DPS915/M-N-M

From CDOT Wiki
Jump to: navigation, search


GPU610/DPS915 | Student List | Group and Project Index | Student Resources | Glossary

Team: M-N-M

Team Members

  1. Muhammad Ahsan
  2. Nitin Prakash Panicker
  3. Mohamed Baig


Email All

Potential Projects

Projects
Project Name Project Description Status
C Compiler Take the compilation process and transfer it to the GPU failed
Galactic Collision Simulation A 2D simulation of two galaxies colliding, with fully simulated gravity effects. cannot benefit from parallel computing,already fast enough
Isomorphism of graphs Check if two graphs are isomorphic in nature. This is a very straight forward program. failed
Image Processing Mathematical operations applied to images for colour negation, rotation, blurring effects, etc. scrapped
Facial Recognition system Speed up time taken to match face with database already CUDA enabled
Steganography Speed up the process of encrypting one type of file into another Profiled
Prime number generator Generate prime numbers Profiled
LZW File compression Compress files using Lempel-Ziv-Welch algorithm Profiled

Progress

Assignment 1

Mohamed Baig: Steganography using Steghide

Steghide has certain dependencies that it uses to complete its function.
Dependencies:
Makefile changes made
  • libmash
  • libcrypt
  • libjpeg
  • zlib
  • Ran the Configure file to see if I have all the Dependencies
  • Installed the all the dependencies
  • Ran Configure again to generate Makefile in the src folder
  • Altered the Makefile to enable profiling
  • Altered some source files to avoid errors

AuSampleValues.cc

#include "AuSampleValues.h"

// AuMuLawSampleValue
template <> //My change
const BYTE AuMuLawSampleValue::MinValue = 0 ;
template <> //My change
const BYTE AuMuLawSampleValue::MaxValue = BYTE_MAX ;

// AuPCM8SampleValue
template <> //My change
const SBYTE AuPCM8SampleValue::MinValue = SBYTE_MIN ;
template <> //My change
const SBYTE AuPCM8SampleValue::MaxValue = SBYTE_MAX ;

// AuPCM16SampleValue
template <> //My change
const SWORD16 AuPCM16SampleValue::MinValue = SWORD16_MIN ;
template <> //My change
const SWORD16 AuPCM16SampleValue::MaxValue = SWORD16_MAX ;

// AuPCM32SampleValue
template <> //My change
const SWORD32 AuPCM32SampleValue::MinValue = SWORD32_MIN ;
template <> //My change
const SWORD32 AuPCM32SampleValue::MaxValue = SWORD32_MAX ;

AuData.h

#ifndef SH_AUDATA_H
#define SH_AUDATA_H

#include "BinaryIO.h"
#include "AudioData.h"

// AuMuLawAudioData
typedef AudioDataImpl<AuMuLaw,BYTE> AuMuLawAudioData ;
template <> //My change
inline BYTE AuMuLawAudioData::readValue (BinaryIO* io) const { return (io->read8()) ; }
template <> //My change
inline void AuMuLawAudioData::writeValue (BinaryIO* io, BYTE v) const { io->write8(v) ; }

// AuPCM8AudioData
typedef AudioDataImpl<AuPCM8,SBYTE> AuPCM8AudioData ;
template <> //My change
inline SBYTE AuPCM8AudioData::readValue (BinaryIO* io) const { return ((SBYTE) io->read8()) ; }
template <> //My change
inline void AuPCM8AudioData::writeValue (BinaryIO* io, SBYTE v) const { io->write8((BYTE) v) ; }

// AuPCM16AudioData
typedef AudioDataImpl<AuPCM16,SWORD16> AuPCM16AudioData ;
template <> //My change
inline SWORD16 AuPCM16AudioData::readValue (BinaryIO* io) const { return ((SWORD16) io->read16_be()) ; }
template <> //My change
inline void AuPCM16AudioData::writeValue (BinaryIO* io, SWORD16 v) const { io->write16_be((UWORD16) v) ; }

// AuPCM32AudioData
typedef AudioDataImpl<AuPCM32,SWORD32> AuPCM32AudioData ;
template <> //My change
inline SWORD32 AuPCM32AudioData::readValue (BinaryIO* io) const { return ((SWORD32) io->read32_be()) ; }
template <> //My change
inline void AuPCM32AudioData::writeValue (BinaryIO* io, SWORD32 v) const { io->write32_be((UWORD32) v) ; }

#endif // ndef SH_AUDATA_H
  • The result of embedding kilobytes of text data into an image
Flat profile:

Each sample counts as 0.01 seconds.
  %   cumulative   self              self     total          
 time   seconds   seconds    calls   s/call   s/call  name   
 13.38      0.21     0.21  1318276     0.00     0.00  Selector::idxX(unsigned int, unsigned int, unsigned int*) const
 11.47      0.39     0.18  4054789     0.00     0.00  Vertex::getDegree() const
  9.55      0.54     0.15   659139     0.00     0.00  Selector::calculate(unsigned int)
  7.64      0.66     0.12   659141     0.00     0.00  JpegFile::getEmbeddedValue(unsigned int) const
  7.01      0.77     0.11   659139     0.00     0.00  __gnu_cxx::hashtable<std::pair<unsigned int const, unsigned int>, unsigned int, __gnu_cxx::hash<unsigned int>, std::_Select1st<std::pair<unsigned int const, unsigned int> >, std::equal_to<unsigned int>, std::allocator<unsigned int> >::resize(unsigned long)
  6.37      0.87     0.10   328293     0.00     0.00  JpegFile::getSampleValue(unsigned int) const
  5.73      0.96     0.09        1     0.09     0.09  Selector::~Selector()
  3.82      1.02     0.06        1     0.06     0.09  JpegFile::read(BinaryIO*)
  • The result of attempting to embed 1.7 GB of data into an image
Flat profile:

Each sample counts as 0.01 seconds.
  %   cumulative   self              self     total          
 time   seconds   seconds    calls   s/call   s/call  name   
 54.43     21.52    21.52 1192388169     0.00     0.00  BitString::_append(bool)
 16.79     28.15     6.64 593200712     0.00     0.00  BitString::operator[](unsigned long) const
  9.30     31.83     3.68 74898412     0.00     0.00  BitString::append(unsigned char, unsigned short)
  4.78     33.72     1.89        3     0.63     6.41  BitString::append(BitString const&)
  3.50     35.10     1.39                             BitString::BitString(unsigned long)
  1.91     35.86     0.76                             BitString::clear()
  1.29     36.37     0.51        1     0.51    37.26  Embedder::Embedder()
  1.20     36.84     0.48 37823336     0.00     0.00  BinaryIO::eof() const

Muhammad Ahsan: Prime Number Generator( 1,000,000,000 primes)

Flat profile

Each sample counts as 0.01 seconds.
  %   cumulative   self              self     total           
 time   seconds   seconds    calls  Ts/call  Ts/call  name    
100.00     53.83    53.83                             get_primes(unsigned long)
  0.00     53.83     0.00       27     0.00     0.00  void std::vector<unsigned long, std::allocator<unsigned long> >::_M_emplace_back_aux<unsigned long const&>(unsigned long const&&&)
  0.00     53.83     0.00        1     0.00     0.00  _GLOBAL__sub_I__Z10get_primesm

/** {{{ http://code.activestate.com/recipes/576559/ (r2) */
/*
 Copyright (c) 2008 Florian Mayer

 Permission is hereby granted, free of charge, to any person obtaining a copy
 of this software and associated documentation files (the "Software"), to deal
 in the Software without restriction, including without limitation the rights
 to use, copy, modify, merge, publish, distribute, sublicense, and/or sell
 copies of the Software, and to permit persons to whom the Software is
 furnished to do so, subject to the following conditions:

 The above copyright notice and this permission notice shall be included in
 all copies or substantial portions of the Software.

 THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
 IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
 FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
 AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
 LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM,
 OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN
 THE SOFTWARE.
*/

#include <iostream>
#include <vector>
#include <string.h>

using namespace std;

vector<unsigned long> get_primes(unsigned long max){
    vector<unsigned long> primes;
    char *sieve;
    sieve = new char[max/8+1];
    // Fill sieve with 1 
    memset(sieve, 0xFF, (max/8+1) * sizeof(char));
    for(unsigned long x = 2; x <= max; x++)
        if(sieve[x/8] & (0x01 << (x % 8))){
            primes.push_back(x);
            // Is prime. Mark multiplicates.
            for(unsigned long j = 2*x; j <= max; j += x)
                sieve[j/8] &= ~(0x01 << (j % 8));
        }
    delete[] sieve;
    return primes;
}

int main(void){
    vector<unsigned long> primes;
    primes = get_primes(1000000000);
    // return 0;
    // Print out result.
    vector<unsigned long>::iterator it;
    for(it=primes.begin(); it < primes.end(); it++)
        cout << *it << " "<<endl;
   
    cout << endl;
    return 0;
}
/** end of http://code.activestate.com/recipes/576559/ }}} */

Nitin Prakash Panicker: LZW File Compression


Flat profile:



Each sample counts as 0.01 seconds.

  %   cumulative   self              self     total           

 time   seconds   seconds    calls  ns/call  ns/call  name    

 99.46     48.19    48.19                             CLZWCompressFile::Compress(char*, char*)

  0.33     48.35     0.16 17122488     9.34     9.34  CLZWCompressFile::getc_src()

  0.21     48.45     0.10  7095561    14.09    14.09  CLZWCompressFile::putc_comp(int)

Source Code for LZW File Compression

lzw.cpp lzw.h

Assignment 2

Source code for prime number generator we will be putting on the gpu


    # include <cmath> // This library enable the use of sqrt.
    # include <iostream>
   
    using namespace std;
   
    void primenum(long double); // Prototype...
   
    int c = 0;
   
    int main(){
        long double x = 0;
        cout<<"\n This program will generate all prime numbers up to the"
        <<"\n number you have entered below...\n";
        cout<<"\n Please enter a number: ";
        cin>> x;
        cout<<"\n Here are all the prime numbers up to "<<x<<".\n";
        primenum(x); //function invocation...
        cout<<endl<<"\nThere are "<<c
        <<" prime numbers less than or equal to "<<x<<".\n\n";
        return 0;
    }
    // This function will determine the primenumbers up to num.
    void primenum(long double x){
    bool prime = true; // Calculates the square-root of 'x'
    int number2;
    number2 =(int) floor (sqrt (x));
    for (int i = 1; i <= x; i++){
        for ( int j = 2; j <= number2; j++){
            if ( i!=j && i % j == 0 ){
                prime = false;
                break;
            }
        }
        if (prime){
            cout <<" "<<i<<" ";
            c += 1;
        }
        prime = true;
    }
    getchar();
    }

Version of prime generator running on GPU


    # include <cmath> // This library enable the use of sqrt.
    # include <iostream>
    # include <cuda_runtime.h>
   
    using namespace std;
   
    void primenum(long double); // Prototype...
   
    int c = 0;
   
    int main(){
        long double x = 0;
        cout<<"\n This program will generate all prime numbers up to the"
        <<"\n number you have entered below...\n";
        cout<<"\n Please enter a number: ";
        cin>> x;
        cout<<"\n Here are all the prime numbers up to "<<x<<".\n";
        primenum(x); //function invocation...
        cout<<endl<<"\nThere are "<<c
        <<" prime numbers less than or equal to "<<x<<".\n\n";
        return 0;
    }
   
    // This function will determine the primenumbers up to num.
    void primenum(long double x){
        //Array to hold generated primes on host
        int *primes_h = new int[x];
        //Device array to hold the primes on the device
        int *primes_d = new int[x];
        //allocate device memory and initialize device memory
        cudaMalloc((void**)&primes_d, x * sizeof(int));
        cudaMemset(&primes_d,sizeof(int),x * sizeof(int);
       
        //Kernal goes here
        //error checking
       
        //copy the array holding primes from device to host
        cudaMemcpy(primes_h, primes_d, x * sizeof(int), cudaMemcpyDeviceToHost);
       
        //display the primes
        for(int i=0; i<x ; i++){
            cout<<primes_h[i]<<endl;
        }
       
        //free allocated memory
        delete [] primes_h;
        cudaFree(primes_d);
       
        getchar();
    }

Almost Final version

# include <cmath> // This library enable the use of sqrt.

    # include <iostream>

	# include <ctime>

	#include<iomanip>

	#include<cstdlib>

    # include <cuda_runtime.h>

	//#include <times.h>

   

    using namespace std;

	

	inline clock_t getMilliSecs() {

		return clock() / (CLOCKS_PER_SEC / 1000);

	}



	__global__ void primegen(bool prime, int number2,int x,int *primes_d)

	{

      int c = 0;



	for (int i = 1; i <= x; i++)

	{

        for ( int j = 2; j <= number2; j++)

		{

            if ( i!=j && i % j == 0 )

			{

                prime = false;

                break;

            }

        }

        if (prime)

		{

           primes_d[c]=i;

            c += 1;

        }

        prime = true;

    }

	

	}	

   

    void primenum(long double); // Prototype...

   

 

   

    int main()

	{

        long double x = 0;

        cout<<"\n This program will generate all prime numbers up to the"<<"\n number you have entered below...\n";

        cout<<"\n Please enter a number: ";

        cin>> x;

        cout<<"\n Here are all the prime numbers up to "<<x<<".\n";

        primenum(x); //function invocation...

        //cout<<endl<<"\nThere are "<<c

        //<<" prime numbers less than or equal to "<<x<<".\n\n";

        return 0;

    }

   

    // This function will determine the primenumbers up to num.

    void primenum(long double x)

	{

		bool prime = true;

		//struct tms start_time, stop_time;

		  int number2;

		  number2 =(int) floor (sqrt (x));

		  clock_t start = getMilliSecs();

        //Array to hold generated primes on host

        int *primes_h = new int[(int)x];

        //Device array to hold the primes on the device

        int *primes_d = new int[(int)x];

        //allocate device memory and initialize device memory

        cudaMalloc((void**)&primes_d, (int)x * sizeof(int));

		// cudaMalloc((void**)&c_d, sizeof(int));

        cudaMemset(&primes_d,0,x * sizeof(int));

        //error checking

        cudaError_t error ;

        //Kernal goes here

		 primegen<<<1,1>>>(prime,number2,(int)x,primes_d);

		

		// extract error code from the kernel's execution

         error = cudaGetLastError();

         if (error != cudaSuccess) {

                 cout << cudaGetErrorString(error) << endl;

         }

       

        //copy the array holding primes from device to host

        error =cudaMemcpy(primes_h, primes_d, ((int)x) * sizeof(int), cudaMemcpyDeviceToHost);

     

         if (error != cudaSuccess) {

                 cout << cudaGetErrorString(error) << endl;

         }

		// cudaMemcpy(c_h, c_d, sizeof(int), cudaMemcpyDeviceToHost);

        //display the primes

		for(int i=0; i<(int)x ; i++){

			if(primes_h[i]>=2 && primes_h[i]<=(int)x){

				cout<<primes_h[i]<<endl;

			}

        }

       cout << "Elapsed time: " << (getMilliSecs() - start) << "ms" << endl;

      // cout<< "time: "<< (stop_s-start_s)/double(CLOCKS_PER_SEC)<<endl;

        //free allocated memory

		

        delete [] primes_h;

        cudaFree(primes_d);

       

        getchar();

    }

Assignment 3

Cuda Version:First Attempt


# include <cmath> // This library enable the use of sqrt.

    # include <iostream>

	# include <ctime>

	#include<iomanip>

	#include<cstdlib>

    # include <cuda_runtime.h>

	//#include <times.h>

   

    using namespace std;

	

	inline clock_t getMilliSecs() {

		return clock() / (CLOCKS_PER_SEC / 1000);

	}



	__global__ void primegen(bool prime, int number2,int x,int *primes_d)

	{

      int c = 0;

	  int idx = blockIdx.x * blockDim.x + threadIdx.x;

	  

	  for ( int i=1; i <= x; i++)

		{

			if( i!= idx && i%idx == 0 )

			{

				prime = false;

				break;

			}

		

			if(prime)

			{

				primes_d[c]=i;

				c += 1;

			}

				prime = true;

		}

		

			

	 } 

	  



	/*for (int i = 1; i <= x; i++)

	{

        for ( int j = 2; j <= number2; j++)

		{

            if ( i!=j && i % j == 0 )

			{

                prime = false;

                break;

            }

        }

        if (prime)

		{

           primes_d[c]=i;

            c += 1;

        }

        prime = true;  

		

    } */

	



   

    void primenum(long double); // Prototype...

   

 

   

    int main()

	{

        long double x = 0;

        cout<<"\n This program will generate all prime numbers up to the"<<"\n number you have entered below...\n";

        cout<<"\n Please enter a number: ";

        cin>> x;

        cout<<"\n Here are all the prime numbers up to "<<x<<".\n";

        primenum(x); //function invocation...

        //cout<<endl<<"\nThere are "<<c

        //<<" prime numbers less than or equal to "<<x<<".\n\n";

        return 0;

    }

   

    // This function will determine the primenumbers up to num.

    void primenum(long double x)

	{

	int   n = x;

	  int   d;

		bool prime = true;

		//struct tms start_time, stop_time;

		  int number2;

		  number2 =(int) floor (sqrt (x));

		  clock_t start = getMilliSecs();

		  

		   cudaDeviceProp prop;

         cudaGetDevice(&d);

         cudaGetDeviceProperties(&prop, d);

         int nThreads = prop.maxThreadsDim[0];

         int n_max = nThreads * prop.maxGridSize[0];

         if ( n> n_max) {

                n = n_max;

                cout << "n reduced to " << n << endl;

         } 

		  

        //Array to hold generated primes on host

        int *primes_h = new int[(int)x];

		

        //Device array to hold the primes on the device

        int *primes_d = new int[(int)x];

		

        //allocate device memory and initialize device memory

        cudaMalloc((void**)&primes_d, (int)x * sizeof(int));

		

		// cudaMalloc((void**)&c_d, sizeof(int));

        cudaMemset(&primes_d,0,x * sizeof(int));

		

        //error checking

        cudaError_t error ;

		

        //Kernal goes here

		 primegen<<<(n + nThreads - 1) / nThreads, nThreads>>>(prime,number2,(int)x,primes_d);

		

		// extract error code from the kernel's execution

		

         error = cudaGetLastError();

         if (error != cudaSuccess) {

                 cout << cudaGetErrorString(error) << endl;

         }

       

        //copy the array holding primes from device to host

		

        error =cudaMemcpy(primes_h, primes_d, ((int)x) * sizeof(int), cudaMemcpyDeviceToHost);

     

         if (error != cudaSuccess) {

                 cout << cudaGetErrorString(error) << endl;

         }

		// cudaMemcpy(c_h, c_d, sizeof(int), cudaMemcpyDeviceToHost);

        //display the primes

		for(int i=0; i<(int)x ; i++){

			if(primes_h[i]>=2 && primes_h[i]<=(int)x){

				cout<<primes_h[i]<<endl;

			}

        }

       cout << "Elapsed time: " << (getMilliSecs() - start) << "ms" << endl;

      // cout<< "time: "<< (stop_s-start_s)/double(CLOCKS_PER_SEC)<<endl;

        //free allocated memory

		

        delete [] primes_h;

        cudaFree(primes_d);

       

        getchar();

    }

Conclusion: Logical Error

GpuA3error.png

The prime number generated seems to have run into some logical error. It does not generate the prime numbers correctly. Instead spits out all numbers.

Cuda Version: Attempt Two

Gives a run time error "invalid argument". Logical error still persists.

Final Cuda version

#include <cstdio>
#include <cstdlib>
#include <iostream>
#include <ctime>
#include <cuda_runtime.h>

using namespace std;

/**
 * This macro checks return value of the CUDA runtime call and exits
 * the application if the call failed.
 */
#define CUDA_CHECK_RETURN(value) {                                            \
    cudaError_t _m_cudaStat = value;                                        \
    if (_m_cudaStat != cudaSuccess) {                                        \
        fprintf(stderr, "Error %s at line %d in file %s\n",                    \
                cudaGetErrorString(_m_cudaStat), __LINE__, __FILE__);        \
        exit(1);                                                            \
    } }

/**
 * Kernel code to generate and detect primes
 */
__global__ void prime(int *num, int blockNum, int threadNum, int size) {
    const int tid = blockIdx.x * blockDim.x + threadIdx.x;
    const int bid = blockIdx.y * blockDim.y + threadIdx.y;
    __syncthreads();

    /**
     * Generate prime numbers and store them in the array.
     * The first element is always 2
     */
    if(tid == 0) {
        num[tid] = 2;
    } else {
        num[tid] = 2 * tid + 1;
    }

    int tmp = bid * threadNum + tid;

    int step1 = 2 * tmp + 3;
    int step2 = tmp + 1;

    while(tmp < size) {
        int i = 1;
        /**
         * Check if an element is not prime, if it isn't set it to 0.
         */
        while((step1 * i + step2) < size) {
            num[step1 * i + step2] = 0;
            i++;
        }
        tmp += blockNum * threadNum;
        __syncthreads();
    }
}

int main(int argc, char* argv[]) {
    if(argc != 2) {
        cout << "Incorrect no of arguments" << endl;
        return 1;
    }
    int n = atoi(argv[1]);

    /**
     * variable declarations
     */
    int *device;
    int host[n];
    int d;
    cudaDeviceProp prop;

    /**
     * Get the properties of the device in use
     */
    cudaGetDevice(&d);
    cudaGetDeviceProperties(&prop, d);
    int numberOfBlocks = 8;
    int maxThreadsPerBlock = prop.maxThreadsPerBlock;
    int numberOfThreads = maxThreadsPerBlock/numberOfBlocks;

    /**
     * Start timer
     */
    clock_t cb, ce;
    cb = clock();

    /**
     * Allocate memory on the device
     */
    CUDA_CHECK_RETURN(cudaMalloc((void**) &device, sizeof(int) * n));

    /**
     * Call kernel with appropriate grid and thread size
     */
    prime<<<numberOfBlocks, numberOfThreads>>>(device, numberOfBlocks, numberOfThreads, n);

    /**
     * Copy results back to host
     */
    CUDA_CHECK_RETURN(cudaMemcpy(&host, device, sizeof(int) * n, cudaMemcpyDeviceToHost));

    /**
     * Free memory on device
     */
    CUDA_CHECK_RETURN(cudaFree(device));

    /**
     * Output values
     */
    for (int i = 0; i < n; i++)
        if (host[i] != 0)
            cout << host[i] << endl;

    /**
     * Stop timer
     */
    ce = clock();
    cout << "Prime generation - took " << double(ce - cb)/CLOCKS_PER_SEC << " seconds" << endl;
}
Manual Delete Warning
Final version's errors, warnings and observations
  • If a number over 515 is entered as the launch argument, the program will display random values at the end of the list of prime numbers
  • When attempting to delete the host array manually in the program, a warning is displayed
Manual Delete Crash
  • The program crashes at the end if the host array is manually deleted
Successful run of Prime generation

PrimeSuccessfulRun.png