In order to avoid a race condition for user events when changing
cpu affinity reset the active flag only when EOI-ing the event.
This is working fine as all user events are lateeoi events. Note that
lateeoi_ack_mask_dynirq() is not modified as there is no explicit call
to xen_irq_lateeoi() expected later.
Reported-by: Julien Grall <[email protected]>
Fixes: b6622798bc50b62 ("xen/events: avoid handling the same event on two cpus at the same time")
Tested-by: Julien Grall <[email protected]>
Signed-off-by: Juergen Gross <[email protected]>
---
drivers/xen/events/events_base.c | 11 ++++++++++-
1 file changed, 10 insertions(+), 1 deletion(-)
diff --git a/drivers/xen/events/events_base.c b/drivers/xen/events/events_base.c
index 7bbfd58958bc..d7e361fb0548 100644
--- a/drivers/xen/events/events_base.c
+++ b/drivers/xen/events/events_base.c
@@ -642,6 +642,9 @@ static void xen_irq_lateeoi_locked(struct irq_info *info, bool spurious)
}
info->eoi_time = 0;
+
+ /* is_active hasn't been reset yet, do it now. */
+ smp_store_release(&info->is_active, 0);
do_unmask(info, EVT_MASK_REASON_EOI_PENDING);
}
@@ -811,6 +814,7 @@ static void xen_evtchn_close(evtchn_port_t port)
BUG();
}
+/* Not called for lateeoi events. */
static void event_handler_exit(struct irq_info *info)
{
smp_store_release(&info->is_active, 0);
@@ -1883,7 +1887,12 @@ static void lateeoi_ack_dynirq(struct irq_data *data)
if (VALID_EVTCHN(evtchn)) {
do_mask(info, EVT_MASK_REASON_EOI_PENDING);
- event_handler_exit(info);
+ /*
+ * Don't call event_handler_exit().
+ * Need to keep is_active non-zero in order to ignore re-raised
+ * events after cpu affinity changes while a lateeoi is pending.
+ */
+ clear_evtchn(evtchn);
}
}
--
2.26.2
On 6/23/21 9:09 AM, Juergen Gross wrote:
> In order to avoid a race condition for user events when changing
> cpu affinity reset the active flag only when EOI-ing the event.
>
> This is working fine as all user events are lateeoi events. Note that
> lateeoi_ack_mask_dynirq() is not modified as there is no explicit call
> to xen_irq_lateeoi() expected later.
>
> Reported-by: Julien Grall <[email protected]>
> Fixes: b6622798bc50b62 ("xen/events: avoid handling the same event on two cpus at the same time")
> Tested-by: Julien Grall <[email protected]>
> Signed-off-by: Juergen Gross <[email protected]>
Reviewed-by: Boris Ostrovsky <[email protected]>
On 05.07.21 11:00, Ross Lagerwall wrote:
> On 2021-06-23 14:09, Juergen Gross wrote:
>> In order to avoid a race condition for user events when changing
>> cpu affinity reset the active flag only when EOI-ing the event.
>>
>> This is working fine as all user events are lateeoi events. Note that
>> lateeoi_ack_mask_dynirq() is not modified as there is no explicit call
>> to xen_irq_lateeoi() expected later.
>>
>> Reported-by: Julien Grall <[email protected]>
>> Fixes: b6622798bc50b62 ("xen/events: avoid handling the same event on two cpus at the same time")
>> Tested-by: Julien Grall <[email protected]>
>> Signed-off-by: Juergen Gross <[email protected]>
>> ---
>
> Hi Juergen,
>
> Are you planning on backporting this fix to releases <= 5.10?
It is on my todo list. :-)
Juergen