On Sun, Oct 26, 2025 at 10:25:02PM +0000, Huang, Kai wrote:
On Sat, 2025-10-25 at 11:59 -0400, Sasha Levin wrote:
From: Kai Huang kai.huang@intel.com
[ Upstream commit 10df8607bf1a22249d21859f56eeb61e9a033313 ]
[...]
LLM Generated explanations, may be completely bogus:
YES
Why this fixes a real bug
- TDX can leave dirty cachelines for private memory with different encryption attributes (C-bit aliases). If kexec interrupts a CPU during a SEAMCALL, its dirty private cachelines can later be flushed in the wrong order and silently corrupt the new kernel’s memory. Marking the CPU’s cache state as “incoherent” before executing SEAMCALL ensures kexec will WBINVD on that CPU and avoid corruption.
Hi,
I don't think we should backport this for 6.17 stable. Kexec/kdump and TDX are mutually exclusive in Kconfig in 6.17, therefore it's not possible for TDX to impact kexec/kdump.
This patch is part of the series which enables kexec/kdump together with TDX in Kconfig (which landed in 6.18) and should not be backported alone.
I'll drop it, thanks for the review!