CS348Project/llama.cpp
5
0
Fork 0
mirror of https://github.com/ggml-org/llama.cpp.git synced 2025-10-31 08:51:55 +00:00
Code Issues Packages Projects Releases Wiki Activity

vulkan: move common FA code to flash_attn_base.comp (#13556)

Browse Source 
* vulkan: move common FA code to flash_attn_base.comp

* vulkan: move common FA index/stride setup code to flash_attn_base.comp

* build fix
This commit is contained in:
Jeff Bolz
2025-05-17 16:14:55 +09:00
committed by GitHub
parent 4f41ee11d6
commit 2f5a4e1e09
4 changed files with 170 additions and 417 deletions
Download Patch File Download Diff File Expand all files Collapse all files

153
ggml/src/ggml-vulkan/vulkan-shaders/flash_attn.comp
View File

@@ -9,60 +9,13 @@
#extension GL_KHR_shader_subgroup_shuffle : enable #extension GL_KHR_shader_subgroup_shuffle : enable
#include "types.comp" #include "types.comp"
#include "flash_attn_base.comp"
layout(local_size_x_id = 0, local_size_y = 1, local_size_z = 1) in;
layout (constant_id = 0) const uint32_t WorkGroupSize = 128;
layout (constant_id = 1) const uint32_t Br = 1;
layout (constant_id = 2) const uint32_t Bc = 32;
layout (constant_id = 3) const uint32_t D = 32;
layout (constant_id = 5) const uint32_t D_split = 16;
const uint32_t D_per_thread = D / D_split; const uint32_t D_per_thread = D / D_split;
const uint32_t cols_per_iter = WorkGroupSize / D_split; const uint32_t cols_per_iter = WorkGroupSize / D_split;
const uint32_t cols_per_thread = Bc / cols_per_iter; const uint32_t cols_per_thread = Bc / cols_per_iter;
layout (push_constant) uniform parameter {
uint32_t N;
uint32_t KV;
uint32_t ne1;
uint32_t ne2;
uint32_t ne3;
uint32_t neq2;
uint32_t neq3;
uint32_t nek2;
uint32_t nek3;
uint32_t nev2;
uint32_t nev3;
uint32_t nem1;
uint32_t nb01;
uint32_t nb02;
uint32_t nb03;
uint32_t nb11;
uint32_t nb12;
uint32_t nb13;
uint32_t nb21;
uint32_t nb22;
uint32_t nb23;
uint32_t nb31;
float scale;
float max_bias;
float logit_softcap;
uint32_t mask;
uint32_t n_head_log2;
float m0;
float m1;
uint32_t gqa_ratio;
uint32_t split_kv;
uint32_t k_num;
} p;
layout (binding = 0) readonly buffer Q {float data_q[];}; layout (binding = 0) readonly buffer Q {float data_q[];};
layout (binding = 0) readonly buffer QV4 {vec4 data_qv4[];}; layout (binding = 0) readonly buffer QV4 {vec4 data_qv4[];};
@@ -71,39 +24,6 @@ layout (binding = 1) readonly buffer KV4 {f16vec4 data_kv4[];};
layout (binding = 2) readonly buffer V {float16_t data_v[];}; layout (binding = 2) readonly buffer V {float16_t data_v[];};
layout (binding = 2) readonly buffer VV4 {f16vec4 data_vv4[];}; layout (binding = 2) readonly buffer VV4 {f16vec4 data_vv4[];};
layout (binding = 3) readonly buffer M {float16_t data_m[];}; layout (binding = 3) readonly buffer M {float16_t data_m[];};
layout (binding = 4) writeonly buffer O {D_TYPE data_o[];};
#if defined(A_TYPE_PACKED16)
#define BINDING_IDX_K 0
#define BINDING_IDX_V 1
layout (binding = 1) readonly buffer KV_PACKED16 {A_TYPE_PACKED16 data_packed16[];} kv_packed[2];
#endif
#if defined(DATA_A_Q4_0)
#define BLOCK_BYTE_SIZE 18
vec4 dequantize4(uint ib, uint iqs, uint a_offset, uint binding_idx) {
uint vui_lo = uint(kv_packed[binding_idx].data_packed16[a_offset + ib].qs[(iqs & 0xF) / 2 + 0]);
uint vui_hi = uint(kv_packed[binding_idx].data_packed16[a_offset + ib].qs[(iqs & 0xF) / 2 + 1]);
uint shift = (iqs & 0x10) >> 2;
vui_lo >>= shift;
vui_hi >>= shift;
return float(kv_packed[binding_idx].data_packed16[a_offset + ib].d) * (vec4(vui_lo & 0xF, (vui_lo >> 8) & 0xF, vui_hi & 0xF, (vui_hi >> 8) & 0xF) - 8.0f);
}
#endif
#if defined(DATA_A_Q8_0)
#define BLOCK_BYTE_SIZE 34
vec4 dequantize4(uint ib, uint iqs, uint a_offset, uint binding_idx) {
const i8vec2 v0 = unpack8(int32_t(kv_packed[binding_idx].data_packed16[a_offset + ib].qs[iqs / 2])).xy; // vec4 used due to #12147
const i8vec2 v1 = unpack8(int32_t(kv_packed[binding_idx].data_packed16[a_offset + ib].qs[iqs / 2 + 1])).xy;
return float(kv_packed[binding_idx].data_packed16[a_offset + ib].d) * vec4(v0.x, v0.y, v1.x, v1.y);
}
#endif
#define CEIL_DIV(a, b) (((a) + (b) - 1) / (b))
// Store the output when doing grouped query attention. // Store the output when doing grouped query attention.
// Rows index by Q's dimension 2, and the first N rows are valid. // Rows index by Q's dimension 2, and the first N rows are valid.
@@ -114,27 +34,6 @@ D_TYPE perElemOpGqaStore(const in uint32_t r, const in uint32_t c, const in D_TY
return elem; return elem;
} }
// Store column zero. This is used to save per-row m and L values for split_k.
ACC_TYPE perElemOpStoreCol0(const in uint32_t r, const in uint32_t c, const in ACC_TYPE elem, const in uint32_t o_offset, const in uint32_t iq2, const in uint32_t N)
{
if (r < N && c == 0) {
uint32_t offset = iq2 + r;
data_o[o_offset + offset] = D_TYPE(elem);
}
return elem;
}
// Load the slope matrix, indexed by Q's dimension 2.
ACC_TYPE perElemOpComputeSlope(const in uint32_t r, const in uint32_t c, const in ACC_TYPE elem, const in uint32_t iq2)
{
const uint32_t h = iq2 + (r % p.gqa_ratio);
const ACC_TYPE base = ACC_TYPE(h < p.n_head_log2 ? p.m0 : p.m1);
const int exph = int(h < p.n_head_log2 ? h + 1 : 2*(h - p.n_head_log2) + 1);
return ACC_TYPE(pow(base, ACC_TYPE(exph)));
}
shared FLOAT_TYPE tmpsh[WorkGroupSize]; shared FLOAT_TYPE tmpsh[WorkGroupSize];
shared vec4 tmpshv4[WorkGroupSize]; shared vec4 tmpshv4[WorkGroupSize];
@@ -146,58 +45,12 @@ void main() {
init_iq_shmem(gl_WorkGroupSize); init_iq_shmem(gl_WorkGroupSize);
#endif #endif
const uint32_t tid = gl_LocalInvocationIndex; init_indices();
const uint32_t N = p.N;
const uint32_t KV = p.KV;
const uint32_t tid = gl_LocalInvocationIndex;
const uint32_t d_tid = gl_LocalInvocationIndex % D_split; const uint32_t d_tid = gl_LocalInvocationIndex % D_split;
const uint32_t col_tid = gl_LocalInvocationIndex / D_split; const uint32_t col_tid = gl_LocalInvocationIndex / D_split;
uint32_t i = gl_WorkGroupID.x;
uint32_t split_k_index = 0;
if (p.k_num > 1) {
i = 0;
split_k_index = gl_WorkGroupID.x;
}
const uint32_t Tr = CEIL_DIV(N, Br);
const uint32_t start_j = split_k_index * p.split_kv / Bc;
const uint32_t end_j = CEIL_DIV(min(KV, (split_k_index + 1) * p.split_kv), Bc);
// When not using grouped query attention, all rows share the same iq2, equal to gl_WorkGroupID.y.
// When using grouped query attention, each workgroup does gqa_ratio consecutive values of iq2.
const uint32_t iq2 = gl_WorkGroupID.y * p.gqa_ratio;
const uint32_t iq3 = gl_WorkGroupID.z;
// broadcast factors
const uint32_t rk2 = p.neq2/p.nek2;
const uint32_t rk3 = p.neq3/p.nek3;
const uint32_t rv2 = p.neq2/p.nev2;
const uint32_t rv3 = p.neq3/p.nev3;
// k indices
const uint32_t ik3 = iq3 / rk3;
const uint32_t ik2 = iq2 / rk2;
// v indices
const uint32_t iv3 = iq3 / rv3;
const uint32_t iv2 = iq2 / rv2;
// nb?1 are already divided by the type size and are in units of elements.
// When using grouped query attention, Q is indexed by iq2, so the stride
// should be nb02 (which is in bytes).
uint32_t q_stride = p.gqa_ratio > 1 ? (p.nb02 / 4) : p.nb01;
uint32_t k_stride = p.nb11;
uint32_t v_stride = p.nb21;
// When using grouped query attention, all rows use the same mask (stride 0).
// "p.gqa_ratio >> 16" is just a roundabout way of writing zero
// that prevents the compiler from folding the "&" through the select
// and breaking the alignment detection.
uint32_t m_stride = (p.gqa_ratio > 1) ? (p.gqa_ratio >> 16) : KV;
uint32_t q_offset = (iq2*p.nb02+iq3*p.nb03) / 4; uint32_t q_offset = (iq2*p.nb02+iq3*p.nb03) / 4;
[[unroll]] for (uint32_t idx = 0; idx < Br * D / 4; idx += gl_WorkGroupSize.x) { [[unroll]] for (uint32_t idx = 0; idx < Br * D / 4; idx += gl_WorkGroupSize.x) {

162
ggml/src/ggml-vulkan/vulkan-shaders/flash_attn_base.comp Normal file
View File

@@ -0,0 +1,162 @@
layout(local_size_x_id = 0, local_size_y = 1, local_size_z = 1) in;
layout (constant_id = 0) const uint32_t WorkGroupSize = 128;
layout (constant_id = 1) const uint32_t Br = 1;
layout (constant_id = 2) const uint32_t Bc = 32;
layout (constant_id = 3) const uint32_t D = 32;
layout (constant_id = 4) const uint32_t Clamp = 0;
layout (constant_id = 5) const uint32_t D_split = 16;
layout (push_constant) uniform parameter {
uint32_t N;
uint32_t KV;
uint32_t ne1;
uint32_t ne2;
uint32_t ne3;
uint32_t neq2;
uint32_t neq3;
uint32_t nek2;
uint32_t nek3;
uint32_t nev2;
uint32_t nev3;
uint32_t nem1;
uint32_t nb01;
uint32_t nb02;
uint32_t nb03;
uint32_t nb11;
uint32_t nb12;
uint32_t nb13;
uint32_t nb21;
uint32_t nb22;
uint32_t nb23;
uint32_t nb31;
float scale;
float max_bias;
float logit_softcap;
uint32_t mask;
uint32_t n_head_log2;
float m0;
float m1;
uint32_t gqa_ratio;
uint32_t split_kv;
uint32_t k_num;
} p;
layout (binding = 4) writeonly buffer O {D_TYPE data_o[];};
#if defined(A_TYPE_PACKED16)
#define BINDING_IDX_K 0
#define BINDING_IDX_V 1
layout (binding = 1) readonly buffer KV_PACKED16 {A_TYPE_PACKED16 data_packed16[];} kv_packed[2];
#endif
#if defined(DATA_A_Q4_0)
#define BLOCK_BYTE_SIZE 18
vec4 dequantize4(uint ib, uint iqs, uint a_offset, uint binding_idx) {
uint vui_lo = uint(kv_packed[binding_idx].data_packed16[a_offset + ib].qs[(iqs & 0xF) / 2 + 0]);
uint vui_hi = uint(kv_packed[binding_idx].data_packed16[a_offset + ib].qs[(iqs & 0xF) / 2 + 1]);
uint shift = (iqs & 0x10) >> 2;
vui_lo >>= shift;
vui_hi >>= shift;
return float(kv_packed[binding_idx].data_packed16[a_offset + ib].d) * (vec4(vui_lo & 0xF, (vui_lo >> 8) & 0xF, vui_hi & 0xF, (vui_hi >> 8) & 0xF) - 8.0f);
}
#endif
#if defined(DATA_A_Q8_0)
#define BLOCK_BYTE_SIZE 34
vec4 dequantize4(uint ib, uint iqs, uint a_offset, uint binding_idx) {
const i8vec2 v0 = unpack8(int32_t(kv_packed[binding_idx].data_packed16[a_offset + ib].qs[iqs / 2])).xy; // vec4 used due to #12147
const i8vec2 v1 = unpack8(int32_t(kv_packed[binding_idx].data_packed16[a_offset + ib].qs[iqs / 2 + 1])).xy;
return float(kv_packed[binding_idx].data_packed16[a_offset + ib].d) * vec4(v0.x, v0.y, v1.x, v1.y);
}
#endif
#define CEIL_DIV(a, b) (((a) + (b) - 1) / (b))
// Store column zero. This is used to save per-row m and L values for split_k.
ACC_TYPE perElemOpStoreCol0(const in uint32_t r, const in uint32_t c, const in ACC_TYPE elem, const in uint32_t o_offset, const in uint32_t iq2, const in uint32_t N)
{
if (r < N && c == 0) {
uint32_t offset = iq2 + r;
data_o[o_offset + offset] = D_TYPE(elem);
}
return elem;
}
// Load the slope matrix, indexed by Q's dimension 2.
ACC_TYPE perElemOpComputeSlope(const in uint32_t r, const in uint32_t c, const in ACC_TYPE elem, const in uint32_t iq2)
{
const uint32_t h = iq2 + (r % p.gqa_ratio);
const ACC_TYPE base = ACC_TYPE(h < p.n_head_log2 ? p.m0 : p.m1);
const int exph = int(h < p.n_head_log2 ? h + 1 : 2*(h - p.n_head_log2) + 1);
return ACC_TYPE(pow(base, ACC_TYPE(exph)));
}
uint32_t i, N, KV, split_k_index, Tr, start_j, end_j,
iq2, iq3, rk2, rk3, rv2, rv3, ik2, ik3, iv2, iv3,
q_stride, k_stride, v_stride, m_stride;
void init_indices()
{
N = p.N;
KV = p.KV;
i = gl_WorkGroupID.x;
split_k_index = 0;
if (p.k_num > 1) {
i = 0;
split_k_index = gl_WorkGroupID.x;
}
Tr = CEIL_DIV(N, Br);
start_j = split_k_index * p.split_kv / Bc;
end_j = CEIL_DIV(min(KV, (split_k_index + 1) * p.split_kv), Bc);
// When not using grouped query attention, all rows share the same iq2, equal to gl_WorkGroupID.y.
// When using grouped query attention, each workgroup does gqa_ratio consecutive values of iq2.
iq2 = gl_WorkGroupID.y * p.gqa_ratio;
iq3 = gl_WorkGroupID.z;
// broadcast factors
rk2 = p.neq2/p.nek2;
rk3 = p.neq3/p.nek3;
rv2 = p.neq2/p.nev2;
rv3 = p.neq3/p.nev3;
// k indices
ik3 = iq3 / rk3;
ik2 = iq2 / rk2;
// v indices
iv3 = iq3 / rv3;
iv2 = iq2 / rv2;
// nb?1 are already divided by the type size and are in units of elements.
// When using grouped query attention, Q is indexed by iq2, so the stride
// should be nb02 (which is in bytes).
q_stride = p.gqa_ratio > 1 ? (p.nb02 / 4) : p.nb01;
k_stride = p.nb11;
v_stride = p.nb21;
// When using grouped query attention, all rows use the same mask (stride 0).
// "p.gqa_ratio >> 16" is just a roundabout way of writing zero
// that prevents the compiler from folding the "&" through the select
// and breaking the alignment detection.
m_stride = (p.gqa_ratio > 1) ? (p.gqa_ratio >> 16) : KV;
}

152
ggml/src/ggml-vulkan/vulkan-shaders/flash_attn_cm1.comp
View File

@@ -11,14 +11,7 @@
#extension GL_KHR_cooperative_matrix : enable #extension GL_KHR_cooperative_matrix : enable
#include "types.comp" #include "types.comp"
#include "flash_attn_base.comp"
layout(local_size_x_id = 0, local_size_y = 1, local_size_z = 1) in;
layout (constant_id = 1) const uint32_t Br = 1;
layout (constant_id = 2) const uint32_t Bc = 32;
layout (constant_id = 3) const uint32_t D = 32;
layout (constant_id = 5) const uint32_t D_split = 16;
const uint32_t D_per_thread = D / D_split; const uint32_t D_per_thread = D / D_split;
const uint32_t row_split = 4; const uint32_t row_split = 4;
@@ -26,46 +19,6 @@ const uint32_t rows_per_thread = Br / row_split;
const uint32_t cols_per_iter = gl_WorkGroupSize.x / D_split / row_split; const uint32_t cols_per_iter = gl_WorkGroupSize.x / D_split / row_split;
const uint32_t cols_per_thread = Bc / cols_per_iter; const uint32_t cols_per_thread = Bc / cols_per_iter;
layout (push_constant) uniform parameter {
uint32_t N;
uint32_t KV;
uint32_t ne1;
uint32_t ne2;
uint32_t ne3;
uint32_t neq2;
uint32_t neq3;
uint32_t nek2;
uint32_t nek3;
uint32_t nev2;
uint32_t nev3;
uint32_t nem1;
uint32_t nb01;
uint32_t nb02;
uint32_t nb03;
uint32_t nb11;
uint32_t nb12;
uint32_t nb13;
uint32_t nb21;
uint32_t nb22;
uint32_t nb23;
uint32_t nb31;
float scale;
float max_bias;
float logit_softcap;
uint32_t mask;
uint32_t n_head_log2;
float m0;
float m1;
uint32_t gqa_ratio;
uint32_t split_kv;
uint32_t k_num;
} p;
layout (binding = 0) readonly buffer Q {float data_q[];}; layout (binding = 0) readonly buffer Q {float data_q[];};
layout (binding = 0) readonly buffer QV4 {vec4 data_qv4[];}; layout (binding = 0) readonly buffer QV4 {vec4 data_qv4[];};
@@ -74,39 +27,6 @@ layout (binding = 1) readonly buffer KV4 {f16vec4 data_kv4[];};
layout (binding = 2) readonly buffer V {float16_t data_v[];}; layout (binding = 2) readonly buffer V {float16_t data_v[];};
layout (binding = 2) readonly buffer VV4 {f16vec4 data_vv4[];}; layout (binding = 2) readonly buffer VV4 {f16vec4 data_vv4[];};
layout (binding = 3) readonly buffer M {float16_t data_m[];}; layout (binding = 3) readonly buffer M {float16_t data_m[];};
layout (binding = 4) writeonly buffer O {D_TYPE data_o[];};
#if defined(A_TYPE_PACKED16)
#define BINDING_IDX_K 0
#define BINDING_IDX_V 1
layout (binding = 1) readonly buffer KV_PACKED16 {A_TYPE_PACKED16 data_packed16[];} kv_packed[2];
#endif
#if defined(DATA_A_Q4_0)
#define BLOCK_BYTE_SIZE 18
vec4 dequantize4(uint ib, uint iqs, uint a_offset, uint binding_idx) {
uint vui_lo = uint(kv_packed[binding_idx].data_packed16[a_offset + ib].qs[(iqs & 0xF) / 2 + 0]);
uint vui_hi = uint(kv_packed[binding_idx].data_packed16[a_offset + ib].qs[(iqs & 0xF) / 2 + 1]);
uint shift = (iqs & 0x10) >> 2;
vui_lo >>= shift;
vui_hi >>= shift;
return float(kv_packed[binding_idx].data_packed16[a_offset + ib].d) * (vec4(vui_lo & 0xF, (vui_lo >> 8) & 0xF, vui_hi & 0xF, (vui_hi >> 8) & 0xF) - 8.0f);
}
#endif
#if defined(DATA_A_Q8_0)
#define BLOCK_BYTE_SIZE 34
vec4 dequantize4(uint ib, uint iqs, uint a_offset, uint binding_idx) {
const i8vec2 v0 = unpack8(int32_t(kv_packed[binding_idx].data_packed16[a_offset + ib].qs[iqs / 2])).xy; // vec4 used due to #12147
const i8vec2 v1 = unpack8(int32_t(kv_packed[binding_idx].data_packed16[a_offset + ib].qs[iqs / 2 + 1])).xy;
return float(kv_packed[binding_idx].data_packed16[a_offset + ib].d) * vec4(v0.x, v0.y, v1.x, v1.y);
}
#endif
#define CEIL_DIV(a, b) (((a) + (b) - 1) / (b))
// Store the output when doing grouped query attention. // Store the output when doing grouped query attention.
// Rows index by Q's dimension 2, and the first N rows are valid. // Rows index by Q's dimension 2, and the first N rows are valid.
@@ -117,27 +37,6 @@ D_TYPE perElemOpGqaStore(const in uint32_t r, const in uint32_t c, const in D_TY
return elem; return elem;
} }
// Store column zero. This is used to save per-row m and L values for split_k.
ACC_TYPE perElemOpStoreCol0(const in uint32_t r, const in uint32_t c, const in ACC_TYPE elem, const in uint32_t o_offset, const in uint32_t iq2, const in uint32_t N)
{
if (r < N && c == 0) {
uint32_t offset = iq2 + r;
data_o[o_offset + offset] = D_TYPE(elem);
}
return elem;
}
// Load the slope matrix, indexed by Q's dimension 2.
ACC_TYPE perElemOpComputeSlope(const in uint32_t r, const in uint32_t c, const in ACC_TYPE elem, const in uint32_t iq2)
{
const uint32_t h = iq2 + (r % p.gqa_ratio);
const ACC_TYPE base = ACC_TYPE(h < p.n_head_log2 ? p.m0 : p.m1);
const int exph = int(h < p.n_head_log2 ? h + 1 : 2*(h - p.n_head_log2) + 1);
return ACC_TYPE(pow(base, ACC_TYPE(exph)));
}
// These need to be supported N,M values for a MatBc x MatBr x 16 coopmatmuladd // These need to be supported N,M values for a MatBc x MatBr x 16 coopmatmuladd
const uint32_t MatBr = 16; const uint32_t MatBr = 16;
const uint32_t MatBc = 16; const uint32_t MatBc = 16;
@@ -162,9 +61,9 @@ void main() {
init_iq_shmem(gl_WorkGroupSize); init_iq_shmem(gl_WorkGroupSize);
#endif #endif
init_indices();
const uint32_t tid = gl_LocalInvocationIndex; const uint32_t tid = gl_LocalInvocationIndex;
const uint32_t N = p.N;
const uint32_t KV = p.KV;
const uint32_t threads_per_rowgroup = gl_WorkGroupSize.x / row_split; const uint32_t threads_per_rowgroup = gl_WorkGroupSize.x / row_split;
const uint32_t row_tid = gl_LocalInvocationIndex / threads_per_rowgroup; const uint32_t row_tid = gl_LocalInvocationIndex / threads_per_rowgroup;
@@ -173,51 +72,6 @@ void main() {
#define tile_row(r) (row_tid * rows_per_thread + (r)) #define tile_row(r) (row_tid * rows_per_thread + (r))
uint32_t i = gl_WorkGroupID.x;
uint32_t split_k_index = 0;
if (p.k_num > 1) {
i = 0;
split_k_index = gl_WorkGroupID.x;
}
const uint32_t Tr = CEIL_DIV(N, Br);
const uint32_t start_j = split_k_index * p.split_kv / Bc;
const uint32_t end_j = CEIL_DIV(min(KV, (split_k_index + 1) * p.split_kv), Bc);
// When not using grouped query attention, all rows share the same iq2, equal to gl_WorkGroupID.y.
// When using grouped query attention, each workgroup does gqa_ratio consecutive values of iq2.
const uint32_t iq2 = gl_WorkGroupID.y * p.gqa_ratio;
const uint32_t iq3 = gl_WorkGroupID.z;
// broadcast factors
const uint32_t rk2 = p.neq2/p.nek2;
const uint32_t rk3 = p.neq3/p.nek3;
const uint32_t rv2 = p.neq2/p.nev2;
const uint32_t rv3 = p.neq3/p.nev3;
// k indices
const uint32_t ik3 = iq3 / rk3;
const uint32_t ik2 = iq2 / rk2;
// v indices
const uint32_t iv3 = iq3 / rv3;
const uint32_t iv2 = iq2 / rv2;
// nb?1 are already divided by the type size and are in units of elements.
// When using grouped query attention, Q is indexed by iq2, so the stride
// should be nb02 (which is in bytes).
uint32_t q_stride = p.gqa_ratio > 1 ? (p.nb02 / 4) : p.nb01;
uint32_t k_stride = p.nb11;
uint32_t v_stride = p.nb21;
// When using grouped query attention, all rows use the same mask (stride 0).
// "p.gqa_ratio >> 16" is just a roundabout way of writing zero
// that prevents the compiler from folding the "&" through the select
// and breaking the alignment detection.
uint32_t m_stride = (p.gqa_ratio > 1) ? (p.gqa_ratio >> 16) : KV;
uint32_t q_offset = (iq2*p.nb02+iq3*p.nb03) / 4; uint32_t q_offset = (iq2*p.nb02+iq3*p.nb03) / 4;
[[unroll]] for (uint32_t idx = 0; idx < Br * D / 4; idx += gl_WorkGroupSize.x) { [[unroll]] for (uint32_t idx = 0; idx < Br * D / 4; idx += gl_WorkGroupSize.x) {

120
ggml/src/ggml-vulkan/vulkan-shaders/flash_attn_cm2.comp
View File

@@ -18,62 +18,12 @@
#include "types.comp" #include "types.comp"
#include "dequant_funcs_cm2.comp" #include "dequant_funcs_cm2.comp"
#include "flash_attn_base.comp"
layout(local_size_x_id = 0, local_size_y = 1, local_size_z = 1) in;
layout (constant_id = 1) const uint32_t Br = 32;
layout (constant_id = 2) const uint32_t Bc = 32;
layout (constant_id = 3) const uint32_t D = 32;
layout (constant_id = 4) const uint32_t Clamp = gl_CooperativeMatrixClampModeConstantNV;
layout (push_constant) uniform parameter {
uint32_t N;
uint32_t KV;
uint32_t ne1;
uint32_t ne2;
uint32_t ne3;
uint32_t neq2;
uint32_t neq3;
uint32_t nek2;
uint32_t nek3;
uint32_t nev2;
uint32_t nev3;
uint32_t nem1;
uint32_t nb01;
uint32_t nb02;
uint32_t nb03;
uint32_t nb11;
uint32_t nb12;
uint32_t nb13;
uint32_t nb21;
uint32_t nb22;
uint32_t nb23;
uint32_t nb31;
float scale;
float max_bias;
float logit_softcap;
uint32_t mask;
uint32_t n_head_log2;
float m0;
float m1;
uint32_t gqa_ratio;
uint32_t split_kv;
uint32_t k_num;
} p;
layout (binding = 0) readonly buffer Q {uint8_t data_q[];}; layout (binding = 0) readonly buffer Q {uint8_t data_q[];};
layout (binding = 1) readonly buffer K {uint8_t data_k[];}; layout (binding = 1) readonly buffer K {uint8_t data_k[];};
layout (binding = 2) readonly buffer V {uint8_t data_v[];}; layout (binding = 2) readonly buffer V {uint8_t data_v[];};
layout (binding = 3) readonly buffer M {uint8_t data_m[];}; layout (binding = 3) readonly buffer M {uint8_t data_m[];};
layout (binding = 4) writeonly buffer O {D_TYPE data_o[];};
#define CEIL_DIV(a, b) (((a) + (b) - 1) / (b))
ACC_TYPE maxReduce(const in ACC_TYPE x, const in ACC_TYPE y) { ACC_TYPE maxReduce(const in ACC_TYPE x, const in ACC_TYPE y) {
return max(x, y); return max(x, y);
@@ -118,67 +68,12 @@ D_TYPE perElemOpGqaStore(const in uint32_t r, const in uint32_t c, const in D_TY
return elem; return elem;
} }
// Store column zero. This is used to save per-row m and L values for split_k.
ACC_TYPE perElemOpStoreCol0(const in uint32_t r, const in uint32_t c, const in ACC_TYPE elem, const in uint32_t o_offset, const in uint32_t iq2, const in uint32_t N)
{
if (r < N && c == 0) {
uint32_t offset = iq2 + r;
data_o[o_offset + offset] = D_TYPE(elem);
}
return elem;
}
// Load the slope matrix, indexed by Q's dimension 2.
ACC_TYPE perElemOpComputeSlope(const in uint32_t r, const in uint32_t c, const in ACC_TYPE elem, const in uint32_t iq2)
{
const uint32_t h = iq2 + (r % p.gqa_ratio);
const ACC_TYPE base = ACC_TYPE(h < p.n_head_log2 ? p.m0 : p.m1);
const int exph = int(h < p.n_head_log2 ? h + 1 : 2*(h - p.n_head_log2) + 1);
return ACC_TYPE(pow(base, ACC_TYPE(exph)));
}
void main() { void main() {
#ifdef NEEDS_INIT_IQ_SHMEM #ifdef NEEDS_INIT_IQ_SHMEM
init_iq_shmem(gl_WorkGroupSize); init_iq_shmem(gl_WorkGroupSize);
#endif #endif
const uint32_t N = p.N; init_indices();
const uint32_t KV = p.KV;
uint32_t i = gl_WorkGroupID.x;
uint32_t split_k_index = 0;
if (p.k_num > 1) {
i = 0;
split_k_index = gl_WorkGroupID.x;
}
const uint32_t Tr = CEIL_DIV(N, Br);
const uint32_t start_j = split_k_index * p.split_kv / Bc;
const uint32_t end_j = CEIL_DIV(min(KV, (split_k_index + 1) * p.split_kv), Bc);
// When not using grouped query attention, all rows share the same iq2, equal to gl_WorkGroupID.y.
// When using grouped query attention, each workgroup does gqa_ratio consecutive values of iq2.
const uint32_t iq2 = gl_WorkGroupID.y * p.gqa_ratio;
const uint32_t iq3 = gl_WorkGroupID.z;
// broadcast factors
const uint32_t rk2 = p.neq2/p.nek2;
const uint32_t rk3 = p.neq3/p.nek3;
const uint32_t rv2 = p.neq2/p.nev2;
const uint32_t rv3 = p.neq3/p.nev3;
// k indices
const uint32_t ik3 = iq3 / rk3;
const uint32_t ik2 = iq2 / rk2;
// v indices
const uint32_t iv3 = iq3 / rv3;
const uint32_t iv2 = iq2 / rv2;
tensorLayoutNV<2, gl_CooperativeMatrixClampModeConstantNV> tensorLayoutQ = createTensorLayoutNV(2, gl_CooperativeMatrixClampModeConstantNV); tensorLayoutNV<2, gl_CooperativeMatrixClampModeConstantNV> tensorLayoutQ = createTensorLayoutNV(2, gl_CooperativeMatrixClampModeConstantNV);
tensorLayoutNV<2, Clamp> tensorLayoutK = createTensorLayoutNV(2, Clamp); tensorLayoutNV<2, Clamp> tensorLayoutK = createTensorLayoutNV(2, Clamp);
@@ -195,17 +90,6 @@ void main() {
tensorLayoutK = setTensorLayoutDimensionNV(tensorLayoutK, KV, D); tensorLayoutK = setTensorLayoutDimensionNV(tensorLayoutK, KV, D);
tensorLayoutV = setTensorLayoutDimensionNV(tensorLayoutV, KV, D); tensorLayoutV = setTensorLayoutDimensionNV(tensorLayoutV, KV, D);
// nb?1 are already divided by the type size and are in units of elements.
// When using grouped query attention, Q is indexed by iq2, so the stride
// should be nb02 (which is in bytes).
uint32_t q_stride = p.gqa_ratio > 1 ? (p.nb02 / 4) : p.nb01;
uint32_t k_stride = p.nb11;
uint32_t v_stride = p.nb21;
// When using grouped query attention, all rows use the same mask (stride 0).
// "p.gqa_ratio >> 16" is just a roundabout way of writing zero
// that prevents the compiler from folding the "&" through the select
// and breaking the alignment detection.
uint32_t m_stride = (p.gqa_ratio > 1) ? (p.gqa_ratio >> 16) : KV;
// hint to the compiler that strides are aligned for the aligned variant of the shader // hint to the compiler that strides are aligned for the aligned variant of the shader
if (Clamp != gl_CooperativeMatrixClampModeConstantNV) if (Clamp != gl_CooperativeMatrixClampModeConstantNV)
{ {