[<prev] [next>] [<thread-prev] [day] [month] [year] [list]
Message-ID: <1f0c8d56-a601-4f52-9d10-76c90f1cc521@intel.com>
Date: Tue, 19 Aug 2025 13:41:50 -0700
From: Sohil Mehta <sohil.mehta@...el.com>
To: Dave Hansen <dave.hansen@...el.com>, Dave Hansen
<dave.hansen@...ux.intel.com>, <x86@...nel.org>
CC: Borislav Petkov <bp@...en8.de>, Thomas Gleixner <tglx@...utronix.de>,
"Ingo Molnar" <mingo@...hat.com>, "H . Peter Anvin" <hpa@...or.com>, Peter
Zijlstra <peterz@...radead.org>, Josh Poimboeuf <jpoimboe@...nel.org>, Pawan
Gupta <pawan.kumar.gupta@...ux.intel.com>, Nikolay Borisov
<nik.borisov@...e.com>, Alex Murray <alex.murray@...onical.com>, Andrew
Cooper <andrew.cooper3@...rix.com>, <linux-kernel@...r.kernel.org>
Subject: Re: [PATCH 2/2] scripts/x86: Add a script to update minimum Intel
ucode revisions
On 8/19/2025 10:14 AM, Dave Hansen wrote:
> But with your version, I can't do that. I have to create the header
> file, then point the script to it. I also have to consume microcode
> files from a single directory.
>
Consuming microcode from multiple directories is easy to fix, just need
to add "nargs='+'" in the argument parser. Will change in the next
revision.
> So, it's less flexible and harder to use with those command-line
> arguments. It's obviously not the end of the world, but why add code to
> the script for this? Why is it needed?
Though the script is less flexible, my goal was to make it easier to
use. When used within the kernel tree, it automatically selects and
updates the header file, since that is always the same.
The usage was supposed to be simple and hard to mess up:
update.py -u /path/to/microcode/files
The header argument is only needed if someone uses the script outside
the tree, which I assumed to be rare. But I understand this limits other
usages.
To make it flexible again, I can get rid of the parameters and make it
print to standard output. Users would then redirect it as needed.
update.py /path/to/files1 /path/to/files2 > /path/to/ucode-def.h
Powered by blists - more mailing lists