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

fix todo

Browse source
This commit is contained in:
hongruichen 2024-06-18 23:07:01 +08:00
parent 9456bba121
commit 65a14d9e9a
1 changed files with 1 additions and 1 deletions
Download patch file Download diff file Expand all files Collapse all files

2
ggml-qnn.cpp
View file

@ -1995,7 +1995,7 @@ public:
if (is_npu) { if (is_npu) {
qnn_instance * instance = ctx->instance; qnn_instance * instance = ctx->instance;
uint8_t *qnn_buffer = static_cast<uint8_t *>(instance->alloc_rpcmem( uint8_t *qnn_buffer = static_cast<uint8_t *>(instance->alloc_rpcmem(
ggml_nbytes(tensor), 4)); // TODO: should we get the align param from device here? ggml_nbytes(tensor), alignof(void*)));
if (!qnn_buffer) { if (!qnn_buffer) {
QNN_LOG_WARN("alloc rpcmem failure, %s\n", strerror(errno)); QNN_LOG_WARN("alloc rpcmem failure, %s\n", strerror(errno));
QNN_LOG_DEBUG("tensor%p name %s", _qnn_tensor, QNN_TENSOR_GET_NAME(*_qnn_tensor)); QNN_LOG_DEBUG("tensor%p name %s", _qnn_tensor, QNN_TENSOR_GET_NAME(*_qnn_tensor));