login
Header Space

 
 

Re: [PATCH 0/5] Boot protocol changes

Score:
Previous message: [thread] [date] [author]
Next message: [thread] [date] [author]
To: Rusty Russell <rusty@...>
Cc: lkml - Kernel Mailing List <linux-kernel@...>, H. Peter Anvin <hpa@...>, Vivek Goyal <vgoyal@...>, lguest <lguest@...>, Eric W. Biederman <ebiederm@...>
Date: Tuesday, October 2, 2007 - 7:46 pm

Rusty Russell wrote:

Ah, good.  I was thinking about reviving this work.  The main problem is
that sticking an ELF header at the 1 meg mark (the address of the
bzImage "payload") breaks 32-bit bootloaders which think they can just
jump to 32-bit code there.  I started a conversation with Eric at KS
about it, but we didn't reach any conclusions.

This series looks like a good start for Xen, but we still need to work
out where to stash the metadata which normally lives in ELF notes.  
Using ELF is convenient for Xen because it lets a large chunk of domain
builder code be reused; on the other hand, loading a plain bzImage is
pretty simple, so maybe it isn't such a big deal.

HPA, Eric: if we don't go the "embed ELF" path, where's a good
backwards-compatible place to stash the note data?  If we do go with
"embed ELF", how should we go about doing it?  Arrange to put the ELF
headers before the 1M mark?

    J
-
Previous message: [thread] [date] [author]
Next message: [thread] [date] [author]

Messages in current thread:
[PATCH 0/5] Boot protocol changes, Rusty Russell, (Tue Oct 2, 7:34 pm)
Re: [PATCH 0/5] Boot protocol changes, Jeremy Fitzhardinge, (Tue Oct 2, 7:46 pm)
Re: [PATCH 0/5] Boot protocol changes, H. Peter Anvin, (Tue Oct 2, 7:53 pm)
Re: [PATCH 0/5] Boot protocol changes, Jeremy Fitzhardinge, (Tue Oct 2, 7:56 pm)
Re: [PATCH 0/5] Boot protocol changes, H. Peter Anvin, (Tue Oct 2, 8:43 pm)
Re: [PATCH 0/5] Boot protocol changes, Jeremy Fitzhardinge, (Tue Oct 2, 8:58 pm)
Re: [PATCH 0/5] Boot protocol changes, H. Peter Anvin, (Tue Oct 2, 9:03 pm)
Re: [PATCH 0/5] Boot protocol changes, H. Peter Anvin, (Tue Oct 2, 8:46 pm)
Re: [PATCH 0/5] Boot protocol changes, H. Peter Anvin, (Tue Oct 2, 7:44 pm)
[PATCH 1/5] update boot spec to 2.07, Rusty Russell, (Tue Oct 2, 7:35 pm)
Re: [PATCH 1/5] update boot spec to 2.07, rae l, (Tue Oct 30, 2:38 am)
[PATCH 2/5] add WEAK() for creating weak asm labels, Rusty Russell, (Tue Oct 2, 7:35 pm)
[PATCH 3/5] i386: paravirt boot sequence, Rusty Russell, (Tue Oct 2, 7:36 pm)
[PATCH 4/5] Revert lguest magic and use hook in head.S, Rusty Russell, (Tue Oct 2, 7:39 pm)
[PATCH 5/5] lguest: loading bzImage directly, Rusty Russell, (Tue Oct 2, 7:40 pm)
Re: [PATCH 5/5] lguest: loading bzImage directly, Chris Malley, (Wed Oct 3, 5:37 am)
Re: [PATCH 5/5] lguest: loading bzImage directly, Rusty Russell, (Wed Oct 3, 8:02 pm)
Re: [PATCH 5/5] lguest: loading bzImage directly, Kjartan Maraas, (Sat Nov 24, 5:54 pm)
Re: [PATCH 5/5] lguest: loading bzImage directly, H. Peter Anvin, (Sat Nov 24, 6:14 pm)
Re: [PATCH 5/5] lguest: loading bzImage directly, Kjartan Maraas, (Sun Nov 25, 8:32 am)
Re: [PATCH 5/5] lguest: loading bzImage directly, H. Peter Anvin, (Wed Oct 3, 8:26 pm)
Re: [PATCH 5/5] lguest: loading bzImage directly, H. Peter Anvin, (Wed Oct 3, 1:12 pm)
speck-geostationary