Baojun Wang

FYI I was able to get around of this by adding `"std::__cmp_cat::.*"` to `blocklist_types`.

@Volker-Weissmann Does adding `pointer` to block list solve this issue?

No worries, I had a similar issue because `bindgen` generates duplicate types (from `typedef`), basically two `pub type iterator` consecutively: ```rust pub type iterator = root::std::_Bit_iterator; ``` I added `iterator`...

@treeowl I’ll try, btw the ci issue should be fixed if you merge master branch

It is true, any `ptrace` operations can fail when `exit_group` happens, this make `ptrace` painful to implement, we'll have to deal with errors in all circumstances. Perhaps wrapping all ptrace...

I believe the issue is that glibc creates some kind of thin wrapper around certain syscalls, if you do `syscall(SYS_timer_create, ...)` there won’t be any issue as I tested. Same...

The first issue is very easy to fix, the second one is harder, will update tomorrow.

minimal reproducer: * run `tty` program ``` ./bin/dettrace tty ``` Report `not a tty` * run ptyprocess (python) ```python #!/usr/bin/env python from ptyprocess import PtyProcessUnicode p = PtyProcessUnicode.spawn(['python']) p.read(20) p.write('6+6\n')...

https://lists.gnu.org/archive/html/screen-devel/2017-03/msg00013.html

should be fixed by 5709e91.