Re: [RFC PATCH v2 07/18] mm: free the pages waiting for a grace period before isolation fails

[Kiryl Shutsemau <kas@kernel.org>]

On Thu, Nov 05, 2026 at 01:03:37PM +0000, sashiko-bot@kernel.org wrote:
> Thank you for your contribution! Sashiko AI review found 3 potential issue(s) to consider:
> - [Medium] mm: premature NR_FREE_RCU_PAGES decrement in rcu_free_pages() skips sync

Only the zone's last batch can make zone_rcu_free_waiting() read 0
while a work is still freeing it, and the pages reach the buddy lists
right after.  alloc_contig_range() fails as it does on a transient
page reference, and offline_pages() takes one more pass.  I don't
think the window is worth closing.

> - [Medium] mm: stale outer_start in alloc_contig_frozen_range_noprof() after RCU free merge

The first case is real.

In v3 alloc_contig_range() finds outer_start again and re-checks
whenever the check fails, not only when pages are waiting.

The second case is a free racing with isolate_freepages_range(), which
can happen without this series too.

> - [Medium] mm: page_is_unmovable() treats freed compound PG_rcu_free folios as unmovable

Valid.

In v3 the flag stays set until the grace period ends, and
page_is_unmovable() counts a large folio with PG_rcu_free and no
references as movable.

-- 
  Kiryl Shutsemau / Kirill A. Shutemov