Skip to content

Commit 71b426e

Browse files
committed
writeback: fix dereferencing NULL mapping->host on writeback_page_template
jira LE-4623 Rebuild_History Non-Buildable kernel-4.18.0-553.81.1.el8_10 commit-author Rafael Aquini <aquini@redhat.com> commit 54abe19 Empty-Commit: Cherry-Pick Conflicts during history rebuild. Will be included in final tarball splat. Ref for failed cherry-pick at: ciq/ciq_backports/kernel-4.18.0-553.81.1.el8_10/54abe19e.failed When commit 19343b5 ("mm/page-writeback: introduce tracepoint for wait_on_page_writeback()") repurposed the writeback_dirty_page trace event as a template to create its new wait_on_page_writeback trace event, it ended up opening a window to NULL pointer dereference crashes due to the (infrequent) occurrence of a race where an access to a page in the swap-cache happens concurrently with the moment this page is being written to disk and the tracepoint is enabled: BUG: kernel NULL pointer dereference, address: 0000000000000040 #PF: supervisor read access in kernel mode #PF: error_code(0x0000) - not-present page PGD 800000010ec0a067 P4D 800000010ec0a067 PUD 102353067 PMD 0 Oops: 0000 [#1] PREEMPT SMP PTI CPU: 1 PID: 1320 Comm: shmem-worker Kdump: loaded Not tainted 6.4.0-rc5+ #13 Hardware name: QEMU Standard PC (Q35 + ICH9, 2009), BIOS edk2-20230301gitf80f052277c8-1.fc37 03/01/2023 RIP: 0010:trace_event_raw_event_writeback_folio_template+0x76/0xf0 Code: 4d 85 e4 74 5c 49 8b 3c 24 e8 06 98 ee ff 48 89 c7 e8 9e 8b ee ff ba 20 00 00 00 48 89 ef 48 89 c6 e8 fe d4 1a 00 49 8b 04 24 <48> 8b 40 40 48 89 43 28 49 8b 45 20 48 89 e7 48 89 43 30 e8 a2 4d RSP: 0000:ffffaad580b6fb60 EFLAGS: 00010246 RAX: 0000000000000000 RBX: ffff90e38035c01c RCX: 0000000000000000 RDX: 0000000000000000 RSI: 0000000000000000 RDI: ffff90e38035c044 RBP: ffff90e38035c024 R08: 0000000000000002 R09: 0000000000000006 R10: ffff90e38035c02e R11: 0000000000000020 R12: ffff90e380bac000 R13: ffffe3a7456d9200 R14: 0000000000001b81 R15: ffffe3a7456d9200 FS: 00007f2e4e8a15c0(0000) GS:ffff90e3fbc80000(0000) knlGS:0000000000000000 CS: 0010 DS: 0000 ES: 0000 CR0: 0000000080050033 CR2: 0000000000000040 CR3: 00000001150c6003 CR4: 0000000000170ee0 Call Trace: <TASK> ? __die+0x20/0x70 ? page_fault_oops+0x76/0x170 ? kernelmode_fixup_or_oops+0x84/0x110 ? exc_page_fault+0x65/0x150 ? asm_exc_page_fault+0x22/0x30 ? trace_event_raw_event_writeback_folio_template+0x76/0xf0 folio_wait_writeback+0x6b/0x80 shmem_swapin_folio+0x24a/0x500 ? filemap_get_entry+0xe3/0x140 shmem_get_folio_gfp+0x36e/0x7c0 ? find_busiest_group+0x43/0x1a0 shmem_fault+0x76/0x2a0 ? __update_load_avg_cfs_rq+0x281/0x2f0 __do_fault+0x33/0x130 do_read_fault+0x118/0x160 do_pte_missing+0x1ed/0x2a0 __handle_mm_fault+0x566/0x630 handle_mm_fault+0x91/0x210 do_user_addr_fault+0x22c/0x740 exc_page_fault+0x65/0x150 asm_exc_page_fault+0x22/0x30 This problem arises from the fact that the repurposed writeback_dirty_page trace event code was written assuming that every pointer to mapping (struct address_space) would come from a file-mapped page-cache object, thus mapping->host would always be populated, and that was a valid case before commit 19343b5. The swap-cache address space (swapper_spaces), however, doesn't populate its ->host (struct inode) pointer, thus leading to the crashes in the corner-case aforementioned. commit 19343b5 ended up breaking the assignment of __entry->name and __entry->ino for the wait_on_page_writeback tracepoint -- both dependent on mapping->host carrying a pointer to a valid inode. The assignment of __entry->name was fixed by commit 68f23b8 ("memcg: fix a crash in wb_workfn when a device disappears"), and this commit fixes the remaining case, for __entry->ino. Link: https://lkml.kernel.org/r/20230606233613.1290819-1-aquini@redhat.com Fixes: 19343b5 ("mm/page-writeback: introduce tracepoint for wait_on_page_writeback()") Signed-off-by: Rafael Aquini <aquini@redhat.com> Reviewed-by: Yafang Shao <laoar.shao@gmail.com> Cc: Aristeu Rozanski <aris@redhat.com> Cc: <stable@vger.kernel.org> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> (cherry picked from commit 54abe19) Signed-off-by: Jonathan Maple <jmaple@ciq.com> # Conflicts: # include/trace/events/writeback.h
1 parent deb0fda commit 71b426e

File tree

1 file changed

+106
-0
lines changed

1 file changed

+106
-0
lines changed
Lines changed: 106 additions & 0 deletions
Original file line numberDiff line numberDiff line change
@@ -0,0 +1,106 @@
1+
writeback: fix dereferencing NULL mapping->host on writeback_page_template
2+
3+
jira LE-4623
4+
Rebuild_History Non-Buildable kernel-4.18.0-553.81.1.el8_10
5+
commit-author Rafael Aquini <aquini@redhat.com>
6+
commit 54abe19e00cfcc5a72773d15cd00ed19ab763439
7+
Empty-Commit: Cherry-Pick Conflicts during history rebuild.
8+
Will be included in final tarball splat. Ref for failed cherry-pick at:
9+
ciq/ciq_backports/kernel-4.18.0-553.81.1.el8_10/54abe19e.failed
10+
11+
When commit 19343b5bdd16 ("mm/page-writeback: introduce tracepoint for
12+
wait_on_page_writeback()") repurposed the writeback_dirty_page trace event
13+
as a template to create its new wait_on_page_writeback trace event, it
14+
ended up opening a window to NULL pointer dereference crashes due to the
15+
(infrequent) occurrence of a race where an access to a page in the
16+
swap-cache happens concurrently with the moment this page is being written
17+
to disk and the tracepoint is enabled:
18+
19+
BUG: kernel NULL pointer dereference, address: 0000000000000040
20+
#PF: supervisor read access in kernel mode
21+
#PF: error_code(0x0000) - not-present page
22+
PGD 800000010ec0a067 P4D 800000010ec0a067 PUD 102353067 PMD 0
23+
Oops: 0000 [#1] PREEMPT SMP PTI
24+
CPU: 1 PID: 1320 Comm: shmem-worker Kdump: loaded Not tainted 6.4.0-rc5+ #13
25+
Hardware name: QEMU Standard PC (Q35 + ICH9, 2009), BIOS edk2-20230301gitf80f052277c8-1.fc37 03/01/2023
26+
RIP: 0010:trace_event_raw_event_writeback_folio_template+0x76/0xf0
27+
Code: 4d 85 e4 74 5c 49 8b 3c 24 e8 06 98 ee ff 48 89 c7 e8 9e 8b ee ff ba 20 00 00 00 48 89 ef 48 89 c6 e8 fe d4 1a 00 49 8b 04 24 <48> 8b 40 40 48 89 43 28 49 8b 45 20 48 89 e7 48 89 43 30 e8 a2 4d
28+
RSP: 0000:ffffaad580b6fb60 EFLAGS: 00010246
29+
RAX: 0000000000000000 RBX: ffff90e38035c01c RCX: 0000000000000000
30+
RDX: 0000000000000000 RSI: 0000000000000000 RDI: ffff90e38035c044
31+
RBP: ffff90e38035c024 R08: 0000000000000002 R09: 0000000000000006
32+
R10: ffff90e38035c02e R11: 0000000000000020 R12: ffff90e380bac000
33+
R13: ffffe3a7456d9200 R14: 0000000000001b81 R15: ffffe3a7456d9200
34+
FS: 00007f2e4e8a15c0(0000) GS:ffff90e3fbc80000(0000) knlGS:0000000000000000
35+
CS: 0010 DS: 0000 ES: 0000 CR0: 0000000080050033
36+
CR2: 0000000000000040 CR3: 00000001150c6003 CR4: 0000000000170ee0
37+
Call Trace:
38+
<TASK>
39+
? __die+0x20/0x70
40+
? page_fault_oops+0x76/0x170
41+
? kernelmode_fixup_or_oops+0x84/0x110
42+
? exc_page_fault+0x65/0x150
43+
? asm_exc_page_fault+0x22/0x30
44+
? trace_event_raw_event_writeback_folio_template+0x76/0xf0
45+
folio_wait_writeback+0x6b/0x80
46+
shmem_swapin_folio+0x24a/0x500
47+
? filemap_get_entry+0xe3/0x140
48+
shmem_get_folio_gfp+0x36e/0x7c0
49+
? find_busiest_group+0x43/0x1a0
50+
shmem_fault+0x76/0x2a0
51+
? __update_load_avg_cfs_rq+0x281/0x2f0
52+
__do_fault+0x33/0x130
53+
do_read_fault+0x118/0x160
54+
do_pte_missing+0x1ed/0x2a0
55+
__handle_mm_fault+0x566/0x630
56+
handle_mm_fault+0x91/0x210
57+
do_user_addr_fault+0x22c/0x740
58+
exc_page_fault+0x65/0x150
59+
asm_exc_page_fault+0x22/0x30
60+
61+
This problem arises from the fact that the repurposed writeback_dirty_page
62+
trace event code was written assuming that every pointer to mapping
63+
(struct address_space) would come from a file-mapped page-cache object,
64+
thus mapping->host would always be populated, and that was a valid case
65+
before commit 19343b5bdd16. The swap-cache address space
66+
(swapper_spaces), however, doesn't populate its ->host (struct inode)
67+
pointer, thus leading to the crashes in the corner-case aforementioned.
68+
69+
commit 19343b5bdd16 ended up breaking the assignment of __entry->name and
70+
__entry->ino for the wait_on_page_writeback tracepoint -- both dependent
71+
on mapping->host carrying a pointer to a valid inode. The assignment of
72+
__entry->name was fixed by commit 68f23b89067f ("memcg: fix a crash in
73+
wb_workfn when a device disappears"), and this commit fixes the remaining
74+
case, for __entry->ino.
75+
76+
Link: https://lkml.kernel.org/r/20230606233613.1290819-1-aquini@redhat.com
77+
Fixes: 19343b5bdd16 ("mm/page-writeback: introduce tracepoint for wait_on_page_writeback()")
78+
Signed-off-by: Rafael Aquini <aquini@redhat.com>
79+
Reviewed-by: Yafang Shao <laoar.shao@gmail.com>
80+
Cc: Aristeu Rozanski <aris@redhat.com>
81+
Cc: <stable@vger.kernel.org>
82+
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
83+
(cherry picked from commit 54abe19e00cfcc5a72773d15cd00ed19ab763439)
84+
Signed-off-by: Jonathan Maple <jmaple@ciq.com>
85+
86+
# Conflicts:
87+
# include/trace/events/writeback.h
88+
diff --cc include/trace/events/writeback.h
89+
index 42c25182279a,54e353c9f919..000000000000
90+
--- a/include/trace/events/writeback.h
91+
+++ b/include/trace/events/writeback.h
92+
@@@ -68,8 -68,8 +68,13 @@@ DECLARE_EVENT_CLASS(writeback_page_temp
93+
strscpy_pad(__entry->name,
94+
bdi_dev_name(mapping ? inode_to_bdi(mapping->host) :
95+
NULL), 32);
96+
++<<<<<<< HEAD
97+
+ __entry->ino = mapping ? mapping->host->i_ino : 0;
98+
+ __entry->index = page->index;
99+
++=======
100+
+ __entry->ino = (mapping && mapping->host) ? mapping->host->i_ino : 0;
101+
+ __entry->index = folio->index;
102+
++>>>>>>> 54abe19e00cf (writeback: fix dereferencing NULL mapping->host on writeback_page_template)
103+
),
104+
105+
TP_printk("bdi %s: ino=%lu index=%lu",
106+
* Unmerged path include/trace/events/writeback.h

0 commit comments

Comments
 (0)