Source Lines of Code
The number of source lines of code in a file.
Interpretation
The attribute that source lines of code is expected to quantify is size. A file that is large is likely to have a high value for the source lines of code metric.
Evidence
Source Lines of Code has been empirically-validated to be associated with historical vulnerabilities in software in the following peer-reviewed research studies:
- Searching for a Needle in a Haystack: Predicting Security Vulnerabilities for Windows Vista [2]
The empirical evidence overwhelming supports the notion that a source code file with high source lines of code is more likely to contain a security vulnerability.
Implications
The security implication(s) of a file having high source lines of code could be one or more of the following:
- Size may make the file unwieldy to review increasing the potential for latent vulnerabilities.
- Size may also make the file difficult to comprehend and subsequently change making likely for developers to introduce vulnerabilities.
Mitigations
The theoretical mitigation to lowering the source lines of code of a file is to have source code files with no code in them at all. However, the theoretical mitigation is not practical, at best, and meaningless, at worst. Therefore, the risk of latent vulnerabilities in a file with high source lines of code could be mitigated using one or more of the following suggestions:
- Modularize the file to distribute the source code contained within it across multiple smaller files.
Implementation
In our implementation of the metric, we use SciTools Understand™ to collect the source lines of code metric from source code files.
The source code of the implementation of the metric will be made available on GitHub. If you need to collect the metric from your project, the implementation will also be made available as a container image on Docker Hub.
Languages
The metric implementation is limited to projects written in C/C++, C#, Ada, Basic, Fortran, Java, Jovial, Pascal, PL/M, Python, VHDL, Cobol, Web.
Example(s)
In this section, we present examples of the metric collected from popular open-source software projects.
Chromium
In this subsection, we present examples of the metric collected from the Chromium, the open-source project behind the Google Chrome web browser.
The metric examples presented here were collected at
6b9bf768231f commit to the master branch of the Chromium source code repository.
Summary
Shown in Figure 1.1 is the distribution of the metric collected from source code files in the Chromium project. Shown in Figure 1.2 is the comparison of the distribution of the metric collected from source code files in the Chromium project that were not historically vulnerable and those that were.
Thresholds
The thresholds of the metric in the Chromium project determined using the approach prescribed by Alves et al. [1] is shown in the table below.
| Metric Range | value < 873 | 873 ≤ value < 1,461 | 1,461 ≤ value < 3,214 | 3,214 ≤ value |
|---|---|---|---|---|
| Risk Level | Low | Medium | High | Critical |
Risky Files
The thresholds are used to classify source code files into appropriate risk levels. Shown below are the top and bottom three source code files from the Chromium project in each of the three non-trivial risk levels.
| Path | Source Lines of Code | Percentile |
|---|---|---|
components/sync/protocol/proto_visitors.h |
873 | 70.0231 |
chrome/browser/chromeos/file_manager/path_util_unittest.cc |
873 | 70.0231 |
third_party/blink/renderer/platform/network/network_state_notifier_test.cc |
873 | 70.0231 | ... |
native_client_sdk/src/libraries/nacl_io/kernel_proxy.cc |
1,459 | 79.9768 |
content/browser/accessibility/dump_accessibility_tree_browsertest.cc |
1,459 | 79.9768 |
gpu/command_buffer/service/gles2_cmd_validation_implementation_autogen.h |
1,460 | 79.9908 |
| Path | Source Lines of Code | Percentile |
|---|---|---|
third_party/protobuf/src/google/protobuf/extension_set.cc |
1,461 | 80.0188 |
components/sync/engine_impl/sync_scheduler_impl_unittest.cc |
1,461 | 80.0188 |
chrome/browser/content_settings/host_content_settings_map_unittest.cc |
1,462 | 80.0329 | ... |
ash/display/display_manager_unittest.cc |
3,157 | 89.9191 |
ash/wm/overview/overview_session_unittest.cc |
3,163 | 89.9494 |
cc/scheduler/scheduler_unittest.cc |
3,172 | 89.9799 |
| Path | Source Lines of Code | Percentile |
|---|---|---|
gpu/command_buffer/client/gles2_cmd_helper_autogen.h |
3,214 | 90.0107 |
content/browser/frame_host/navigation_controller_impl_unittest.cc |
3,223 | 90.0416 |
content/browser/appcache/appcache_update_job_unittest.cc |
3,255 | 90.0728 | ... |
third_party/libxml/src/testapi.c |
31,132 | 98.9888 |
third_party/hunspell/fuzz/hunspell_fuzzer_hunspell_dictionary.h |
37,181 | 99.3454 |
third_party/sqlite/amalgamation/sqlite3.c |
68,243 | 100 |
OpenBSD
In this subsection, we present examples of the metric collected from the UNIX-like operating system developed by the OpenBSD project.
The metric examples presented here were collected at dbdab68da3b
commit to the master branch of the OpenBSD source code repository.
Summary
Shown in Figure 2.1 is the distribution of the metric collected from source code files in the OpenBSD project. Shown in Figure 2.2 is the comparison of the distribution of the metric collected from source code files in the OpenBSD project that were not historically vulnerable and those that were.
Thresholds
The thresholds of the metric in the OpenBSD project determined using the approach prescribed by Alves et al. [1] is shown in the table below.
| Metric Range | value < 2,530 | 2,530 ≤ value < 4,299 | 4,299 ≤ value < 7,755 | 7,755 ≤ value |
|---|---|---|---|---|
| Risk Level | Low | Medium | High | Critical |
Risky Files
The thresholds are used to classify source code files into appropriate risk levels. Shown below are the top and bottom three source code files from the OpenBSD project in each of the three non-trivial risk levels.
| Path | Source Lines of Code | Percentile |
|---|---|---|
gnu/llvm/tools/clang/lib/Parse/ParseObjc.cpp |
2,530 | 70.0137 |
gnu/gcc/gcc/cp/semantics.c |
2,532 | 70.0416 |
sys/dev/pci/drm/i915/intel_ddi.c |
2,533 | 70.0695 | ... |
gnu/gcc/gcc/cp/call.c |
4,273 | 79.8920 |
sys/dev/microcode/symbol/spectrum24t_cf.h |
4,286 | 79.9392 |
gnu/usr.bin/gcc/gcc/c-common.c |
4,295 | 79.9865 |
| Path | Source Lines of Code | Percentile |
|---|---|---|
gnu/usr.bin/binutils-2.17/bfd/xcofflink.c |
4,299 | 80.0338 |
gnu/usr.bin/binutils-2.17/bfd/elfxx-ia64.c |
4,318 | 80.0814 |
gnu/llvm/lib/Target/PowerPC/PPCISelDAGToDAG.cpp |
4,319 | 80.1290 | ... |
sys/dev/pci/if_em_hw.c |
7,579 | 89.7582 |
sys/dev/microcode/myx/ethp_z8e.h |
7,587 | 89.8418 |
gnu/gcc/gcc/combine.c |
7,712 | 89.9267 |
| Path | Source Lines of Code | Percentile |
|---|---|---|
gnu/llvm/lib/Analysis/ScalarEvolution.cpp |
7,755 | 90.0121 |
gnu/llvm/lib/Target/AArch64/AArch64ISelLowering.cpp |
7,803 | 90.0981 |
gnu/gcc/gcc/config/sh/sh.c |
7,811 | 90.1841 | ... |
gnu/usr.bin/binutils-2.17/opcodes/m32c-desc.c |
49,967 | 98.5500 |
sys/dev/microcode/udl/udl_huffman.h |
65,542 | 99.2720 |
gnu/usr.bin/binutils-2.17/opcodes/m32c-opc.c |
66,091 | 100 |
Reference(s)
[1] Tiago L. Alves, Christiaan Ypma, and Joost Visser. 2010. Deriving Metric Thresholds From Benchmark Data. In Proceedings of the 26th International Conference on Software Maintenance (ICSM '10). 1-10. https://doi.org/10.1109/ICSM.2010.5609747
[2] Thomas Zimmermann, Nachiappan Nagappan, and Laurie Williams. 2010. Searching for a Needle in a Haystack: Predicting Security Vulnerabilities for Windows Vista. In Proceedings of the 3rd International Conference on Software Testing, Verification and Validation (ICST '10). 421-428. https://doi.org/10.1109/ICST.2010.32