-
Notifications
You must be signed in to change notification settings - Fork 2.4k
Commit
This commit does not belong to any branch on this repository, and may belong to a fork outside of the repository.
[Snippets][CPU] Optimized reg spills in BrgemmCopyBKernel (#28635)
### Details: - *BrgemmCopyBKernel is JIT compiled kernel which emits blocking loop by `N` dimension with binary call of the kernel `dnnl::x64::brgemm_copy_b_kernel` inside. Currently, we spill all registers before the each call of this kernel. It means that if the dimension `N` is quite big, there will be many binary calls and reg spills. These regular reg spills may lead to perf degradation for Subgraphs with small `M`, `K,` and big `N` dimensions. This PR uses feature of binary call optimizations - spills only live registers* ### Tickets: - *N/A*
- Loading branch information
1 parent
9930aea
commit 7e8c945
Showing
2 changed files
with
13 additions
and
1 deletion.
There are no files selected for viewing
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters