static cl_int queue_whirlcoin_kernel(struct __clState *clState, struct _dev_blk_ctx *blk, __maybe_unused cl_uint threads)
{
cl_kernel *kernel;
cl_ulong le_target;
cl_int status = 0;
le_target = *(cl_ulong *)(blk->work->device_target + 24);
flip80(clState->cldata, blk->work->data);
status = clEnqueueWriteBuffer(clState->commandQueue, clState->CLbuffer0, true, 0, 80, clState->cldata, 0, NULL, NULL);
//clbuffer, hashes
kernel = &clState->kernel;
CL_SET_ARG_N(0, clState->CLbuffer0);
CL_SET_ARG_N(1, clState->padbuffer8);
kernel = clState->extra_kernels;
CL_SET_ARG_N(0, clState->padbuffer8);
CL_NEXTKERNEL_SET_ARG_N(0, clState->padbuffer8);
//hashes, output, target
CL_NEXTKERNEL_SET_ARG_N(0, clState->padbuffer8);
CL_SET_ARG_N(1, clState->outputBuffer);
CL_SET_ARG_N(2, le_target);
return status;
}
static cl_int queue_whirlpoolx_kernel(struct __clState *clState, struct _dev_blk_ctx *blk, __maybe_unused cl_uint threads)
{
uint64_t midblock[8], key[8] = { 0 }, tmp[8] = { 0 };
cl_ulong le_target;
cl_int status;
le_target = *(cl_ulong *)(blk->work->device_target + 24);
flip80(clState->cldata, blk->work->data);
memcpy(midblock, clState->cldata, 64);
// midblock = n, key = h
for (int i = 0; i < 10; ++i) {
tmp[0] = WHIRLPOOL_ROUND_CONSTANTS[i];
whirlpool_round(key, tmp);
tmp[0] = 0;
whirlpool_round(midblock, tmp);
for (int x = 0; x < 8; ++x) {
midblock[x] ^= key[x];
}
}
for (int i = 0; i < 8; ++i) {
midblock[i] ^= ((uint64_t *)(clState->cldata))[i];
}
status = clSetKernelArg(clState->kernel, 0, sizeof(cl_ulong8), (cl_ulong8 *)&midblock);
status |= clSetKernelArg(clState->kernel, 1, sizeof(cl_ulong), (void *)(((uint64_t *)clState->cldata) + 8));
status |= clSetKernelArg(clState->kernel, 2, sizeof(cl_ulong), (void *)(((uint64_t *)clState->cldata) + 9));
status |= clSetKernelArg(clState->kernel, 3, sizeof(cl_mem), (void *)&clState->outputBuffer);
status |= clSetKernelArg(clState->kernel, 4, sizeof(cl_ulong), (void *)&le_target);
return status;
}
static cl_int queue_yescrypt_kernel(_clState *clState, dev_blk_ctx *blk, __maybe_unused cl_uint threads)
{
cl_kernel *kernel = &clState->kernel;
unsigned int num = 0;
cl_uint le_target;
cl_int status = 0;
// le_target = (*(cl_uint *)(blk->work->device_target + 28));
le_target = (cl_uint)le32toh(((uint32_t *)blk->work->/*device_*/target)[7]);
// le_target = (cl_uint)((uint32_t *)blk->work->target)[7];
// memcpy(clState->cldata, blk->work->data, 80);
flip80(clState->cldata, blk->work->data);
status = clEnqueueWriteBuffer(clState->commandQueue, clState->CLbuffer0, true, 0, 80, clState->cldata, 0, NULL, NULL);
CL_SET_ARG(clState->CLbuffer0);
CL_SET_ARG(clState->outputBuffer);
CL_SET_ARG(clState->padbuffer8);
CL_SET_ARG(clState->buffer1);
CL_SET_ARG(clState->buffer2);
CL_SET_ARG(le_target);
return status;
}
static cl_int queue_fresh_kernel(struct __clState *clState, struct _dev_blk_ctx *blk, __maybe_unused cl_uint threads)
{
cl_kernel *kernel;
unsigned int num;
cl_ulong le_target;
cl_int status = 0;
le_target = *(cl_ulong *)(blk->work->device_target + 24);
flip80(clState->cldata, blk->work->data);
status = clEnqueueWriteBuffer(clState->commandQueue, clState->CLbuffer0, true, 0, 80, clState->cldata, 0, NULL, NULL);
// shavite 1 - search
kernel = &clState->kernel;
num = 0;
CL_SET_ARG(clState->CLbuffer0);
CL_SET_ARG(clState->padbuffer8);
// smid 1 - search1
kernel = clState->extra_kernels;
CL_SET_ARG_0(clState->padbuffer8);
// shavite 2 - search2
CL_NEXTKERNEL_SET_ARG_0(clState->padbuffer8);
// smid 2 - search3
CL_NEXTKERNEL_SET_ARG_0(clState->padbuffer8);
// echo - search4
num = 0;
CL_NEXTKERNEL_SET_ARG(clState->padbuffer8);
CL_SET_ARG(clState->outputBuffer);
CL_SET_ARG(le_target);
return status;
}
int test_nonce(struct mm_work *mw, uint32_t nonce2, uint32_t nonce)
{
/* Generate the work base on nonce2 */
struct work work;
debug32("Test: %08x %08x\n", nonce2, nonce);
miner_gen_nonce2_work(mw, nonce2, &work);
/* Write the nonce to block header */
uint32_t *work_nonce = (uint32_t *)(work.header + 64 + 12);
*work_nonce = bswap_32(nonce);
/* Regen hash */
uint32_t *data32 = (uint32_t *)(work.header);
unsigned char swap[80];
uint32_t *swap32 = (uint32_t *)swap;
unsigned char hash1[32];
uint32_t *hash_32 = (uint32_t *)(hash1 + 28);
flip80(swap32, data32);
dsha256(swap, 80, hash1);
if (*hash_32 != 0)
return NONCE_HW;
/* Compare hash with target */
return fulltest(hash1, mw->target);
}
void skeincoin_prepare_work(dev_blk_ctx *blk, uint32_t *state, uint32_t *pdata)
{
uint32_t be_data[20];
flip80(be_data, pdata);
skeinmid((unsigned char *)blk->ulongMidstate, (const unsigned char *)be_data);
memcpy(blk->ulongData, be_data + 16, 12);
}
void skeincoin_regenhash(struct work *work)
{
uint32_t be_data[20];
unsigned char hash1[64];
flip80(be_data, work->data);
skeincoin(hash1, (const unsigned char *)be_data, 80);
sha256(hash1, 64, work->hash);
}
static cl_int queue_lyra2rev2_kernel(struct __clState *clState, struct _dev_blk_ctx *blk, __maybe_unused cl_uint threads)
{
cl_kernel *kernel;
unsigned int num;
cl_int status = 0;
cl_ulong le_target;
// le_target = *(cl_uint *)(blk->work->device_target + 28);
le_target = *(cl_ulong *)(blk->work->device_target + 24);
flip80(clState->cldata, blk->work->data);
status = clEnqueueWriteBuffer(clState->commandQueue, clState->CLbuffer0, true, 0, 80, clState->cldata, 0, NULL, NULL);
// blake - search
kernel = &clState->kernel;
num = 0;
// CL_SET_ARG(clState->CLbuffer0);
CL_SET_ARG(clState->buffer1);
CL_SET_ARG(blk->work->blk.ctx_a);
CL_SET_ARG(blk->work->blk.ctx_b);
CL_SET_ARG(blk->work->blk.ctx_c);
CL_SET_ARG(blk->work->blk.ctx_d);
CL_SET_ARG(blk->work->blk.ctx_e);
CL_SET_ARG(blk->work->blk.ctx_f);
CL_SET_ARG(blk->work->blk.ctx_g);
CL_SET_ARG(blk->work->blk.ctx_h);
CL_SET_ARG(blk->work->blk.cty_a);
CL_SET_ARG(blk->work->blk.cty_b);
CL_SET_ARG(blk->work->blk.cty_c);
// keccak - search1
kernel = clState->extra_kernels;
CL_SET_ARG_0(clState->buffer1);
// cubehash - search2
num = 0;
CL_NEXTKERNEL_SET_ARG_0(clState->buffer1);
// lyra - search3
num = 0;
CL_NEXTKERNEL_SET_ARG_N(0, clState->buffer1);
CL_SET_ARG_N(1, clState->padbuffer8);
// skein -search4
num = 0;
CL_NEXTKERNEL_SET_ARG_0(clState->buffer1);
// cubehash - search5
num = 0;
CL_NEXTKERNEL_SET_ARG_0(clState->buffer1);
// bmw - search6
num = 0;
CL_NEXTKERNEL_SET_ARG(clState->buffer1);
CL_SET_ARG(clState->outputBuffer);
CL_SET_ARG(le_target);
return status;
}
static cl_int queue_bitblock_kernel(struct __clState *clState, struct _dev_blk_ctx *blk, __maybe_unused cl_uint threads)
{
cl_kernel *kernel;
unsigned int num;
cl_ulong le_target;
cl_int status = 0;
le_target = *(cl_ulong *)(blk->work->device_target + 24);
flip80(clState->cldata, blk->work->data);
status = clEnqueueWriteBuffer(clState->commandQueue, clState->CLbuffer0, true, 0, 80, clState->cldata, 0, NULL, NULL);
// blake - search
kernel = &clState->kernel;
num = 0;
CL_SET_ARG(clState->CLbuffer0);
CL_SET_ARG(clState->padbuffer8);
// bmw - search1
kernel = clState->extra_kernels;
CL_SET_ARG_0(clState->padbuffer8);
// groestl - search2
CL_NEXTKERNEL_SET_ARG_0(clState->padbuffer8);
// skein - search3
CL_NEXTKERNEL_SET_ARG_0(clState->padbuffer8);
// jh - search4
CL_NEXTKERNEL_SET_ARG_0(clState->padbuffer8);
// keccak - search5
CL_NEXTKERNEL_SET_ARG_0(clState->padbuffer8);
// luffa - search6
CL_NEXTKERNEL_SET_ARG_0(clState->padbuffer8);
// cubehash - search7
CL_NEXTKERNEL_SET_ARG_0(clState->padbuffer8);
// shavite - search8
CL_NEXTKERNEL_SET_ARG_0(clState->padbuffer8);
// simd - search9
CL_NEXTKERNEL_SET_ARG_0(clState->padbuffer8);
// echo - search10
CL_NEXTKERNEL_SET_ARG_0(clState->padbuffer8);
// hamsi - search11
CL_NEXTKERNEL_SET_ARG_0(clState->padbuffer8);
// fugue - search12
CL_NEXTKERNEL_SET_ARG_0(clState->padbuffer8);
// hamsi - search11
CL_NEXTKERNEL_SET_ARG_0(clState->padbuffer8);
// fugue - search12
num = 0;
CL_NEXTKERNEL_SET_ARG(clState->padbuffer8);
CL_SET_ARG(clState->outputBuffer);
CL_SET_ARG(le_target);
return status;
}
static cl_int queue_maxcoin_kernel(struct __clState *clState, struct _dev_blk_ctx *blk, __maybe_unused cl_uint threads)
{
cl_kernel *kernel = &clState->kernel;
unsigned int num = 0;
cl_int status = 0;
flip80(clState->cldata, blk->work->data);
status = clEnqueueWriteBuffer(clState->commandQueue, clState->CLbuffer0, true, 0, 80, clState->cldata, 0, NULL, NULL);
CL_SET_ARG(clState->CLbuffer0);
CL_SET_ARG(clState->outputBuffer);
return status;
}
void hefty_regenhash(struct work *work)
{
uint32_t result[8];
unsigned int data[20], datacopy[20]; // 32-aligned
memcpy(datacopy, work->data, 80);
flip80(data, datacopy);
char *hdata = bin2hex((const unsigned char *)data, 80);
applog(LOG_DEBUG, "Verifying hefty data %s", hdata);
free(hdata);
heavycoin_hash((const char*)data, 80, (char*)&result);
memcpy(work->hash, &result, 32);
}
static cl_int queue_sph_kernel(_clState *clState, dev_blk_ctx *blk, __maybe_unused cl_uint threads)
{
unsigned char *midstate = blk->work->midstate;
cl_kernel *kernel = &clState->kernel;
unsigned int num = 0;
cl_ulong le_target;
cl_int status = 0;
le_target = *(cl_ulong *)(blk->work->device_target + 24);
flip80(clState->cldata, blk->work->data);
status = clEnqueueWriteBuffer(clState->commandQueue, clState->CLbuffer0, true, 0, 80, clState->cldata, 0, NULL,NULL);
CL_SET_ARG(clState->CLbuffer0);
CL_SET_ARG(clState->outputBuffer);
CL_SET_ARG(le_target);
return status;
}
