CVE-2021-47256: mm/memory-failure: make sure wait for page writeback in memory_failure
In the Linux kernel, the following vulnerability has been resolved:
mm/memory-failure: make sure wait for page writeback in memory_failure
Our syzkaller trigger the "BUG_ON(!list_empty(&inode->i_wb_list))" in
clear_inode:
kernel BUG at fs/inode.c:519!
Internal error: Oops - BUG: 0 [#1] SMP
Modules linked in:
Process syz-executor.0 (pid: 249, stack limit = 0x00000000a12409d7)
CPU: 1 PID: 249 Comm: syz-executor.0 Not tainted 4.19.95
Hardware name: linux,dummy-virt (DT)
pstate: 80000005 (Nzcv daif -PAN -UAO)
pc : clear_inode+0x280/0x2a8
lr : clear_inode+0x280/0x2a8
Call trace:
clear_inode+0x280/0x2a8
ext4_clear_inode+0x38/0xe8
ext4_free_inode+0x130/0xc68
ext4_evict_inode+0xb20/0xcb8
evict+0x1a8/0x3c0
iput+0x344/0x460
do_unlinkat+0x260/0x410
__arm64_sys_unlinkat+0x6c/0xc0
el0_svc_common+0xdc/0x3b0
el0_svc_handler+0xf8/0x160
el0_svc+0x10/0x218
Kernel panic - not syncing: Fatal exception
A crash dump of this problem show that someone called __munlock_pagevec
to clear page LRU without lock_page: do_mmap -> mmap_region -> do_munmap
-> munlock_vma_pages_range -> __munlock_pagevec.
As a result memory_failure will call identify_page_state without
wait_on_page_writeback. And after truncate_error_page clear the mapping
of this page. end_page_writeback won't call sb_clear_inode_writeback to
clear inode->i_wb_list. That will trigger BUG_ON in clear_inode!
Fix it by checking PageWriteback too to help determine should we skip
wait_on_page_writeback.
Security readout for executives and security teams
Plain-English summary
This Linux kernel flaw can crash a vulnerable system when the memory-failure handler mishandles a page that is still being written back. The public record describes a kernel panic found by syzkaller, not data theft or remote compromise. Business urgency depends on whether affected kernel builds are used on production or multi-tenant systems.
Executive priority
Treat as a reliability and availability issue. It should be scheduled into normal kernel patching, with faster handling for shared infrastructure or systems running untrusted workloads. Current sources do not justify emergency response based on exploitation evidence.
Technical view
The bug is in mm/memory-failure handling. The fix adds PageWriteback consideration so memory_failure waits appropriately before truncate_error_page clears page mapping. Without that wait, end_page_writeback may not clear inode writeback state, leading to BUG_ON in clear_inode and kernel panic.
Likely exposure
Exposure is likely limited to systems running affected Linux kernel versions or downstream builds without the stable fix. The bundle lists affected Linux versions including 3.16, 4.14.238, 4.19.196, 5.4.128, 5.10.46, 5.12.13, and 5.13; distribution backports must be verified separately.
Exploitation context
No active exploitation is supported by the provided sources. KEV is false. The evidence describes a syzkaller-triggered crash path and kernel panic. The bundle does not identify remote exploitation, privilege escalation, public weaponization, or confirmed real-world abuse.
Researcher notes
The source evidence is kernel-resolution text plus stable commit references. Impact is demonstrated as a kernel panic from incorrect writeback waiting in memory_failure. Affected-version metadata is broad and should be reconciled with distro backports before declaring individual hosts vulnerable.
Mitigation direction
Apply vendor kernel updates that include the referenced stable fixes.
Verify distribution backports rather than relying only on upstream version numbers.
Prioritize production, multi-tenant, and untrusted-code Linux hosts.
If patching is delayed, check vendor guidance for supported mitigations.
Validation and detection
Inventory running kernel versions across Linux hosts.
Confirm whether the relevant stable commits or vendor backports are present.
Review crash logs for clear_inode, inode i_wb_list, or memory_failure panic indicators.
Check CVE tracking against distribution advisories for each deployed kernel line.
Generated from the cited source records. This long-tail analysis has not been individually reviewed by a named human.
Potential ATT&CK relevance
Conservative CVE-to-ATT&CK context
These mappings and lookup hints may be relevant to the vulnerability behavior, CWE, affected product, or exposure path. Glexia-inferred context is not an official MITRE, ATT&CK, CWE, or CVE Program mapping.
ATT&CK lookup starting points
Use these exact CWE pages and searches to review the Glexia ATT&CK library from this CVE's weakness and description context.
cve · low confidence lookup
CVE-2021-47256 mapping review
Open the CVE-to-ATT&CK bridge for reviewed, inferred, or future official mappings tied to this CVE.
These fields come from the CVE record and ADP containers, not from Glexia's Take. They preserve time-varying source decisions such as CISA SSVC, KEV status, CVSS metrics, and provider references.