You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
This breaks when used with event tracing: connection records are created artificially in script land but these do do not have a backing Connection object.
Seems reasonable to check the GetOrigin() return value and triggering a runtime error and return early if it's a nullptr. Currently with event tracing, any bif trying to access the underlying Connection ends up with a null-pointer-deref crash.
We may in the future extend Zeek to be able to create a backing connection in the core for a given connection record. The check seems useful still to prevent a proliferation of null checks in individual BIFs and give a signal when something is totally off.
Today, these BIFs provide an easy way to directly crash Zeek from script land:
As discussed with @vpax externally:
A BIF with a
c: connection
parameter transparently gets access to the underlying/backingConnection *
object of a givenconnection
record.This breaks when used with event tracing:
connection
records are created artificially in script land but these do do not have a backingConnection
object.Seems reasonable to check the
GetOrigin()
return value and triggering a runtime error and return early if it's a nullptr. Currently with event tracing, any bif trying to access the underlyingConnection
ends up with a null-pointer-deref crash.We may in the future extend Zeek to be able to create a backing connection in the core for a given
connection
record. The check seems useful still to prevent a proliferation of null checks in individual BIFs and give a signal when something is totally off.Today, these BIFs provide an easy way to directly crash Zeek from script land:
The text was updated successfully, but these errors were encountered: