From: Vlastimil Babka <vbabka@suse.cz>
Date: Mon, 14 Oct 2019 14:11:40 -0700
Subject: mm, page_owner: fix off-by-one error in __set_page_owner_handle()
Git-commit: 5556cfe8d994d5e7b4d50fd91597b8dc0b3a82fd
Patch-mainline: v5.5 or v5.4-rc4 (next release)
References: jsc#SLE-8956, bsc#1144653, VM Debug Functionality
Patch series "followups to debug_pagealloc improvements through
page_owner", v3.
These are followups to [1] which made it to Linus meanwhile. Patches 1
and 3 are based on Kirill's review, patch 2 on KASAN request [2]. It
would be nice if all of this made it to 5.4 with [1] already there (or
at least Patch 1).
This patch (of 3):
As noted by Kirill, commit 7e2f2a0cd17c ("mm, page_owner: record page
owner for each subpage") has introduced an off-by-one error in
__set_page_owner_handle() when looking up page_ext for subpages. As a
result, the head page page_owner info is set twice, while for the last
tail page, it's not set at all.
Fix this and also make the code more efficient by advancing the page_ext
pointer we already have, instead of calling lookup_page_ext() for each
subpage. Since the full size of struct page_ext is not known at compile
time, we can't use a simple page_ext++ statement, so introduce a
page_ext_next() inline function for that.
Link: http://lkml.kernel.org/r/20190930122916.14969-2-vbabka@suse.cz
Fixes: 7e2f2a0cd17c ("mm, page_owner: record page owner for each subpage")
Signed-off-by: Vlastimil Babka <vbabka@suse.cz>
Reported-by: Kirill A. Shutemov <kirill@shutemov.name>
Reported-by: Miles Chen <miles.chen@mediatek.com>
Acked-by: Kirill A. Shutemov <kirill.shutemov@linux.intel.com>
Cc: Andrey Ryabinin <aryabinin@virtuozzo.com>
Cc: Dmitry Vyukov <dvyukov@google.com>
Cc: Walter Wu <walter-zh.wu@mediatek.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
---
include/linux/page_ext.h | 8 ++++++++
mm/page_ext.c | 23 +++++++++--------------
mm/page_owner.c | 15 +++++++--------
3 files changed, 24 insertions(+), 22 deletions(-)
--- a/include/linux/page_ext.h
+++ b/include/linux/page_ext.h
@@ -47,6 +47,7 @@ struct page_ext {
unsigned long flags;
};
+extern unsigned long page_ext_size;
extern void pgdat_page_ext_init(struct pglist_data *pgdat);
#ifdef CONFIG_SPARSEMEM
@@ -63,6 +64,13 @@ static inline void page_ext_init(void)
struct page_ext *lookup_page_ext(struct page *page);
+static inline struct page_ext *page_ext_next(struct page_ext *curr)
+{
+ void *next = curr;
+ next += page_ext_size;
+ return next;
+}
+
#else /* !CONFIG_PAGE_EXTENSION */
struct page_ext;
--- a/mm/page_ext.c
+++ b/mm/page_ext.c
@@ -67,8 +67,9 @@ static struct page_ext_operations *page_
#endif
};
+unsigned long page_ext_size = sizeof(struct page_ext);
+
static unsigned long total_usage;
-static unsigned long extra_mem;
static bool __init invoke_need_callbacks(void)
{
@@ -78,9 +79,8 @@ static bool __init invoke_need_callbacks
for (i = 0; i < entries; i++) {
if (page_ext_ops[i]->need && page_ext_ops[i]->need()) {
- page_ext_ops[i]->offset = sizeof(struct page_ext) +
- extra_mem;
- extra_mem += page_ext_ops[i]->size;
+ page_ext_ops[i]->offset = page_ext_size;
+ page_ext_size += page_ext_ops[i]->size;
need = true;
}
}
@@ -99,14 +99,9 @@ static void __init invoke_init_callbacks
}
}
-static unsigned long get_entry_size(void)
-{
- return sizeof(struct page_ext) + extra_mem;
-}
-
static inline struct page_ext *get_entry(void *base, unsigned long index)
{
- return base + get_entry_size() * index;
+ return base + page_ext_size * index;
}
#if !defined(CONFIG_SPARSEMEM)
@@ -156,7 +151,7 @@ static int __init alloc_node_page_ext(in
!IS_ALIGNED(node_end_pfn(nid), MAX_ORDER_NR_PAGES))
nr_pages += MAX_ORDER_NR_PAGES;
- table_size = get_entry_size() * nr_pages;
+ table_size = page_ext_size * nr_pages;
base = memblock_virt_alloc_try_nid_nopanic(
table_size, PAGE_SIZE, __pa(MAX_DMA_ADDRESS),
@@ -237,7 +232,7 @@ static int __meminit init_section_page_e
if (section->page_ext)
return 0;
- table_size = get_entry_size() * PAGES_PER_SECTION;
+ table_size = page_ext_size * PAGES_PER_SECTION;
base = alloc_page_ext(table_size, nid);
/*
@@ -257,7 +252,7 @@ static int __meminit init_section_page_e
* we need to apply a mask.
*/
pfn &= PAGE_SECTION_MASK;
- section->page_ext = (void *)base - get_entry_size() * pfn;
+ section->page_ext = (void *)base - page_ext_size * pfn;
total_usage += table_size;
return 0;
}
@@ -270,7 +265,7 @@ static void free_page_ext(void *addr)
struct page *page = virt_to_page(addr);
size_t table_size;
- table_size = get_entry_size() * PAGES_PER_SECTION;
+ table_size = page_ext_size * PAGES_PER_SECTION;
BUG_ON(PageReserved(page));
free_pages_exact(addr, table_size);
--- a/mm/page_owner.c
+++ b/mm/page_owner.c
@@ -170,10 +170,10 @@ void __reset_page_owner(struct page *pag
handle = save_stack(GFP_NOWAIT | __GFP_NOWARN);
#endif
+ page_ext = lookup_page_ext(page);
+ if (unlikely(!page_ext))
+ return;
for (i = 0; i < (1 << order); i++) {
- page_ext = lookup_page_ext(page + i);
- if (unlikely(!page_ext))
- continue;
__clear_bit(PAGE_EXT_OWNER_ACTIVE, &page_ext->flags);
#ifdef CONFIG_DEBUG_PAGEALLOC
if (debug_pagealloc_enabled()) {
@@ -181,6 +181,7 @@ void __reset_page_owner(struct page *pag
page_owner->free_handle = handle;
}
#endif
+ page_ext = page_ext_next(page_ext);
}
}
@@ -200,7 +201,7 @@ static inline void __set_page_owner_hand
__set_bit(PAGE_EXT_OWNER, &page_ext->flags);
__set_bit(PAGE_EXT_OWNER_ACTIVE, &page_ext->flags);
- page_ext = lookup_page_ext(page + i);
+ page_ext = page_ext_next(page_ext);
}
}
@@ -238,12 +239,10 @@ void __split_page_owner(struct page *pag
if (unlikely(!page_ext))
return;
- page_owner = get_page_owner(page_ext);
- page_owner->order = 0;
- for (i = 1; i < (1 << order); i++) {
- page_ext = lookup_page_ext(page + i);
+ for (i = 0; i < (1 << order); i++) {
page_owner = get_page_owner(page_ext);
page_owner->order = 0;
+ page_ext = page_ext_next(page_ext);
}
}