$(ROCM_PATH)/lib -lhsa-runtime64 -lamdhip64
+    HIPFLAGS = -Wall -x hip -D__HIP_PLATFORM_AMD__ -D__HIPCC__ $(GPU_TARGETS_FLAGS)
+    ifneq ($(strip $(ROCM_DEVICE_LIB_PATH)),)
+      HIPFLAGS += --rocm-device-lib-path=$(ROCM_DEVICE_LIB_PATH)
+    endif
   endif
-  NVFLAGS  = -x cu -lnuma -arch=native
 
   ifeq ($(SINGLE_KERNEL), 1)
-    CXXFLAGS += -DSINGLE_KERNEL
+    COMMON_FLAGS += -DSINGLE_KERNEL
   endif
 
   ifeq ($(DEBUG), 0)
@@ -79,7 +84,7 @@ ifeq ($(filter clean,$(MAKECMDGOALS)),)
     else ifeq ("$(shell echo '#include <infiniband/verbs.h>' | $(CXX) -E - 2>/dev/null | grep -c 'infiniband/verbs.h')", "0")
       $(info infiniband/verbs.h not found)
     else
-      CXXFLAGS += -DNIC_EXEC_ENABLED
+      COMMON_FLAGS += -DNIC_EXEC_ENABLED
       LDFLAGS += -libverbs
       NIC_ENABLED = 1
     endif
@@ -101,7 +106,7 @@ ifeq ($(filter clean,$(MAKECMDGOALS)),)
       $(info Unable to find mpi.h at $(MPI_PATH)/include.  Please specify appropriate MPI_PATH)
     else
       MPI_ENABLED = 1
-      CXXFLAGS += -DMPI_COMM_ENABLED -I$(MPI_PATH)/include
+      COMMON_FLAGS += -DMPI_COMM_ENABLED -I$(MPI_PATH)/include
       LDFLAGS += -L/$(MPI_PATH)/lib -lmpi
       ifeq ($(DEBUG), 1)
         LDFLAGS += -lmpi_cxx
diff --git a/src/client/EnvVars.hpp b/src/client/EnvVars.hpp
index 8346e02..5f6bbe7 100644
--- a/src/client/EnvVars.hpp
+++ b/src/client/EnvVars.hpp
@@ -40,7 +40,7 @@ THE SOFTWARE.
 #include <random>
 #include <time.h>
 
-#define CLIENT_VERSION "00"
+#define CLIENT_VERSION "01"
 
 #include "TransferBench.hpp"
 using namespace TransferBench;
diff --git a/src/header/TransferBench.hpp b/src/header/TransferBench.hpp
index 3906930..a8dd6cb 100644
--- a/src/header/TransferBench.hpp
+++ b/src/header/TransferBench.hpp
@@ -1176,7 +1176,10 @@ namespace {
       }
 
       // Parse memory type
-      ERR_CHECK(CharToMemType(*ptr, w.memType));
+      ErrResult err = CharToMemType(*ptr, w.memType);
+      if (err.errType != ERR_NONE) {
+        return {err.errType, "Error parsing token [%s]: %s\n", token.c_str(), err.errMsg.c_str()};
+      }
       ptr++; // Skip memory type
 
       // Parse memory index
@@ -1184,6 +1187,8 @@ namespace {
         ptr = ParseRange(ptr, -1, w.memIndices);
         if (!ptr) return {ERR_FATAL, "Unable to parse device index in memory token %s", token.c_str()};
         memDevices.push_back(w);
+      } else {
+        break;
       }
     }
     return ERR_NONE;