Two KVM x86 backports for 5.15. Patch 2 is the primary motivation (fix for potential guest data corruption after live migration). Patch 1 is a (very) soft dependency to resolve a conflict. It's not strictly necessary (manually resolving the conflict wouldn't be difficult), but it is a fix that has been in upstream for a long time. The only reason I didn't tag it for stable from the get-go is that the bug it fixes is very theoretical. At this point, the odds of the patch causing problems are lower than the odds of me botching a manual backport. Sean Christopherson (2): KVM: x86: Bail to userspace if emulation of atomic user access faults KVM: x86: Mark target gfn of emulated atomic instruction as dirty arch/x86/kvm/x86.c | 12 +++++++++++- 1 file changed, 11 insertions(+), 1 deletion(-) base-commit: 9465fef4ae351749f7068da8c78af4ca27e61928 -- 2.44.0.478.gd926399ef9-goog
