[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: [Xen-devel] [PATCH 0/4] mini-os: initial ARM support

Hi Oleksandr,

The "proposed" branch currently contain's only Karim Raslan's initial
ARM work. I have a series of patches on top of this (on the next
branch), but I was planning to get this initial support in first.

Possibly my "Fixed initialisation code" commit should be merged with
Karim's code into a single patch. On the other hand, it might be
clearer to keep them separate (especially for other people who have
built on Karim's original work, e.g. I see on github that Jonathan
Daugherty has a fork from there).

I have applied your commit "Fixed initialisation code" on top of "proposed" branch -> Mini-OS boots without crashes
and works as expected. It is very important commit!)

I have only one question. Why do you use those addresses for GIC for now:

ÂÂÂ gic.gicd_base = (char *)0x2c001000ULL;
ÂÂÂ gic.gicc_base = (char *)0x2c002000ULL;

With this I have a hypervisor crash(

I let the Mini-OS to read them from cpu regs. And in my case they are:
- 0x48211000
- 0x48212000

> Also, could you, please, explain me some generic things about Mini-OS:
> 1. Is HAVE_LIBC a necessary config? I mean if I build Mini-OS without proper
> LIBC support will the Mini-OS to operate fully
> (are there any restrictions?)

You only need HAVE_LIBC if you want to use your own libc. Without
HAVE_LIBC, minios provides its own versions of some basic functions
(memcmp, sprintf, etc). Mirage works without it, for example.

OK. It is clear.
Thank you.


Oleksandr Tyshchenko | Embedded Dev
Xen-devel mailing list



Lists.xenproject.org is hosted with RackSpace, monitoring our
servers 24x7x365 and backed by RackSpace's Fanatical Support®.