Ghidra-based AVR Lifter #1228
Open
Ghidra-based AVR Lifter #1228
Conversation
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
ivg
force-pushed
the
avr
branch
4 times, most recently
from
018ff17
to
671b859
Compare
ivg
force-pushed
the
avr
branch
3 times, most recently
from
aa81f60
to
5f0b9ac
Compare
ivg
force-pushed
the
avr
branch
2 times, most recently
from
248a8fd
to
8894ec1
Compare
Current State ------------- 1) Works with only the latest versions of LLVM (11 and above) 2) the instructions so far implemented: ADC, ADD, LDI Design Issues ------------- A unique feature of AVR is that it maps registers directly to RAM, thus we theoretically can't represent registers as variables, but instead shall model them as memory addresses[^1]. Treating this AVR feature fairly, will render quite hard to read and to analyze code. After digging through the forums, datasheets, compilers source code, and grepping avr-objdump outputs, we can presume that C compilers are not leveraging this feature and use normal reads and writes to access registers, instead of loads and stores. And the memory-mapping of registers is mostly reserved RAM-less AVR boards, where 32 register plays the role of minimal RAM. Of course, it doesn't mean that the malicious code can't exploit this. With this in mind, we still decided to model AVR as normal register-based machine but add a command-line option later, that will enable a conservative model, that lacks registers. Formally, our current lifter implementation (the one with registers) just assumes that all addresses in load and stores operation are greater than 32. [^1]: And yes, `*0 = 'B'` is a perfectly valid code on AVR that writes `0x42` into the `R0` register.
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Add this suggestion to a batch that can be applied as a single commit.
This suggestion is invalid because no changes were made to the code.
Suggestions cannot be applied while the pull request is closed.
Suggestions cannot be applied while viewing a subset of changes.
Only one suggestion per line can be applied in a batch.
Add this suggestion to a batch that can be applied as a single commit.
Applying suggestions on deleted lines is not supported.
You must change the existing code in this line in order to create a valid suggestion.
Outdated suggestions cannot be applied.
This suggestion has been applied or marked resolved.
Suggestions cannot be applied from pending reviews.
Suggestions cannot be applied on multi-line comments.
Suggestions cannot be applied while the pull request is queued to merge.
Suggestion cannot be applied right now. Please check back later.
Current State
Since LLVM is not yet ready for AVR we switched to Ghidra. We still have to use LLVM as a loader (and it proves to fail on AVR as well, but we addressed it). The lifter and decode work well so far, but we lack good testing binaries to be sure about that. The data model of the lifter (namely the registers) is not yet clear.
We will try adding some tests before merging.