oscarbg

// -*- compile-command: "nvcc -m 32 -arch sm_50 -Xptxas=-v,-abi=no -cubin int_mul.cu" ; -*-

#include <stdint.h>

//
//
//

#define KERNEL_QUALIFIERS                   __global__
#define KERNEL_QUALIFIERS_EXTERN            extern     KERNEL_QUALIFIERS

oscarbg

//goes well until ptxas which says something about invalid arguments so at least f16x2 modified atom instruction is recognized?
u32
atomf16x2(u32 a, u32 b)
{
	u32 d;

	asm("atom.global.add.f16x2 	%0, [%1], %2;" : "=r"(d) : "r"(a), "r"(b));
	//atom.global.add.u32 	%r5, [%rd2], 10;
	//asm("mul.wide.s16 %0, %1, %2;" : "=r"(d) : "h"(a), "h"(b));

oscarbg

# Approximate recipe on compiling and running OpenSWR on AWS Instance (Ubuntu 14.04)
# Questions regarding this recipe: @__rej__
# http://openswr.org



# prerequisites
sudo apt-get update
sudo apt-get install git
sudo apt-get install build-essential

oscarbg

# Pre requisites
- Enable DRI3 as described in https://vulkan.lunarg.com/app/docs/v1.0.3.1/getting_started_linux
- Install a library for SHA, e.g. sudo apt-get install libgcrypt11-dev (if not already present)
  - Otherwise the driver may throw an error with "_mesa_sha1_compute" when loading SPIR-V shaders

# Building
- Clone Mesa Master : git clone git://anongit.freedesktop.org/mesa/mesa -b Master
- cd mesa
- autoreconf -vfi
- ./configure --with-dri-drivers=i965 --with-gallium-drivers= --with-sha1= --with-vulkan-drivers=intel

oscarbg

#include <wrl/client.h>
#include <winml.h>
#include <stdio.h>
using Microsoft::WRL::ComPtr;

#include"cnpy.h"

#define PRINTDBG

int main(int argc,char**argv)

oscarbg

directml.dll metacommands found:
copytensor
reduce
gemm
pooling
roipooling
convolution
normalization
mvn
rnn

oscarbg

name of display: :0
display: :0  screen: 0
direct rendering: Yes
server glx vendor string: SGI
server glx version string: 1.4
server glx extensions:
    GLX_ARB_create_context, GLX_ARB_create_context_no_error, 
    GLX_ARB_create_context_profile, GLX_ARB_create_context_robustness, 
    GLX_ARB_fbconfig_float, GLX_ARB_framebuffer_sRGB, GLX_ARB_multisample, 
    GLX_EXT_create_context_es2_profile, GLX_EXT_create_context_es_profile, 

oscarbg

name of display: :0
display: :0  screen: 0
direct rendering: Yes
server glx vendor string: SGI
server glx version string: 1.4
server glx extensions:
    GLX_ARB_create_context, GLX_ARB_create_context_no_error, 
    GLX_ARB_create_context_profile, GLX_ARB_create_context_robustness, 
    GLX_ARB_fbconfig_float, GLX_ARB_framebuffer_sRGB, GLX_ARB_multisample, 
    GLX_EXT_create_context_es2_profile, GLX_EXT_create_context_es_profile, 

oscarbg

name of display: :0
display: :0  screen: 0
direct rendering: Yes
server glx vendor string: SGI
server glx version string: 1.4
server glx extensions:
    GLX_ARB_create_context, GLX_ARB_create_context_no_error, 
    GLX_ARB_create_context_profile, GLX_ARB_create_context_robustness, 
    GLX_ARB_fbconfig_float, GLX_ARB_framebuffer_sRGB, GLX_ARB_multisample, 
    GLX_EXT_create_context_es2_profile, GLX_EXT_create_context_es_profile, 

oscarbg

==========
VULKANINFO
==========

Vulkan Instance Version: 1.1.82

WARNING: [Loader Message] Code 0 : loader_scanned_icd_add: Using deprecated ICD interface of 'vkGetInstanceProcAddr' instead of 'vk_icdGetInstanceProcAddr' for ICD talvos-vulkan.dll


Instance Extensions: