Messages in this thread | | | Date | Sun, 29 Mar 2015 17:59:54 -0700 | From | Alexei Starovoitov <> | Subject | Re: [PATCH v11 tip 6/9] samples: bpf: simple non-portable kprobe filter example |
| |
On 3/29/15 5:34 PM, Jovi Zhangwei wrote: >> + snprintf(buf, sizeof(buf), >> + "echo '%c:%s %s' >> /sys/kernel/debug/tracing/kprobe_events", >> + is_kprobe ? 'p' : 'r', event, event); >> + err = system(buf); > > Maybe we need to remember cleanup the kprobe_events in debugfs?
the real tracing tool should be cleaning it up. This is sample code. I didn't want to overcomplicate it with a chain of ctrl-c handlers. Notice patch 7 is simply doing signal(SIGINT, int_exit) and prints histogram when process is terminated. The kprobe cleaning logic would have interfere with this and overall would have made these samples unnecessary complex.
| |