From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: X-Spam-Checker-Version: SpamAssassin 3.4.0 (2014-02-07) on aws-us-west-2-korg-lkml-1.web.codeaurora.org Received: from lists.ozlabs.org (lists.ozlabs.org [112.213.38.117]) (using TLSv1.2 with cipher ECDHE-RSA-AES256-GCM-SHA384 (256/256 bits)) (No client certificate requested) by smtp.lore.kernel.org (Postfix) with ESMTPS id 54BD2CA5FC1 for ; Wed, 30 Sep 2026 08:54:12 +0000 (UTC) Received: from boromir.ozlabs.org (localhost [127.0.0.1]) by lists.ozlabs.org (Postfix) with ESMTP id 4hvpkV48zxz2yqT; Wed, 30 Sep 2026 18:54:10 +1000 (AEST) Authentication-Results: lists.ozlabs.org; arc=none smtp.remote-ip=95.215.58.64 ARC-Seal: i=1; a=rsa-sha256; d=lists.ozlabs.org; s=201707; t=1790758450; cv=none; b=fnJrKAWUclM5IfOtYHKPx3QUfQASuWsL2LXdhOqRihG9cA1uyWJ2qxxICV+6kSB8/PZpLyrPthPSn6kLQUB7Mx+3ooOiMthonytwdLFpKegbtpuCXrxqZ5OBuQO4EYTQC4DPnpt53/pGndV5vWG6ICT4pZ39FHSWycCnANqfDxarGuJV/VRMwalZ8UwzJvcbV+C0D+bW7q6Upoj+8vliTfKYsvg62xfeeDinKiobxrWvn5NUWvcmrCQ8hPMSnOia65vMMw8ykjlbruzEPBI8bsFAg/57IwV1Eyj2QSHfeL3pVFPJ8yaHrdHwfL3S+dE8qTxZ/Pnp27GjnkpbWJRm8Q== ARC-Message-Signature: i=1; a=rsa-sha256; d=lists.ozlabs.org; s=201707; t=1790758450; c=relaxed/relaxed; bh=ASh1LR56EfBikA9krZgbBpC09kFpBHQaRZVce4w2D60=; h=From:To:Cc:Subject:Date:Message-ID:In-Reply-To:References: MIME-Version:Content-Type; b=F2PTQOJQVZrRw3rsm5RNrzzTXMwRL9cNHVDv843ULC4gG9HVl2LS8E5mMXUU9PZGVhNM52RbuYd/2h00WQ9xWNVnlNeXmpy7iQdBvHdDa3L16TQr9KIuY+1ZyaGleZGNM5bOCfUUSQMW+wqWX4zLgUATmA8GPBRiF8okRDizj7FA17BtzzlgtS+JPd8IlHA9AILdqTA017cEz6ZjTij19aFFscE2XMSNrBkZ+aW1n+KzzOHB+sEtMNU2dxDniBKxU85GUiO5jhDDor7yQwJj9q119PCumVVA1p5wV5Ow3rmv3jMvCrXJa0TzCe/oI3vD1bVC55jHDG0poGHmw3m/7A== ARC-Authentication-Results: i=1; lists.ozlabs.org; dmarc=pass (p=none dis=none) header.from=linux.dev; dkim=pass (1024-bit key; unprotected) header.d=linux.dev header.i=@linux.dev header.a=rsa-sha256 header.s=key1 header.b=chNeUyIU; dkim-atps=neutral; spf=pass (client-ip=95.215.58.64; helo=mta1.migadu.com; envelope-from=lance.yang@linux.dev; receiver=lists.ozlabs.org) smtp.mailfrom=linux.dev Authentication-Results: lists.ozlabs.org; dmarc=pass (p=none dis=none) header.from=linux.dev Authentication-Results: lists.ozlabs.org; dkim=pass (1024-bit key; unprotected) header.d=linux.dev header.i=@linux.dev header.a=rsa-sha256 header.s=key1 header.b=chNeUyIU; dkim-atps=neutral Authentication-Results: lists.ozlabs.org; spf=pass (sender SPF authorized) smtp.mailfrom=linux.dev (client-ip=95.215.58.64; helo=mta1.migadu.com; envelope-from=lance.yang@linux.dev; receiver=lists.ozlabs.org) Received: from mta1.migadu.com (out-64.mta1.migadu.com [95.215.58.64]) (using TLSv1.3 with cipher TLS_AES_256_GCM_SHA384 (256/256 bits) key-exchange x25519 server-signature RSA-PSS (2048 bits) server-digest SHA256) (No client certificate requested) by lists.ozlabs.org (Postfix) with ESMTPS id 4hvpkP3Jbnz2y24 for ; Wed, 30 Sep 2026 18:53:58 +1000 (AEST) X-Envelope-To: linuxppc-dev@lists.ozlabs.org DKIM-Signature: a=rsa-sha256; bh=x3C9eFMj4s4C+4L8a6ck2RX+0F5Vx+DLli30NnboIh4=; c=simple/simple; d=linux.dev; h=from:to:subject:date:message-id:mime-version:content-type; s=key1; t=1790758418; v=1; x=1791363218; b=chNeUyIUcRJiEYe/HzjokB4VaM4k1Frk2WTry5hanTEkS2tnrOXiHZfITnCCETi2CMJOaiMO a9onNnLbzGsVFzfoiyTFg60d1ZrM5Cu4ff3Nybfa2HFrZRe8xX0droNFettjFw8mGlH/4wzy2hu qR9xeDShk5xtbFQHMl2IcyM0= X-Envelope-To: linuxppc-dev@lists.ozlabs.org Received: by smtp.migadu.com with ESMTPS id d8889dd78a97ab2f; Wed, 30 Sep 2026 08:53:38 +0000 X-Mizu-Trace-ID: d8889dd78a97ab2f X-Migadu-Flow: FLOW_OUT From: Lance Yang To: songmuchun@bytedance.com Cc: maddy@linux.ibm.com, rppt@kernel.org, akpm@linux-foundation.org, david@kernel.org, mpe@ellerman.id.au, npiggin@gmail.com, chleroy@kernel.org, ritesh.list@gmail.com, sshegde@linux.ibm.com, ljs@kernel.org, liam@infradead.org, vbabka@kernel.org, surenb@google.com, mhocko@suse.com, qi.zheng@linux.dev, linuxppc-dev@lists.ozlabs.org, linux-kernel@vger.kernel.org, linux-mm@kvack.org, muchun.song@linux.dev, Lance Yang Subject: Re: [PATCH v3 2/6] mm/sparse-vmemmap: support device DAX in common vmemmap path Date: Wed, 30 Sep 2026 16:53:29 +0800 Message-ID: <20260930085329.17337-1-lance.yang@linux.dev> X-Mailer: git-send-email 2.49.0 In-Reply-To: <20260929053231.66085-3-songmuchun@bytedance.com> References: <20260929053231.66085-3-songmuchun@bytedance.com> X-Mailing-List: linuxppc-dev@lists.ozlabs.org List-Id: List-Help: List-Owner: List-Post: List-Archive: , List-Subscribe: , , List-Unsubscribe: Precedence: list MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit On Tue, Sep 29, 2026 at 01:32:27PM +0800, Muchun Song wrote: >The common vmemmap population path cannot yet handle optimized Device DAX >mappings on its own. It uses pfn_to_zone() to find the shared tail page, >but Device DAX populates its vmemmap at runtime before the ZONE_DEVICE span >is initialized. > >Teach the common path to use device_zone() for runtime optimized vmemmap >population while retaining pfn_to_zone() for early boot. This allows the >same path to support both early boot mappings and Device DAX. > >The backing PFN supplied by the Device DAX-specific population path is no >longer used, allowing the redundant lookup and population code to be >removed later. > >Signed-off-by: Muchun Song >Acked-by: Qi Zheng >--- >v3: >- Collect Acked-by from Qi Zheng > >v2: >- Expand comments around slab initialization to explain zone lookup and > page refcounting (suggested by Qi Zheng) >--- > mm/sparse-vmemmap.c | 64 +++++++++++++++++++++++++-------------------- > 1 file changed, 35 insertions(+), 29 deletions(-) > >diff --git a/mm/sparse-vmemmap.c b/mm/sparse-vmemmap.c >index 8219abc6c3e5..ee4c113ca938 100644 >--- a/mm/sparse-vmemmap.c >+++ b/mm/sparse-vmemmap.c >@@ -237,18 +237,43 @@ static __meminit void *vmemmap_alloc_pte(unsigned long pfn, int node, > struct page *page; > const unsigned int order = pfn_to_section_compound_order(pfn); > >- /* >- * Device DAX still relies on vmemmap_populate_compound_pages() for >- * head/first-tail allocation and tail-page reuse. >- */ > if (!vmemmap_optimizable_pfn(pfn)) > return vmemmap_alloc_block_buf(PAGE_SIZE, node, altmap); > >- zone = pfn_to_zone(pfn, node); >+ /* >+ * Before slab is available, vmemmap optimization is used for early >+ * system RAM, whose zone can be determined from the PFN. >+ * >+ * Once slab is available, only ZONE_DEVICE memory reaches this >+ * optimized population path. Its zone span has not been initialized >+ * while its vmemmap is being populated, so pfn_to_zone() cannot be >+ * used. Obtain ZONE_DEVICE directly from the node instead. >+ */ >+ zone = slab_is_available() ? device_zone(node) : pfn_to_zone(pfn, node); > page = vmemmap_shared_tail_page(order, zone); > if (!page) > return NULL; > >+ /* >+ * During early vmemmap population, the shared tail vmemmap backing >+ * page is allocated from memblock before its struct page can safely >+ * participate in page refcounting. Therefore, no reference can be >+ * held for each shared PTE mapping, and the mappings must be unshared >+ * before the vmemmap is depopulated. >+ * >+ * Once slab is available, the shared backing page is allocated from >+ * the buddy allocator and can be refcounted. Hold one reference for >+ * each shared PTE mapping. The architecture vmemmap teardown drops >+ * the reference through __free_pages() when removing the mapping, >+ * preventing the backing page from being freed while it is shared. >+ * >+ * The backing page may be shared by enough PTE mappings to exhaust >+ * the positive range of its reference count. Stop populating the >+ * vmemmap if another reference cannot be acquired. >+ */ >+ if (slab_is_available() && !try_get_page(page)) >+ return NULL; BTW, shouldn't __add_pages() undo the earlier sections on a population failure? Say the first section is added successfully, but populating the next one fails, e.g. due to an allocation failure: void *memremap_pages(struct dev_pagemap *pgmap, int nid) { ... const int nr_range = pgmap->nr_range; int error, i; ... pgmap->nr_range = 0; error = 0; for (i = 0; i < nr_range; i++) { error = pagemap_range(pgmap, ¶ms, i, nid); if (error) break; pgmap->nr_range++; } if (i < nr_range) { memunmap_pages(pgmap); pgmap->nr_range = nr_range; return ERR_PTR(error); } ... } We still need to undo the sections already added in the failed range, though ... memunmap_pages() won't touch those, since it only removes completed ranges. If the range starts at a section boundary, we'd hit -EEXIST in fill_subsection_map() on retry while those subsection bits are still set. The old DAX path had this issue too. Could we roll back [start_pfn, pfn) in __add_pages() as a separate fix? Something like this: ---8<--- diff --git a/mm/memory_hotplug.c b/mm/memory_hotplug.c index 796af1028ee2..16a0a2c885bc 100644 --- a/mm/memory_hotplug.c +++ b/mm/memory_hotplug.c @@ -380,6 +380,7 @@ EXPORT_SYMBOL_GPL(pfn_to_online_page); int __add_pages(int nid, unsigned long pfn, unsigned long nr_pages, struct mhp_params *params) { + const unsigned long start_pfn = pfn; const unsigned long end_pfn = pfn + nr_pages; unsigned long cur_nr_pages; int err; @@ -417,6 +418,10 @@ int __add_pages(int nid, unsigned long pfn, unsigned long nr_pages, break; cond_resched(); } + + /* Roll back the sections added before the failure. */ + if (err && pfn != start_pfn) + __remove_pages(start_pfn, pfn - start_pfn, altmap, params->pgmap); vmemmap_populate_print_last(); return err; } -- Hope I haven't missed anything :) Cheers, Lance >+ > return page_address(page); > } > >@@ -260,31 +285,12 @@ static pte_t * __meminit vmemmap_pte_populate(pmd_t *pmd, unsigned long addr, in > > if (pte_none(ptep_get(pte))) { > pte_t entry; >+ void *p = vmemmap_alloc_pte(pfn, node, altmap); > >- if (ptpfn == (unsigned long)-1) { >- void *p = vmemmap_alloc_pte(pfn, node, altmap); >- >- if (!p) >- return NULL; >- ptpfn = PHYS_PFN(__pa(p)); >- } else { >- /* >- * When a PTE/PMD entry is freed from the init_mm >- * there's a free_pages() call to this page allocated >- * above. Thus this try_get_page() is paired with the >- * put_page_testzero() on the freeing path. >- * This can only called by certain ZONE_DEVICE path, >- * and through vmemmap_populate_compound_pages() when >- * slab is available. >- * >- * Use try_get_page() to prevent the shared page refcount >- * from overflowing. >- */ >- if (slab_is_available() && >- !try_get_page(pfn_to_page(ptpfn))) >- return NULL; >- } >- entry = pfn_pte(ptpfn, PAGE_KERNEL); >+ if (!p) >+ return NULL; >+ >+ entry = pfn_pte(PHYS_PFN(__pa(p)), PAGE_KERNEL); > set_pte_at(&init_mm, addr, pte, entry); > } else if (WARN_ON_ONCE(vmemmap_optimizable_pfn(pfn))) > return NULL; >-- >2.54.0 > >