Why does the timestamp in ftrace output jump around?
Issue
- Why does the timestamp in ftrace output jump around?
Example:
15d.... 84936.642635: funcgraph_entry: 0.498 us | native_read_cr0();
15d.... 84936.642636: funcgraph_entry: 0.279 us | native_write_cr0();
--> 13d.... 84936.720054: sched_wakeup: 25221:?:? + 25222:39:? java [015] Success
13d.... 84936.721371: sched_wakeup: 25221:?:? + 25222:39:? java [015] Success
14d.... 84936.735276: sched_wakeup: 25210:?:? + 25222:39:? java [015] Success
13d.... 84936.749354: funcgraph_exit: 1.171 us | }
13d.... 84936.749354: funcgraph_entry: | __enqueue_rt_entity() {
13d.... 84936.749354: funcgraph_entry: | cpupri_set() {
--> 13d.... 84936.642664: funcgraph_entry: 0.533 us | _spin_lock_irqsave();
13d.... 84936.642664: funcgraph_entry: 0.237 us | _spin_unlock_irqrestore();
Environment
Red Hat Enterprise Linux (RHEL) 6
Red Hat Enterprise MRG Realtime 1.3
Red Hat Enterprise MRG Realtime 2.0
Subscriber exclusive content
A Red Hat subscription provides unlimited access to our knowledgebase, tools, and much more.