diff options
author | Nicholas Piggin <npiggin@gmail.com> | 2022-11-27 22:49:26 +1000 |
---|---|---|
committer | Michael Ellerman <mpe@ellerman.id.au> | 2022-11-30 20:40:17 +1100 |
commit | 6c645b01e536757a9e1a9f72c13767f9b3f8559f (patch) | |
tree | 30bc5fbdfc2e6ea5f820015292163f248b948b50 /arch/powerpc/include/asm/kvm_ppc.h | |
parent | a96b20758b23be7e9f693218908228d6100c3c26 (diff) |
KVM: PPC: Book3E: Fix CONFIG_TRACE_IRQFLAGS support
32-bit does not trace_irqs_off() to match the trace_irqs_on() call in
kvmppc_fix_ee_before_entry(). This can lead to irqs being enabled twice
in the trace, and the irqs-off region between guest exit and the host
enabling local irqs again is not properly traced.
64-bit code does call this, but from asm code where volatiles are live
and so incorrectly get clobbered.
Move the irq reconcile into C to fix both problems.
Signed-off-by: Nicholas Piggin <npiggin@gmail.com>
Signed-off-by: Michael Ellerman <mpe@ellerman.id.au>
Link: https://lore.kernel.org/r/20221127124942.1665522-2-npiggin@gmail.com
Diffstat (limited to 'arch/powerpc/include/asm/kvm_ppc.h')
-rw-r--r-- | arch/powerpc/include/asm/kvm_ppc.h | 12 |
1 files changed, 12 insertions, 0 deletions
diff --git a/arch/powerpc/include/asm/kvm_ppc.h b/arch/powerpc/include/asm/kvm_ppc.h index bfacf12784dd..eae9619b6190 100644 --- a/arch/powerpc/include/asm/kvm_ppc.h +++ b/arch/powerpc/include/asm/kvm_ppc.h @@ -1014,6 +1014,18 @@ static inline void kvmppc_fix_ee_before_entry(void) #endif } +static inline void kvmppc_fix_ee_after_exit(void) +{ +#ifdef CONFIG_PPC64 + /* Only need to enable IRQs by hard enabling them after this */ + local_paca->irq_happened = PACA_IRQ_HARD_DIS; + irq_soft_mask_set(IRQS_ALL_DISABLED); +#endif + + trace_hardirqs_off(); +} + + static inline ulong kvmppc_get_ea_indexed(struct kvm_vcpu *vcpu, int ra, int rb) { ulong ea; |