[<prev] [next>] [thread-next>] [day] [month] [year] [list]
Message-ID: <20260107182615.488194-1-Simon.Richter@hogyros.de>
Date: Thu, 8 Jan 2026 03:26:01 +0900
From: Simon Richter <Simon.Richter@...yros.de>
To: linux-kernel@...r.kernel.org,
linux-next@...r.kernel.org
Cc: Simon Richter <Simon.Richter@...yros.de>,
Alexander Usyskin <alexander.usyskin@...el.com>,
Greg Kroah-Hartman <gregkh@...uxfoundation.org>
Subject: [PATCH v2 0/4] Intel MEI Kconfig cleanups
Hi,
this enables GPU firmware updates on Intel dGPUs plugged into non-x86
machines. The most important patches are the first and last.
The second patch switches the ME interfaces for GPUs to use !=n tests, so
it is possible for them to be builtins when the GPU drivers themselves are
modules (as is likely). MEI is built-in by default on x86_64 (and Atom).
This might need some discussion what is desirable here:
- this is a strong dependency from an "early" directory (misc/mei) to a
later one (gpu/drm), causing backtracking
- does it make sense for MEI to be built-in by default, or should it
default to being a module?
- is the strong dependency desired because if the code is compiled as
built-in, it will not be used until the module is loaded?
There is no strong functional dependency here, no symbol references, as
these devices are enumerated over an auxiliary bus.
Also, enable building without the GPU drivers for compile testing.
No stable backport.
Simon
Simon Richter (4):
misc/mei: Allow building Intel ME interface on non-x86
misc/mei: Decouple ME interfaces from GPU drivers
misc/mei: Allow building standalone for compile testing
misc/mei: gsc_proxy: add dependency on Xe driver
drivers/misc/mei/Kconfig | 6 +++---
drivers/misc/mei/gsc_proxy/Kconfig | 2 +-
drivers/misc/mei/hdcp/Kconfig | 2 +-
drivers/misc/mei/pxp/Kconfig | 2 +-
4 files changed, 6 insertions(+), 6 deletions(-)
base-commit: f8f9c1f4d0c7a64600e2ca312dec824a0bc2f1da
--
2.47.3
Powered by blists - more mailing lists