There's a new Ghidra release last week! Lots of improvements to the debugger, which is awesome. But this brings up some thoughts that have been triggering my vulnerability-and-exploitation-specific OCD for some time now.
Behind every good RE tool is a crappy crappy database. Implicitly we, as a community, understand there is no good reason that every reverse engineering project needs to implement a key-value store, or a B-Tree https://github.com/NationalSecurityAgency/ghidra/tree/master/Ghidra/Framework/DB/src/main/java/db, or partner with a colony of bees which maintain tool state by various wiggly dances. But yet each and every tool has a developer with decades of reverse engineering experience on rare embedded platforms either building custom indexes in a pale imitation of a real DB structure or engaging in insect-based diplomacy efforts.
I think the Ghidra team (and Binja/IDA teams!) are geniuses, but they are probably NOT geniuses at building database engines. And reading through the issues https://github.com/NationalSecurityAgency/ghidra/issues/985 with ANY reverse engineering product you find that performance even for the base feature-set is a difficult ask.
My plea is this: We need to port Ghidra to Neo4j as soon as possible. Having a real Graph DB store underneath Ghidra solves the scalability issues. I understand the difficulty here is: There are few engineers who understand both Neo4j and reverse engineering to the point where this can be done. I mean, why do it in Neo4j and not PostGres? An argument can be made for both, in the sense that PostGres is truly Free and the most solid DB on the market. The pluses for Neo4j are that RE data is typically graph-based more than linear.
I spent the last two years learning graph dbs, out of some masochistic desire and ended up getting certified - and I can still RE a little bit. I will manage the team porting Ghidra to Neo4j if someone funds it. :)
Either way, sooner is better than later. There are so many companies and people relying on these tools that it seems silly to do anything else.
-dave P.S. Yes, I remember BinNavi used MsSQL installs for its data, and this was annoying to install but ... I get why Halvar did it at the time. It's because he had real work to do and building a DB was not it. I can only assume Reven doesn't use their own DB? I mean the benefits for interoperability would be huge between tools. . . like literally everything you want to do with these tools is better with a real DB underneath.
There is joernio's ghidra2cpg, not sure why they now seem to be pushing a forked set of patches https://github.com/joernio/ghidra, probably the DB format changes too rapidly or some other "we automatically intake unknown relationships lost statically". That might get part of what you're looking for, even though, it isn't an exact fit, bringing in some higher level tooling, like all the graphql UI's that contextualize queries with type context are so helpful, whenever I don't have context aware syntax support, thar barrier to actually do anything limit's my enthusiasm so that only the most impactful (perceived before getting too far) get my attention (and I'm often wrong so :). I forget if joern still uses Neo4j, I am confident that it's the best FOSS available for describing code/binaries right now.
Getting more tools in this space is a great initiative that deserves attention. Being able to communicate so expressively, codifying knowledge for bugs some helpers around supporting guided generation of queries for arbitrary conditions, the benefits for invariant analysis (as can been seen with Semmle/CodeQL) are extreme.
On Mon, Jun 26, 2023 at 3:46 PM Dave Aitel via Dailydave < dailydave@lists.aitelfoundation.org> wrote:
There's a new Ghidra release last week! Lots of improvements to the debugger, which is awesome. But this brings up some thoughts that have been triggering my vulnerability-and-exploitation-specific OCD for some time now.
Behind every good RE tool is a crappy crappy database. Implicitly we, as a community, understand there is no good reason that every reverse engineering project needs to implement a key-value store, or a B-Tree https://github.com/NationalSecurityAgency/ghidra/tree/master/Ghidra/Framework/DB/src/main/java/db, or partner with a colony of bees which maintain tool state by various wiggly dances. But yet each and every tool has a developer with decades of reverse engineering experience on rare embedded platforms either building custom indexes in a pale imitation of a real DB structure or engaging in insect-based diplomacy efforts.
I think the Ghidra team (and Binja/IDA teams!) are geniuses, but they are probably NOT geniuses at building database engines. And reading through the issues https://github.com/NationalSecurityAgency/ghidra/issues/985 with ANY reverse engineering product you find that performance even for the base feature-set is a difficult ask.
My plea is this: We need to port Ghidra to Neo4j as soon as possible. Having a real Graph DB store underneath Ghidra solves the scalability issues. I understand the difficulty here is: There are few engineers who understand both Neo4j and reverse engineering to the point where this can be done. I mean, why do it in Neo4j and not PostGres? An argument can be made for both, in the sense that PostGres is truly Free and the most solid DB on the market. The pluses for Neo4j are that RE data is typically graph-based more than linear.
I spent the last two years learning graph dbs, out of some masochistic desire and ended up getting certified - and I can still RE a little bit. I will manage the team porting Ghidra to Neo4j if someone funds it. :)
Either way, sooner is better than later. There are so many companies and people relying on these tools that it seems silly to do anything else.
-dave P.S. Yes, I remember BinNavi used MsSQL installs for its data, and this was annoying to install but ... I get why Halvar did it at the time. It's because he had real work to do and building a DB was not it. I can only assume Reven doesn't use their own DB? I mean the benefits for interoperability would be huge between tools. . . like literally everything you want to do with these tools is better with a real DB underneath.
Dailydave mailing list -- dailydave@lists.aitelfoundation.org To unsubscribe send an email to dailydave-leave@lists.aitelfoundation.org
Maybe not from a vuln hunter perspective, but from a threat hunter view we usually mill out to targeting platforms, e.g., Vertex Synapse (commercial) and/or MISP (open-source) for graph-ready pivoting and analysis
For example, TheDFIRReport hosts a MISP that contexts the process objects template -- https://github.com/MISP/misp-objects/blob/main/objects/process/definition.js... -- and fills it up with ATT&CK tags Sure, I'd also context that with YARA or capa rules, but MISP etc also supports embedding and mapping those. While the VirusTotal capa support is not the full-context kind you get from the entire tool (VTE doesn't show the `-vv`output, only the top-level jank), and it's also not capa-explorer level where you get the function mappings, but I'd be curious to see the behavior profiling and livehunting (which the YARA vt module now supports Sigma-style rule checking against the behavior profiles) across, say, the Immunity Security CANVAS leaked binaries that made it up on VT years back, or other VT-visible exploit codes
What are the core problems attempting to solve here? For malware analysis, it's typically mapping in changes, historical views, popular function analysis, and evolution For function analysis, all of the reversing platforms, e.g., IDA Pro (sub BinaryNinja, sub Ghidra, sub Radare2/Cutter) have a capa-explorer plugin of sorts, and usually commonly-found function analysis routines (for IDA Pro it's called Lumina, but each of these also supports the concept, i.e., radare2 zignatures. Then there's also Writeprint and other more classic techniques that work very-broadly
In terms of changes and evolution, malware analysis on the reversing platforms hasn't seen much intooling, but they are capable. There's ASM2VEC (of course, very different compared to node2vec/graph2vec but having relevance) as seen in IDA Pro plugins such as -- https://github.com/McGill-DMaS/Kam1n0-Community -- (more on kam1n0 here -- https://mcgillnews.mcgill.ca/s/1762/news/interior.aspx?sid=1762&gid=2&am... )
Is there a knowledgebase approach that is cross-functional and cross vuln-/mal-/exploit-/threat- efforts to be had here?
On Tue, Jun 27, 2023 at 9:24 AM Shane Macaulay via Dailydave < dailydave@lists.aitelfoundation.org> wrote:
There is joernio's ghidra2cpg, not sure why they now seem to be pushing a forked set of patches https://github.com/joernio/ghidra, probably the DB format changes too rapidly or some other "we automatically intake unknown relationships lost statically". That might get part of what you're looking for, even though, it isn't an exact fit, bringing in some higher level tooling, like all the graphql UI's that contextualize queries with type context are so helpful, whenever I don't have context aware syntax support, thar barrier to actually do anything limit's my enthusiasm so that only the most impactful (perceived before getting too far) get my attention (and I'm often wrong so :). I forget if joern still uses Neo4j, I am confident that it's the best FOSS available for describing code/binaries right now.
Getting more tools in this space is a great initiative that deserves attention. Being able to communicate so expressively, codifying knowledge for bugs some helpers around supporting guided generation of queries for arbitrary conditions, the benefits for invariant analysis (as can been seen with Semmle/CodeQL) are extreme.
On Mon, Jun 26, 2023 at 3:46 PM Dave Aitel via Dailydave < dailydave@lists.aitelfoundation.org> wrote:
There's a new Ghidra release last week! Lots of improvements to the debugger, which is awesome. But this brings up some thoughts that have been triggering my vulnerability-and-exploitation-specific OCD for some time now.
Behind every good RE tool is a crappy crappy database. Implicitly we, as a community, understand there is no good reason that every reverse engineering project needs to implement a key-value store, or a B-Tree https://github.com/NationalSecurityAgency/ghidra/tree/master/Ghidra/Framework/DB/src/main/java/db, or partner with a colony of bees which maintain tool state by various wiggly dances. But yet each and every tool has a developer with decades of reverse engineering experience on rare embedded platforms either building custom indexes in a pale imitation of a real DB structure or engaging in insect-based diplomacy efforts.
I think the Ghidra team (and Binja/IDA teams!) are geniuses, but they are probably NOT geniuses at building database engines. And reading through the issues https://github.com/NationalSecurityAgency/ghidra/issues/985 with ANY reverse engineering product you find that performance even for the base feature-set is a difficult ask.
My plea is this: We need to port Ghidra to Neo4j as soon as possible. Having a real Graph DB store underneath Ghidra solves the scalability issues. I understand the difficulty here is: There are few engineers who understand both Neo4j and reverse engineering to the point where this can be done. I mean, why do it in Neo4j and not PostGres? An argument can be made for both, in the sense that PostGres is truly Free and the most solid DB on the market. The pluses for Neo4j are that RE data is typically graph-based more than linear.
I spent the last two years learning graph dbs, out of some masochistic desire and ended up getting certified - and I can still RE a little bit. I will manage the team porting Ghidra to Neo4j if someone funds it. :)
Either way, sooner is better than later. There are so many companies and people relying on these tools that it seems silly to do anything else.
-dave P.S. Yes, I remember BinNavi used MsSQL installs for its data, and this was annoying to install but ... I get why Halvar did it at the time. It's because he had real work to do and building a DB was not it. I can only assume Reven doesn't use their own DB? I mean the benefits for interoperability would be huge between tools. . . like literally everything you want to do with these tools is better with a real DB underneath.
Dailydave mailing list -- dailydave@lists.aitelfoundation.org To unsubscribe send an email to dailydave-leave@lists.aitelfoundation.org
Dailydave mailing list -- dailydave@lists.aitelfoundation.org To unsubscribe send an email to dailydave-leave@lists.aitelfoundation.org
I've been using Joern with Ghidra for a couple years. For my needs, the ghidra2cpg project is too limiting. It works at the disassembly level by mapping various CPU instructions for a few common archs via a pre-patched bundled Ghidra build (ie, no custom loaders, processor definitions, etc). As a work-around, I've hacked together an alternative (worse) approach that operates on the decompiler level: a Ghidra script that just enumerate every function and adds the decompiler psuedo-C to a Joern project, using Joern's fuzzy C parser. It's a lossy hack, but it mostly just works. But both options undermine a lot of the value Ghidra provides.
A more ideal yet reachable solution that I've been toying with is to write a proper Joern CPG frontend parser for PCODE. This could provide a graphdb representations of the AST, CFG, CDG, DDG, and PDG, plus the CPG glue tying them (and the parent PCODE) all together.
Joern switched away from using native neo4j a while back for performance reasons, but they do provide an export utility to generate a neo4j-csv so you can use graphviz, import into neo4j, graphml, etc.
The big question: is the CPG format an ideal graph representation? Has anyone besides the authors' companies been working with at a high level? What known limitations does it present? I'd be interested in seeing what changes ShiftLeft and QwietAI have made. For reference, here's the original academic research: https://www.sec.cs.tu-bs.de/pubs/2014-ieeesp.pdf
On 6/27/23 10:08, Shane Macaulay via Dailydave wrote:
There is joernio's ghidra2cpg, not sure why they now seem to be pushing a forked set of patches https://github.com/joernio/ghidra, probably the DB format changes too rapidly or some other "we automatically intake unknown relationships lost statically". That might get part of what you're looking for, even though, it isn't an exact fit, bringing in some higher level tooling, like all the graphql UI's that contextualize queries with type context are so helpful, whenever I don't have context aware syntax support, thar barrier to actually do anything limit's my enthusiasm so that only the most impactful (perceived before getting too far) get my attention (and I'm often wrong so :). I forget if joern still uses Neo4j, I am confident that it's the best FOSS available for describing code/binaries right now.
Getting more tools in this space is a great initiative that deserves attention. Being able to communicate so expressively, codifying knowledge for bugs some helpers around supporting guided generation of queries for arbitrary conditions, the benefits for invariant analysis (as can been seen with Semmle/CodeQL) are extreme.
On Mon, Jun 26, 2023 at 3:46 PM Dave Aitel via Dailydave dailydave@lists.aitelfoundation.org wrote:
There's a new Ghidra release last week! Lots of improvements to the debugger, which is awesome. But this brings up some thoughts that have been triggering my vulnerability-and-exploitation-specific OCD for some time now. Behind every good RE tool is a crappy crappy database. Implicitly we, as a community, understand there is no good reason that every reverse engineering project needs to implement a key-value store, or a B-Tree <https://github.com/NationalSecurityAgency/ghidra/tree/master/Ghidra/Framework/DB/src/main/java/db>, or partner with a colony of bees which maintain tool state by various wiggly dances. But yet each and every tool has a developer with decades of reverse engineering experience on rare embedded platforms either building custom indexes in a pale imitation of a real DB structure or engaging in insect-based diplomacy efforts. I think the Ghidra team (and Binja/IDA teams!) are geniuses, but they are probably NOT geniuses at building database engines. And reading through the issues <https://github.com/NationalSecurityAgency/ghidra/issues/985> with ANY reverse engineering product you find that performance even for the base feature-set is a difficult ask. My plea is this: We need to port Ghidra to Neo4j as soon as possible. Having a real Graph DB store underneath Ghidra solves the scalability issues. I understand the difficulty here is: There are few engineers who understand both Neo4j and reverse engineering to the point where this can be done. I mean, why do it in Neo4j and not PostGres? An argument can be made for both, in the sense that PostGres is truly Free and the most solid DB on the market. The pluses for Neo4j are that RE data is typically graph-based more than linear. I spent the last two years learning graph dbs, out of some masochistic desire and ended up getting certified - and I can still RE a little bit. I will manage the team porting Ghidra to Neo4j if someone funds it. :) Either way, sooner is better than later. There are so many companies and people relying on these tools that it seems silly to do anything else. -dave P.S. Yes, I remember BinNavi used MsSQL installs for its data, and this was annoying to install but ... I get why Halvar did it at the time. It's because he had real work to do and building a DB was not it. I can only assume Reven doesn't use their own DB? I mean the benefits for interoperability would be huge between tools. . . like literally everything you want to do with these tools is better with a real DB underneath. _______________________________________________ Dailydave mailing list -- dailydave@lists.aitelfoundation.org To unsubscribe send an email to dailydave-leave@lists.aitelfoundation.org
Dailydave mailing list -- dailydave@lists.aitelfoundation.org To unsubscribe send an email to dailydave-leave@lists.aitelfoundation.org
Would it be easier to go from Ghidra -> BinExport -> Neo4j instead of Ghidra -> Neo4j? The BinExport format seems a bit more stable/documented to me (I have not played around with either in too much depth though).
dailydave@lists.aitelfoundation.org
-
Andre Gironda -
Dave Aitel -
David Manouchehri -
Jason Lash -
Shane Macaulay