From 3c5194eb2e644b8360a368a1637bb2851be0a9c3 Mon Sep 17 00:00:00 2001 From: Christoph Hellwig Date: Thu, 23 Jul 2026 07:46:04 +0200 Subject: [PATCH] mm/swap: revert to single-folio writes for synchronous swap devices Patch series "swap_ops updates", v2. This series is a follow on to the swap ops series now in mm-unstable. The first patch reintroduces direct folio writes for synchronous swap files, the other two remove the double indirect for file system based swap. This patch (of 3): Kairui Song reported that zram benefits from submitting each folio directly instead of batching up I/O because the classic LRU scanning benefits from clearing the folio writeback bit in the scan loop. Accommodate that by kicking off writes for synchronous devices for each iteration. Link: https://lore.kernel.org/20260723054622.3460249-1-hch@lst.de Link: https://lore.kernel.org/20260723054622.3460249-2-hch@lst.de Signed-off-by: Christoph Hellwig Acked-by: Usama Arif Acked-by: Chris Li Cc: Baoquan He Cc: Kairui Song Cc: Kairui Song Cc: Kemeng Shi Cc: Nhat Pham Cc: Steve French Signed-off-by: Andrew Morton --- mm/page_io.c | 11 ++++++++++- 1 file changed, 10 insertions(+), 1 deletion(-) diff --git a/mm/page_io.c b/mm/page_io.c index e4fa7ffffe8b..c984a4023a65 100644 --- a/mm/page_io.c +++ b/mm/page_io.c @@ -358,7 +358,16 @@ static void swap_add_folio(struct swap_io_ctx *ctx, struct folio *folio, int rw) } bvec_set_folio(&sio->bvecs[sio->nr_bvecs], folio, folio_size(folio), 0); sio->len += folio_size(folio); - if (++sio->nr_bvecs == ARRAY_SIZE(sio->bvecs)) { + + /* + * Write out the iocb if we filled it, or if the device is synchronous. + * + * The latter is to work around expectations in the classic LRU code + * which make synchronous clearing of the folio writeback flag in the + * reclaim path beneficial. + */ + if (++sio->nr_bvecs == ARRAY_SIZE(sio->bvecs) || + (rw == WRITE && (sis->flags & SWP_SYNCHRONOUS_IO))) { if (rw == WRITE) swap_write_submit(ctx); else