On 2018-11-19 10:30:52 +0100, Vincent Lefevre wrote: > On 2018-11-19 09:58:04 +0100, Vincent Lefevre wrote: > > Or rebuilding atop with -fsanitize=undefined -fno-sanitize-recover > > would allow one to see where the problem occurs, since one can get a > > negative value: > > > > Malloc failed for -11968 exited processes > > > > probably as the cause of some undefined behavior.
The negative value was due to a type mismatch in a format string. I've fixed 2 of them in my fork and did a pull request. > I've done that, but then, I can't reproduce the problem! > > It might be due to a compiler bug, or because the code is too slow > in this case. I've also noticed that the failure of system/library calls is not checked (see my comments in the upstream bug). In case there's a failure due to some race condition, this could explain the issue. But it may also be due to something else. -- Vincent Lefèvre <vinc...@vinc17.net> - Web: <https://www.vinc17.net/> 100% accessible validated (X)HTML - Blog: <https://www.vinc17.net/blog/> Work: CR INRIA - computer arithmetic / AriC project (LIP, ENS-Lyon)
