OKUJI Yoshinori:
> > (1) I had to change 90 to 120 in stage1.S. This is the number of
> > sectors which stage2 is allowed to occupy. I think it's a shame that
> > there is no automatic check whether stage2 hasn't turned out larger.
> > Could this be added to the TODO list, please?
>
> There IS already the check. Just run "make check".
Oh good. Thanks. I'll try it ...
> > (3) Fetching compressed files didn't work. I haven't yet investigated.
>
> Please tell us what you did in details.
It's working now. I have succeeded in netbooting the Linux kernel,
both compressed and uncompressed, using both an ISA and a PCI network
card.
The problem with compressed files is that the gunzip method wants to
check the last 8 bytes of the file, where the CRC and original file
length are kept. But tftp only lets you download the whole file, and
it doesn't tell you the length of the file in advance, I think. But it
seems easy to work around. You don't really need to know the length in
advance.
> > Can anyone thing of why
> > the change from 90 to 120 (see (1) above) might have broken something
> > else?
>
> I don't think so. If you show us your patch, we will tell you more
> meaningful answers.
I still can't boot Fiasco/L4-Linux with my modified GRUB. It might be
a Fiasco problem, or it might have to do with modules ...
And I'm still using a work-around instead of currticks().
I'll make a patch available soon. I just want to get rid of all my
debug printfs ...
Edmund
