From mboxrd@z Thu Jan 1 00:00:00 1970 Authentication-Results: smtp.subspace.kernel.org; arc=none smtp.remote-ip=100.103.45.18 ARC-Seal: i=1; d=subspace.kernel.org; s=arc20261015; a=rsa-sha256; cv=none; t=1794796810; b=3PbWQswCGaYiezESTwtv1pBxmwUsmd/BwKsNDy1mjqYJ4G4Hx+60tDApqXc1SwHhqLuV eHIKuX68ULONzeaZFtYlD8tBWAVSp9v/MGF3yLsl1KSggLSRtXFdrUeBmpf5lXlbTWaCL lcytTLXgW66qhhwFC4tNjT2kzuVva1GpcXqfAvZgT5mx5DV5TefR76Fd6Kg/qu6zTMiGO brZgOdK1R8n153AKH5Om7e9ODPCGOBHRNGyEw2d2geRq6M62PXf9V8OLFLxAvkmrHIrzM idPXyJdbuywIsOjhzY6fnLEJvljZrRhDwBQx5jz7o5vC+KRvk0y8tuZJLNRaUJBlcOA== ARC-Message-Signature: i=1; d=subspace.kernel.org; s=arc20261015; a=rsa-sha256; c=relaxed/relaxed; t=1794796810; h=Received:DKIM-Signature:From:To:Cc:Subject:Date:Message-ID:X-Mailer: In-Reply-To:References:MIME-Version:Content-Transfer-Encoding; bh=ZNAXlgmAqdbj8M748qwNcBn2E70EGcrFHX6OURkv0Z8=; b=SfG/5tHMHo527b4ElVnTD0GPNhLuXUuUfmWSPIOrRvD9J+oMjKjmWd1HPBXe83Tt/HSO ZbXoDuPOrMzLERtcc+uxWr5SU1W9dZA+AfFiFqy8tLoZrULQvEVYVOz1HKO8Gz/Jdx9vU EHFtB87hDH49CEQTGAxaxBzelWs2F05MxRJfpULI9uZWckudZjXEDaqt/O/N80jk/kW5m McRDJJzZfjOOzcxoJCMHtlnNFRcTJrX2ItUxwaB2kXtdYa8vC7dmcdJP1XUpvzT0cv7yX xjyaS1FllbF9z/kvabx6yMMR5cJi7LcxCbNWi0uzV//S2QnO34dXeLt19rIJ7nkeNdQ== ARC-Authentication-Results: i=1; smtp.subspace.kernel.org; dkim=pass header.d=kernel.org header.i=@kernel.org header.a=rsa-sha256 header.s=k20260515 header.b=Su1WPxRb; arc=none smtp.remote-ip=100.103.45.18 Received: from smtp.kernel.org (quasi.space.kernel.org [100.103.45.18]) (using TLSv1.3 with cipher TLS_AES_256_GCM_SHA384 (256/256 bits) key-exchange X25519 server-signature ECDSA (prime256v1)) (No client certificate requested) by smtp.subspace.kernel.org (Postfix) with ESMTPS id B715C606A31F for ; Mon, 16 Nov 2026 02:40:10 +0000 (UTC) Authentication-Results: smtp.subspace.kernel.org; dkim=pass (2048-bit key, unprotected) header.d=kernel.org header.i=@kernel.org header.a=rsa-sha256 header.s=k20260515 header.b=Su1WPxRb Received: by smtp.kernel.org (Postfix) with ESMTPSA id 4312D1F000C4; Mon, 16 Nov 2026 02:40:09 +0000 (UTC) DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=kernel.org; s=k20260515; t=1794796810; bh=ZNAXlgmAqdbj8M748qwNcBn2E70EGcrFHX6OURkv0Z8=; h=From:To:Cc:Subject:Date:In-Reply-To:References; b=Su1WPxRb4uW2JTmQgkBqQ5q6N4irjW61vcw2KaBnR6/xQkT4/m8MM4QUph4boa6jp EbSdjKAJi5uZHNf8LlaihCxDeZZ875Qg5WQOibJ+PIpzhvfVzcnryCysi0PnpCvu1m V4n0u8SwFTKtQdLT0L/8yqHP44TJbjoeTqhpnJZGZtnChudDuYCEchSAmhBVNXCPHg TxBeRm2VWu113+l0rMjl/sq+1DoAHsZ5iu8IlBKSmgajOOoNYI9zTzEXmV7HVVv5FE 8sOsQJ3NgDvXEwUDdXCxtAShBxbvfJWNWEPNJe7h+Mj1svnctQ0rSFT6hBh4I2Uf2C tsbS0z/K6x6Nw== From: Sasha Levin To: stable@vger.kernel.org Cc: Chris Mason , Pedro Falcato , "Gregory Price (Meta)" , Muhammad Usama Anjum , Lance Yang , Baolin Wang , "David Hildenbrand (Arm)" , Sasha Levin Subject: [PATCH 5.10.y] mm/khugepaged: preserve dirty bits when collapsing PTE-mapped THPs Date: Sun, 15 Nov 2026 21:40:07 -0500 Message-ID: <20261116023942.2399755-1-sashal@kernel.org> X-Mailer: git-send-email 2.53.0 In-Reply-To: <20261109202751-failed-5.10-fe2817e3d5ea@kernel.org> References: <20261109202751-failed-5.10-fe2817e3d5ea@kernel.org> X-Mailing-List: stable@vger.kernel.org List-Id: List-Post: List-Help: List-Subscribe: List-Unsubscribe: List-Owner: Precedence: list MIME-Version: 1.0 Content-Transfer-Encoding: 8bit From: Chris Mason [ Upstream commit fe2817e3d5ea347214342c9a5cd8bcbff20c0574 ] try_collapse_pte_mapped_thp() clears PTEs without transferring their dirty bits to the folio. A shmem read fault can map a clean folio writable, so a later store only dirties the PTE. Losing that dirty bit allows reclaim to discard modified data. The assumption that regular files are read-only is also obsolete since commit 1b6444331bebb ("mm/khugepaged: enable clean pagecache folio collapse for writable files"). In any case, transferring PTE dirty bits to the folio when removing mappings (and while the folio is still mapped) is the right thing to do. Use get_and_clear_ptes() to collect the dirty bits and mark the folio dirty before unmapping it. Update the comment. This was reproduced with shmem using MADV_COLLAPSE followed by MADV_PAGEOUT. Fix generated by an LLM and sent by Chris Mason in private. Fixes: 27e1f8273113 ("khugepaged: enable collapse pmd for pte-mapped THP") Fixes: 1b6444331bebb ("mm/khugepaged: enable clean pagecache folio collapse for writable files") Cc: stable@vger.kernel.org Signed-off-by: Chris Mason [ david: adjust comment; simplify description; cover file-thp as well ] Reviewed-by: Pedro Falcato Reviewed-by: Gregory Price (Meta) Tested-by: Gregory Price (Meta) Tested-by: Muhammad Usama Anjum Link: https://patch.msgid.link/20261031-fixes_core-v3-17-d6437171ecc3@kernel.org Reviewed-by: Lance Yang Reviewed-by: Baolin Wang Signed-off-by: David Hildenbrand (Arm) [ sashal: adapted batched folio handling to per-PTE clearing with page-based dirty marking ] Signed-off-by: Sasha Levin --- mm/khugepaged.c | 4 ++++ 1 file changed, 4 insertions(+) diff --git a/mm/khugepaged.c b/mm/khugepaged.c index e523bb9381189..0b10b04f5f61f 100644 --- a/mm/khugepaged.c +++ b/mm/khugepaged.c @@ -1518,10 +1518,14 @@ void collapse_pte_mapped_thp(struct mm_struct *mm, unsigned long addr) for (i = 0, addr = haddr, pte = start_pte; i < HPAGE_PMD_NR; i++, addr += PAGE_SIZE, pte++) { struct page *page; + pte_t ptent; if (pte_none(*pte)) continue; page = vm_normal_page(vma, addr, *pte); + ptent = ptep_get_and_clear(mm, addr, pte); + if (pte_dirty(ptent)) + set_page_dirty(hpage); page_remove_rmap(page, false); } -- 2.53.0