8:45 a.m.
Hello,
I recently posted a message, where I asked how to create a profile which
could be used with ArchLinux, but now I just solved this by having
installed Lubuntu 16.04 (4.4.0-31-generic 64-bit), so that I was able to
analyze my system's dumped memory using the pre-built Ubuntu 16.04 image I
found on GitHub (the image I created by myself couldn't be used by
Volatility, although I definitely followed each step precisely).
The checks I performed confirmed my suspicion that my system would be
compromised, as one can see by taking a look at the results I uploaded on
GoogleDrive:
https://drive.google.com/open?id=0B62Y5Qk_rdbWTnNYWlJRWXpsZUE
E.g.:
linux_check_afinfo
Symbol Name Member
Address
------------------------------------------ ------------------------------
------------------
udplite6_seq_afinfo next
0xffffffff81effcc8
udplite6_seq_afinfo stop
0xffffffff81eff808
udplite4_seq_afinfo next
0xffffffff81efeac8
udplite4_seq_afinfo stop
0xffffffff81efbce8
udp4_seq_afinfo start
0xffffffff81efae00
udp4_seq_afinfo stop
0xffffffff81efa3a0
linux_check_inline_kernel
Name Member Hook Type
Hook Address
------------------------------------------------ ---------------- ---------
------------------
udp4_seq_afinfo stop JMP
0x0000000000000000
[A huge number of hooks shown by linux_check_syscall]
Using the netstat plugin shows no result at all (none of the connections
shown by using the normal Linux netstat command). Neither linux_lsmod nor
linux_hidden_modules give any output as well.
I assume that my system is infected by an ACPI rootkit, which is able to
compromise both Linux and Windows systems. After having submitted the
extracted the ACPI tables' code to malwr.com, where it gets executed on a
Windows sandbox, it shows that the system gets manipulated in the following
way:
https://malwr.com/analysis/ODkxOThjOTk1MDAzNGE4M2JhOWNhNzk1ZTJjM2IyYWQ/
It might be interesting that the ACPI code of four different systems being
used by me seems to have been manipulated in the same way, since the
extracted code found on one of the other systems leads to the same result
when submitting it to malwr.com. E.g., the link above shows the result for
the analysis of ACPI code on my AMD 64-bit desktop computer (Asus-M4N68T-M
LE mainboard), while the ACPI code extracted from my Lenovo G710 notebook
leads to the same when executed on a Windows system:
https://malwr.com/analysis/MjZkOGU4Y2ZmMGM5NDQ1Njg5OTc4NTVlOTQ5NThiMmY/
I guess everyone can see that the results show how the Windows system gets
compromised for being able to monitor it and gaining remote access over it,
if you take a look at the file and registry activities (just googling some
of the file names makes that clear).
Since a Linux system running on the same machine gets compromised as well,
it would be reasonable to assume that this also takes place by the ACPI
code's execution. Taking a look at the dmesg output, which I also uploaded
on GoogleDrive, seems to confirm this assumption:
[ 0.225468] ACPI: Added _OSI(Module Device)
[ 0.225470] ACPI: Added _OSI(Processor Device)
[ 0.225471] ACPI: Added _OSI(3.0 _SCP Extensions)
[ 0.225472] ACPI: Added _OSI(Processor Aggregator Device)
[ 0.227433] ACPI: Executed 1 blocks of module-level executable AML code
[ 0.293448] ACPI: Interpreter enabled
[ 0.293458] ACPI Exception: AE_NOT_FOUND, While evaluating Sleep State
[\_S2_] (20150930/hwxface-580)
[ 0.293469] ACPI: (supports S0 S1 S3 S4 S5)
[ 0.293470] ACPI: Using IOAPIC for interrupt routing
[ 0.293491] PCI: MMCONFIG for domain 0000 [bus 00-ff] at [mem
0xe0000000-0xefffffff] (base 0xe0000000)
[ 0.294509] PCI: MMCONFIG at [mem 0xe0000000-0xefffffff] reserved in
ACPI motherboard resources
[ 0.294522] PCI: Using host bridge windows from ACPI; if necessary, use
"pci=nocrs" and report a bug
[ 0.298938] ACPI: PCI Root Bridge [PCI0] (domain 0000 [bus 00-ff])
[ 0.298943] acpi PNP0A03:00: _OSC: OS supports [ExtendedConfig ASPM
ClockPM Segments MSI]
[ 0.299051] acpi PNP0A03:00: _OSC: platform does not support
[PCIeHotplug PCIeCapability]
[ 0.299099] acpi PNP0A03:00: _OSC: not requesting control; platform does
not support [PCIeCapability]
[ 0.299101] acpi PNP0A03:00: _OSC: OS requested [PCIeHotplug PME AER
PCIeCapability]
[ 0.299103] acpi PNP0A03:00: _OSC: platform willing to grant [PME AER]
[ 0.299104] acpi PNP0A03:00: _OSC failed (AE_SUPPORT); disabling ASPM
[ 8.647712] ACPI Warning: SystemIO range
0x0000000000000600-0x000000000000063F conflicts with OpRegion
0x0000000000000600-0x00000000000006FF (\_SB_.PCI0.SBRG.ASOC.SMRG)
(20150930/utaddress-254)
The extracted and disassembled ACPI code of my AMD system can be downloaded
from here:
https://drive.google.com/open?id=0B62Y5Qk_rdbWYzhPTHhHM1RxRTg
So I would appreciate it, if anyone would have an idea on how to proceed
with a further analysis. It would be interesting, if one would be able to
see how exactly the ACPI code's execution interacts with the kernel in
order to compromise the system. And of course it would be interesting to
discover where the relevant network traffic gets forwarded to / comes in
from (for remote access), since the checks I already performed showed that
the networking structure got manipulated, so that the usage of Wireshark
etc. won't show anything.
Kind regards and thanks in advance
David