> reset via netdev timeout [1]. The patch addresses that by setting the Tx
> ring buffer pointer to where the driver thinks the chip should continue.
You shouldn't use virt_to_bus().
On x86 it doesn't matter(?) but on other hw it does.
See Documentation/DMA-mapping.txt.
The tx_ring array is allocated with pci_alloc_consistent and the addresses
you are supposed to send to the hardware are the rx_ring_dma and
tx_ring_dma. tx_ring and rx_ring are for the driver.
The initialisation of tx_ring_dma assumes the area allocted by
pci_alloc_consistent is contiguous and that it is ok to use any address
within that area (within any alignment constrinats of the hw). If that
code is correct you could copy it and do:
int entry = np->dirty_tx % TX_RING_SIZE;
writel(np->tx_ring_dma + entry * sizeof(struct tx_desc),
dev->base_addr + TxRingPtr);
(Or something prettier ...)
/Urban
-
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@vger.kernel.org
More majordomo info at http://vger.kernel.org/majordomo-info.html
Please read the FAQ at http://www.tux.org/lkml/