Skip to content

kvm-5.6-1

tagged this 30 Jan 17:57
PPC: Bugfixes

x86:
* Support for mapping DAX areas with large nested page table entries.
* Cleanups and bugfixes here too.  A particularly important one is
a fix for FPU load when the thread has TIF_NEED_FPU_LOAD.  There is
also a race condition which could be used in guest userspace to exploit
the guest kernel, for which the embargo expired today.
* Fast path for IPI delivery vmexits, shaving about 200 clock cycles
from IPI latency.
* Protect against "Spectre-v1/L1TF" (bring data in the cache via
speculative out of bound accesses, use L1TF on the sibling hyperthread
to read it), which unfortunately is an even bigger whack-a-mole game
than SpectreV1.

Sean continues his mission to rewrite KVM.  In addition to a sizable
number of x86 patches, this time he contributed a pretty large refactoring
of vCPU creation that affects all architectures but should not have any
visible effect.

s390 will come next week together with some more x86 patches.
Assets 2
Loading