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=1794538872; b=AIWeTDSmhiGkmN8mwLQab1DG2GpoA+OMs1FCzs/nY4aJmgAQY/eM4sWTjcBLQTzrZ+Rz DhMt/wtynC2elIyDqR5VVW4C2Fcgdl+BnJlJ6CzKtOxHl2/rLliAI9YAFVzhDH+kduRjh dEnoUu4pDwRU32xeY8xQiAPnmB0NHOZRGXmMYjAk/Tq77Ae0c5NwAdyKO0otItUEgI5QQ lubDRg65Fv7NUDymCNSRG8mfEInRzDRfelhBnleB/pmiEwf8DK5JAsDV6wuUqyEIk7S3U 94OLlS/loQIRVBWCixQTYEImvzEd4M0puBX/cwLJ1HWhCKARnOTyyQTnZOnCdt4M/og== ARC-Message-Signature: i=1; d=subspace.kernel.org; s=arc20261015; a=rsa-sha256; c=relaxed/relaxed; t=1794538872; h=Received:DKIM-Signature:From:To:Cc:Subject:Date:Message-ID:X-Mailer: In-Reply-To:References:MIME-Version:Content-Transfer-Encoding; bh=oBCPQ56tDVZFZjzhbLYtS2SpeoPJaV1DcVuq9Q5q8D0=; b=Y0HMbf1SbcayYDzl+RNsZ/w0VWUMPYaoNhcSA8hLkcUt7scWxcVnfNoHaMVIMlYZYVpX wsQ+OBj+hrrFWDwKddCRrXtQjNfwqMAeFDwewvcYWHoPJ6FaF0i3I3A65paFdLfKTzUzv +OvQBmBpAALMFTyikbB7BQUJvhnHP/QueAachGkK4ZyA6TzBYMZoMPFngU2dVFXdmG/NH SN58/DCN9FlX7U9SFBc0Xr0WkrEIkjP0BAFuXqK3nKqGFhgMp9/be8yZp/+/dLmLD4rL7 X12jdceQveAp3PQE7X8Q8psIZlCdOAvYqzZfug8SPapMO5rSzSqyv7R6XnU/gUULd0Q== 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=labq4gSb; 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 A5883606A321 for ; Fri, 13 Nov 2026 03:01:12 +0000 (UTC) Authentication-Results: smtp.subspace.kernel.org; dkim=pass (2048-bit key) header.d=kernel.org header.i=@kernel.org header.a=rsa-sha256 header.s=k20260515 header.b=labq4gSb Received: by smtp.kernel.org (Postfix) with ESMTPSA id 9593D1F000C4; Fri, 13 Nov 2026 03:01:10 +0000 (UTC) DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=kernel.org; s=k20260515; t=1794538872; bh=oBCPQ56tDVZFZjzhbLYtS2SpeoPJaV1DcVuq9Q5q8D0=; h=From:To:Cc:Subject:Date:In-Reply-To:References; b=labq4gSbOO415PBtlYRt2fulSHxO0PrNCL9vQ9bhlZuxhZU1q/LQ05Sh2LxknC8Fg iWfKDSiIH/flqFcXX4TQ7wL8JBjLFXnIl9upT47ILNgx1ryNpv9kd7kJ6ZP5Vi/Bnq kgWQspoUz/T5KUik2qNMVcdvCcH2HSpGuKNeVUo7r4tZupO6wkUiO4aFP4MNr6gaKx boiZEy5QVkJqD0+R0Y4BBaDP9uceRjG9eDyv/mmp+Yrn5yZOICh+skA2D/BUtWq3WN v6MC2Cg3vbhr+UEErlQfl2OR7X+I8vxTlAJg/ea2eF7Dk4N8CTUKfqflBpnJP2p/OW J6vRv9R6vOA3w== From: Sasha Levin To: stable@vger.kernel.org Cc: Johannes Weiner , Andrew Morton , Sashiko , Balbir Singh , "Vlastimil Babka (SUSE)" , "Matthew Wilcox (Oracle)" , Salvatore Dipietro , Christoph Hellwig , Dave Chinner , Matt Fleming , Mel Gorman , Michal Hocko , Suren Baghdasaryan , Brendan Jackman , Zi Yan , Sasha Levin Subject: [PATCH 5.15.y] mm: page_alloc: do not give all non-blocking requests reserve access Date: Thu, 12 Nov 2026 22:01:09 -0500 Message-ID: <20261113030048.989971-1-sashal@kernel.org> X-Mailer: git-send-email 2.53.0 In-Reply-To: <20261109202217-failed-5.15-6ee3d87b9408@kernel.org> References: <20261109202217-failed-5.15-6ee3d87b9408@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: Johannes Weiner [ Upstream commit 6ee3d87b940848aaed5d4e8d005c5f674bf2637a ] Patch series "mm: page_alloc: fix reserve access for non-blocking requests", v2. Two fixes for how non-blocking allocations interact with the highatomic reserve. There are prerequisites for Salvatore's "mm/page_alloc: avoid direct reclaim and compaction for costly __GFP_NORETRY allocations" [0]. Patch 1 narrows which requests may physically take highatomic pages, patch 2 narrows which requests may count them in their watermark check. Both are prerequisites for Salvatore's patch: it clears __GFP_DIRECT_RECLAIM for costly __GFP_NORETRY requests, which turns every large page cache folio allocation that reaches the slowpath into a non-blocking one. Without patch 1, those requests may fall back to the highatomic freelists reserved for atomic requests. Without patch 2, they count the free highatomic pages in their watermark check while being served from regular memory, so they can consume regular high-order blocks below the min watermark by the size of the reserve. GFP_NOWAIT requests hit both paths today, so these are existing problems worthy of their own stable backports; but Salvatore's patch would widen them significantly. Patch 2 reproduction: a GFP_NOWAIT burst under memory pressure can drain regular free memory to zero while the highatomic reserve sits idle. Measured on a 32G box with an anonymous working set, swap, and a full 290M reserve, three synthetic bursts drove the 28G Normal zone's regular free memory to 28M, 0.8M and 0.6M against a 60M min watermark. Swapout failed its swap table allocation inside reclaim, reclaim scanned 13M pages to free 200k, and page faults stalled for tens to hundreds of milliseconds. With the fix the burst stops at the min watermark and nothing else happens. This was found in review and reproduced synthetically; it has not been observed in production. GFP_ATOMIC order-0 requests have a milder version of the same problem: They are entitled to the reserve but are served regular memory first, so the fix there has to change the watermark check and the freelist order together. Salvatore's change doesn't interact with this, though; fixing that is left for a separate patch. This patch (of 2): 281dd25c1a01 ("mm/page_alloc: let GFP_ATOMIC order-0 allocs access highatomic reserves") accidentally gave all non-blocking allocation requests access to highatomic reserves, which includes GFP_NOWAIT and other sites clearing __GFP_DIRECT_RECLAIM. While this improved atomic request success rate, it's unintentionally broad and can actually worsen highatomic requests by squandering the reserves on requests that don't need it. There are concurrent efforts to clear __GFP_DIRECT_RECLAIM altogether for costly order __GFP_NORETRY requests, which would then also fall into this exemption and deplete reserves even faster. What GFP_ATOMIC has over the others is __GFP_HIGH, which alloc_flags_slowpath() translates to ALLOC_MIN_RESERVE. Narrow the exemption to contexts that have both ALLOC_NON_BLOCK | ALLOC_MIN_RESERVE. Link: https://lore.kernel.org/20261001082152.2879289-1-dipiets@amazon.it [0] Link: https://lore.kernel.org/linux-mm/arFBWvzXPYrKuIYi@cmpxchg.org/ [1] Link: https://lore.kernel.org/linux-mm/20261001082152.2879289-1-dipiets@amazon.it/ [2] Link: https://lore.kernel.org/20261028171719.2043421-1-hannes@cmpxchg.org Link: https://lore.kernel.org/20261028171719.2043421-2-hannes@cmpxchg.org Fixes: 281dd25c1a01 ("mm/page_alloc: let GFP_ATOMIC order-0 allocs access highatomic reserves") Signed-off-by: Johannes Weiner Signed-off-by: Andrew Morton Reported-by: Sashiko Closes: https://sashiko.dev/#/patchset/20260904115629.3993331-1-dipiets@amazon.it Acked-by: Balbir Singh Reviewed-by: Vlastimil Babka (SUSE) Cc: Matthew Wilcox (Oracle) Cc: Salvatore Dipietro Cc: Christoph Hellwig Cc: Dave Chinner Cc: Matt Fleming Cc: Mel Gorman Cc: Michal Hocko Cc: Suren Baghdasaryan Cc: Brendan Jackman Cc: Zi Yan Cc: regular free memory to zero while the highatomic reserve sits idle. Measured on a 32G box with an anonymous working set, swap, and a full 290M reserve, three synthetic bursts drove the 28G Normal zone's regular free memory to 28M, 0.8M and 0.6M against a 60M min watermark. Swapout failed its swap table allocation inside reclaim, reclaim scanned 13M pages to free 200k, and page faults stalled for tens to hundreds of milliseconds. [ sashal: placed ALLOC_MASK_ATOMIC in the existing internal header instead of the dedicated page allocator header ] Signed-off-by: Sasha Levin --- mm/internal.h | 3 +++ mm/page_alloc.c | 4 +++- 2 files changed, 6 insertions(+), 1 deletion(-) diff --git a/mm/internal.h b/mm/internal.h index 34b3a16aa01f3..fa8f43cdc39e2 100644 --- a/mm/internal.h +++ b/mm/internal.h @@ -631,6 +631,9 @@ unsigned int reclaim_clean_pages_from_list(struct zone *zone, /* Flags that allow allocations below the min watermark. */ #define ALLOC_RESERVES (ALLOC_NON_BLOCK|ALLOC_MIN_RESERVE|ALLOC_HIGHATOMIC|ALLOC_OOM) +/* Flags that mean GFP_ATOMIC */ +#define ALLOC_MASK_ATOMIC (ALLOC_NON_BLOCK|ALLOC_MIN_RESERVE) + enum ttu_flags; struct tlbflush_unmap_batch; diff --git a/mm/page_alloc.c b/mm/page_alloc.c index 396961cdf93be..b31f50daad319 100644 --- a/mm/page_alloc.c +++ b/mm/page_alloc.c @@ -3633,7 +3633,9 @@ struct page *rmqueue_buddy(struct zone *preferred_zone, struct zone *zone, * reserves as failing now is worse than failing a * high-order atomic allocation in the future. */ - if (!page && (alloc_flags & (ALLOC_OOM|ALLOC_NON_BLOCK))) + if (!page && + ((alloc_flags & ALLOC_OOM) || + (alloc_flags & ALLOC_MASK_ATOMIC) == ALLOC_MASK_ATOMIC)) page = __rmqueue_smallest(zone, order, MIGRATE_HIGHATOMIC); if (!page) { -- 2.53.0