Hi Masahiro,
On 15 December 2014 at 21:38, Masahiro Yamada yamada.m@jp.panasonic.com wrote:
Hi Simon,
On Mon, 15 Dec 2014 07:19:52 -0700 Simon Glass sjg@chromium.org wrote:
This came up in a discussion on the mailing list here:
https://patchwork.ozlabs.org/patch/384613/
My concerns at the time were:
- it doesn't need to be written in assembler
Assembler can guarantee that stack is not used.
So can C if written correctly, at least on PowerPC, ARM and a few other archs.
- it doesn't need to be ARM-specific
OK.
This patch provides a possible alternative. It works by allowing any serial driver to export one init function and provide a putc() function. These can be used to output debug data before the real serial driver is available.
This implementation does not depend on driver model,
OK.
and it is possible for it to operate without a stack on some architectures (e.g. PowerPC, ARM).
It depends on the optimization option -O2 / -Os.
OK, fair enough.
In 18/19, you add "static inline" to calc_divisor().
"static inline" does not force-inline the code, but still leaves it to the compiler decision. (Marek Vasut also mentioned this in commit 4425e62856)
Sure, I understand.
For me this implementation works on ARM and PowerPC, and can be applied for any UART in U-Boot. We don't need to write assembler in each UART for each architecture we support. That's my point.
I'm not suggesting we drop the assembler version, just that we have another option. Of course if a particular UART driver wants to include assembler it can.
Regards, Simon