Christian Brauner
Christian Brauner
The kernel will refuse to mount sysfs as plain read-write in the container if sys is mounted as nodev, nosuid, noexec on the host but the container doesn't specify the...
Ah, it's missing NOATIME
That's on the host. My point is that when we mount sysfs in the container we need to make sure that we mirror specific mount options that the kernel wants...
Can you try: ``` lxc.mount.auto = proc:rw lxc.mount.entry = sysfs sysfs rw,nosuid,nodev,noexec,noatime ```
@jtlayton was interested in this so I'm pinging him as well.
> > ``` > > /* test_id */23:49 > > Found cgroup2 on /sys/fs/cgroup/unified, unified hierarchy for systemd controller23:49 > > Assertion 'path_equal(p, "/sys/fs/cgroup")' failed at src/test/test-cgroup.c:142, function test_id(). Aborting....
> > > ``` > > > /* test_id */23:49 > > > Found cgroup2 on /sys/fs/cgroup/unified, unified hierarchy for systemd controller23:49 > > > Assertion 'path_equal(p, "/sys/fs/cgroup")' failed at...
> @metan-ucw I got the distro version wrong, the build is done in Fedora rawhide. This is with gcc 14. > > Reproduce instructions: > > ``` > $ git...
Since you're closing all fds you could call it with the `CLOSE_RANGE_UNSHARE` flag, i.e. ``` close_range(4, UINT_MAX, CLOSE_RANGE_UNSHARE) ``` The kernel will detect that you're closing all file descriptors and...
So I've looked through git history and I'm a bit puzzled that so many new keymap additions have a `Co-developed-by:` tag attached to them. As a kernel maintainer and developer...