zeromq / libzmq

75%

LAST BUILD ON BRANCH master

branch: SELECT

CHANGE BRANCH

x

• No branch selected
• 260121-cmake-install-dest
• 2607-include-new
• WindowsServer2003
• add-unittests-mtrie
• ci
• dev1
• dev2
• dev3
• dev4
• fix-cacheline-size-undefined
• fix-corrupt-re-subs
• fix-gss-leak
• fix-gssapi-auth
• fix-install-cmake-config-on-mingw
• fix-interrupted-calls
• fix-ws-addr-path
• fix/out-event-io-error-assert
• fix/pipe-term-ack-double-free
• fix/stream-hwm-disconnect-4828
• fix/udp-msg-leak-early-return
• fix/udp-out-buffer-overflow
• fix/udp-signed-char-sign-extension
• fix_ctx_get
• increase_msleep_test_mock_pub_sub
• jemc-relicense
• latest_release
• libera
• ls-fix-input-stopped
• make-failing-test-not-hang
• master
• msg_dontroute
• no-allocation-sends
• p-stable-poller
• patch-1
• pre_reformat
• refs/tags/v4.3.5
• revert-1975-master
• revert-2572-rep_leak
• revert-2974-static_init_gcc
• rp42-fix-windows-mt-build
• secpolicy
• show-ref
• test_loopback
• tmp_obs_release_branch
• v4.0.2-test
• v4.2.0
• v4.2.0-rc1
• v4.2.1
• v4.2.2
• v4.2.3
• v4.2.4
• v4.2.5
• v4.3.0
• v4.3.1
• v4.3.2
• v4.3.2-win
• v4.3.3
• v4.3.4
• vsock
• zap-domain-docs

Build # 28712569049

Build Type

push

github

Committed by bluca

Commit Message

Fix GSSAPI ZAP auth: honour negative ZAP responses

When a ZAP handler returned 300/400/500, gssapi_server_t ignored the
denial and transitioned to send_ready anyway, allowing the connection
to proceed unauthenticated.

Root cause: unlike PLAIN/CURVE (which use zap_client_common_handshake_t
and override handle_zap_status_code to update the state machine),
gssapi_server_t inherits directly from zap_client_t whose
handle_zap_status_code only fires a monitor event without touching
the GSSAPI state machine.

Fix both paths where ZAP is consulted:

1. Synchronous reply (process_handshake_command): check status_code
   after receive_and_process_zap_reply() succeeds; go to send_ready
   only on "2xx", otherwise go to the new `closing` state.

2. Asynchronous reply (zap_msg_available): same status_code check
   instead of always setting send_ready.

Add a `closing` state to the state machine and make status() return
mechanism_t::error for it so the session tears down the connection.

Coverage Stats

15151 of 20093 relevant lines covered (75.4%)

30955.76 hits per line