http: explicitly blocking CONNECT for HTTP/1.1 #9637
Conversation
Signed-off-by: Alyssa Wilk <alyssar@chromium.org>
| @@ -29,9 +29,9 @@ constexpr const char* runtime_features[] = { | |||
| "envoy.reloadable_features.buffer_filter_populate_content_length", | |||
| "envoy.reloadable_features.outlier_detection_support_for_grpc_status", | |||
| "envoy.reloadable_features.connection_header_sanitization", | |||
| "envoy.reloadable_features.strict_header_validation", | |||
There was a problem hiding this comment.
Did you mean to delete this? Don't we still need this since it exists in the code? Or should we clean it up in the code also?
There was a problem hiding this comment.
"Also removing a spurious "envoy.reloadable_features.strict_header_validation" as it was enabled twice." :-P
Github truncated but the file was
"envoy.reloadable_features.test_feature_true",
"envoy.reloadable_features.strict_header_validation",
"envoy.reloadable_features.buffer_filter_populate_content_length",
"envoy.reloadable_features.outlier_detection_support_for_grpc_status",
"envoy.reloadable_features.connection_header_sanitization",
"envoy.reloadable_features.strict_header_validation",
There was a problem hiding this comment.
Oops sorry I should read more. :)
| if (downstreamProtocol() == Http::CodecClient::Type::HTTP1) { | ||
| return; | ||
| } |
There was a problem hiding this comment.
nit: I would just do this at the top of the test.
This is continued clean up from red emu, unifying HTTP/1 and HTTP/2 behavior.
Guarded by envoy.reloadable_features.strict_method_validation
Also removing a spurious "envoy.reloadable_features.strict_header_validation" as it was enabled twice.
Risk Level: Medium (someone might depend on current broken behavior)
Testing: new integration tests
Docs Changes: n/a
Release Notes: inline