Vortex 2.0 changes:

+ Microarchitecture optimizations
+ 64-bit support
+ Xilinx FPGA support
+ LLVM-16 support
+ Refactoring and quality control fixes

minor update

minor update

minor update

minor update

minor update

minor update

cleanup

cleanup

cache bindings and memory perf refactory

minor update

minor update

hw unit tests fixes

minor update

minor update

minor update

minor update

minor update

minor udpate

minor update

minor update

minor update

minor update

minor update

minor update

minor update

minor updates

minor updates

minor update

minor update

minor update

minor update

minor update

minor update

minor updates

minor updates

minor updates

minor updates

minor update

minor update

tests/regression/printf/main.cpp

@@ -1,6 +1,7 @@
 #include <iostream>
 #include <unistd.h>
 #include <string.h>
+#include <vector>
 #include <vortex.h>
 #include "common.h"
 
@@ -20,8 +21,8 @@ const char* kernel_file = "kernel.bin";
 uint32_t count = 4;
 
 vx_device_h device = nullptr;
-vx_buffer_h staging_buf = nullptr;
-kernel_arg_t kernel_arg;
+std::vector<uint8_t> staging_buf;
+kernel_arg_t kernel_arg = {};
 
 static void show_usage() {
    std::cout << "Vortex Test." << std::endl;
@@ -51,9 +52,6 @@ static void parse_args(int argc, char **argv) {
 }
 
 void cleanup() {
-  if (staging_buf) {
-    vx_buf_free(staging_buf);
-  }
   if (device) {
     vx_mem_free(device, kernel_arg.src_addr);
     vx_dev_close(device);
@@ -67,14 +65,12 @@ int run_test() {
 
   // wait for completion
   std::cout << "wait for completion" << std::endl;
-  RT_CHECK(vx_ready_wait(device, MAX_TIMEOUT));
+  RT_CHECK(vx_ready_wait(device, VX_MAX_TIMEOUT));
 
   return 0;
 }
 
-int main(int argc, char *argv[]) {
-  size_t value;
-  
+int main(int argc, char *argv[]) {  
   // parse command arguments
   parse_args(argc, argv);
 
@@ -86,10 +82,9 @@ int main(int argc, char *argv[]) {
   std::cout << "open device connection" << std::endl;  
   RT_CHECK(vx_dev_open(&device));
 
-  uint64_t max_cores, max_warps, max_threads;
-  RT_CHECK(vx_dev_caps(device, VX_CAPS_MAX_CORES, &max_cores));
-  RT_CHECK(vx_dev_caps(device, VX_CAPS_MAX_WARPS, &max_warps));
-  RT_CHECK(vx_dev_caps(device, VX_CAPS_MAX_THREADS, &max_threads));
+  uint64_t num_warps, num_threads;
+  RT_CHECK(vx_dev_caps(device, VX_CAPS_NUM_WARPS, &num_warps));
+  RT_CHECK(vx_dev_caps(device, VX_CAPS_NUM_THREADS, &num_threads));
 
   uint32_t num_points = count;
   uint32_t buf_size = count * sizeof(int32_t);
@@ -102,37 +97,35 @@ int main(int argc, char *argv[]) {
   RT_CHECK(vx_upload_kernel_file(device, kernel_file));
 
   // allocate device memory
-  std::cout << "allocate device memory" << std::endl;  
-
-  RT_CHECK(vx_mem_alloc(device, buf_size, &value));
-  kernel_arg.src_addr = value;
+  std::cout << "allocate device memory" << std::endl;
+  RT_CHECK(vx_mem_alloc(device, buf_size, VX_MEM_TYPE_GLOBAL, &kernel_arg.src_addr));
 
   kernel_arg.num_points = num_points;
 
-  std::cout << "dev_src=" << std::hex << kernel_arg.src_addr << std::endl;
+  std::cout << "dev_src=0x" << std::hex << kernel_arg.src_addr << std::endl;
   
-  // allocate shared memory  
-  std::cout << "allocate shared memory" << std::endl;    
+  // allocate staging buffer  
+  std::cout << "allocate staging buffer" << std::endl;    
   uint32_t alloc_size = std::max<uint32_t>(buf_size, sizeof(kernel_arg_t));
-  RT_CHECK(vx_buf_alloc(device, alloc_size, &staging_buf));
+  staging_buf.resize(alloc_size);
   
-  // upload kernel argument
-  std::cout << "upload kernel argument" << std::endl;
+  // upload kernel argument  
   {
-    auto buf_ptr = (void*)vx_host_ptr(staging_buf);
+    std::cout << "upload kernel argument" << std::endl;
+    auto buf_ptr = (void*)staging_buf.data();
     memcpy(buf_ptr, &kernel_arg, sizeof(kernel_arg_t));
-    RT_CHECK(vx_copy_to_dev(staging_buf, KERNEL_ARG_DEV_MEM_ADDR, sizeof(kernel_arg_t), 0));
+    RT_CHECK(vx_copy_to_dev(device, KERNEL_ARG_DEV_MEM_ADDR, staging_buf.data(), sizeof(kernel_arg_t)));
   }
 
   // upload source buffer0
   {
-    auto buf_ptr = (int*)vx_host_ptr(staging_buf);
+    std::cout << "upload source buffer" << std::endl;
+    auto buf_ptr = (int32_t*)staging_buf.data();
     for (uint32_t i = 0; i < num_points; ++i) {
       buf_ptr[i] = i;
-    }
+    }    
+    RT_CHECK(vx_copy_to_dev(device, kernel_arg.src_addr, staging_buf.data(), buf_size));  
   }
-  std::cout << "upload source buffer" << std::endl;      
-  RT_CHECK(vx_copy_to_dev(staging_buf, kernel_arg.src_addr, buf_size, 0));  
 
   // run tests
   std::cout << "run tests" << std::endl;