We needn't to account WRITEBACK_TEMP in balance_dirty_pages anymore because of "strictlimit" feature: fuse marks its bdi with BDI_CAP_STRICTLIMIT, so balance_dirty_pages knows that for fuse it should always pay attention to per-bdi limits and counters.
Kudos to Andrey Ryabinin for pointing out. https://jira.sw.ru/browse/PSBM-60701 Signed-off-by: Maxim Patlasov <mpatla...@virtuozzo.com> --- mm/page-writeback.c | 3 +-- 1 file changed, 1 insertion(+), 2 deletions(-) diff --git a/mm/page-writeback.c b/mm/page-writeback.c index fd900c0..fcc0c1b 100644 --- a/mm/page-writeback.c +++ b/mm/page-writeback.c @@ -1534,8 +1534,7 @@ static void balance_dirty_pages(struct address_space *mapping, */ nr_reclaimable = global_page_state(NR_FILE_DIRTY) + global_page_state(NR_UNSTABLE_NFS); - nr_dirty = nr_reclaimable + global_page_state(NR_WRITEBACK) + - global_page_state(NR_WRITEBACK_TEMP); + nr_dirty = nr_reclaimable + global_page_state(NR_WRITEBACK); global_dirty_limits(&background_thresh, &dirty_thresh); _______________________________________________ Devel mailing list Devel@openvz.org https://lists.openvz.org/mailman/listinfo/devel