I have several client devices that use the same AR9160 card that are
receiving PCI FATAL interrupts several times a day.
The pattern is the same in all cases I've been able to capture logs for:
kernel: ath: phy1: transmitting packet, skb: ffff88013b5cc0c0
kernel: ath: phy1: qnum: 2, txq depth: 0
kernel: ath: phy1: link[2] (ffff88003786a1b0)=3786bc18 (ffff88003786bc18)
kernel: ath: phy1: Enable TXE on queue: 2
kernel: ath: phy1: disable IER
kernel: ath: phy1: tx queue 0 (37860dd0), link ffff880037860dd0
kernel: ath: phy1: tx queue 2 (3786bc18), link ffff88003786bc18
kernel: ath: phy1: enable IER
kernel: ath: phy1: AR_IMR 0x918404b0 IER 0x1
kernel: ath: phy1: received PCI FATAL interrupt
kernel: ath: phy1: disable IER
kernel: ath: phy1: enable IER
kernel: ath: phy1: AR_IMR 0x918404b0 IER 0x1
kernel: ath: phy1: Stopping ANI
... snip (lots of "disable IER" msgs) ...
kernel: Failed to stop TX DMA, queues=0x004!
kernel: ath: phy1:
kernel: ath: phy1: disable IER
kernel: TX complete: skb: ffff88013b5cc0c0 <--- same skb from right
before the PCI FATAL interrupt
kernel: ath: phy1: disable IER
kernel: ath: phy1: disable IER
kernel: ath: phy1: Disable MIB counters
... snip (lots of "disable IER" msgs) ...
kernel: DMA failed to stop in 10 ms AR_CR=0x00000024
AR_DIAG_SW=0x42000020 DMADBG_7=0x000062c0
... snip (lots of "disable IER" msgs) ...
kernel: Could not stop RX, we could be confusing the DMA engine when
we start RX up
... snip (lots of "disable IER" msgs) ...
kernel: ath: phy1: Reset to 2462 MHz, HT40: 0 fastcc: 0
Is there any way to get the reason for the fatal interrupt?