The patch titled
Subject: Revert "mm: use early_pfn_to_nid in page_ext_init"
has been removed from the -mm tree. Its filename was
revert-mm-use-early_pfn_to_nid-in-page_ext_init.patch
This patch was dropped because it was merged into mainline or a subsystem tree
------------------------------------------------------
From: Qian Cai <cai@xxxxxx>
Subject: Revert "mm: use early_pfn_to_nid in page_ext_init"
This reverts fe53ca54270a7.
When booting a system with "page_owner=on",
start_kernel
page_ext_init
invoke_init_callbacks
init_section_page_ext
init_page_owner
init_early_allocated_pages
init_zones_in_node
init_pages_in_zone
lookup_page_ext
page_to_nid
The issue here is that page_to_nid() will not work since some page flags
have no node information until later in page_alloc_init_late() due to
DEFERRED_STRUCT_PAGE_INIT. Hence, it could trigger an out-of-bounds
access with an invalid nid.
[ 8.666047] UBSAN: Undefined behaviour in ./include/linux/mm.h:1104:50
[ 8.672603] index 7 is out of range for type 'zone [5]'
Also, kernel will panic since flags were poisoned earlier with,
CONFIG_DEBUG_VM_PGFLAGS=y
CONFIG_NODE_NOT_IN_PAGE_FLAGS=n
start_kernel
setup_arch
pagetable_init
paging_init
sparse_init
sparse_init_nid
memblock_alloc_try_nid_raw
It did not handle it well in init_pages_in_zone() which ends up calling
page_to_nid().
page:ffffea0004200000 is uninitialized and poisoned
raw: ffffffffffffffff ffffffffffffffff ffffffffffffffff ffffffffffffffff
raw: ffffffffffffffff ffffffffffffffff ffffffffffffffff ffffffffffffffff
page dumped because: VM_BUG_ON_PAGE(PagePoisoned(p))
page_owner info is not active (free page?)
kernel BUG at include/linux/mm.h:990!
RIP: 0010:init_page_owner+0x486/0x520
This means that assumptions behind commit fe53ca54270a ("mm: use
early_pfn_to_nid in page_ext_init") are incomplete. Therefore, revert the
commit for now. A proper way to move the page_owner initialization to
sooner is to hook into memmap initialization.
Link: http://lkml.kernel.org/r/20190115202812.75820-1-cai@xxxxxx
Signed-off-by: Qian Cai <cai@xxxxxx>
Acked-by: Michal Hocko <mhocko@xxxxxxxxxx>
Cc: Pasha Tatashin <Pavel.Tatashin@xxxxxxxxxxxxx>
Cc: Mel Gorman <mgorman@xxxxxxxxxxxxxxxxxxx>
Cc: Yang Shi <yang.shi@xxxxxxxxxx>
Cc: Joonsoo Kim <iamjoonsoo.kim@xxxxxxx>
Signed-off-by: Andrew Morton <akpm@xxxxxxxxxxxxxxxxxxxx>
---
init/main.c | 3 ++-
mm/page_ext.c | 4 +---
2 files changed, 3 insertions(+), 4 deletions(-)
--- a/init/main.c~revert-mm-use-early_pfn_to_nid-in-page_ext_init
+++ a/init/main.c
@@ -695,7 +695,6 @@ asmlinkage __visible void __init start_k
initrd_start = 0;
}
#endif
- page_ext_init();
kmemleak_init();
setup_per_cpu_pageset();
numa_policy_init();
@@ -1131,6 +1130,8 @@ static noinline void __init kernel_init_
sched_init_smp();
page_alloc_init_late();
+ /* Initialize page ext after all struct pages are initialized. */
+ page_ext_init();
do_basic_setup();
--- a/mm/page_ext.c~revert-mm-use-early_pfn_to_nid-in-page_ext_init
+++ a/mm/page_ext.c
@@ -398,10 +398,8 @@ void __init page_ext_init(void)
* We know some arch can have a nodes layout such as
* -------------pfn-------------->
* N0 | N1 | N2 | N0 | N1 | N2|....
- *
- * Take into account DEFERRED_STRUCT_PAGE_INIT.
*/
- if (early_pfn_to_nid(pfn) != nid)
+ if (pfn_to_nid(pfn) != nid)
continue;
if (init_section_page_ext(pfn, nid))
goto oom;
_
Patches currently in -mm which might be from cai@xxxxxx are
slub-fix-slab_consistency_checks-kasan_sw_tags.patch
page_poison-play-nicely-with-kasan.patch
slab-kmemleak-no-scan-alien-caches.patch
slub-remove-an-unused-addr-argument.patch
mm-compaction-be-selective-about-what-pageblocks-to-clear-skip-hints-fix.patch
signal-allow-the-null-signal-in-rt_sigqueueinfo.patch
