vbatts/llama.cpp
1
0
Fork 0
Code Issues Pull requests Projects Releases Packages Wiki Activity Actions

rm trailing spaces

Browse source
This commit is contained in:
Xuan Son Nguyen 2025-01-27 15:26:01 +01:00
parent e5aeb423a5
commit 226d59270b
1 changed files with 16 additions and 16 deletions
Download patch file Download diff file Expand all files Collapse all files

32
ggml/src/ggml-cpu/ggml-cpu-quants.c
View file

@ -2190,7 +2190,7 @@ void ggml_vec_dot_q4_0_q8_0(int n, float * restrict s, size_t bs, const void * r
// Accumulate results with scaling // Accumulate results with scaling
float scale0 = GGML_FP16_TO_FP32(x0->d) * GGML_FP16_TO_FP32(y0->d); float scale0 = GGML_FP16_TO_FP32(x0->d) * GGML_FP16_TO_FP32(y0->d);
float scale1 = GGML_FP16_TO_FP32(x1->d) * GGML_FP16_TO_FP32(y1->d); float scale1 = GGML_FP16_TO_FP32(x1->d) * GGML_FP16_TO_FP32(y1->d);
sumv = wasm_f32x4_add(sumv, wasm_f32x4_mul(wasm_f32x4_convert_i32x4(dp0), wasm_f32x4_splat(scale0))); sumv = wasm_f32x4_add(sumv, wasm_f32x4_mul(wasm_f32x4_convert_i32x4(dp0), wasm_f32x4_splat(scale0)));
sumv = wasm_f32x4_add(sumv, wasm_f32x4_mul(wasm_f32x4_convert_i32x4(dp1), wasm_f32x4_splat(scale1))); sumv = wasm_f32x4_add(sumv, wasm_f32x4_mul(wasm_f32x4_convert_i32x4(dp1), wasm_f32x4_splat(scale1)));
} }
@ -4685,19 +4685,19 @@ void ggml_vec_dot_q2_K_q8_K(int n, float * restrict s, size_t bs, const void * r
{ {
v128_t sc_vec = wasm_v128_load(sc); v128_t sc_vec = wasm_v128_load(sc);
v128_t sc_upper = wasm_u8x16_shr(sc_vec, 4); v128_t sc_upper = wasm_u8x16_shr(sc_vec, 4);
v128_t sc_low = wasm_u16x8_extend_low_u8x16(sc_upper); v128_t sc_low = wasm_u16x8_extend_low_u8x16(sc_upper);
v128_t sc_high = wasm_u16x8_extend_high_u8x16(sc_upper); v128_t sc_high = wasm_u16x8_extend_high_u8x16(sc_upper);
v128_t bsums1 = wasm_v128_load(&y[i].bsums[0]); v128_t bsums1 = wasm_v128_load(&y[i].bsums[0]);
v128_t bsums2 = wasm_v128_load(&y[i].bsums[8]); v128_t bsums2 = wasm_v128_load(&y[i].bsums[8]);
summs_vec = wasm_i32x4_add( summs_vec = wasm_i32x4_add(
wasm_i32x4_add(wasm_i32x4_dot_i16x8(sc_low, bsums1), wasm_i32x4_add(wasm_i32x4_dot_i16x8(sc_low, bsums1),
wasm_i32x4_dot_i16x8(sc_high, bsums2)), wasm_i32x4_dot_i16x8(sc_high, bsums2)),
summs_vec summs_vec
); );
summs_vec = wasm_i32x4_add(summs_vec, wasm_i32x4_shuffle(summs_vec, summs_vec, 2, 3, 0, 1)); summs_vec = wasm_i32x4_add(summs_vec, wasm_i32x4_shuffle(summs_vec, summs_vec, 2, 3, 0, 1));
summs_vec = wasm_i32x4_add(summs_vec, wasm_i32x4_shuffle(summs_vec, summs_vec, 1, 0, 3, 2)); summs_vec = wasm_i32x4_add(summs_vec, wasm_i32x4_shuffle(summs_vec, summs_vec, 1, 0, 3, 2));
} }
@ -4707,11 +4707,11 @@ void ggml_vec_dot_q2_K_q8_K(int n, float * restrict s, size_t bs, const void * r
int32_t isum = 0; int32_t isum = 0;
const uint8_t * sc_ptr = sc; const uint8_t * sc_ptr = sc;
const int k_iters = QK_K/128; const int k_iters = QK_K/128;
for (int k = 0; k < k_iters; ++k) { for (int k = 0; k < k_iters; ++k) {
v128_t isum_vec = wasm_i32x4_splat(0); v128_t isum_vec = wasm_i32x4_splat(0);
int shift = 0; int shift = 0;
for (int j = 0; j < 4; ++j) { for (int j = 0; j < 4; ++j) {
const int d0 = (sc_ptr[0] & 0xF); const int d0 = (sc_ptr[0] & 0xF);
const int d1 = (sc_ptr[1] & 0xF); const int d1 = (sc_ptr[1] & 0xF);
@ -4722,7 +4722,7 @@ void ggml_vec_dot_q2_K_q8_K(int n, float * restrict s, size_t bs, const void * r
v128_t q8_0 = wasm_v128_load(q8); v128_t q8_0 = wasm_v128_load(q8);
v128_t q2_shift_0 = wasm_u8x16_shr(q2_0, shift); v128_t q2_shift_0 = wasm_u8x16_shr(q2_0, shift);
v128_t q2_bits_0 = wasm_v128_and(q2_shift_0, wasm_i8x16_splat(0x03)); v128_t q2_bits_0 = wasm_v128_and(q2_shift_0, wasm_i8x16_splat(0x03));
// Process next 16 elements // Process next 16 elements
v128_t q2_1 = wasm_v128_load(q2 + 16); v128_t q2_1 = wasm_v128_load(q2 + 16);
v128_t q8_1 = wasm_v128_load(q8 + 16); v128_t q8_1 = wasm_v128_load(q8 + 16);
@ -4752,7 +4752,7 @@ void ggml_vec_dot_q2_K_q8_K(int n, float * restrict s, size_t bs, const void * r
wasm_i32x4_mul(wasm_i32x4_add(p0, p1), wasm_i32x4_splat(d0)), wasm_i32x4_mul(wasm_i32x4_add(p0, p1), wasm_i32x4_splat(d0)),
wasm_i32x4_mul(wasm_i32x4_add(p2, p3), wasm_i32x4_splat(d1)) wasm_i32x4_mul(wasm_i32x4_add(p2, p3), wasm_i32x4_splat(d1))
); );
isum_vec = wasm_i32x4_add(isum_vec, scaled); isum_vec = wasm_i32x4_add(isum_vec, scaled);
q8 += 32; q8 += 32;
shift += 2; shift += 2;
@ -5464,7 +5464,7 @@ void ggml_vec_dot_q3_K_q8_K(int n, float * restrict s, size_t bs, const void * r
const uint8_t * restrict q3 = x[i].qs; const uint8_t * restrict q3 = x[i].qs;
const uint8_t * restrict hm = x[i].hmask; const uint8_t * restrict hm = x[i].hmask;
const int8_t * restrict q8 = y[i].qs; const int8_t * restrict q8 = y[i].qs;
// Process blocks with SIMD // Process blocks with SIMD
int8_t * a = aux8; int8_t * a = aux8;
uint8_t m = 1; uint8_t m = 1;
@ -5475,11 +5475,11 @@ void ggml_vec_dot_q3_K_q8_K(int n, float * restrict s, size_t bs, const void * r
v128_t v_q3 = wasm_v128_load(q3 + l); v128_t v_q3 = wasm_v128_load(q3 + l);
v128_t v_shift = wasm_i8x16_shr(v_q3, shift); v128_t v_shift = wasm_i8x16_shr(v_q3, shift);
v128_t v_low2 = wasm_v128_and(v_shift, wasm_i8x16_splat(0x03)); v128_t v_low2 = wasm_v128_and(v_shift, wasm_i8x16_splat(0x03));
v128_t v_hm = wasm_v128_load(hm + l); v128_t v_hm = wasm_v128_load(hm + l);
v128_t v_mask = wasm_v128_and(v_hm, v_m); v128_t v_mask = wasm_v128_and(v_hm, v_m);
v_mask = wasm_i8x16_ne(v_mask, wasm_i8x16_splat(0)); v_mask = wasm_i8x16_ne(v_mask, wasm_i8x16_splat(0));
v_low2 = wasm_i8x16_sub(v_low2, wasm_v128_and(wasm_i8x16_splat(4), wasm_v128_not(v_mask))); v_low2 = wasm_i8x16_sub(v_low2, wasm_v128_and(wasm_i8x16_splat(4), wasm_v128_not(v_mask)));
wasm_v128_store(a + l, v_low2); wasm_v128_store(a + l, v_low2);
} }
@ -5509,13 +5509,13 @@ void ggml_vec_dot_q3_K_q8_K(int n, float * restrict s, size_t bs, const void * r
for (int k = 0; k < 2; ++k) { for (int k = 0; k < 2; ++k) {
const v128_t v_q8 = wasm_i16x8_load8x8(q8); const v128_t v_q8 = wasm_i16x8_load8x8(q8);
const v128_t v_a = wasm_i16x8_load8x8(a); const v128_t v_a = wasm_i16x8_load8x8(a);
v128_t v_prod = wasm_i16x8_mul(v_q8, v_a); v128_t v_prod = wasm_i16x8_mul(v_q8, v_a);
v_prod = wasm_i16x8_mul(v_prod, v_scale); v_prod = wasm_i16x8_mul(v_prod, v_scale);
v_acc0 = wasm_i32x4_add(v_acc0, wasm_i32x4_extend_low_i16x8(v_prod)); v_acc0 = wasm_i32x4_add(v_acc0, wasm_i32x4_extend_low_i16x8(v_prod));
v_acc1 = wasm_i32x4_add(v_acc1, wasm_i32x4_extend_high_i16x8(v_prod)); v_acc1 = wasm_i32x4_add(v_acc1, wasm_i32x4_extend_high_i16x8(v_prod));
q8 += 8; q8 += 8;
a += 8; a += 8;
} }
@ -5528,7 +5528,7 @@ void ggml_vec_dot_q3_K_q8_K(int n, float * restrict s, size_t bs, const void * r
wasm_f32x4_mul(wasm_f32x4_convert_i32x4(v_acc0), v_d), wasm_f32x4_mul(wasm_f32x4_convert_i32x4(v_acc0), v_d),
wasm_f32x4_mul(wasm_f32x4_convert_i32x4(v_acc1), v_d) wasm_f32x4_mul(wasm_f32x4_convert_i32x4(v_acc1), v_d)
); );
// Accumulate into sums vector // Accumulate into sums vector
wasm_v128_store(sums, wasm_f32x4_add(wasm_v128_load(sums), v_sum)); wasm_v128_store(sums, wasm_f32x4_add(wasm_v128_load(sums), v_sum));
} }