github/darkfi
1
1
Fork 0
mirror of https://github.com/darkrenaissance/darkfi.git synced 2026-01-06 21:34:00 -05:00
Code Issues Packages Projects Releases Wiki Activity

fuzz/README.md: fuzzing explicitly requires nightly now

Browse Source 
While we've moving (almost) everything stable by default now,
fuzzing still requires nightly.

Since stable is now the default we need to add "+nightly" to
various fuzzing commands.
This commit is contained in:
Paul Otten
2025-04-14 16:10:28 -04:00
parent 0a240bea37
commit be2ced6543
2 changed files with 16 additions and 11 deletions
Download Patch File Download Diff File Expand all files Collapse all files

6
doc/src/dev/contrib/contrib.md
View File

@@ -79,16 +79,16 @@ Run the following from the DarkFi repo folder:
```
$ cd fuzz/
$ cargo fuzz list
$ cargo +nightly fuzz list
```
This will list the available fuzzing targets. Choose one and run it with:
### Run
```
# format: cargo fuzz run TARGET
# format: cargo +nightly fuzz run TARGET
# e.g. if `serial` is your target:
$ cargo fuzz run --all-features -s none --jobs $(nproc) serial
$ cargo +nightly fuzz run --all-features -s none --jobs $(nproc) serial
```
This process will run infinitely until a crash occurs or until it is cancelled by the user.

21
fuzz/README.md
View File

@@ -11,14 +11,19 @@ tool `honggfuzz` and its related files are located in `fuzz/honggfuzz`.
cargo install cargo-fuzz
```
You will also need Rust's nightly toolchain installed.
```sh
rustup toolchain install nightly
```
## Usage
```sh
# List available targets
$ cargo fuzz list
$ cargo +nightly fuzz list
# Run fuzzer on a target
# format: cargo fuzz run TARGET
# format: cargo +nightly fuzz run TARGET
# e.g. if `serial` is your target:
$ cargo fuzz run serial
$ cargo +nightly fuzz run serial
```
This process will run infinitely until a crash occurs or until it is cancelled by the user.
@@ -55,7 +60,7 @@ We store dictionaries in the `dictionaries/` directory.
A more efficient way to fuzz safe Rust code is the following:
```sh
cargo fuzz run --jobs $(nproc) -s none --all-features TARGET -- -dict=dictionaries/SOMEDICT.dict
cargo +nightly fuzz run --jobs $(nproc) -s none --all-features TARGET -- -dict=dictionaries/SOMEDICT.dict
```
## Fuzzing Corpora
@@ -82,8 +87,8 @@ easier to share corpora between libfuzzer and honggfuzz.)
e.g. scenario: while testing ZkBinary's decode() function, you find
that an empty input causes a panic.
* Identify your fuzz target (`cargo fuzz list` or whatever you used
for `cargo fuzz run TARGET`
* Identify your fuzz target (`cargo +nightly fuzz list` or whatever
you used for `cargo +nightly fuzz run TARGET`)
* Examine the fuzzing artifacts: `ls artifacts/TARGET/`
* `cat` the file and check that it matches the error message from
the fuzzer. The filename's prefix will match the kind of error
@@ -124,7 +129,7 @@ It is possible to increase the amount of memory libFuzzer is allowed to use by p
to it via libFuzzer like so:
```sh
cargo fuzz run --all-features zkas-decoder -- "-rss_limit_mb=4096"
cargo +nightly fuzz run --all-features zkas-decoder -- "-rss_limit_mb=4096"
```
To disable memory limits entirely, pass the argument:
@@ -165,7 +170,7 @@ rustup component add llvm-tools-preview
# Generate coverage files. Run this from fuzz/
# This step will be faster if you minimize the corpus first.
cargo fuzz coverage zkas-compile
cargo +nightly fuzz coverage zkas-compile
# Manually create a .profdata file. (One is generated by the above command, but it appears to be broken)
llvm-profdata merge -sparse coverage/zkas-compile/raw/* -o zkas-compile.profdata