matejaputic

matejaputic

#define TILEX 4
#define TILEX_SHIFT 2
#define TILEY 4
#define TILEY_SHIFT 2

/* Matrix A is cached into local memory block */
/* Required global threads = (widthC / 4, heightC / 4) */
__kernel void mmmKernel_local(__global float4 *matrixA,
                              __global float4 *matrixB,
                              __global float4* matrixC,

matejaputic

diff --git a/src/samples/example_sgemm.c b/src/samples/example_sgemm.c
index fe47ba3..ea3c1be 100644
--- a/src/samples/example_sgemm.c
+++ b/src/samples/example_sgemm.c
@@ -17,6 +17,7 @@
 
 #include <sys/types.h>
 #include <stdio.h>
+#include <stdlib.h>
 #include <string.h>

matejaputic

#ProfileFileVersion=3.1
#ProfilerVersion=3.1.7247
#Application=/home/users/mputic/persistent/Projects/clBLAS/build/samples/example_sgemm
#ApplicationArgs=
#WorkingDirectory=
#Device AMD A10-7850K Radeon R7, 12 Compute Cores 4C+8G Platform Vendor=Advanced Micro Devices, Inc.
#Device AMD A10-7850K Radeon R7, 12 Compute Cores 4C+8G Platform Name=AMD Accelerated Parallel Processing
#Device AMD A10-7850K Radeon R7, 12 Compute Cores 4C+8G Platform Version=OpenCL 1.2 AMD-APP (1445.5)
#Device AMD A10-7850K Radeon R7, 12 Compute Cores 4C+8G CLDriver Version=1445.5 (sse2,avx,fma4)
#Device AMD A10-7850K Radeon R7, 12 Compute Cores 4C+8G CLRuntime Version=OpenCL 1.2 AMD-APP (1445.5)

matejaputic

#0  0x0000000000000000 in ?? ()
#1  0x00007ffff2e3435d in ?? () from /opt/AMDAPPSDK-2.9-1/lib/x86_64/libamdocl64.so
#2  0x00007ffff2df2a95 in ?? () from /opt/AMDAPPSDK-2.9-1/lib/x86_64/libamdocl64.so
#3  0x00007ffff2e075a6 in ?? () from /opt/AMDAPPSDK-2.9-1/lib/x86_64/libamdocl64.so
#4  0x00007ffff2de7cdd in clBuildProgram () from /opt/AMDAPPSDK-2.9-1/lib/x86_64/libamdocl64.so
#5  0x00007ffff651f692 in makeGemmKernel (clKernel=0x7fffffffd970, clQueue=0x87d6e0,
    kernelSource=0x7ffff68634a0 "__attribute__((reqd_work_group_size(16,16,1))) __kernel void sgemm_Col_NN_B1_MX096_NX096_KX16 ( __global float const * restrict A, __global float const * restrict B, __global float * C, float const al"...,
    sourceBuildOptions=0x7ffff685009e "-cl-std=CL2.0", kernelBinary=0x7fffffffd908,
    kernelBinarySize=0x7ffff7bc5da0 <sgemm_Col_NN_B1_MX096_NX096_KX16_binSize>,
    binaryBuildOptions=0x7ffff6850090 "-cl-std=CL2.0")

matejaputic

/usr/bin/cmake -H/home/users/mputic/persistent/Projects/clBLAS/src -B/home/users/mputic/persistent/Projects/clBLAS/build --check-build-system CMakeFiles/Makefile.cmake 0
Re-run cmake file: Makefile older than: /home/users/mputic/persistent/Projects/clBLAS/src/library/CMakeLists.txt
-- Build static Hawaii kernels.
-- Uncheck OCL_OFFLINE_BUILD_HAWAII_KERNEL to build kernls at run-time
-- Please ensure the presence of Hawaii device in the system. With certain driver/compiler flags, this might result in compile-time error.
-- Build dynamic Bonaire kernels.
-- Check OCL_OFFLINE_BUILD_BONAIRE_KERNEL to build kernls at compile-time. This will eliminates clBuildProgram() overhead and better kernel performance with certain driver.
-- Build dynamic Tahiti kernels.
-- Check OCL_OFFLINE_BUILD_TAHITI_KERNEL to build kernls at compile-time. This will eliminates clBuildProgram() overhead and better kernel performance with certain driver.
-- You have confirmed OpenCL 2.0 is supported in your system

matejaputic

#0  0x0000000000000000 in ?? ()
#1  0x00007ffff2e4435d in ?? () from /opt/AMDAPPSDK-2.9-1/lib/x86_64/libamdocl64.so
#2  0x00007ffff2e02a95 in ?? () from /opt/AMDAPPSDK-2.9-1/lib/x86_64/libamdocl64.so
#3  0x00007ffff2e175a6 in ?? () from /opt/AMDAPPSDK-2.9-1/lib/x86_64/libamdocl64.so
#4  0x00007ffff2df7cdd in clBuildProgram () from /opt/AMDAPPSDK-2.9-1/lib/x86_64/libamdocl64.so
#5  0x00007ffff652383d in makeGemmKernel (clKernel=0x7ffff7bc1350 <sgemm_Col_NN_B1_ML016_NL016_KX01_clKernel>, 
    clQueue=0x87d6e0, 
    kernelSource=0x7ffff69bafc8 "\n/* sgemm_Col_NN_B1_ML016_NL016_KX01 */\n\n/* kernel parameters */\n#define WG_NUM_ROWS          16\n#define WG_NUM_COLS          16\n#define MICRO_TILE_NUM_ROWS  1\n#define MICRO_TILE_NUM_COLS  1\n#define M"..., 
    sourceBuildOptions=0x7ffff683a73e "-cl-std=CL2.0", kernelBinary=0x7fffffffded0, 
    kernelBinarySize=0x7ffff7bc60d0 <sgemm_Col_NN_B1_ML016_NL016_KX01_binSize>, 

matejaputic

#0  0x0000000000000000 in ?? ()
No symbol table info available.
#1  0x00007ffff479335d in ?? () from /opt/AMDAPPSDK-2.9-1/lib/x86_64/libamdocl64.so
No symbol table info available.
#2  0x00007ffff4751a95 in ?? () from /opt/AMDAPPSDK-2.9-1/lib/x86_64/libamdocl64.so
No symbol table info available.
#3  0x00007ffff47665a6 in ?? () from /opt/AMDAPPSDK-2.9-1/lib/x86_64/libamdocl64.so
No symbol table info available.
#4  0x00007ffff4746cdd in clBuildProgram () from /opt/AMDAPPSDK-2.9-1/lib/x86_64/libamdocl64.so
No symbol table info available.

matejaputic

<button type="button" class="btn btn-default" id="save">Save Image</button>
<script>
var a = document.createElement("a");

d3.select("#save").on("click", function() {
  var canvas = document.querySelector("canvas"),
      context = canvas.getContext("2d");

  context.clearRect(0, 0, canvas.width, canvas.height);

matejaputic

<anml version="1.0"  xmlns:xsi="http://www.w3.org/2001/XMLSchema-instance">
  <automata-network id="netlist">
    <counter id="n0" target="53" at-target="latch">
      <report-on-target/>
    </counter>
    <state-transition-element id="n8t0" symbol-set="*" start="start-of-data">
      <activate-on-match element="n8t0"/>
      <activate-on-match element="n7t1"/>
    </state-transition-element>
    <state-transition-element id="n2t0" symbol-set="[\xff]" start="start-of-data">