Superminaren/llama.cpp
1
0
Fork 0
mirror of https://github.com/ggml-org/llama.cpp.git synced 2026-09-24 13:37:01 +02:00
Code Issues Packages Projects Releases Wiki Activity

ggml-meta: resolve multi buffer views (#29266)

Browse Source 
* ggml-meta: resolve multi buffer views

* add TODO to revisit if graph allocator gets refactored
This commit is contained in:
Ruben Ortlam
2026-09-23 07:35:24 +02:00
committed by GitHub
parent 1a679828f3
commit 94256114c2
1 changed files with 13 additions and 7 deletions
Download Patch File Download Diff File Expand all files Collapse all files
ggml/src/ggml-backend-meta.cpp
+13 -7
View File
@@ -1200,11 +1200,6 @@ static enum ggml_status ggml_backend_meta_buffer_init_tensor_impl(ggml_backend_m
ggml_context * simple_ctx = stc.ctxs[j].get();
ggml_backend_buffer_t simple_buf = buf_ctx->bufs[j].get();
if ((simple_buf != nullptr) && ggml_backend_buffer_is_multi_buffer(simple_buf)) {
// see https://github.com/ggml-org/llama.cpp/issues/22197
GGML_ABORT("multi buffers are not supported by the meta backend");
}
if (split_dim >= 0 && split_dim < GGML_MAX_DIMS) {
// TODO: the following assert fails for llama-parallel even though the results are correct:
// GGML_ASSERT(ggml_is_contiguously_allocated(tensor));
@@ -1252,16 +1247,27 @@ static enum ggml_status ggml_backend_meta_buffer_init_tensor_impl(ggml_backend_m
}
}
}
// TODO: revisit once the graph allocator has been refactored, see https://github.com/ggml-org/llama.cpp/pull/25051#issuecomment-4842873396
ggml_backend_buffer_t init_buf = simple_buf;
if (t_ij->view_src != nullptr) {
t_ij->data = (char *) t_ij->view_src->data + t_ij->view_offs;
// views inherit the source slice's concrete sub-buffer (issue 22197)
if (tensor->view_src != nullptr && ggml_backend_buffer_is_meta(tensor->view_src->buffer)
&& t_ij->view_src->buffer != nullptr) {
t_ij->buffer = t_ij->view_src->buffer;
init_buf = t_ij->view_src->buffer;
}
} else if (simple_buf != nullptr) {
if (ggml_backend_buffer_is_multi_buffer(simple_buf)) {
GGML_ABORT("multi buffers are not supported by the meta backend");
}
t_ij->data = (char *) ggml_backend_buffer_get_base(simple_buf)
+ size_t(tensor->data) - size_t(ggml_backend_buffer_get_base(tensor->buffer));
}
if (simple_buf) {
if (init_buf) {
// the backend that owns the buffer will set .extra
ggml_backend_buffer_init_tensor(simple_buf, t_ij);
ggml_backend_buffer_init_tensor(init_buf, t_ij);
} else {
t_ij->extra = tensor->extra;
}