bug in task server startup code

[Bas Wijnen]

I spend most of today and part of yesterday tracing a bug regarding the task
server.  When wortel_putchar didn't work, I tried to force a crash, in which I
was unsuccessful.  I didn't quite find the bug, but found a pointer which
should help.  I'm writing now because I might not work on it for a few weeks
and if someone else does, I want them to know where to look :-)

The startup code (wortel/startup.c) is copied and executed as expected.
Entry_point is called from there.  In my case this was 0x080480a0.  However,
the code that was there is what should be at 0x080490a0, so it seems the code
is placed 0x1000 too low in memory, or something.  AFAIK all memory is mapped
at its real address, so it shouldn't be an address translation problem.

This is the bug which makes physmem pagefault at 0xa when used as task,
instead of asking wortel for memory, or whatever physmem would normally do.

That's it for now, if anyone has a hint on what the problem might be, please
let me know.

Thanks,
Bas

-- 
I encourage sending me encrypted e-mail.
Please send the central message of e-mails as plain text in the message body,
   not as HTML and definitely not as MS Word.
Please do not use the MS Word format for attachments either.
for more information, see http://129.125.47.90/e-mail.html

pgpHP__irWfoL.pgp
Description: PGP signature