Hi Valentin,
On 12/11/2020 12:46 PM, Valentin Schneider wrote:
On 03/12/20 23:25, Reinette Chatre wrote:
Valentin's series in [2] ends by adding memory barriers to support the
updating of the task_struct from one CPU and the usage of the task_struct data
from another CPU. This work is still needed and as discussed with Valentin in
that thread the work would be re-evaluated by him after seeing how this series
turns out.
Thank you very much for taking a look.
So the "problematic" pattern is still there: a context switch can happen
concurrently with a write to the switching-to-tasks's {closid, rmid}.
Accesses to these fields would thus need to be wrapped by READ_ONCE() &
WRITE_ONCE().
ok.
Thinking a bit more (too much?) about it, we could limit ourselves to
wrapping only reads not protected by the rdtgroup_mutex: the only two
task_struct {closid, rmid} writers are
- rdtgroup_move_task()
- rdt_move_group_tasks()
and they are both invoked while holding said mutex. Thus, a reader holding
the mutex cannot race with a write, so load tearing ought to be safe.
The reads that are not protected by the rdtgroup_mutex can be found in
__resctrl_sched_in(). It thus sounds to me like your proposed changes to
this function found in your patch [1] is what is needed? It is not clear
to me how the pairing would work in this case though. If I understand
correctly the goal is for the write to the closid/rmid in the functions
you mention above to be paired with the reads in resctrl_sched_in() and
it is not clear how adding a single READ_ONCE would accomplish this
pairing by itself.
It is also not entirely clear to me what the problematic scenario could
be. If I understand correctly, the risk is (as you explained in your
commit message), that a CPU could have its {closid, rmid} fields read
locally (resctrl_sched_in()) while they are concurrently being written
to from another CPU (in rdtgroup_move_task() and rdt_move_group_tasks()
as you state above). If this happens then a task being moved may be
scheduled in with its old closid/rmid. The update of closid/rmid in
rdtgroup_move_task()/rdt_move_group_tasks() is followed by
smp_call_function_xx() where the registers are updated with preemption
disabled and thus protected against __switch_to. If a task was thus
incorrectly scheduled in with old closid/rmid, would it not be corrected
at this point?
Thank you
Reinette
[1]
https://lore.kernel.org/lkml/20201123022433.17905-4-valentin.schnei...@arm.com/
