Merge branch 'bpf: add percpu stats for bpf_map'

Anton Protopopov says:

====================
This series adds a mechanism for maps to populate per-cpu counters on
insertions/deletions. The sum of these counters can be accessed by a new kfunc
from map iterator and tracing programs.

The following patches are present in the series:

  * Patch 1 adds a generic per-cpu counter to struct bpf_map
  * Patch 2 adds a new kfunc to access the sum of per-cpu counters
  * Patch 3 utilizes this mechanism for hash-based maps
  * Patch 4 extends the preloaded map iterator to dump the sum
  * Patch 5 adds a self-test for the change

The reason for adding this functionality in our case (Cilium) is to get signals
about how full some heavy-used maps are and what the actual dynamic profile of
map capacity is. In the case of LRU maps this is impossible to get this
information anyhow else. The original presentation can be found here [1].

  [1] https://lpc.events/event/16/contributions/1368/

v4 -> v5:
* don't pass useless empty opts when creating a link, pass NULL (Hou)
* add a debug message (Hou)
* make code more readable (Alexei)
* remove the selftest which only checked that elem_count != NULL

v3 -> v4:
* fix selftests:
  * added test code for batch map operations
  * added a test for BPF_MAP_TYPE_HASH_OF_MAPS (Hou)
  * added tests for BPF_MAP_TYPE_LRU* with BPF_F_NO_COMMON_LRU (Hou)
  * map_info was called multiple times unnecessarily (Hou)
  * small fixes + some memory leaks (Hou)
* fixed wrong error path for freeing a non-prealloc map (Hou)
* fixed counters for batch delete operations (Hou)

v2 -> v3:
- split commits to better represent update logic (Alexei)
- remove filter from kfunc to allow all tracing programs (Alexei)
- extend selftests (Alexei)

v1 -> v2:
- make the counters generic part of struct bpf_map (Alexei)
- don't use map_info and /proc/self/fdinfo in favor of a kfunc (Alexei)
====================

Signed-off-by: Alexei Starovoitov <ast@kernel.org>