561 lines
23 KiB
C++
561 lines
23 KiB
C++
#include <stdint.h>
|
|
#include <vx_intrinsics.h>
|
|
#include <vx_print.h>
|
|
#include <vx_spawn.h>
|
|
#include "common.h"
|
|
#include "include/gemmini.h"
|
|
#include "gemmini_mmio.h"
|
|
|
|
#define TILE_M 32
|
|
#define TILE_N 32
|
|
#define TILE_K 32
|
|
#define TILE_MN 1024
|
|
#define TILE_MK 1024
|
|
#define TILE_NK 1024
|
|
|
|
#define NUM_CLUSTERS 1
|
|
#define NUM_THREADS_IN_CLUSTER 128
|
|
|
|
#define SMEM_ADDR_0K ((float * const) 0xff000000)
|
|
#define SMEM_ADDR_4K ((float * const) 0xff001000)
|
|
#define SMEM_ADDR_8K ((float * const) 0xff002000)
|
|
#define SMEM_ADDR_12K ((float * const) 0xff003000)
|
|
#define SPAD_ADDR_0K 0x0
|
|
#define SPAD_ADDR_4K 0x80
|
|
#define SPAD_ADDR_8K 0x100
|
|
#define SPAD_ADDR_12K 0x180
|
|
|
|
//#define DEBUG_PRINT
|
|
//#define EXT_ACCUMULATE
|
|
#define HARDCODE
|
|
#define REGBLOCK
|
|
#define DBUF
|
|
//#define DETAILED_PERF
|
|
//#define ACTIVATE
|
|
#define CISC
|
|
|
|
#define rd_cycles_force(x) asm volatile ("csrr %0, mcycle" : "=r" (x))
|
|
#ifdef DETAILED_PERF
|
|
#define rd_cycles(x) rd_cycles_force(x)
|
|
#else
|
|
#define rd_cycles(x)
|
|
#endif
|
|
#define HW_TID() ({uint32_t gtid; asm volatile ("csrr %0, mhartid" : "=r" (gtid)); gtid;})
|
|
#define PRINTF(...) sprintf(PRINT_BUF, __VA_ARGS__)
|
|
// #define PRINTF(...) vx_printf(__VA_ARGS__)
|
|
#define SWISH(beta, x) ((x) / (1 + exp(-(beta) * (x))))
|
|
|
|
inline void threadblock_barrier(unsigned int barrier_id, unsigned int count) {
|
|
vx_fence();
|
|
vx_barrier(barrier_id, count);
|
|
}
|
|
|
|
void thread_block_matmul_gemmini(kernel_arg_t *__UNIFORM__ arg,
|
|
const uint32_t threadblock_id,
|
|
const uint32_t tid_in_threadblock) {
|
|
__asm__("matmul_start:");
|
|
const float * const A = (const float * const) arg->addr_a;
|
|
const float * const B = (const float * const) arg->addr_b;
|
|
float * const C = (float * const) arg->addr_c;
|
|
|
|
if (HW_TID() == 0) {
|
|
gemmini_config_ld(0);
|
|
gemmini_extended_config_ex(WEIGHT_STATIONARY, 0, 0, 1, 0, 0);
|
|
gemmini_config_st(0);
|
|
PRINTF("start\n");
|
|
}
|
|
|
|
vx_fence();
|
|
|
|
uint32_t marker0, marker1, marker2, marker3, marker4;
|
|
uint32_t marker5, marker6, marker7, marker8, marker9;
|
|
#ifdef ACTIVATE
|
|
uint32_t swish_dur = 0;
|
|
#endif
|
|
rd_cycles_force(marker0);
|
|
|
|
const uint32_t dim_m = arg->dim_m;
|
|
const uint32_t dim_n = arg->dim_n;
|
|
const uint32_t dim_k = arg->dim_k;
|
|
const uint32_t num_tiles_m = dim_m / TILE_M;
|
|
const uint32_t num_tiles_n = dim_n / TILE_N;
|
|
const uint32_t num_tiles_k = dim_k / TILE_K;
|
|
constexpr uint32_t num_threads_in_cluster = NUM_THREADS_IN_CLUSTER;
|
|
constexpr uint32_t a_elems_per_thread = TILE_MK / num_threads_in_cluster;
|
|
constexpr uint32_t b_elems_per_thread = TILE_NK / num_threads_in_cluster;
|
|
constexpr uint32_t c_elems_per_thread = TILE_MN / num_threads_in_cluster;
|
|
const uint32_t hw_tid = tid_in_threadblock % num_threads_in_cluster;
|
|
|
|
// the dram coordinates are (i1 + i0, j1 + j0). i0 and j0 are both spatially mapped only.
|
|
const uint32_t j0 = HW_TID() % DIM;
|
|
const uint32_t i0 = (HW_TID() / DIM) % DIM;
|
|
|
|
// j1 is both spatially and temporally mapped. j1 increases every iteration.
|
|
const uint32_t j1_idx = (HW_TID() / DIM / DIM) * DIM; // A: % TILE_K, B: % TILE_N, C: % TILE_N
|
|
// every iteratioon, j1 increases by j1_stride
|
|
constexpr uint32_t j1_stride = (num_threads_in_cluster / DIM / DIM) * DIM; // mod TILE_W after stride
|
|
|
|
// i1 is only temporally mapped. i1 increments every one or more iterations
|
|
constexpr uint32_t i1_stride = DIM; // step per increment (increment doesnt happen every iteration)
|
|
constexpr uint32_t i1_iters = (DIM * DIM * (TILE_K / DIM)) / num_threads_in_cluster; // num of iters before striding
|
|
|
|
const uint32_t num_tile_rows_per_tb = num_tiles_m / NUM_CLUSTERS;
|
|
for (uint32_t tile_i = num_tile_rows_per_tb * threadblock_id;
|
|
tile_i < num_tile_rows_per_tb * (threadblock_id + 1);
|
|
tile_i += 1) {
|
|
__asm__("i_loop:");
|
|
for (int tile_j = 0; tile_j < num_tiles_n; tile_j += 1) {
|
|
__asm__("j_loop:");
|
|
float * const smem_c_tile_start = SMEM_ADDR_4K;
|
|
#ifndef EXT_ACCUMULATE
|
|
float * const smem_acc_tile_start = SMEM_ADDR_0K + HW_TID();
|
|
#else
|
|
float * const smem_acc_tile_start = SMEM_ADDR_8K + hw_tid;
|
|
#endif
|
|
|
|
__asm__("k_loop:");
|
|
for (int tile_k = 0; tile_k < num_tiles_k; tile_k += 1) {
|
|
// TODO: double buffer
|
|
rd_cycles(marker1);
|
|
|
|
#ifdef HARDCODE
|
|
#if (TILE_MK / NUM_THREADS / NUM_WARPS / CORES_PER_CLUSTER) != 8
|
|
#error CANNOT UNROLL
|
|
#endif
|
|
|
|
constexpr uint32_t every_iter = j1_stride;
|
|
const uint32_t every_2iters_a = i1_stride * dim_k;
|
|
const uint32_t runtime_const_a = i0 * dim_k + j1_idx + j0;
|
|
const uint32_t every_2iters_b = i1_stride * dim_n;
|
|
const uint32_t runtime_const_b = i0 * dim_n + j1_idx + j0;
|
|
|
|
const float * const dram_a_tile_start = A + tile_i * TILE_M * dim_k + tile_k * TILE_K + runtime_const_a;
|
|
const float * const dram_b_tile_start = B + tile_k * TILE_K * dim_n + tile_j * TILE_N + runtime_const_b;
|
|
#ifdef DBUF
|
|
float * const smem_a_tile_start = ((tile_k & 1) ? SMEM_ADDR_4K : SMEM_ADDR_0K) + HW_TID();
|
|
float * const smem_b_tile_start = ((tile_k & 1) ? SMEM_ADDR_12K : SMEM_ADDR_8K) + HW_TID();
|
|
#else
|
|
float * const smem_a_tile_start = SMEM_ADDR_0K + HW_TID();
|
|
float * const smem_b_tile_start = SMEM_ADDR_12K + HW_TID();
|
|
#endif
|
|
|
|
{
|
|
#ifndef REGBLOCK
|
|
smem_a_tile_start[0 * num_threads_in_cluster + hw_tid] = \
|
|
dram_a_tile_start[every_iter * 0 + every_2iters_a * 0];
|
|
smem_a_tile_start[1 * num_threads_in_cluster + hw_tid] = \
|
|
dram_a_tile_start[every_iter * 1 + every_2iters_a * 0];
|
|
smem_a_tile_start[2 * num_threads_in_cluster + hw_tid] = \
|
|
dram_a_tile_start[every_iter * 0 + every_2iters_a * 1];
|
|
smem_a_tile_start[3 * num_threads_in_cluster + hw_tid] = \
|
|
dram_a_tile_start[every_iter * 1 + every_2iters_a * 1];
|
|
smem_a_tile_start[4 * num_threads_in_cluster + hw_tid] = \
|
|
dram_a_tile_start[every_iter * 0 + every_2iters_a * 2];
|
|
smem_a_tile_start[5 * num_threads_in_cluster + hw_tid] = \
|
|
dram_a_tile_start[every_iter * 1 + every_2iters_a * 2];
|
|
smem_a_tile_start[6 * num_threads_in_cluster + hw_tid] = \
|
|
dram_a_tile_start[every_iter * 0 + every_2iters_a * 3];
|
|
smem_a_tile_start[7 * num_threads_in_cluster + hw_tid] = \
|
|
dram_a_tile_start[every_iter * 1 + every_2iters_a * 3];
|
|
|
|
smem_b_tile_start[0 * num_threads_in_cluster + hw_tid] = \
|
|
dram_b_tile_start[every_iter * 0 + every_2iters_b * 0];
|
|
smem_b_tile_start[1 * num_threads_in_cluster + hw_tid] = \
|
|
dram_b_tile_start[every_iter * 1 + every_2iters_b * 0];
|
|
smem_b_tile_start[2 * num_threads_in_cluster + hw_tid] = \
|
|
dram_b_tile_start[every_iter * 0 + every_2iters_b * 1];
|
|
smem_b_tile_start[3 * num_threads_in_cluster + hw_tid] = \
|
|
dram_b_tile_start[every_iter * 1 + every_2iters_b * 1];
|
|
smem_b_tile_start[4 * num_threads_in_cluster + hw_tid] = \
|
|
dram_b_tile_start[every_iter * 0 + every_2iters_b * 2];
|
|
smem_b_tile_start[5 * num_threads_in_cluster + hw_tid] = \
|
|
dram_b_tile_start[every_iter * 1 + every_2iters_b * 2];
|
|
smem_b_tile_start[6 * num_threads_in_cluster + hw_tid] = \
|
|
dram_b_tile_start[every_iter * 0 + every_2iters_b * 3];
|
|
smem_b_tile_start[7 * num_threads_in_cluster + hw_tid] = \
|
|
dram_b_tile_start[every_iter * 1 + every_2iters_b * 3];
|
|
#else
|
|
__asm__("load_ab:");
|
|
float v0 = dram_a_tile_start[every_iter * 0 + every_2iters_a * 0];
|
|
float v1 = dram_a_tile_start[every_iter * 1 + every_2iters_a * 0];
|
|
float v2 = dram_a_tile_start[every_iter * 0 + every_2iters_a * 1];
|
|
float v3 = dram_a_tile_start[every_iter * 1 + every_2iters_a * 1];
|
|
smem_a_tile_start[0 * num_threads_in_cluster] = v0;
|
|
smem_a_tile_start[1 * num_threads_in_cluster] = v1;
|
|
smem_a_tile_start[2 * num_threads_in_cluster] = v2;
|
|
smem_a_tile_start[3 * num_threads_in_cluster] = v3;
|
|
|
|
__asm__("load_ab1:");
|
|
v0 = dram_b_tile_start[every_iter * 0 + every_2iters_b * 0];
|
|
v1 = dram_b_tile_start[every_iter * 1 + every_2iters_b * 0];
|
|
v2 = dram_b_tile_start[every_iter * 0 + every_2iters_b * 1];
|
|
v3 = dram_b_tile_start[every_iter * 1 + every_2iters_b * 1];
|
|
smem_b_tile_start[0 * num_threads_in_cluster] = v0;
|
|
smem_b_tile_start[1 * num_threads_in_cluster] = v1;
|
|
smem_b_tile_start[2 * num_threads_in_cluster] = v2;
|
|
smem_b_tile_start[3 * num_threads_in_cluster] = v3;
|
|
|
|
__asm__("load_ab2:");
|
|
v0 = dram_a_tile_start[every_iter * 0 + every_2iters_a * 2];
|
|
v1 = dram_a_tile_start[every_iter * 1 + every_2iters_a * 2];
|
|
v2 = dram_a_tile_start[every_iter * 0 + every_2iters_a * 3];
|
|
v3 = dram_a_tile_start[every_iter * 1 + every_2iters_a * 3];
|
|
smem_a_tile_start[4 * num_threads_in_cluster] = v0;
|
|
smem_a_tile_start[5 * num_threads_in_cluster] = v1;
|
|
smem_a_tile_start[6 * num_threads_in_cluster] = v2;
|
|
smem_a_tile_start[7 * num_threads_in_cluster] = v3;
|
|
|
|
__asm__("load_ab3:");
|
|
v0 = dram_b_tile_start[every_iter * 0 + every_2iters_b * 2];
|
|
v1 = dram_b_tile_start[every_iter * 1 + every_2iters_b * 2];
|
|
v2 = dram_b_tile_start[every_iter * 0 + every_2iters_b * 3];
|
|
v3 = dram_b_tile_start[every_iter * 1 + every_2iters_b * 3];
|
|
smem_b_tile_start[4 * num_threads_in_cluster] = v0;
|
|
smem_b_tile_start[5 * num_threads_in_cluster] = v1;
|
|
smem_b_tile_start[6 * num_threads_in_cluster] = v2;
|
|
smem_b_tile_start[7 * num_threads_in_cluster] = v3;
|
|
|
|
__asm__("end_loadab:");
|
|
#endif
|
|
}
|
|
#else
|
|
__asm__("loop_load_ab:");
|
|
|
|
const float * const dram_a_tile_start = A + tile_i * TILE_M * dim_k + tile_k * TILE_K;
|
|
const float * const dram_b_tile_start = B + tile_k * TILE_K * dim_n + tile_j * TILE_N;
|
|
float * const smem_a_tile_start = SMEM_ADDR_0K;
|
|
float * const smem_b_tile_start = SMEM_ADDR_12K;
|
|
|
|
/* for (uint32_t thread_i = 0, j1 = 0, i1 = 0;
|
|
thread_i < a_elems_per_thread;
|
|
thread_i += 1,
|
|
j1 = (j1 + j1_stride) % TILE_K,
|
|
i1 = (thread_i % i1_iters == 0) ? i1 + i1_stride : i1) {
|
|
smem_a_tile_start[thread_i * num_threads_in_cluster + hw_tid] = \
|
|
dram_a_tile_start[(0 + i0) * dim_k + j1 + j1_idx + j0];
|
|
} */
|
|
#pragma clang loop unroll(disable)
|
|
for (int thread_i = 0; thread_i < a_elems_per_thread; thread_i++) {
|
|
uint32_t elem_offset = hw_tid + num_threads_in_cluster * thread_i;
|
|
smem_a_tile_start[SMEM_MAT_OFFSET(elem_offset / TILE_K, elem_offset % TILE_K, TILE_K)] = \
|
|
dram_a_tile_start[elem_offset / TILE_K * dim_k + elem_offset % TILE_K];
|
|
}
|
|
__asm__("loop_load_a_end:");
|
|
#pragma clang loop unroll(disable)
|
|
for (int thread_i = 0; thread_i < b_elems_per_thread; thread_i++) {
|
|
uint32_t elem_offset = hw_tid + num_threads_in_cluster * thread_i;
|
|
smem_b_tile_start[SMEM_MAT_OFFSET(elem_offset / TILE_N, elem_offset % TILE_N, TILE_N)] = \
|
|
dram_b_tile_start[elem_offset / TILE_N * dim_n + elem_offset % TILE_N];
|
|
}
|
|
#endif
|
|
|
|
#ifdef DEBUG_PRINT
|
|
if (hw_tid == 0) {
|
|
PRINTF("\nA %d %d\n", tile_i, tile_k);
|
|
for (int i = 0; i < TILE_M; i += 8) {
|
|
for (int j = 0; j < TILE_K; j += 8) {
|
|
uint32_t mat_offset = SMEM_MAT_OFFSET(i, j, TILE_K);
|
|
PRINTF("%x %x ",
|
|
(int) (smem_a_tile_start[mat_offset]),
|
|
(int) (smem_a_tile_start[mat_offset + 4])
|
|
);
|
|
}
|
|
PRINTF("\n");
|
|
}
|
|
PRINTF("\nB %d %d\n", tile_k, tile_j);
|
|
for (int i = 0; i < TILE_K; i += 8) {
|
|
for (int j = 0; j < TILE_N; j += 8) {
|
|
uint32_t mat_offset = SMEM_MAT_OFFSET(i, j, TILE_N);
|
|
PRINTF("%x %x ",
|
|
(int) (smem_b_tile_start[mat_offset]),
|
|
(int) (smem_b_tile_start[mat_offset + 4])
|
|
);
|
|
}
|
|
PRINTF("\n");
|
|
}
|
|
}
|
|
#endif
|
|
|
|
|
|
rd_cycles(marker2);
|
|
// cluster wide barrier to wait for A and B loads to complete
|
|
threadblock_barrier(/*barrier_id=*/0, /*count=*/NUM_WARPS);
|
|
rd_cycles(marker3);
|
|
__asm__("gemmini:");
|
|
if (HW_TID() == 0) {
|
|
#ifdef DBUF
|
|
gemmini_fence();
|
|
#endif
|
|
#ifdef CISC
|
|
#ifndef DBUF
|
|
#error MUST ENABLE DBUF
|
|
#endif
|
|
#ifdef EXT_ACCUMULATE
|
|
#error MUST DISABLE EXT ACCUMULATE
|
|
#endif
|
|
if (tile_k == 0) {
|
|
GEMMINI_CISC_CMD_I(0);
|
|
} else if (tile_k & 1) {
|
|
GEMMINI_CISC_CMD_I(2);
|
|
} else {
|
|
GEMMINI_CISC_CMD_I(1);
|
|
}
|
|
#else
|
|
sp_tiled_matmul_full_spad_ws(
|
|
#ifdef DBUF
|
|
(tile_k & 1) ? SPAD_ADDR_4K : SPAD_ADDR_0K, (tile_k & 1) ? SPAD_ADDR_12K : SPAD_ADDR_8K,
|
|
#else
|
|
SPAD_ADDR_0K, SPAD_ADDR_12K,
|
|
#endif
|
|
/*spad_D=*/0, /*spad_C=*/SPAD_ADDR_4K,
|
|
/*I=*/TILE_M / DIM, /*J=*/TILE_N / DIM, /*K=*/TILE_K / DIM, /*pad_I=*/0, /*pad_J=*/0, /*pad_K=*/0,
|
|
/*a_transpose=*/0, /*b_transpose=*/0, /*full_C=*/0, /*low_D=*/0,
|
|
#ifdef EXT_ACCUMULATE
|
|
/*acc=*/0, /*act=*/NO_ACTIVATION, /*skips=*/0x38U)
|
|
#else
|
|
/*acc=*/tile_k != 0, /*act=*/NO_ACTIVATION, /*skips=*/0xB8U)
|
|
#endif
|
|
|
|
#endif
|
|
#ifndef DBUF
|
|
gemmini_fence();
|
|
#endif
|
|
}
|
|
__asm__("end_gemmini:");
|
|
rd_cycles(marker4);
|
|
// threadblock_barrier(/*barrier_id=*/0, /*count=*/NUM_WARPS);
|
|
rd_cycles(marker5);
|
|
|
|
// accumulate C matrix
|
|
#ifdef EXT_ACCUMULATE
|
|
__asm__("accumulate:");
|
|
if (tile_k == 0) {
|
|
#pragma GCC ivdep
|
|
#pragma GCC unroll 8
|
|
for (int thread_i = 0; thread_i < c_elems_per_thread; thread_i++) {
|
|
constexpr uint32_t s = num_threads_in_cluster;
|
|
smem_acc_tile_start[thread_i * s] = smem_c_tile_start[hw_tid + s * thread_i];
|
|
}
|
|
} else {
|
|
#if (TILE_NK / NUM_THREADS / NUM_WARPS / CORES_PER_CLUSTER) != 8
|
|
#error CANNOT UNROLL
|
|
#endif
|
|
for (int thread_i = 0; thread_i < c_elems_per_thread; thread_i += 8) {
|
|
constexpr uint32_t s = num_threads_in_cluster;
|
|
smem_acc_tile_start[s * 0] += smem_c_tile_start[hw_tid + s * 0];
|
|
smem_acc_tile_start[s * 1] += smem_c_tile_start[hw_tid + s * 1];
|
|
smem_acc_tile_start[s * 2] += smem_c_tile_start[hw_tid + s * 2];
|
|
smem_acc_tile_start[s * 3] += smem_c_tile_start[hw_tid + s * 3];
|
|
smem_acc_tile_start[s * 4] += smem_c_tile_start[hw_tid + s * 4];
|
|
smem_acc_tile_start[s * 5] += smem_c_tile_start[hw_tid + s * 5];
|
|
smem_acc_tile_start[s * 6] += smem_c_tile_start[hw_tid + s * 6];
|
|
smem_acc_tile_start[s * 7] += smem_c_tile_start[hw_tid + s * 7];
|
|
}
|
|
}
|
|
__asm__("end_accumulate:");
|
|
#endif
|
|
|
|
#ifdef DEBUG_PRINT
|
|
if (hw_tid == 0) {
|
|
PRINTF("\nC %d %d %d\n", tile_i, tile_j, tile_k);
|
|
for (int i = 0; i < TILE_M; i += 8) {
|
|
for (int j = 0; j < TILE_N; j += 8) {
|
|
uint32_t mat_offset = SMEM_MAT_OFFSET(i, j, TILE_N);
|
|
PRINTF("%d %d ",
|
|
(int) (smem_c_tile_start[mat_offset]),
|
|
(int) (smem_c_tile_start[mat_offset + 4])
|
|
);
|
|
}
|
|
PRINTF("\n");
|
|
}
|
|
}
|
|
#endif
|
|
rd_cycles(marker6);
|
|
|
|
/* if (HW_TID() == 0) {
|
|
PRINTF("\ntile start: %d\n", marker1);
|
|
PRINTF("single tile cycles: %d\n", marker6 - marker1);
|
|
PRINTF("A/B tile load cycles: %d\n", marker2 - marker1);
|
|
PRINTF("first barrier: %d\n", marker3 - marker2);
|
|
PRINTF("gemmini cycles: %d\n", marker4 - marker3);
|
|
PRINTF("second barrier: %d\n", marker5 - marker4);
|
|
} */
|
|
|
|
}
|
|
|
|
#ifndef EXT_ACCUMULATE
|
|
threadblock_barrier(/*barrier_id=*/0, /*count=*/NUM_WARPS);
|
|
rd_cycles(marker6);
|
|
__asm__("mvout_spad_ser:");
|
|
// mvout to scratchpad for activation
|
|
if (HW_TID() == 0) {
|
|
__asm__("mvout_spad:");
|
|
// #ifdef DBUF
|
|
// gemmini_fence();
|
|
// #endif
|
|
#ifdef CISC
|
|
GEMMINI_CISC_CMD_I(9);
|
|
#else
|
|
ROCC_INSTRUCTION_RS1_RS2(XCUSTOM_ACC, 0, (4ULL << 32) | (4ULL << 16) | 4ULL, k_LOOP_WS_CONFIG_BOUNDS)
|
|
ROCC_INSTRUCTION_RS1_RS2(XCUSTOM_ACC, 0, 0x278U, k_LOOP_WS)
|
|
#endif
|
|
__asm__("mvout_spad_fence:");
|
|
gemmini_fence();
|
|
}
|
|
__asm__("mvout_spad_bar:");
|
|
threadblock_barrier(/*barrier_id=*/0, /*count=*/NUM_WARPS);
|
|
__asm__("end_mvout_spad:");
|
|
#endif
|
|
rd_cycles(marker7);
|
|
|
|
// move out to dram
|
|
__asm__("mvout_dram:");
|
|
#ifdef HARDCODE
|
|
#if (TILE_MN / NUM_THREADS / NUM_WARPS / CORES_PER_CLUSTER) != 8
|
|
#error CANNOT UNROLL
|
|
#endif
|
|
constexpr uint32_t every_iter = j1_stride;
|
|
const uint32_t every_2iters = i1_stride * dim_n;
|
|
const uint32_t runtime_const = i0 * dim_n + j1_idx + j0;
|
|
float * const dram_c_tile_start = C + tile_i * TILE_M * dim_n + tile_j * TILE_N + runtime_const;
|
|
|
|
#ifdef REGBLOCK
|
|
float v0 = smem_acc_tile_start[0 * num_threads_in_cluster];
|
|
float v1 = smem_acc_tile_start[1 * num_threads_in_cluster];
|
|
float v2 = smem_acc_tile_start[2 * num_threads_in_cluster];
|
|
float v3 = smem_acc_tile_start[3 * num_threads_in_cluster];
|
|
#ifdef ACTIVATE
|
|
uint32_t swish_start, swish_end;
|
|
rd_cycles_force(swish_start);
|
|
v0 = SWISH(1, v0);
|
|
v1 = SWISH(1, v1);
|
|
v2 = SWISH(1, v2);
|
|
v3 = SWISH(1, v3);
|
|
rd_cycles_force(swish_end);
|
|
swish_dur += swish_end - swish_start;
|
|
#endif
|
|
dram_c_tile_start[every_iter * 0 + every_2iters * 0] = v0;
|
|
dram_c_tile_start[every_iter * 1 + every_2iters * 0] = v1;
|
|
dram_c_tile_start[every_iter * 0 + every_2iters * 1] = v2;
|
|
dram_c_tile_start[every_iter * 1 + every_2iters * 1] = v3;
|
|
|
|
v0 = smem_acc_tile_start[4 * num_threads_in_cluster];
|
|
v1 = smem_acc_tile_start[5 * num_threads_in_cluster];
|
|
v2 = smem_acc_tile_start[6 * num_threads_in_cluster];
|
|
v3 = smem_acc_tile_start[7 * num_threads_in_cluster];
|
|
#ifdef ACTIVATE
|
|
rd_cycles_force(swish_start);
|
|
v0 = SWISH(1, v0);
|
|
v1 = SWISH(1, v1);
|
|
v2 = SWISH(1, v2);
|
|
v3 = SWISH(1, v3);
|
|
rd_cycles_force(swish_end);
|
|
swish_dur += swish_end - swish_start;
|
|
#endif
|
|
dram_c_tile_start[every_iter * 0 + every_2iters * 2] = v0;
|
|
dram_c_tile_start[every_iter * 1 + every_2iters * 2] = v1;
|
|
dram_c_tile_start[every_iter * 0 + every_2iters * 3] = v2;
|
|
dram_c_tile_start[every_iter * 1 + every_2iters * 3] = v3;
|
|
#else
|
|
dram_c_tile_start[every_iter * 0 + every_2iters * 0] = \
|
|
smem_acc_tile_start[0 * num_threads_in_cluster];
|
|
dram_c_tile_start[every_iter * 1 + every_2iters * 0] = \
|
|
smem_acc_tile_start[1 * num_threads_in_cluster];
|
|
dram_c_tile_start[every_iter * 0 + every_2iters * 1] = \
|
|
smem_acc_tile_start[2 * num_threads_in_cluster];
|
|
dram_c_tile_start[every_iter * 1 + every_2iters * 1] = \
|
|
smem_acc_tile_start[3 * num_threads_in_cluster];
|
|
dram_c_tile_start[every_iter * 0 + every_2iters * 2] = \
|
|
smem_acc_tile_start[4 * num_threads_in_cluster];
|
|
dram_c_tile_start[every_iter * 1 + every_2iters * 2] = \
|
|
smem_acc_tile_start[5 * num_threads_in_cluster];
|
|
dram_c_tile_start[every_iter * 0 + every_2iters * 3] = \
|
|
smem_acc_tile_start[6 * num_threads_in_cluster];
|
|
dram_c_tile_start[every_iter * 1 + every_2iters * 3] = \
|
|
smem_acc_tile_start[7 * num_threads_in_cluster];
|
|
#endif
|
|
|
|
#else
|
|
|
|
float * const dram_c_tile_start = C + tile_i * TILE_M * dim_n + tile_j * TILE_N;
|
|
#pragma clang loop unroll(disable)
|
|
for (int thread_i = 0; thread_i < c_elems_per_thread; thread_i++) {
|
|
uint32_t elem_offset = hw_tid + num_threads_in_cluster * thread_i;
|
|
dram_c_tile_start[elem_offset / TILE_N * dim_n + elem_offset % TILE_N] = \
|
|
*(SMEM_ADDR_8K + SMEM_MAT_OFFSET(elem_offset / TILE_N, elem_offset % TILE_N, TILE_N));
|
|
}
|
|
#endif
|
|
__asm__("end_mvout_dram:");
|
|
|
|
// rd_cycles_force(marker8);
|
|
}
|
|
}
|
|
// last thread block complete
|
|
if (threadblock_id == NUM_CLUSTERS - 1) {
|
|
threadblock_barrier(/*barrier_id=*/0, /*count=*/NUM_WARPS);
|
|
rd_cycles_force(marker9);
|
|
if (HW_TID() == 0) {
|
|
PRINTF("\ncomplete\n");
|
|
PRINTF("total cycles: %d\n", marker9 - marker0);
|
|
}
|
|
#ifdef ACTIVATE
|
|
if (HW_TID() == 0) {
|
|
PRINTF("swish cycles: %d\n", swish_dur);
|
|
}
|
|
#endif
|
|
#ifdef DETAILED_PERF
|
|
vx_tmc(0x81);
|
|
for (int x = 0; x < num_threads_in_cluster; x += num_threads_in_cluster - 1) {
|
|
if (HW_TID() == x) {
|
|
PRINTF("\ntile start: %d\n", marker1);
|
|
PRINTF("single tile cycles: %d\n", marker6 - marker1);
|
|
PRINTF("A/B tile load cycles: %d\n", marker2 - marker1);
|
|
PRINTF("first barrier: %d\n", marker3 - marker2);
|
|
PRINTF("gemmini cycles: %d\n", marker4 - marker3);
|
|
PRINTF("second barrier: %d\n", marker5 - marker4);
|
|
#ifdef EXT_ACCUMULATE
|
|
PRINTF("accumulation cycles: %d\n", marker6 - marker5);
|
|
#else
|
|
PRINTF("smem mvout cycles: %d %d-%d\n", marker7 - marker6, marker7, marker6);
|
|
#endif
|
|
PRINTF("dram mvout cycles: %d\n", marker8 - marker7);
|
|
}
|
|
threadblock_barrier(/*barrier_id=*/1, /*count=*/NUM_WARPS);
|
|
}
|
|
#endif
|
|
if (HW_TID() == 0) {
|
|
for (int i = 0; i < dim_m; i += 8) {
|
|
for (int j = 0; j < dim_n; j += 8) {
|
|
PRINTF("%d %d ", (int) (C[i * dim_n + j]), (int) (C[i * dim_n + j + 4]));
|
|
}
|
|
PRINTF("\n");
|
|
}
|
|
}
|
|
}
|
|
vx_tmc(0);
|
|
}
|
|
|
|
void kernel_body(int task_id, kernel_arg_t *__UNIFORM__ arg) {
|
|
// @perf: All threads are running these compute whose result is mostly same
|
|
// across the threadblock
|
|
|
|
const int threadblock_id = task_id / NUM_THREADS_IN_CLUSTER;
|
|
const int tid_in_threadblock = task_id % NUM_THREADS_IN_CLUSTER;
|
|
|
|
thread_block_matmul_gemmini(arg, threadblock_id, tid_in_threadblock);
|
|
}
|
|
|
|
int main() {
|
|
kernel_arg_t *arg = (kernel_arg_t *)KERNEL_ARG_DEV_MEM_ADDR;
|
|
|
|
const uint32_t num_threads_in_cluster = vx_num_threads() * vx_num_warps() * CORES_PER_CLUSTER;
|
|
const uint32_t grid_size = num_threads_in_cluster * NUM_CLUSTERS;
|
|
#ifdef RADIANCE
|
|
vx_spawn_tasks_cluster(grid_size, (vx_spawn_tasks_cb)kernel_body, arg);
|
|
#else
|
|
// NOTE: This kernel assumes contiguous thread scheduling for efficient shared
|
|
// memory allocation, and therefore does not work with original vx_spawn_tasks
|
|
vx_spawn_tasks_contiguous(grid_size, (vx_spawn_tasks_cb)kernel_body, arg);
|
|
#endif
|
|
return 0;
|
|
} |