Patchwork [1/3] apic: Resolve potential endless loop around apic_update_irq

login
register
mail settings
Submitter Jan Kiszka
Date July 9, 2012, 2:42 p.m.
Message ID <d837433e9f810c55bb6e25c659bca05df5991a03.1341844944.git.jan.kiszka@siemens.com>
Download mbox | patch
Permalink /patch/169865/
State New
Headers show

Comments

Jan Kiszka - July 9, 2012, 2:42 p.m.
Commit d96e173769 refactored the reinjection of pending PIC interrupts.
However, it missed the potential loop of apic_update_irq ->
apic_deliver_pic_intr -> apic_local_deliver -> apic_set_irq ->
apic_update_irq that /could/ occur if LINT0 is injected as APIC_DM_FIXED
and that vector is currently blocked via TPR.

Resolve this by reinjecting only where it matters: inside
apic_get_interrupt. This function may clear a vector while a
PIC-originated reason still exists.

Signed-off-by: Jan Kiszka <jan.kiszka@siemens.com>
---
 hw/apic.c |   10 +++++++---
 1 files changed, 7 insertions(+), 3 deletions(-)

Patch

diff --git a/hw/apic.c b/hw/apic.c
index 60552df..e65a35f 100644
--- a/hw/apic.c
+++ b/hw/apic.c
@@ -363,9 +363,6 @@  static void apic_update_irq(APICCommonState *s)
     }
     if (apic_irq_pending(s) > 0) {
         cpu_interrupt(s->cpu_env, CPU_INTERRUPT_HARD);
-    } else if (apic_accept_pic_intr(&s->busdev.qdev) &&
-               pic_get_output(isa_pic)) {
-        apic_deliver_pic_intr(&s->busdev.qdev, 1);
     }
 }
 
@@ -560,7 +557,14 @@  int apic_get_interrupt(DeviceState *d)
     reset_bit(s->irr, intno);
     set_bit(s->isr, intno);
     apic_sync_vapic(s, SYNC_TO_VAPIC);
+
+    /* re-inject if there is still a pending PIC interrupt */
+    if (apic_accept_pic_intr(&s->busdev.qdev) && pic_get_output(isa_pic)) {
+        apic_deliver_pic_intr(&s->busdev.qdev, 1);
+    }
+
     apic_update_irq(s);
+
     return intno;
 }