On Wed, May 4, 2011 at 3:55 PM, Tony Lindgren [off-list ref] wrote:

* Govindraj [off-list ref] [110504 03:03]:

quoted

On Wed, May 4, 2011 at 3:25 PM, Tony Lindgren [off-list ref] wrote:

quoted

* Govindraj.R [off-list ref] [110429 05:39]:

quoted

@@ -250,7 +251,7 @@ static void __init omap_2430sdp_init(void)

? ? ? omap2430_i2c_init();

? ? ? platform_add_devices(sdp2430_devices, ARRAY_SIZE(sdp2430_devices));
- ? ? omap_serial_init();
+ ? ? omap_serial_init(NULL);
? ? ? omap2_hsmmc_init(mmc);
? ? ? omap2_usbfs_init(&sdp2430_usb_config);

...

quoted

?static inline void board_serial_init(void)
?{
- ? ? omap_serial_init();
+ ? ? omap_serial_init(NULL);
?}

This change seems like "crazy churn" and probably not needed
if it's always null. Boards using platform_data can use
omap_serial_init_port instead, right?

We can intiliaze only one uart port with omap_serial_init_port.

omap_serial_init initializes all uart port.

During any client device integration using dma mode
(for example Bluetooth) we can pass the dma config
params based on the board & client device.

But you're changing all omap_serial_init() calls to
omap_serial_init(NULL) so there are no users for that.

It should be a separate patch if it's needed.

Agree.

Will add seperate func. in serial.c file for board that need to pass these
params and retain the old func. omap_serial_init() syntax to avoid all the
additions in board-*.c files.

--
Thanks,
Govindraj.R

Tony