aboutsummaryrefslogtreecommitdiff
path: root/projects/libxml2
diff options
context:
space:
mode:
authorMax Moroz <mmoroz@chromium.org>2019-09-06 09:56:42 -0700
committerGitHub <noreply@github.com>2019-09-06 09:56:42 -0700
commit16125ac2bce61d29aa5e418444dd328a555c3d90 (patch)
tree5b263ac0259e7ac7834bc34600581cce74f8b10d /projects/libxml2
parent227ee9f107b5f7f7110a3e30d8b3562407722fa3 (diff)
downloadoss-fuzz-16125ac2bce61d29aa5e418444dd328a555c3d90.tar.gz
[libxml2] Use FuzzedDataProvider isntead of byte_stream. (#2809)
Diffstat (limited to 'projects/libxml2')
-rw-r--r--projects/libxml2/byte_stream.h142
-rw-r--r--projects/libxml2/libxml2_xml_reader_for_file_fuzzer.cc17
2 files changed, 9 insertions, 150 deletions
diff --git a/projects/libxml2/byte_stream.h b/projects/libxml2/byte_stream.h
deleted file mode 100644
index da72d5573..000000000
--- a/projects/libxml2/byte_stream.h
+++ /dev/null
@@ -1,142 +0,0 @@
-// Copyright 2018 Google Inc.
-//
-// 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.
-
-#ifndef BYTE_STREAM_H_
-#define BYTE_STREAM_H_
-
-#include <algorithm>
-#include <cstddef>
-#include <cstdint>
-#include <limits>
-#include <string>
-
-// Wrapper for fuzzer input strings that helps consume and interpret the data
-// as a sequence of values, such as strings and PODs.
-class ByteStream {
- public:
- // Does not take ownership of data.
- ByteStream(const uint8_t* data, size_t size)
- : data_(data), size_(size), position_(0) {}
-
- ByteStream(const ByteStream&) = delete;
- ByteStream& operator=(const ByteStream&) = delete;
-
- // Returns a pointer to the chunk of data of |size| bytes, where |size| is
- // either a requested value or all the bytes that are available. If the
- // requested |size| is 0, return all the bytes that are available.
- const uint8_t* GetNextChunk(size_t* size);
-
- // Returns a string. Strings are obtained from the byte stream by reading a
- // size_t N followed by N char elements. If there are fewer than N bytes left
- // in the stream, this returns as many bytes as are available.
- std::string GetNextString();
-
- // The following GetNext{integer type} functions all return the next
- // sizeof(integer type) bytes in the stream or 0 if there is insufficient
- // capacity.
- size_t GetNextSizeT() { return ConsumeCopyOrDefault<size_t>(0); }
- int GetNextInt() { return ConsumeCopyOrDefault<int>(0); }
- uint8_t GetNextUint8() { return ConsumeCopyOrDefault<uint8_t>(0); }
- int64_t GetNextInt64() { return ConsumeCopyOrDefault<int64_t>(0); }
-
- // Returns an integer in the range [0,n) for n > 0 and consumes up to
- // sizeof(int) bytes. For n<=0, returns 0 and consumes 0 bytes.
- int GetNextInt(int n);
-
- private:
- // The remaining capacity of the ByteStream.
- size_t capacity() const { return size_ - position_; }
-
- // Returns data_ + position_ and then advances position_ by requested bytes.
- //
- // This is the canonical way for the class to request regions of memory
- // or to advance the position by requested bytes. This operation is unchecked
- // for maintaining that position_ <= size_. Requesting 0 bytes always
- // succeeds.
- const uint8_t* UncheckedConsume(size_t requested) {
- const uint8_t* region = data_ + position_;
- position_ += requested;
- return region;
- }
-
- // Directly initialize T by copying sizeof(T) bytes into results if there is
- // sufficient capacity in the stream. If there is not sufficient capacity
- // result is unmodified.
- template <class T>
- void ConsumeBytesByCopy(T* result) {
- constexpr size_t type_size = sizeof(T);
- if (type_size <= capacity()) {
- const uint8_t* region = UncheckedConsume(type_size);
- memcpy(static_cast<void*>(result), region, type_size);
- } else {
- // Consume the remainder of data_.
- UncheckedConsume(capacity());
- }
- }
-
- // A helper function for using ConsumeBytesByCopy and returning a default
- // value `t` if there is insufficient capacity to read a full `T`. T should
- // probably be a primitive type.
- template <class T>
- T ConsumeCopyOrDefault(T t) {
- ConsumeBytesByCopy(&t);
- return t;
- }
-
- const uint8_t* data_;
- const size_t size_;
- size_t position_;
-};
-
-inline const uint8_t* ByteStream::GetNextChunk(size_t* size) {
- if (*size)
- *size = std::min(*size, capacity());
- else
- *size = capacity();
-
- return UncheckedConsume(*size);
-}
-
-inline std::string ByteStream::GetNextString() {
- const size_t requested_size = GetNextSizeT();
- const size_t consumed_size = std::min(requested_size, capacity());
- const uint8_t* selection = UncheckedConsume(consumed_size);
- return std::string(reinterpret_cast<const char*>(selection), consumed_size);
-}
-
-inline int ByteStream::GetNextInt(int n) {
- if (n <= 0) {
- return 0;
- }
- // We grab as few bytes as possible as n will often be fixed.
- int selection = 0;
- if (n <= std::numeric_limits<uint8_t>::max()) {
- selection = static_cast<int>(GetNextUint8());
- } else if (n <= std::numeric_limits<uint16_t>::max()) {
- selection = ConsumeCopyOrDefault<uint16_t>(0);
- } else {
- selection = GetNextInt();
- }
-
- // Take the absolute value of selection w/o undefined behavior.
- // If selection is INT_MIN, return 0.
- if (selection == std::numeric_limits<int>::min()) {
- selection = 0;
- } else if (selection < 0) {
- selection = -selection;
- }
- return selection % n;
-}
-
-#endif // BYTE_STREAM_H_
diff --git a/projects/libxml2/libxml2_xml_reader_for_file_fuzzer.cc b/projects/libxml2/libxml2_xml_reader_for_file_fuzzer.cc
index ab296d834..4a71aa5a9 100644
--- a/projects/libxml2/libxml2_xml_reader_for_file_fuzzer.cc
+++ b/projects/libxml2/libxml2_xml_reader_for_file_fuzzer.cc
@@ -12,11 +12,12 @@
// See the License for the specific language governing permissions and
// limitations under the License.
+#include <fuzzer/FuzzedDataProvider.h>
+
#include <cstddef>
#include <cstdint>
#include <string>
-#include "byte_stream.h"
#include "fuzzer_temp_file.h"
#include "libxml/xmlreader.h"
@@ -28,14 +29,14 @@ void ignore (void* ctx, const char* msg, ...) {
extern "C" int LLVMFuzzerTestOneInput(const uint8_t* data, size_t size) {
xmlSetGenericErrorFunc(NULL, &ignore);
- ByteStream stream(data, size);
- const int options = stream.GetNextInt();
- const std::string encoding = stream.GetNextString();
- size_t file_contents_size = 0;
- const uint8_t* file_contents = stream.GetNextChunk(&file_contents_size);
+ FuzzedDataProvider provider(data, size);
+ const int options = provider.ConsumeIntegral<int>();
+
+ // libxml does not expect more than 100 characters, let's go beyond that.
+ const std::string encoding = provider.ConsumeRandomLengthString(128);
+ auto file_contents = provider.ConsumeRemainingBytes<uint8_t>();
- // Intentionally pass raw data as the API does not require trailing \0.
- FuzzerTemporaryFile file(file_contents, file_contents_size);
+ FuzzerTemporaryFile file(file_contents.data(), file_contents.size());
xmlTextReaderPtr xmlReader =
xmlReaderForFile(file.filename(), encoding.c_str(), options);