- Oct 29, 2024
-
-
Marcel Cornu authored
Signed-off-by:
Marcel Cornu <marcel.d.cornu@intel.com>
-
Marcel Cornu authored
Signed-off-by:
Marcel Cornu <marcel.d.cornu@intel.com>
-
Pablo de Lara authored
Signed-off-by:
Pablo de Lara <pablo.de.lara.guarch@intel.com>
-
Pablo de Lara authored
Signed-off-by:
Pablo de Lara <pablo.de.lara.guarch@intel.com>
-
Tomasz Kantecki authored
-
Pablo de Lara authored
Job/burst API can be used instead. Signed-off-by:
Pablo de Lara <pablo.de.lara.guarch@intel.com>
-
- Oct 25, 2024
-
-
Tomasz Kantecki authored
Co-authored-by:
Adrian Stanciu <adrian.stanciu@intel.com> Signed-off-by:
Tomasz Kantecki <tomasz.kantecki@intel.com>
-
Tomasz Kantecki authored
Signed-off-by:
Tomasz Kantecki <tomasz.kantecki@intel.com>
-
Tomasz Kantecki authored
Signed-off-by:
Tomasz Kantecki <tomasz.kantecki@intel.com>
-
Pablo de Lara authored
Algorithms using the counter mode were using the term "CNTR" in the cipher mode enums. "CTR" is more widely used, so more enums are added replacing CNTR with CTR, leaving the previous ones untouched, to avoid API breakages. Signed-off-by:
Pablo de Lara <pablo.de.lara.guarch@intel.com>
-
Pablo de Lara authored
Except for IMB_CIPHER_SM4_CNTR cipher mode enum, which remains as it is, to avoid API breakage on frameworks like DPDK that are already using this enum (IMB_CIPHER_SM4_CTR is also defined, though). Signed-off-by:
Pablo de Lara <pablo.de.lara.guarch@intel.com>
-
Tomasz Kantecki authored
avx2: [aes-gcm] unify temporary SIMD register layout in INITIAL_BLOCKS_PARTIAL() and INITIAL_BLOCKS() Signed-off-by:
Tomasz Kantecki <tomasz.kantecki@intel.com>
-
Tomasz Kantecki authored
All SIMD moves to and from the stack frame are aligned. Signed-off-by:
Tomasz Kantecki <tomasz.kantecki@intel.com>
-
Tomasz Kantecki authored
Signed-off-by:
Tomasz Kantecki <tomasz.kantecki@intel.com>
-
Tomasz Kantecki authored
Signed-off-by:
Tomasz Kantecki <tomasz.kantecki@intel.com>
-
Tomasz Kantecki authored
Signed-off-by:
Tomasz Kantecki <tomasz.kantecki@intel.com>
-
- Oct 24, 2024
-
-
Marcel Cornu authored
Required as some algorithms do not support all default buffer sizes Signed-off-by:
Marcel Cornu <marcel.d.cornu@intel.com>
-
Marcel Cornu authored
Signed-off-by:
Marcel Cornu <marcel.d.cornu@intel.com>
-
Tomasz Kantecki authored
-
Tomasz Kantecki authored
Move aes-gcm context definition to a separate include file.
-
Marcel Cornu authored
Signed-off-by:
Marcel Cornu <marcel.d.cornu@intel.com>
-
Marcel Cornu authored
Signed-off-by:
Marcel Cornu <marcel.d.cornu@intel.com>
-
Marcel Cornu authored
Signed-off-by:
Marcel Cornu <marcel.d.cornu@intel.com>
-
Marcel Cornu authored
Signed-off-by:
Marcel Cornu <marcel.d.cornu@intel.com>
-
Marcel Cornu authored
Signed-off-by:
Marcel Cornu <marcel.d.cornu@intel.com>
-
Marcel Cornu authored
Signed-off-by:
Marcel Cornu <marcel.d.cornu@intel.com>
-
Kamila Lipinska authored
-
- Oct 10, 2024
-
-
Tomasz Kantecki authored
-
vkarpenk authored
-
- Oct 09, 2024
-
-
Marcel Cornu authored
Signed-off-by:
Marcel Cornu <marcel.d.cornu@intel.com>
-
Tomasz Kantecki authored
In some cases, it could have included fragment of sensitive data.
-
Tomasz Kantecki authored
-
- Oct 08, 2024
-
-
Tomasz Kantecki authored
This is to further decrease chance of false positive hits.
-
- Oct 04, 2024
-
-
Tomasz Kantecki authored
- remove stack frame clear operation from AES-GCM AVX2 type 2 - add small comment about register usage in AES-GCM AVX512 type 2
-
Tomasz Kantecki authored
-
Pablo de Lara authored
Signed-off-by:
Pablo de Lara <pablo.de.lara.guarch@intel.com>
-
- Oct 03, 2024
-
-
Marcel Cornu authored
Signed-off-by:
Marcel Cornu <marcel.d.cornu@intel.com>
-
Pablo de Lara authored
Signed-off-by:
Pablo de Lara <pablo.de.lara.guarch@intel.com>
-
Tomasz Kantecki authored
- Remove AVX architecture type from the library and applications - Products like Sandy Bridge or Ivy Bridge will use SSE implementations now - Remove AVX AES-GCM symbols. They were already pointing at SSE implementations.
-
Tomasz Kantecki authored
- AVX type1 switches to use SSE implementation for these algorithms - memory copy function, used for storing computed tag value, moved from AVX to AVX2 - removed AVX512 and AVX2 variants of AES-CFB-ONE functions
-