From 2b04c2ebd0f7ac9fe1bf48cba7bc5bbb6dd88b75 Mon Sep 17 00:00:00 2001 From: =?UTF-8?q?Micha=C5=82=20G=C3=B3rny?= Date: Mon, 26 May 2025 17:08:18 +0200 Subject: [PATCH] profiles/desc: Add avx_vnni to CPU_FLAGS_X86 MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit As needed by `sci-ml/ollama::guru`. Signed-off-by: Michał Górny Part-of: https://github.com/gentoo/gentoo/pull/42265 Signed-off-by: Michał Górny --- profiles/desc/cpu_flags_x86.desc | 1 + 1 file changed, 1 insertion(+) diff --git a/profiles/desc/cpu_flags_x86.desc b/profiles/desc/cpu_flags_x86.desc index 9fc7a47752f48..3de7350d8c48c 100644 --- a/profiles/desc/cpu_flags_x86.desc +++ b/profiles/desc/cpu_flags_x86.desc @@ -9,6 +9,7 @@ 3dnowext - Use the Enhanced 3DNow! instruction set aes - Enable support for Intel's AES instruction set (AES-NI) avx - Adds support for Advanced Vector Extensions instructions +avx_vnni - Use AVX (VEX-encoded) versions of the Vector Neural Network Instructions avx2 - Adds support for Advanced Vector Extensions 2 instructions avx512_4fmaps - Use AVX-512 Fused Multiply-Accumulate Packed Single Precision instruction set avx512_4vnniw - Use AVX-512 Vector Neural Network Instructions Word Variable Precision