Documentation ¶
Index ¶
Constants ¶
View Source
const ( MSG_OP_UNDEF = 0 // MSG_OP_EXECVE event indicates a process was created. The 'PID' // and 'Common' fields will be populated. This event is positioned // after exec() calls have been validated so failed syscalls will // not be reported. To generate events provide the '-e' option to // tetragon, without the option the hook only populates the data // store for later use by above events. MSG_OP_EXECVE = 5 MSG_OP_EXIT = 7 MSG_OP_GENERIC_KPROBE = 13 MSG_OP_GENERIC_TRACEPOINT = 14 MSG_OP_GENERIC_UPROBE = 15 MSG_OP_GENERIC_LSM = 16 // MSG_OP_CLONE notifies user-space that a clone() event has occurred. MSG_OP_CLONE = 23 MSG_OP_DATA = 24 MSG_OP_CGROUP = 25 MSG_OP_LOADER = 26 MSG_OP_THROTTLE = 27 // just for testing MSG_OP_TEST = 254 )
OpCodes must be in sync with msg_ops enum in bpf/lib/msg_types.h and should have a human-readable representation in OpCodeStrings.
Variables ¶
View Source
var OpCodeStrings = map[OpCode]string{ MSG_OP_UNDEF: "Undef", MSG_OP_EXECVE: "Execve", MSG_OP_EXIT: "Exit", MSG_OP_GENERIC_KPROBE: "GenericKprobe", MSG_OP_GENERIC_TRACEPOINT: "GenericTracepoint", MSG_OP_GENERIC_UPROBE: "GenericUprobe", MSG_OP_GENERIC_LSM: "GenericLSM", MSG_OP_CLONE: "Clone", MSG_OP_DATA: "Data", MSG_OP_CGROUP: "Cgroup", MSG_OP_LOADER: "Loader", MSG_OP_THROTTLE: "Throttle", MSG_OP_TEST: "Test", }
Functions ¶
This section is empty.
Types ¶
type CgroupOpCode ¶ added in v0.8.4
type CgroupOpCode int
const ( MSG_OP_CGROUP_UNDEF CgroupOpCode = iota MSG_OP_CGROUP_MKDIR CgroupOpCode = 1 MSG_OP_CGROUP_RMDIR CgroupOpCode = 2 MSG_OP_CGROUP_RELEASE CgroupOpCode = 3 MSG_OP_CGROUP_ATTACH_TASK CgroupOpCode = 10 )
Cgroup Operations that are sent from BPF side. Right now they are used only for logging and debugging, except for CGROUP_ATTACH_TASK which will be used to detect cgroup configuration.
func (CgroupOpCode) String ¶ added in v0.8.4
func (op CgroupOpCode) String() string
type CgroupState ¶ added in v0.8.4
type CgroupState int
const ( CGROUP_UNTRACKED CgroupState = iota // Cgroup was created but we did not track it CGROUP_NEW CgroupState = 1 // Cgroup was just created CGROUP_RUNNING CgroupState = 2 // Cgroup from new => running (fork,exec task inside) CGROUP_RUNNING_PROC CgroupState = 3 // Cgroups that were generated from pids of procfs )
Different cgroup states.
func (CgroupState) String ¶ added in v0.8.4
func (st CgroupState) String() string
Click to show internal directories.
Click to hide internal directories.