root/llama.cpp
1
0
Fork 0
mirror of https://github.com/ggerganov/llama.cpp.git synced 2024-12-29 04:44:34 +00:00
Code Issues Actions 2 Packages Projects Releases Wiki Activity

kompute : fix compile warnings

Browse Source
This commit is contained in:
Jared Van Bortel 2023-12-13 17:30:38 -05:00
parent 747e1eafcf
commit b906e126ca
1 changed files with 36 additions and 34 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

70
ggml-kompute.cpp
View File

@ -64,7 +64,7 @@ struct ggml_kompute_context {
// we *have* to have the kompute manager no matter what for device discovery, but the kompute context
// is only created when a device is set and vulkan is explicitly turned on.
ggml_kompute_context *s_kompute_context = nullptr;
kp::Manager *komputeManager() {
static kp::Manager *komputeManager() {
static kp::Manager *s_mgr = nullptr;
if (s_mgr && !s_mgr->hasInstance()) {
delete s_mgr;
@ -551,7 +551,7 @@ void ggml_vk_d2h_tensor(struct ggml_kompute_context * ctx, struct ggml_tensor *
komputeManager()->sequence()->eval<kp::OpTensorSyncLocal>({res});
}
std::vector<uint32_t> getSpirvShader(const unsigned char* rawData, size_t size) {
static std::vector<uint32_t> getSpirvShader(const unsigned char* rawData, size_t size) {
if (size % sizeof(uint32_t) != 0) {
throw std::runtime_error("Invalid size: must be divisible by sizeof(uint32_t)");
}
@ -573,7 +573,7 @@ uint32_t safe_divide(uint32_t a, uint32_t b) {
return a / b;
}
void ggml_vk_add(
static void ggml_vk_add(
kp::Sequence& seq,
const std::shared_ptr<kp::Tensor>& inA,
const std::shared_ptr<kp::Tensor>& inB,
@ -621,7 +621,7 @@ void ggml_vk_add(
seq.record<kp::OpAlgoDispatch>(s_algo);
}
void ggml_vk_addrow(kp::Sequence& seq,
static void ggml_vk_addrow(kp::Sequence& seq,
const std::shared_ptr<kp::Tensor>& inA,
const std::shared_ptr<kp::Tensor>& inB,
const std::shared_ptr<kp::Tensor>& out,
@ -652,7 +652,7 @@ void ggml_vk_addrow(kp::Sequence& seq,
seq.record<kp::OpAlgoDispatch>(s_algo);
}
void ggml_vk_mul(kp::Sequence& seq,
static void ggml_vk_mul(kp::Sequence& seq,
const std::shared_ptr<kp::Tensor>& inA,
const std::shared_ptr<kp::Tensor>& inB,
const std::shared_ptr<kp::Tensor>& out,
@ -681,7 +681,7 @@ void ggml_vk_mul(kp::Sequence& seq,
seq.record<kp::OpAlgoDispatch>(s_algo);
}
void ggml_vk_mulrow(kp::Sequence& seq,
static void ggml_vk_mulrow(kp::Sequence& seq,
const std::shared_ptr<kp::Tensor>& inA,
const std::shared_ptr<kp::Tensor>& inB,
const std::shared_ptr<kp::Tensor>& out,
@ -712,7 +712,7 @@ void ggml_vk_mulrow(kp::Sequence& seq,
seq.record<kp::OpAlgoDispatch>(s_algo);
}
void ggml_vk_scale(kp::Sequence& seq,
static void ggml_vk_scale(kp::Sequence& seq,
const std::shared_ptr<kp::Tensor>& in,
const std::shared_ptr<kp::Tensor>& out,
uint32_t inOff, uint32_t outOff,
@ -753,7 +753,7 @@ void ggml_vk_scale(kp::Sequence& seq,
seq.record<kp::OpAlgoDispatch>(s_algo);
}
void ggml_vk_xxlu(const std::vector<uint32_t>& spirv, kp::Sequence& seq,
static void ggml_vk_xxlu(const std::vector<uint32_t>& spirv, kp::Sequence& seq,
const std::shared_ptr<kp::Tensor>& in,
const std::shared_ptr<kp::Tensor>& out,
uint32_t inOff, uint32_t outOff,
@ -778,7 +778,7 @@ void ggml_vk_xxlu(const std::vector<uint32_t>& spirv, kp::Sequence& seq,
}
template <typename... Args>
void ggml_vk_silu(Args&&... args) {
static void ggml_vk_silu(Args&&... args) {
const static auto spirv = getSpirvShader(kp::shader_data::op_silu_comp_spv,
kp::shader_data::op_silu_comp_spv_len);
@ -786,7 +786,7 @@ void ggml_vk_silu(Args&&... args) {
}
template <typename... Args>
void ggml_vk_relu(Args&&... args) {
static void ggml_vk_relu(Args&&... args) {
const static auto spirv = getSpirvShader(kp::shader_data::op_relu_comp_spv,
kp::shader_data::op_relu_comp_spv_len);
@ -794,14 +794,14 @@ void ggml_vk_relu(Args&&... args) {
}
template <typename... Args>
void ggml_vk_gelu(Args&&... args) {
static void ggml_vk_gelu(Args&&... args) {
const static auto spirv = getSpirvShader(kp::shader_data::op_gelu_comp_spv,
kp::shader_data::op_gelu_comp_spv_len);
ggml_vk_xxlu(spirv, std::forward<Args>(args)...);
}
void ggml_vk_soft_max(kp::Sequence& seq,
static void ggml_vk_soft_max(kp::Sequence& seq,
const std::shared_ptr<kp::Tensor>& in,
const std::shared_ptr<kp::Tensor>& out,
uint32_t inOff, uint32_t outOff,
@ -833,7 +833,7 @@ void ggml_vk_soft_max(kp::Sequence& seq,
seq.record<kp::OpAlgoDispatch>(s_algo);
}
void ggml_vk_norm_(const std::vector<uint32_t>& spirv, kp::Sequence& seq,
static void ggml_vk_norm_(const std::vector<uint32_t>& spirv, kp::Sequence& seq,
const std::shared_ptr<kp::Tensor>& in,
const std::shared_ptr<kp::Tensor>& out,
uint32_t inOff, uint32_t outOff,
@ -865,7 +865,7 @@ void ggml_vk_norm_(const std::vector<uint32_t>& spirv, kp::Sequence& seq,
}
template <typename... Args>
void ggml_vk_norm(Args&&... args) {
static void ggml_vk_norm(Args&&... args) {
const static auto spirv = getSpirvShader(kp::shader_data::op_norm_comp_spv,
kp::shader_data::op_norm_comp_spv_len);
@ -873,14 +873,14 @@ void ggml_vk_norm(Args&&... args) {
}
template <typename... Args>
void ggml_vk_rms_norm(Args&&... args) {
static void ggml_vk_rms_norm(Args&&... args) {
const static auto spirv = getSpirvShader(kp::shader_data::op_rmsnorm_comp_spv,
kp::shader_data::op_rmsnorm_comp_spv_len);
ggml_vk_norm_(spirv, std::forward<Args>(args)...);
}
void ggml_vk_diag_mask_inf(kp::Sequence& seq,
static void ggml_vk_diag_mask_inf(kp::Sequence& seq,
const std::shared_ptr<kp::Tensor>& in,
const std::shared_ptr<kp::Tensor>& out,
uint32_t inOff, uint32_t outOff,
@ -912,7 +912,7 @@ void ggml_vk_diag_mask_inf(kp::Sequence& seq,
seq.record<kp::OpAlgoDispatch>(s_algo);
}
void ggml_vk_mul_mat_f16(kp::Sequence& seq,
static void ggml_vk_mul_mat_f16(kp::Sequence& seq,
const std::shared_ptr<kp::Tensor>& inA,
const std::shared_ptr<kp::Tensor>& inB,
const std::shared_ptr<kp::Tensor>& out,
@ -951,7 +951,7 @@ void ggml_vk_mul_mat_f16(kp::Sequence& seq,
seq.record<kp::OpAlgoDispatch>(s_algo);
}
void ggml_vk_mul_mat_q8_0(kp::Sequence& seq,
static void ggml_vk_mul_mat_q8_0(kp::Sequence& seq,
const std::shared_ptr<kp::Tensor>& inA,
const std::shared_ptr<kp::Tensor>& inB,
const std::shared_ptr<kp::Tensor>& out,
@ -989,7 +989,7 @@ void ggml_vk_mul_mat_q8_0(kp::Sequence& seq,
}
void ggml_vk_mul_mat_mat_f32(kp::Sequence& seq,
static void ggml_vk_mul_mat_mat_f32(kp::Sequence& seq,
const std::shared_ptr<kp::Tensor>& inA,
const std::shared_ptr<kp::Tensor>& inB,
const std::shared_ptr<kp::Tensor>& out,
@ -1039,7 +1039,7 @@ void ggml_vk_mul_mat_mat_f32(kp::Sequence& seq,
seq.record<kp::OpAlgoDispatch>(s_algo);
}
void ggml_vk_mul_mat_q4_x(const std::vector<uint32_t>& spirv, uint32_t block_size, kp::Sequence& seq,
static void ggml_vk_mul_mat_q4_x(const std::vector<uint32_t>& spirv, uint32_t block_size, kp::Sequence& seq,
const std::shared_ptr<kp::Tensor>& inA,
const std::shared_ptr<kp::Tensor>& inB,
const std::shared_ptr<kp::Tensor>& out,
@ -1069,7 +1069,7 @@ void ggml_vk_mul_mat_q4_x(const std::vector<uint32_t>& spirv, uint32_t block_siz
}
template <typename... Args>
void ggml_vk_mul_mat_q4_0(Args&&... args) {
static void ggml_vk_mul_mat_q4_0(Args&&... args) {
const static auto spirv = getSpirvShader(kp::shader_data::op_mul_mat_q4_0_comp_spv,
kp::shader_data::op_mul_mat_q4_0_comp_spv_len);
@ -1077,14 +1077,14 @@ void ggml_vk_mul_mat_q4_0(Args&&... args) {
}
template <typename... Args>
void ggml_vk_mul_mat_q4_1(Args&&... args) {
static void ggml_vk_mul_mat_q4_1(Args&&... args) {
const static auto spirv = getSpirvShader(kp::shader_data::op_mul_mat_q4_1_comp_spv,
kp::shader_data::op_mul_mat_q4_1_comp_spv_len);
ggml_vk_mul_mat_q4_x(spirv, 1/*We access blocks unaligned*/, std::forward<Args>(args)...);
}
void ggml_vk_mul_mat_q6_k(kp::Sequence& seq,
static void ggml_vk_mul_mat_q6_k(kp::Sequence& seq,
const std::shared_ptr<kp::Tensor>& inA,
const std::shared_ptr<kp::Tensor>& inB,
const std::shared_ptr<kp::Tensor>& out,
@ -1116,7 +1116,7 @@ void ggml_vk_mul_mat_q6_k(kp::Sequence& seq,
seq.record<kp::OpAlgoDispatch>(s_algo);
}
void ggml_vk_get_rows(const std::vector<uint32_t>& spirv,
static void ggml_vk_get_rows(const std::vector<uint32_t>& spirv,
unsigned element_size, unsigned qk,
kp::Sequence& seq,
const std::shared_ptr<kp::Tensor>& inA,
@ -1151,7 +1151,7 @@ void ggml_vk_get_rows(const std::vector<uint32_t>& spirv,
}
template <typename... Args>
void ggml_vk_get_rows_f16(Args&&... args) {
static void ggml_vk_get_rows_f16(Args&&... args) {
const static auto spirv = getSpirvShader(kp::shader_data::op_getrows_f16_comp_spv,
kp::shader_data::op_getrows_f16_comp_spv_len);
@ -1159,7 +1159,7 @@ void ggml_vk_get_rows_f16(Args&&... args) {
}
template <typename... Args>
void ggml_vk_get_rows_q4_0(Args&&... args) {
static void ggml_vk_get_rows_q4_0(Args&&... args) {
const static auto spirv = getSpirvShader(kp::shader_data::op_getrows_q4_0_comp_spv,
kp::shader_data::op_getrows_q4_0_comp_spv_len);
@ -1167,7 +1167,7 @@ void ggml_vk_get_rows_q4_0(Args&&... args) {
}
template <typename... Args>
void ggml_vk_get_rows_q4_1(Args&&... args) {
static void ggml_vk_get_rows_q4_1(Args&&... args) {
const static auto spirv = getSpirvShader(kp::shader_data::op_getrows_q4_1_comp_spv,
kp::shader_data::op_getrows_q4_1_comp_spv_len);
@ -1175,13 +1175,13 @@ void ggml_vk_get_rows_q4_1(Args&&... args) {
}
template <typename... Args>
void ggml_vk_get_rows_q6_k(Args&&... args) {
static void ggml_vk_get_rows_q6_k(Args&&... args) {
const static auto spirv = getSpirvShader(kp::shader_data::op_getrows_q6_k_comp_spv,
kp::shader_data::op_getrows_q6_k_comp_spv_len);
ggml_vk_get_rows(spirv, 1/*We access blocks unaligned*/, QK_NL, std::forward<Args>(args)...);
}
void ggml_vk_rope(
static void ggml_vk_rope(
kp::Sequence& seq,
const std::shared_ptr<kp::Tensor>& inA,
const std::shared_ptr<kp::Tensor>& inB,
@ -1249,7 +1249,7 @@ void ggml_vk_rope(
}
template<uint32_t in_element_size, uint32_t out_element_size>
void ggml_vk_cpy(const std::vector<uint32_t>& spirv,
static void ggml_vk_cpy(const std::vector<uint32_t>& spirv,
kp::Sequence& seq,
const std::shared_ptr<kp::Tensor>& in,
const std::shared_ptr<kp::Tensor>& out,
@ -1289,28 +1289,28 @@ void ggml_vk_cpy(const std::vector<uint32_t>& spirv,
}
template <typename... Args>
void ggml_vk_cpy_f32_f16(Args&&... args) {
static void ggml_vk_cpy_f32_f16(Args&&... args) {
const static auto spirv = getSpirvShader(kp::shader_data::op_cpy_f32_f16_comp_spv,
kp::shader_data::op_cpy_f32_f16_comp_spv_len);
ggml_vk_cpy<4, 2>(spirv, std::forward<Args>(args)...);
}
template <typename... Args>
void ggml_vk_cpy_f32_f32(Args&&... args) {
static void ggml_vk_cpy_f32_f32(Args&&... args) {
const static auto spirv = getSpirvShader(kp::shader_data::op_cpy_f32_f32_comp_spv,
kp::shader_data::op_cpy_f32_f32_comp_spv_len);
ggml_vk_cpy<4, 4>(spirv, std::forward<Args>(args)...);
}
template <typename... Args>
void ggml_vk_cpy_f16_f16(Args&&... args) {
static void ggml_vk_cpy_f16_f16(Args&&... args) {
const static auto spirv = getSpirvShader(kp::shader_data::op_cpy_f16_f16_comp_spv,
kp::shader_data::op_cpy_f16_f16_comp_spv_len);
ggml_vk_cpy<2, 2>(spirv, std::forward<Args>(args)...);
}
template <typename... Args>
void ggml_vk_cpy_f16_f32(Args&&... args) {
static void ggml_vk_cpy_f16_f32(Args&&... args) {
const static auto spirv = getSpirvShader(kp::shader_data::op_cpy_f16_f32_comp_spv,
kp::shader_data::op_cpy_f16_f32_comp_spv_len);
ggml_vk_cpy<2, 4>(spirv, std::forward<Args>(args)...);
@ -1349,6 +1349,8 @@ void ggml_vk_graph_compute(struct ggml_kompute_context * ctx, struct ggml_cgraph
case GGML_OP_TRANSPOSE:
case GGML_OP_PERMUTE:
continue; // noop -> next node
default:
break;
}
const int32_t ne00 = src0 ? src0->ne[0] : 0;