bioem_cuda.cu 26.8 KB
Newer Older
Pilar Cossio's avatar
License  
Pilar Cossio committed
1
/* ++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++
2 3 4 5
   < BioEM software for Bayesian inference of Electron Microscopy images>
   Copyright (C) 2016 Pilar Cossio, David Rohr, Fabio Baruffa, Markus Rampp, 
        Volker Lindenstruth and Gerhard Hummer.
   Max Planck Institute of Biophysics, Frankfurt, Germany.
6

7
   See license statement for terms of distribution.
Pilar Cossio's avatar
License  
Pilar Cossio committed
8 9 10

   ++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++*/

11 12 13 14 15 16 17 18 19 20 21
#define BIOEM_GPUCODE

#if defined(_WIN32)
#include <windows.h>
#endif

#include <iostream>
using namespace std;

#include "bioem_cuda_internal.h"
#include "bioem_algorithm.h"
Pilar Cossio's avatar
Pilar Cossio committed
22
//#include "helper_cuda.h"
23

24 25 26 27 28 29 30
#define checkCudaErrors(error) \
{ \
	if ((error) != cudaSuccess) \
	{ \
		printf("CUDA Error %d / %s (%s: %d)\n", error, cudaGetErrorString(error), __FILE__, __LINE__); \
		exit(1); \
	} \
31 32
}

David Rohr's avatar
David Rohr committed
33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69
static const char *cufftGetErrorStrung(cufftResult error)
{
    switch (error)
    {
        case CUFFT_SUCCESS:
            return "CUFFT_SUCCESS";

        case CUFFT_INVALID_PLAN:
            return "CUFFT_INVALID_PLAN";

        case CUFFT_ALLOC_FAILED:
            return "CUFFT_ALLOC_FAILED";

        case CUFFT_INVALID_TYPE:
            return "CUFFT_INVALID_TYPE";

        case CUFFT_INVALID_VALUE:
            return "CUFFT_INVALID_VALUE";

        case CUFFT_INTERNAL_ERROR:
            return "CUFFT_INTERNAL_ERROR";

        case CUFFT_EXEC_FAILED:
            return "CUFFT_EXEC_FAILED";

        case CUFFT_SETUP_FAILED:
            return "CUFFT_SETUP_FAILED";

        case CUFFT_INVALID_SIZE:
            return "CUFFT_INVALID_SIZE";

        case CUFFT_UNALIGNED_DATA:
            return "CUFFT_UNALIGNED_DATA";
    }
    return "UNKNOWN";
}

Luka Stanisic's avatar
Luka Stanisic committed
70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133
/* Handing CUDA Driver errors */

#define cuErrorCheck(call) \
  do { \
    CUresult __error__; \
    if ((__error__ = (call)) != CUDA_SUCCESS) { \
      printf("CUDA Driver Error %d / %s (%s %d)\n", __error__, cuGetError(__error__),__FILE__, __LINE__); \
      return __error__; \
    } \
  } while (false)

static const char * cuGetError(CUresult result) {
  switch (result) {
    case CUDA_SUCCESS:                              return "No errors";
    case CUDA_ERROR_INVALID_VALUE:                  return "Invalid value";
    case CUDA_ERROR_OUT_OF_MEMORY:                  return "Out of memory";
    case CUDA_ERROR_NOT_INITIALIZED:                return "Driver not initialized";
    case CUDA_ERROR_DEINITIALIZED:                  return "Driver deinitialized";
    case CUDA_ERROR_PROFILER_DISABLED:              return "Profiler disabled";
    case CUDA_ERROR_PROFILER_NOT_INITIALIZED:       return "Profiler not initialized";
    case CUDA_ERROR_PROFILER_ALREADY_STARTED:       return "Profiler already started";
    case CUDA_ERROR_PROFILER_ALREADY_STOPPED:       return "Profiler already stopped";
    case CUDA_ERROR_NO_DEVICE:                      return "No CUDA-capable device available";
    case CUDA_ERROR_INVALID_DEVICE:                 return "Invalid device";
    case CUDA_ERROR_INVALID_IMAGE:                  return "Invalid kernel image";
    case CUDA_ERROR_INVALID_CONTEXT:                return "Invalid context";
    case CUDA_ERROR_CONTEXT_ALREADY_CURRENT:        return "Context already current";
    case CUDA_ERROR_MAP_FAILED:                     return "Map failed";
    case CUDA_ERROR_UNMAP_FAILED:                   return "Unmap failed";
    case CUDA_ERROR_ARRAY_IS_MAPPED:                return "Array is mapped";
    case CUDA_ERROR_ALREADY_MAPPED:                 return "Already mapped";
    case CUDA_ERROR_NO_BINARY_FOR_GPU:              return "No binary for GPU";
    case CUDA_ERROR_ALREADY_ACQUIRED:               return "Already acquired";
    case CUDA_ERROR_NOT_MAPPED:                     return "Not mapped";
    case CUDA_ERROR_NOT_MAPPED_AS_ARRAY:            return "Not mapped as array";
    case CUDA_ERROR_NOT_MAPPED_AS_POINTER:          return "Not mapped as pointer";
    case CUDA_ERROR_ECC_UNCORRECTABLE:              return "Uncorrectable ECC error";
    case CUDA_ERROR_UNSUPPORTED_LIMIT:              return "Unsupported CUlimit";
    case CUDA_ERROR_CONTEXT_ALREADY_IN_USE:         return "Context already in use";
    case CUDA_ERROR_INVALID_SOURCE:                 return "Invalid source";
    case CUDA_ERROR_FILE_NOT_FOUND:                 return "File not found";
    case CUDA_ERROR_SHARED_OBJECT_SYMBOL_NOT_FOUND: return "Shared object symbol not found";
    case CUDA_ERROR_SHARED_OBJECT_INIT_FAILED:      return "Shared object initialization failed";
    case CUDA_ERROR_OPERATING_SYSTEM:               return "Operating System call failed";
    case CUDA_ERROR_INVALID_HANDLE:                 return "Invalid handle";
    case CUDA_ERROR_NOT_FOUND:                      return "Not found";
    case CUDA_ERROR_NOT_READY:                      return "CUDA not ready";
    case CUDA_ERROR_LAUNCH_FAILED:                  return "Launch failed";
    case CUDA_ERROR_LAUNCH_OUT_OF_RESOURCES:        return "Launch exceeded resources";
    case CUDA_ERROR_LAUNCH_TIMEOUT:                 return "Launch exceeded timeout";
    case CUDA_ERROR_LAUNCH_INCOMPATIBLE_TEXTURING:  return "Launch with incompatible texturing";
    case CUDA_ERROR_PEER_ACCESS_ALREADY_ENABLED:    return "Peer access already enabled";
    case CUDA_ERROR_PEER_ACCESS_NOT_ENABLED:        return "Peer access not enabled";
    case CUDA_ERROR_PRIMARY_CONTEXT_ACTIVE:         return "Primary context active";
    case CUDA_ERROR_CONTEXT_IS_DESTROYED:           return "Context is destroyed";
    case CUDA_ERROR_ASSERT:                         return "Device assert failed";
    case CUDA_ERROR_TOO_MANY_PEERS:                 return "Too many peers";
    case CUDA_ERROR_HOST_MEMORY_ALREADY_REGISTERED: return "Host memory already registered";
    case CUDA_ERROR_HOST_MEMORY_NOT_REGISTERED:     return "Host memory not registered";
    case CUDA_ERROR_UNKNOWN:                        return "Unknown error";
    default:                                        return "Unknown error code";
  }
}

134 135 136 137
bioem_cuda::bioem_cuda()
{
	deviceInitialized = 0;
	GPUAlgo = getenv("GPUALGO") == NULL ? 2 : atoi(getenv("GPUALGO"));
138 139
	GPUAsync = getenv("GPUASYNC") == NULL ? 1 : atoi(getenv("GPUASYNC"));
	GPUWorkload = getenv("GPUWORKLOAD") == NULL ? 100 : atoi(getenv("GPUWORKLOAD"));
140
	GPUDualStream = getenv("GPUDUALSTREAM") == NULL ? 1 : atoi(getenv("GPUDUALSTREAM"));
141 142 143 144 145 146 147
}

bioem_cuda::~bioem_cuda()
{
	deviceExit();
}

148 149 150
__global__ void compareRefMap_kernel(const int iOrient, const int iConv,  const myfloat_t amp, const myfloat_t pha, const myfloat_t env, const myfloat_t sumC,
                                                const myfloat_t sumsquareC, const myfloat_t* pMap, bioem_Probability pProb, 
						const bioem_param_device param, const bioem_RefMap_Mod RefMap, const int cent_x, const int cent_y, const int maxRef)
151 152
{
	const int iRefMap = myBlockIdxX * myBlockDimX + myThreadIdxX;
153
	if (iRefMap < maxRef)
154
	{
155
		compareRefMap<0>(iRefMap, iOrient, iConv, amp, pha, env, sumC, sumsquareC, pMap, pProb, param, RefMap, cent_x, cent_y);
156 157 158
	}
}

Pilar Cossio's avatar
Pilar Cossio committed
159
__global__ void compareRefMapShifted_kernel(const int iOrient, const int iConv, const myfloat_t amp, const myfloat_t pha, const myfloat_t env, const myfloat_t sumC, const myfloat_t sumsquareC, const myfloat_t* pMap, bioem_Probability pProb, const bioem_param_device param, const bioem_RefMap_Mod RefMap, const int maxRef)
160 161
{
	const int iRefMap = myBlockIdxX * myBlockDimX + myThreadIdxX;
162
	if (iRefMap < maxRef)
163
	{
164
		compareRefMapShifted<1>(iRefMap, iOrient, iConv, amp, pha, env, sumC, sumsquareC, pMap, pProb, param, RefMap);
165 166 167
	}
}

168 169 170 171 172 173
__global__ void cudaZeroMem(void* ptr, size_t size)
{
	int* myptr = (int*) ptr;
	int mysize = size / sizeof(int);
	int myid = myBlockDimX * myBlockIdxX + myThreadIdxX;
	int mygrid = myBlockDimX * myGridDimX;
174
	for (int i = myid; i < mysize; i += mygrid) myptr[i] = 0;
175 176
}

Pilar Cossio's avatar
Pilar Cossio committed
177
__global__ void compareRefMapLoopShifts_kernel(const int iOrient, const int iConv, const myfloat_t amp, const myfloat_t pha, const myfloat_t env, const myfloat_t sumC, const myfloat_t sumsquareC, const myfloat_t* pMap, bioem_Probability pProb, const bioem_param_device param, const bioem_RefMap RefMap, const int blockoffset, const int nShifts, const int nShiftBits, const int maxRef)
178 179 180 181 182 183 184 185 186
{
	const size_t myid = (myBlockIdxX + blockoffset) * myBlockDimX + myThreadIdxX;
	const int iRefMap = myid >> (nShiftBits << 1);
	const int myRef = myThreadIdxX >> (nShiftBits << 1);
	const int myShiftIdx = (myid >> nShiftBits) & (nShifts - 1);
	const int myShiftIdy = myid & (nShifts - 1);
	const int myShift = myid & (nShifts * nShifts - 1);
	const int cent_x = myShiftIdx * param.GridSpaceCenter - param.maxDisplaceCenter;
	const int cent_y = myShiftIdy * param.GridSpaceCenter - param.maxDisplaceCenter;
187

188
	const bool threadActive = myShiftIdx < nShifts && myShiftIdy < nShifts && iRefMap < maxRef;
189

Pilar Cossio's avatar
Pilar Cossio committed
190
	compareRefMap<2>(iRefMap, iOrient, iConv, amp, pha, env, sumC, sumsquareC, pMap, pProb, param, RefMap, cent_x, cent_y, myShift, nShifts * nShifts, myRef, threadActive);
191 192
}

193
__global__ void multComplexMap(const mycomplex_t* convmap, const mycomplex_t* refmap, mycuComplex_t* out, const int NumberPixelsTotal, const int MapSize, const int NumberMaps, const int Offset)
194 195
{
	if (myBlockIdxX >= NumberMaps) return;
196
	const mycuComplex_t* myin = (mycuComplex_t*) &refmap[(myBlockIdxX + Offset) * MapSize];
197
	const mycuComplex_t* myconv = (mycuComplex_t*) convmap;
198
	mycuComplex_t* myout = &out[myBlockIdxX * MapSize];
199
	for(int i = myThreadIdxX; i < NumberPixelsTotal; i += myBlockDimX)
200
	{
201 202 203 204 205 206 207
		mycuComplex_t val;
		const mycuComplex_t conv = myconv[i];
		const mycuComplex_t in = myin[i];

		val.x = conv.x * in.x + conv.y * in.y;
		val.y = conv.y * in.x - conv.x * in.y;
		myout[i] = val;
208 209 210
	}
}

211
__global__ void cuDoRefMapsFFT(const int iOrient, const int iConv, const myfloat_t amp, const myfloat_t pha, const myfloat_t env, const myfloat_t* lCC, const myfloat_t sumC, const myfloat_t sumsquareC, bioem_Probability pProb, const bioem_param_device param, const bioem_RefMap RefMap, const int maxRef, const int Offset)
212
{
213
	if (myBlockIdxX * myBlockDimX + myThreadIdxX >= maxRef) return;
214 215
	const int iRefMap = myBlockIdxX * myBlockDimX + myThreadIdxX + Offset;
	const myfloat_t* mylCC = &lCC[(myBlockIdxX * myBlockDimX + myThreadIdxX) * param.NumberPixels * param.NumberPixels];
216
	doRefMapFFT(iRefMap, iOrient, iConv, amp, pha, env, mylCC, sumC, sumsquareC, pProb, param, RefMap);
217 218
}

219 220 221 222 223 224 225 226 227 228 229 230 231
template <class T> static inline T divup(T num, T divider) {return((num + divider - 1) / divider);}
static inline bool IsPowerOf2(int x) {return ((x > 0) && ((x & (x - 1)) == 0));}
#if defined(_WIN32)
static inline int ilog2 (int value)
{
	DWORD index;
	_BitScanReverse (&index, value);
	return(value);
}
#else
static inline int ilog2(int value) {return 31 - __builtin_clz(value);}
#endif

232
int bioem_cuda::compareRefMaps(int iOrient, int iConv, myfloat_t amp, myfloat_t pha, myfloat_t env, const myfloat_t* conv_map, mycomplex_t* localmultFFT, myfloat_t sumC, myfloat_t sumsquareC, const int startMap)
233
{
234 235 236 237 238
	if (startMap)
	{
		cout << "Error startMap not implemented for GPU Code\n";
		exit(1);
	}
Luka Stanisic's avatar
Luka Stanisic committed
239 240 241 242 243 244 245
#ifdef DEBUG_GPU
	float time;
	cudaEvent_t start, stop;
	checkCudaErrors(cudaEventCreate(&start));
	checkCudaErrors(cudaEventCreate(&stop));
	checkCudaErrors(cudaEventRecord(start, 0));
#endif
246 247 248 249
	if (GPUAsync)
	{
		checkCudaErrors(cudaEventSynchronize(cudaEvent[iConv & 1]));
	}
Luka Stanisic's avatar
Luka Stanisic committed
250 251 252 253 254 255 256
#ifdef DEBUG_GPU
	checkCudaErrors(cudaEventRecord(stop, 0));
	checkCudaErrors(cudaEventSynchronize(stop));
	checkCudaErrors(cudaEventElapsedTime(&time, start, stop));
	printf("\t\t\tGPU: time to synch projections %1.6f sec\n", time/1000);
	checkCudaErrors(cudaEventRecord(start, 0));
#endif
257
	if (FFTAlgo)
258
	{
259
		memcpy(&pConvMapFFT_Host[(iConv & 1) * param.FFTMapSize], localmultFFT, param.FFTMapSize * sizeof(mycomplex_t));
260
		checkCudaErrors(cudaMemcpyAsync(&pConvMapFFT[(iConv & 1) * param.FFTMapSize], &pConvMapFFT_Host[(iConv & 1) * param.FFTMapSize], param.FFTMapSize * sizeof(mycomplex_t), cudaMemcpyHostToDevice, cudaStream[GPUAsync ? 2 : 0]));
Luka Stanisic's avatar
Luka Stanisic committed
261 262 263 264 265 266 267
#ifdef DEBUG_GPU
		checkCudaErrors(cudaEventRecord(stop, 0));
		checkCudaErrors(cudaEventSynchronize(stop));
		checkCudaErrors(cudaEventElapsedTime(&time, start, stop));
		printf("\t\t\tGPU: time for memcpy %1.6f sec\n", time/1000);
		checkCudaErrors(cudaEventRecord(start, 0));
#endif
268 269 270 271 272
		if (GPUAsync)
		{
			checkCudaErrors(cudaEventRecord(cudaEvent[2], cudaStream[2]));
			checkCudaErrors(cudaStreamWaitEvent(cudaStream[0], cudaEvent[2], 0));
		}
273
		if (GPUDualStream)
274
		{
275 276 277 278 279 280
			checkCudaErrors(cudaEventRecord(cudaFFTEvent[0], cudaStream[0]));
			checkCudaErrors(cudaStreamWaitEvent(cudaStream[1], cudaFFTEvent[0], 0));
		}
		for (int i = 0, j = 0; i < maxRef; i += CUDA_FFTS_AT_ONCE, j++)
		{
			if (!GPUDualStream) j = 0;
281
			const int num = min(CUDA_FFTS_AT_ONCE, maxRef - i);
282 283
			multComplexMap<<<num, CUDA_THREAD_COUNT, 0, cudaStream[j & 1]>>>(&pConvMapFFT[(iConv & 1) * param.FFTMapSize], pRefMapsFFT, pFFTtmp2[j & 1], param.param_device.NumberPixels * param.param_device.NumberFFTPixels1D, param.FFTMapSize, num, i);
			cufftResult err = mycufftExecC2R(i + CUDA_FFTS_AT_ONCE > maxRef ? plan[1][j & 1] : plan[0][j & 1], pFFTtmp2[j & 1], pFFTtmp[j & 1]);
David Rohr's avatar
David Rohr committed
284
			if (err != CUFFT_SUCCESS)
285
			{
David Rohr's avatar
David Rohr committed
286
				cout << "Error running CUFFT " << cufftGetErrorStrung(err) << "\n";
287 288
				exit(1);
			}
289
			cuDoRefMapsFFT<<<divup(num, CUDA_THREAD_COUNT), CUDA_THREAD_COUNT, 0, cudaStream[j & 1]>>>(iOrient, iConv,  amp, pha, env, pFFTtmp[j & 1], sumC, sumsquareC, pProb_device, param.param_device, *gpumap, num, i);
290
		}
Luka Stanisic's avatar
Luka Stanisic committed
291
		checkCudaErrors(cudaPeekAtLastError());
292 293 294 295 296
		if (GPUDualStream)
		{
			checkCudaErrors(cudaEventRecord(cudaFFTEvent[1], cudaStream[1]));
			checkCudaErrors(cudaStreamWaitEvent(cudaStream[0], cudaFFTEvent[1], 0));
		}
297 298 299
	}
	else
	{
300
		checkCudaErrors(cudaMemcpyAsync(pConvMap_device[iConv & 1], conv_map, sizeof(myfloat_t) * RefMap.refMapSize, cudaMemcpyHostToDevice, cudaStream[0]));
Luka Stanisic's avatar
Luka Stanisic committed
301 302 303 304 305 306 307
#ifdef DEBUG_GPU
		checkCudaErrors(cudaEventRecord(stop, 0));
		checkCudaErrors(cudaEventSynchronize(stop));
		checkCudaErrors(cudaEventElapsedTime(&time, start, stop));
		printf("\t\t\tGPU: time for memcpy %1.6f sec\n", time/1000);
		checkCudaErrors(cudaEventRecord(start, 0) );
#endif
308
		if (GPUAlgo == 2) //Loop over shifts
309
		{
310 311 312 313 314 315 316 317 318 319 320 321 322 323 324 325 326 327 328
			const int nShifts = 2 * param.param_device.maxDisplaceCenter / param.param_device.GridSpaceCenter + 1;
			if (!IsPowerOf2(nShifts))
			{
				cout << "Invalid number of displacements, no power of two\n";
				exit(1);
			}
			if (CUDA_THREAD_COUNT % (nShifts * nShifts))
			{
				cout << "CUDA Thread count (" << CUDA_THREAD_COUNT << ") is no multiple of number of shifts (" << (nShifts * nShifts) << ")\n";
				exit(1);
			}
			if (nShifts > CUDA_MAX_SHIFT_REDUCE)
			{
				cout << "Too many displacements for CUDA reduction\n";
				exit(1);
			}
			const int nShiftBits = ilog2(nShifts);
			size_t totalBlocks = divup((size_t) maxRef * (size_t) nShifts * (size_t) nShifts, (size_t) CUDA_THREAD_COUNT);
			size_t nBlocks = CUDA_BLOCK_COUNT;
329
			for (size_t i = 0; i < totalBlocks; i += nBlocks)
330
			{
Pilar Cossio's avatar
Pilar Cossio committed
331
				compareRefMapLoopShifts_kernel<<<min(nBlocks, totalBlocks - i), CUDA_THREAD_COUNT, (CUDA_THREAD_COUNT * 2 + CUDA_THREAD_COUNT / (nShifts * nShifts) * 4) * sizeof(myfloat_t), cudaStream[0] >>> (iOrient, iConv, amp, pha, env, sumC, sumsquareC, pConvMap_device[iConv & 1], pProb_device, param.param_device, *gpumap, i, nShifts, nShiftBits, maxRef);
332
			}
333
		}
334
		else if (GPUAlgo == 1) //Split shifts in multiple kernels
335
		{
336
			for (int cent_x = -param.param_device.maxDisplaceCenter; cent_x <= param.param_device.maxDisplaceCenter; cent_x = cent_x + param.param_device.GridSpaceCenter)
337
			{
338
				for (int cent_y = -param.param_device.maxDisplaceCenter; cent_y <= param.param_device.maxDisplaceCenter; cent_y = cent_y + param.param_device.GridSpaceCenter)
339
				{
Pilar Cossio's avatar
Pilar Cossio committed
340
					compareRefMap_kernel<<<divup(maxRef, CUDA_THREAD_COUNT), CUDA_THREAD_COUNT, 0, cudaStream[0]>>> (iOrient, iConv, amp, pha, env, sumC, sumsquareC, pConvMap_device[iConv & 1], pProb_device, param.param_device, *pRefMap_device_Mod, cent_x, cent_y, maxRef);
341 342
				}
			}
343
		}
344
		else if (GPUAlgo == 0) //All shifts in one kernel
345
		{
346
			compareRefMapShifted_kernel<<<divup(maxRef, CUDA_THREAD_COUNT), CUDA_THREAD_COUNT, 0, cudaStream[0]>>> (iOrient, iConv, amp, pha, env, sumC, sumsquareC, pConvMap_device[iConv & 1], pProb_device, param.param_device, *pRefMap_device_Mod, maxRef);
347
		}
348
		else
349
		{
350 351
			cout << "Invalid GPU Algorithm selected\n";
			exit(1);
352
		}
353
	}
Luka Stanisic's avatar
Luka Stanisic committed
354 355 356 357 358 359 360
#ifdef DEBUG_GPU
	checkCudaErrors(cudaEventRecord(stop, 0));
	checkCudaErrors(cudaEventSynchronize(stop));
	checkCudaErrors(cudaEventElapsedTime(&time, start, stop));
	printf("\t\t\tGPU: time to run CUDA %1.6f sec\n", time/1000);
	checkCudaErrors(cudaEventRecord(start, 0));
#endif
361 362
	if (GPUWorkload < 100)
	{
363
		bioem::compareRefMaps(iOrient, iConv, amp, pha, env, conv_map, localmultFFT, sumC, sumsquareC, maxRef);
364
	}
Luka Stanisic's avatar
Luka Stanisic committed
365 366 367 368 369 370
#ifdef DEBUG_GPU
	checkCudaErrors(cudaEventRecord(stop, 0));
	checkCudaErrors(cudaEventSynchronize(stop));
	checkCudaErrors(cudaEventElapsedTime(&time, start, stop));
	printf("\t\t\tGPU: time to run OMP %1.6f sec\n", time/1000);
#endif
371 372
	if (GPUAsync)
	{
373
		checkCudaErrors(cudaEventRecord(cudaEvent[iConv & 1], cudaStream[0]));
374
	}
375 376
	else
	{
377
		checkCudaErrors(cudaStreamSynchronize(cudaStream[0]));
378 379 380 381
	}
	return(0);
}

David Rohr's avatar
David Rohr committed
382 383 384
int bioem_cuda::selectCudaDevice()
{
	int count;
385
	int bestDevice = 0;
David Rohr's avatar
David Rohr committed
386
	cudaDeviceProp deviceProp;
387 388 389 390 391

	/* Initializing CUDA driver API */
	cuErrorCheck(cuInit(0));

	/* Get number of available CUDA devices */
David Rohr's avatar
David Rohr committed
392 393 394 395 396 397
	checkCudaErrors(cudaGetDeviceCount(&count));
	if (count == 0)
	{
		printf("No CUDA device detected\n");
		return(1);
	}
398 399 400 401

	/* Find the best GPU */
	long long int bestDeviceSpeed = -1, deviceSpeed = -1;
	for (int i = 0; i < count; i++)
David Rohr's avatar
David Rohr committed
402
	{
403 404
		cudaGetDeviceProperties(&deviceProp, i);
		deviceSpeed = (long long int) deviceProp.multiProcessorCount * (long long int) deviceProp.clockRate * (long long int) deviceProp.warpSize;
David Rohr's avatar
David Rohr committed
405 406 407 408 409 410
		if (deviceSpeed > bestDeviceSpeed)
		{
			bestDevice = i;
			bestDeviceSpeed = deviceSpeed;
		}
	}
411 412

	/* Get user-specified GPU choice */
413 414 415 416 417 418 419 420 421 422 423 424 425 426 427 428 429
	if (getenv("GPUDEVICE"))
	{
		int device = atoi(getenv("GPUDEVICE"));
		if (device > count)
		{
			printf("Invalid CUDA device specified, max device number is %d\n", count);
			exit(1);
		}
#ifdef WITH_MPI
		if (device == -1)
		{
			device = mpi_rank % count;
		}
#endif
		if (device < 0)
		{
			printf("Negative CUDA device specified: %d, invalid!\n", device);
430
			exit(1);
431 432 433
		}
		bestDevice = device;
	}
David Rohr's avatar
David Rohr committed
434

435 436 437 438 439 440 441 442 443 444 445 446 447 448 449 450 451
	/* Set CUDA processes to appropriate devices */
	cudaGetDeviceProperties(&deviceProp, bestDevice);
	if (deviceProp.computeMode == 0)
	{
		checkCudaErrors(cudaSetDevice(bestDevice));
	}
	else
	{
		if (DebugOutput >= 1)
		{
			printf("CUDA device %d is not set in DEFAULT mode, make sure that CUDA processes are pinned as planned!\n", bestDevice);
			printf("Pinning process %d to CUDA device %d\n", mpi_rank, bestDevice);
		}
		checkCudaErrors(cudaSetDevice(bestDevice));
		/* This synchronization is needed in order to detect bogus silent errors from cudaSetDevice call */
		checkCudaErrors(cudaDeviceSynchronize());
	}
David Rohr's avatar
David Rohr committed
452

453
	/* Debugging information about CUDA devices used by the current process */
David Rohr's avatar
David Rohr committed
454
	if (DebugOutput >= 3)
David Rohr's avatar
David Rohr committed
455
	{
David Rohr's avatar
David Rohr committed
456 457 458 459 460 461 462 463 464 465 466 467 468 469 470 471
		printf("Using CUDA Device %s with Properties:\n", deviceProp.name);
		printf("totalGlobalMem = %lld\n", (unsigned long long int) deviceProp.totalGlobalMem);
		printf("sharedMemPerBlock = %lld\n", (unsigned long long int) deviceProp.sharedMemPerBlock);
		printf("regsPerBlock = %d\n", deviceProp.regsPerBlock);
		printf("warpSize = %d\n", deviceProp.warpSize);
		printf("memPitch = %lld\n", (unsigned long long int) deviceProp.memPitch);
		printf("maxThreadsPerBlock = %d\n", deviceProp.maxThreadsPerBlock);
		printf("maxThreadsDim = %d %d %d\n", deviceProp.maxThreadsDim[0], deviceProp.maxThreadsDim[1], deviceProp.maxThreadsDim[2]);
		printf("maxGridSize = %d %d %d\n", deviceProp.maxGridSize[0], deviceProp.maxGridSize[1], deviceProp.maxGridSize[2]);
		printf("totalConstMem = %lld\n", (unsigned long long int) deviceProp.totalConstMem);
		printf("major = %d\n", deviceProp.major);
		printf("minor = %d\n", deviceProp.minor);
		printf("clockRate = %d\n", deviceProp.clockRate);
		printf("memoryClockRate = %d\n", deviceProp.memoryClockRate);
		printf("multiProcessorCount = %d\n", deviceProp.multiProcessorCount);
		printf("textureAlignment = %lld\n", (unsigned long long int) deviceProp.textureAlignment);
472 473 474 475 476 477 478 479 480 481 482 483 484 485 486 487 488 489 490 491
		printf("computeMode = %d\n", deviceProp.computeMode);
#if CUDA_VERSION > 3010
		size_t free, total;
#else
		unsigned int free, total;
#endif
		if (deviceProp.computeMode == 0)
		{
			CUdevice tmpDevice;
			cuErrorCheck(cuDeviceGet(&tmpDevice, bestDevice));
			CUcontext tmpContext;
			cuErrorCheck(cuCtxCreate(&tmpContext, 0, tmpDevice));
			cuErrorCheck(cuMemGetInfo(&free, &total));
			cuErrorCheck(cuCtxDestroy(tmpContext));
		}
		else
		{
			cuErrorCheck(cuMemGetInfo(&free, &total));
		}
		printf("free memory = %lld; total memory = %lld\n", free, total);
David Rohr's avatar
David Rohr committed
492
	}
493

David Rohr's avatar
David Rohr committed
494 495
	if (DebugOutput >= 1)
	{
David Rohr's avatar
David Rohr committed
496
		printf("BioEM for CUDA initialized (MPI Rank %d), %d GPUs found, using GPU %d\n", mpi_rank, count, bestDevice);
David Rohr's avatar
David Rohr committed
497
	}
498

David Rohr's avatar
David Rohr committed
499 500 501
	return(0);
}

502 503 504
int bioem_cuda::deviceInit()
{
	deviceExit();
David Rohr's avatar
David Rohr committed
505
	
506
	selectCudaDevice();
507

508 509
	if (FFTAlgo) GPUAlgo = 2;

510 511 512 513 514
	gpumap = new bioem_RefMap;
	memcpy(gpumap, &RefMap, sizeof(bioem_RefMap));
	if (FFTAlgo == 0)
	{
		checkCudaErrors(cudaMalloc(&maps, sizeof(myfloat_t) * RefMap.ntotRefMap * RefMap.refMapSize));
515 516 517 518 519 520 521 522 523 524 525 526 527

		if (GPUAlgo == 0 || GPUAlgo == 1)
		{
			pRefMap_device_Mod = (bioem_RefMap_Mod*) gpumap;
			bioem_RefMap_Mod* RefMapGPU = new bioem_RefMap_Mod;
			RefMapGPU->init(RefMap);
			checkCudaErrors(cudaMemcpy(maps, RefMapGPU->maps, sizeof(myfloat_t) * RefMap.ntotRefMap * RefMap.refMapSize, cudaMemcpyHostToDevice));
			delete RefMapGPU;
		}
		else
		{
			checkCudaErrors(cudaMemcpy(maps, RefMap.maps, sizeof(myfloat_t) * RefMap.ntotRefMap * RefMap.refMapSize, cudaMemcpyHostToDevice));
		}
528 529 530 531 532 533 534 535 536
	}
	checkCudaErrors(cudaMalloc(&sum, sizeof(myfloat_t) * RefMap.ntotRefMap));
	checkCudaErrors(cudaMemcpy(sum, RefMap.sum_RefMap, sizeof(myfloat_t) * RefMap.ntotRefMap, cudaMemcpyHostToDevice));
	checkCudaErrors(cudaMalloc(&sumsquare, sizeof(myfloat_t) * RefMap.ntotRefMap));
	checkCudaErrors(cudaMemcpy(sumsquare, RefMap.sumsquare_RefMap, sizeof(myfloat_t) * RefMap.ntotRefMap, cudaMemcpyHostToDevice));
	gpumap->maps = maps;
	gpumap->sum_RefMap = sum;
	gpumap->sumsquare_RefMap = sumsquare;

537
	checkCudaErrors(cudaMalloc(&pProb_memory, pProb_device.get_size(RefMap.ntotRefMap, param.nTotGridAngles, param.nTotCC, param.param_device.writeAngles, param.param_device.writeCC)));
538

539
	for (int i = 0; i < 2; i++)
540
	{
541
		checkCudaErrors(cudaStreamCreate(&cudaStream[i]));
542
		checkCudaErrors(cudaEventCreate(&cudaEvent[i]));
543
		checkCudaErrors(cudaEventCreate(&cudaFFTEvent[i]));
544
		checkCudaErrors(cudaMalloc(&pConvMap_device[i], sizeof(myfloat_t) * RefMap.refMapSize));
545
	}
546 547 548 549 550
	if (GPUAsync)
	{
		checkCudaErrors(cudaStreamCreate(&cudaStream[2]));
		checkCudaErrors(cudaEventCreate(&cudaEvent[2]));
	}
551

552 553
	if (FFTAlgo)
	{
554
		checkCudaErrors(cudaMalloc(&pRefMapsFFT, RefMap.ntotRefMap * param.FFTMapSize * sizeof(mycomplex_t)));
555 556 557 558
		checkCudaErrors(cudaMalloc(&pFFTtmp2[0], CUDA_FFTS_AT_ONCE * param.FFTMapSize * 2 * sizeof(mycomplex_t)));
		checkCudaErrors(cudaMalloc(&pFFTtmp[0], CUDA_FFTS_AT_ONCE * param.param_device.NumberPixels * param.param_device.NumberPixels * 2 * sizeof(myfloat_t)));
		pFFTtmp2[1] = pFFTtmp2[0] + CUDA_FFTS_AT_ONCE * param.FFTMapSize;
		pFFTtmp[1] = pFFTtmp[0] + CUDA_FFTS_AT_ONCE * param.param_device.NumberPixels * param.param_device.NumberPixels;
559
		checkCudaErrors(cudaMalloc(&pConvMapFFT, param.FFTMapSize * sizeof(mycomplex_t) * 2));
560
		checkCudaErrors(cudaHostAlloc(&pConvMapFFT_Host, param.FFTMapSize * sizeof(mycomplex_t) * 2, 0));
561
		checkCudaErrors(cudaMemcpy(pRefMapsFFT, RefMap.RefMapsFFT, RefMap.ntotRefMap * param.FFTMapSize * sizeof(mycomplex_t), cudaMemcpyHostToDevice));
562 563
	}

564 565 566 567 568 569 570
	deviceInitialized = 1;
	return(0);
}

int bioem_cuda::deviceExit()
{
	if (deviceInitialized == 0) return(0);
571

572

David Rohr's avatar
David Rohr committed
573
	cudaFree(pProb_memory);
574 575
	cudaFree(sum);
	cudaFree(sumsquare);
576
	for (int i = 0; i < 2; i++)
577
	{
578
		cudaStreamDestroy(cudaStream[i]);
579
		cudaEventDestroy(cudaEvent[i]);
580
		cudaEventDestroy(cudaFFTEvent[i]);
581
		cudaFree(pConvMap_device[i]);
582
	}
583 584 585 586
	if (FFTAlgo)
	{
		cudaFree(pRefMapsFFT);
		cudaFree(pConvMapFFT);
587
		cudaFreeHost(pConvMapFFT_Host);
588 589
		cudaFree(pFFTtmp[0]);
		cudaFree(pFFTtmp2[0]);
590
	}
591 592 593 594 595 596 597 598
	else
	{
		cudaFree(maps);
	}
	if (GPUAlgo == 0 || GPUAlgo == 1)
	{
		cudaFree(pRefMap_device_Mod);
	}
599 600 601 602 603 604
	if (GPUAsync)
	{
		cudaStreamDestroy(cudaStream[2]);
		cudaEventDestroy(cudaEvent[2]);
	}

605
	delete gpumap;
606
	cudaThreadExit();
607

608 609 610 611 612 613
	deviceInitialized = 0;
	return(0);
}

int bioem_cuda::deviceStartRun()
{
David Rohr's avatar
David Rohr committed
614 615 616 617 618 619 620
	if (GPUWorkload >= 100)
	{
		maxRef = RefMap.ntotRefMap;
		pProb_host = &pProb;
	}
	else
	{
621
		maxRef = RefMap.ntotRefMap == 1 ? (size_t) RefMap.ntotRefMap : (size_t) RefMap.ntotRefMap * (size_t) GPUWorkload / 100;
David Rohr's avatar
David Rohr committed
622
		pProb_host = new bioem_Probability;
623
		pProb_host->init(maxRef, param.nTotGridAngles, param.nTotCC, *this);
David Rohr's avatar
David Rohr committed
624 625
		pProb_host->copyFrom(&pProb, *this);
	}
626

David Rohr's avatar
David Rohr committed
627 628 629
	pProb_device = *pProb_host;
	pProb_device.ptr = pProb_memory;
	pProb_device.set_pointers();
630
	checkCudaErrors(cudaMemcpyAsync(pProb_device.ptr, pProb_host->ptr, pProb_host->get_size(maxRef, param.nTotGridAngles, param.nTotCC, param.param_device.writeAngles, param.param_device.writeCC), cudaMemcpyHostToDevice, cudaStream[0]));
631 632 633

	if (FFTAlgo)
	{
634
		for (int j = 0;j < 2;j++)
635
		{
636
			for (int i = 0; i < 2; i++)
637
			{
638
				if (i && maxRef % CUDA_FFTS_AT_ONCE == 0) continue;
639 640 641 642 643 644
				int n[2] = {param.param_device.NumberPixels, param.param_device.NumberPixels};
				if (cufftPlanMany(&plan[i][j], 2, n, NULL, 1, param.FFTMapSize, NULL, 1, 0, MY_CUFFT_C2R, i ? (maxRef % CUDA_FFTS_AT_ONCE) : CUDA_FFTS_AT_ONCE) != CUFFT_SUCCESS)
				{
					cout << "Error planning CUFFT\n";
					exit(1);
				}
645
			        if (cufftSetCompatibilityMode(plan[i][j], CUFFT_COMPATIBILITY_FFTW_PADDING) != CUFFT_SUCCESS)
646 647 648 649 650 651 652 653 654
				{
					cout << "Error planning CUFFT compatibility\n";
					exit(1);
				}
				if (cufftSetStream(plan[i][j], cudaStream[j]) != CUFFT_SUCCESS)
				{
					cout << "Error setting CUFFT stream\n";
					exit(1);
				}
655
			}
656
			if (!GPUDualStream) break;
657 658
		}
	}
659 660 661 662 663
	return(0);
}

int bioem_cuda::deviceFinishRun()
{
664
	if (GPUAsync) cudaStreamSynchronize(cudaStream[0]);
665
	checkCudaErrors(cudaMemcpyAsync(pProb_host->ptr, pProb_device.ptr, pProb_host->get_size(maxRef, param.nTotGridAngles, param.nTotCC, param.param_device.writeAngles, param.param_device.writeCC), cudaMemcpyDeviceToHost, cudaStream[0]));
666

667 668
	if (FFTAlgo)
	{
669 670
		for (int j = 0;j < 2;j++)
		{
671 672 673 674 675
			for (int i = 0; i < 2; i++)
			{
				if (i && maxRef % CUDA_FFTS_AT_ONCE == 0) continue;
				cufftDestroy(plan[i][j]);
			}
676 677
			if (!GPUDualStream) break;
		}
678
	}
David Rohr's avatar
David Rohr committed
679 680 681 682
	cudaThreadSynchronize();
	if (GPUWorkload < 100)
	{
		pProb.copyFrom(pProb_host, *this);
683
		free_device_host(pProb_host->ptr);
David Rohr's avatar
David Rohr committed
684 685
		delete[] pProb_host;
	}
686

687 688 689
	return(0);
}

690 691 692 693 694 695 696 697 698 699 700 701
void* bioem_cuda::malloc_device_host(size_t size)
{
	void* ptr;
	checkCudaErrors(cudaHostAlloc(&ptr, size, 0));
	return(ptr);
}

void bioem_cuda::free_device_host(void* ptr)
{
	cudaFreeHost(ptr);
}

702 703
bioem* bioem_cuda_create()
{
David Rohr's avatar
David Rohr committed
704 705 706 707 708 709 710 711 712
	int count;
	
	if (cudaGetDeviceCount(&count) != cudaSuccess) count = 0;
	if (count == 0)
	{
		printf("No CUDA device available, using fallback to CPU version\n");
		return new bioem;
	}

713 714
	return new bioem_cuda;
}