91 lines
2.7 KiB
Common Lisp
91 lines
2.7 KiB
Common Lisp
// super super slow sgemm kernel by heehoon
|
|
#define TS 32
|
|
#define WPT 32
|
|
#define RTS 1
|
|
|
|
#define PADDINGX TS
|
|
#define PADDINGY TS
|
|
|
|
#define MIN(a, b) ((a) < (b) ? (a) : (b))
|
|
|
|
__kernel void sgemm(__global float *A, __global float *B, __global float *C, int M, int N, int K) {
|
|
const int row = get_local_id(0);
|
|
const int col = get_local_id(1);
|
|
const int global_row = TS * get_group_id(0) + row;
|
|
const int global_col = TS * get_group_id(1) + col;
|
|
|
|
__local float Asub[TS][TS];
|
|
__local float Bsub[TS][TS];
|
|
|
|
float Csub[WPT];
|
|
for (int w = 0; w < WPT; w++) {
|
|
Csub[w] = 0.0f;
|
|
}
|
|
|
|
const int num_tiles = (K + TS - 1) / TS;
|
|
|
|
for (int t = 0; t < num_tiles; t++) {
|
|
#pragma unroll
|
|
for (int w = 0; w < WPT; w++) {
|
|
const int t_row = TS * t + row;
|
|
const int t_col = TS * t + col;
|
|
Asub[row + w*RTS][col] = A[(global_row + w*RTS) * K + t_col];
|
|
Bsub[row + w*RTS][col] = B[(t_row + w*RTS) * N + global_col];
|
|
}
|
|
|
|
barrier(CLK_LOCAL_MEM_FENCE);
|
|
|
|
for (int k = 0; k < TS; k++) {
|
|
for (int w = 0; w < WPT; w++) {
|
|
Csub[w] += Asub[row + w*RTS][k] * Bsub[k][col];
|
|
}
|
|
}
|
|
|
|
barrier(CLK_LOCAL_MEM_FENCE);
|
|
}
|
|
|
|
#pragma unroll
|
|
for (int w = 0; w < WPT; w++) {
|
|
C[(global_row + w*RTS) * N + global_col] = Csub[w];
|
|
}
|
|
}
|
|
|
|
// Pad the P * Q matrix with zeroes to form a P_XL * Q_XL matrix
|
|
__kernel void paddingAddZeroes(const int P, const int Q,
|
|
const __global float* input,
|
|
const int P_XL, const int Q_XL,
|
|
__global float* output) {
|
|
|
|
const int tx = get_group_id(0) * PADDINGX + get_local_id(0);
|
|
const int ty = get_group_id(1) * PADDINGY + get_local_id(1);
|
|
|
|
// Check whether we are within bounds of the XL matrix
|
|
if (tx < P_XL && ty < Q_XL) {
|
|
|
|
// Copy the input or pad a zero
|
|
float value;
|
|
if (tx < P && ty < Q) {
|
|
value = input[tx*Q + ty];
|
|
}
|
|
else {
|
|
value = 0.0f;
|
|
}
|
|
|
|
output[tx*Q_XL + ty] = value;
|
|
}
|
|
}
|
|
|
|
// Remove padded values from a P_XL * Q_XL matrix to form a P * Q matrix
|
|
__kernel void paddingRemoveZeroes(const int P_XL, const int Q_XL,
|
|
const __global float* input,
|
|
const int P, const int Q,
|
|
__global float* output) {
|
|
|
|
const int tx = get_group_id(0) * PADDINGX + get_local_id(0);
|
|
const int ty = get_group_id(1) * PADDINGY + get_local_id(1);
|
|
|
|
// Only store the result if within P * Q bounds
|
|
if (tx < P && ty < Q) {
|
|
output[tx*Q + ty] = input[tx*Q_XL + ty];
|
|
}
|
|
} |