mm: fs: invalidate bh_lrus for only cold path

mm: fs: invalidate bh_lrus for only cold path

The kernel test robot reported the regression of fio.write_iops[1] with
commit 8cc621d2f45d ("mm: fs: invalidate BH LRU during page migration").

Since lru_add_drain is called frequently, invalidate bh_lrus there could
increase bh_lrus cache miss ratio, which needs more IO in the end.

This patch moves the bh_lrus invalidation from the hot path( e.g.,
zap_page_range, pagevec_release) to cold path(i.e., lru_add_drain_all,
lru_cache_disable).

Zhengjun Xing confirmed
 "I test the patch, the regression reduced to -2.9%"

[1] https://lore.kernel.org/lkml/20210520083144.GD14190@xsang-OptiPlex-9020/
[2] 8cc621d2f45d, mm: fs: invalidate BH LRU during page migration

Link: https://lkml.kernel.org/r/[email protected]
Signed-off-by: Minchan Kim <[email protected]>
Reported-by: kernel test robot <[email protected]>
Reviewed-by: Chris Goldsworthy <[email protected]>
Tested-by: "Xing, Zhengjun" <[email protected]>
Signed-off-by: Andrew Morton <[email protected]>
Signed-off-by: Linus Torvalds <[email protected]>