In the program I'm working on, it seems to be triggered by a popen call. If I remove that, the program starts in GDB as it should. So I suspect there's something wrong with one of the lower-level calls used to implement popen, probably vfork or exec.
]]>Starting program: /home/robbie/X-Plane11/X-Plane-x86_64
[Thread debugging using libthread_db enabled]
Using host libthread_db library "/usr/lib/libthread_db.so.1".
[New Thread 0x7fffede3a640 (LWP 169759)]
[Thread 0x7fffede3a640 (LWP 169759) exited]
[New Thread 0x7fffede3a640 (LWP 169760)]
[New Thread 0x7fffecbd1640 (LWP 169761)]
MicroProfile: Web server started on 127.0.0.1:1338
MicroProfile: Web server started on 10.10.101.213:1338
[New Thread 0x7fffe7fff640 (LWP 169762)]
[New Thread 0x7fffe729e640 (LWP 169763)]
..... I deleted some
[New Thread 0x7fff9553b640 (LWP 169782)]
[New Thread 0x7fff7ffff640 (LWP 169783)]
[Detaching after vfork from child process 169784]
[Thread 0x7fff7ffff640 (LWP 169783) exited]
[Thread 0x7fff9553b640 (LWP 169782) exited]
..... I deleted some
[Thread 0x7fffe7fff640 (LWP 169762) exited]
[Thread 0x7fffecbd1640 (LWP 169761) exited]
[Thread 0x7fffede3a640 (LWP 169760) exited]
Program terminated with signal SIGTRAP, Trace/breakpoint trap.
The program no longer exists.
This is pretty quick and during initialization before any plugins are evaluated or anything.
I posted this to Reddit and got one response. Figured I'd run it here before submitting a bug.
https://www.reddit.com/r/archlinux/comments/r0lgkc
Seems like it might be a bug in 5.15.4 and downgrading to 5.15.2 fixes it? I haven't downgraded and will wait because I won't be debugging anything for the holiday weekend and next week anyways due to business trip.
]]>