In gicv3_redist_lpi_pending() we update cs->hpplpi to indicate the
new highest priority pending LPI after changing the requested LPI
pending bit. However the overall highest priority pending interrupt
information won't be updated unless we call gicv3_redist_update().
We do that from the callsite in gicv3-redist_process_lpi(), but not
from the callsite in icc_activate_irq(). The effect is that when the
guest acknowledges an LPI by reading ICC_IAR1_EL1, we mark it as not
pending in the data structure but still leave it in cs->hppi so will
offer it to the guest again.
The effect is that if we are using an emulated GICv3 and ITS and
using devices which use LPIs (ie PCI devices) then Linux will
complain "irq 54: nobody cared" and then hang (probably because the
stale bogus interrupt info meant we never tried to deliver some other
real interrupt).
Add the missing gicv3_redist_update() call.
Signed-off-by: Peter Maydell <peter.mayd...@linaro.org>
---
Marked for-6.2 because this is a bug fix to the ITS support
which is new in this release. At least for me this is necessary
to boot Debian on the virt board, since the ITS is default-enabled.
The failure seemed to be somewhat intermittent; I haven't bothered
to try to work out why.
hw/intc/arm_gicv3_cpuif.c | 1 +
1 file changed, 1 insertion(+)
diff --git a/hw/intc/arm_gicv3_cpuif.c b/hw/intc/arm_gicv3_cpuif.c
index 3fe5de8ad7d..2d9f2ad2b6c 100644
--- a/hw/intc/arm_gicv3_cpuif.c
+++ b/hw/intc/arm_gicv3_cpuif.c
@@ -906,6 +906,7 @@ static void icc_activate_irq(GICv3CPUState *cs, int irq)
gicv3_update(cs->gic, irq, 1);
} else {
gicv3_redist_lpi_pending(cs, irq, 0);
+ gicv3_redist_update(cs);
}
}
--
2.25.1
