happyz
/
llama.cpp
mirror of https://github.com/ggerganov/llama.cpp.git
1
0
Fork 0
Code Issues Packages Projects Releases Wiki Activity

ggml-cpu: remove ifunc for rvv vec dot

This commit is contained in:
taimur-10x 2025-12-19 18:02:24 +05:00
parent 3805783fc3
commit c646e71982
1 changed files with 25 additions and 37 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

62
ggml/src/ggml-cpu/arch/riscv/quants.c
View File

@ -2274,24 +2274,19 @@ void ggml_vec_dot_iq2_xxs_q8_K_vl256(int n, float * GGML_RESTRICT s, size_t bs,
*s = 0.125f * sumf;
}
static void (*resolve_ggml_vec_dot_iq2_xxs_q8_K(void))(int n, float * GGML_RESTRICT s, size_t bs, const void * GGML_RESTRICT vx, size_t bx, const void * GGML_RESTRICT vy, size_t by, int nrc){
void ggml_vec_dot_iq2_xxs_q8_K(int n, float * GGML_RESTRICT s, size_t bs, const void * GGML_RESTRICT vx, size_t bx, const void * GGML_RESTRICT vy, size_t by, int nrc) {
#if defined __riscv_v_intrinsic
size_t vlen = __riscv_vlenb() * 8;
switch (vlen) {
switch (__riscv_vlenb() * 8) {
case 128:
return ggml_vec_dot_iq2_xxs_q8_K_vl128;
case 256:
return ggml_vec_dot_iq2_xxs_q8_K_vl256;
return ggml_vec_dot_iq2_xxs_q8_K_vl128(n, s, bs, vx, bx, vy, by, nrc);
default:
return ggml_vec_dot_iq2_xxs_q8_K_vl256;
return ggml_vec_dot_iq2_xxs_q8_K_vl256(n, s, bs, vx, bx, vy, by, nrc);
}
#else
return ggml_vec_dot_iq2_xxs_q8_K(n, s, bs, vx, bx, vy, by, nrc);
#endif
return ggml_vec_dot_iq2_xxs_q8_K_generic;
}
// Declare "ggml_vec_dot_iq2_xxs_q8_K" as an ifunc.
void ggml_vec_dot_iq2_xxs_q8_K(int n, float * GGML_RESTRICT s, size_t bs, const void * GGML_RESTRICT vx, size_t bx, const void * GGML_RESTRICT vy, size_t by, int nrc) __attribute__((ifunc("resolve_ggml_vec_dot_iq2_xxs_q8_K")));
void ggml_vec_dot_iq4_nl_q8_0_vl128(int n, float * GGML_RESTRICT s, size_t bs, const void * GGML_RESTRICT vx, size_t bx, const void * GGML_RESTRICT vy, size_t by, int nrc) {
assert(nrc == 1);
UNUSED(nrc);
@ -2402,22 +2397,19 @@ void ggml_vec_dot_iq4_nl_q8_0_vl256(int n, float * GGML_RESTRICT s, size_t bs, c
*s = sumf;
}
static void (*resolve_ggml_vec_dot_iq4_nl_q8_0(void))(int n, float * GGML_RESTRICT s, size_t bs, const void * GGML_RESTRICT vx, size_t bx, const void * GGML_RESTRICT vy, size_t by, int nrc) {
void ggml_vec_dot_iq4_nl_q8_0(int n, float * GGML_RESTRICT s, size_t bs, const void * GGML_RESTRICT vx, size_t bx, const void * GGML_RESTRICT vy, size_t by, int nrc) {
#if defined __riscv_v_intrinsic
size_t vlen = __riscv_vlenb() * 8;
switch (vlen) {
switch (__riscv_vlenb() * 8) {
case 128:
return ggml_vec_dot_iq4_nl_q8_0_vl128;
return ggml_vec_dot_iq4_nl_q8_0_vl128(n, s, bs, vx, bx, vy, by, nrc);
default:
return ggml_vec_dot_iq4_nl_q8_0_vl256;
return ggml_vec_dot_iq4_nl_q8_0_vl256(n, s, bs, vx, bx, vy, by, nrc);
}
#else
return ggml_vec_dot_iq4_nl_q8_0_generic(n, s, bs, vx, bx, vy, by, nrc);
#endif
return ggml_vec_dot_iq4_nl_q8_0_generic;
}
// Declare "ggml_vec_dot_iq4_nl_q8_0" as an ifunc.
void ggml_vec_dot_iq4_nl_q8_0(int n, float * GGML_RESTRICT s, size_t bs, const void * GGML_RESTRICT vx, size_t bx, const void * GGML_RESTRICT vy, size_t by, int nrc) __attribute__((ifunc("resolve_ggml_vec_dot_iq4_nl_q8_0")));
void ggml_vec_dot_mxfp4_q8_0_vl128(int n, float * GGML_RESTRICT s, size_t bs, const void * GGML_RESTRICT vx, size_t bx, const void * GGML_RESTRICT vy, size_t by, int nrc) {
assert(nrc == 1);
UNUSED(nrc);
@ -2528,22 +2520,19 @@ void ggml_vec_dot_mxfp4_q8_0_vl256(int n, float * GGML_RESTRICT s, size_t bs, co
*s = sumf;
}
static void (*resolve_ggml_vec_dot_mxfp4_q8_0(void))(int n, float * GGML_RESTRICT s, size_t bs, const void * GGML_RESTRICT vx, size_t bx, const void * GGML_RESTRICT vy, size_t by, int nrc) {
void ggml_vec_dot_mxfp4_q8_0(int n, float * GGML_RESTRICT s, size_t bs, const void * GGML_RESTRICT vx, size_t bx, const void * GGML_RESTRICT vy, size_t by, int nrc) {
#if defined __riscv_v_intrinsic
size_t vlen = __riscv_vlenb() * 8;
switch (vlen) {
switch (__riscv_vlenb() * 8) {
case 128:
return ggml_vec_dot_mxfp4_q8_0_vl128;
return ggml_vec_dot_mxfp4_q8_0_vl128(n, s, bs, vx, bx, vy, by, nrc);
default:
return ggml_vec_dot_mxfp4_q8_0_vl256;
return ggml_vec_dot_mxfp4_q8_0_vl256(n, s, bs, vx, bx, vy, by, nrc);
}
#else
return ggml_vec_dot_mxfp4_q8_0_generic(n, s, bs, vx, bx, vy, by, nrc);
#endif
return ggml_vec_dot_mxfp4_q8_0_generic;
}
// Declare "ggml_vec_dot_mxfp4_q8_0" as an ifunc.
void ggml_vec_dot_mxfp4_q8_0(int n, float * GGML_RESTRICT s, size_t bs, const void * GGML_RESTRICT vx, size_t bx, const void * GGML_RESTRICT vy, size_t by, int nrc) __attribute__((ifunc("resolve_ggml_vec_dot_mxfp4_q8_0")));
void ggml_vec_dot_iq4_xs_q8_K_vl256(int n, float * GGML_RESTRICT s, size_t bs, const void * GGML_RESTRICT vx, size_t bx, const void * GGML_RESTRICT vy, size_t by, int nrc) {
assert(nrc == 1);
UNUSED(nrc);
@ -2628,16 +2617,15 @@ void ggml_vec_dot_iq4_xs_q8_K_vl256(int n, float * GGML_RESTRICT s, size_t bs, c
#endif
}
static void (*resolve_ggml_vec_dot_iq4_xs_q8_K(void))(int n, float * GGML_RESTRICT s, size_t bs, const void * GGML_RESTRICT vx, size_t bx, const void * GGML_RESTRICT vy, size_t by, int nrc){
void ggml_vec_dot_iq4_xs_q8_K(int n, float * GGML_RESTRICT s, size_t bs, const void * GGML_RESTRICT vx, size_t bx, const void * GGML_RESTRICT vy, size_t by, int nrc) {
#if defined __riscv_v_intrinsic
size_t vlen = __riscv_vlenb() * 8;
switch (vlen) {
switch (__riscv_vlenb() * 8) {
case 256:
return ggml_vec_dot_iq4_xs_q8_K_vl256;
return ggml_vec_dot_iq4_xs_q8_K_vl256(n, s, bs, vx, bx, vy, by, nrc);
default:
return ggml_vec_dot_iq4_xs_q8_K_generic(n, s, bs, vx, bx, vy, by, nrc);
}
#else
return ggml_vec_dot_iq4_xs_q8_K_generic(n, s, bs, vx, bx, vy, by, nrc);
#endif
return ggml_vec_dot_iq4_xs_q8_K_generic;
}
// Declare "ggml_vec_dot_iq4_xs_q8_K" as an ifunc.
void ggml_vec_dot_iq4_xs_q8_K(int n, float * GGML_RESTRICT s, size_t bs, const void * GGML_RESTRICT vx, size_t bx, const void * GGML_RESTRICT vy, size_t by, int nrc) __attribute__((ifunc("resolve_ggml_vec_dot_iq4_xs_q8_K")));