79 	bool protection = track_protection(counter);  in page_counter_charge()  local
85 		if (protection)  in page_counter_charge()
123 	bool protection = track_protection(counter);  in page_counter_try_charge()  local
155 		if (protection)  in page_counter_try_charge()
296  * protection which is derived from its own memory.min/low, its
300  * The following rules apply to the effective protection values:
302  * 1. At the first level of reclaim, effective protection is equal to
303  *    the declared protection in memory.min and memory.low.
305  * 2. To enable safe delegation of the protection configuration, at
306  *    subsequent levels the effective protection is capped to the
307  *    parent's effective protection.
310  *    user is allowed to overcommit the declared protection at a given
311  *    level. If that is the case, the parent's effective protection is
312  *    distributed to the children in proportion to how much protection
316  *    if one counter claims much more protection than it uses memory,
319  * 4. Conversely, when the declared protection is undercommitted at a
320  *    given level, the distribution of the larger parental protection
321  *    budget is NOT proportional. A counter's protection from a sibling
326  *    of the ancestor's claim to protection, any unutilized -
327  *    "floating" - protection from up the tree is distributed in
328  *    proportion to each counter's *usage*. This makes the protection
330  *    the shared parental protection budget, but it protects the
350 	 * protection than what the parent affords them, distribute  in effective_protection()
354 	 * claimed protection in order to be work-conserving: claimed  in effective_protection()
355 	 * but unused protection is available to siblings that would  in effective_protection()
362 	 * Ok, utilized protection of all children is within what the  in effective_protection()
369 	 * If there is unutilized protection, the cgroup will be fully  in effective_protection()
371 	 * protection than what the group could enjoy in theory. This  in effective_protection()
373 	 * protection is always dependent on how memory is actually  in effective_protection()
379 	 * If the children aren't claiming (all of) the protection  in effective_protection()
387 	 * some cgroups DO claim explicit protection, we don't protect  in effective_protection()