Cyril Hrubis

We are in the middle of cgroup tests rewrite, this script should be removed completely and new test should be written from scratch.

@richiejp And this one. I suppose that we want to re-create the test coverage from scratch then delete these scripts.

@mdoucha I've been thinking along the same lines in issue #462 and I think that the most sane way is to fork for each test iteration. That would fix some...

@mdoucha That's where I ended up as well when I was trying to figure out the semantics. - As for the cleanup, I think that the cleanest solution is to...

I guess that we need a write barrier after the tail update. Does following patch fix the failures for you? ``` diff --git a/testcases/kernel/syscalls/io_uring/io_uring01.c b/testcases/kernel/syscalls/io_uring/io_uring01.c index 70151bb85..c324455aa 100644 --- a/testcases/kernel/syscalls/io_uring/io_uring01.c...

The value is arbitrary indeed and was chosen big enough. Having buf1.tms_stime = 6 just after a fork means that system spend 6 seconds doing something after forking a new...

Ah, right, we have to divide the values with sysconf(_SC_CLK_TCK) to get seconds. I guess that one thing that can be tried is to fork() a fresh process in the...

AFAIK the open posix testsuite build never worked with out-of-tree build. It's mostly separate part that is not well integrated into the rest of the LTP. But looking into the...

I just had a quick look on this issue and found out that this is far more broken than that. The whole open_posix_testsuite/Makefile would have to be fixed so that...

Actually we should probably add a code to detect if the `ip` utility is from busybox or iproute2. Peter can you please have a look at this later on?