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

SYCL: Migrate away from deprecated ggml_tensor->backend (#10840)

Browse source 
* Migrate to tensor->buffer for checking backend buffer type: 1

* SYCL: common.cpp try to migrate away from tensor->backend

* SYCL: fix assertions and add proper comments

* SYCL: remove extra space

* SYCL: Add back static to ggml_backend_buffer_is_sycl_split function

* SYCL: Add pragma directive to suppress warning spam

* SYCL: Integrate debug logs with GGML_LOG and other fixes

* Revert "SYCL: Integrate debug logs with GGML_LOG and other fixes"

This reverts commit 2607b7de0f.
Let's keep the current SYCL specific logging mechanism for now

* SYCL: Use GGML_SYCL_DEBUG after reverting

* SYCL: reg_get_proc_address func, update to the current func signature

* SYCL: Refactor SYCL buffer checks in ggml_sycl_cpy_tensor_2d
This commit is contained in:
Akarshan Biswas 2024-12-20 21:01:28 +05:30 • committed by GitHub
parent 0ca416c91a
commit eb5c3dc64b
No known key found for this signature in database
GPG key ID: B5690EEEBB952194
3 changed files with 35 additions and 23 deletions
Download patch file Download diff file Expand all files Collapse all files

8
ggml/src/ggml-sycl/common.cpp
View file

@ -11,6 +11,8 @@
//
#include "common.hpp"
#include "ggml-backend-impl.h"
#include "ggml-impl.h"
int get_current_device_id() {
@ -65,9 +67,9 @@ void ggml_sycl_op_flatten(ggml_backend_sycl_context & ctx, const ggml_tensor *sr
const ggml_sycl_op_flatten_t op) try {
const bool use_src1 = src1 != nullptr;
GGML_ASSERT(!use_src1 || src1->backend != GGML_BACKEND_TYPE_GPU_SPLIT);
GGML_ASSERT( dst->backend != GGML_BACKEND_TYPE_GPU_SPLIT);
if(use_src1)
GGML_ASSERT(strcmp(src1->buffer->buft->iface.get_name(src1->buffer->buft), GGML_SYCL_NAME "_Split") != 0);
GGML_ASSERT(strcmp(dst->buffer->buft->iface.get_name(dst->buffer->buft), GGML_SYCL_NAME "_Split") != 0);
// dd = data device
float * src0_ddf = (float *) src0->data;