Repository navigation
coverage: forward rss_limit_mb from target options - #16464
Open
speranza-survivor wants to merge 1 commit into
Open
speranza-survivor wants to merge 1 commit into
speranza-survivor wants to merge 1 commit into
Conversation
|
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. |
The coverage runner ignores each target's .options file while run_fuzzer applies it, so a target cannot raise the RSS cap for corpus replay. Forward only rss_limit_mb, preserving the default when the option is absent. coverage_test.sh runs the real script against a stub target and asserts the forwarding matrix: present -> forwarded; absent/malformed -> nothing; other options (max_len) never forwarded; existing args preserved.
speranza-survivor
force-pushed
the
coverage-rss-limit-options
branch
from
November 5, 2026 16:05
d8dace6 to
985b2b6
Compare
This branch has not been deployed
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
The coverage runner ignores each target's
.optionsfile:run_fuzz_targetininfra/base-images/base-runner/coveragehardcodeswhile fuzzing runs (
run_fuzzer) do apply.optionsviaparse_options.py.So a target cannot raise the RSS cap for corpus replay, and one currently
sits at 0.0% on the introspector because its coverage task dies at libFuzzer's
built-in 2048 MB default:
(libpcap's fuzz_filter; the libpcap side sets
rss_limit_mbin its.optionsin the-tcpdump-group/libpcap#1803, which only takes effect for the coverage
task once this change lands.)
This change forwards only
rss_limit_mbfrom the target's.optionsto thereplay args, reusing
parse_options.pythe same wayrun_fuzzerdoes. Otherfuzzing options may be unsuitable for corpus replay and are not forwarded.
When the option is absent, or malformed, the previous behavior is unchanged.
coverage_test.sh(added next tocoverage) runs the real script against astub target and asserts:
.optionswithrss_limit_mb = 4096->-rss_limit_mb=4096is passed.options, only other options, or a malformed value -> nothing extramax_len) are never forwarded-merge=1,-timeout=100) are preservedOn the unpatched script the two forwarding cases fail; on the patched script
all five pass.
End-to-end check with a real libpcap coverage build inside base-runner: the
patched script reads the post-#1803
.options, passes-rss_limit_mb=4096to
fuzz_filter, and the full pipeline (merge -> llvm-profdata -> llvm-cov)completes with 8,048 of 15,811 lines covered.
Scope note: this change is verified to make coverage honor the target's
configured RSS limit and to complete the public-corpus coverage pipeline. The
production OOM that motivated the libpcap
.optionschange did not reproducein our environment, so the effect of 4096 MB on it is unverified; the limit is
sized as headroom over the ClusterFuzz-observed 2346 MB peak.