On Tue, 19 Nov 2024 15:24:00 +0100 Jean-Michel Hautbois <jeanmichel.hautbois@xxxxxxxxxx> wrote:
Indeed printf gave me clues ! It appears to be a bug in libtracefs (v1.8.1). rtla segfaults when calling tracefs_local_events() in trace_instance_init(). Debugging libtracefs pointed me to the load_events() function, and the segfault happens after tep_parse_event() is called for "/sys/kernel/debug/tracing/events/vmscan/mm_vmscan_write_folio/format". Going through the calls I get to event_read_print_args(). I changed libtraceevent log level to get the warnings, and it says: libtraceevent: Resource temporarily unavailable unknown op '.' Segmentation fault
Can you do me a favor and send me privately a tarball of: # cp -r /sys/kernel/tracing/events /tmp/events # cd /tmp # tar -cvjf events.tar.bz2 events You can't call tar on the /sys/kernel/tracing files as those are pseudo files with size of zero, and tar will just record empty files :-p -- Steve
