GBE/libocl: Add __gen_ocl_get_timestamp() to get timestamp.

Gen provide tm0 register for intra-kernel profiling.
Here we provide an API __gen_ocl_get_timestamp() to return
the timestamp in TM.

The return type is defined as:
struct time_stamp {
  ulong tick;
  uint event;
};

'tick' is a 64bit time tick. 'event' stores a value which means
whether a tmEvent has occured (non-zero) or not (0). tmEvent includes
time-impacting event such as context switch or frequency change
since last time tm0 was read.

I add a sample in the kernels/compiler_time_stamp.cl. Hope it
would help you understand how to use it.

V2:
Introduce ir::ARFRegister to avoid directly use of nr/subnr in Gen IR.
Rename __gen_ocl_extract_reg to __gen_ocl_region.
Rename beignet_get_time_stamp to __gen_ocl_get_timestamp.

Signed-off-by: Ruiling Song <ruiling.song@intel.com>
Reviewed-by: Zhigang Gong <zhigang.gong@linux.intel.com>

1 files changed, 52 insertions, 0 deletions

diff --git a/utests/compiler_time_stamp.cpp b/utests/compiler_time_stamp.cpp
new file mode 100644
index 00000000..4da57523
--- /dev/null
+++ b/utests/compiler_time_stamp.cpp
@@ -0,0 +1,52 @@
+#include "utest_helper.hpp"
+
+static void cpu(int global_id, int *src, int *dst) {
+  int i;
+  int final[16];
+  for (i = 0; i < 16; ++i) {
+    int array[16], j;
+    for (j = 0; j < 16; ++j)
+      array[j] = global_id;
+    for (j = 0; j < src[0]; ++j)
+      array[j] = 1+src[j];
+    final[i] = array[i];
+  }
+  dst[global_id] = final[global_id];
+}
+
+void compiler_time_stamp(void)
+{
+  const size_t n = 16;
+  int cpu_dst[16], cpu_src[16];
+
+  // Setup kernel and buffers
+  OCL_CREATE_KERNEL("compiler_time_stamp");
+  OCL_CREATE_BUFFER(buf[0], 0, n * sizeof(uint32_t), NULL);
+  OCL_CREATE_BUFFER(buf[1], 0, n * sizeof(uint32_t), NULL);
+  OCL_SET_ARG(0, sizeof(cl_mem), &buf[0]);
+  OCL_SET_ARG(1, sizeof(cl_mem), &buf[1]);
+  globals[0] = 16;
+  locals[0] = 16;
+
+  // Run random tests
+  for (uint32_t pass = 0; pass < 1; ++pass) {
+    OCL_MAP_BUFFER(0);
+    for (int32_t i = 0; i < (int32_t) n; ++i)
+      cpu_src[i] = ((int32_t*)buf_data[0])[i] = rand() % 16;
+    OCL_UNMAP_BUFFER(0);
+
+    // Run the kernel on GPU
+    OCL_NDRANGE(1);
+
+    // Run on CPU
+    for (int32_t i = 0; i <(int32_t) n; ++i) cpu(i, cpu_src, cpu_dst);
+
+    // Compare
+    OCL_MAP_BUFFER(1);
+    for (int32_t i = 0; i < 11; ++i)
+      OCL_ASSERT(((int32_t*)buf_data[1])[i] == cpu_dst[i]);
+    OCL_UNMAP_BUFFER(1);
+  }
+}
+
+MAKE_UTEST_FROM_FUNCTION(compiler_time_stamp);