On Tue, Dec 11, 2012 at 04:24:37PM -0600, Scott Wood wrote:
On 12/10/2012 07:16:50 PM, Tom Rini wrote:
On Mon, Dec 10, 2012 at 07:09:55PM -0600, Scott Wood wrote:
On 12/10/2012 09:24:32 AM, Pantelis Antoniou wrote:
- sprintf(cmd_buf, "nand %s %p %llx %llx",
op == DFU_OP_READ ? "read" : "write",buf, start, count);- debug("%s: %s 0x%p\n", __func__, cmd_buf, cmd_buf);
- ret = run_command(cmd_buf, 0);
Why not use the C interface to NAND?
- /* find out how much actual bytes have been written */
- /* the difference is the amount of skip we must add from now on
*/
- actual = nand_extent_skip_bad(&nand_info[dev], start, count);
...especially since you already need to interact with it here?
I've been talking with Pantelis about this as well and in short, this series adds NAND support ala MMC (which is to say, (ab)using the command interface). I think he was thinking we need a bit more generic help to avoid having to duplicate the code the command interface also uses (state, sanity checking), iirc.
Some elaboration on what exactly he's relying on from the command line interface would be nice.
My gmane-fu is failing me, but the actual 4/7 parts of http://search.gmane.org/?query=dfu+mmc+%224%2F7%22&author=&group=gma... show the discussion. In short, for MMC writing it's more complex because we have a number of layouts we need to deal with (raw, FAT, other), but yes, we should in the end migrate things to using the API rather than abusing the command infrastructure.