diff mbox series

hw/intc/arm_gicv3: ICC_HPPIR* return SPURIOUS if int group is disabled

Message ID 20240328153333.2522667-1-peter.maydell@linaro.org (mailing list archive)
State New, archived
Headers show
Series hw/intc/arm_gicv3: ICC_HPPIR* return SPURIOUS if int group is disabled | expand

Commit Message

Peter Maydell March 28, 2024, 3:33 p.m. UTC
If the group of the highest priority pending interrupt is disabled
via ICC_IGRPEN*, the ICC_HPPIR* registers should return
INTID_SPURIOUS, not the interrupt ID.  (See the GIC architecture
specification pseudocode functions ICC_HPPIR1_EL1[] and
HighestPriorityPendingInterrupt().)

Make HPPIR reads honour the group disable, the way we already do
when determining whether to preempt in icc_hppi_can_preempt().

Cc: qemu-stable@nongnu.org
Signed-off-by: Peter Maydell <peter.maydell@linaro.org>
---
Pre-existing bug which I happened to notice while working
on review of the FEAT_NMI patches. I don't suppose real world
code disables interrupt groups which it's actually using, which
is why nobody's noticed it. Still, it's a safe bugfix so might
as well go to stable too.
---
 hw/intc/arm_gicv3_cpuif.c | 4 ++--
 1 file changed, 2 insertions(+), 2 deletions(-)

Comments

Richard Henderson March 28, 2024, 6:38 p.m. UTC | #1
On 3/28/24 05:33, Peter Maydell wrote:
> If the group of the highest priority pending interrupt is disabled
> via ICC_IGRPEN*, the ICC_HPPIR* registers should return
> INTID_SPURIOUS, not the interrupt ID.  (See the GIC architecture
> specification pseudocode functions ICC_HPPIR1_EL1[] and
> HighestPriorityPendingInterrupt().)
> 
> Make HPPIR reads honour the group disable, the way we already do
> when determining whether to preempt in icc_hppi_can_preempt().
> 
> Cc:qemu-stable@nongnu.org
> Signed-off-by: Peter Maydell<peter.maydell@linaro.org>
> ---
> Pre-existing bug which I happened to notice while working
> on review of the FEAT_NMI patches. I don't suppose real world
> code disables interrupt groups which it's actually using, which
> is why nobody's noticed it. Still, it's a safe bugfix so might
> as well go to stable too.
> ---
>   hw/intc/arm_gicv3_cpuif.c | 4 ++--
>   1 file changed, 2 insertions(+), 2 deletions(-)

Reviewed-by: Richard Henderson <richard.henderson@linaro.org>

r~
diff mbox series

Patch

diff --git a/hw/intc/arm_gicv3_cpuif.c b/hw/intc/arm_gicv3_cpuif.c
index e1a60d8c15b..67d8fd07b7f 100644
--- a/hw/intc/arm_gicv3_cpuif.c
+++ b/hw/intc/arm_gicv3_cpuif.c
@@ -1067,7 +1067,7 @@  static uint64_t icc_hppir0_value(GICv3CPUState *cs, CPUARMState *env)
      */
     bool irq_is_secure;
 
-    if (cs->hppi.prio == 0xff) {
+    if (icc_no_enabled_hppi(cs)) {
         return INTID_SPURIOUS;
     }
 
@@ -1104,7 +1104,7 @@  static uint64_t icc_hppir1_value(GICv3CPUState *cs, CPUARMState *env)
      */
     bool irq_is_secure;
 
-    if (cs->hppi.prio == 0xff) {
+    if (icc_no_enabled_hppi(cs)) {
         return INTID_SPURIOUS;
     }