Fixed a bug where Linux task address spaces wouldn't inherit the profile.

Fixed a bug where Linux task address spaces wouldn't inherit the profile.
Fixed a bug where get_available_addresses in the XEN AS would return MFNs, breaking plugins that depend on it, like memdump.
Fixed an issue where phys_addr in Linux profiles would return negative values.
BUG=
R=scudette@gmail.com

Committed: 4cdeec0bb7c04e0d6fa65cc66c608a89980a80c8

[Jordi, 2015-09-09 00:07:51]

https://codereview.appspot.com/264190043/diff/60001/rekall-core/rekall/plugin...
File rekall-core/rekall/plugins/addrspaces/amd64.py (right):

https://codereview.appspot.com/264190043/diff/60001/rekall-core/rekall/plugin...
rekall-core/rekall/plugins/addrspaces/amd64.py:400: p2m_top_location =
self.session.profile.get_constant_object(
I've been looking into changing this code into a hook as discussed offline, but
it's tricky.

At the moment I'm doing things like this get_constant_object the hook doesn't
have any AS other than the physical one because I'm trying to create the kernel
one. 

I'd have to change this line and many others into:

p2m_top_location_phys =
self.session.profile.phys_addr(self.session.profile.get_constant("p2m_top"))
p2m_top_location =
self.session.profile.get_constant_object(p2m_top_location_phys, "Pointer")

Currently, I do this automatically by defaulting to the PAGE_OFFSET translation
in this AS's vtop while there's no mapping.

The code would become extremely verbose. Though it may not be a bad thing to be
very explicit about everything that recreating the mapping entails since there's
a few intrinsics that are hidden by the use of self as an AS.

Thoughts?

[Jordi, 2015-09-09 05:51:33]

To not delay the release any longer, I'm happy to submit these bugfixes now and
I'll convert the code into hooks later.

[scudette, 2015-09-09 07:59:48]

Lgtm. Awesome!
Am 09.09.2015 07:51 schrieb <parki.san@gmail.com>:

> To not delay the release any longer, I'm happy to submit these bugfixes
> now and I'll convert the code into hooks later.
>
> https://codereview.appspot.com/264190043/
>

[Jordi, 2015-09-09 15:31:37]

I've added an additional small fix (that doesn't currently affect any piece of
real code) but I found while debugging the Xen AS.

[Jordi, 2015-09-09 15:31:38]

I've added an additional small fix (that doesn't currently affect any piece of
real code) but I found while debugging the Xen AS.

[scudette, 2015-09-09 15:36:30]

LGTM

[Jordi, 2015-09-09 15:38:52]

Committed patchset #5 (id:70001) manually as
4cdeec0bb7c04e0d6fa65cc66c608a89980a80c8 (presubmit successful).