[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]
Message-ID: <90d5a756-e534-490b-b451-7c855183ebc3@gmail.com>
Date: Tue, 17 Sep 2024 21:33:36 +0000
From: Hanabishi <i.r.e.c.c.a.k.u.n+kernel.org@...il.com>
To: Borislav Petkov <bp@...en8.de>, John <therealgraysky@...ton.me>
Cc: "linux-kernel@...r.kernel.org" <linux-kernel@...r.kernel.org>,
Thomas Gleixner <tglx@...utronix.de>, Ingo Molnar <mingo@...hat.com>,
Dave Hansen <dave.hansen@...ux.intel.com>, Unknown <x86@...nel.org>,
"H. Peter Anvin" <hpa@...or.com>
Subject: Re: [PATCH] x86: add more x86-64 micro-architecture levels
On 9/15/24 12:49, Borislav Petkov wrote:
> Patches like this one appear off and on on the mailing list and each
> time I ask what's the upside of maintaining this complexity?
Besides, there are already well-known patches exist for years. So why reinventing the wheel here?
E.g. graysky patch used by ZEN kernel:
https://github.com/zen-kernel/zen-kernel/commit/6f32b8af8ccdb56ef2856db3631eea55b79378c6
It contains way more architectures, includig ISA levels.
On 9/15/24 11:05, John wrote:
> GCC 11.1 and Clang 12.0[1] allow for the following new generic
> 64-bit levels: x86-64-v2, x86-64-v3, and x86-64-v4. This commit
> adds them as options accessible under:
> Processor type and features --->
> Processor family --->
Anyway, this whole thing is actually more complicated than simply setting '-march'.
Vector instructions are known to be problematic for the kernel, so they are disabled by KBUILD_CFLAGS.
If you want to go with higher ISA levels than the kernel expects, an additional patch like this is required:
https://github.com/zen-kernel/zen-kernel/commit/addc601c58e035e28153deeb6d441b91f1a50247
Powered by blists - more mailing lists