SBFT24 Competition

[phi-go]

This PR combines all fuzzers submitted to SBFT24 and the mutation measurer to allow experiments for the competition.

[google-cla[bot]]

Thanks for your pull request! It looks like this may be your first contribution to a Google open source project. Before we can look at your pull request, you'll need to sign a Contributor License Agreement (CLA).

View this failed invocation of the CLA check for more information.

For the most up to date status, view the checks section at the bottom of the pull request.

[phi-go]

Hey, @Alan32Liu, we would be ready to start experiments. Probably best to start at least the first run with a single benchmark.

The following benchmarks should work, I would start with a smaller one like jsoncpp. zlib_zlib_uncompress_fuzzer, stb_stbi_read_fuzzer, vorbis_decode_fuzzer, jsoncpp_jsoncpp_fuzzer, libpcap_fuzz_both, lcms_cms_transform_fuzzer, libxml2_xml, freetype2_ftfuzzer

/gcbrun run_experiment.py -a --experiment-config /opt/fuzzbench/service/experiment-config.yaml --experiment-name sbft24-jsoncpp --fuzzers aflplusplus libfuzzer mystique pastis bandfuzz tunefuzz fox --benchmarks jsoncpp_jsoncpp_fuzzer --mutation-analysis

Also I can't seem to add a tag for this PR or I would add the SBFT one.

[DonggeLiu]

/gcbrun run_experiment.py -a --experiment-config /opt/fuzzbench/service/experiment-config.yaml --experiment-name sbft24-jsoncpp-01-14 --fuzzers aflplusplus libfuzzer mystique pastis bandfuzz tunefuzz fox --benchmarks jsoncpp_jsoncpp_fuzzer --mutation-analysis

[DonggeLiu]

Experiment sbft24-jsoncpp-01-14 data and results will be available later at: The experiment data. The experiment report.

[phi-go]

@Alan32Liu do you know if there a known problem with aflplusplus? We hit the following exception in two separate experiments:

ERROR:root:Error occurred when generating coverage report. Extras: 
    traceback: Traceback (most recent call last):
  File "/work/src/experiment/measurer/coverage_utils.py", line 74, in generate_coverage_report
    coverage_reporter.generate_coverage_summary_json()
  File "/work/src/experiment/measurer/coverage_utils.py", line 141, in generate_coverage_summary_json
    result = generate_json_summary(coverage_binary,
  File "/work/src/experiment/measurer/coverage_utils.py", line 280, in generate_json_summary
    with open(output_file, 'w', encoding='utf-8') as dst_file:
FileNotFoundError: [Errno 2] No such file or directory: '/work/measurement-folders/jsoncpp_jsoncpp_fuzzer-aflplusplus/merged.json'

[DonggeLiu]

@Alan32Liu do you know if there a known problem with aflplusplus? We hit the following exception in two separate experiments:

Not that I know of. Would you like to test out a different version?

I believe this is the last time it was updated: https://github.com/google/fuzzbench/pull/1936

[phi-go]

Oh, actually there is a build failure for afl++, build-log. Even for the run you started yesterday. Seems like apt update is not part of the dockerfile. So a older version probably won't help.

[phi-go]

@Alan32Liu Would it be possible to start a experiment for all the default benchmarks without using mutation testing as well? Would be nice to have the default experiments at least, as we support only a limited set of benchmarks and sadly won't have time to really expand that until the deadline. Though, I need to fix something before we can do that using this branch.

[DonggeLiu]

@Alan32Liu Would it be possible to start a experiment for all the default benchmarks without using mutation testing as well? Would be nice to have the default experiments at least, as we support only a limited set of benchmarks and sadly won't have time to really expand that until the deadline. Though, I need to fix something before we can do that using this branch.

Sure, we have a pretty recent one at https://github.com/google/fuzzbench/pull/1929#issuecomment-1849164925. Will run another one in this PR and at https://github.com/google/fuzzbench/pull/1945 as a comparison.

[DonggeLiu]

Could you please set this to false? https://github.com/google/fuzzbench/blob/64da23a158642b039751291e3422938dc333e6c6/service/experiment-config.yaml#L18

This merges new exp results with old ones. It's better to disable it for our case because we want the latest pure results without mutation. It's also necessary to disable it for later experiments with mutation.

[phi-go]

@Alan32Liu this version should now work without the --mutation-analysis flag. Could you do a standard coverage run with the competitors fuzzers?

/gcbrun run_experiment.py -a --experiment-config /opt/fuzzbench/service/experiment-config.yaml --experiment-name sbft-standard-cov-01-15 --fuzzers libafl libfuzzer mystique pastis bandfuzz tunefuzz fox

Let's use libafl as a baseline instead of aflplusplus, libafl seems to work fine.

merge_with_nonprivate: true

Is now also set to false.

[DonggeLiu]

/gcbrun run_experiment.py -a --experiment-config /opt/fuzzbench/service/experiment-config.yaml --experiment-name sbft-standard-cov-01-16 --fuzzers libafl libfuzzer mystique pastis bandfuzz tunefuzz fox

[DonggeLiu]

Experiment sbft-standard-cov-01-16 data and results will be available later at: The experiment data. The experiment report.

[kdsjZh]

Hi, I found that tunefuzz on jsoncpp_jsoncpp_fuzzer run failed in sbft standard cov. But it succeed on the last mutation analysis you run, which should be the same program and same setup. Given that I didn't change any setting for the fuzzer, I presume there might be something wrong in the last sbft standard cov trail.

Besides, if I understand correctly, the soft-standard-cov should contain 23 public benchmarks, while there are only 9 available.

Could you check it out a bit? Thanks!

[phi-go]

@kdsjZh I think the experiment is still running, from what I understand not all trials are started at once, there is a limit of parallel runs so it might take a bit until all results show up. Tunefuzz seemed to build jsoncpp without a problem build logs, so I wouldn't worry yet. @Alan32Liu, I hope you can confirm :slightly_smiling_face:

It also says in the report linked that the experiment is still in progress.

[DonggeLiu]

Hi @kdsjZh, thanks for reporting this.

@phi-go

I think the experiment is still running, from what I understand not all trials are started at once, there is a limit of parallel runs so it might take a bit until all results show up.

Yep, by looking at the gcloud logs, I can confirm that the experiment is running.

Tunefuzz seemed to build jsoncpp without a problem build logs, so I wouldn't worry yet.

That's also true.

However, something did draw my attention:

1. Tunefuzz error in doing trails:

Here is an example error log entry:

{
  "insertId": "1m02pjzfgwjx29",
  "jsonPayload": {
    "traceback": "Traceback (most recent call last):\n  File \"/src/experiment/runner.py\", line 483, in experiment_main\n    runner.conduct_trial()\n  File \"/src/experiment/runner.py\", line 280, in conduct_trial\n    self.set_up_corpus_directories()\n  File \"/src/experiment/runner.py\", line 265, in set_up_corpus_directories\n    _unpack_clusterfuzz_seed_corpus(target_binary, input_corpus)\n  File \"/src/experiment/runner.py\", line 136, in _unpack_clusterfuzz_seed_corpus\n    seed_corpus_archive_path = get_clusterfuzz_seed_corpus_path(\n  File \"/src/experiment/runner.py\", line 102, in get_clusterfuzz_seed_corpus_path\n    fuzz_target_without_extension = os.path.splitext(fuzz_target_path)[0]\n  File \"/usr/local/lib/python3.10/posixpath.py\", line 118, in splitext\n    p = os.fspath(p)\nTypeError: expected str, bytes or os.PathLike object, not NoneType\n",
    "component": "runner",
    "trial_id": "2721051",
    "instance_name": "r-sbft-standard-cov-01-16-2721051",
    "fuzzer": "tunefuzz",
    "benchmark": "jsoncpp_jsoncpp_fuzzer",
    "message": "Error doing trial.",
    "experiment": "sbft-standard-cov-01-16"
  },
  "resource": {
    "type": "gce_instance",
    "labels": {
      "zone": "projects/1097086166031/zones/us-central1-c",
      "instance_id": "3117243333354390038",
      "project_id": "fuzzbench"
    }
  },
  "timestamp": "2024-01-16T17:33:02.824333536Z",
  "severity": "ERROR",
  "logName": "projects/fuzzbench/logs/fuzzbench",
  "receiveTimestamp": "2024-01-16T17:33:02.824333536Z"
}

2. Many build failures of other fuzzers. Normally, I would expect libFuzzer to be able to build on all benchmarks, but:

Example:

{
  "insertId": "1m9tb2iffruamu",
  "jsonPayload": {
    "message": "Failed to build benchmark: woff2_convert_woff2ttf_fuzzer, fuzzer: libfuzzer.",
    "experiment": "sbft-standard-cov-01-16",
    "traceback": "Traceback (most recent call last):\n  File \"/work/src/experiment/build/builder.py\", line 191, in build_fuzzer_benchmark\n    buildlib.build_fuzzer_benchmark(fuzzer, benchmark)\n  File \"/work/src/experiment/build/gcb_build.py\", line 140, in build_fuzzer_benchmark\n    _build(config, config_name)\n  File \"/work/src/experiment/build/gcb_build.py\", line 124, in _build\n    raise subprocess.CalledProcessError(result.retcode, command)\nsubprocess.CalledProcessError: Command '['gcloud', 'builds', 'submit', '/work/src', '--config=/tmp/tmpjbm4vf7x', '--timeout=14400s', '--worker-pool=projects/fuzzbench/locations/us-central1/workerPools/buildpool-e2-std-32']' returned non-zero exit status 1.\n",
    "instance_name": "d-sbft-standard-cov-01-16",
    "component": "dispatcher"
  },
  "resource": {
    "type": "gce_instance",
    "labels": {
      "project_id": "fuzzbench",
      "zone": "projects/1097086166031/zones/us-central1-c",
      "instance_id": "7879014797801950091"
    }
  },
  "timestamp": "2024-01-16T13:01:38.517591212Z",
  "severity": "ERROR",
  "logName": "projects/fuzzbench/logs/fuzzbench",
  "receiveTimestamp": "2024-01-16T13:01:38.517591212Z"
}

Same with libAFL:

[DonggeLiu]

@phi-go: Feel free to let me know if you'd like to look into these logs together.

[phi-go]

Yeah, it would be good to look into this. I'm not quite sure what could be the root cause at the moment.

[DonggeLiu]

Yeah, it would be good to look into this. I'm not quite sure what could be the root cause at the moment.

For TuneFuzz, would this help debugging?

Traceback (most recent call last):\n  File \"/src/experiment/runner.py\", line 483, in experiment_main\n    runner.conduct_trial()\n  File \"/src/experiment/runner.py\", line 280, in conduct_trial\n    self.set_up_corpus_directories()\n  File \"/src/experiment/runner.py\", line 265, in set_up_corpus_directories\n    _unpack_clusterfuzz_seed_corpus(target_binary, input_corpus)\n  File \"/src/experiment/runner.py\", line 136, in _unpack_clusterfuzz_seed_corpus\n    seed_corpus_archive_path = get_clusterfuzz_seed_corpus_path(\n  File \"/src/experiment/runner.py\", line 102, in get_clusterfuzz_seed_corpus_path\n    fuzz_target_without_extension = os.path.splitext(fuzz_target_path)[0]\n  File \"/usr/local/lib/python3.10/posixpath.py\", line 118, in splitext\n    p = os.fspath(p)\nTypeError: expected str, bytes or os.PathLike object, not NoneType\n

[phi-go]

The clusterfuzz seed corpus is a source for the initial corpus, correct? While I added timestamps to the corpus collected during the trial I did nothing to the clusterfuzz part. I'll need to take a closer look at the code. Can we confirm that this only happens in this branch and not master?

[DonggeLiu]

Can we confirm that this only happens in this branch and not master?

I think we can. Earlier I prepared these two experiments on master as a comparison: https://github.com/google/fuzzbench/pull/1945#issuecomment-1892842376

TuneFuzz had no such error in that experiment:

BTW, libFuzzer had no build error either:

There were some Coverage run failed., but they were fuzzer runtime errors like ERROR: libFuzzer: out-of-memory (used: 2065Mb; limit: 2048Mb), which is unrelated.

[phi-go]

So it seems that in this code:

    def set_up_corpus_directories(self):
        """Set up corpora for fuzzing. Set up the input corpus for use by the
        fuzzer and set up the output corpus for the first sync so the initial
        seeds can be measured."""
        fuzz_target_name = environment.get('FUZZ_TARGET')
        target_binary = fuzzer_utils.get_fuzz_target_binary(
            FUZZ_TARGET_DIR, fuzz_target_name)
        input_corpus = environment.get('SEED_CORPUS_DIR')
        os.makedirs(input_corpus, exist_ok=True)
        if not environment.get('CUSTOM_SEED_CORPUS_DIR'):
            _unpack_clusterfuzz_seed_corpus(target_binary, input_corpus)
        else:
            _copy_custom_seed_corpus(input_corpus)

The variable target_binary is set to None. Responsible function get_fuzz_target_binary is here: https://github.com/phi-go/fuzzbench/blob/72926c0bdf8614f16adaef2b4cd658e1908f6186/common/fuzzer_utils.py#L73.

In that function the target binary path is only returned if the file exists. So this is probably a build error not specifically a corpus thing. This is under the assumption that FUZZ_TARGET is set.

[phi-go]

I modified part of the Makefile generation to support the mutation testing docker builds, maybe I broke something there. @Alan32Liu could you take a look at the following changes to the files, I thought those should be fine so maybe another pair of eyes would be good:

https://github.com/google/fuzzbench/pull/1941/files#diff-9ba00a2744edb4b6e8a4768b520cd4b147e26ddec73c13337aac6a79ccfa99a0

• docker/generate_makefile.py
• docker/image_types.yaml
• experiment/build/gcb_build.py

[phi-go]

Oh, I see now. On the cloud it seems the mutation analysis build process is used for the fuzzer builds, which is definitely wrong... Though, I don't yet understand why that happens.

https://www.googleapis.com/download/storage/v1/b/fuzzbench-data/o/sbft-standard-cov-01-16%2Fbuild-logs%2Fbenchmark-bloaty_fuzz_target-fuzzer-libfuzzer.txt?generation=1705424198204020&alt=media

[DonggeLiu]

I modified part of the Makefile generation to support the mutation testing docker builds, maybe I broke something there. @Alan32Liu could you take a look at the following changes to the files, I thought those should be fine so maybe another pair of eyes would be good:

https://github.com/google/fuzzbench/pull/1941/files#diff-9ba00a2744edb4b6e8a4768b520cd4b147e26ddec73c13337aac6a79ccfa99a0

• docker/generate_makefile.py
• docker/image_types.yaml
• experiment/build/gcb_build.py

I did not notice anything either: They replicate coverage, and nothing seems too strange.

However, I noticed that TuneFuzz works fine on some benchmarks (e.g., freetype2_ftfuzzer):

In fact, ignoring the build errors, the error in doing trails only occur on benchmark jsoncpp_jsoncpp_fuzzer:

Not too sure about the build errors, though. The log is not very useful:

{
  "insertId": "yys9kff991ita",
  "jsonPayload": {
    "component": "dispatcher",
    "traceback": "Traceback (most recent call last):\n  File \"/work/src/experiment/build/builder.py\", line 191, in build_fuzzer_benchmark\n    buildlib.build_fuzzer_benchmark(fuzzer, benchmark)\n  File \"/work/src/experiment/build/gcb_build.py\", line 140, in build_fuzzer_benchmark\n    _build(config, config_name)\n  File \"/work/src/experiment/build/gcb_build.py\", line 124, in _build\n    raise subprocess.CalledProcessError(result.retcode, command)\nsubprocess.CalledProcessError: Command '['gcloud', 'builds', 'submit', '/work/src', '--config=/tmp/tmprsixo8fm', '--timeout=14400s', '--worker-pool=projects/fuzzbench/locations/us-central1/workerPools/buildpool-e2-std-32']' returned non-zero exit status 1.\n",
    "message": "Failed to build benchmark: curl_curl_fuzzer_http, fuzzer: tunefuzz.",
    "experiment": "sbft-standard-cov-01-16",
    "instance_name": "d-sbft-standard-cov-01-16"
  },
  "resource": {
    "type": "gce_instance",
    "labels": {
      "project_id": "fuzzbench",
      "zone": "projects/1097086166031/zones/us-central1-c",
      "instance_id": "7879014797801950091"
    }
  },
  "timestamp": "2024-01-16T12:42:50.595247299Z",
  "severity": "ERROR",
  "logName": "projects/fuzzbench/logs/fuzzbench",
  "receiveTimestamp": "2024-01-16T12:42:50.595247299Z"
}

I suppose one thing we can do is to add extensive logs related to these build errors (which seem to have more victims and deserve a higher priority). Then we run a simple test experiment and debug with the logs.

[phi-go]

Luckily the build-logs do show something, see my other comment. Though, they are not available on a local build, I'll try to patch that in, maybe I messed something up in the dockerfile dependencies. I can test that locally for now.

Also thank you for taking a look.

[phi-go]

Ok, I can confirm that mutation testing is not used locally to build bloaty_fuzz_target-libfuzzer and more imporantly it builds without a problem. So it should be something in the gcb specific code, which I do not really understand that well. Also the build-logs are truncated so we do not see the remaining info: https://www.googleapis.com/download/storage/v1/b/fuzzbench-data/o/sbft-standard-cov-01-16%2Fbuild-logs%2Fbenchmark-bloaty_fuzz_target-fuzzer-libfuzzer.txt?generation=1705424198204020&alt=media.

However, before I dig deeper we can still complete the evaluation without fixing this. For now we planned to do the mutation analysis part locally on our blades, we do not support that many benchmarks so this is fine. The coverage for all benchmarks we could do an a branch without our changes and only the fuzzer PRs.

[phi-go]

I suppose one thing we can do is to add extensive logs related to these build errors (which seem to have more victims and deserve a higher priority). Then we run a simple test experiment and debug with the logs.

The missing information seems to just be truncated from the build-log, I changed the code a bit to allow storing everything for the gcb_build execute call, I hope that should reveal the missing info. Let's try the simple test experiment, if you feel comfortable with that.

[DonggeLiu]

How about:

--fuzzers libafl libfuzzer tunefuzz pastis
--benchmarks freetype2_ftfuzzer jsoncpp_jsoncpp_fuzzer bloaty_fuzz_target lcms_cms_transform_fuzzer

Because it includes good success/failure comparisons on both fuzzers and benchmarks.

benchmark \ fuzzer	libafl	libfuzzer	tunefuzz	pastis
freetype2_ftfuzzer	Can Run	Can Run	Can Run	Can Run but NaN
jsoncpp_jsoncpp_fuzzer	Can Run	Can Run	Error doing trails	Error doing trails
bloaty_fuzz_target	Cannot Build	Cannot Build	Cannot Build	Cannot Build
lcms_cms_transform_fuzzer	Can Run	Can Run	Can Run	Can Run

Let me know if you'd like to add more.

[phi-go]

Thank you for looking into this so thoroughly. This sounds like a plan. If you want to reduce compute more, even one hour runs and a few trials should give us enough to debug, though, I don't know how to do this with flags.

/gcbrun run_experiment.py -a --mutation-analysis --experiment-config /opt/fuzzbench/service/experiment-config.yaml --experiment-name sbft-dev-01-18 --fuzzers libafl libfuzzer tunefuzz pastis --benchmarks freetype2_ftfuzzer jsoncpp_jsoncpp_fuzzer bloaty_fuzz_target lcms_cms_transform_fuzzer