Some cards don't update the PISR properly when all SISR bits for Tx interrupts are being cleared and as a result we get interrupt storm. Since we handle all tx queues all together (so we don't really use the SISR bits to do per-queue interrupt handling), we can manualy update PISR by doing a write-to-clear on its Tx interrupt bits. Signed-off-by: Nick Kossifidis <mickflemm@xxxxxxxxx> --- drivers/net/wireless/ath/ath5k/dma.c | 11 ++++++++++- 1 file changed, 10 insertions(+), 1 deletion(-) diff --git a/drivers/net/wireless/ath/ath5k/dma.c b/drivers/net/wireless/ath/ath5k/dma.c index ce86f15..84c01c5 100644 --- a/drivers/net/wireless/ath/ath5k/dma.c +++ b/drivers/net/wireless/ath/ath5k/dma.c @@ -616,7 +616,16 @@ ath5k_hw_get_isr(struct ath5k_hw *ah, enum ath5k_int *interrupt_mask) * SISRs will also clear PISR so no need to worry here. */ - pisr_clear = pisr & ~AR5K_ISR_BITS_FROM_SISRS; + /* XXX: There seems to be an issue on some cards + * with tx interrupt flags not being updated + * on PISR despite that all Tx interrupt bits + * are cleared on SISRs. Since we handle all + * Tx queues all together it shouldn't be an + * issue if we clear Tx interrupt flags also + * on PISR to avoid that. + */ + pisr_clear = (pisr & ~AR5K_ISR_BITS_FROM_SISRS) | + (pisr & AR5K_INT_TX_ALL); /* * Write to clear them... -- 1.8.4.4 -- To unsubscribe from this list: send the line "unsubscribe linux-wireless" in the body of a message to majordomo@xxxxxxxxxxxxxxx More majordomo info at http://vger.kernel.org/majordomo-info.html
