An open source memory analysis tool built on top of Volatility. It is meant as a proving ground for interesting new techniques to be made available to the community. These techniques are an attempt to speed up the investigation process through data reduction and codifying some expert knowledge.
##Table of Contents
- ~30 Volatility plugins combined into ~20 DAMM plugins (e.g., pslist, psxview and other elements are combined into a 'processes' plugin)
- Can run multiple plugins in one invocation
- The option to store plugin results in SQLite databases for preservation or for "cached" analysis
- A filtering/type system that allows easily filtering on attributes like pids to see all information related to some process and exact or partial matching for strings, etc.
- The ability to show the differences between two databases of results for the same or similar machines and manipulate from the cmdline how the differencing operates
- The ability to warn on certain types of suspicious behavior
- Output for terminal, tsv or grepable
NOTE: Most DAMM output looks better piped through 'less -S' (upper 'S') as in:
#python damm.py <some DAMM functionality> | less -S (for default output format)
python damm.py -h
usage: damm.py [-h] [-d DIR] [-p PLUGIN [PLUGIN ...]] [-f FILE] [-k KDBG]
[--db DB] [--profile PROFILE] [--debug] [--info] [--tsv]
[--grepable] [--filter FILTER] [--filtertype FILTERTYPE]
[--diff BASELINE] [-u FIELD [FIELD ...]] [--warnings] [-q]
DAMM v1.0 Beta
optional arguments:
-h, --help show this help message and exit
-d DIR Path to additional plugin directory
-p PLUGIN [PLUGIN ...]
Plugin(s) to run. For a list of options use --info
-f FILE Memory image file to run plugin on
-k KDBG KDBG address for the images (in hex)
--db DB SQLite db file, for efficient input/output
--profile PROFILE Volatility profile for the images (e.g. WinXPSP2x86)
--debug Print debugging statements
--info Print available volatility profiles, plugins
--tsv Print screen formatted output.
--grepable Print in grepable text format
--filter FILTER Filter results on name:value pair, e.g., pid:42
--filtertype FILTERTYPE
Filter match type; either "exact" or "partial",
defaults to partial
--diff BASELINE Diff the imageFile|db with this db file as a baseline
-u FIELD [FIELD ...] Use the specified fields to determine uniqueness of
memobjs when diffing
--warnings Look for suspicious objects.
-q Query the supplied db (via --db).
See #python damm.py --info
apihooks callbacks connections devicetree dlls evtlogs handles idt injections messagehooks mftentries modules mutants privileges processes services sids timers
Supply a profile as in Volatility, a memory image and a list of plugins to run (or 'all') to get terminal output:
python damm.py --profile WinXPSP2x86 -f memory.dmp -p processes | less -S
(or python damm.py --profile WinXPSP2x86 -f memory.dmp -p processes dlls modules)
(or python damm.py --profile WinXPSP2x86 -f memory.dmp -p all)
processes
offset name pid ppid prio image_path_name create_time exit_time threads session_id handles is_wow64 pslist psscan thrdproc pspcid csrss session deskthrd command_line
0x25c8830 System 4 0 8 59 403 False True True True True False False False
0x225ada0 alg.exe 188 668 8 C:\WINDOWS\System32\alg.exe 2010-10-29 17:09:09 UTC+0000 6 0 107 False True True True True True True True C:\WINDOWS\System32\alg.exe
0x2114938 ipconfig.exe 304 968 8 2011-06-03 04:31:35 UTC+0000 2011-06-03 04:31:36 UTC+0000 0 0 False True True False True False False False
0x2086978 TSVNCache.exe 324 1196 8 C:\Program Files\TortoiseSVN\bin\TSVNCache.exe 2010-10-29 17:11:49 UTC+0000 7 0 54 False True True True True True True True "C:\Program Files\TortoiseSVN\bin\TSVNCache.exe"
0x22df020 smss.exe 376 4 11 \SystemRoot\System32\smss.exe 2010-10-29 17:08:53 UTC+0000 3 19 False True True True True False False False \SystemRoot\System32\smss.exe
...
To make these results persist in a SQLite db, just supply a filename for the db:
python damm.py --profile WinXPSP2x86 -f memory.dmp -p processes --db my_results.db
This will print results to the terminal as well as store them in 'my_results.db'
To see the results again:
python damm.py -p processes --db my_results.db
(Note that you no longer need the memory image or to specify a profile, and the listing will come out pretty close to instantly regardless of how long the original processing took.)
If you later wish to see processes and other plugins:
python damm.py --profile WinXPSP2x86 -p processes dlls modules --db my_results.db
Will:
- consult the db for the 'processes' output
- run the 'dlls' and 'modules' plugins
- display the results
- store the new results in the db
Once you have stored some data in a db, you can query it with the -q switch
python damm.py -q --db my_results.db
profile: WinXPSP2x86
memimg: WinXPSP2x86/stuxnet.vmem
COMPUTERNAME: JAN-DF663B3DBF1
plugins: processes dlls modules
Plugins have attributes that can have types for filtering, e.g., for processes: (use --info to see for all plugin attributes)
offset
name : string
pid : pid
ppid : pid
image_path_name : string
command_line : string
create_time
exit_time
threads
session_id
handles
is_wow64
pslist
psscan
thrdproc
pspcid
csrss
session
deskthrd
These attributes and types can be leveraged by the differencing and filtering functions of DAMM
To use the differencing engine, create 2 databases from 2 distinct memory images, such as one from before and one from after a piece of malware is executed
python damm.py --profile WinXPSP2x86-f before.dmp -p processes --db before.db
python damm.py --profile WinXPSP2x86 -f after.dmp -p processes --db after.db
Then use the --diff option for the baseline db (here, the db from the uninfected memory image).
python damm.py -p processes --db after.db --diff before.db
processes
Status offset name pid ppid prio image_path_name create_time exit_time threads session_id handles is_wow64 pslist psscan thrdproc pspcid csrss session deskthrd command_line
New 0x17d22e0 pythonw.exe 1256 1940 8 2013-10-31 23:23:14 UTC+0000 2013-10-31 23:23:19 UTC+0000 0 -268370093 False False True False False False False False
Changed 0x18b4d38 svchost.exe 1080 692 8 2013-10-31 17:21:26 UTC+0000 66->71 False False True False False False False False
Changed 0x1915198 winlogon.exe 648 376 13 2013-10-31 17:21:25 UTC+0000 24->26 False False True False False False False False
Changed 0x1900120 services.exe 692 648 9 2013-10-31 17:21:25 UTC+0000 16->18 False False True False False False False False
Changed 0x18b0360 svchost.exe 1124 692 8 2013-10-31 17:21:26 UTC+0000 5->6 False False True False False False False False
Changed 0x1875490 explorer.exe 1636 1596 8 2013-10-31 17:21:27 UTC+0000 13->14 False False True False False False False False
...
The results look similar to the 'processes' plugin output above, but there are some differences:
- Only results that are new in 'after.db' or that are in both dbs but have some attributes that changed from 'before.db' are displayed (the output here is snipped).
- Results that are only in the 'after.db' have 'New' in the first ('Status') column
- Results in that have changed between the dbs have a 'Status' of 'Changed', and, importantly, denote the changes DAMM detected with '->': in the last line of output above the number of threads has changed.
In order to determine which processes exist in both memory captures above, behind the scenes certain attributes of processes are used to make a unique identifier for each. For example, by default DAMM used the pid, ppid, name, and start time as the unique identifier of a process. This makes sense as these things are unlikely to (shouldn't? can't?) change over the life of the process, as opposed to attributes like the number of threads and handles, which change constantly. This default set works fine for comparisons of objects from memory images from the same boot of the same machine (e.g., using VM snapshots), but what about comparing across memory images taken from different boots of the machine? Or even other machines? The pid, and ppid will likely not be the same, but the name, image path and command line should be.
Comparing a stock XPSP2x86 memory image with our image after some malware ran:
python damm.py -p processes --diff stock_WinXPSP2x86_processes.db --db after_malware.db
...
New 0x1874da0 explorer.exe 1636 1596 C:\WINDOWS\Explorer.EXE C:\WINDOWS\Explorer.EXE 2013-10-31 17:21:27 UTC+0000 None 12 0 316 False True False True True True True True
New 0x1983020 smss.exe 376 4 \SystemRoot\System32\smss.exe \SystemRoot\System32\smss.exe 2013-10-31 17:21:24 UTC+0000 None 3 19 False True False True True False False
New 0x182cda0 wpabaln.exe 1812 648 C:\WINDOWS\system32\wpabaln.exe C:\WINDOWS\system32\wpabaln.exe 2013-10-31 23:10:13 UTC+0000 None 1 0 58 False True False True True True True
New 0x1883308 spoolsv.exe 1500 692 C:\WINDOWS\system32\spoolsv.exe C:\WINDOWS\system32\spoolsv.exe 2013-10-31 17:21:27 UTC+0000 None 14 0 113 False True False True True True True
Changed 0x1bcc830->0x1bcc9c8 System 4 0 None None 60->71 209->266 False True True->False True True False False False
Results in every process being tagged as 'New' (except System) becuase by default we use the pid and ppid to make the unique identifier.
Because of this, DAMM allows the user to specify which attributes of some object can be used to create a unique identifier. If we tell DAMM to just use process name, image_path_name, and command_line, we get much more reasonable results:
python damm.py -p processes --diff stock_WinXPSP2x86_processes.db --db after_malware.db -u name image_path_name command_line
Status offset name pid ppid image_path_name command_line create_time exit_time threads session_id handles is_wow64 pslist psscan thrdproc pspcid csrss session deskthrd
New 0x1860020 wuauclt.exe 548 1080 C:\WINDOWS\system32\wuauclt.exe "C:\WINDOWS\system32\wuauclt.exe" /RunStoreAsComServer Local\[438]SUSDS109850d1d4659d4590c0302d99249922 2013-10-31 17:22:20 UTC+0000 None 7
New 0x1891da0 VBoxTray.exe 1932 1636 C:\WINDOWS\system32\VBoxTray.exe "C:\WINDOWS\system32\VBoxTray.exe" 2013-10-31 17:21:29 UTC+0000 None 7 0 65 False True False True True
New 0x195bbf0 pythonw.exe 1256 1940 C:\Python27\pythonw.exe C:\Python27\pythonw.exe C:\hkurkt\analyzer.py 2013-10-31 23:09:24 UTC+0000 None 5 0 114 False True False True True True
New 0x1877448 MagicDisc.exe 1960 1636 C:\Program Files\MagicDisc\MagicDisc.exe "C:\Program Files\MagicDisc\MagicDisc.exe" 2013-10-31 17:21:29 UTC+0000 None 1 0 24 False True False
New 0x182cda0 wpabaln.exe 1812 648 C:\WINDOWS\system32\wpabaln.exe C:\WINDOWS\system32\wpabaln.exe 2013-10-31 23:10:13 UTC+0000 None 1 0 58 False True False True True True True
New 0x1877940 pythonw.exe 1940 1636 C:\Python27\pythonw.exe "C:\Python27\pythonw.exe" "C:\Documents and Settings\jawauser\Start Menu\Programs\Startup\agent.pyw" 2013-10-31 17:21:29 UTC+0000 None 1 0
New 0x1875718 tdl3 1344 1256 C:\DOCUME~1\jawauser\LOCALS~1\Temp\tdl3 "C:\DOCUME~1\jawauser\LOCALS~1\Temp\tdl3" 2013-10-31 23:09:25 UTC+0000 None 1 0 37 False True False True True
New 0x18ee360 VBoxService.exe 860 692 C:\WINDOWS\system32\VBoxService.exe system32\VBoxService.exe 2013-10-31 17:21:26 UTC+0000 None 8 0 106 False True False True True True
Changed 0x1bcc830->0x1bcc9c8 System 4 0 None None 60->71 209->266 False True True->False True True False False False
Changed 0x18b7020->0x18b4648 svchost.exe 1076->1080 680->692 C:\WINDOWS\System32\svchost.exe C:\WINDOWS\System32\svchost.exe -k netsvcs 2011-09-26 01:33:36 UTC+0000->2013-10-31 17:21:26 UTC+0000 None 87->7
Changed 0x1994d08->0x18ffa30 services.exe 680->692 636->648 C:\WINDOWS\system32\services.exe C:\WINDOWS\system32\services.exe 2011-09-26 01:33:35 UTC+0000->2013-10-31 17:21:25 UTC+0000 None 15->1
Changed 0x16a2cd0->0x18fd648 lsass.exe 692->704 636->648 C:\WINDOWS\system32\lsass.exe C:\WINDOWS\system32\lsass.exe 2011-09-26 01:33:35 UTC+0000->2013-10-31 17:21:25 UTC+0000 None 24->22 0 356->
Changed 0x1aefda0->0x1983020 smss.exe 384->376 4 \SystemRoot\System32\smss.exe \SystemRoot\System32\smss.exe 2011-09-26 01:33:32 UTC+0000->2013-10-31 17:21:24 UTC+0000 None 3 19 False
Changed 0x189a1d0->0x18a7a60 svchost.exe 1336->1152 680->692 C:\WINDOWS\system32\svchost.exe C:\WINDOWS\system32\svchost.exe -k LocalService 2011-09-26 01:33:37 UTC+0000->2013-10-31 17:21:26 UTC+0000 None 14->1
Changed 0x16c9b40->0x1914aa8 winlogon.exe 636->648 384->376 \??\C:\WINDOWS\system32\winlogon.exe winlogon.exe 2011-09-26 01:33:35 UTC+0000->2013-10-31 17:21:25 UTC+0000 None 16->18 0 498->509
Changed 0x1ab5248->0x18c1020 svchost.exe 944->992 680->692 C:\WINDOWS\system32\svchost.exe C:\WINDOWS\system32\svchost -k rpcss 2011-09-26 01:33:36 UTC+0000->2013-10-31 17:21:26 UTC+0000 None 11->9 0
Changed 0x14b03e0->0x183e620 alg.exe 2272->1888 680->692 C:\WINDOWS\System32\alg.exe C:\WINDOWS\System32\alg.exe 2011-09-26 01:33:55 UTC+0000->2013-10-31 17:21:37 UTC+0000 None 7->6 0 112->105
Changed 0x1af5cd0->0x1874da0 explorer.exe 1752->1636 1696->1596 C:\WINDOWS\Explorer.EXE C:\WINDOWS\Explorer.EXE 2011-09-26 01:33:45 UTC+0000->2013-10-31 17:21:27 UTC+0000 None 32->12 0 680->316 False
Changed 0x1670020->0x18e4020 svchost.exe 868->904 680->692 C:\WINDOWS\system32\svchost.exe C:\WINDOWS\system32\svchost -k DcomLaunch 2011-09-26 01:33:35 UTC+0000->2013-10-31 17:21:26 UTC+0000 None 17
Changed 0x15685e0->0x1883308 spoolsv.exe 1516->1500 680->692 C:\WINDOWS\system32\spoolsv.exe C:\WINDOWS\system32\spoolsv.exe 2011-09-26 01:33:39 UTC+0000->2013-10-31 17:21:27 UTC+0000 None 14 0 159->
Changed 0x1816ab8->0x190c020 csrss.exe 612->624 384->376 \??\C:\WINDOWS\system32\csrss.exe C:\WINDOWS\system32\csrss.exe ObjectDirectory=\Windows SharedSection=1024,3072,512 Windows=On SubSystemType=Windows S
Changed 0x19f7548->0x18afc70 svchost.exe 1200->1124 680->692 C:\WINDOWS\system32\svchost.exe C:\WINDOWS\system32\svchost.exe -k NetworkService 2011-09-26 01:33:37 UTC+0000->2013-10-31 17:21:26 UTC+0000 None
DAMM now identifies fewer processes as 'New' (including the malware process), allowing the investigator to focus their efforts on these processes.
With all plugins run on a small memory sample, we get ~14,000 memory objects: processes, dlls, modules, etc. What if we have already identified some process or string of interest? Grep can be problematic, especially when searching for pids, so DAMM includes a simple type and filtering system. To filter on objects that have a pid attribute of a certain value:
python damm.py -p processes dlls connections handles --db after_malware.db --filter pid:1344
processes
offset name pid ppid image_path_name command_line create_time exit_time threads session_id handles is_wow64 pslist psscan thrdproc pspcid csrss session deskthrd
0x1875718 tdl3 1344 1256 C:\DOCUME~1\jawauser\LOCALS~1\Temp\tdl3 "C:\DOCUME~1\jawauser\LOCALS~1\Temp\tdl3" 2013-10-31 23:09:25 UTC+0000 None 1 0 37 False True False True True TrueTrue True
dlls
proc_pid dll_base size_of_image load_count full_dll_name
1344 0x73000000 155648 0x1 C:\WINDOWS\system32\WINSPOOL.DRV
1344 0x400000 77824 0xffff C:\DOCUME~1\jawauser\LOCALS~1\Temp\tdl3
1344 0x77f10000 299008 0xffff C:\WINDOWS\system32\GDI32.dll
1344 0x7e410000 593920 0xffff C:\WINDOWS\system32\user32.dll
...
connections
offset pid local_ip local_port remote_ip remote_port allocated
0x1853580 1344 192.168.56.101 1035 192.168.43.171 2042 True
handles
offset pid handle_value granted_access object_type name
0xe1007ff0 1344 0x4 0xf0003 KeyedEvent CritSecOutOfMemoryEvent
0x81a43310 1344 0x3c 0x1f03ff Thread TID 384 PID 1344
0x81902878 1344 0x6c 0x1f01ff File \Device\Tcp
0x81902240 1344 0xc 0x100020 File \Device\HarddiskVolume1\DOCUME~1\jawauser\LOCALS~1\Temp
0x81906158 1344 0x38 0x1f0003 Semaphore shell.{A48F1A32-A340-11D1-BC6B-00A0C90312E1}
0x81a43310 1344 0x64 0x1f03ff Thread TID 384 PID 1344
0x81851900 1344 0x68 0x1f01ff File \Device\Afd\Endpoint
0x81953f78 1344 0x20 0xf01ff Desktop Default
0xe106b648 1344 0x44 0xf003f Key MACHINE\SYSTEM\CONTROLSET001\SERVICES\WINSOCK2\PARAMETERS\PROTOCOL_CATALOG9
0x81902950 1344 0x70 0x1f01ff File \Device\Tcp
0x81902cd0 1344 0x7c 0x100001 File \Device\KsecDD
0xe1aa0ca0 1344 0x80 0x2001f Key USER\S-1-5-21-1644491937-789336058-854245398-1003\SOFTWARE\MICROSOFT\WINDOWS\CURRENTVERSION\INTERNET SETTINGS
(many lines removed for brevity)
This can give a nice overview of the objects associated with a process.
Even more powerful, diff and filtering can be used in conjunction. I have a memory sample from before a tdl3 infection and one after. Searching for the string 'tdl' in the before db results in ~600 hits. In the after infection db, there are ~730 hits. (Note that ntdll.dll contain the string tdl.) Using diff and filtering in conjunction as below results in only ~180 hits - a significant reduction. Note that for string and pid filtering, DAMM defaults to exact matching. Using the --filtertype partial option changes the fileterming to partial matching.
python damm.py -p all --diff before_tdl3.db --db after_tdl3.db --filter string:tdl --filtertype partial > string_tdl_diff.txt
In an attempt to make the triage process even easier, DAMM has an experimental warning system built in to sniff out signs of malicious activity including:
For certain Windows processes:
- incorrect parent/child relationships
- hidden processes
- incorrect binary path
- incorrect default priority
- incorrect session
For all processes, and loaded DLLs and modules:
- loaded/run from temp directory
For DLLs:
- bogus extensions
- hidden DLLs
Plus more!
- PE headers in injections
- SIDs giving domain access
- debug privileges ...
python damm.py --db after_tdl3.db --warnings
See the warnings.py file for much more information on what DAMM checks for.
Thanks to the Volatility team for the Art of Memory Forensics book as well as the Volatility cheat sheet where many of these warning ideas came from!
For questions or comments: damm@504ensics.com For bug reports, please use the github issue tracker.