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

fix error of select non-zero device, format device list

Browse source
This commit is contained in:
jianyuzh 2024-01-08 14:23:55 +08:00 • committed by Meng, Hengyu
parent c2ef7a9cb9
commit 69d76c8b58
1 changed files with 37 additions and 23 deletions
Download patch file Download diff file Expand all files Collapse all files

60
ggml-sycl.cpp
View file

@ -63,6 +63,10 @@ static int g_ggml_sycl_debug=0;
static_assert(sizeof(sycl::half) == sizeof(ggml_fp16_t), "wrong fp16 size");
static void crash(){
int *ptr = NULL;
*ptr = 0;
}
static void ggml_sycl_error(const char * stmt, const char * func, const char * file, const int line, const char * msg) {
fprintf(stderr, "SYCL error: %s: %s\n", stmt, msg);
@ -358,22 +362,21 @@ struct ggml_tensor_extra_gpu {
inline dpct::err0 ggml_sycl_set_device(const int device) try {
int current_device;
SYCL_CHECK(CHECK_TRY_ERROR(
current_device = dpct::dev_mgr::instance().current_device_id()));
// GGML_SYCL_DEBUG("ggml_sycl_set_device device=%d, current_device=%d\n", device, current_device);
if (device == current_device) {
return 0;
}
/*
DPCT1093:53: The "device" device may be not the one intended for use. Adjust
the selected device if needed.
*/
return CHECK_TRY_ERROR(dpct::select_device(device));
}
catch (sycl::exception const &exc) {
std::cerr << exc.what() << "Exception caught at file:" << __FILE__
<< ", line:" << __LINE__ << std::endl;
crash();
std::exit(1);
}
@ -6391,6 +6394,13 @@ int get_device_index_by_id(int id){
return res;
}
int get_device_id_by_index(int index){
int res = g_device_caps[index].device_id;
GGML_ASSERT(res>=0);
return res;
}
int get_current_device_index(){
return get_device_index_by_id(dpct::dev_mgr::instance().current_device_id());
}
@ -8258,12 +8268,13 @@ bool ggml_sycl_loaded(void) {
}
void print_devices(){
int device_count = dpct::dev_mgr::instance().device_count();
fprintf(stderr, "%s: found %d SYCL devices:\n", __func__, device_count);
fprintf(stderr, "\n%s: found %d SYCL devices:\n", __func__, device_count);
for (int id = 0; id < device_count; ++id) {
dpct::device_info prop;
SYCL_CHECK(CHECK_TRY_ERROR(dpct::get_device_info(
prop, dpct::dev_mgr::instance().get_device(id))));
fprintf(stderr, " Device %d: %s, compute capability %d.%d, max compute_units %d, max work group size %d, max sub group size %d, global mem size %lu\n", id,
sycl::device cur_device = dpct::dev_mgr::instance().get_device(id);
fprintf(stderr, " Device %d: %s,\tcompute capability %d.%d,\n\tmax compute_units %d,\tmax work group size %d,\tmax sub group size %d,\tglobal mem size %lu\n", id,
prop.get_name(), prop.get_major_version(),
prop.get_minor_version(),
prop.get_max_compute_units(),
@ -8272,6 +8283,7 @@ void print_devices(){
prop.get_global_mem_size()
);
}
fprintf(stderr, "\n");
}
int get_sycl_env(const char* env_name, int default_val){
@ -8302,7 +8314,7 @@ void ggml_init_sycl() try {
printf("GGML_SYCL_DEBUG=%d\n", g_ggml_sycl_debug);
int user_device_number = get_sycl_env("GGML_SYCL_DEVICE", 0);
int user_device_id = get_sycl_env("GGML_SYCL_DEVICE", 0);
print_devices();
@ -8341,7 +8353,7 @@ void ggml_init_sycl() try {
int device_inx = -1;
for (int id = 0; id < g_all_sycl_device_count; ++id) {
if(id!=user_device_number) continue;
if(id!=user_device_id) continue;
device_inx++;
int device_vmm = 0;
@ -8369,14 +8381,14 @@ void ggml_init_sycl() try {
}
device_inx = -1;
for (int id = 0; id < g_all_sycl_device_count; ++id) {
if(id!=user_device_number) continue;
if(id!=user_device_id) continue;
device_inx++;
g_tensor_split[device_inx] /= total_vram;
}
device_inx = -1;
for (int id = 0; id < g_all_sycl_device_count; ++id) {
if(id!=user_device_number) continue;
if(id!=user_device_id) continue;
device_inx++;
SYCL_CHECK(ggml_sycl_set_device(id));
@ -8406,9 +8418,9 @@ void ggml_init_sycl() try {
//hardcode, force set to 1 device
g_device_count = 1;
ggml_sycl_set_main_device(user_device_number);
ggml_sycl_set_device(user_device_number);
fprintf(stderr, "Using Device %d\n", user_device_number);
ggml_sycl_set_main_device(user_device_id);
ggml_sycl_set_device(user_device_id);
fprintf(stderr, "Using Device %d\n", user_device_id);
// for (int id = 0; id < g_all_sycl_device_count; ++id) {
// GGML_SYCL_DEBUG("id=%d g_device_caps[%d].device_id=%d g_sycl_device_id2index[%d].index=%d ", id, id,
@ -9583,7 +9595,8 @@ static void ggml_sycl_op_flatten(const ggml_tensor *src0,
ggml_sycl_set_device(g_main_device);
dpct::queue_ptr main_stream = g_syclStreams[g_main_device_index][0];
GGML_SYCL_DEBUG("g_main_device_index=%d, src0=%p\n", g_main_device_index, src0);
GGML_SYCL_DEBUG("g_main_device_index=%d, src0=%p main_stream=%p src0_on_device=%d\n",
g_main_device_index, src0, main_stream, src0_on_device);
if (src0_on_device) {
src0_ddf = (float *) src0_extra->data_device[g_main_device_index];
@ -9609,6 +9622,8 @@ static void ggml_sycl_op_flatten(const ggml_tensor *src0,
dst_ddf = dst_f.alloc(ggml_nelements(dst));
}
GGML_SYCL_DEBUG("op src0=%p, src1=%p, dst=%p, src0_ddf=%p, src1_ddf=%p, dst_ddf=%p, main_stream=%p\n",
src0, src1, dst, src0_ddf, src1_ddf, dst_ddf, main_stream);
// do the computation
op(src0, src1, dst, src0_ddf, src1_ddf, dst_ddf, main_stream);
/*
@ -9648,12 +9663,12 @@ static void ggml_sycl_set_peer_access(const int n_tokens) {
#ifdef NDEBUG
for (int id = 0; id < g_device_count; ++id) {
SYCL_CHECK(ggml_sycl_set_device(g_device_caps[id].device_id));
SYCL_CHECK(ggml_sycl_set_device(get_device_id_by_index(id)));
// SYCL_CHECK(syclDeviceSynchronize());
}
for (int id = 0; id < g_device_count; ++id) {
SYCL_CHECK(ggml_sycl_set_device(g_device_caps[id].device_id));
SYCL_CHECK(ggml_sycl_set_device(get_device_id_by_index(id)));
int device_id = g_device_caps[id].device_id;
for (int id_other = 0; id_other < g_device_count; ++id_other) {
@ -9786,7 +9801,7 @@ static void ggml_sycl_op_mul_mat(const ggml_tensor *src0,
const bool src1_on_device = src1->backend == GGML_BACKEND_GPU && id == g_main_device_index;
const bool dst_on_device = dst->backend == GGML_BACKEND_GPU && id == g_main_device_index;
ggml_sycl_set_device(id);
ggml_sycl_set_device(get_device_id_by_index(id));
const dpct::queue_ptr stream = g_syclStreams[id][0];
if (src0_on_device && src0_is_contiguous) {
@ -9852,7 +9867,7 @@ static void ggml_sycl_op_mul_mat(const ggml_tensor *src0,
const bool dst_on_device = dst->backend == GGML_BACKEND_GPU && id == g_main_device_index;
const int64_t row_diff = row_high[id] - row_low[id];
ggml_sycl_set_device(id);
ggml_sycl_set_device(get_device_id_by_index(id));
const dpct::queue_ptr stream = g_syclStreams[id][is];
// wait for main GPU data if necessary
@ -9983,7 +9998,7 @@ static void ggml_sycl_op_mul_mat(const ggml_tensor *src0,
if ((!split && id != g_main_device_index) || row_low[id] == row_high[id]) {
continue;
}
SYCL_CHECK(ggml_sycl_set_device(id));
SYCL_CHECK(ggml_sycl_set_device(get_device_id_by_index(id)));
// free buffers again when done
if (dst_as[id] > 0) {
@ -10978,8 +10993,7 @@ void ggml_sycl_transform_tensor(void *data, struct ggml_tensor *tensor) try {
if (backend == GGML_BACKEND_GPU && id != g_main_device_index) {
continue;
}
ggml_sycl_set_device(id);
ggml_sycl_set_device(get_device_id_by_index(id));
int64_t row_low, row_high;
if (backend == GGML_BACKEND_GPU) {
@ -11059,14 +11073,14 @@ void ggml_sycl_free_data(struct ggml_tensor *tensor) try {
for (int64_t id = 0; id < g_device_count; ++id) {
if (extra->data_device[id] != nullptr) {
SYCL_CHECK(ggml_sycl_set_device(id));
SYCL_CHECK(ggml_sycl_set_device(get_device_id_by_index(id)));
SYCL_CHECK(CHECK_TRY_ERROR(sycl::free(
extra->data_device[id], dpct::get_in_order_queue())));
}
for (int64_t is = 0; is < MAX_STREAMS; ++is) {
if (extra->events[id][is] != nullptr) {
SYCL_CHECK(ggml_sycl_set_device(id));
SYCL_CHECK(ggml_sycl_set_device(get_device_id_by_index(id)));
SYCL_CHECK(CHECK_TRY_ERROR(
dpct::destroy_event(extra->events[id][is])));
}