Signed-off-by: Anton Bondarenko <anton.bondarenko@arm.com>

Approved-by: Jens Elofsson <jens.elofsson@arm.com>

Anitha Raj authored Oct 03, 2024 and Felix Johnny Thomasmathibalan committed Oct 03, 2024

Signed-off-by: Anitha Raj <anitha.raj@arm.com>

Reviewed-by: Felix Johnny Thomasmathibalan <felixjohnny.thomasmathibalan@arm.com>
Reviewed-by: Anton Bondarenko <anton.bondarenko@arm.com>
Reviewed-by: Gian Marco Iodice <gianmarco.iodice@arm.com>
Approved-by: Felix Johnny Thomasmathibalan <felixjohnny.thomasmathibalan@arm.com>

Signed-off-by: Anton Bondarenko <anton.bondarenko@arm.com>

Approved-by: Jakub Sujak <jakub.sujak@arm.com>

Signed-off-by: Anton Bondarenko <anton.bondarenko@arm.com>

Approved-by: Jakub Sujak <jakub.sujak@arm.com>

Previously, multiple build directories were not being ignored.

Signed-off-by: Jakub Sujak <jakub.sujak@arm.com>

Approved-by: Viet-Hoa Do <viet-hoa.do@arm.com>

Jakub Sujak authored Sep 27, 2024 and Felix Johnny Thomasmathibalan committed Sep 27, 2024

Signed-off-by: Jakub Sujak <jakub.sujak@arm.com>

Approved-by: Felix Johnny Thomasmathibalan <felixjohnny.thomasmathibalan@arm.com>

Signed-off-by: Felix Thomasmathibalan <felixjohnny.thomasmathibalan@arm.com>

Approved-by: Jakub Sujak <jakub.sujak@arm.com>

Anitha Raj authored Sep 27, 2024 and Felix Johnny Thomasmathibalan committed Sep 27, 2024

Signed-off-by: Anitha Raj <anitha.raj@arm.com>

Approved-by: Felix Johnny Thomasmathibalan <felixjohnny.thomasmathibalan@arm.com>

Signed-off-by: Felix Thomasmathibalan <felixjohnny.thomasmathibalan@arm.com>

Approved-by: Jakub Sujak <jakub.sujak@arm.com>

Signed-off-by: Felix Thomasmathibalan <felixjohnny.thomasmathibalan@arm.com>

Approved-by: Jakub Sujak <jakub.sujak@arm.com>

Signed-off-by: Anton Bondarenko <anton.bondarenko@arm.com>

Approved-by: Jakub Sujak <jakub.sujak@arm.com>

Some toolchains might be missing CPU capability in <bits/hwcap.h> so
alternative definition needed to make sure library would compile
properly in a such case.

Signed-off-by: Anton Bondarenko <anton.bondarenko@arm.com>

Closes #2

Approved-by: Felix Johnny Thomasmathibalan <felixjohnny.thomasmathibalan@arm.com>

Signed-off-by: Felix Thomasmathibalan <felixjohnny.thomasmathibalan@arm.com>

Signed-off-by: Anitha Raj <anitha.raj@arm.com>

Reviewed-by: Anitha Raj <anitha.raj@arm.com>
Reviewed-by: Felix Johnny Thomasmathibalan <felixjohnny.thomasmathibalan@arm.com>
Reviewed-by: Anton Bondarenko <anton.bondarenko@arm.com>
Reviewed-by: Jakub Sujak <jakub.sujak@arm.com>
Approved-by: Anton Bondarenko <anton.bondarenko@arm.com>

Additional checks for all used and some future CPU features added
to unit tests to avoid crash if instruction not supported by running CPU.

Signed-off-by: Anton Bondarenko <anton.bondarenko@arm.com>

Approved-by: Jakub Sujak <jakub.sujak@arm.com>

Test panels where M=1 and N=1.

Signed-off-by: Jakub Sujak <jakub.sujak@arm.com>

Approved-by: Anton Bondarenko <anton.bondarenko@arm.com>

This reverts commit cddf991a

The existing FP16 GEMM micro-kernel (matmul_clamp_f16_f16_f16p16x1biasf16_6x16x8_neon_mla) has a dedicated path for M=1 (a "GEMV" operation).

Signed-off-by: Jakub Sujak <jakub.sujak@arm.com>

Approved-by: Felix Johnny Thomasmathibalan <felixjohnny.thomasmathibalan@arm.com>

Jakub Sujak authored Sep 25, 2024 and Felix Johnny Thomasmathibalan committed Sep 25, 2024

This reverts commit f4f59599

The existing FP32 GEMM micro-kernel (matmul_clamp_f32_f32_f32p8x1biasf32_6x8x4_neon_mla) has a dedicated path for M=1 (a "GEMV" operation).

Signed-off-by: Jakub Sujak <jakub.sujak@arm.com>

Approved-by: Felix Johnny Thomasmathibalan <felixjohnny.thomasmathibalan@arm.com>

Signed-off-by: Viet-Hoa Do <viet-hoa.do@arm.com>

Approved-by: Felix Johnny Thomasmathibalan <felixjohnny.thomasmathibalan@arm.com>

Some text editors like neovim have trouble parsing EditorConfig when comments are on the same line as the setting.

Signed-off-by: Jakub Sujak <jakub.sujak@arm.com>

Approved-by: Jens Elofsson <jens.elofsson@arm.com>

Michael Kozlov authored Sep 19, 2024 and Viet-Hoa Do committed Sep 19, 2024

Signed-off-by: Michael Kozlov <michael.kozlov@arm.com>

Approved-by: Viet-Hoa Do <viet-hoa.do@arm.com>

Anitha Raj authored Sep 18, 2024 and Jakub Sujak committed Sep 18, 2024

* Fix the function name kai_get_rhs_packed_stride_rhs_pack_kxn_qsi4c32p_qsu4c32s1s0

Signed-off-by: Anitha Raj <anitha.raj@arm.com>

Approved-by: Jakub Sujak <jakub.sujak@arm.com>

* Linux kernel from 6.9.4 to 6.9.12
* Toolchain from 13.2.rel1 to 13.3.rel1

Signed-off-by: Anton Bondarenko <anton.bondarenko@arm.com>

Approved-by: Jakub Sujak <jakub.sujak@arm.com>

Signed-off-by: Jens Elofsson <jens.elofsson@arm.com>

Approved-by: Viet-Hoa Do <viet-hoa.do@arm.com>
Approved-by: Anton Bondarenko <anton.bondarenko@arm.com>

* We already have a CI job with MemTag sanitizer which relies on
  hardware support to provide fast memory sanitization.
  It also works with assembly kernels.

Signed-off-by: Viet-Hoa Do <viet-hoa.do@arm.com>

Approved-by: Jakub Sujak <jakub.sujak@arm.com>

Anitha Raj authored Sep 16, 2024 and Jakub Sujak committed Sep 16, 2024

Signed-off-by: Anitha Raj <anitha.raj@arm.com>

Approved-by: Jakub Sujak <jakub.sujak@arm.com>

Gian Marco Iodice authored Sep 12, 2024 and Viet-Hoa Do committed Sep 12, 2024

- The LHS matrix is Quantized (Q) Asymmetric (A) Signed 8-bit (I8) with per-row
(DX) quantization parameters
- The RHS matrix is quantized (Q) Symmetric (S) Signed 4-bit (I4) with
per-block quantization
- The destination is F32
- Implement micro-kernels to perform the matrix multiplication
- Implement a micro-kernel to pack the RHS matrix

Signed-off-by: Gian Marco Iodice <gianmarco.iodice@arm.com>

Signed-off-by: Anitha Raj <anitha.raj@arm.com>

Signed-off-by: Viet-Hoa Do <viet-hoa.do@arm.com>

Reviewed-by: Viet-Hoa Do <viet-hoa.do@arm.com>
Reviewed-by: Anitha Raj <anitha.raj@arm.com>
Reviewed-by: Felix Johnny Thomasmathibalan <felixjohnny.thomasmathibalan@arm.com>
Reviewed-by: Jakub Sujak <jakub.sujak@arm.com>
Reviewed-by: Max Ren <maxren@meta.com>
Approved-by: Viet-Hoa Do <viet-hoa.do@arm.com>

* This is useful for building and testing with `+memtag` option.

Signed-off-by: Viet-Hoa Do <viet-hoa.do@arm.com>

Approved-by: Jakub Sujak <jakub.sujak@arm.com>

Signed-off-by: Jens Elofsson <jens.elofsson@arm.com>

Reviewed-by: Jakub Sujak <jakub.sujak@arm.com>
Approved-by: Jakub Sujak <jakub.sujak@arm.com>

* Introduce cache for big downloads to improve image build time
* Utilize ccache for Linux kernel compilation to improve image build time

Signed-off-by: Anton Bondarenko <anton.bondarenko@arm.com>

Approved-by: Jens Elofsson <jens.elofsson@arm.com>
Approved-by: Jakub Sujak <jakub.sujak@arm.com>

Nikhil Gupta authored Sep 04, 2024 and Anton Bondarenko committed Sep 04, 2024

GTest targets clash with parent gtest targets when KLEIDIAI_BUILD_TESTS
is turned off

Signed-off-by: Nikhil Gupta <nikhil.gupta2@arm.com>

Approved-by: Anton Bondarenko <anton.bondarenko@arm.com>

Signed-off-by: Anton Bondarenko <anton.bondarenko@arm.com>

Approved-by: Jens Elofsson <jens.elofsson@arm.com>

Signed-off-by: Anton Bondarenko <anton.bondarenko@arm.com>

Approved-by: Felix Johnny Thomasmathibalan <felixjohnny.thomasmathibalan@arm.com>

Felix Johnny Thomasmathibalan authored Aug 30, 2024 and Jakub Sujak committed Aug 30, 2024

Signed-off-by: Felix Thomasmathibalan <felixjohnny.thomasmathibalan@arm.com>

Approved-by: Jakub Sujak <jakub.sujak@arm.com>

Anton Bondarenko authored Aug 30, 2024 and Jakub Sujak committed Aug 30, 2024

Code coverage tools used in the project has a strong limitation
for location of source code used to compile with coverage. This
source code must be in exactly same absolute path when running
coverage processing tool. It's not easy to guarantee when several
GitLab runners are used in project CI as these have different paths
for source code location. Improve coverity pipeline where source code,
execution and report generation could be in different folders.

Signed-off-by: Anton Bondarenko <anton.bondarenko@arm.com>

Approved-by: Jakub Sujak <jakub.sujak@arm.com>

Signed-off-by: Felix Thomasmathibalan <felixjohnny.thomasmathibalan@arm.com>

Approved-by: Anton Bondarenko <anton.bondarenko@arm.com>

Signed-off-by: Anton Bondarenko <anton.bondarenko@arm.com>

Approved-by: Felix Johnny Thomasmathibalan <felixjohnny.thomasmathibalan@arm.com>

Michael Kozlov authored Aug 28, 2024 and Felix Johnny Thomasmathibalan committed Aug 28, 2024

Signed-off-by: Michael Kozlov <michael.kozlov@arm.com>

Approved-by: Anton Bondarenko <anton.bondarenko@arm.com>

Examples is a vital part of KleidiAI delivery and need to be checked
in a same way as core library and unit tests.

Signed-off-by: Anton Bondarenko <anton.bondarenko@arm.com>

Approved-by: Jakub Sujak <jakub.sujak@arm.com>

Jakub Sujak authored Aug 19, 2024 and Felix Johnny Thomasmathibalan committed Aug 19, 2024

* Add test coverage for the matmul_clamp_f32_qsi8d32p1x8_qsi4c32p4x8_1x4x32_neon_dotprod micro-kernel.

Signed-off-by: Jakub Sujak <jakub.sujak@arm.com>

Approved-by: Felix Johnny Thomasmathibalan <felixjohnny.thomasmathibalan@arm.com>

Jakub Sujak authored Aug 19, 2024 and Felix Johnny Thomasmathibalan committed Aug 19, 2024

Signed-off-by: Jakub Sujak <jakub.sujak@arm.com>

Approved-by: Felix Johnny Thomasmathibalan <felixjohnny.thomasmathibalan@arm.com>