mirrors/llama.cpp
1
0
Fork 0
mirror of https://github.com/ggerganov/llama.cpp.git synced 2025-04-14 10:36:07 +00:00
Code Issues Packages Projects Releases Wiki Activity

Revert "sycl:remove redundant memcopy in function ggml_backend_sycl_buffer_set_tensor" (#12812)

Browse Source 
* Revert "sycl: remove redundant memcopy in function ggml_backend_sycl_buffer_s…"

This reverts commit 518a01480eb3a7c80a4951b430db9dee55428310.

* Update ggml/src/ggml-sycl/ggml-sycl.cpp

* Update ggml/src/ggml-sycl/ggml-sycl.cpp

* rm tail space
This commit is contained in:
Neo Zhang Jianyu 2025-04-08 15:03:21 +08:00 committed by GitHub
parent a226bc7a9a
commit 656babd6c2
No known key found for this signature in database
GPG Key ID: B5690EEEBB952194
1 changed files with 6 additions and 1 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

7
ggml/src/ggml-sycl/ggml-sycl.cpp
View File

@ -372,9 +372,14 @@ static void ggml_backend_sycl_buffer_set_tensor(ggml_backend_buffer_t buffer,
auto stream = &(dpct::dev_mgr::instance().get_device(ctx->device).default_queue());
SYCL_CHECK(
CHECK_TRY_ERROR(dpct::dev_mgr::instance().get_device(ctx->device).queues_wait_and_throw()));
// Note: Use host buffer to save the data from mmap(), then copy to device. It's workaround for mmap() issue on PVC GPU.
// This function will be called during load model from disk. Use memory buffer replace dynamic won't save more time and brings potential memory leak risk here.
char* host_buf = (char*)malloc(size);
memcpy(host_buf, data, size);
SYCL_CHECK(
CHECK_TRY_ERROR((*stream).memcpy((char *)tensor->data + offset, data, size)
CHECK_TRY_ERROR((*stream).memcpy((char *)tensor->data + offset, host_buf, size)
.wait()));
free(host_buf);
}
catch (sycl::exception const &exc) {
std::cerr << exc.what() << "Exception caught at file:" << __FILE__