Ivan Kelly
Ivan Kelly
Ah, for some reason I thought stretch was also supported. I'm currently writing a test against stretch and a custom OS implementation. If it shows itself to work, I'll create...
the problem appears to be in utils.cpp, get_pids_for_program. The call to ```std_filesystem::canonical(program);``` is what is throwing the exception. Changing that call to a call to abs_path() from the same file...
broken stacktrace appears to be an ebpf issue. the returned stacktraces only have one element.
stacktrace issue was just some silliness with docker build not picking up the latest image. all working as expected now, with the abs_path change
The change to abs_path works for me. I can submit a PR, but I'd like to add some testing to it but not sure where in the tests to add....
I've been looking at writing a runtime test for this, but it doesn't look like it's possible without a significant change to the engine and adding a dependency on docker...
PR opened: https://github.com/iovisor/bpftrace/pull/2662
The workaround for this is to set the env variable BPFTRACE_CACHE_USER_SYMBOLS=NONE
@Johnlon for some reason github didn't send me a notification for this. I'm updating the tutorial for the new api, so will add this fix as I go.
The iterator itself doesn't get dropped, but the future returned by ```iterator.next()``` can be dropped if ```iterval.tick``` complete first. I don't think this is normally a problem because the iterator...