-
Notifications
You must be signed in to change notification settings - Fork 4.8k
New issue
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? Sign in to your account
ext_proc fuzzer test trigger ENVOY_BUG when clear route cache #27657
ext_proc fuzzer test trigger ENVOY_BUG when clear route cache #27657
Conversation
…tream response Signed-off-by: Yanjun Xiang <[email protected]>
/assign @tyxia As codeowner /wait For CI |
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
Signed-off-by: Yanjun Xiang <[email protected]>
…roc_clear_route_cache Signed-off-by: Yanjun Xiang <[email protected]>
/assign @yanavlasov @htuch @mpwarres @stevenzzzz |
This is the OSS fuzzer report: Here is the crash trace back decode: Command: /mnt/scratch0/clusterfuzz/bot/builds/clusterfuzz-g3-builds_third_party-envoy-src-test-extensions-filters-http-ext_proc_libfuzzer_memory_ext_proc_grpc_fuzz_test_77651789446b3c3a04b9f492ff141f003d437347/revisions/ext_proc_grpc_fuzz_test -runs=100 /mnt/scratch0/clusterfuzz/bot/inputs/fuzzer-testcases/input.test | Time ran: 57.5866904258728 Command: /mnt/scratch0/clusterfuzz/bot/builds/clusterfuzz-g3-builds_third_party-envoy-src-test-extensions-filters-http-ext_proc_libfuzzer_memory_ext_proc_grpc_fuzz_test_77651789446b3c3a04b9f492ff141f003d437347/revisions/ext_proc_grpc_fuzz_test -runs=100 /mnt/scratch0/clusterfuzz/bot/inputs/fuzzer-testcases/input.test |
This is the fuzzer test case: ext_proc_data: "scterpc_cre:csp_yilooo0\n\000*!pV1:ae!FoFFF,F\n" |
Nope, the refresh in response phase still may changes the roue and effects the per filter config/metadata... etc that filters may need to use. Apparently it's wrong, so, we should never do this refresh in the response phase. |
Yes, per filter config can be used in response path. What I meant about "no-op" is that before your PR #26045, such behavior will not trigger the Envoy_Bug. Thanks for sharing more context! |
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
/wait
Signed-off-by: Yanjun Xiang <[email protected]>
CC @envoyproxy/api-shepherds: Your approval is needed for changes made to |
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
/wait
Signed-off-by: Yanjun Xiang <[email protected]>
Signed-off-by: Yanjun Xiang <[email protected]>
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
LGTM, modulo nit.
Thanks!
Signed-off-by: Yanjun Xiang <[email protected]>
Signed-off-by: Yanjun Xiang <[email protected]>
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
Thanks!
@envoyproxy/api-shepherds PTAL |
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
/wait-any
… it completely anyway Signed-off-by: Yanjun Xiang <[email protected]>
Kind ping! |
LGTM |
Kind ping! |
…roxy#27657) * ext_proc fuzzer test trigger ENVOY_BUG when clear route cache for upstream response Signed-off-by: Yanjun Xiang <[email protected]> Signed-off-by: asheryer <[email protected]>
…roxy#27657) * ext_proc fuzzer test trigger ENVOY_BUG when clear route cache for upstream response Signed-off-by: Yanjun Xiang <[email protected]> Signed-off-by: Ryan Eskin <[email protected]>
This PR is to address an ENVOY_BUG crash exposed by ext_proc fuzzer test reported by:
https://clusterfuzz.corp.google.com/testcase-detail/4756668218736640
The steps to trigger this ENVOY_BUG is below:
Commit Message:
Additional Description:
Risk Level:
Testing:
Docs Changes:
Release Notes:
Platform Specific Features:
[Optional Runtime guard:]
[Optional Fixes #Issue]
[Optional Fixes commit #PR or SHA]
[Optional Deprecated:]
[Optional API Considerations:]