From: Henry Yen <[email protected]>
It's observed that host driver might occasionally receive
interrupts from unexpected Rx ring, whose Rx NAPI hasn't been
prepared yet. Under such situation, __napi_poll crash issue
would occur, so we add a sanity check to prevent it.
Without this patch, we might encounter kernel crash issue
especially in WED-on & RRO-on software path.
Signed-off-by: Henry Yen <[email protected]>
Signed-off-by: Shayne Chen <[email protected]>
---
drivers/net/wireless/mediatek/mt76/mt7996/mmio.c | 2 +-
1 file changed, 1 insertion(+), 1 deletion(-)
diff --git a/drivers/net/wireless/mediatek/mt76/mt7996/mmio.c b/drivers/net/wireless/mediatek/mt76/mt7996/mmio.c
index 304e5fd14803..6de5b7de4ac8 100644
--- a/drivers/net/wireless/mediatek/mt76/mt7996/mmio.c
+++ b/drivers/net/wireless/mediatek/mt76/mt7996/mmio.c
@@ -560,7 +560,7 @@ static void mt7996_irq_tasklet(struct tasklet_struct *t)
napi_schedule(&dev->mt76.tx_napi);
for (i = 0; i < __MT_RXQ_MAX; i++) {
- if ((intr & MT_INT_RX(i)))
+ if ((intr & MT_INT_RX(i)) && dev->mt76.napi[i].poll)
napi_schedule(&dev->mt76.napi[i]);
}
--
2.39.2
On 20.03.24 12:09, Shayne Chen wrote:
> From: Henry Yen <[email protected]>
>
> It's observed that host driver might occasionally receive
> interrupts from unexpected Rx ring, whose Rx NAPI hasn't been
> prepared yet. Under such situation, __napi_poll crash issue
> would occur, so we add a sanity check to prevent it.
>
> Without this patch, we might encounter kernel crash issue
> especially in WED-on & RRO-on software path.
Does this issue occur even with Lorenzo's fix "wifi: mt76: mt7996: fix
uninitialized variable in mt7996_irq_tasklet()"?
- Felix
On Wed, 2024-04-03 at 12:57 +0200, Felix Fietkau wrote:
>
> External email : Please do not click links or open attachments until
> you have verified the sender or the content.
> On 20.03.24 12:09, Shayne Chen wrote:
> > From: Henry Yen <[email protected]>
> >
> > It's observed that host driver might occasionally receive
> > interrupts from unexpected Rx ring, whose Rx NAPI hasn't been
> > prepared yet. Under such situation, __napi_poll crash issue
> > would occur, so we add a sanity check to prevent it.
> >
> > Without this patch, we might encounter kernel crash issue
> > especially in WED-on & RRO-on software path.
>
> Does this issue occur even with Lorenzo's fix "wifi: mt76: mt7996:
> fix
> uninitialized variable in mt7996_irq_tasklet()"?
>
They achieve the same purpose, please drop this patch.
Thanks,
Shayne
> - Felix