Ping :)
On 9/5/26 3:41 AM, Daniil Tatianin wrote:
This fixes a few problems that eventually lead to the guest receiving a
bogus #GP due to a stale interrupt still being asserted. Either patch
fixes the problem, but the first one fixes the actual root cause, and
the second one simply aligns interrupt handling with KVM, which made
the bug invisible by simply dropping the IRQ targeting a bogus vector.
Daniil Tatianin (2):
hw/intc/apic: don't drop irq line re-evaluation when LINT0 is masked
target/i386: drop attempts to deliver an interrupt without a vector
hw/intc/apic.c | 3 ++-
target/i386/tcg/system/seg_helper.c | 3 +++
2 files changed, 5 insertions(+), 1 deletion(-)
