Hi, I'm not sure if this is the correct list to post this; sory if it isn't. I'm writting a network driver for a DMA-enabled Fast Ethernet card, for linux 2.4. If I'm not mistaken the kernel, whenever it wants to transmit a packet, calls the driver's "hard_start_xmit()" function, passing it a pointer to the "skb" holding the packet. My question is: is there any guarantee that the packet's data at "skb->data" are stored in physically-continuous memory locations? If it is so, then I can pass the "skb->data" pointer directly to the card and let it do all the work (using of-course "pci_map_single()" or similar means, in order to synchronize it, and get its physical address). If not then I have to copy the packet to a known-to-be-continuous area (e.g. a kmalloc-allocated one); this caries a certain performance penalty. Existing drivers I've studied (namely: "eepro100.c") seem to make this assumption, but I have also heard that "zero-copy" networking was added to the kernel at some point. Zero-copy indicates that data come directly for user-space and, hence, they might be non-continuous. Thanks in advance /npat -- What strange patches have you been smoking? -- Linus Torvalds - : send the line "unsubscribe linux-net" in the body of a message to majordomo@vger.kernel.org More majordomo info at http://vger.kernel.org/majordomo-info.html
