kawakami-k

Sorry for inconvenience. I'll check the issue at this week end. If you can provide additional information, as nSircombe points out, could you please give me.

> hi sir, I'm trying to using my own thread_pool to replace the omp_thread_pool. During debug, I find that `get_max_threads_to_use()` in `oneDNN-2.2\src\cpu\platform.cpp` always return 1 in aarch64. Is that any...

I am aware that this Issue has been assigned to me. Please give me some more time so I can add the feature.

Hi @igorsafo, Thanks for the heads-up. It's LGTM. Are there any other RFCs or documents regarding API changes in version 3.0?

Thank you for your comment. README.md may be incorrect. The below may be correct. ``` ld4((v0.b - v3.b), ptr(x0)); ``` I'll check and fix it.

Sorry, `ld4((v0.b - v3.b), ptr(x0));` is typo. `ld4((v0.b16 - v3.b16), ptr(x0));` and `ld4((v0.b8 - v3.b8), ptr(x0));` are correct. Syntax of Xbyak_aarch64 is defined to be as close as possible to...

Although the default branch of xbyak_aarch64 is set to the fjmaster, please use the main branch. We plan to use the main branch as main stream. Please try the following...

It takes less than a second to generate the code. The overhead of code generation is negligible, as deep learning takes a order of hours to days.

Please use "main" branch. We plan to overwrite "fj_master" branch with "main" branch in the near future. You can use 'xbyak_aarch64_code_array.h:dd(uint32_t code)' to write single 32-bit data at the current...

Yes, it does. Please be careful not to overflow the code size. The initial code size can be given in the constructor. ```C++ #include "xbyak_aarch64.h" using namespace Xbyak_aarch64; class Generator...