On Wed, 7 Feb 2001 [EMAIL PROTECTED] wrote:
> >> M Init C display variables. Required before printk()
>
> Isn't this done first in hardware startup so that we can print debug
> stuff when configuring cache and the like?
no, it is not. There are a few variables used to manage a log_buf style
buffer which need to be set up.
> Can you give examples? I only see MTRRs used to config L2 cache. This
> seems like a board and/or application dependent thing.
no, the only example I have is MTRR. But I don't want to rule it out.
> My concern about one size fits all is that it doesn't. And when you
> are forced to go outside the system it gets hard to support and hard to
> share results. A much more general system certainly can include a one
> size fits most function that could be called by the board/application
> dependent initializer function.
Tyson, this is not an attempt at one size fits all. Eric Biederman asked
for a task list, I didn't know why at first, but he convinced me. I'm
trying to get a feel for what the basic flow is -- yes, it will vary.
All those steps labeled 'O(ptional)' will be in some mainboard.c and not
others. But I'm currently wondering if we shouldn't generate mainboard.c
dynamically. I worry about lots of mainboard.c floating around, some
having a recent fix, others not; there are one or two common pieces. If
you're using a pentium, for example, you really do want the cache on early
to avoid those 30-second boots.
To give you an example of how to get into trouble. At some point I half
way thought all the crt0.S would vary, mainboard to mainboard. You can
still find bits of crt0.S in various places in the tree that I have not
cleaned up yet. But there were enough fixes that were common to all crt0.S
that I realized that approach would fail, and hence moved to the current
config tool which generates crt0.S. Based on what you've told me so far,
we need main() to vary a lot. Based on what I've seen so far, I don't want
lots and lots of main()s floating around. I'm trying to find a way to make
this work out ...
thanks
ron
