On Friday 08 June 2018 11:56 AM, Tony Lindgren wrote:

* Geert Uytterhoeven [off-list ref] [180607 13:40]:

quoted

The divisions (and multiplications) can be avoided by changing the loops
to use increments of mux_bytes instead of 1.
While at it, remove the unneeded casts when assigning void pointers.

This saves +100 bytes of kernel size on arm32/arm64.

Signed-off-by: Geert Uytterhoeven <geert+renesas@glider.be>
---
Compile-tested only.

Makes sense to me:

Acked-by: Tony Lindgren <tony@atomide.com>

Keerthy can you please test?

Tested on AM437x-gp-evm for Deep Sleep0.

Tested-by: Keerthy <j-keerthy@ti.com>
Reviewed-by: Keerthy <j-keerthy@ti.com>

Thanks Geert!

Regards,

Tony