]> git.ipfire.org Git - people/arne_f/kernel.git/commit
KVM: PPC: Book3S HV: Make idle_kvm_start_guest() return 0 if it went to guest
authorMichael Ellerman <mpe@ellerman.id.au>
Fri, 15 Oct 2021 12:02:08 +0000 (23:02 +1100)
committerGreg Kroah-Hartman <gregkh@linuxfoundation.org>
Wed, 27 Oct 2021 07:56:52 +0000 (09:56 +0200)
commit197ec50b2df12dbfb17929eda643b16117b6f0ca
treea92f4585133152d7e5afb56710f3a9582d8f5529
parentfbd724c49bead048ae9fc1a5b7bff2fb3e54f855
KVM: PPC: Book3S HV: Make idle_kvm_start_guest() return 0 if it went to guest

commit cdeb5d7d890e14f3b70e8087e745c4a6a7d9f337 upstream.

We call idle_kvm_start_guest() from power7_offline() if the thread has
been requested to enter KVM. We pass it the SRR1 value that was returned
from power7_idle_insn() which tells us what sort of wakeup we're
processing.

Depending on the SRR1 value we pass in, the KVM code might enter the
guest, or it might return to us to do some host action if the wakeup
requires it.

If idle_kvm_start_guest() is able to handle the wakeup, and enter the
guest it is supposed to indicate that by returning a zero SRR1 value to
us.

That was the behaviour prior to commit 10d91611f426 ("powerpc/64s:
Reimplement book3s idle code in C"), however in that commit the
handling of SRR1 was reworked, and the zeroing behaviour was lost.

Returning from idle_kvm_start_guest() without zeroing the SRR1 value can
confuse the host offline code, causing the guest to crash and other
weirdness.

Fixes: 10d91611f426 ("powerpc/64s: Reimplement book3s idle code in C")
Cc: stable@vger.kernel.org # v5.2+
Signed-off-by: Michael Ellerman <mpe@ellerman.id.au>
Link: https://lore.kernel.org/r/20211015133929.832061-2-mpe@ellerman.id.au
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
arch/powerpc/kvm/book3s_hv_rmhandlers.S