irq_work's use of the DEC SPR is racy with guest<->host switch and guest
entry which flips the DEC interrupt to guest, which could lose a host
work interrupt.

This patch closes one race, and attempts to comment several others.

(XXX: should think a bit harder about this)

Signed-off-by: Nicholas Piggin <npig...@gmail.com>
---
 arch/powerpc/include/asm/paca.h |  1 +
 arch/powerpc/kvm/book3s_hv.c    | 15 ++++++++++++++-
 2 files changed, 15 insertions(+), 1 deletion(-)

diff --git a/arch/powerpc/include/asm/paca.h b/arch/powerpc/include/asm/paca.h
index ec18ac818e3a..23c12048fbc9 100644
--- a/arch/powerpc/include/asm/paca.h
+++ b/arch/powerpc/include/asm/paca.h
@@ -174,6 +174,7 @@ struct paca_struct {
        u8 irq_happened;                /* irq happened while soft-disabled */
        u8 irq_work_pending;            /* IRQ_WORK interrupt while 
soft-disable */
 #ifdef CONFIG_KVM_BOOK3S_HV_POSSIBLE
+       /* Could have irq_work_using_hdec here, but what about nested HV entry 
modifying DEC? Could have a pointer to the hv struct time limit */
        u8 pmcregs_in_use;              /* pseries puts this in lppaca */
 #endif
        u64 sprg_vdso;                  /* Saved user-visible sprg */
diff --git a/arch/powerpc/kvm/book3s_hv.c b/arch/powerpc/kvm/book3s_hv.c
index d98958b78830..1997cf347d3e 100644
--- a/arch/powerpc/kvm/book3s_hv.c
+++ b/arch/powerpc/kvm/book3s_hv.c
@@ -3689,6 +3689,18 @@ static int kvmhv_p9_guest_entry(struct kvm_vcpu *vcpu, 
u64 time_limit,
        if (!(vcpu->arch.ctrl & 1))
                mtspr(SPRN_CTRLT, mfspr(SPRN_CTRLF) & ~1);
 
+       /*
+        * XXX: must always deal with irq_work_raise via NMI vs setting DEC.
+        * The problem occurs right as we switch into guest mode if a NMI
+        * hits and sets pending work and sets DEC, then that will apply to
+        * the guest and not bring us back to the host.
+        *
+        * irq_work_raise could check a flag (or possibly LPCR[HDICE] for
+        * example) and set HDEC to 1? That wouldn't solve the nested hv
+        * case which needs to abort the hcall or zero the time limit.
+        *
+        * Another day's problem.
+        */
        mtspr(SPRN_DEC, vcpu->arch.dec_expires - tb);
 
        if (kvmhv_on_pseries()) {
@@ -3822,7 +3834,8 @@ static int kvmhv_p9_guest_entry(struct kvm_vcpu *vcpu, 
u64 time_limit,
        vc->entry_exit_map = 0x101;
        vc->in_guest = 0;
 
-       mtspr(SPRN_DEC, local_paca->kvm_hstate.dec_expires - tb);
+       set_dec_or_work(local_paca->kvm_hstate.dec_expires - tb);
+
        mtspr(SPRN_SPRG_VDSO_WRITE, local_paca->sprg_vdso);
 
        kvmhv_load_host_pmu();
-- 
2.23.0

Reply via email to