powerpc/64s: Fix may_hard_irq_enable() for PMI soft masking
authorNicholas Piggin <npiggin@gmail.com>
Sat, 3 Feb 2018 07:17:50 +0000 (17:17 +1000)
committerMichael Ellerman <mpe@ellerman.id.au>
Thu, 8 Feb 2018 12:56:10 +0000 (23:56 +1100)
commit6cc3f91bf69fc8c1719704607474f9b9df56f348
tree05bbb665ea70b6ddec598b1c3dd53aa139c1f911
parent5c11d1e52d996749897a8616860b18a084c894f0
powerpc/64s: Fix may_hard_irq_enable() for PMI soft masking

The soft IRQ masking code has to hard-disable interrupts in cases
where the exception is not cleared by the masked handler. External
interrupts used this approach for soft masking. Now recently PMU
interrupts do the same thing.

The soft IRQ masking code additionally allowed for interrupt handlers
to hard-enable interrupts after soft-disabling them. The idea is to
allow PMU interrupts through to profile interrupt handlers.

So when interrupts are being replayed when there is a pending
interrupt that requires hard-disabling, there is a test to prevent
those handlers from hard-enabling them if there is a pending external
interrupt. may_hard_irq_enable() handles this.

After f442d00480 ("powerpc/64s: Add support to mask perf interrupts
and replay them"), may_hard_irq_enable() could prematurely enable
MSR[EE] when a PMU exception exists, which would result in the
interrupt firing again while masked, and MSR[EE] being disabled again.

I haven't seen that this could cause a serious problem, but it's
more consistent to handle these soft-masked interrupts in the same
way. So introduce a define for all types of interrupts that require
MSR[EE] masking in their soft-disable handlers, and use that in
may_hard_irq_enable().

Fixes: f442d004806e ("powerpc/64s: Add support to mask perf interrupts and replay them")
Signed-off-by: Nicholas Piggin <npiggin@gmail.com>
Reviewed-by: Madhavan Srinivasan <maddy@linux.vnet.ibm.com>
Signed-off-by: Michael Ellerman <mpe@ellerman.id.au>
arch/powerpc/include/asm/hw_irq.h
arch/powerpc/kernel/exceptions-64e.S
arch/powerpc/kernel/exceptions-64s.S