I agree. Such workarounds take boot time and create situations when
first action of kernel is to undo last action of booter.
Sometimes it can't be avoided but if we can we should try. Also it adds
some flexibility since last action of booter is
performed in nice unix-like environment and kernel s
This is small enough that it could make sense to use inline asm instead,
so we can avoid a separate file. What do you think?
I agree with this. Actually in similar place in my xnu booting I used 4
lines of inline assembly. The only reason I kept it in .S is that I was
in one of my lazy states
On Wed, Feb 04, 2009 at 01:08:26PM +0100, phcoder wrote:
> Hello. On the IRC discussion it was said that removing call to
> grub_dl_unload_all_modules isn't a good thing. I would like to discuss
> this on the list
> 1) Call to this functon is present only on i386-pc
> 2) I would suggest an interf
On Thu, Jan 29, 2009 at 04:57:08PM +0100, phcoder wrote:
> Index: loader/i386/bsd_helper.S
> ===
> --- loader/i386/bsd_helper.S (revision 0)
> +++ loader/i386/bsd_helper.S (revision 0)
> @@ -0,0 +1,26 @@
> +#include
> +
> +
Hello. On the IRC discussion it was said that removing call to
grub_dl_unload_all_modules isn't a good thing. I would like to discuss
this on the list
1) Call to this functon is present only on i386-pc
2) I would suggest an interface for preboot hooks if modules needs to do
something before boo
