(Resending and including the cgroups mailing list this time oops)
While going through the cgroup namespace I found a couple of significant
bugs. The first bug I fix could cause a kernel deadlock. The second
fixes a rare race that if it happens we get insane semantics. The third
removes an allowance that could not possibly be used.
The patches are against v4.7-rc2 where I have been doing my testing
and does not include the recent spin_lock_bh(&css_set_lock) to
spin_lock_irq(&css_set_lock) change. But that should be trivial to
resolve.
If you need me to respin things against your for-4.7-fixes branch I will
be happy to oblige. Likewise if another strategy for fixing the cgroup
namespace locking issues is necessary just let me know.
Eric W. Biederman (3):
cgroupns: Fix the locking in copy_cgroup_ns
cgroupns: Close race between cgroup_post_fork and copy_cgroup_ns
cgroupns: Only allow creation of hierarchies in the initial cgroup namespace
kernel/cgroup.c | 18 ++++++++----------
1 file changed, 8 insertions(+), 10 deletions(-)
Thank you,
Eric Biederman
--
To unsubscribe from this list: send the line "unsubscribe cgroups" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at http://vger.kernel.org/majordomo-info.html
