On Fri, Oct 17, 2008 at 11:08:25PM +0900, Atsushi Nemoto wrote:
> This is the driver for the Toshiba TX4939 SoC ATA controller.
> This controller has standard ATA taskfile registers and DMA
> command/status registers, but the register layout is swapped on big
> endian. There are some other endian issue and some special registers
> which requires many custom dma_ops/tp_ops routines and build_dmatable.
> Signed-off-by: Atsushi Nemoto <firstname.lastname@example.org>
> This patch is against current linux-mips tree.
> Changes since v3:
> * more consistent symbol naming
> * handle only DMA mode in set_dma_mode
> * rename tx4939ide_read_and_clear_dma_status to tx4939ide_clear_dma_status
> * use standard ide_read_sff_dma_status in LE mode
> * remove CS5530 workaround from tx4939ide_build_dmatable
> * use ide_host_alloc/ide_host_register instead of ide_host_alloc
> * fold tx4939ide_insw_swap into tx4939ide_input_data_swap
> * more informative printk
> * whitespace cleanups and spelling fixes
> drivers/ide/Kconfig | 6 +
> drivers/ide/mips/Makefile | 1 +
> drivers/ide/mips/tx4939ide.c | 755
Btw, I don't think architecture specific subdirectories in subsystems are
generally usefull. Just as in this example this IDE controller happens
only to be in use on a particular MIPS-based SOC but there is nothing
really architecture specific in most such devices.