As we deployed vm.defrag_mode=1 in Meta production, some workloads
regressed with recurring pressure spikes and swap storms (which in turn
triggered userspace OOM rules on pressure and swap utilization levels).

Tracing pinned this to non-movable requests spinning and reclaiming
unproductively when kswapd/kcompactd are overwhelmed. Direct reclaim
predominantly frees up pages in movable blocks, but those requests
cannot use that space under defrag_mode rules; and it is unlikely to
free up whole blocks incidentally for __rmqueue_claim() to work.

This series fixes it by making non-movable requests participate in
pageblock production in the allocator slowpath - meaning, they will
invoke direct reclaim and direct compaction with pageblock_order.

That requires some small-ish adjustments up front in the allocator and
the compaction code: three prep patches and the fix last.

The series has been in production against one of the affected workloads
for several weeks and restores the OOM kill rate to !defrag_mode baseline.

v1: https://lore.kernel.org/20260626182215.1107966-1-hannes@cmpxchg.org/

Changes in v2:
- Patch 3 is now Vlastimil's version of the capture_control move: the
  allocator owns and arms the capture control entirely, and compaction
  just aims it at the zone being compacted.
- Comments on the migration source scan restrictions in patch 2, and
  spelled out "small" as <pageblock_order (Vlastimil, Zi Yan)
- Corrected the Fixes tag in patch 4 - the referenced SHA was from an
  old development branch (Zi Yan)
- Clarified in the cover letter what block production means (Shakeel)
- Added review/ack tags
- Cc stable on the fix (Vlastimil)
- Rebased onto mm-stable

Based on mm-stable (2026-07-06).

 include/linux/compaction.h |  3 +-
 mm/compaction.c            | 96 +++++++++++++++++++++++++---------------------
 mm/internal.h              | 10 ++++-
 mm/page_alloc.c            | 76 ++++++++++++++++++++++++++++++------
 4 files changed, 129 insertions(+), 56 deletions(-)