Mirrors/llama.cpp
1
0
Fork 0
mirror of https://github.com/ggerganov/llama.cpp.git synced 2024-12-25 22:08:46 +01:00
Code Issues Packages Projects Releases Wiki Activity

[SYCL] conext add name (#5624)

Browse Source 
* [SYCL] conext add name

* name should start with SYCL*
This commit is contained in:
Meng, Hengyu 2024-02-21 17:52:06 +08:00 committed by GitHub
parent a14679cc30
commit 88c46cbdac
No known key found for this signature in database
GPG Key ID: B5690EEEBB952194
1 changed files with 12 additions and 14 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

26
ggml-sycl.cpp
View File

@ -14642,7 +14642,8 @@ GGML_CALL static const char * ggml_backend_sycl_buffer_type_name(ggml_backend_bu
static ggml_backend_buffer_t static ggml_backend_buffer_t
ggml_backend_sycl_buffer_type_alloc_buffer(ggml_backend_buffer_type_t buft, ggml_backend_sycl_buffer_type_alloc_buffer(ggml_backend_buffer_type_t buft,
size_t size) try { size_t size) try {
int device = (int) (intptr_t) buft->context; ggml_backend_sycl_buffer_type_context * buft_ctx = (ggml_backend_sycl_buffer_type_context *)buft->context;
int device = (int) buft_ctx->device;
ggml_sycl_set_device(device); ggml_sycl_set_device(device);
int device_index = get_device_index_by_id(device); int device_index = get_device_index_by_id(device);
@ -14720,7 +14721,7 @@ ggml_backend_buffer_type_t ggml_backend_sycl_buffer_type(int device) {
for (int i = 0; i < GGML_SYCL_MAX_DEVICES; i++) { for (int i = 0; i < GGML_SYCL_MAX_DEVICES; i++) {
ggml_backend_sycl_buffer_types[i] = { ggml_backend_sycl_buffer_types[i] = {
/* .iface = */ ggml_backend_sycl_buffer_type_interface, /* .iface = */ ggml_backend_sycl_buffer_type_interface,
/* .context = */ (ggml_backend_buffer_type_context_t) (intptr_t) i, /* .context = */ new ggml_backend_sycl_buffer_type_context{i, GGML_SYCL_NAME + std::to_string(i)},
}; };
} }
ggml_backend_sycl_buffer_type_initialized = true; ggml_backend_sycl_buffer_type_initialized = true;
@ -14782,10 +14783,6 @@ ggml_backend_buffer_type_t ggml_backend_sycl_host_buffer_type() {
// backend // backend
struct ggml_backend_context_sycl {
int device;
};
static const char * ggml_backend_sycl_name(ggml_backend_t backend) { static const char * ggml_backend_sycl_name(ggml_backend_t backend) {
return GGML_SYCL_NAME; return GGML_SYCL_NAME;
@ -14793,14 +14790,14 @@ static const char * ggml_backend_sycl_name(ggml_backend_t backend) {
} }
static void ggml_backend_sycl_free(ggml_backend_t backend) { static void ggml_backend_sycl_free(ggml_backend_t backend) {
ggml_backend_context_sycl * sycl_ctx = (ggml_backend_context_sycl *)backend->context; ggml_backend_sycl_context * sycl_ctx = (ggml_backend_sycl_context *)backend->context;
delete sycl_ctx; delete sycl_ctx;
delete backend; delete backend;
} }
static ggml_backend_buffer_type_t ggml_backend_sycl_get_default_buffer_type(ggml_backend_t backend) { static ggml_backend_buffer_type_t ggml_backend_sycl_get_default_buffer_type(ggml_backend_t backend) {
ggml_backend_context_sycl * sycl_ctx = (ggml_backend_context_sycl *)backend->context; ggml_backend_sycl_context * sycl_ctx = (ggml_backend_sycl_context *)backend->context;
return ggml_backend_sycl_buffer_type(sycl_ctx->device); return ggml_backend_sycl_buffer_type(sycl_ctx->device);
} }
@ -14809,7 +14806,7 @@ static void ggml_backend_sycl_set_tensor_async(ggml_backend_t backend,
ggml_tensor *tensor, ggml_tensor *tensor,
const void *data, size_t offset, const void *data, size_t offset,
size_t size) try { size_t size) try {
ggml_backend_context_sycl * sycl_ctx = (ggml_backend_context_sycl *)backend->context; ggml_backend_sycl_context * sycl_ctx = (ggml_backend_sycl_context *)backend->context;
GGML_ASSERT(tensor->buffer->buft == ggml_backend_sycl_buffer_type(sycl_ctx->device) && "unsupported buffer type"); GGML_ASSERT(tensor->buffer->buft == ggml_backend_sycl_buffer_type(sycl_ctx->device) && "unsupported buffer type");
GGML_ASSERT(tensor->backend == GGML_BACKEND_GPU); GGML_ASSERT(tensor->backend == GGML_BACKEND_GPU);
@ -14827,7 +14824,7 @@ static void ggml_backend_sycl_get_tensor_async(ggml_backend_t backend,
const ggml_tensor *tensor, const ggml_tensor *tensor,
void *data, size_t offset, void *data, size_t offset,
size_t size) try { size_t size) try {
ggml_backend_context_sycl * sycl_ctx = (ggml_backend_context_sycl *)backend->context; ggml_backend_sycl_context * sycl_ctx = (ggml_backend_sycl_context *)backend->context;
GGML_ASSERT(tensor->buffer->buft == ggml_backend_sycl_buffer_type(sycl_ctx->device) && "unsupported buffer type"); GGML_ASSERT(tensor->buffer->buft == ggml_backend_sycl_buffer_type(sycl_ctx->device) && "unsupported buffer type");
GGML_ASSERT(tensor->backend == GGML_BACKEND_GPU); GGML_ASSERT(tensor->backend == GGML_BACKEND_GPU);
@ -14842,7 +14839,7 @@ catch (sycl::exception const &exc) {
} }
static void ggml_backend_sycl_synchronize(ggml_backend_t backend) try { static void ggml_backend_sycl_synchronize(ggml_backend_t backend) try {
ggml_backend_context_sycl * sycl_ctx = (ggml_backend_context_sycl *)backend->context; ggml_backend_sycl_context * sycl_ctx = (ggml_backend_sycl_context *)backend->context;
SYCL_CHECK(CHECK_TRY_ERROR(g_syclStreams[sycl_ctx->device][0]->wait())); SYCL_CHECK(CHECK_TRY_ERROR(g_syclStreams[sycl_ctx->device][0]->wait()));
@ -14878,7 +14875,7 @@ static void ggml_backend_sycl_graph_plan_compute(ggml_backend_t backend, ggml_ba
} }
static bool ggml_backend_sycl_graph_compute(ggml_backend_t backend, ggml_cgraph * cgraph) { static bool ggml_backend_sycl_graph_compute(ggml_backend_t backend, ggml_cgraph * cgraph) {
ggml_backend_context_sycl * sycl_ctx = (ggml_backend_context_sycl *)backend->context; ggml_backend_sycl_context * sycl_ctx = (ggml_backend_sycl_context *)backend->context;
ggml_sycl_set_main_device(sycl_ctx->device); ggml_sycl_set_main_device(sycl_ctx->device);
@ -15092,8 +15089,9 @@ ggml_backend_t ggml_backend_sycl_init(int device) {
// not strictly necessary, but it may reduce the overhead of the first graph_compute // not strictly necessary, but it may reduce the overhead of the first graph_compute
ggml_sycl_set_main_device(device); ggml_sycl_set_main_device(device);
ggml_backend_context_sycl * ctx = new ggml_backend_context_sycl { ggml_backend_sycl_context * ctx = new ggml_backend_sycl_context {
/* .device = */ device /* .device = */ device,
/* .name = */ GGML_SYCL_NAME + std::to_string(device),
}; };
ggml_backend_t sycl_backend = new ggml_backend { ggml_backend_t sycl_backend = new ggml_backend {