Sancov based ngram & ctx implementation (#1864)

* ngram

* ctx

* push stuff

* passing cargo check

* save stuf

* add

* no default

* fuzzbench ready

* formatt

* aaaaaaaa

* adaptive map

* add all

* chg

* fix

* deleting stuff

* fmt

* clang18

* fuck

* save space

* fix?

* feature name

* fucking nightly only

* harness

* fix libafl_cc

---------

Co-authored-by: Dominik Maier <domenukk@gmail.com>

.github/workflows/build_and_test.yml

@@ -168,13 +168,26 @@ jobs:
   ubuntu-check:
     runs-on: ubuntu-22.04
     steps:
-    - name: Remove Dotnet & Haskell
-      run: rm -rf /usr/share/dotnet && rm -rf /opt/ghc
     - uses: actions-rs/toolchain@v1
       with:
         profile: minimal
         toolchain: stable
         components: llvm-tools  
+    - name: Free Disk Space (Ubuntu)
+      uses: jlumbroso/free-disk-space@main
+      with:
+        # this might remove tools that are actually needed,
+        # if set to "true" but frees about 6 GB
+        tool-cache: false
+
+        # all of these default to true, but feel free to set to
+        # "false" if necessary for your workflow
+        android: true
+        dotnet: true
+        haskell: true
+        large-packages: false
+        docker-images: true
+        swap-storage: true
     - name: Remove existing clang and LLVM
       run: sudo apt purge llvm* clang*
     - name: Install and cache deps

fuzzers/baby_fuzzer_swap_differential/Cargo.toml

@@ -21,7 +21,7 @@ debug = true
 
 [build-dependencies]
 anyhow = "1"
-bindgen = "0.63"
+bindgen = "0.69.4"
 cc = "1.0"
 
 [dependencies]

fuzzers/forkserver_libafl_cc/Cargo.toml

@@ -26,7 +26,7 @@ nix = "0.27"
 libafl = { path = "../../libafl/" }
 libafl_bolts = { path = "../../libafl_bolts/" }
 libafl_cc = { path = "../../libafl_cc/" }
-libafl_targets = { path = "../../libafl_targets/" }
+libafl_targets = { path = "../../libafl_targets/", features = ["sancov_pcguard_hitcounts", "libfuzzer"] }
 
 [lib]
 name = "libforkserver_libafl_cc"

fuzzers/forkserver_libafl_cc/src/bin/libafl_cc.rs

@@ -24,9 +24,6 @@ pub fn main() {
             .parse_args(&args)
             .expect("Failed to parse the command line")
             // Enable libafl's coverage instrumentation
-            .add_pass(LLVMPasses::AFLCoverage)
-            .add_arg("-mllvm")
-            .add_arg("-ctx") // Context sensitive coverage
             // Imitate afl-cc's compile definitions 
             .add_arg("-D__AFL_FUZZ_INIT()=int __afl_sharedmem_fuzzing = 1;extern unsigned int *__afl_fuzz_len;extern unsigned char *__afl_fuzz_ptr;unsigned char __afl_fuzz_alt[1048576];unsigned char *__afl_fuzz_alt_ptr = __afl_fuzz_alt;void libafl_start_forkserver(void)")
             .add_arg("-D__AFL_FUZZ_TESTCASE_BUF=(__afl_fuzz_ptr ? __afl_fuzz_ptr : __afl_fuzz_alt_ptr)")

fuzzers/fuzzbench_ctx/Cargo.toml

@@ -1,5 +1,5 @@
 [package]
-name = "libfuzzer_libpng_ctx"
+name = "fuzzbench"
 version = "0.11.2"
 authors = ["Andrea Fioraldi <andreafioraldi@gmail.com>", "Dominik Maier <domenukk@gmail.com>"]
 edition = "2021"
@@ -7,6 +7,7 @@ edition = "2021"
 [features]
 default = ["std"]
 std = []
+no_link_main = ["libafl_targets/libfuzzer_no_link_main"]
 
 [profile.release]
 lto = true
@@ -14,19 +15,25 @@ codegen-units = 1
 opt-level = 3
 debug = true
 
+[profile.release-fuzzbench]
+inherits = "release"
+debug = false
+strip = true
+
 [build-dependencies]
 cc = { version = "1.0", features = ["parallel"] }
 which = "4.4"
 
 [dependencies]
-libafl = { path = "../../libafl/", features = ["std", "derive", "llmp_compression", "introspection"] }
+libafl = { path = "../../libafl/" }
 libafl_bolts = { path = "../../libafl_bolts/" }
-libafl_targets = { path = "../../libafl_targets/", features = ["libfuzzer"] }
+libafl_targets = { path = "../../libafl_targets/", features = ["sancov_pcguard_hitcounts", "sancov_cmplog", "libfuzzer", "sancov_ctx"] }
 # TODO Include it only when building cc
 libafl_cc = { path = "../../libafl_cc/" }
-clap = { version = "4.0", features = ["derive"] }
+clap = { version = "4.0", features = ["default"] }
+nix = { version = "0.27", features = ["fs"] }
 mimalloc = { version = "*", default-features = false }
 
 [lib]
-name = "libfuzzer_libpng"
+name = "fuzzbench"
 crate-type = ["staticlib"]

fuzzers/fuzzbench_ctx/Makefile.toml

@@ -0,0 +1,108 @@
+[env]
+PROJECT_DIR = { script = ["pwd"] }
+CARGO_TARGET_DIR = { value = "${PROJECT_DIR}/target", condition = { env_not_set = ["CARGO_TARGET_DIR"] } }
+FUZZER_NAME="fuzzer"
+PROFILE = { value = "release", condition = {env_not_set = ["PROFILE"]} }
+PROFILE_DIR = {value = "release", condition = {env_not_set = ["PROFILE_DIR"] }}
+
+[tasks.unsupported]
+script_runner="@shell"
+script='''
+echo "Cargo-make not integrated yet on this"
+'''
+
+# Compilers
+[tasks.cxx]
+linux_alias = "cxx_unix"
+mac_alias = "cxx_unix"
+windows_alias = "unsupported"
+
+[tasks.cxx_unix]
+command = "cargo"
+args = ["build", "--profile", "${PROFILE}"]
+
+[tasks.cc]
+linux_alias = "cc_unix"
+mac_alias = "cc_unix"
+windows_alias = "unsupported"
+
+[tasks.cc_unix]
+command = "cargo"
+args = ["build", "--profile", "${PROFILE}"]
+
+# fuzz.o File
+[tasks.fuzz_o]
+linux_alias = "fuzz_o_unix"
+mac_alias = "fuzz_o_unix"
+windows_alias = "unsupported"
+
+[tasks.fuzz_o_unix]
+command = "${CARGO_TARGET_DIR}/${PROFILE_DIR}/libafl_cc"
+args = ["--libafl-no-link", "-O3", "-c", "fuzz.c", "-o", "fuzz.o"]
+dependencies = ["cc", "cxx"]
+
+# Fuzzer
+[tasks.fuzzer]
+linux_alias = "fuzzer_unix"
+mac_alias = "fuzzer_unix"
+windows_alias = "unsupported"
+
+[tasks.fuzzer_unix]
+command = "${CARGO_TARGET_DIR}/${PROFILE_DIR}/libafl_cxx"
+args = ["--libafl", "fuzz.o", "-o", "${FUZZER_NAME}", "-lm", "-lz"]
+dependencies = ["cc", "cxx", "fuzz_o"]
+
+# Run
+[tasks.run]
+linux_alias = "run_unix"
+mac_alias = "run_unix"
+windows_alias = "unsupported"
+
+[tasks.run_unix]
+script_runner="@shell"
+script='''
+rm -rf libafl_unix_shmem_server || true
+mkdir in || true
+echo a > in/a
+./${FUZZER_NAME} -o out -i in
+'''
+dependencies = ["fuzzer"]
+
+
+# Test
+[tasks.test]
+linux_alias = "test_unix"
+mac_alias = "test_unix"
+windows_alias = "unsupported"
+
+[tasks.test_unix]
+script_runner="@shell"
+script='''
+rm -rf libafl_unix_shmem_server || true
+mkdir in || true
+echo a > in/a
+# Allow sigterm as exit code 
+timeout 31s ./${FUZZER_NAME} -o out -i in >fuzz_stdout.log || true
+if grep -qa "objectives: 1" fuzz_stdout.log; then
+    echo "Fuzzer is working"
+else
+    echo "Fuzzer does not generate any testcases or any crashes"
+    exit 1
+fi
+rm -rf out || true
+rm -rf in || true
+'''
+dependencies = ["fuzzer"]
+
+# Clean
+[tasks.clean]
+linux_alias = "clean_unix"
+mac_alias = "clean_unix"
+windows_alias = "unsupported"
+
+[tasks.clean_unix]
+script_runner="@shell"
+script='''
+rm ./${FUZZER_NAME} || true
+rm fuzz.o || true
+'''

fuzzers/fuzzbench_ctx/fuzz.c

@@ -0,0 +1,19 @@
+#include <stdint.h>
+#include <stdlib.h>
+#include <string.h>
+
+int LLVMFuzzerTestOneInput(const uint8_t *Data, size_t Size) {
+  if (Size >= 8 && *(uint32_t *)Data == 0xaabbccdd) { abort(); }
+  char buf[8] = {'a', 'b', 'c', 'd'};
+
+  if (memcmp(Data, buf, 4) == 0) { abort(); }
+  return 0;
+}
+
+/*
+int main() {
+
+  char buf [10] = {0};
+  LLVMFuzzerTestOneInput(buf, 10);
+
+}*/

fuzzers/fuzzbench_ctx/src/bin/libafl_cc.rs

@@ -3,7 +3,7 @@ use std::env;
 use libafl_cc::{ClangWrapper, CompilerWrapper, LLVMPasses, ToolWrapper};
 
 pub fn main() {
-    let args: Vec<String> = env::args().collect();
+    let mut args: Vec<String> = env::args().collect();
     if args.len() > 1 {
         let mut dir = env::current_exe().unwrap();
         let wrapper_name = dir.file_name().unwrap().to_str().unwrap();
@@ -16,16 +16,25 @@ pub fn main() {
 
         dir.pop();
 
+        // Must be always present, even without --libafl
+        args.push("-fsanitize-coverage=trace-pc-guard,trace-cmp".into());
+
         let mut cc = ClangWrapper::new();
+
+        #[cfg(any(target_os = "linux", target_vendor = "apple"))]
+        cc.add_pass(LLVMPasses::AutoTokens);
+
         if let Some(code) = cc
             .cpp(is_cpp)
             // silence the compiler wrapper output, needed for some configure scripts.
             .silence(true)
+            // add arguments only if --libafl or --libafl-no-link are present
+            .need_libafl_arg(true)
             .parse_args(&args)
             .expect("Failed to parse the command line")
-            .add_pass(LLVMPasses::AFLCoverage)
+            .link_staticlib(&dir, "fuzzbench")
-            .add_passes_arg("-ctx") // Context sensitive coverage
+            .add_pass(LLVMPasses::CmpLogRtn)
-            .link_staticlib(&dir, "libfuzzer_libpng")
+            .add_pass(LLVMPasses::Ctx)
             .run()
             .expect("Failed to run the wrapped compiler")
         {

fuzzers/fuzzbench_ctx/src/lib.rs

@@ -0,0 +1,407 @@
+//! A singlethreaded libfuzzer-like fuzzer that can auto-restart.
+use mimalloc::MiMalloc;
+#[global_allocator]
+static GLOBAL: MiMalloc = MiMalloc;
+
+use core::{cell::RefCell, time::Duration};
+#[cfg(unix)]
+use std::os::unix::io::{AsRawFd, FromRawFd};
+use std::{
+    env,
+    fs::{self, File, OpenOptions},
+    io::{self, Read, Write},
+    path::PathBuf,
+    process,
+};
+
+use clap::{Arg, Command};
+use libafl::{
+    corpus::{Corpus, InMemoryOnDiskCorpus, OnDiskCorpus},
+    events::SimpleRestartingEventManager,
+    executors::{inprocess::HookableInProcessExecutor, ExitKind},
+    feedback_or,
+    feedbacks::{CrashFeedback, MaxMapFeedback, TimeFeedback},
+    fuzzer::{Fuzzer, StdFuzzer},
+    inputs::{BytesInput, HasTargetBytes},
+    monitors::SimpleMonitor,
+    mutators::{
+        scheduled::havoc_mutations, token_mutations::I2SRandReplace, tokens_mutations,
+        StdMOptMutator, StdScheduledMutator, Tokens,
+    },
+    observers::{HitcountsMapObserver, StdMapObserver, TimeObserver},
+    schedulers::{
+        powersched::PowerSchedule, IndexesLenTimeMinimizerScheduler, StdWeightedScheduler,
+    },
+    stages::{
+        calibrate::CalibrationStage, power::StdPowerMutationalStage, StdMutationalStage,
+        TracingStage,
+    },
+    state::{HasCorpus, HasMetadata, StdState},
+    Error,
+};
+use libafl_bolts::{
+    current_nanos, current_time,
+    os::dup2,
+    ownedref::OwnedMutSlice,
+    rands::StdRand,
+    shmem::{ShMemProvider, StdShMemProvider},
+    tuples::{tuple_list, Merge},
+    AsSlice,
+};
+#[cfg(any(target_os = "linux", target_vendor = "apple"))]
+use libafl_targets::autotokens;
+use libafl_targets::{
+    edges_map_mut_ptr, edges_max_num, libfuzzer_initialize, libfuzzer_test_one_input,
+    std_edges_map_observer, CmpLogObserver, CtxHook, EDGES_MAP_SIZE,
+};
+#[cfg(unix)]
+use nix::{self, unistd::dup};
+
+/// The fuzzer main (as `no_mangle` C function)
+#[no_mangle]
+pub extern "C" fn libafl_main() {
+    // Registry the metadata types used in this fuzzer
+    // Needed only on no_std
+    // unsafe { RegistryBuilder::register::<Tokens>(); }
+
+    let res = match Command::new(env!("CARGO_PKG_NAME"))
+        .version(env!("CARGO_PKG_VERSION"))
+        .author("AFLplusplus team")
+        .about("LibAFL-based fuzzer for Fuzzbench")
+        .arg(
+            Arg::new("out")
+                .short('o')
+                .long("output")
+                .help("The directory to place finds in ('corpus')"),
+        )
+        .arg(
+            Arg::new("in")
+                .short('i')
+                .long("input")
+                .help("The directory to read initial inputs from ('seeds')"),
+        )
+        .arg(
+            Arg::new("tokens")
+                .short('x')
+                .long("tokens")
+                .help("A file to read tokens from, to be used during fuzzing"),
+        )
+        .arg(
+            Arg::new("logfile")
+                .short('l')
+                .long("logfile")
+                .help("Duplicates all output to this file")
+                .default_value("libafl.log"),
+        )
+        .arg(
+            Arg::new("timeout")
+                .short('t')
+                .long("timeout")
+                .help("Timeout for each individual execution, in milliseconds")
+                .default_value("1200"),
+        )
+        .arg(Arg::new("remaining"))
+        .try_get_matches()
+    {
+        Ok(res) => res,
+        Err(err) => {
+            println!(
+                "Syntax: {}, [-x dictionary] -o corpus_dir -i seed_dir\n{:?}",
+                env::current_exe()
+                    .unwrap_or_else(|_| "fuzzer".into())
+                    .to_string_lossy(),
+                err,
+            );
+            return;
+        }
+    };
+
+    println!(
+        "Workdir: {:?}",
+        env::current_dir().unwrap().to_string_lossy().to_string()
+    );
+
+    if let Some(filenames) = res.get_many::<String>("remaining") {
+        let filenames: Vec<&str> = filenames.map(String::as_str).collect();
+        if !filenames.is_empty() {
+            run_testcases(&filenames);
+            return;
+        }
+    }
+
+    // For fuzzbench, crashes and finds are inside the same `corpus` directory, in the "queue" and "crashes" subdir.
+    let mut out_dir = PathBuf::from(
+        res.get_one::<String>("out")
+            .expect("The --output parameter is missing")
+            .to_string(),
+    );
+    if fs::create_dir(&out_dir).is_err() {
+        println!("Out dir at {:?} already exists.", &out_dir);
+        if !out_dir.is_dir() {
+            println!("Out dir at {:?} is not a valid directory!", &out_dir);
+            return;
+        }
+    }
+    let mut crashes = out_dir.clone();
+    crashes.push("crashes");
+    out_dir.push("queue");
+
+    let in_dir = PathBuf::from(
+        res.get_one::<String>("in")
+            .expect("The --input parameter is missing")
+            .to_string(),
+    );
+    if !in_dir.is_dir() {
+        println!("In dir at {:?} is not a valid directory!", &in_dir);
+        return;
+    }
+
+    let tokens = res.get_one::<String>("tokens").map(PathBuf::from);
+
+    let logfile = PathBuf::from(res.get_one::<String>("logfile").unwrap().to_string());
+
+    let timeout = Duration::from_millis(
+        res.get_one::<String>("timeout")
+            .unwrap()
+            .to_string()
+            .parse()
+            .expect("Could not parse timeout in milliseconds"),
+    );
+
+    fuzz(out_dir, crashes, &in_dir, tokens, &logfile, timeout)
+        .expect("An error occurred while fuzzing");
+}
+
+fn run_testcases(filenames: &[&str]) {
+    // The actual target run starts here.
+    // Call LLVMFUzzerInitialize() if present.
+    let args: Vec<String> = env::args().collect();
+    if libfuzzer_initialize(&args) == -1 {
+        println!("Warning: LLVMFuzzerInitialize failed with -1");
+    }
+
+    println!(
+        "You are not fuzzing, just executing {} testcases",
+        filenames.len()
+    );
+    for fname in filenames {
+        println!("Executing {fname}");
+
+        let mut file = File::open(fname).expect("No file found");
+        let mut buffer = vec![];
+        file.read_to_end(&mut buffer).expect("Buffer overflow");
+
+        libfuzzer_test_one_input(&buffer);
+    }
+}
+
+/// The actual fuzzer
+#[allow(clippy::too_many_lines)]
+fn fuzz(
+    corpus_dir: PathBuf,
+    objective_dir: PathBuf,
+    seed_dir: &PathBuf,
+    tokenfile: Option<PathBuf>,
+    logfile: &PathBuf,
+    timeout: Duration,
+) -> Result<(), Error> {
+    let log = RefCell::new(OpenOptions::new().append(true).create(true).open(logfile)?);
+
+    #[cfg(unix)]
+    let mut stdout_cpy = unsafe {
+        let new_fd = dup(io::stdout().as_raw_fd())?;
+        File::from_raw_fd(new_fd)
+    };
+    #[cfg(unix)]
+    let file_null = File::open("/dev/null")?;
+
+    // 'While the monitor are state, they are usually used in the broker - which is likely never restarted
+    let monitor = SimpleMonitor::new(|s| {
+        #[cfg(unix)]
+        writeln!(&mut stdout_cpy, "{s}").unwrap();
+        #[cfg(windows)]
+        println!("{s}");
+        writeln!(log.borrow_mut(), "{:?} {s}", current_time()).unwrap();
+    });
+
+    // We need a shared map to store our state before a crash.
+    // This way, we are able to continue fuzzing afterwards.
+    let mut shmem_provider = StdShMemProvider::new()?;
+
+    let (state, mut mgr) = match SimpleRestartingEventManager::launch(monitor, &mut shmem_provider)
+    {
+        // The restarting state will spawn the same process again as child, then restarted it each time it crashes.
+        Ok(res) => res,
+        Err(err) => match err {
+            Error::ShuttingDown => {
+                return Ok(());
+            }
+            _ => {
+                panic!("Failed to setup the restarter: {err}");
+            }
+        },
+    };
+
+    // Create an observation channel using the coverage map
+    // We don't use the hitcounts (see the Cargo.toml, we use pcguard_edges)
+    let edges_observer = HitcountsMapObserver::new(unsafe {
+        StdMapObserver::from_mut_slice(
+            "edges",
+            OwnedMutSlice::from_raw_parts_mut(edges_map_mut_ptr(), EDGES_MAP_SIZE),
+        )
+    });
+
+    // Create an observation channel to keep track of the execution time
+    let time_observer = TimeObserver::new("time");
+
+    let cmplog_observer = CmpLogObserver::new("cmplog", true);
+
+    let map_feedback = MaxMapFeedback::tracking(&edges_observer, true, false);
+
+    let calibration = CalibrationStage::new(&map_feedback);
+
+    // Feedback to rate the interestingness of an input
+    // This one is composed by two Feedbacks in OR
+    let mut feedback = feedback_or!(
+        // New maximization map feedback linked to the edges observer and the feedback state
+        map_feedback,
+        // Time feedback, this one does not need a feedback state
+        TimeFeedback::with_observer(&time_observer)
+    );
+
+    // A feedback to choose if an input is a solution or not
+    let mut objective = CrashFeedback::new();
+
+    // If not restarting, create a State from scratch
+    let mut state = state.unwrap_or_else(|| {
+        StdState::new(
+            // RNG
+            StdRand::with_seed(current_nanos()),
+            // Corpus that will be evolved, we keep it in memory for performance
+            InMemoryOnDiskCorpus::new(corpus_dir).unwrap(),
+            // Corpus in which we store solutions (crashes in this example),
+            // on disk so the user can get them after stopping the fuzzer
+            OnDiskCorpus::new(objective_dir).unwrap(),
+            // States of the feedbacks.
+            // The feedbacks can report the data that should persist in the State.
+            &mut feedback,
+            // Same for objective feedbacks
+            &mut objective,
+        )
+        .unwrap()
+    });
+
+    println!("Let's fuzz :)");
+
+    // The actual target run starts here.
+    // Call LLVMFUzzerInitialize() if present.
+    let args: Vec<String> = env::args().collect();
+    if libfuzzer_initialize(&args) == -1 {
+        println!("Warning: LLVMFuzzerInitialize failed with -1");
+    }
+
+    // Setup a randomic Input2State stage
+    let i2s = StdMutationalStage::new(StdScheduledMutator::new(tuple_list!(I2SRandReplace::new())));
+
+    // Setup a MOPT mutator
+    let mutator = StdMOptMutator::new(
+        &mut state,
+        havoc_mutations().merge(tokens_mutations()),
+        7,
+        5,
+    )?;
+
+    let power = StdPowerMutationalStage::new(mutator);
+
+    // A minimization+queue policy to get testcasess from the corpus
+    let scheduler = IndexesLenTimeMinimizerScheduler::new(StdWeightedScheduler::with_schedule(
+        &mut state,
+        &edges_observer,
+        Some(PowerSchedule::FAST),
+    ));
+
+    // A fuzzer with feedbacks and a corpus scheduler
+    let mut fuzzer = StdFuzzer::new(scheduler, feedback, objective);
+
+    // The wrapped harness function, calling out to the LLVM-style harness
+    let mut harness = |input: &BytesInput| {
+        let target = input.target_bytes();
+        let buf = target.as_slice();
+        libfuzzer_test_one_input(buf);
+        ExitKind::Ok
+    };
+
+    let mut tracing_harness = harness;
+    let ctx_hook = CtxHook::default();
+    // Create the executor for an in-process function with one observer for edge coverage and one for the execution time
+    let mut executor = HookableInProcessExecutor::with_timeout_generic(
+        tuple_list!(ctx_hook),
+        &mut harness,
+        tuple_list!(edges_observer, time_observer),
+        &mut fuzzer,
+        &mut state,
+        &mut mgr,
+        timeout,
+    )?;
+
+    // Setup a tracing stage in which we log comparisons
+    let tracing = TracingStage::new(
+        HookableInProcessExecutor::with_timeout_generic(
+            tuple_list!(ctx_hook),
+            &mut tracing_harness,
+            tuple_list!(cmplog_observer),
+            &mut fuzzer,
+            &mut state,
+            &mut mgr,
+            timeout * 10,
+        )?,
+        // Give it more time!
+    );
+
+    // The order of the stages matter!
+    let mut stages = tuple_list!(calibration, tracing, i2s, power);
+
+    // Read tokens
+    if state.metadata_map().get::<Tokens>().is_none() {
+        let mut toks = Tokens::default();
+        if let Some(tokenfile) = tokenfile {
+            toks.add_from_file(tokenfile)?;
+        }
+        #[cfg(any(target_os = "linux", target_vendor = "apple"))]
+        {
+            toks += autotokens()?;
+        }
+
+        if !toks.is_empty() {
+            state.add_metadata(toks);
+        }
+    }
+
+    // In case the corpus is empty (on first run), reset
+    if state.must_load_initial_inputs() {
+        state
+            .load_initial_inputs(&mut fuzzer, &mut executor, &mut mgr, &[seed_dir.clone()])
+            .unwrap_or_else(|_| {
+                println!("Failed to load initial corpus at {:?}", &seed_dir);
+                process::exit(0);
+            });
+        println!("We imported {} inputs from disk.", state.corpus().count());
+    }
+
+    // Remove target output (logs still survive)
+    #[cfg(unix)]
+    {
+        let null_fd = file_null.as_raw_fd();
+        // dup2(null_fd, io::stdout().as_raw_fd())?;
+        if std::env::var("LIBAFL_FUZZBENCH_DEBUG").is_err() {
+            dup2(null_fd, io::stderr().as_raw_fd())?;
+        }
+    }
+    // reopen file to make sure we're at the end
+    log.replace(OpenOptions::new().append(true).create(true).open(logfile)?);
+
+    fuzzer.fuzz_loop(&mut stages, &mut executor, &mut state, &mut mgr)?;
+
+    // Never reached
+    Ok(())
+}

fuzzers/fuzzbench_ctx/stub_rt.c

@@ -0,0 +1,34 @@
+#include <stdint.h>
+
+__attribute__((weak)) void __sanitizer_cov_trace_pc_guard_init(uint32_t *start,
+                                                               uint32_t *stop) {
+}
+
+__attribute__((weak)) void __sanitizer_cov_trace_pc_guard(uint32_t *guard) {
+}
+
+__attribute__((weak)) void __cmplog_rtn_hook(uint8_t *ptr1, uint8_t *ptr2) {
+}
+
+__attribute__((weak)) void __cmplog_rtn_gcc_stdstring_cstring(
+    uint8_t *stdstring, uint8_t *cstring) {
+}
+
+__attribute__((weak)) void __cmplog_rtn_gcc_stdstring_stdstring(
+    uint8_t *stdstring1, uint8_t *stdstring2) {
+}
+
+__attribute__((weak)) void __cmplog_rtn_llvm_stdstring_cstring(
+    uint8_t *stdstring, uint8_t *cstring) {
+}
+
+__attribute__((weak)) void __cmplog_rtn_llvm_stdstring_stdstring(
+    uint8_t *stdstring1, uint8_t *stdstring2) {
+}
+
+extern void libafl_main(void);
+
+int main(int argc, char **argv) {
+  libafl_main();
+  return 0;
+}

fuzzers/libfuzzer_libpng_ctx/.gitignore

@@ -1 +0,0 @@
-libpng-*

fuzzers/libfuzzer_libpng_ctx/Makefile.toml

@@ -1,134 +0,0 @@
-# Variables
-[env]
-FUZZER_NAME='fuzzer_libpng_ctx'
-PROJECT_DIR = { script = ["pwd"] }
-CARGO_TARGET_DIR = { value = "${PROJECT_DIR}/target", condition = { env_not_set = ["CARGO_TARGET_DIR"] } }
-PROFILE = { value = "release", condition = {env_not_set = ["PROFILE"]} }
-PROFILE_DIR = {value = "release", condition = {env_not_set = ["PROFILE_DIR"] }}
-LIBAFL_CC = '${CARGO_TARGET_DIR}/${PROFILE_DIR}/libafl_cc'
-LIBAFL_CXX = '${CARGO_TARGET_DIR}/${PROFILE}/libafl_cxx'
-FUZZER = '${CARGO_TARGET_DIR}/${PROFILE_DIR}/${FUZZER_NAME}'
-
-[tasks.unsupported]
-script_runner="@shell"
-script='''
-echo "Cargo-make not integrated yet on this platform"
-'''
-
-# libpng
-[tasks.libpng]
-linux_alias = "libpng_unix"
-mac_alias = "libpng_unix"
-windows_alias = "unsupported"
-
-[tasks.libpng_unix]
-condition = { files_not_exist = ["./libpng-1.6.37"]}
-script_runner="@shell"
-script='''
-wget https://github.com/glennrp/libpng/archive/refs/tags/v1.6.37.tar.gz
-tar -xvf v1.6.37.tar.gz
-'''
-
-# Compilers
-[tasks.cxx]
-linux_alias = "cxx_unix"
-mac_alias = "cxx_unix"
-windows_alias = "unsupported"
-
-[tasks.cxx_unix]
-command = "cargo"
-args = ["build" , "--profile", "${PROFILE}"]
-
-[tasks.cc]
-linux_alias = "cc_unix"
-mac_alias = "cc_unix"
-windows_alias = "unsupported"
-
-[tasks.cc_unix]
-command = "cargo"
-args = ["build" , "--profile", "${PROFILE}"]
-
-# Library
-[tasks.lib]
-linux_alias = "lib_unix"
-mac_alias = "lib_unix"
-windows_alias = "unsupported"
-
-[tasks.lib_unix]
-script_runner="@shell"
-script='''
-cd libpng-1.6.37 && ./configure --enable-shared=no --with-pic=yes --enable-hardware-optimizations=yes
-cd "${PROJECT_DIR}"
-make -C libpng-1.6.37 CC="${CARGO_TARGET_DIR}/${PROFILE_DIR}/libafl_cc" CXX="${CARGO_TARGET_DIR}/${PROFILE_DIR}/libafl_cxx"
-'''
-dependencies = [ "libpng", "cxx", "cc" ]
-
-
-# Harness
-[tasks.fuzzer]
-linux_alias = "fuzzer_unix"
-mac_alias = "fuzzer_unix"
-windows_alias = "unsupported"
-
-[tasks.fuzzer_unix]
-command = "${CARGO_TARGET_DIR}/${PROFILE_DIR}/libafl_cxx"
-args = ["${PROJECT_DIR}/harness.cc", "${PROJECT_DIR}/libpng-1.6.37/.libs/libpng16.a", "-I", "${PROJECT_DIR}/libpng-1.6.37/", "-o", "${FUZZER_NAME}", "-lm", "-lz"]
-dependencies = [ "lib", "cxx", "cc" ]
-
-# Run the fuzzer
-[tasks.run]
-linux_alias = "run_unix"
-mac_alias = "run_unix"
-windows_alias = "unsupported"
-
-[tasks.run_unix]
-script_runner = "@shell"
-script='''
-./${FUZZER_NAME} --cores 0 --input ./corpus
-'''
-dependencies = [ "fuzzer" ]
-
-# Test
-[tasks.test]
-linux_alias = "test_unix"
-mac_alias = "test_mac"
-windows_alias = "unsupported"
-
-[tasks.test_unix]
-script_runner = "@shell"
-script='''
-rm -rf libafl_unix_shmem_server || true
-timeout 31s ./${FUZZER_NAME} --cores 0 --input ./corpus >fuzz_stdout.log 2>/dev/null || true
-if grep -qa "corpus: 30" fuzz_stdout.log; then
-    echo "Fuzzer is working"
-elif grep -qa "objectives: 1" fuzz_stdout.log; then
-    echo "Fuzzer finds timeout or crash"
-else
-    echo "Fuzzer does not generate any testcases or any crashes"
-    exit 1
-fi
-'''
-dependencies = [ "fuzzer" ]
-
-[tasks.test_mac]
-script_runner = "@shell"
-script='''
-rm -rf libafl_unix_shmem_server || true
-timeout 31s ./${FUZZER_NAME} --cores 0 --input ./corpus >fuzz_stdout.log 2>/dev/null || true
-'''
-
-# Clean up
-[tasks.clean]
-linux_alias = "clean_unix"
-mac_alias = "clean_unix"
-windows_alias = "unsupported"
-
-[tasks.clean_unix]
-# Disable default `clean` definition
-clear = true
-script_runner="@shell"
-script='''
-rm -f ./${FUZZER_NAME}
-make -C libpng-1.6.37 clean
-cargo clean
-'''

fuzzers/libfuzzer_libpng_ctx/README.md

@@ -1,47 +0,0 @@
-# Libfuzzer for libpng, with launcher
-
-This folder contains an example fuzzer for libpng, using LLMP for fast multi-process fuzzing and crash detection.
-To show off crash detection, we added a `ud2` instruction to the harness, edit harness.cc if you want a non-crashing example.
-It has been tested on Linux.
-
-In contrast to the normal libfuzzer libpng example, this uses the `launcher` feature, that automatically spawns `n` child processes, and binds them to a free core.
-
-## Build
-
-To build this example, run
-
-```bash
-cargo build --release
-```
-
-This will build the library with the fuzzer (src/lib.rs) with the libfuzzer compatibility layer and the SanitizerCoverage runtime functions for coverage feedback.
-In addition, it will also build two C and C++ compiler wrappers (bin/libafl_c(libafl_c/xx).rs) that you must use to compile the target.
-
-Then download libpng, and unpack the archive:
-```bash
-wget https://github.com/glennrp/libpng/archive/refs/tags/v1.6.37.tar.gz
-tar -xvf v1.6.37.tar.gz
-```
-
-Now compile libpng, using the libafl_cc compiler wrapper:
-
-```bash
-cd libpng-1.6.37
-./configure
-make CC=../target/release/libafl_cc CXX=../target/release/libafl_cxx -j `nproc`
-```
-
-You can find the static lib at `libpng-1.6.37/.libs/libpng16.a`.
-
-Now, we have to build the libfuzzer harness and link all together to create our fuzzer binary.
-
-```
-cd ..
-./target/release/libafl_cxx ./harness.cc libpng-1.6.37/.libs/libpng16.a -I libpng-1.6.37/ -o fuzzer_libpng -lz -lm
-```
-
-Afterwards, the fuzzer will be ready to run.
-
-## Run
-
-Just run once, the launcher feature should do the rest.

fuzzers/libfuzzer_libpng_ctx/harness.cc

@@ -1,188 +0,0 @@
-// libpng_read_fuzzer.cc
-// Copyright 2017-2018 Glenn Randers-Pehrson
-// Copyright 2015 The Chromium Authors. All rights reserved.
-// Use of this source code is governed by a BSD-style license that may
-// be found in the LICENSE file https://cs.chromium.org/chromium/src/LICENSE
-
-// Last changed in libpng 1.6.35 [July 15, 2018]
-
-// The modifications in 2017 by Glenn Randers-Pehrson include
-// 1. addition of a PNG_CLEANUP macro,
-// 2. setting the option to ignore ADLER32 checksums,
-// 3. adding "#include <string.h>" which is needed on some platforms
-//    to provide memcpy().
-// 4. adding read_end_info() and creating an end_info structure.
-// 5. adding calls to png_set_*() transforms commonly used by browsers.
-
-#include <stddef.h>
-#include <stdint.h>
-#include <string.h>
-
-#include <vector>
-
-#define PNG_INTERNAL
-#include "png.h"
-
-#define PNG_CLEANUP                                                        \
-  if (png_handler.png_ptr) {                                               \
-    if (png_handler.row_ptr)                                               \
-      png_free(png_handler.png_ptr, png_handler.row_ptr);                  \
-    if (png_handler.end_info_ptr)                                          \
-      png_destroy_read_struct(&png_handler.png_ptr, &png_handler.info_ptr, \
-                              &png_handler.end_info_ptr);                  \
-    else if (png_handler.info_ptr)                                         \
-      png_destroy_read_struct(&png_handler.png_ptr, &png_handler.info_ptr, \
-                              nullptr);                                    \
-    else                                                                   \
-      png_destroy_read_struct(&png_handler.png_ptr, nullptr, nullptr);     \
-    png_handler.png_ptr = nullptr;                                         \
-    png_handler.row_ptr = nullptr;                                         \
-    png_handler.info_ptr = nullptr;                                        \
-    png_handler.end_info_ptr = nullptr;                                    \
-  }
-
-struct BufState {
-  const uint8_t *data;
-  size_t         bytes_left;
-};
-
-struct PngObjectHandler {
-  png_infop   info_ptr = nullptr;
-  png_structp png_ptr = nullptr;
-  png_infop   end_info_ptr = nullptr;
-  png_voidp   row_ptr = nullptr;
-  BufState   *buf_state = nullptr;
-
-  ~PngObjectHandler() {
-    if (row_ptr) { png_free(png_ptr, row_ptr); }
-    if (end_info_ptr)
-      png_destroy_read_struct(&png_ptr, &info_ptr, &end_info_ptr);
-    else if (info_ptr)
-      png_destroy_read_struct(&png_ptr, &info_ptr, nullptr);
-    else
-      png_destroy_read_struct(&png_ptr, nullptr, nullptr);
-    delete buf_state;
-  }
-};
-
-void user_read_data(png_structp png_ptr, png_bytep data, size_t length) {
-  BufState *buf_state = static_cast<BufState *>(png_get_io_ptr(png_ptr));
-  if (length > buf_state->bytes_left) { png_error(png_ptr, "read error"); }
-  memcpy(data, buf_state->data, length);
-  buf_state->bytes_left -= length;
-  buf_state->data += length;
-}
-
-static const int kPngHeaderSize = 8;
-
-// Entry point for LibFuzzer.
-// Roughly follows the libpng book example:
-// http://www.libpng.org/pub/png/book/chapter13.html
-extern "C" int LLVMFuzzerTestOneInput(const uint8_t *data, size_t size) {
-  if (size < kPngHeaderSize) { return 0; }
-
-  std::vector<unsigned char> v(data, data + size);
-  if (png_sig_cmp(v.data(), 0, kPngHeaderSize)) {
-    // not a PNG.
-    return 0;
-  }
-
-  PngObjectHandler png_handler;
-  png_handler.png_ptr = nullptr;
-  png_handler.row_ptr = nullptr;
-  png_handler.info_ptr = nullptr;
-  png_handler.end_info_ptr = nullptr;
-
-  png_handler.png_ptr =
-      png_create_read_struct(PNG_LIBPNG_VER_STRING, nullptr, nullptr, nullptr);
-  if (!png_handler.png_ptr) { return 0; }
-
-  png_handler.info_ptr = png_create_info_struct(png_handler.png_ptr);
-  if (!png_handler.info_ptr) {
-    PNG_CLEANUP
-    return 0;
-  }
-
-  png_handler.end_info_ptr = png_create_info_struct(png_handler.png_ptr);
-  if (!png_handler.end_info_ptr) {
-    PNG_CLEANUP
-    return 0;
-  }
-
-  png_set_crc_action(png_handler.png_ptr, PNG_CRC_QUIET_USE, PNG_CRC_QUIET_USE);
-#ifdef PNG_IGNORE_ADLER32
-  png_set_option(png_handler.png_ptr, PNG_IGNORE_ADLER32, PNG_OPTION_ON);
-#endif
-
-  // Setting up reading from buffer.
-  png_handler.buf_state = new BufState();
-  png_handler.buf_state->data = data + kPngHeaderSize;
-  png_handler.buf_state->bytes_left = size - kPngHeaderSize;
-  png_set_read_fn(png_handler.png_ptr, png_handler.buf_state, user_read_data);
-  png_set_sig_bytes(png_handler.png_ptr, kPngHeaderSize);
-
-  if (setjmp(png_jmpbuf(png_handler.png_ptr))) {
-    PNG_CLEANUP
-    return 0;
-  }
-
-  // Reading.
-  png_read_info(png_handler.png_ptr, png_handler.info_ptr);
-
-  // reset error handler to put png_deleter into scope.
-  if (setjmp(png_jmpbuf(png_handler.png_ptr))) {
-    PNG_CLEANUP
-    return 0;
-  }
-
-  png_uint_32 width, height;
-  int         bit_depth, color_type, interlace_type, compression_type;
-  int         filter_type;
-
-  if (!png_get_IHDR(png_handler.png_ptr, png_handler.info_ptr, &width, &height,
-                    &bit_depth, &color_type, &interlace_type, &compression_type,
-                    &filter_type)) {
-    PNG_CLEANUP
-    return 0;
-  }
-
-  // This is going to be too slow.
-  if (width && height > 100000000 / width) {
-    PNG_CLEANUP
-#ifdef HAS_DUMMY_CRASH
-  #ifdef __aarch64__
-    asm volatile(".word 0xf7f0a000\n");
-  #else
-    asm("ud2");
-  #endif
-#endif
-    return 0;
-  }
-
-  // Set several transforms that browsers typically use:
-  png_set_gray_to_rgb(png_handler.png_ptr);
-  png_set_expand(png_handler.png_ptr);
-  png_set_packing(png_handler.png_ptr);
-  png_set_scale_16(png_handler.png_ptr);
-  png_set_tRNS_to_alpha(png_handler.png_ptr);
-
-  int passes = png_set_interlace_handling(png_handler.png_ptr);
-
-  png_read_update_info(png_handler.png_ptr, png_handler.info_ptr);
-
-  png_handler.row_ptr =
-      png_malloc(png_handler.png_ptr,
-                 png_get_rowbytes(png_handler.png_ptr, png_handler.info_ptr));
-
-  for (int pass = 0; pass < passes; ++pass) {
-    for (png_uint_32 y = 0; y < height; ++y) {
-      png_read_row(png_handler.png_ptr,
-                   static_cast<png_bytep>(png_handler.row_ptr), nullptr);
-    }
-  }
-
-  png_read_end(png_handler.png_ptr, png_handler.end_info_ptr);
-
-  PNG_CLEANUP
-  return 0;
-}

fuzzers/libfuzzer_libpng_ctx/src/lib.rs

@@ -1,243 +0,0 @@
-//! A libfuzzer-like fuzzer with llmp-multithreading support and restarts
-//! The example harness is built for libpng.
-//! In this example, you will see the use of the `launcher` feature.
-//! The `launcher` will spawn new processes for each cpu core.
-use mimalloc::MiMalloc;
-#[global_allocator]
-static GLOBAL: MiMalloc = MiMalloc;
-
-use core::time::Duration;
-use std::{env, net::SocketAddr, path::PathBuf};
-
-use clap::{self, Parser};
-use libafl::{
-    corpus::{Corpus, InMemoryCorpus, OnDiskCorpus},
-    events::{launcher::Launcher, EventConfig},
-    executors::{inprocess::InProcessExecutor, ExitKind},
-    feedback_or, feedback_or_fast,
-    feedbacks::{CrashFeedback, MaxMapFeedback, TimeFeedback, TimeoutFeedback},
-    fuzzer::{Fuzzer, StdFuzzer},
-    inputs::{BytesInput, HasTargetBytes},
-    monitors::MultiMonitor,
-    mutators::{
-        scheduled::{havoc_mutations, tokens_mutations, StdScheduledMutator},
-        token_mutations::Tokens,
-    },
-    observers::{HitcountsMapObserver, StdMapObserver, TimeObserver},
-    schedulers::{IndexesLenTimeMinimizerScheduler, QueueScheduler},
-    stages::mutational::StdMutationalStage,
-    state::{HasCorpus, HasMetadata, StdState},
-    Error,
-};
-use libafl_bolts::{
-    core_affinity::Cores,
-    current_nanos,
-    rands::StdRand,
-    shmem::{ShMemProvider, StdShMemProvider},
-    tuples::{tuple_list, Merge},
-    AsSlice,
-};
-use libafl_targets::{libfuzzer_initialize, libfuzzer_test_one_input, std_edges_map_observer};
-
-fn timeout_from_millis_str(time: &str) -> Result<Duration, Error> {
-    Ok(Duration::from_millis(time.parse()?))
-}
-
-#[derive(Debug, Parser)]
-#[command(
-    name = "libfuzzer_libpng_ctx",
-    about = "A clone of libfuzzer using LibAFL for a libpng harness",
-    author = "Andrea Fioraldi <andreafioraldi@gmail.com>, Dominik Maier <domenukk@gmail.com>"
-)]
-struct Opt {
-    #[arg(
-        short,
-        long,
-        value_parser = Cores::from_cmdline,
-        help = "Spawn a client in each of the provided cores. Broker runs in the 0th core. 'all' to select all available cores. 'none' to run a client without binding to any core. eg: '1,2-4,6' selects the cores 1,2,3,4,6.",
-        name = "CORES"
-    )]
-    cores: Cores,
-
-    #[arg(
-        short = 'p',
-        long,
-        help = "Choose the broker TCP port, default is 1337",
-        name = "PORT",
-        default_value = "1337"
-    )]
-    broker_port: u16,
-
-    #[arg(short = 'a', long, help = "Specify a remote broker", name = "REMOTE")]
-    remote_broker_addr: Option<SocketAddr>,
-
-    #[arg(short, long, help = "Set an initial corpus directory", name = "INPUT")]
-    input: Vec<PathBuf>,
-
-    #[arg(
-        short,
-        long,
-        help = "Set the output directory, default is ./out",
-        name = "OUTPUT",
-        default_value = "./out"
-    )]
-    output: PathBuf,
-
-    #[arg(
-        short,
-        long,
-        value_parser = timeout_from_millis_str,
-        help = "Set the exeucution timeout in milliseconds, default is 10000",
-        name = "TIMEOUT",
-        default_value = "10000",
-    )]
-    timeout: Duration,
-    /*
-        // The tokens are hardcoded in this example.
-        #[arg(
-
-        short = "x",
-        long,
-        help = "Feed the fuzzer with an user-specified list of tokens (often called \"dictionary\"",
-        name = "TOKENS",
-        multiple = true
-    )]
-    tokens: Vec<PathBuf>,*/
-}
-
-/// The main fn, `no_mangle` as it is a C symbol
-#[no_mangle]
-pub extern "C" fn libafl_main() {
-    // Registry the metadata types used in this fuzzer
-    // Needed only on no_std
-    // unsafe { RegistryBuilder::register::<Tokens>(); }
-    let opt = Opt::parse();
-
-    let broker_port = opt.broker_port;
-
-    let cores = opt.cores;
-
-    println!(
-        "Workdir: {:?}",
-        env::current_dir().unwrap().to_string_lossy().to_string()
-    );
-
-    let shmem_provider = StdShMemProvider::new().expect("Failed to init shared memory");
-
-    let monitor = MultiMonitor::new(|s| println!("{s}"));
-
-    let mut run_client = |state: Option<_>, mut restarting_mgr, _core_id| {
-        // Create an observation channel using the coverage map
-        let edges_observer = unsafe { HitcountsMapObserver::new(std_edges_map_observer("edges")) };
-
-        // Create an observation channel to keep track of the execution time
-        let time_observer = TimeObserver::new("time");
-
-        // Feedback to rate the interestingness of an input
-        // This one is composed by two Feedbacks in OR
-        let mut feedback = feedback_or!(
-            // New maximization map feedback linked to the edges observer and the feedback state
-            MaxMapFeedback::tracking(&edges_observer, true, false),
-            // Time feedback, this one does not need a feedback state
-            TimeFeedback::with_observer(&time_observer)
-        );
-
-        // A feedback to choose if an input is a solution or not
-        let mut objective = feedback_or_fast!(CrashFeedback::new(), TimeoutFeedback::new());
-
-        // If not restarting, create a State from scratch
-        let mut state = state.unwrap_or_else(|| {
-            StdState::new(
-                // RNG
-                StdRand::with_seed(current_nanos()),
-                // Corpus that will be evolved, we keep it in memory for performance
-                InMemoryCorpus::new(),
-                // Corpus in which we store solutions (crashes in this example),
-                // on disk so the user can get them after stopping the fuzzer
-                OnDiskCorpus::new(opt.output.clone()).unwrap(),
-                // States of the feedbacks.
-                // The feedbacks can report the data that should persist in the State.
-                &mut feedback,
-                // Same for objective feedbacks
-                &mut objective,
-            )
-            .unwrap()
-        });
-
-        println!("We're a client, let's fuzz :)");
-
-        // Create a PNG dictionary if not existing
-        if state.metadata_map().get::<Tokens>().is_none() {
-            state.add_metadata(Tokens::from([
-                vec![137, 80, 78, 71, 13, 10, 26, 10], // PNG header
-                "IHDR".as_bytes().to_vec(),
-                "IDAT".as_bytes().to_vec(),
-                "PLTE".as_bytes().to_vec(),
-                "IEND".as_bytes().to_vec(),
-            ]));
-        }
-
-        // Setup a basic mutator with a mutational stage
-        let mutator = StdScheduledMutator::new(havoc_mutations().merge(tokens_mutations()));
-        let mut stages = tuple_list!(StdMutationalStage::new(mutator));
-
-        // A minimization+queue policy to get testcasess from the corpus
-        let scheduler = IndexesLenTimeMinimizerScheduler::new(QueueScheduler::new());
-
-        // A fuzzer with feedbacks and a corpus scheduler
-        let mut fuzzer = StdFuzzer::new(scheduler, feedback, objective);
-
-        // The wrapped harness function, calling out to the LLVM-style harness
-        let mut harness = |input: &BytesInput| {
-            let target = input.target_bytes();
-            let buf = target.as_slice();
-            libfuzzer_test_one_input(buf);
-            ExitKind::Ok
-        };
-
-        // Create the executor for an in-process function with one observer for edge coverage and one for the execution time
-        let mut executor = InProcessExecutor::with_timeout(
-            &mut harness,
-            tuple_list!(edges_observer, time_observer),
-            &mut fuzzer,
-            &mut state,
-            &mut restarting_mgr,
-            opt.timeout,
-        )?;
-
-        // The actual target run starts here.
-        // Call LLVMFUzzerInitialize() if present.
-        let args: Vec<String> = env::args().collect();
-        if libfuzzer_initialize(&args) == -1 {
-            println!("Warning: LLVMFuzzerInitialize failed with -1");
-        }
-
-        // In case the corpus is empty (on first run), reset
-        if state.must_load_initial_inputs() {
-            state
-                .load_initial_inputs(&mut fuzzer, &mut executor, &mut restarting_mgr, &opt.input)
-                .unwrap_or_else(|_| panic!("Failed to load initial corpus at {:?}", &opt.input));
-            println!("We imported {} inputs from disk.", state.corpus().count());
-        }
-
-        fuzzer.fuzz_loop(&mut stages, &mut executor, &mut state, &mut restarting_mgr)?;
-        Ok(())
-    };
-
-    match Launcher::builder()
-        .shmem_provider(shmem_provider)
-        .configuration(EventConfig::from_name("default"))
-        .monitor(monitor)
-        .run_client(&mut run_client)
-        .cores(&cores)
-        .broker_port(broker_port)
-        .remote_broker_addr(opt.remote_broker_addr)
-        .stdout_file(Some("/dev/null"))
-        .build()
-        .launch()
-    {
-        Ok(()) => (),
-        Err(Error::ShuttingDown) => println!("Fuzzing stopped by user. Good bye."),
-        Err(err) => panic!("Failed to run launcher: {err:?}"),
-    }
-}

libafl/src/executors/inprocess.rs

@@ -44,6 +44,9 @@ use crate::{
 /// The process executor simply calls a target function, as mutable reference to a closure
 pub type InProcessExecutor<'a, H, OT, S> = GenericInProcessExecutor<H, &'a mut H, (), OT, S>;
 
+/// The inprocess executor that allows hooks
+pub type HookableInProcessExecutor<'a, H, HT, OT, S> =
+    GenericInProcessExecutor<H, &'a mut H, HT, OT, S>;
 /// The process executor simply calls a target function, as boxed `FnMut` trait object
 pub type OwnedInProcessExecutor<OT, S> = GenericInProcessExecutor<
     dyn FnMut(&<S as UsesInput>::Input) -> ExitKind,

libafl/src/stages/calibrate.rs

@@ -254,6 +254,7 @@ where
                 .ok_or_else(|| Error::key_not_found("MapObserver not found".to_string()))?;
 
             let mut bitmap_size = map.count_bytes();
+            assert!(bitmap_size != 0);
             bitmap_size = bitmap_size.max(1); // just don't make it 0 because we take log2 of it later.
             let psmeta = state
                 .metadata_map_mut()

libafl_cc/build.rs

@@ -404,10 +404,10 @@ pub const LIBAFL_CC_LLVM_VERSION: Option<usize> = None;
 
     for pass in &[
         "cmplog-routines-pass.cc",
-        "afl-coverage-pass.cc",
         "autotokens-pass.cc",
         "coverage-accounting-pass.cc",
         "cmplog-instructions-pass.cc",
+        "ctx-pass.cc",
     ] {
         build_pass(
             bindir_path,

libafl_cc/src/afl-coverage-pass.cc

@@ -1,872 +0,0 @@
-/*
-   american fuzzy lop++ - LLVM-mode instrumentation pass
-   ---------------------------------------------------
-
-   Written by Laszlo Szekeres <lszekeres@google.com>,
-              Adrian Herrera <adrian.herrera@anu.edu.au>,
-              Michal Zalewski
-
-   LLVM integration design comes from Laszlo Szekeres. C bits copied-and-pasted
-   from afl-as.c are Michal's fault.
-
-   NGRAM previous location coverage comes from Adrian Herrera.
-
-   Copyright 2015, 2016 Google Inc. All rights reserved.
-   Copyright 2019-2020 AFLplusplus Project. All rights reserved.
-
-   Licensed under the Apache License, Version 2.0 (the "License");
-   you may not use this file except in compliance with the License.
-   You may obtain a copy of the License at:
-
-     http://www.apache.org/licenses/LICENSE-2.0
-
-   This library is plugged into LLVM when invoking clang through afl-clang-fast.
-   It tells the compiler to add code roughly equivalent to the bits discussed
-   in ../afl-as.h.
-
- */
-
-#include "common-llvm.h"
-
-#include <time.h>
-#include <stdio.h>
-#include <stdlib.h>
-#ifndef _WIN32
-  #include <unistd.h>
-  #include <sys/time.h>
-#else
-  #include <io.h>
-#endif
-#include <sys/types.h>
-#include <sys/stat.h>
-#include <fcntl.h>
-#include <ctype.h>
-
-#include <list>
-#include <string>
-#include <fstream>
-
-#include "llvm/Support/CommandLine.h"
-#include "llvm/IR/IRBuilder.h"
-#include "llvm/IR/BasicBlock.h"
-#include "llvm/IR/Module.h"
-#include "llvm/Support/Debug.h"
-#include "llvm/Support/MathExtras.h"
-#include "llvm/ADT/DenseMap.h"
-#include "llvm/Support/FormatVariadic.h"
-
-// Without this, Can't build with llvm-14 & old PM
-#if LLVM_VERSION_MAJOR >= 14 && !defined(USE_NEW_PM)
-  #include "llvm/Pass.h"
-#endif
-
-#if LLVM_VERSION_MAJOR > 3 || \
-    (LLVM_VERSION_MAJOR == 3 && LLVM_VERSION_MINOR > 4)
-  #include "llvm/IR/DebugInfo.h"
-  #include "llvm/IR/CFG.h"
-#else
-  #include "llvm/DebugInfo.h"
-  #include "llvm/Support/CFG.h"
-#endif
-
-typedef uint32_t prev_loc_t;
-
-/* Maximum ngram size */
-#define NGRAM_SIZE_MAX 16U
-
-/* Maximum K for top-K context sensitivity */
-#define CTX_MAX_K 32U
-
-#define MAP_SIZE LIBAFL_EDGES_MAP_SIZE
-
-using namespace llvm;
-
-static cl::opt<bool>     Debug("debug_afl_coverage", cl::desc("Debug prints"),
-                               cl::init(false), cl::NotHidden);
-static cl::opt<uint32_t> InstRatio(
-    "inst_ratio_afl_coverage", cl::desc("Instrumentation ratio in percentage"),
-    cl::init(100), cl::NotHidden);
-static cl::opt<bool>     NotZero("not_zero",
-                                 cl::desc("Never hit 0 again in the hitcount"),
-                                 cl::init(true), cl::NotHidden);
-static cl::opt<uint32_t> Ngram(
-    "ngram", cl::desc("Size of the Ngram instrumentation (0 to disable)"),
-    cl::init(0), cl::NotHidden);
-static cl::opt<uint32_t> CtxK(
-    "ctx_k",
-    cl::desc(
-        "Size of the context for K-Ctx context sensitivity (0 to disable)"),
-    cl::init(0), cl::NotHidden);
-static cl::opt<bool> Ctx("ctx",
-                         cl::desc("Enable full context sensitive coverage"),
-                         cl::init(false), cl::NotHidden);
-static cl::opt<bool> ThreadSafe("thread_safe_afl_coverage",
-                                cl::desc("Use the thread safe instrumentation"),
-                                cl::init(false), cl::NotHidden);
-static cl::opt<bool> DumpCFG(
-    "dump_afl_cfg", cl::desc("Dump CFG containing AFL-style edge index"),
-    cl::init(false), cl::NotHidden);
-static cl::opt<std::string> DumpCFGPath(
-    "dump_afl_cfg_path",
-    cl::desc("Path to dump CFG containing AFL-style edge index"),
-    cl::init(".cfg"), cl::NotHidden);
-
-namespace {
-
-#ifdef USE_NEW_PM
-class AFLCoverage : public PassInfoMixin<AFLCoverage> {
- public:
-  AFLCoverage() {
-#else
-class AFLCoverage : public ModulePass {
- public:
-  static char ID;
-  AFLCoverage() : ModulePass(ID) {
-#endif
-
-    // initInstrumentList();
-  }
-
-#ifdef USE_NEW_PM
-  PreservedAnalyses run(Module &M, ModuleAnalysisManager &MAM);
-#else
-  bool runOnModule(Module &M) override;
-#endif
-
- protected:
-  uint32_t                          map_size = MAP_SIZE;
-  uint32_t                          function_minimum_size = 1;
-  DenseMap<BasicBlock *, int32_t>   bb_to_cur_loc;
-  DenseMap<StringRef, BasicBlock *> entry_bb;
-};
-
-}  // namespace
-
-#ifdef USE_NEW_PM
-extern "C" ::llvm::PassPluginLibraryInfo LLVM_ATTRIBUTE_WEAK
-llvmGetPassPluginInfo() {
-  return {LLVM_PLUGIN_API_VERSION, "AFLCoverage", "v0.1",
-          /* lambda to insert our pass into the pass pipeline. */
-          [](PassBuilder &PB) {
-  #if 1
-    #if LLVM_VERSION_MAJOR <= 13
-            using OptimizationLevel = typename PassBuilder::OptimizationLevel;
-    #endif
-            PB.registerOptimizerLastEPCallback(
-                [](ModulePassManager &MPM, OptimizationLevel OL) {
-                  MPM.addPass(AFLCoverage());
-                });
-  /* TODO LTO registration */
-  #else
-            using PipelineElement = typename PassBuilder::PipelineElement;
-            PB.registerPipelineParsingCallback([](StringRef          Name,
-                                                  ModulePassManager &MPM,
-                                                  ArrayRef<PipelineElement>) {
-              if (Name == "AFLCoverage") {
-                MPM.addPass(AFLCoverage());
-                return true;
-              } else {
-                return false;
-              }
-            });
-  #endif
-          }};
-}
-#else
-
-char AFLCoverage::ID = 0;
-#endif
-
-#ifdef USE_NEW_PM
-PreservedAnalyses AFLCoverage::run(Module &M, ModuleAnalysisManager &MAM) {
-#else
-bool AFLCoverage::runOnModule(Module &M) {
-#endif
-  if (Ctx && DumpCFG) {
-    FATAL(
-        "Does not support dumping CFG with full context sensitive coverage "
-        "enabled.");
-  }
-  LLVMContext &C = M.getContext();
-
-  IntegerType *Int8Ty = IntegerType::getInt8Ty(C);
-  IntegerType *Int32Ty = IntegerType::getInt32Ty(C);
-#ifdef HAVE_VECTOR_INTRINSICS
-  IntegerType *IntLocTy =
-      IntegerType::getIntNTy(C, sizeof(prev_loc_t) * CHAR_BIT);
-#endif
-  uint32_t     rand_seed;
-  unsigned int cur_loc = 0;
-
-#ifdef USE_NEW_PM
-  auto PA = PreservedAnalyses::all();
-#endif
-
-  /* Setup random() so we get Actually Random(TM) */
-  rand_seed = time(NULL);
-  srand(rand_seed);
-
-  /*
-    char *ptr;
-    if ((ptr = getenv("AFL_MAP_SIZE")) || (ptr = getenv("AFL_MAPSIZE"))) {
-
-      map_size = atoi(ptr);
-      if (map_size < 8 || map_size > (1 << 29)) {
-        FATAL("illegal AFL_MAP_SIZE %u, must be between 2^3 and 2^30",
-      }
-    map_size); if (map_size % 8) {map_size = (((map_size >> 3) + 1) << 3);}
-
-    }
-
-  */
-
-  /* Decide instrumentation ratio */
-
-  if (!InstRatio || InstRatio > 100) {
-    FATAL("Bad value of the instrumentation ratio (must be between 1 and 100)");
-  }
-
-  unsigned PrevLocSize = 0;
-  unsigned PrevCallerSize = 0;
-
-  bool instrument_ctx = Ctx || CtxK > 0;
-  bool instrument_caller = false;
-
-#ifdef HAVE_VECTOR_INTRINSICS
-  /* Decide previous location vector size (must be a power of two) */
-  VectorType *PrevLocTy = NULL;
-
-  if (Ngram && (Ngram < 2 || Ngram > NGRAM_SIZE_MAX)) {
-    FATAL(
-        "Bad value of the Ngram size (must be between 2 and NGRAM_SIZE_MAX "
-        "(%u))",
-        NGRAM_SIZE_MAX);
-  }
-
-  if (Ngram) {
-    PrevLocSize = Ngram - 1;
-  } else {
-    PrevLocSize = 1;
-  }
-
-  /* Decide K-Ctx vector size (must be a power of two) */
-  VectorType *PrevCallerTy = NULL;
-
-  if (CtxK > CTX_MAX_K) {
-    FATAL(
-        "Bad value of K for K-context sensitivity (must be between 1 and "
-        "CTX_MAX_K (%u))",
-        CTX_MAX_K);
-  }
-
-  if (CtxK == 1) {
-    CtxK = 0;
-    instrument_ctx = true;
-    instrument_caller = true;  // Enable CALLER instead
-  }
-
-  if (CtxK) {
-    PrevCallerSize = CtxK;
-    instrument_ctx = true;
-  }
-
-#else
-  if (Ngram)
-  #ifndef LLVM_VERSION_PATCH
-    FATAL(
-        "Sorry, NGRAM branch coverage is not supported with llvm version "
-        "%d.%d.%d!",
-        LLVM_VERSION_MAJOR, LLVM_VERSION_MINOR, 0);
-  #else
-    FATAL(
-        "Sorry, NGRAM branch coverage is not supported with llvm version "
-        "%d.%d.%d!",
-        LLVM_VERSION_MAJOR, LLVM_VERSION_MINOR, LLVM_VERSION_PATCH);
-  #endif
-  if (CtxK)
-  #ifndef LLVM_VERSION_PATCH
-    FATAL(
-        "Sorry, K-CTX branch coverage is not supported with llvm version "
-        "%d.%d.%d!",
-        LLVM_VERSION_MAJOR, LLVM_VERSION_MINOR, 0);
-  #else
-    FATAL(
-        "Sorry, K-CTX branch coverage is not supported with llvm version "
-        "%d.%d.%d!",
-        LLVM_VERSION_MAJOR, LLVM_VERSION_MINOR, LLVM_VERSION_PATCH);
-  #endif
-  PrevLocSize = 1;
-#endif
-
-#ifdef HAVE_VECTOR_INTRINSICS
-  int PrevLocVecSize = PowerOf2Ceil(PrevLocSize);
-  if (Ngram)
-    PrevLocTy = VectorType::get(IntLocTy, PrevLocVecSize
-  #if LLVM_VERSION_MAJOR >= 12
-                                ,
-                                false
-  #endif
-    );
-#endif
-
-#ifdef HAVE_VECTOR_INTRINSICS
-  int PrevCallerVecSize = PowerOf2Ceil(PrevCallerSize);
-  if (CtxK)
-    PrevCallerTy = VectorType::get(IntLocTy, PrevCallerVecSize
-  #if LLVM_VERSION_MAJOR >= 12
-                                   ,
-                                   false
-  #endif
-    );
-#endif
-
-  /* Get globals for the SHM region and the previous location. Note that
-     __afl_prev_loc is thread-local. */
-
-  GlobalVariable *AFLMapPtr =
-      new GlobalVariable(M, PointerType::get(Int8Ty, 0), false,
-                         GlobalValue::ExternalLinkage, 0, "__afl_area_ptr");
-  GlobalVariable *AFLPrevLoc;
-  GlobalVariable *AFLPrevCaller;
-  GlobalVariable *AFLContext = NULL;
-
-  if (Ctx || instrument_caller)
-#if defined(__ANDROID__) || defined(__HAIKU__)
-    AFLContext = new GlobalVariable(
-        M, Int32Ty, false, GlobalValue::ExternalLinkage, 0, "__afl_prev_ctx");
-#else
-    AFLContext = new GlobalVariable(
-        M, Int32Ty, false, GlobalValue::ExternalLinkage, 0, "__afl_prev_ctx", 0,
-        GlobalVariable::GeneralDynamicTLSModel, 0, false);
-#endif
-
-#ifdef HAVE_VECTOR_INTRINSICS
-  if (Ngram)
-  #if defined(__ANDROID__) || defined(__HAIKU__)
-    AFLPrevLoc = new GlobalVariable(
-        M, PrevLocTy, /* isConstant */ false, GlobalValue::ExternalLinkage,
-        /* Initializer */ nullptr, "__afl_prev_loc");
-  #else
-    AFLPrevLoc = new GlobalVariable(
-        M, PrevLocTy, /* isConstant */ false, GlobalValue::ExternalLinkage,
-        /* Initializer */ nullptr, "__afl_prev_loc",
-        /* InsertBefore */ nullptr, GlobalVariable::GeneralDynamicTLSModel,
-        /* AddressSpace */ 0, /* IsExternallyInitialized */ false);
-  #endif
-  else
-#endif
-#if defined(__ANDROID__) || defined(__HAIKU__)
-    AFLPrevLoc = new GlobalVariable(
-        M, Int32Ty, false, GlobalValue::ExternalLinkage, 0, "__afl_prev_loc");
-#else
-  AFLPrevLoc = new GlobalVariable(
-      M, Int32Ty, false, GlobalValue::ExternalLinkage, 0, "__afl_prev_loc", 0,
-      GlobalVariable::GeneralDynamicTLSModel, 0, false);
-#endif
-
-#ifdef HAVE_VECTOR_INTRINSICS
-  if (CtxK)
-  #if defined(__ANDROID__) || defined(__HAIKU__)
-    AFLPrevCaller = new GlobalVariable(
-        M, PrevCallerTy, /* isConstant */ false, GlobalValue::ExternalLinkage,
-        /* Initializer */ nullptr, "__afl_prev_caller");
-  #else
-    AFLPrevCaller = new GlobalVariable(
-        M, PrevCallerTy, /* isConstant */ false, GlobalValue::ExternalLinkage,
-        /* Initializer */ nullptr, "__afl_prev_caller",
-        /* InsertBefore */ nullptr, GlobalVariable::GeneralDynamicTLSModel,
-        /* AddressSpace */ 0, /* IsExternallyInitialized */ false);
-  #endif
-  else
-#endif
-#if defined(__ANDROID__) || defined(__HAIKU__)
-    AFLPrevCaller =
-        new GlobalVariable(M, Int32Ty, false, GlobalValue::ExternalLinkage, 0,
-                           "__afl_prev_caller");
-#else
-  AFLPrevCaller = new GlobalVariable(
-      M, Int32Ty, false, GlobalValue::ExternalLinkage, 0, "__afl_prev_caller",
-      0, GlobalVariable::GeneralDynamicTLSModel, 0, false);
-#endif
-
-#ifdef HAVE_VECTOR_INTRINSICS
-  /* Create the vector shuffle mask for updating the previous block history.
-     Note that the first element of the vector will store cur_loc, so just set
-     it to undef to allow the optimizer to do its thing. */
-
-  SmallVector<Constant *, 32> PrevLocShuffle = {UndefValue::get(Int32Ty)};
-
-  for (unsigned I = 0; I < PrevLocSize - 1; ++I) {
-    PrevLocShuffle.push_back(ConstantInt::get(Int32Ty, I));
-  }
-
-  for (int I = PrevLocSize; I < PrevLocVecSize; ++I) {
-    PrevLocShuffle.push_back(ConstantInt::get(Int32Ty, PrevLocSize));
-  }
-
-  Constant *PrevLocShuffleMask = ConstantVector::get(PrevLocShuffle);
-
-  Constant                   *PrevCallerShuffleMask = NULL;
-  SmallVector<Constant *, 32> PrevCallerShuffle = {UndefValue::get(Int32Ty)};
-
-  if (CtxK) {
-    for (unsigned I = 0; I < PrevCallerSize - 1; ++I) {
-      PrevCallerShuffle.push_back(ConstantInt::get(Int32Ty, I));
-    }
-
-    for (int I = PrevCallerSize; I < PrevCallerVecSize; ++I) {
-      PrevCallerShuffle.push_back(ConstantInt::get(Int32Ty, PrevCallerSize));
-    }
-
-    PrevCallerShuffleMask = ConstantVector::get(PrevCallerShuffle);
-  }
-
-#endif
-
-  // other constants we need
-  ConstantInt *One = ConstantInt::get(Int8Ty, 1);
-
-  Value    *PrevCtx = NULL;     // CTX sensitive coverage
-  LoadInst *PrevCaller = NULL;  // K-CTX coverage
-
-  /* Instrument all the things! */
-
-  int inst_blocks = 0;
-  // scanForDangerousFunctions(&M);
-
-  for (auto &F : M) {
-    int has_calls = 0;
-    if (Debug)
-      fprintf(stderr, "FUNCTION: %s (%zu)\n", F.getName().str().c_str(),
-              F.size());
-
-    if (isIgnoreFunction(&F)) { continue; }
-
-    if (F.size() < function_minimum_size) { continue; }
-    if (DumpCFG) { entry_bb[F.getName()] = &F.getEntryBlock(); }
-
-    std::list<Value *> todo;
-    for (auto &BB : F) {
-      BasicBlock::iterator IP = BB.getFirstInsertionPt();
-      IRBuilder<>          IRB(&(*IP));
-
-      // Context sensitive coverage
-      if (instrument_ctx && &BB == &F.getEntryBlock()) {
-#ifdef HAVE_VECTOR_INTRINSICS
-        if (CtxK) {
-          PrevCaller = IRB.CreateLoad(
-  #if LLVM_VERSION_MAJOR >= 14
-              PrevCallerTy,
-  #endif
-              AFLPrevCaller);
-          PrevCaller->setMetadata(M.getMDKindID("nosanitize"),
-                                  MDNode::get(C, None));
-          PrevCtx =
-              IRB.CreateZExt(IRB.CreateXorReduce(PrevCaller), IRB.getInt32Ty());
-
-        } else
-#endif
-        {
-
-          // load the context ID of the previous function and write to to a
-          // local variable on the stack
-          LoadInst *PrevCtxLoad = IRB.CreateLoad(
-#if LLVM_VERSION_MAJOR >= 14
-              IRB.getInt32Ty(),
-#endif
-              AFLContext);
-          PrevCtxLoad->setMetadata(M.getMDKindID("nosanitize"),
-                                   MDNode::get(C, None));
-          PrevCtx = PrevCtxLoad;
-        }
-
-        // does the function have calls? and is any of the calls larger than one
-        // basic block?
-        for (auto &BB_2 : F) {
-          if (has_calls) { break; }
-          for (auto &IN : BB_2) {
-            CallInst *callInst = nullptr;
-            if ((callInst = dyn_cast<CallInst>(&IN))) {
-              Function *Callee = callInst->getCalledFunction();
-              if (!Callee || Callee->size() < function_minimum_size)
-                continue;
-              else {
-                has_calls = 1;
-                break;
-              }
-            }
-          }
-        }
-
-        // if yes we store a context ID for this function in the global var
-        if (has_calls) {
-          Value *NewCtx = ConstantInt::get(Int32Ty, RandBelow(map_size));
-#ifdef HAVE_VECTOR_INTRINSICS
-          if (CtxK) {
-            Value *ShuffledPrevCaller = IRB.CreateShuffleVector(
-                PrevCaller, UndefValue::get(PrevCallerTy),
-                PrevCallerShuffleMask);
-            Value *UpdatedPrevCaller = IRB.CreateInsertElement(
-                ShuffledPrevCaller, NewCtx, (uint64_t)0);
-
-            StoreInst *Store =
-                IRB.CreateStore(UpdatedPrevCaller, AFLPrevCaller);
-            Store->setMetadata(M.getMDKindID("nosanitize"),
-                               MDNode::get(C, None));
-
-          } else
-#endif
-          {
-
-            if (Ctx) { NewCtx = IRB.CreateXor(PrevCtx, NewCtx); }
-            StoreInst *StoreCtx = IRB.CreateStore(NewCtx, AFLContext);
-            StoreCtx->setMetadata(M.getMDKindID("nosanitize"),
-                                  MDNode::get(C, None));
-          }
-        }
-      }
-
-      if (RandBelow(100) >= InstRatio) { continue; }
-
-      /* Make up cur_loc */
-
-      // cur_loc++;
-      cur_loc = RandBelow(map_size);
-      if (DumpCFG) { bb_to_cur_loc[&BB] = cur_loc; }
-/* There is a problem with Ubuntu 18.04 and llvm 6.0 (see issue #63).
-   The inline function successors() is not inlined and also not found at runtime
-   :-( As I am unable to detect Ubuntu18.04 heree, the next best thing is to
-   disable this optional optimization for LLVM 6.0.0 and Linux */
-#if !(LLVM_VERSION_MAJOR == 6 && LLVM_VERSION_MINOR == 0) || !defined __linux__
-      // only instrument if this basic block is the destination of a previous
-      // basic block that has multiple successors
-      // this gets rid of ~5-10% of instrumentations that are unnecessary
-      // result: a little more speed and less map pollution
-      int more_than_one = -1;
-      // fprintf(stderr, "BB %u: ", cur_loc);
-      for (pred_iterator PI = pred_begin(&BB), E = pred_end(&BB); PI != E;
-           ++PI) {
-        BasicBlock *Pred = *PI;
-
-        int count = 0;
-        if (more_than_one == -1) { more_than_one = 0; }
-        // fprintf(stderr, " %p=>", Pred);
-
-        for (succ_iterator SI = succ_begin(Pred), E = succ_end(Pred); SI != E;
-             ++SI) {
-          BasicBlock *Succ = *SI;
-
-          // if (count > 0)
-          //  fprintf(stderr, "|");
-          if (Succ != NULL) { count++; }
-          // fprintf(stderr, "%p", Succ);
-        }
-
-        if (count > 1) { more_than_one = 1; }
-      }
-
-      // fprintf(stderr, " == %d\n", more_than_one);
-      if (F.size() > 1 && more_than_one != 1) {
-        // in CTX mode we have to restore the original context for the caller -
-        // she might be calling other functions which need the correct CTX
-        if (instrument_ctx && has_calls) {
-          Instruction *Inst = BB.getTerminator();
-          if (isa<ReturnInst>(Inst) || isa<ResumeInst>(Inst)) {
-            IRBuilder<> Post_IRB(Inst);
-
-            StoreInst *RestoreCtx;
-  #ifdef HAVE_VECTOR_INTRINSICS
-            if (CtxK)
-              RestoreCtx = IRB.CreateStore(PrevCaller, AFLPrevCaller);
-            else
-  #endif
-              RestoreCtx = Post_IRB.CreateStore(PrevCtx, AFLContext);
-            RestoreCtx->setMetadata(M.getMDKindID("nosanitize"),
-                                    MDNode::get(C, None));
-          }
-        }
-
-        continue;
-      }
-
-#endif
-
-      ConstantInt *CurLoc;
-
-#ifdef HAVE_VECTOR_INTRINSICS
-      if (Ngram)
-        CurLoc = ConstantInt::get(IntLocTy, cur_loc);
-      else
-#endif
-        CurLoc = ConstantInt::get(Int32Ty, cur_loc);
-
-      /* Load prev_loc */
-
-      LoadInst *PrevLoc;
-
-      if (Ngram) {
-        PrevLoc = IRB.CreateLoad(
-#if LLVM_VERSION_MAJOR >= 14
-            PrevLocTy,
-#endif
-            AFLPrevLoc);
-      } else {
-        PrevLoc = IRB.CreateLoad(
-#if LLVM_VERSION_MAJOR >= 14
-            IRB.getInt32Ty(),
-#endif
-            AFLPrevLoc);
-      }
-      PrevLoc->setMetadata(M.getMDKindID("nosanitize"), MDNode::get(C, None));
-      Value *PrevLocTrans;
-
-#ifdef HAVE_VECTOR_INTRINSICS
-      /* "For efficiency, we propose to hash the tuple as a key into the
-         hit_count map as (prev_block_trans << 1) ^ curr_block_trans, where
-         prev_block_trans = (block_trans_1 ^ ... ^ block_trans_(n-1)" */
-
-      if (Ngram)
-        PrevLocTrans =
-            IRB.CreateZExt(IRB.CreateXorReduce(PrevLoc), IRB.getInt32Ty());
-      else
-#endif
-        PrevLocTrans = PrevLoc;
-
-      if (instrument_ctx)
-        PrevLocTrans =
-            IRB.CreateZExt(IRB.CreateXor(PrevLocTrans, PrevCtx), Int32Ty);
-      else
-        PrevLocTrans = IRB.CreateZExt(PrevLocTrans, IRB.getInt32Ty());
-
-      /* Load SHM pointer */
-
-      LoadInst *MapPtr = IRB.CreateLoad(
-#if LLVM_VERSION_MAJOR >= 14
-          PointerType::get(Int8Ty, 0),
-#endif
-          AFLMapPtr);
-      MapPtr->setMetadata(M.getMDKindID("nosanitize"), MDNode::get(C, None));
-
-      Value *MapPtrIdx;
-#ifdef HAVE_VECTOR_INTRINSICS
-      if (Ngram)
-        MapPtrIdx = IRB.CreateGEP(
-  #if LLVM_VERSION_MAJOR >= 14
-            Int8Ty,
-  #endif
-            MapPtr,
-            IRB.CreateZExt(
-                IRB.CreateXor(PrevLocTrans, IRB.CreateZExt(CurLoc, Int32Ty)),
-                Int32Ty));
-      else
-#endif
-        MapPtrIdx = IRB.CreateGEP(
-#if LLVM_VERSION_MAJOR >= 14
-            Int8Ty,
-#endif
-            MapPtr, IRB.CreateXor(PrevLocTrans, CurLoc));
-
-      /* Update bitmap */
-
-      if (ThreadSafe) { /* Atomic */
-                        /*
-                        #if LLVM_VERSION_MAJOR < 9
-                                if (neverZero_counters_str !=
-                                    NULL) {  // with llvm 9 we make this the default as the bug
-                        in llvm
-                                             // is then fixed
-                        #else
-                                if (NotZero) {
-                
-                        #endif
-                                  // register MapPtrIdx in a todo list
-                                  todo.push_back(MapPtrIdx);
-                
-                                } else {
-                
-                        */
-        IRB.CreateAtomicRMW(llvm::AtomicRMWInst::BinOp::Add, MapPtrIdx, One,
-#if LLVM_VERSION_MAJOR >= 13
-                            llvm::MaybeAlign(1),
-#endif
-                            llvm::AtomicOrdering::Monotonic);
-        /*
-
-                }
-
-        */
-
-      } else {
-        LoadInst *Counter = IRB.CreateLoad(
-#if LLVM_VERSION_MAJOR >= 14
-            IRB.getInt8Ty(),
-#endif
-            MapPtrIdx);
-        Counter->setMetadata(M.getMDKindID("nosanitize"), MDNode::get(C, None));
-
-        Value *Incr = IRB.CreateAdd(Counter, One);
-
-#if LLVM_VERSION_MAJOR < 9
-        if (neverZero_counters_str !=
-            NULL) {  // with llvm 9 we make this the default as the bug in llvm
-                     // is then fixed
-#else
-        if (NotZero) {
-
-#endif
-          /* hexcoder: Realize a counter that skips zero during overflow.
-           * Once this counter reaches its maximum value, it next increments to
-           * 1
-           *
-           * Instead of
-           * Counter + 1 -> Counter
-           * we inject now this
-           * Counter + 1 -> {Counter, OverflowFlag}
-           * Counter + OverflowFlag -> Counter
-           */
-
-          ConstantInt *Zero = ConstantInt::get(Int8Ty, 0);
-          auto         cf = IRB.CreateICmpEQ(Incr, Zero);
-          auto         carry = IRB.CreateZExt(cf, Int8Ty);
-          Incr = IRB.CreateAdd(Incr, carry);
-        }
-
-        IRB.CreateStore(Incr, MapPtrIdx)
-            ->setMetadata(M.getMDKindID("nosanitize"), MDNode::get(C, None));
-
-      } /* non atomic case */
-
-      /* Update prev_loc history vector (by placing cur_loc at the head of the
-         vector and shuffle the other elements back by one) */
-
-      StoreInst *Store;
-
-#ifdef HAVE_VECTOR_INTRINSICS
-      if (Ngram) {
-        Value *ShuffledPrevLoc = IRB.CreateShuffleVector(
-            PrevLoc, UndefValue::get(PrevLocTy), PrevLocShuffleMask);
-        Value *UpdatedPrevLoc = IRB.CreateInsertElement(
-            ShuffledPrevLoc, IRB.CreateLShr(CurLoc, (uint64_t)1), (uint64_t)0);
-
-        Store = IRB.CreateStore(UpdatedPrevLoc, AFLPrevLoc);
-        Store->setMetadata(M.getMDKindID("nosanitize"), MDNode::get(C, None));
-
-      } else
-#endif
-      {
-
-        Store = IRB.CreateStore(ConstantInt::get(Int32Ty, cur_loc >> 1),
-                                AFLPrevLoc);
-        Store->setMetadata(M.getMDKindID("nosanitize"), MDNode::get(C, None));
-      }
-
-      // in CTX mode we have to restore the original context for the caller -
-      // she might be calling other functions which need the correct CTX.
-      // Currently this is only needed for the Ubuntu clang-6.0 bug
-      if (instrument_ctx && has_calls) {
-        Instruction *Inst = BB.getTerminator();
-        if (isa<ReturnInst>(Inst) || isa<ResumeInst>(Inst)) {
-          IRBuilder<> Post_IRB(Inst);
-
-          StoreInst *RestoreCtx;
-#ifdef HAVE_VECTOR_INTRINSICS
-          if (CtxK)
-            RestoreCtx = IRB.CreateStore(PrevCaller, AFLPrevCaller);
-          else
-#endif
-            RestoreCtx = Post_IRB.CreateStore(PrevCtx, AFLContext);
-          RestoreCtx->setMetadata(M.getMDKindID("nosanitize"),
-                                  MDNode::get(C, None));
-        }
-      }
-
-      inst_blocks++;
-    }
-  }
-  if (DumpCFG) {
-    int fd;
-#ifndef _WIN32
-    if ((fd = open(DumpCFGPath.c_str(), O_WRONLY | O_CREAT | O_APPEND, 0644)) <
-        0)
-#else
-    if ((fd = _open(DumpCFGPath.c_str(), O_WRONLY | O_CREAT | O_APPEND, 0644)) <
-        0)
-#endif
-      FATAL("Could not open/create CFG dump file.");
-    std::string cfg = "";
-    for (auto record = entry_bb.begin(); record != entry_bb.end(); record++) {
-      // Dump function BB entry points
-      cfg += formatv("$${0}+{1}\n", record->getFirst(),
-                     bb_to_cur_loc[record->getSecond()]);
-    }
-    for (auto record = bb_to_cur_loc.begin(); record != bb_to_cur_loc.end();
-         record++) {
-      // Dump CFG information
-      auto      current_bb = record->getFirst();
-      Function *calling_func = current_bb->getParent();
-      if (calling_func) {
-        auto function_name = calling_func->getName().str();
-        cfg += formatv("%%{0}", function_name);
-      } else
-        cfg += "%%__";
-      auto current_cur_loc = record->getSecond();
-      cfg += formatv("+{0}\n", current_cur_loc);
-      for (auto bb_successor = succ_begin(current_bb);
-           bb_successor != succ_end(current_bb); bb_successor++) {
-        cfg += formatv("->{0}\n", bb_to_cur_loc[*bb_successor]).str();
-      }
-    }
-    if (Debug) { errs() << "CFG: \n" << cfg; }
-    if (cfg.size() > 0 && write(fd, cfg.c_str(), cfg.length()) <= 0)
-      FATAL("Failed to dump CFG.\n");
-  }
-
-  /* Say something nice. */
-
-  /*if (!be_quiet) {
-
-    if (!inst_blocks)
-      WARNF("No instrumentation targets found.");
-    else {
-
-      char modeline[100];
-      snprintf(modeline, sizeof(modeline), "%s%s%s%s%s",
-               getenv("AFL_HARDEN") ? "hardened" : "non-hardened",
-               getenv("AFL_USE_ASAN") ? ", ASAN" : "",
-               getenv("AFL_USE_MSAN") ? ", MSAN" : "",
-               getenv("AFL_USE_CFISAN") ? ", CFISAN" : "",
-               getenv("AFL_USE_UBSAN") ? ", UBSAN" : "");
-      OKF("Instrumented %d locations (%s mode, ratio %u%%).", inst_blocks,
-          modeline, InstRatio);
-
-    }
-
-  }*/
-
-  if (Debug) {
-    if (!inst_blocks)
-      fprintf(stderr, "No instrumentation targets found.\n");
-    else
-      fprintf(stderr, "Instrumented %d locations (ratio %u%%).\n", inst_blocks,
-              (unsigned)InstRatio);
-  }
-
-#ifdef USE_NEW_PM
-  return PA;
-#else
-  return true;
-#endif
-}
-
-#ifndef USE_NEW_PM
-static void registerAFLPass(const PassManagerBuilder &,
-                            legacy::PassManagerBase &PM) {
-  PM.add(new AFLCoverage());
-}
-
-static RegisterStandardPasses RegisterAFLPass(
-    PassManagerBuilder::EP_OptimizerLast, registerAFLPass);
-
-static RegisterStandardPasses RegisterAFLPass0(
-    PassManagerBuilder::EP_EnabledOnOptLevel0, registerAFLPass);
-#endif

libafl_cc/src/clang.rs

@@ -30,8 +30,6 @@ pub enum LLVMPasses {
     //CmpLogIns,
     /// The `CmpLog` pass
     CmpLogRtn,
-    /// The AFL coverage pass
-    AFLCoverage,
     /// The Autotoken pass
     AutoTokens,
     /// The Coverage Accouting (BB metric) pass
@@ -41,6 +39,8 @@ pub enum LLVMPasses {
     #[cfg(unix)]
     /// The `CmpLog` Instruction pass
     CmpLogInstructions,
+    /// Instrument caller for sancov coverage
+    Ctx,
 }
 
 impl LLVMPasses {
@@ -50,8 +50,6 @@ impl LLVMPasses {
         match self {
             LLVMPasses::CmpLogRtn => PathBuf::from(env!("OUT_DIR"))
                 .join(format!("cmplog-routines-pass.{}", dll_extension())),
-            LLVMPasses::AFLCoverage => PathBuf::from(env!("OUT_DIR"))
-                .join(format!("afl-coverage-pass.{}", dll_extension())),
             LLVMPasses::AutoTokens => {
                 PathBuf::from(env!("OUT_DIR")).join(format!("autotokens-pass.{}", dll_extension()))
             }
@@ -63,6 +61,9 @@ impl LLVMPasses {
             #[cfg(unix)]
             LLVMPasses::CmpLogInstructions => PathBuf::from(env!("OUT_DIR"))
                 .join(format!("cmplog-instructions-pass.{}", dll_extension())),
+            LLVMPasses::Ctx => {
+                PathBuf::from(env!("OUT_DIR")).join(format!("ctx-pass.{}", dll_extension()))
+            }
         }
     }
 }

libafl_cc/src/cmplog-routines-pass.cc

@@ -71,7 +71,7 @@ class CmpLogRoutines : public ModulePass {
 #if USE_NEW_PM
   PreservedAnalyses run(Module &M, ModuleAnalysisManager &MAM);
 #else
-  bool        runOnModule(Module &M) override;
+  bool runOnModule(Module &M) override;
 
   #if LLVM_VERSION_MAJOR < 4
   const char *getPassName() const override {

libafl_cc/src/ctx-pass.cc

@@ -0,0 +1,224 @@
+/*
+   LibAFL - Ctx LLVM pass
+   --------------------------------------------------
+
+   Written by Dongjia Zhang <toka@aflplus.plus>
+
+   Copyright 2022-2023 AFLplusplus Project. All rights reserved.
+
+   Licensed under the Apache License, Version 2.0 (the "License");
+   you may not use this file except in compliance with the License.
+   You may obtain a copy of the License at:
+
+     http://www.apache.org/licenses/LICENSE-2.0
+
+*/
+
+#include <stdio.h>
+#include <stdlib.h>
+#include "common-llvm.h"
+#ifndef _WIN32
+  #include <unistd.h>
+  #include <sys/time.h>
+#else
+  #include <io.h>
+#endif
+#include <string.h>
+#include <sys/types.h>
+#include <sys/stat.h>
+#include <fcntl.h>
+#include <ctype.h>
+
+#include <list>
+#include <string>
+#include <fstream>
+#include <set>
+
+#include "llvm/Config/llvm-config.h"
+#include "llvm/ADT/Statistic.h"
+#include "llvm/IR/IRBuilder.h"
+
+#if USE_NEW_PM
+  #include "llvm/Passes/PassPlugin.h"
+  #include "llvm/Passes/PassBuilder.h"
+  #include "llvm/IR/PassManager.h"
+#else
+  #include "llvm/IR/LegacyPassManager.h"
+  #include "llvm/Transforms/IPO/PassManagerBuilder.h"
+#endif
+
+#include "llvm/IR/BasicBlock.h"
+#include "llvm/IR/Module.h"
+#include "llvm/IR/DebugInfo.h"
+#include "llvm/IR/CFG.h"
+#include "llvm/IR/Verifier.h"
+#include "llvm/Support/Debug.h"
+#include "llvm/Support/raw_ostream.h"
+#include "llvm/Transforms/Utils/BasicBlockUtils.h"
+#include "llvm/Analysis/LoopInfo.h"
+#include "llvm/Analysis/ValueTracking.h"
+#include "llvm/Pass.h"
+#include "llvm/IR/Constants.h"
+
+#include <iostream>
+
+using namespace llvm;
+
+#define MAP_SIZE LIBAFL_EDGES_MAP_SIZE
+
+namespace {
+
+#if USE_NEW_PM
+class CtxPass : public PassInfoMixin<CtxPass> {
+ public:
+  CtxPass() {
+#else
+class CtxPass : public ModulePass {
+ public:
+  static char ID;
+
+  CtxPass() : ModulePass(ID) {
+#endif
+  }
+
+#if USE_NEW_PM
+  PreservedAnalyses run(Module &M, ModuleAnalysisManager &MAM);
+#else
+  bool runOnModule(Module &M) override;
+#endif
+
+ protected:
+  uint32_t map_size = MAP_SIZE;
+
+ private:
+  bool isLLVMIntrinsicFn(StringRef &n) {
+    // Not interested in these LLVM's functions
+    if (n.startswith("llvm.")) {
+      return true;
+    } else {
+      return false;
+    }
+  }
+};
+
+}  // namespace
+
+#if USE_NEW_PM
+extern "C" ::llvm::PassPluginLibraryInfo LLVM_ATTRIBUTE_WEAK
+llvmGetPassPluginInfo() {
+  return {LLVM_PLUGIN_API_VERSION, "CtxPass", "v0.1",
+          /* lambda to insert our pass into the pass pipeline. */
+          [](PassBuilder &PB) {
+
+  #if LLVM_VERSION_MAJOR <= 13
+            using OptimizationLevel = typename PassBuilder::OptimizationLevel;
+  #endif
+            PB.registerOptimizerLastEPCallback(
+                [](ModulePassManager &MPM, OptimizationLevel OL) {
+                  MPM.addPass(CtxPass());
+                });
+          }};
+}
+#else
+char CtxPass::ID = 0;
+#endif
+
+#if USE_NEW_PM
+PreservedAnalyses CtxPass::run(Module &M, ModuleAnalysisManager &MAM) {
+#else
+bool CtxPass::runOnModule(Module &M) {
+
+#endif
+  LLVMContext &C = M.getContext();
+  auto         moduleName = M.getName();
+  IntegerType *Int8Ty = IntegerType::getInt8Ty(C);
+  IntegerType *Int32Ty = IntegerType::getInt32Ty(C);
+
+  uint32_t rand_seed;
+
+  rand_seed = time(NULL);
+  srand(rand_seed);
+
+  GlobalVariable *AFLContext = new GlobalVariable(
+      M, Int32Ty, false, GlobalValue::ExternalLinkage, 0, "__afl_prev_ctx");
+  Value *PrevCtx =
+      NULL;  // the ctx value up until now that we save on the stack
+
+  for (auto &F : M) {
+    int has_calls = 0;
+
+    if (isIgnoreFunction(&F)) { continue; }
+    if (F.size() < 1) { continue; }
+    for (auto &BB : F) {
+      BasicBlock::iterator IP = BB.getFirstInsertionPt();
+      IRBuilder<>          IRB(&(*IP));
+      if (&BB == &F.getEntryBlock()) {
+        // if this is the first block..
+        LoadInst *PrevCtxLoad = IRB.CreateLoad(IRB.getInt32Ty(), AFLContext);
+        PrevCtxLoad->setMetadata(M.getMDKindID("nosanitize"),
+                                 MDNode::get(C, None));
+        PrevCtx = PrevCtxLoad;
+
+        // now check for if calls exists
+        for (auto &BB_2 : F) {
+          if (has_calls) { break; }
+          for (auto &IN : BB_2) {
+            CallInst *callInst = nullptr;
+            if ((callInst = dyn_cast<CallInst>(&IN))) {
+              Function *Callee = callInst->getCalledFunction();
+              if (!Callee || Callee->size() < 1) {
+                continue;
+              } else {
+                has_calls = 1;
+                break;
+              }
+            }
+          }
+        }
+
+        if (has_calls) {
+          Value *NewCtx = ConstantInt::get(Int32Ty, RandBelow(map_size));
+          NewCtx = IRB.CreateXor(PrevCtx, NewCtx);
+          StoreInst *StoreCtx = IRB.CreateStore(NewCtx, AFLContext);
+          StoreCtx->setMetadata(M.getMDKindID("nosanitize"),
+                                MDNode::get(C, None));
+        }
+      }
+      // Restore the ctx at the end of BB
+      Instruction *Inst = BB.getTerminator();
+      if (has_calls) {
+        if (isa<ReturnInst>(Inst) || isa<ResumeInst>(Inst)) {
+          IRBuilder<> Post_IRB(Inst);
+          StoreInst  *RestoreCtx;
+          RestoreCtx = Post_IRB.CreateStore(PrevCtx, AFLContext);
+          RestoreCtx->setMetadata(M.getMDKindID("nosanitize"),
+                                  MDNode::get(C, None));
+        }
+      }
+    }
+  }
+
+#if USE_NEW_PM
+  auto PA = PreservedAnalyses::all();
+  return PA;
+#else
+  return true;
+#endif
+}
+
+#if USE_NEW_PM
+
+#else
+static void registerCtxPass(const PassManagerBuilder &,
+                            legacy::PassManagerBase &PM) {
+  PM.add(new CtxPass());
+}
+
+static RegisterPass<CtxPass> X("ctx", "ctx instrumentation pass", false, false);
+
+static RegisterStandardPasses RegisterCtxPass(
+    PassManagerBuilder::EP_OptimizerLast, registerCtxPass);
+
+static RegisterStandardPasses RegisterCtxPass0(
+    PassManagerBuilder::EP_EnabledOnOptLevel0, registerCtxPass);
+#endif

libafl_concolic/symcc_runtime/Cargo.toml

@@ -30,7 +30,7 @@ libafl_bolts = { path = "../../libafl_bolts", version = "0.11.2", default-featur
 
 [build-dependencies]
 cmake = "0.1"
-bindgen = "0.69"
+bindgen = "0.69.4"
 regex = "1"
 which = "4.4"
 symcc_libafl = { path = "../symcc_libafl", version = "0.11.2" }

libafl_libfuzzer/libafl_libfuzzer_runtime/Cargo.toml

@@ -48,7 +48,7 @@ utf8-chars = "3.0.1"
 env_logger = "0.10"
 
 [build-dependencies]
-bindgen = "0.68.1"
+bindgen = "0.69.4"
 cc = { version = "1.0", features = ["parallel"] }
 
 [workspace]

libafl_qemu/libafl_qemu_build/Cargo.toml

@@ -27,7 +27,7 @@ slirp = [] # build qemu with host libslirp (for user networking)
 clippy = [] # special feature for clippy, don't use in normal projects§
 
 [dependencies]
-bindgen = "0.68"
+bindgen = "0.69.4"
 which = "4.4"
 json = "0.12"
 shell-words = "1.1"

libafl_qemu/libqasan/hooks.c

@@ -480,7 +480,7 @@ char *strndup(const char *s, size_t n) {
   QASAN_LOAD(s, l + 1);
   void *r = __libqasan_malloc(l + 1);
   __libqasan_memcpy(r, s, l);
-  ((char*)r)[l] = 0;
+  ((char *)r)[l] = 0;
   QASAN_DEBUG("\t\t = %p\n", r);
 
   return r;

libafl_targets/Cargo.toml

@@ -42,6 +42,8 @@ sancov_pcguard_edges = ["coverage"]
 sancov_pcguard_hitcounts = ["coverage"]
 sancov_value_profile = ["common"]
 sancov_8bit = []
+sancov_ngram4 = ["coverage"]
+sancov_ctx = ["coverage"]
 sancov_cmplog = ["common"] # Defines cmp and __sanitizer_weak_hook functions. Use libfuzzer_interceptors to define interceptors (only compatible with Linux)
 sancov_pcguard = ["sancov_pcguard_hitcounts"]
 sanitizer_interfaces = []
@@ -56,7 +58,7 @@ whole_archive = [] # use +whole-archive to ensure the presence of weak symbols
 cmplog_extended_instrumentation = [] # support for aflpp cmplog map, we will remove this once aflpp and libafl cmplog shares the same LLVM passes.
 
 [build-dependencies]
-bindgen = "0.68"
+bindgen = "0.69.4"
 cc = { version = "1.0", features = ["parallel"] }
 rustversion = "1.0"
 
@@ -65,6 +67,7 @@ libafl = { path = "../libafl", version = "0.11.2", default-features = false, fea
 libafl_bolts = { path = "../libafl_bolts", version = "0.11.2", default-features = false, features = [] }
 libc = "0.2"
 log = "0.4.20"
+rustversion = "1.0"
 
 rangemap = "1.3"
 serde = { version = "1.0", default-features = false, features = ["alloc"] } # serialization lib

libafl_targets/build.rs

@@ -5,8 +5,17 @@ use std::{env, fs::File, io::Write, path::Path};
 const TWO_MB: usize = 2_621_440;
 const SIXTY_FIVE_KB: usize = 65_536;
 
+#[rustversion::nightly]
+fn enable_nightly() {
+    println!("cargo:rustc-cfg=nightly");
+}
+
+#[rustversion::not(nightly)]
+fn enable_nightly() {}
+
 #[allow(clippy::too_many_lines)]
 fn main() {
+    enable_nightly();
     let out_dir = env::var_os("OUT_DIR").unwrap();
     let out_dir = out_dir.to_string_lossy().to_string();
     //let out_dir_path = Path::new(&out_dir);

libafl_targets/src/coverage.c

@@ -33,7 +33,5 @@ uint8_t *__token_stop = &__stop_libafl_token;
 #endif
 
 // #if defined(__ANDROID__) || defined(__HAIKU__)
-MAYBE_THREAD_LOCAL prev_loc_t __afl_prev_loc[NGRAM_SIZE_MAX];
+uint32_t                      __afl_prev_ctx;
-MAYBE_THREAD_LOCAL prev_loc_t __afl_prev_caller[CTX_MAX_K];
-MAYBE_THREAD_LOCAL uint32_t   __afl_prev_ctx;
 MAYBE_THREAD_LOCAL prev_loc_t __afl_acc_prev_loc;

libafl_targets/src/lib.rs

@@ -3,6 +3,7 @@
 #![deny(rustdoc::broken_intra_doc_links)]
 #![deny(clippy::all)]
 #![deny(clippy::pedantic)]
+#![cfg_attr(nightly, feature(portable_simd))]
 #![allow(
     clippy::unreadable_literal,
     clippy::type_repetition_in_bounds,
@@ -66,9 +67,19 @@ extern crate alloc;
 
 include!(concat!(env!("OUT_DIR"), "/constants.rs"));
 
-#[cfg(any(feature = "sancov_pcguard_edges", feature = "sancov_pcguard_hitcounts",))]
+#[cfg(any(
+    feature = "sancov_pcguard_edges",
+    feature = "sancov_pcguard_hitcounts",
+    feature = "sancov_ngram4",
+    feature = "sancov_ctx"
+))]
 pub mod sancov_pcguard;
-#[cfg(any(feature = "sancov_pcguard_edges", feature = "sancov_pcguard_hitcounts",))]
+#[cfg(any(
+    feature = "sancov_pcguard_edges",
+    feature = "sancov_pcguard_hitcounts",
+    feature = "sancov_ngram4",
+    feature = "sancov_ctx"
+))]
 pub use sancov_pcguard::*;
 
 #[cfg(any(feature = "sancov_cmplog", feature = "sancov_value_profile"))]

libafl_targets/src/sancov_pcguard.rs

@@ -1,8 +1,18 @@
 //! [`LLVM` `PcGuard`](https://clang.llvm.org/docs/SanitizerCoverage.html#tracing-pcs-with-guards) runtime for `LibAFL`.
 
-use crate::coverage::{EDGES_MAP, MAX_EDGES_NUM};
+#[rustversion::nightly]
+#[cfg(feature = "sancov_ngram4")]
+use core::simd::num::SimdUint;
+
+#[cfg(any(feature = "sancov_ngram4", feature = "sancov_ctx"))]
+use libafl::executors::{hooks::ExecutorHook, HasObservers};
+
 #[cfg(feature = "pointer_maps")]
 use crate::coverage::{EDGES_MAP_PTR, EDGES_MAP_PTR_NUM};
+use crate::{
+    coverage::{EDGES_MAP, MAX_EDGES_NUM},
+    EDGES_MAP_SIZE,
+};
 
 #[cfg(all(feature = "sancov_pcguard_edges", feature = "sancov_pcguard_hitcounts"))]
 #[cfg(not(any(doc, feature = "clippy")))]
@@ -10,14 +20,122 @@ compile_error!(
     "the libafl_targets `sancov_pcguard_edges` and `sancov_pcguard_hitcounts` features are mutually exclusive."
 );
 
+#[cfg(feature = "sancov_ngram4")]
+#[rustversion::nightly]
+type Ngram4 = core::simd::u32x4;
+
+/// The array holding the previous locs. This is required for NGRAM-4 instrumentation
+#[cfg(feature = "sancov_ngram4")]
+#[rustversion::nightly]
+pub static mut PREV_ARRAY: Ngram4 = Ngram4::from_array([0, 0, 0, 0]);
+
+/// The hook to initialize ngram everytime we run the harness
+#[cfg(feature = "sancov_ngram4")]
+#[rustversion::nightly]
+#[derive(Default, Debug, Clone, Copy)]
+pub struct NgramHook {}
+
+/// The hook to initialize ctx everytime we run the harness
+#[cfg(feature = "sancov_ctx")]
+#[derive(Default, Debug, Clone, Copy)]
+pub struct CtxHook {}
+
+#[cfg(feature = "sancov_ngram4")]
+#[rustversion::nightly]
+impl ExecutorHook for NgramHook {
+    fn init<E: HasObservers, S>(&mut self, _state: &mut S) {}
+    fn pre_exec<EM, I, S, Z>(
+        &mut self,
+        _fuzzer: &mut Z,
+        _state: &mut S,
+        _mgr: &mut EM,
+        _input: &I,
+    ) {
+        unsafe {
+            PREV_ARRAY = Ngram4::from_array([0, 0, 0, 0]);
+        }
+    }
+    fn post_exec<EM, I, S, Z>(
+        &mut self,
+        _fuzzer: &mut Z,
+        _state: &mut S,
+        _mgr: &mut EM,
+        _input: &I,
+    ) {
+    }
+}
+
+#[cfg(feature = "sancov_ctx")]
+impl ExecutorHook for CtxHook {
+    fn init<E: HasObservers, S>(&mut self, _state: &mut S) {}
+    fn pre_exec<EM, I, S, Z>(
+        &mut self,
+        _fuzzer: &mut Z,
+        _state: &mut S,
+        _mgr: &mut EM,
+        _input: &I,
+    ) {
+        unsafe {
+            __afl_prev_ctx = 0;
+        }
+    }
+    fn post_exec<EM, I, S, Z>(
+        &mut self,
+        _fuzzer: &mut Z,
+        _state: &mut S,
+        _mgr: &mut EM,
+        _input: &I,
+    ) {
+    }
+}
+
+#[rustversion::nightly]
+#[cfg(feature = "sancov_ngram4")]
+unsafe fn update_ngram(mut pos: usize) -> usize {
+    #[cfg(feature = "sancov_ngram4")]
+    {
+        PREV_ARRAY = PREV_ARRAY.rotate_elements_right::<1>();
+        PREV_ARRAY.as_mut_array()[0] = pos as u32;
+        let reduced = PREV_ARRAY.reduce_xor() as usize;
+        pos ^= reduced;
+        pos %= EDGES_MAP_SIZE;
+    }
+    pos
+}
+
+#[rustversion::not(nightly)]
+#[cfg(feature = "sancov_ngram4")]
+unsafe fn update_ngram(pos: usize) -> usize {
+    pos
+}
+
+extern "C" {
+    /// The ctx variable
+    pub static mut __afl_prev_ctx: u32;
+}
+
 /// Callback for sancov `pc_guard` - usually called by `llvm` on each block or edge.
 ///
 /// # Safety
 /// Dereferences `guard`, reads the position from there, then dereferences the [`EDGES_MAP`] at that position.
 /// Should usually not be called directly.
 #[no_mangle]
+#[allow(unused_assignments)]
 pub unsafe extern "C" fn __sanitizer_cov_trace_pc_guard(guard: *mut u32) {
-    let pos = *guard as usize;
+    let mut pos = *guard as usize;
+
+    #[cfg(feature = "sancov_ngram4")]
+    {
+        pos = update_ngram(pos);
+        // println!("Wrinting to {} {}", pos, EDGES_MAP_SIZE);
+    }
+
+    #[cfg(feature = "sancov_ctx")]
+    {
+        pos ^= __afl_prev_ctx as usize;
+        // println!("Wrinting to {} {}", pos, EDGES_MAP_SIZE);
+    }
+
     #[cfg(feature = "pointer_maps")]
     {
         #[cfg(feature = "sancov_pcguard_edges")]
@@ -72,7 +190,7 @@ pub unsafe extern "C" fn __sanitizer_cov_trace_pc_guard_init(mut start: *mut u32
         #[cfg(not(feature = "pointer_maps"))]
         {
             MAX_EDGES_NUM = MAX_EDGES_NUM.wrapping_add(1);
-            assert!((MAX_EDGES_NUM <= EDGES_MAP.len()), "The number of edges reported by SanitizerCoverage exceed the size of the edges map ({}). Use the LIBAFL_EDGES_MAP_SIZE env to increase it at compile time.", EDGES_MAP.len());
+            // assert!((MAX_EDGES_NUM <= EDGES_MAP.len()), "The number of edges reported by SanitizerCoverage exceed the size of the edges map ({}). Use the LIBAFL_EDGES_MAP_SIZE env to increase it at compile time.", EDGES_MAP.len());
         }
     }
 }