OSS-Fuzz: OSS-Fuzz fuzzing integration

Signed-off-by: Arthur Chan <[email protected]>

fuzz/.gitignore

@@ -0,0 +1,4 @@
+target
+corpus
+artifacts
+coverage

fuzz/Cargo.toml

@@ -0,0 +1,36 @@
+[package]
+name = "tar-fuzz"
+version = "0.0.0"
+publish = false
+edition = "2018"
+
+[package.metadata]
+cargo-fuzz = true
+
+[dependencies]
+libfuzzer-sys = "0.4"
+tempfile = "3.3"
+
+[dependencies.tar]
+path = ".."
+
+[[bin]]
+name = "archive"
+path = "fuzz_targets/archive.rs"
+test = false
+doc = false
+bench = false
+
+[[bin]]
+name = "builder"
+path = "fuzz_targets/builder.rs"
+test = false
+doc = false
+bench = false
+
+[[bin]]
+name = "tar"
+path = "fuzz_targets/tar.rs"
+test = false
+doc = false
+bench = false

fuzz/fuzz_targets/archive.rs

@@ -0,0 +1,139 @@
+// Copyright 2024 Google LLC
+//
+// 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
+//
+// Unless required by applicable law or agreed to in writing, software
+// distributed under the License is distributed on an "AS IS" BASIS,
+// WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
+// See the License for the specific language governing permissions and
+// limitations under the License.
+
+#![no_main]
+
+use libfuzzer_sys::fuzz_target;
+
+use std::fs::{File, OpenOptions};
+use std::io::{Cursor, Write, Read};
+use tar::{Archive, Builder, EntryType, Header};
+use tempfile::tempdir;
+use std::convert::TryInto;
+use std::str;
+
+fuzz_target!(|data: &[u8]| {
+    // Skip this iteration when data is not enough
+    if data.len() < 10 {
+        return;
+    }
+
+    // Create temp file and dir
+    let temp_dir = tempdir().unwrap();
+    let file_name = match str::from_utf8(&data[0..data.len().min(10)]) {
+        Ok(name) => name.to_string(),
+        Err(_) => "default_file_name".to_string(),
+    };
+    let dir_name = match str::from_utf8(&data[data.len().min(10)..data.len().min(20)]) {
+        Ok(name) => name.to_string(),
+        Err(_) => "default_dir_name".to_string(),
+    };
+    let temp_file_path = temp_dir.path().join(format!("{}_file.tar", file_name));
+
+    // Initialise builder and cursor
+    let mut builder = Builder::new(Vec::new());
+    let mut cursor = Cursor::new(data.to_vec());
+
+    // Choose an etnry type
+    let entry_type_byte = data[0];
+    let entry_type = match entry_type_byte % 5 {
+        0 => EntryType::Regular,
+        1 => EntryType::Directory,
+        2 => EntryType::Symlink,
+        3 => EntryType::hard_link(),
+        _ => EntryType::character_special(),
+    };
+
+    // Initilaise header
+    let mut header = Header::new_gnu();
+    let file_size = u64::from_le_bytes(
+        data.get(1..9)
+            .unwrap_or(&[0; 8])
+            .try_into()
+            .unwrap_or([0; 8]),
+    );
+    header.set_size(file_size);
+    header.set_entry_type(entry_type);
+    header.set_cksum();
+
+    // Prepare sample tar file
+    let tar_file_path = format!("{}/{}", dir_name, file_name);
+    let _ = builder.append_data(&mut header, tar_file_path.clone(), &mut cursor).ok();
+    cursor.set_position(0);
+    for i in 1..5 {
+        let start = i * 10 % data.len();
+        let end = std::cmp::min(start + 10, data.len());
+        let entry_data = &data[start..end];
+        let entry_name = match str::from_utf8(&entry_data) {
+            Ok(name) => name.to_string(),
+            Err(_) => format!("entry_{}", i),
+        };
+
+        let mut entry_header = Header::new_gnu();
+        entry_header.set_size(entry_data.len() as u64);
+        entry_header.set_entry_type(entry_type);
+        entry_header.set_cksum();
+
+        let mut entry_cursor = Cursor::new(entry_data.to_vec());
+        let _ = builder.append_data(&mut entry_header, entry_name, &mut entry_cursor).ok();
+    }
+
+    // Prepare malformed tar header
+    if data.len() > 512 {
+        let corrupt_header_data = &data[data.len() - 512..];
+        let corrupt_header = Header::from_byte_slice(corrupt_header_data);
+        let mut corrupt_cursor = Cursor::new(data.to_vec());
+        let corrupt_entry_name = "corrupt_entry.txt";
+        let _ = builder.append_data(&mut corrupt_header.clone(), corrupt_entry_name, &mut corrupt_cursor).ok();
+    }
+
+    if let Ok(mut tar_file) = File::create(&temp_file_path) {
+        if let Ok(tar_data) = builder.into_inner() {
+            let _ = tar_file.write_all(&tar_data);
+        }
+    }
+
+    // Fuzz archive and builder unpack with malformed tar archvie
+    if let Ok(mut tar_file) = OpenOptions::new().read(true).open(&temp_file_path) {
+        let mut tar_data = Vec::new();
+        let _ = tar_file.read_to_end(&mut tar_data);
+        let mut tar_cursor = Cursor::new(tar_data);
+        let mut archive = Archive::new(&mut tar_cursor);
+        let _ = archive.unpack(temp_dir.path()).ok();
+    }
+
+    // Fuzz archive and builder
+    for i in 0..3 {
+        let name_data = &data[i * 5 % data.len()..(i * 5 + 5) % data.len()];
+        let name = match str::from_utf8(name_data) {
+            Ok(n) => n.to_string(),
+            Err(_) => format!("random_name_{}", i),
+        };
+        let path = temp_dir.path().join(name);
+        if i % 2 == 0 {
+            // Create a file
+            if let Ok(mut file) = File::create(&path) {
+                let _ = file.write_all(data);
+            }
+        } else {
+            // Create a directory
+            let _ = std::fs::create_dir(&path);
+        }
+    }
+
+    // Fuzz unpacking
+    let mut data_cursor = Cursor::new(data.to_vec());
+    let mut data_archive = Archive::new(&mut data_cursor);
+    let _ = data_archive.unpack(temp_dir.path()).ok();
+});

fuzz/fuzz_targets/builder.rs

@@ -0,0 +1,67 @@
+// Copyright 2024 Google LLC
+//
+// 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
+//
+// Unless required by applicable law or agreed to in writing, software
+// distributed under the License is distributed on an "AS IS" BASIS,
+// WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
+// See the License for the specific language governing permissions and
+// limitations under the License.
+
+#![no_main]
+
+use libfuzzer_sys::fuzz_target;
+
+use std::io::Cursor;
+use tar::Builder;
+use tempfile::{tempdir, tempfile};
+
+fuzz_target!(|data: &[u8]| {
+    // Initialization
+    let random_bool = data.first().map(|&b| b % 2 == 0).unwrap_or(false);
+    let temp_dir = tempdir().expect("");
+
+    // Create a temporary file for testing
+    if let Ok(temp_file) = tempfile() {
+        let mut builder = Builder::new(temp_file);
+
+        // Randomly choose a function target from builder to fuzz
+        match data.first().map(|&b| b % 8) {
+            Some(0) => {
+                builder.mode(if random_bool { tar::HeaderMode::Deterministic } else { tar::HeaderMode::Complete });
+            }
+            Some(1) => {
+                if let Ok(mut file) = tempfile() {
+                    let _ = builder.append_file("testfile.txt", &mut file);
+                }
+            }
+            Some(2) => {
+                let _ = builder.append_data(&mut tar::Header::new_old(), "randomfile", Cursor::new(data));
+            }
+            Some(3) => {
+                if let Ok(mut file) = tempfile() {
+                    let _ = builder.append_data(&mut tar::Header::new_old(), "testwrite.txt", &mut file);
+                }
+            }
+            Some(4) => {
+                let link_path = temp_dir.path().join("testlink");
+                let _ = builder.append_link(&mut tar::Header::new_old(), "testlink.txt", &link_path);
+            }
+            Some(5) => {
+                let _ = builder.append_path(temp_dir.path());
+            }
+            Some(6) => {
+                let link_path = temp_dir.path().join("testlink_with_path");
+                let _ = builder.append_link(&mut tar::Header::new_old(), temp_dir.path(), &link_path);
+            }
+            Some(7) => {
+                let _ = builder.append_dir_all("testdir", temp_dir.path());
+            }
+            _ => {}
+        }
+    }
+});

fuzz/fuzz_targets/tar.rs

@@ -0,0 +1,99 @@
+// Copyright 2024 Google LLC
+//
+// 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
+//
+// Unless required by applicable law or agreed to in writing, software
+// distributed under the License is distributed on an "AS IS" BASIS,
+// WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
+// See the License for the specific language governing permissions and
+// limitations under the License.
+
+#![no_main]
+
+use libfuzzer_sys::fuzz_target;
+
+use tar::{Builder, Header, Archive, EntryType};
+use std::io::{Cursor, Read, Write, Seek};
+use tempfile::{tempdir, NamedTempFile};
+
+fuzz_target!(|data: &[u8]| {
+    // Setup temporary directory and path
+    let temp_dir = tempdir().unwrap();
+    let archive_data = Cursor::new(data);
+    let mut builder = Builder::new(Cursor::new(Vec::new()));
+    let mut header = Header::new_gnu();
+
+    // Set header metadata
+    header.set_size(data.len() as u64);
+    header.set_cksum();
+    header.set_entry_type(EntryType::file());
+
+    // Append data and a temp file to tar
+    let _ = builder.append_data(&mut header, "fuzzed/file", archive_data);
+    let mut temp_file = NamedTempFile::new().unwrap();
+    let _ = temp_file.write_all(data);
+    let _ = builder.append_file("fuzzed/file2", temp_file.as_file_mut()).ok();
+
+    #[cfg(unix)]
+    let _ = builder.append_link(&mut header, "symlink/path", "target/path").ok();
+
+    let _ = builder.finish();
+
+    // Fuzzing Archive and Entry logic
+    let mut archive = Archive::new(Cursor::new(data));
+    if let Ok(mut entries) = archive.entries() {
+        while let Some(Ok(mut entry)) = entries.next() {
+            let _ = entry.path().map(|p| p.to_owned());
+            let _ = entry.link_name().map(|l| l.map(|ln| ln.to_owned()));
+            let _ = entry.size();
+            let _ = entry.header();
+            let _ = entry.raw_header_position();
+            let _ = entry.raw_file_position();
+
+            match entry.header().entry_type() {
+                EntryType::Regular => { /* Do nothing */ }
+                EntryType::Directory => {
+                    let _ = entry.unpack_in(temp_dir.path()).ok();
+                }
+                EntryType::Symlink => {
+                    let _ = entry.unpack_in(temp_dir.path()).ok();
+                }
+                EntryType::Link => {
+                    let _ = entry.unpack_in(temp_dir.path()).ok();
+                }
+                EntryType::Fifo => { /* Do nothing */ }
+                _ => { /* Do nothing */ }
+            }
+
+            let mut buffer = Vec::new();
+            let _ = entry.read_to_end(&mut buffer).ok();
+            entry.set_mask(0o755);
+            entry.set_unpack_xattrs(true);
+            entry.set_preserve_permissions(true);
+            entry.set_preserve_mtime(true);
+
+            // Fuzz unpack
+            let dst_path = temp_dir.path().join("unpacked_file");
+            let _ = entry.unpack(&dst_path).ok();
+            let _ = entry.unpack_in(temp_dir.path()).ok();
+
+            // Fuzz PaxExtensions
+            if let Ok(Some(pax_extensions)) = entry.pax_extensions() {
+                for ext in pax_extensions {
+                    let _ = ext.ok();
+                }
+            }
+
+            // Fuzzing file search with tar entry position
+            if entry.size() > 0 {
+                let mut data_cursor = Cursor::new(data);
+                let _ = data_cursor.seek(std::io::SeekFrom::Start(entry.raw_file_position())).ok();
+                let _ = data_cursor.read(&mut buffer).ok();
+            }
+        }
+    }
+});