erasure_code: fix unaligned free error in perf apps on windows

Signed-off-by: Cornu, Marcel D <marcel.d.cornu@intel.com>

erasure_code: reformat using new code style

Signed-off-by: Marcel Cornu <marcel.d.cornu@intel.com>

Avoid using x18 register

Signed-off-by: Taiju Yamada <tyamada@bi.a.u-tokyo.ac.jp>

erasure_code: fix wrong return type

erasure_code/ppc64le/gf_vect_mul_vsx.c: In function '_gf_vect_mul_base':
erasure_code/ppc64le/gf_vect_mul_vsx.c:14:16: error: 'return' with a value, in function r

erasure_code: fix wrong return type

erasure_code/ppc64le/gf_vect_mul_vsx.c: In function '_gf_vect_mul_base':
erasure_code/ppc64le/gf_vect_mul_vsx.c:14:16: error: 'return' with a value, in function returning void [-Wreturn-mismatch]
14 | return 0;
| ^
erasure_code/ppc64le/gf_vect_mul_vsx.c:6:13: note: declared here
6 | static void _gf_vect_mul_base(int len, unsigned char *a, unsigned char *src,
| ^~~~~~~~~~~~~~~~~

Signed-off-by: Mattias Ellert <mattias.ellert@physics.uu.se>

show more ...

erasure_code: use internal gf_vect_mul_base for ppc64le encoding

gf_vect_mul_base is expected to work for all buffer sizes.
However, this function is checking for size alignment to 32 bytes,
to foll

erasure_code: use internal gf_vect_mul_base for ppc64le encoding

gf_vect_mul_base is expected to work for all buffer sizes.
However, this function is checking for size alignment to 32 bytes,
to follow the other gf_vect_mul implementations.
Therefore, another implementation for this function is included
inside ppc64le folder to be used by the encoding functions.

Signed-off-by: Pablo de Lara <pablo.de.lara.guarch@intel.com>

show more ...

erasure_code: check for size alignment on powerpc gf_vect_mul_vsx implementation

Follows the rest of the gf_vect_mul implementations for other architectures,
and checks for size alignment, stated in

erasure_code: check for size alignment on powerpc gf_vect_mul_vsx implementation

Follows the rest of the gf_vect_mul implementations for other architectures,
and checks for size alignment, stated in the documentation.

Signed-off-by: Pablo de Lara <pablo.de.lara.guarch@intel.com>

show more ...

erasure_code: check for size on gf_vect_mul_sse/avx

gf_vect_mul requires length to be multiple of 32 bytes,
so this check is added in the SSE/AVX implementations.

Signed-off-by: Pablo de Lara <pabl

erasure_code: check for size on gf_vect_mul_sse/avx

gf_vect_mul requires length to be multiple of 32 bytes,
so this check is added in the SSE/AVX implementations.

Signed-off-by: Pablo de Lara <pablo.de.lara.guarch@intel.com>

show more ...

gf_vect_mul_sve: fix error and enable unit tests for aarch64

Signed-off-by: liuqinfei <lucas.liuqinfei@huawei.com>

erasure_code: disable unit tests temporarily for aarch64/ppc64le

Some aarch64 and ppc64le implementations of gf_vect_mul do not check
for invalid sizes, so the unit test checking for negative return

erasure_code: disable unit tests temporarily for aarch64/ppc64le

Some aarch64 and ppc64le implementations of gf_vect_mul do not check
for invalid sizes, so the unit test checking for negative return value
from this function is disabled temporarily on these architectures.

Signed-off-by: Pablo de Lara <pablo.de.lara.guarch@intel.com>

show more ...

erasure_code: add missing aarch64 and powerpc interface for ec_init_tables

ec_init_tables is now a multi-implementation function,
so it requires a dispatcher for all architectures.

Signed-off-by: P

erasure_code: add missing aarch64 and powerpc interface for ec_init_tables

ec_init_tables is now a multi-implementation function,
so it requires a dispatcher for all architectures.

Signed-off-by: Pablo de Lara <pablo.de.lara.guarch@intel.com>

show more ...

erasure_code: various fixes for static code analysis issues

Signed-off-by: Tomasz Kantecki <tomasz.kantecki@intel.com>

erasure_code: [test] fix memory leak

Signed-off-by: Pablo de Lara <pablo.de.lara.guarch@intel.com>

erasure_code: check for size in gf_Xvect_mad_avx512_gfni

Length of data was not checked in implementation with AVX512+GFNI,
at the start of the gf_Xvect_mad_avx512_gfni functions, resulting
in buffe

erasure_code: check for size in gf_Xvect_mad_avx512_gfni

Length of data was not checked in implementation with AVX512+GFNI,
at the start of the gf_Xvect_mad_avx512_gfni functions, resulting
in buffer overflow if length was less than 64 bytes.

Signed-off-by: Pablo de Lara <pablo.de.lara.guarch@intel.com>

show more ...

erasure_code: fix modules using incorrect unsigned jump

Signed-off-by: Marcel Cornu <marcel.d.cornu@intel.com>

erasure_code: add AVX2 5vect mad with GFNI implementation

Signed-off-by: Marcel Cornu <marcel.d.cornu@intel.com>

erasure_code: add AVX2 4vect mad with GFNI implementation

Signed-off-by: Marcel Cornu <marcel.d.cornu@intel.com>

erasure_code: add AVX2 3vect mad with GFNI implementation

Signed-off-by: Marcel Cornu <marcel.d.cornu@intel.com>

erasure_code: optimize AVX2-GFNI single vector mad implementation

Signed-off-by: Marcel Cornu <marcel.d.cornu@intel.com>

erasure_code: add AVX2 2vect mad with GFNI implementation

Signed-off-by: Marcel Cornu <marcel.d.cornu@intel.com>

erasure_code: optimize AVX2 GFNI 2 vector dot product

Signed-off-by: Marcel Cornu <marcel.d.cornu@intel.com>

erasure_code: optimize AVX2 GFNI single vector dot product

Signed-off-by: Marcel Cornu <marcel.d.cornu@intel.com>

erasure_code: add 2 vector AVX2 dot product with GFNI implementation

Signed-off-by: Marcel Cornu <marcel.d.cornu@intel.com>

erasure_code: add 3 vector AVX2 dot product with GFNI implementation

Signed-off-by: Marcel Cornu <marcel.d.cornu@intel.com>

lib: reduce verbosity by default in tests

Signed-off-by: Pablo de Lara <pablo.de.lara.guarch@intel.com>

erasure_code: add initial AVX2 mad with GFNI implementation

Signed-off-by: Marcel Cornu <marcel.d.cornu@intel.com>