Getting started with MojoLPM


Note: Using MojoLPM to fuzz your Mojo interfaces is intended to be simple, but there are edge-cases that may require a very detailed understanding of the Mojo implementation to fix. If you run into problems that you can‘t understand readily, send an email to markbrand@google.com and cc fuzzing@chromium.org and we’ll try and help.

Prerequisites: Knowledge of libfuzzer and basic understanding of Protocol Buffers and libprotobuf-mutator. Basic understanding of testing in Chromium.


This document will walk you through:

  • An overview of MojoLPM and what it's used for.
  • Adding a fuzzer to an existing Mojo interface using MojoLPM.

Overview of MojoLPM

MojoLPM is a toolchain for automatically generating structure-aware fuzzers for Mojo interfaces using libprotobuf-mutator as the fuzzing engine.

This tool works by using the existing “grammar” for the interface provided by the .mojom files, and translating that into a Protocol Buffer format that can be fuzzed by libprotobuf-mutator. These protocol buffers are then interpreted by a generated runtime as a sequence of mojo method calls on the targeted interface.

The intention is that using these should be as simple as plugging the generated code in to the existing unittests for those interfaces - so if you've already implemented the necessary mocks to unittest your code, the majority of the work needed to get quite effective fuzzing of your interfaces is already complete!

Choose the Mojo interface(s) to fuzz

If you‘re a developer looking to add fuzzing support for an interface that you’re developing, then this should be very easy for you!

If not, then a good starting point is to search for interfaces in codesearch. The most interesting interfaces from a security perspective are those which are implemented in the browser process and exposed to the renderer process, but there isn't a very simple way to enumerate these, so you may need to look through some of the source code to find an interesting one.

A few of the places which bind many of these cross-privilege interfaces are content/browser/browser_interface_binders.cc and content/browser/render_process_host_impl.cc, specifically RenderProcessHostImpl::RegisterMojoInterfaces.

For the rest of this guide, we'll write a new fuzzer for blink.mojom.CodeCacheHost, which is defined in third_party/blink/public/mojom/loader/code_cache.mojom.

We then need to find the relevant GN build target for this mojo interface so that we know how to refer to it later - in this case that is //third_party/blink/public/mojom:mojom_platform.

Find the implementations of the interfaces

If you are developing these interfaces, then you already know where to find the implementations.

Otherwise a good starting point is to search for references to “public blink::mojom::CodeCacheHost”. Usually there is only a single implementation of a given Mojo interface (there are a few exceptions where the interface abstracts platform specific details, but this is less common). This leads us to content/browser/renderer_host/code_cache_host_impl.h and CodeCacheHostImpl.

Find the unittest for the implementation

Specifically, we're looking for a browser-process side unittest (so not in //third_party/blink). We want the unittest for the browser side implementation of that Mojo interface - in many cases if such exists, it will be directly next to the implementation source, ie. in this case we would be most likely to find them in content/browser/renderer_host/code_cache_host_impl_unittest.cc.

Unfortunately, it doesn‘t look like CodeCacheHostImpl has a unittest, so we’ll have to go through the process of understanding how to create a valid instance ourselves in order to fuzz this interface.

Since this implementation runs in the Browser process, and is part of /content, we're going to create our new fuzzer in /content/test/fuzzer.

Add our testcase proto

First we'll add a proto source file, code_cache_host_mojolpm_fuzzer.proto, which is going to define the structure of our testcases. This is basically boilerplate, but it allows creating fuzzers which interact with multiple Mojo interfaces to uncover more complex issues. For our case, this will be a simple file:

// Copyright 2020 The Chromium Authors. All rights reserved.
// Use of this source code is governed by a BSD-style license that can be
// found in the LICENSE file.

// Message format for the MojoLPM fuzzer for the CodeCacheHost interface.

syntax = "proto2";

package content.fuzzing.code_cache_host.proto;

import "third_party/blink/public/mojom/loader/code_cache.mojom.mojolpm.proto";

// Bind a new CodeCacheHost remote
message NewCodeCacheHostAction {
  required uint32 id = 1;
}

// Run the specific sequence for (an indeterminate) period. This is not
// intended to create a specific ordering, but to allow the fuzzer to delay a
// later task until previous tasks have completed.
message RunThreadAction {
  enum ThreadId {
    IO = 0;
    UI = 1;
  }

  required ThreadId id = 1;
}

// Actions that can be performed by the fuzzer.
message Action {
  oneof action {
    NewCodeCacheHostAction new_code_cache_host = 1;
    RunThreadAction run_thread = 2;
    mojolpm.blink.mojom.CodeCacheHost.RemoteAction
        code_cache_host_remote_action = 3;
  }
}

// Sequence provides a level of indirection which allows Testcase to compactly
// express repeated sequences of actions.
message Sequence {
  repeated uint32 action_indexes = 1 [packed = true];
}

// Testcase is the top-level message type interpreted by the fuzzer.
message Testcase {
  repeated Action actions = 1;
  repeated Sequence sequences = 2;
  repeated uint32 sequence_indexes = 3 [packed = true];
}

This specifies all of the actions that the fuzzer will be able to take - it will be able to create a new CodeCacheHost instance, perform sequences of interface calls on those instances, and wait for various threads to be idle.

In order to build this proto file, we'll need to copy it into the out/ directory so that it can reference the proto files generated by MojoLPM - this will be handled for us by the mojolpm_fuzzer_test build rule.

Add our fuzzer source

Now we're ready to create the fuzzer c++ source file, code_cache_host_mojolpm_fuzzer.cc and the fuzzer build target. This target is going to depend on both our proto file, and on the c++ source file. Most of the necessary dependencies will be handled for us, but we do still need to add some directly.

Note especially the dependency on mojom_platform_mojolpm in blink, this is an autogenerated target where the target containing the generated fuzzer protocol buffer descriptions will be the name of the mojom target with _mojolpm appended.

mojolpm_fuzzer_test("code_cache_host_mojolpm_fuzzer") {
  sources = [
    "code_cache_host_mojolpm_fuzzer.cc"
  ]

  proto_source = "code_cache_host_mojolpm_fuzzer.proto"

   deps = [
    "//base/test:test_support",
    "//content/browser:for_content_tests",
    "//content/public/browser:browser_sources",
    "//content/test:test_support",
    "//services/network:test_support",
    "//storage/browser:test_support",
  ]

  proto_deps = [
    "//third_party/blink/public/mojom:mojom_platform_mojolpm",
  ]
}

Now, the minimal source code to do load our testcases:

// Copyright 2020 The Chromium Authors. All rights reserved.
// Use of this source code is governed by a BSD-style license that can be
// found in the LICENSE file.

#include <stdint.h>
#include <utility>

#include "code_cache_host_mojolpm_fuzzer.pb.h"
#include "mojo/core/embedder/embedder.h"
#include "third_party/blink/public/mojom/loader/code_cache.mojom-mojolpm.h"
#include "third_party/libprotobuf-mutator/src/src/libfuzzer/libfuzzer_macro.h"

DEFINE_BINARY_PROTO_FUZZER(
    const content::fuzzing::code_cache_host::proto::Testcase& testcase) {
}

You should now be able to build and run this fuzzer (it, of course, won't do very much) to check that everything is lined up right so far. Recommended GN arguments:

# DCHECKS are really useful when getting your fuzzer up and running correctly,
# but will often get in the way when running actual fuzzing, so we will disable
# this later.
dcheck_always_on = true
# Without this flag, our fuzzer target won't exist.
enable_mojom_fuzzer = true
# ASAN is super useful for fuzzing, but in this case we just want it to help us
# debug the inevitable lifetime issues while we get everything set-up correctly!
is_asan = true
is_component_build = true
is_debug = false
optimize_for_fuzzing = true
use_goma = true
use_libfuzzer = true

Handle global process setup

Now we need to add some basic setup code so that our process has something that mostly resembles a normal Browser process; if you look in the file this is ContentFuzzerEnvironment, which adds a global environment instance that will handle setting up this basic environment, which will be reused for all of our testcases, since starting threads is expensive and slow. This code should also be responsible for setting up any “stateless” (or more-or-less stateless) code that is required for your interface to run - examples are initializing the Mojo Core, and loading ICU datafiles.

A key difference between our needs here and those of a normal unittest is that we very likely do not want to be running in a special single-threaded mode. We want to be able to trigger issues related to threading, sequencing and ordering, and making sure that the UI, IO and threadpool threads behave as close to a normal browser process as possible is desirable.

It‘s likely better to be conservative here - while it might appear that an interface to be tested has no interaction with the UI thread, and so we could save some resources by only having a real IO thread, it’s often very difficult to establish this with certainty.

In practice, the most efficient way forward will be to copy the existing Environment setup from another MojoLPM fuzzer and adapting that to the context in which the interface to be fuzzed will actually run.

Handle per-testcase setup

We next need to handle the necessary setup to instantiate CodeCacheHostImpl, so that we can actually run the testcases. At this point, we realise that it's likely that we want to be able to have multiple CodeCacheHostImpl's with different render_process_ids and different backing origins, so we need to modify our proto file to reflect this:

message NewCodeCacheHost {
  enum OriginId {
    ORIGIN_A = 0;
    ORIGIN_B = 1;
    ORIGIN_OPAQUE = 2;
    ORIGIN_EMPTY = 3;
  }

  required uint32 id = 1;
  required uint32 render_process_id = 2;
  required OriginId origin_id = 3;
}

Note that we‘re using an enum to represent the origin, rather than a string; it’s unlikely that the true value of the origin is going to be important, so we've instead chosen a few select values based on the cases mentioned in the source.

The first thing that we need to do is set-up the basic Browser process environment; this is what ContentFuzzerEnvironment is doing - this has a basic setup suitable for fuzzing interfaces in /content. A few things to be careful of are that we need to make sure that mojo::core::Init() is called (only once) and we probably want as much freedom as possible in terms of scheduling, so we want to use slightly different threading options than the average unittest. This is a singleton type that will live for the entire duration of the fuzzer process so we don't want to be holding any testcase-specific data here.

The next thing that we need to do is to figure out the basic setup needed to instantiate the interface we‘re interested in. Looking at the constructor for CodeCacheHostImpl we need three things; a valid render_process_id, an instance of CacheStorageContextImpl and an instance of GeneratedCodeCacheContext. CodeCacheHostFuzzerContext is our container for these per-testcase instances; and will handle creating and binding the instances of the Mojo interfaces that we’re going to fuzz. The most important thing to be careful of here is that everything happens on the correct thread/sequence. Many Browser-process objects have specific expectations, and will end up with very different behaviour if they are created or used from the wrong context.

If your test case requires the existence of a RenderFrameHost and similar structures, see content/test/fuzzer/presentation_service_mojolpm_fuzzer.cc for a fuzzer which already sets them up (in particular, inheritance from RenderViewHostTestHarness).

The most important thing to be careful of here is that everything happens on the correct thread/sequence. Many Browser-process objects have specific expectations, and will end up with very different behaviour if they are created or used from the wrong context. Test code doesn't always behave the same way, so try to check the behaviour in the real Browser.

The second most important thing to be aware of is to make sure that the fuzzer has the same control over lifetimes of objects that a renderer process would normally have - the best way to check this is to make sure that you've found and understood the browser process code that would usually bind that interface.

Integrate with the generated MojoLPM fuzzer code

Finally, we need to do a little bit more plumbing, to rig up this infrastructure that we've built together with the autogenerated code that MojoLPM gives us to interpret and run our testcases. This is the CodeCacheHostTestcase, and the part where the magic happens is here:

void CodeCacheHostTestcase::NextAction() {
  if (next_idx_ < testcase_.sequence_indexes_size()) {
    auto sequence_idx = testcase_.sequence_indexes(next_idx_++);
    const auto& sequence =
      testcase_.sequences(sequence_idx % testcase_.sequences_size());
    for (auto action_idx : sequence.action_indexes()) {
      if (!testcase_.actions_size() || ++action_count_ > MAX_ACTION_COUNT) {
        return;
      }
      const auto& action =
          testcase_.actions(action_idx % testcase_.actions_size());
      if (action.ByteSizeLong() > max_action_size_) {
        return;
      }
      switch (action.action_case()) {
        case Action::kNewCodeCacheHost:
          cch_context_.AddCodeCacheHost(
            action.new_code_cache_host().id(),
            action.new_code_cache_host().render_process_id(),
            action.new_code_cache_host().origin_id());
          break;

        case Action::kRunUntilIdle:
          if (action.run_until_idle().id()) {
            content::RunUIThreadUntilIdle();
          } else {
            content::RunIOThreadUntilIdle();
          }
          break;

        case Action::kCodeCacheHostCall:
          mojolpm::HandleRemoteMethodCall(action.code_cache_host_call());
          break;

        case Action::ACTION_NOT_SET:
          break;
      }
    }
  }
}

The key line here in integration with MojoLPM is the last case, kCodeCacheHostCall, where we're asking MojoLPM to treat this incoming proto entry as a call to a method on the CodeCacheHost interface.

There's just a little bit more boilerplate in the bottom of the file to tidy up concurrency loose ends, making sure that the fuzzer components are all running on the correct threads; those are more-or-less common to any fuzzer using MojoLPM.

Resulting structure

Overall, the structure of your fuzzer is likely approximately to reflect that of the content/test/fuzzer/presentation_service_mojolpm_fuzzer.cc, shown here:

alt text

(drawing source here )

Test it!

Make a corpus directory and fire up your shiny new fuzzer!

 ~/chromium/src% out/Default/code_cache_host_mojolpm_fuzzer /dev/shm/corpus
INFO: Seed: 3273881842
INFO: Loaded 1 modules   (1121912 inline 8-bit counters): 1121912 [0x559151a1aea8, 0x559151b2cd20),
INFO: Loaded 1 PC tables (1121912 PCs): 1121912 [0x559151b2cd20,0x559152c4b4a0),
INFO:      146 files found in /dev/shm/corpus
INFO: -max_len is not provided; libFuzzer will not generate inputs larger than 4096 bytes
INFO: seed corpus: files: 146 min: 2b max: 268b total: 8548b rss: 88Mb
#147  INITED cov: 4633 ft: 10500 corp: 138/8041b exec/s: 0 rss: 91Mb
#152  NEW    cov: 4633 ft: 10501 corp: 139/8139b lim: 4096 exec/s: 0 rss: 91Mb L: 98/268 MS: 8 Custom-ChangeByte-Custom-EraseBytes-Custom-ShuffleBytes-Custom-Custom-
#154  NEW    cov: 4634 ft: 10510 corp: 140/8262b lim: 4096 exec/s: 0 rss: 91Mb L: 123/268 MS: 3 CustomCrossOver-ChangeBit-Custom-
#157  NEW    cov: 4634 ft: 10512 corp: 141/8384b lim: 4096 exec/s: 0 rss: 91Mb L: 122/268 MS: 3 CustomCrossOver-Custom-CustomCrossOver-
#158  NEW    cov: 4634 ft: 10514 corp: 142/8498b lim: 4096 exec/s: 0 rss: 91Mb L: 114/268 MS: 1 CustomCrossOver-
#159  NEW    cov: 4634 ft: 10517 corp: 143/8601b lim: 4096 exec/s: 0 rss: 91Mb L: 103/268 MS: 1 Custom-
#160  NEW    cov: 4634 ft: 10526 corp: 144/8633b lim: 4096 exec/s: 0 rss: 91Mb L: 32/268 MS: 1 Custom-
#164  NEW    cov: 4634 ft: 10528 corp: 145/8851b lim: 4096 exec/s: 0 rss: 91Mb L: 218/268 MS: 4 CustomCrossOver-Custom-CustomCrossOver-Custom-

Wait for it...

Let the fuzzer run for a while, and keep periodically checking in in case it‘s fallen over. It’s likely you‘ll have made a few mistakes somewhere along the way but hopefully soon you’ll have the fuzzer running ‘clean’ for a few hours.

If you run into DCHECK failures in deserialization, see the section below marked [triage].

Expand it to include all relevant interfaces

CodeCacheHost is a very simple interface, and it doesn‘t have any dependencies on other interfaces. In reality, most Mojo interfaces are much more complex, and fuzzing their implementations thoroughly will require more work. We’ll take a quick look at a more complex interface, the BlobRegistry. If we look at blob_registry.mojom:

// This interface is the primary access point from renderer to the browser's
// blob system. This interface provides methods to register new blobs and get
// references to existing blobs.
interface BlobRegistry {
  // Registers a new blob with the blob registry.
  // TODO(mek): Make this method non-sync and get rid of the UUID parameter once
  // enough of the rest of the system doesn't rely on the UUID anymore.
  [Sync] Register(pending_receiver<blink.mojom.Blob> blob, string uuid,
                  string content_type, string content_disposition,
                  array<DataElement> elements) => ();

  // Creates a new blob out of a data pipe.
  // |length_hint| is only used as a hint, to decide if the blob should be
  // stored in memory or on disk. Registration will still succeed even if less
  // or more bytes are read from the pipe. The resulting SerializedBlob can be
  // inspected to see how many bytes actually did end up being read from
  // the pipe. Pass 0 if nothing is known about the expected size.
  // If something goes wrong (for example the blob system doesn't have enough
  // available space to store all the data from the stream) null will be
  // returned.
  RegisterFromStream(string content_type, string content_disposition,
                     uint64 length_hint,
                     handle<data_pipe_consumer> data,
                     pending_associated_remote<ProgressClient>? progress_client)
      => (SerializedBlob? blob);

  // Returns a reference to an existing blob. Should not be used by new code,
  // is only exposed to make converting existing blob using code easier.
  // TODO(mek): Remove when crbug.com/740744 is resolved.
  [Sync] GetBlobFromUUID(pending_receiver<Blob> blob, string uuid) => ();

  // Returns a BlobURLStore for a specific origin.
  URLStoreForOrigin(url.mojom.Origin origin,
                    pending_associated_receiver<blink.mojom.BlobURLStore> url_store);
};

We can see that this interface references multiple other interfaces; there are several different kinds of reference that we need to worry about:

Additional fuzzable interfaces - if an interface method can return a pending_remote<> or take a pending_receiver<> to an interface Foo, then we want our fuzzer to fuzz those interfaces too.

Here we would want to add blink.mojom.Blob.RemoteAction and blink.mojom.BlobURLStore.AssociatedRemoteAction to the possible actions that our fuzzer protobufs can take.

Renderer-hosted interfaces - if an interface method takes a pending_remote<> (or returns a pending_receiver<>), then we'll also want to add response handling to our fuzzer. This lets the fuzzer send fuzzer-side implementations of mojo interfaces, and handle fuzzing the values returned if those methods are called.

Here we can see blink.mojom.ProgressClient is needed, but we can also see that we pass blink.mojom.DataElement structures to BlobRegistry.Register. These can contain remote<blink.mojom.Blob>, so we also need to support blink.mojom.Blob.

These are handled similarly to the RemoteActions, but the type that we need to add to our proto is instead blink.mojom.ProgressClient.ReceiverAction, and so on.

We can continue applying this logic recursively to all of the interfaces that might be accessed - this comes down to a question of what dependencies are most likely to be important in getting good coverage, so the later step of examining code coverage may also help in guiding the addition of new interfaces here.

blob_registry_mojolpm_fuzzer.proto illustrates how these responses can be added to the testcase proto.

Start fuzzing

Once the fuzzer is up and running, we probably want to remove dcheck_always_on.

enable_mojom_fuzzer = true
is_asan = true
is_component_build = true
is_debug = false
optimize_for_fuzzing = true
use_goma = true
use_libfuzzer = true

The reason for this is that while DCHECKs are often useful when fuzzing (and a good indication of potential bugs), the Mojo serialization code often contains quite a few DCHECKs, and our fuzzer is essentially serializing untrusted data before it can deserialize that data on the Browser-process side. This means that we can easily get blocked by a “completely valid” DCHECK during serialisation that a compromised renderer would bypass. Removing DCHECKs will sometimes let the fuzzer continue in these situations, and will reduce spurious results, but if your fuzzer doesn‘t trigger any of these cases it may be beneficial to also fuzz with DCHECKs enabled. We’ll discuss this below under triage.

If your coverage isn‘t going up at all, then you’ve probably made a mistake and it likely isn‘t managing to actually interact with the interface you’re trying to fuzz - try using the code coverage output from the next step to debug what's going wrong.

(Optional) Run coverage

In many cases it's useful to check the code coverage to see if we can benefit from adding some manual testcases to get deeper coverage. For this example I used the following gn arguments and command:

enable_mojom_fuzzer = true
is_component_build = false
is_debug = false
use_clang_coverage = true
use_goma = true
use_libfuzzer = true
python tools/code_coverage/coverage.py code_cache_host_mojolpm_fuzzer -b out/Coverage -o ManualReport -c "out/Coverage/code_cache_host_mojolpm_fuzzer -ignore_timeouts=1 -timeout=4 -runs=0 /dev/shm/corpus" -f content

With the CodeCacheHost, looking at the coverage after a few hours we could see that there's definitely some room for improvement:

/* 55       */ base::Optional<GURL> GetSecondaryKeyForCodeCache(const GURL& resource_url,
/* 56 53.6k */ int render_process_id) {
/* 57 53.6k */    if (!resource_url.is_valid() || !resource_url.SchemeIsHTTPOrHTTPS())
/* 58 53.6k */      return base::nullopt;
/* 59 0     */
/* 60 0     */    GURL origin_lock =
/* 61 0     */        ChildProcessSecurityPolicyImpl::GetInstance()->GetOriginLock(
/* 62 0     */            render_process_id);

(Optional) Improve corpus manually

It's fairly easy to improve the corpus manually, since our corpus files are just protobuf files that describe the sequence of interface calls to make.

There are a couple of approaches that we can take here - we‘ll try building a small manual seed corpus that we’ll use to kick-start our fuzzer. Since it's easier to edit text protos, MojoLPM can automatically convert our seed corpus from text protos to binary protos during the build, making this slightly less painful for us, and letting us store our corpus in-tree in a readable format.

So, we'll create a new folder to hold this seed corpus, and craft our first file:

actions {
  new_code_cache_host {
    id: 1
    render_process_id: 0
    origin_id: ORIGIN_A
  }
}
actions {
  code_cache_host_remote_action {
    id: 1
    m_did_generate_cacheable_metadata {
      m_cache_type: CodeCacheType_kJavascript
      m_url {
        new {
          id: 1
          m_url: "http://aaa.com/test"
        }
      }
      m_data {
        new {
          id: 1
          m_bytes {
          }
        }
      }
      m_expected_response_time {
      }
    }
  }
}
sequences {
  action_indexes: 0
  action_indexes: 1
}
sequence_indexes: 0

We can then add some new entries to our build target to have the corpus converted to binary proto directly during build.

  testcase_proto_kind = "content.fuzzing.code_cache_host.proto.Testcase"

  seed_corpus_sources = [
    "code_cache_host_mojolpm_fuzzer_corpus/did_generate_cacheable_metadata.textproto",
  ]

If we now run a new coverage report using this single file seed corpus: (note that the binary corpus files will be output in your output directory, in this case code_cache_host_mojolpm_fuzzer_seed_corpus.zip):

autoninja -C out/Coverage chrome
rm -rf /tmp/corpus; mkdir /tmp/corpus; unzip out/Coverage/code_cache_host_mojolpm_fuzzer_seed_corpus.zip -d /tmp/corpus
python tools/code_coverage/coverage.py code_cache_host_mojolpm_fuzzer -b out/Coverage -o ManualReport -c "out/Coverage/code_cache_host_mojolpm_fuzzer -ignore_timeouts=1 -timeout=4 -runs=0 /tmp/corpus" -f content

We can see that we're now getting some more coverage:

/* 118   */ void CodeCacheHostImpl::DidGenerateCacheableMetadata(
/* 119   */     blink::mojom::CodeCacheType cache_type,
/* 120   */     const GURL& url,
/* 121   */     base::Time expected_response_time,
/* 122 2 */       mojo_base::BigBuffer data) {
/* 123 2 */     if (!url.SchemeIsHTTPOrHTTPS()) {
/* 124 0 */       mojo::ReportBadMessage("Invalid URL scheme for code cache.");
/* 125 0 */       return;
/* 126 0 */     }
/* 127 2 */
/* 128 2 */     DCHECK_CURRENTLY_ON(BrowserThread::UI);
/* 129 2 */
/* 130 2 */     GeneratedCodeCache* code_cache = GetCodeCache(cache_type);
/* 131 2 */     if (!code_cache)
/* 132 0 */       return;
/* 133 2 */
/* 134 2 */     base::Optional<GURL> origin_lock =
/* 135 2 */         GetSecondaryKeyForCodeCache(url, render_process_id_);
/* 136 2 */     if (!origin_lock)
/* 137 0 */       return;
/* 138 2 */
/* 139 2 */     code_cache->WriteEntry(url, *origin_lock, expected_response_time,
/* 140 2 */                            std::move(data));
/* 141 2 */ }

Much better!

Triage notes

MojoLPM fuzzers have a number of common failure modes that are fairly easy to distinguish from real bugs in the implementation being fuzzed.

The first of these is any crash on the fuzzer_thread. Code in the implementation should never, under any circumstances be running on this thread, so any crash on this thread is the result of a bug in the fuzzer itself, or one of the other causes mentioned below.

The second is DCHECK or other failures during Mojo serialization. Various traits assert that they are serializing reasonable values - since we need to reuse this serialization code in the fuzzer to produce input to the implementation, we can trigger these on the fuzzer_thread while processing input to send to the implementation.

The example ASAN error output below illustrates an example of both of these cases - the error happens on the fuzzer_thread, and during serialization.

==2940792==ERROR: AddressSanitizer: ILL on unknown address 0x7fbd9391d0f9 (pc 0x7fbd9391d0f9 bp 0x7fbd24deb3e0 sp 0x7fbd24deb3e0 T5)
    #0 0x7fbd9391d0f9 in unsigned int base::internal::CheckOnFailure::HandleFailure<unsigned int>() base/numerics/safe_conversions_impl.h:122:5
    #1 0x7fbd9391ba78 in unsigned int base::internal::checked_cast<unsigned int, base::internal::CheckOnFailure, unsigned long>(unsigned long) base/numerics/safe_conversions.h:114:16
    #2 0x7fbd9391ba28 in mojo::StructTraits<mojo_base::mojom::BigBufferSharedMemoryRegionDataView, mojo_base::internal::BigBufferSharedMemoryRegion>::size(mojo_base::internal::BigBufferSharedMemoryRegion const&) mojo/public/cpp/base/big_buffer_mojom_traits.cc:17:10
    #3 0x7fbd7f62fc2e in mojo::internal::Serializer<mojo_base::mojom::BigBufferSharedMemoryRegionDataView, mojo_base::internal::BigBufferSharedMemoryRegion>::Serialize(mojo_base::internal::BigBufferSharedMemoryRegion&, mojo::internal::Buffer*, mojo_base::mojom::internal::BigBufferSharedMemoryRegion_Data::BufferWriter*, mojo::internal::SerializationContext*) gen/mojo/public/mojom/base/big_buffer.mojom-shared.h:182:23
...
    #41 0x7fbd955376e8 in base::RunLoop::Run() base/run_loop.cc:124:14
    #42 0x7fbd95707f83 in base::Thread::Run(base::RunLoop*) base/threading/thread.cc:311:13
    #43 0x7fbd95708427 in base::Thread::ThreadMain() base/threading/thread.cc:382:3
    #44 0x7fbd957dfb40 in base::(anonymous namespace)::ThreadFunc(void*) base/threading/platform_thread_posix.cc:81:13
    #45 0x7fbd403866b9 in start_thread /build/glibc-LK5gWL/glibc-2.23/nptl/pthread_create.c:333
AddressSanitizer can not provide additional info.
SUMMARY: AddressSanitizer: ILL (/mnt/scratch0/clusterfuzz/bot/builds/chromium-browser-libfuzzer_linux-release-asan_ae530a86793cd6b8b56ce9af9159ac101396e802/revisions/libfuzzer-linux-release-807440/libmojo_base_shared_typemap_traits.so+0x190f9)
Thread T5 (fuzzer_thread) created by T0 here:
    #0 0x56433ef70b3a in pthread_create third_party/llvm/compiler-rt/lib/asan/asan_interceptors.cpp:214:3
...
    #14 0x56433f15380c in main third_party/libFuzzer/src/FuzzerMain.cpp:19:10
    #15 0x7fbd3c38a82f in __libc_start_main /build/glibc-LK5gWL/glibc-2.23/csu/libc-start.c:291
==2940792==ABORTING