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

Re: [PATCH 4/4] If anaconda is not run from loader or liveinst, fail.

> We'll have to investigate that further I guess. It's my
> understanding that Anaconda and all the python code live in the
> stage2 ramdisk? We currently only rebuild the kernel initrd, so
> rebuilding stage2 also could be problematic.

That's correct - all the python stuff is in stage2.

> Would it be possible to either allow both  /sbin/loader, and
> /tmp/updates/anaconda?
> Or to walk further (maybe only one level more?) up the process tree
> looking for /sbin/loader?
> I imagine there maybe other valid uses for /tmp/updates/anaconda to
> be a wrapper.

Yes I could probably walk process parents looking for something that's
the loader, but that's starting to get pretty complicated.  This looks
like the most controversial patch in the set, so maybe I'll just skip
it in favor of pushing all the others.

> We're looking into adding support for RHEL6 soon - We generally
> don't start working on your new releases till the betas are
> available. It's my understanding that RHEL6 is still in 'alpha'?
> Given this though I'll suggest we start looking at it earlier rather
> than later.
> Do you have any specifics in mind for why you think it might already
> be broken for RHEL6?

Ah, I think I was misunderstanding your setup a little.  You're not in
any way starting anaconda from a command line.  You're using the updates
mechanism to start another program that's coincidentally called
anaconda, that in turn starts the real anaconda.  I think this should
work, then.

- Chris

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