From nobody Tue Sep 9 21:30:25 2025 Received: from smtp.kernel.org (aws-us-west-2-korg-mail-1.web.codeaurora.org [10.30.226.201]) (using TLSv1.2 with cipher ECDHE-RSA-AES256-GCM-SHA384 (256/256 bits)) (No client certificate requested) by smtp.subspace.kernel.org (Postfix) with ESMTPS id 1E5FD13D521 for ; Sun, 7 Sep 2025 07:00:32 +0000 (UTC) Authentication-Results: smtp.subspace.kernel.org; arc=none smtp.client-ip=10.30.226.201 ARC-Seal: i=1; a=rsa-sha256; d=subspace.kernel.org; s=arc-20240116; t=1757228433; cv=none; b=YETZufxmyZ4GYZq108yXMoo04+FUB59+flWMB4Gf2guFxkCyT/i0WSiJ4vxth8qFK9S0nOWfjfVZhOPEO+tQqbmgfPzwJvTbyqGi98N+GXPSxcbOSvGoKJuXZjiwx2gmfsldY2hAn2gwuCrpZjLY/YhzLZIfFbM7P66cmLzPO1o= ARC-Message-Signature: i=1; a=rsa-sha256; d=subspace.kernel.org; s=arc-20240116; t=1757228433; c=relaxed/simple; bh=7kJZgdLvlkty1R/matrNEswbE0YZDCA8+y+K2Ilbxag=; h=From:To:Cc:Subject:Date:Message-ID:In-Reply-To:References: MIME-Version; b=OVA03dUYbB6wn5Nchpv+aqz0/Ro89RxL14O0QSNlSbCLkV8aHP9yn64YsuNMLrmkwmmAxrO3ZeWNi1zFkDXPTGpCkFelDBi8nyIJJgbmPgBSFch9WJvtHWhHqisKVOfyT3KpE5Nxv7I8+J0qPK0xg4Y7QDWT0gzcNUOEC0Kgoiw= ARC-Authentication-Results: i=1; smtp.subspace.kernel.org; dkim=pass (2048-bit key) header.d=kernel.org header.i=@kernel.org header.b=VlOT3EFe; arc=none smtp.client-ip=10.30.226.201 Authentication-Results: smtp.subspace.kernel.org; dkim=pass (2048-bit key) header.d=kernel.org header.i=@kernel.org header.b="VlOT3EFe" Received: by smtp.kernel.org (Postfix) with ESMTPSA id 8C9F1C4CEF8; Sun, 7 Sep 2025 07:00:29 +0000 (UTC) DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/simple; d=kernel.org; s=k20201202; t=1757228432; bh=7kJZgdLvlkty1R/matrNEswbE0YZDCA8+y+K2Ilbxag=; h=From:To:Cc:Subject:Date:In-Reply-To:References:From; b=VlOT3EFeKTyres4rPOzS3AcP7VlATID/HLbJnnDh+6EybF7zeQv7+4diYrdOTInvV w4me702fgy6FjrPR4FwzHTTFmr3lVzJMmNRCZnqh0NkqX7TC+yzVxBXk1C62NF7sxS yn/HfG1Np32F2F0BESMRSBuw51pYAOVWYEWiF3BDEyt36cwAXoIxrNj+ODbX+v7ZPV pyFVeF0F+isNGfwuRMNgO6oUDW++IlIXsSV5TQBC3QZ6CSg6wC7lR93ggtg3I/Zx7X qJQ5+1W3VTPFZOdeCGaLbIng+d6XQYBGvbnPyLkDDuLmlzW0ERUXdGsasuVtHbptLj dIQQo34KASMOw== From: Mike Rapoport To: Andrew Morton Cc: Alexander Graf , Baoquan He , Changyuan Lyu , Chris Li , Jason Gunthorpe , Mike Rapoport , Pasha Tatashin , Pratyush Yadav , kexec@lists.infradead.org, linux-mm@kvack.org, linux-kernel@vger.kernel.org Subject: [PATCH v2 1/2] kho: add support for preserving vmalloc allocations Date: Sun, 7 Sep 2025 10:00:18 +0300 Message-ID: <20250907070022.2177974-2-rppt@kernel.org> X-Mailer: git-send-email 2.50.1 In-Reply-To: <20250907070022.2177974-1-rppt@kernel.org> References: <20250907070022.2177974-1-rppt@kernel.org> Precedence: bulk X-Mailing-List: linux-kernel@vger.kernel.org List-Id: List-Subscribe: List-Unsubscribe: MIME-Version: 1.0 Content-Transfer-Encoding: quoted-printable Content-Type: text/plain; charset="utf-8" From: "Mike Rapoport (Microsoft)" A vmalloc allocation is preserved using binary structure similar to global KHO memory tracker. It's a linked list of pages where each page is an array of physical address of pages in vmalloc area. kho_preserve_vmalloc() hands out the physical address of the head page to the caller. This address is used as the argument to kho_vmalloc_restore() to restore the mapping in the vmalloc address space and populate it with the preserved pages. Signed-off-by: Mike Rapoport (Microsoft) --- include/linux/kexec_handover.h | 12 ++ kernel/kexec_handover.c | 200 +++++++++++++++++++++++++++++++++ 2 files changed, 212 insertions(+) diff --git a/include/linux/kexec_handover.h b/include/linux/kexec_handover.h index 348844cffb13..b7bf3bf11019 100644 --- a/include/linux/kexec_handover.h +++ b/include/linux/kexec_handover.h @@ -42,8 +42,10 @@ struct kho_serialization; bool kho_is_enabled(void); =20 int kho_preserve_folio(struct folio *folio); +int kho_preserve_vmalloc(void *ptr, phys_addr_t *preservation); int kho_preserve_phys(phys_addr_t phys, size_t size); struct folio *kho_restore_folio(phys_addr_t phys); +void *kho_restore_vmalloc(phys_addr_t preservation); int kho_add_subtree(struct kho_serialization *ser, const char *name, void = *fdt); int kho_retrieve_subtree(const char *name, phys_addr_t *phys); =20 @@ -70,11 +72,21 @@ static inline int kho_preserve_phys(phys_addr_t phys, s= ize_t size) return -EOPNOTSUPP; } =20 +static inline int kho_preserve_vmalloc(void *ptr, phys_addr_t *preservatio= n) +{ + return -EOPNOTSUPP; +} + static inline struct folio *kho_restore_folio(phys_addr_t phys) { return NULL; } =20 +static inline void *kho_restore_vmalloc(phys_addr_t preservation) +{ + return NULL; +} + static inline int kho_add_subtree(struct kho_serialization *ser, const char *name, void *fdt) { diff --git a/kernel/kexec_handover.c b/kernel/kexec_handover.c index ecd1ac210dbd..c4560ff9b1fc 100644 --- a/kernel/kexec_handover.c +++ b/kernel/kexec_handover.c @@ -18,6 +18,7 @@ #include #include #include +#include =20 #include =20 @@ -733,6 +734,205 @@ int kho_preserve_phys(phys_addr_t phys, size_t size) } EXPORT_SYMBOL_GPL(kho_preserve_phys); =20 +struct kho_vmalloc_chunk; + +struct kho_vmalloc_hdr { + DECLARE_KHOSER_PTR(next, struct kho_vmalloc_chunk *); + unsigned int total_pages; /* only valid in the first chunk */ + unsigned int flags; /* only valid in the first chunk */ + unsigned short order; /* only valid in the first chunk */ + unsigned short num_elms; +}; + +#define KHO_VMALLOC_SIZE \ + ((PAGE_SIZE - sizeof(struct kho_vmalloc_hdr)) / \ + sizeof(phys_addr_t)) + +struct kho_vmalloc_chunk { + struct kho_vmalloc_hdr hdr; + phys_addr_t phys[KHO_VMALLOC_SIZE]; +}; + +static_assert(sizeof(struct kho_vmalloc_chunk) =3D=3D PAGE_SIZE); + +#define KHO_VMALLOC_FLAGS_MASK (VM_ALLOC | VM_ALLOW_HUGE_VMAP) + +static struct kho_vmalloc_chunk *new_vmalloc_chunk(struct kho_vmalloc_chun= k *cur) +{ + struct kho_vmalloc_chunk *chunk; + int err; + + chunk =3D kzalloc(PAGE_SIZE, GFP_KERNEL); + if (!chunk) + return NULL; + + err =3D kho_preserve_phys(virt_to_phys(chunk), PAGE_SIZE); + if (err) + goto err_free; + if (cur) + KHOSER_STORE_PTR(cur->hdr.next, chunk); + return chunk; + +err_free: + kfree(chunk); + return NULL; +} + +static void kho_vmalloc_free_chunks(struct kho_vmalloc_chunk *first_chunk) +{ + struct kho_mem_track *track =3D &kho_out.ser.track; + struct kho_vmalloc_chunk *chunk =3D first_chunk; + + while (chunk) { + unsigned long pfn =3D PHYS_PFN(virt_to_phys(chunk)); + struct kho_vmalloc_chunk *tmp =3D chunk; + + __kho_unpreserve(track, pfn, pfn + 1); + + chunk =3D KHOSER_LOAD_PTR(chunk->hdr.next); + kfree(tmp); + } +} + +/** + * kho_preserve_vmalloc - preserve memory allocated with vmalloc() across = kexec + * @ptr: pointer to the area in vmalloc address space + * @preservation: returned physical address of preservation metadata + * + * Instructs KHO to preserve the area in vmalloc address space at @ptr. The + * physical pages mapped at @ptr will be preserved and on successful return + * @preservation will hold the physical address of a structure that descri= bes + * the preservation. + * + * NOTE: The memory allocated with vmalloc_node() variants cannot be relia= bly + * restored on the same node + * + * Return: 0 on success, error code on failure + */ +int kho_preserve_vmalloc(void *ptr, phys_addr_t *preservation) +{ + struct kho_mem_track *track =3D &kho_out.ser.track; + struct kho_vmalloc_chunk *chunk, *first_chunk; + struct vm_struct *vm =3D find_vm_area(ptr); + unsigned int order, flags; + int err; + + if (!vm) + return -EINVAL; + + if (vm->flags & ~KHO_VMALLOC_FLAGS_MASK) + return -EOPNOTSUPP; + + flags =3D vm->flags & KHO_VMALLOC_FLAGS_MASK; + order =3D get_vm_area_page_order(vm); + + chunk =3D new_vmalloc_chunk(NULL); + if (!chunk) + return -ENOMEM; + first_chunk =3D chunk; + first_chunk->hdr.total_pages =3D vm->nr_pages; + first_chunk->hdr.flags =3D flags; + first_chunk->hdr.order =3D order; + + for (int i =3D 0; i < vm->nr_pages; i +=3D (1 << order)) { + phys_addr_t phys =3D page_to_phys(vm->pages[i]); + + err =3D __kho_preserve_order(track, PHYS_PFN(phys), order); + if (err) + goto err_free; + + chunk->phys[chunk->hdr.num_elms] =3D phys; + chunk->hdr.num_elms++; + if (chunk->hdr.num_elms =3D=3D ARRAY_SIZE(chunk->phys)) { + chunk =3D new_vmalloc_chunk(chunk); + if (!chunk) + goto err_free; + } + } + + *preservation =3D virt_to_phys(first_chunk); + return 0; + +err_free: + kho_vmalloc_free_chunks(first_chunk); + return err; +} +EXPORT_SYMBOL_GPL(kho_preserve_vmalloc); + +/** + * kho_restore_vmalloc - recreates and populates an area in vmalloc address + * space from the preserved memory. + * @preservation: physical address of the preservation metadata. + * + * Recreates an area in vmalloc address space and populates it with memory= that + * was preserved using kho_preserve_vmalloc(). + * + * Return: pointer to the area in the vmalloc address space, NULL on failu= re. + */ +void *kho_restore_vmalloc(phys_addr_t preservation) +{ + struct kho_vmalloc_chunk *chunk =3D phys_to_virt(preservation); + unsigned int align, order, shift, flags; + unsigned int idx =3D 0, nr; + unsigned long addr, size; + struct vm_struct *area; + struct page **pages; + int err; + + flags =3D chunk->hdr.flags; + if (flags & ~KHO_VMALLOC_FLAGS_MASK) + return NULL; + + nr =3D chunk->hdr.total_pages; + pages =3D kvmalloc_array(nr, sizeof(*pages), GFP_KERNEL); + if (!pages) + return NULL; + order =3D chunk->hdr.order; + shift =3D PAGE_SHIFT + order; + align =3D 1 << shift; + + while (chunk) { + struct page *page; + + for (int i =3D 0; i < chunk->hdr.num_elms; i++) { + phys_addr_t phys =3D chunk->phys[i]; + + for (int j =3D 0; j < (1 << order); j++) { + page =3D phys_to_page(phys); + kho_restore_page(page, 0); + pages[idx++] =3D page; + phys +=3D PAGE_SIZE; + } + } + + page =3D virt_to_page(chunk); + chunk =3D KHOSER_LOAD_PTR(chunk->hdr.next); + kho_restore_page(page, 0); + __free_page(page); + } + + area =3D __get_vm_area_node(nr * PAGE_SIZE, align, shift, flags, + VMALLOC_START, VMALLOC_END, NUMA_NO_NODE, + GFP_KERNEL, __builtin_return_address(0)); + if (!area) + goto err_free_pages_array; + + addr =3D (unsigned long)area->addr; + size =3D get_vm_area_size(area); + err =3D vmap_pages_range(addr, addr + size, PAGE_KERNEL, pages, shift); + if (err) + goto err_free_vm_area; + + return area->addr; + +err_free_vm_area: + free_vm_area(area); +err_free_pages_array: + kvfree(pages); + return NULL; +} +EXPORT_SYMBOL_GPL(kho_restore_vmalloc); + /* Handling for debug/kho/out */ =20 static struct dentry *debugfs_root; --=20 2.50.1