On Thu, Oct 24, 2019 at 12:23:16AM +0200, Lorenzo Bianconi wrote:
> mt76 dma layer is supposed to unmap skb data buffers while keep txwi
> mapped on hw dma ring. At the moment mt76 wrongly unmap txwi or does
> not unmap data fragments in even positions for non-linear skbs. This
> issue may result in hw hangs with A-MSDU if the system relies on IOMMU
> or SWIOTLB. Fix this behaviour properly unmapping data fragments on
> non-linear skbs.
If we have to keep txwi mapped, before unmap fragments, when then
txwi is unmaped ?
Stanislaw
> Fixes: 17f1de56df05 ("mt76: add common code shared between multiple chipsets")
> Signed-off-by: Lorenzo Bianconi <lorenzo@xxxxxxxxxx>
> ---
> drivers/net/wireless/mediatek/mt76/dma.c | 10 +++++++---
> 1 file changed, 7 insertions(+), 3 deletions(-)
>
> diff --git a/drivers/net/wireless/mediatek/mt76/dma.c b/drivers/net/wireless/mediatek/mt76/dma.c
> index c747eb24581c..8c27956875e7 100644
> --- a/drivers/net/wireless/mediatek/mt76/dma.c
> +++ b/drivers/net/wireless/mediatek/mt76/dma.c
> @@ -93,11 +93,14 @@ static void
> mt76_dma_tx_cleanup_idx(struct mt76_dev *dev, struct mt76_queue *q, int idx,
> struct mt76_queue_entry *prev_e)
> {
> - struct mt76_queue_entry *e = &q->entry[idx];
> __le32 __ctrl = READ_ONCE(q->desc[idx].ctrl);
> + struct mt76_queue_entry *e = &q->entry[idx];
> u32 ctrl = le32_to_cpu(__ctrl);
> + bool mcu = e->skb && !e->txwi;
> + bool first = e->skb == DMA_DUMMY_DATA || e->txwi == DMA_DUMMY_DATA ||
> + (e->skb && !skb_is_nonlinear(e->skb));
>
> - if (!e->txwi || !e->skb) {
> + if (!first || mcu) {
> __le32 addr = READ_ONCE(q->desc[idx].buf0);
> u32 len = FIELD_GET(MT_DMA_CTL_SD_LEN0, ctrl);
>
> @@ -105,7 +108,8 @@ mt76_dma_tx_cleanup_idx(struct mt76_dev *dev, struct mt76_queue *q, int idx,
> DMA_TO_DEVICE);
> }
>
> - if (!(ctrl & MT_DMA_CTL_LAST_SEC0)) {
> + if (!(ctrl & MT_DMA_CTL_LAST_SEC0) ||
> + e->txwi == DMA_DUMMY_DATA) {
> __le32 addr = READ_ONCE(q->desc[idx].buf1);
> u32 len = FIELD_GET(MT_DMA_CTL_SD_LEN1, ctrl);
>
> --
> 2.21.0
>
