CS348Project/llama.cpp
5
0
Fork 0
mirror of https://github.com/ggml-org/llama.cpp.git synced 2025-11-13 10:57:15 +00:00
Code Issues Packages Projects Releases Wiki Activity

ggml-backend : add GGML_BACKEND_DEVICE_TYPE_IGPU device type (#15797)

Browse Source 
* ggml-backend : add GGML_BACKEND_DEVICE_TYPE_IGPU device type

ggml-backend : add device id to device props

llama : only use iGPU devices if there are no GPU devices

llama : do not use multiple devices from different backends with the same device id
This commit is contained in:
Diego Devesa
2025-09-11 13:47:38 -07:00
committed by GitHub
parent 0e6ff0046f
commit 360d6533db
7 changed files with 81 additions and 14 deletions
Download Patch File Download Diff File Expand all files Collapse all files

8
ggml/src/ggml-cuda/ggml-cuda.cu
View File

@@ -3210,6 +3210,7 @@ struct ggml_backend_cuda_device_context {
int device;
std::string name;
std::string description;
std::string pci_bus_id;
};
static const char * ggml_backend_cuda_device_get_name(ggml_backend_dev_t dev) {
@@ -3234,9 +3235,12 @@ static enum ggml_backend_dev_type ggml_backend_cuda_device_get_type(ggml_backend
}
static void ggml_backend_cuda_device_get_props(ggml_backend_dev_t dev, ggml_backend_dev_props * props) {
ggml_backend_cuda_device_context * ctx = (ggml_backend_cuda_device_context *)dev->context;
props->name = ggml_backend_cuda_device_get_name(dev);
props->description = ggml_backend_cuda_device_get_description(dev);
props->type = ggml_backend_cuda_device_get_type(dev);
props->device_id = ctx->pci_bus_id.empty() ? nullptr : ctx->pci_bus_id.c_str();
ggml_backend_cuda_device_get_memory(dev, &props->memory_free, &props->memory_total);
bool host_buffer = getenv("GGML_CUDA_NO_PINNED") == nullptr;
@@ -3804,6 +3808,10 @@ ggml_backend_reg_t ggml_backend_cuda_reg() {
CUDA_CHECK(cudaGetDeviceProperties(&prop, i));
dev_ctx->description = prop.name;
char pci_bus_id[16] = {};
snprintf(pci_bus_id, sizeof(pci_bus_id), "%04x:%02x:%02x.0", prop.pciDomainID, prop.pciBusID, prop.pciDeviceID);
dev_ctx->pci_bus_id = pci_bus_id;
ggml_backend_dev_t dev = new ggml_backend_device {
/* .iface = */ ggml_backend_cuda_device_interface,
/* .reg = */ &reg,