[syzbot] [mm?] INFO: task hung in __rmap_walk_file

[syzbot]

Hello,

syzbot found the following issue on:

HEAD commit: 5472d60c129f Merge tag 'trace-v6.18-2' of git://git.kernel..
git tree: upstream
console output: https://syzkaller.appspot.com/x/log.txt?x=16d69304580000
kernel config: https://syzkaller.appspot.com/x/.config?x=5b213914b883d014
dashboard link: https://syzkaller.appspot.com/bug?extid=2d9c96466c978346b55f
compiler: Debian clang version 20.1.8 (++20250708063551+0c9f909b7976-1~exp1~20250708183702.136), Debian LLD 20.1.8
syz repro: https://syzkaller.appspot.com/x/repro.syz?x=133e89e2580000
C reproducer: https://syzkaller.appspot.com/x/repro.c?x=17f3ba7c580000

Downloadable assets:
disk image (non-bootable): https://storage.googleapis.com/syzbot-assets/d900f083ada3/non_bootable_disk-5472d60c.raw.xz
vmlinux: https://storage.googleapis.com/syzbot-assets/af61e8db8b22/vmlinux-5472d60c.xz
kernel image: https://storage.googleapis.com/syzbot-assets/a2c11e401d8a/bzImage-5472d60c.xz

IMPORTANT: if you fix the issue, please add the following tag to the commit:
Reported-by: syzbot+2d9c96...@syzkaller.appspotmail.com

INFO: task syz.5.48:5749 blocked for more than 143 seconds.
Not tainted syzkaller #0
"echo 0 > /proc/sys/kernel/hung_task_timeout_secs" disables this message.
task:syz.5.48 state:D stack:27656 pid:5749 tgid:5747 ppid:5477 task_flags:0x400040 flags:0x00080002
Call Trace:
<TASK>
context_switch kernel/sched/core.c:5325 [inline]
__schedule+0x1798/0x4cc0 kernel/sched/core.c:6929
__schedule_loop kernel/sched/core.c:7011 [inline]
schedule+0x165/0x360 kernel/sched/core.c:7026
schedule_preempt_disabled+0x13/0x30 kernel/sched/core.c:7083
rwsem_down_read_slowpath+0x5fd/0x8f0 kernel/locking/rwsem.c:1086
__down_read_common kernel/locking/rwsem.c:1261 [inline]
__down_read kernel/locking/rwsem.c:1274 [inline]
down_read+0x98/0x2e0 kernel/locking/rwsem.c:1539
i_mmap_lock_read include/linux/fs.h:568 [inline]
__rmap_walk_file+0x227/0x620 mm/rmap.c:2905
remove_migration_ptes mm/migrate.c:471 [inline]
unmap_and_move_huge_page mm/migrate.c:1520 [inline]
migrate_hugetlbs mm/migrate.c:1641 [inline]
migrate_pages+0xc98/0x2930 mm/migrate.c:2080
do_mbind mm/mempolicy.c:1539 [inline]
kernel_mbind mm/mempolicy.c:1682 [inline]
__do_sys_mbind mm/mempolicy.c:1756 [inline]
__se_sys_mbind+0xa47/0xc40 mm/mempolicy.c:1752
do_syscall_x64 arch/x86/entry/syscall_64.c:63 [inline]
do_syscall_64+0xfa/0xfa0 arch/x86/entry/syscall_64.c:94
entry_SYSCALL_64_after_hwframe+0x77/0x7f
RIP: 0033:0x7fd36cd8eec9
RSP: 002b:00007fd36dccd038 EFLAGS: 00000246 ORIG_RAX: 00000000000000ed
RAX: ffffffffffffffda RBX: 00007fd36cfe5fa0 RCX: 00007fd36cd8eec9
RDX: 0000000000000000 RSI: 0000000000800000 RDI: 0000200000001000
RBP: 00007fd36ce11f91 R08: 0000000000000040 R09: 0000000000000002
R10: 0000000000000000 R11: 0000000000000246 R12: 0000000000000000
R13: 00007fd36cfe6038 R14: 00007fd36cfe5fa0 R15: 00007ffec3386808
</TASK>
INFO: task syz.5.48:5754 blocked for more than 143 seconds.
Not tainted syzkaller #0
"echo 0 > /proc/sys/kernel/hung_task_timeout_secs" disables this message.
task:syz.5.48 state:D stack:26920 pid:5754 tgid:5747 ppid:5477 task_flags:0x400040 flags:0x00080002
Call Trace:
<TASK>
context_switch kernel/sched/core.c:5325 [inline]
__schedule+0x1798/0x4cc0 kernel/sched/core.c:6929
__schedule_loop kernel/sched/core.c:7011 [inline]
schedule+0x165/0x360 kernel/sched/core.c:7026
io_schedule+0x80/0xd0 kernel/sched/core.c:7871
folio_wait_bit_common+0x6b0/0xb80 mm/filemap.c:1330
__folio_lock mm/filemap.c:1706 [inline]
folio_lock include/linux/pagemap.h:1141 [inline]
__filemap_get_folio+0x139/0xaf0 mm/filemap.c:1960
filemap_lock_folio include/linux/pagemap.h:820 [inline]
filemap_lock_hugetlb_folio include/linux/hugetlb.h:814 [inline]
hugetlbfs_zero_partial_page+0xae/0x610 fs/hugetlbfs/inode.c:654
hugetlbfs_punch_hole fs/hugetlbfs/inode.c:708 [inline]
hugetlbfs_fallocate+0xb91/0x1100 fs/hugetlbfs/inode.c:741
vfs_fallocate+0x666/0x7e0 fs/open.c:342
madvise_remove mm/madvise.c:1049 [inline]
madvise_vma_behavior+0x31b3/0x3a10 mm/madvise.c:1346
madvise_walk_vmas+0x51c/0xa30 mm/madvise.c:1669
madvise_do_behavior+0x38e/0x550 mm/madvise.c:1885
do_madvise+0x1bc/0x270 mm/madvise.c:1978
__do_sys_madvise mm/madvise.c:1987 [inline]
__se_sys_madvise mm/madvise.c:1985 [inline]
__x64_sys_madvise+0xa7/0xc0 mm/madvise.c:1985
do_syscall_x64 arch/x86/entry/syscall_64.c:63 [inline]
do_syscall_64+0xfa/0xfa0 arch/x86/entry/syscall_64.c:94
entry_SYSCALL_64_after_hwframe+0x77/0x7f
RIP: 0033:0x7fd36cd8eec9
RSP: 002b:00007fd36dcac038 EFLAGS: 00000246 ORIG_RAX: 000000000000001c
RAX: ffffffffffffffda RBX: 00007fd36cfe6090 RCX: 00007fd36cd8eec9
RDX: 0000000000000009 RSI: 0000000000600002 RDI: 0000200000000000
RBP: 00007fd36ce11f91 R08: 0000000000000000 R09: 0000000000000000
R10: 0000000000000000 R11: 0000000000000246 R12: 0000000000000000
R13: 00007fd36cfe6128 R14: 00007fd36cfe6090 R15: 00007ffec3386808
</TASK>

Showing all locks held in the system:
3 locks held by kworker/u4:0/12:
#0: ffff88801a479948 ((wq_completion)events_unbound#2){+.+.}-{0:0}, at: process_one_work kernel/workqueue.c:3238 [inline]
#0: ffff88801a479948 ((wq_completion)events_unbound#2){+.+.}-{0:0}, at: process_scheduled_works+0x9b4/0x17b0 kernel/workqueue.c:3346
#1: ffffc900001e7ba0 ((linkwatch_work).work){+.+.}-{0:0}, at: process_one_work kernel/workqueue.c:3239 [inline]
#1: ffffc900001e7ba0 ((linkwatch_work).work){+.+.}-{0:0}, at: process_scheduled_works+0x9ef/0x17b0 kernel/workqueue.c:3346
#2: ffffffff8f4e1ac8 (rtnl_mutex){+.+.}-{4:4}, at: linkwatch_event+0xe/0x60 net/core/link_watch.c:303
1 lock held by khungtaskd/26:
#0: ffffffff8e13d320 (rcu_read_lock){....}-{1:3}, at: rcu_lock_acquire include/linux/rcupdate.h:331 [inline]
#0: ffffffff8e13d320 (rcu_read_lock){....}-{1:3}, at: rcu_read_lock include/linux/rcupdate.h:867 [inline]
#0: ffffffff8e13d320 (rcu_read_lock){....}-{1:3}, at: debug_show_all_locks+0x2e/0x180 kernel/locking/lockdep.c:6775
5 locks held by kworker/u4:6/1041:
#0: ffff88801b6f8948 ((wq_completion)netns){+.+.}-{0:0}, at: process_one_work kernel/workqueue.c:3238 [inline]
#0: ffff88801b6f8948 ((wq_completion)netns){+.+.}-{0:0}, at: process_scheduled_works+0x9b4/0x17b0 kernel/workqueue.c:3346
#1: ffffc900025b7ba0 (net_cleanup_work){+.+.}-{0:0}, at: process_one_work kernel/workqueue.c:3239 [inline]
#1: ffffc900025b7ba0 (net_cleanup_work){+.+.}-{0:0}, at: process_scheduled_works+0x9ef/0x17b0 kernel/workqueue.c:3346
#2: ffffffff8f4d4c50 (pernet_ops_rwsem){++++}-{4:4}, at: cleanup_net+0xf7/0x820 net/core/net_namespace.c:669
#3: ffffffff8f4e1ac8 (rtnl_mutex){+.+.}-{4:4}, at: default_device_exit_batch+0xdc/0x890 net/core/dev.c:12807
#4: ffffffff8e142db8 (rcu_state.exp_mutex){+.+.}-{4:4}, at: exp_funnel_lock kernel/rcu/tree_exp.h:311 [inline]
#4: ffffffff8e142db8 (rcu_state.exp_mutex){+.+.}-{4:4}, at: synchronize_rcu_expedited+0x2f6/0x730 kernel/rcu/tree_exp.h:957
1 lock held by dhcpcd/5017:
#0: ffffffff8f4e1ac8 (rtnl_mutex){+.+.}-{4:4}, at: rtnl_net_lock include/linux/rtnetlink.h:130 [inline]
#0: ffffffff8f4e1ac8 (rtnl_mutex){+.+.}-{4:4}, at: devinet_ioctl+0x323/0x1b50 net/ipv4/devinet.c:1120
2 locks held by getty/5110:
#0: ffff88803597c0a0 (&tty->ldisc_sem){++++}-{0:0}, at: tty_ldisc_ref_wait+0x25/0x70 drivers/tty/tty_ldisc.c:243
#1: ffffc9000018e2f0 (&ldata->atomic_read_lock){+.+.}-{4:4}, at: n_tty_read+0x43e/0x1400 drivers/tty/n_tty.c:2222
1 lock held by syz.5.48/5749:
#0: ffff888011d3f618 (&hugetlbfs_i_mmap_rwsem_key){++++}-{4:4}, at: i_mmap_lock_read include/linux/fs.h:568 [inline]
#0: ffff888011d3f618 (&hugetlbfs_i_mmap_rwsem_key){++++}-{4:4}, at: __rmap_walk_file+0x227/0x620 mm/rmap.c:2905
3 locks held by syz.5.48/5754:
#0: ffff88801ebe2420 (sb_writers#12){.+.+}-{0:0}, at: file_start_write include/linux/fs.h:3111 [inline]
#0: ffff88801ebe2420 (sb_writers#12){.+.+}-{0:0}, at: vfs_fallocate+0x5f0/0x7e0 fs/open.c:341
#1: ffff888011d3f348 (&sb->s_type->i_mutex_key#20){+.+.}-{4:4}, at: inode_lock include/linux/fs.h:980 [inline]
#1: ffff888011d3f348 (&sb->s_type->i_mutex_key#20){+.+.}-{4:4}, at: hugetlbfs_punch_hole fs/hugetlbfs/inode.c:683 [inline]
#1: ffff888011d3f348 (&sb->s_type->i_mutex_key#20){+.+.}-{4:4}, at: hugetlbfs_fallocate+0x3cc/0x1100 fs/hugetlbfs/inode.c:741
#2: ffff888011d3f618 (&hugetlbfs_i_mmap_rwsem_key){++++}-{4:4}, at: i_mmap_lock_write include/linux/fs.h:548 [inline]
#2: ffff888011d3f618 (&hugetlbfs_i_mmap_rwsem_key){++++}-{4:4}, at: hugetlbfs_punch_hole fs/hugetlbfs/inode.c:691 [inline]
#2: ffff888011d3f618 (&hugetlbfs_i_mmap_rwsem_key){++++}-{4:4}, at: hugetlbfs_fallocate+0x4b5/0x1100 fs/hugetlbfs/inode.c:741
2 locks held by syz-executor/20402:
#0: ffffffff8ec74400 (&ops->srcu#2){.+.+}-{0:0}, at: rcu_lock_acquire include/linux/rcupdate.h:331 [inline]
#0: ffffffff8ec74400 (&ops->srcu#2){.+.+}-{0:0}, at: rcu_read_lock include/linux/rcupdate.h:867 [inline]
#0: ffffffff8ec74400 (&ops->srcu#2){.+.+}-{0:0}, at: rtnl_link_ops_get+0x23/0x250 net/core/rtnetlink.c:570
#1: ffffffff8f4e1ac8 (rtnl_mutex){+.+.}-{4:4}, at: rtnl_lock net/core/rtnetlink.c:80 [inline]
#1: ffffffff8f4e1ac8 (rtnl_mutex){+.+.}-{4:4}, at: rtnl_nets_lock net/core/rtnetlink.c:341 [inline]
#1: ffffffff8f4e1ac8 (rtnl_mutex){+.+.}-{4:4}, at: rtnl_newlink+0x8e9/0x1c80 net/core/rtnetlink.c:4064
1 lock held by syz.4.7446/21125:
#0: ffffffff8e142db8 (rcu_state.exp_mutex){+.+.}-{4:4}, at: exp_funnel_lock kernel/rcu/tree_exp.h:343 [inline]
#0: ffffffff8e142db8 (rcu_state.exp_mutex){+.+.}-{4:4}, at: synchronize_rcu_expedited+0x3b9/0x730 kernel/rcu/tree_exp.h:957

=============================================

NMI backtrace for cpu 0
CPU: 0 UID: 0 PID: 26 Comm: khungtaskd Not tainted syzkaller #0 PREEMPT(full)
Hardware name: QEMU Standard PC (Q35 + ICH9, 2009), BIOS 1.16.3-debian-1.16.3-2~bpo12+1 04/01/2014
Call Trace:
<TASK>
dump_stack_lvl+0x189/0x250 lib/dump_stack.c:120
nmi_cpu_backtrace+0x39e/0x3d0 lib/nmi_backtrace.c:113
nmi_trigger_cpumask_backtrace+0x17a/0x300 lib/nmi_backtrace.c:62
trigger_all_cpu_backtrace include/linux/nmi.h:160 [inline]
check_hung_uninterruptible_tasks kernel/hung_task.c:332 [inline]
watchdog+0xf60/0xfa0 kernel/hung_task.c:495
kthread+0x711/0x8a0 kernel/kthread.c:463
ret_from_fork+0x4bc/0x870 arch/x86/kernel/process.c:158
ret_from_fork_asm+0x1a/0x30 arch/x86/entry/entry_64.S:245
</TASK>


---
This report is generated by a bot. It may contain errors.
See https://goo.gl/tpsmEJ for more information about syzbot.
syzbot engineers can be reached at syzk...@googlegroups.com.

syzbot will keep track of this issue. See:
https://goo.gl/tpsmEJ#status for how to communicate with syzbot.

If the report is already addressed, let syzbot know by replying with:
#syz fix: exact-commit-title

If you want syzbot to run the reproducer, reply with:
#syz test: git://repo/address.git branch-or-commit-hash
If you attach or paste a git patch, syzbot will apply it before testing.

If you want to overwrite report's subsystems, reply with:
#syz set subsystems: new-subsystem
(See the list of subsystem names on the web dashboard)

If the report is a duplicate of another one, reply with:
#syz dup: exact-subject-of-another-report

If you want to undo deduplication, reply with:
#syz undup

[Lance Yang]

It would be great if CONFIG_DETECT_HUNG_TASK_BLOCKER was set, as it could
point directly to the blocker. However, the lockdep output at the end of
the report already gives us the crucial clue :)

```

1 lock held by syz.5.48/5749:
#0: ffff888011d3f618 (&hugetlbfs_i_mmap_rwsem_key){++++}-{4:4}, at:
i_mmap_lock_read include/linux/fs.h:568 [inline]
#0: ffff888011d3f618 (&hugetlbfs_i_mmap_rwsem_key){++++}-{4:4}, at:
__rmap_walk_file+0x227/0x620 mm/rmap.c:2905

3 locks held by syz.5.48/5754:

...

#2: ffff888011d3f618 (&hugetlbfs_i_mmap_rwsem_key){++++}-{4:4}, at:
i_mmap_lock_write include/linux/fs.h:548 [inline]
#2: ffff888011d3f618 (&hugetlbfs_i_mmap_rwsem_key){++++}-{4:4}, at:
hugetlbfs_punch_hole fs/hugetlbfs/inode.c:691 [inline]
#2: ffff888011d3f618 (&hugetlbfs_i_mmap_rwsem_key){++++}-{4:4}, at:
hugetlbfs_fallocate+0x4b5/0x1100

```

1) One task (5754) holds the i_mmap_rwsem write lock for hugetlbfs_fallocate.
2) While holding that lock, it blocks waiting for a folio_lock (according
to its call trace).
3) This starves the other task (5749), which is waiting for the i_mmap_rwsem
read lock to perform migrate_pages.

From the report, I cannot tell who originally held the folio_lock that
blocked task 5754. I hope this analysis is useful ...

Thanks,
Lance

[Lance Yang]

I believe task (5749) is the one holding that folio_lock. IIUC, that's a
A-B-B-A deadlock:

1) Task (5749): Holds folio_lock, then tries to acquire i_mmap_rwsem(read lock).
2) Task (5754): Holds i_mmap_rwsem(write lock), then tries to acquire
folio_lock.

# Task (5749)
migrate_pages()
-> migrate_hugetlbs()
-> unmap_and_move_huge_page() <- Takes folio_lock!
-> remove_migration_ptes()
-> __rmap_walk_file()
-> i_mmap_lock_read() <- Waits for i_mmap_rwsem(read lock)!

# Task (5754)
hugetlbfs_fallocate()
-> hugetlbfs_punch_hole() <- Takes i_mmap_rwsem(write lock)!
-> hugetlbfs_zero_partial_page()
-> filemap_lock_hugetlb_folio()
-> filemap_lock_folio()
-> __filemap_get_folio <- Waits for folio_lock!

Thanks,
Lance

[Harry Yoo]

[+Cc HUGETLB SUBSYSTEM folks]

> # Task (5749)
> migrate_pages()
> -> migrate_hugetlbs()
> -> unmap_and_move_huge_page() <- Takes folio_lock!
> -> remove_migration_ptes()
> -> __rmap_walk_file()
> -> i_mmap_lock_read() <- Waits for i_mmap_rwsem(read lock)!
>
> # Task (5754)
> hugetlbfs_fallocate()
> -> hugetlbfs_punch_hole() <- Takes i_mmap_rwsem(write lock)!
> -> hugetlbfs_zero_partial_page()
> -> filemap_lock_hugetlb_folio()
> -> filemap_lock_folio()
> -> __filemap_get_folio <- Waits for folio_lock!

So this is against the locking order mentioned in mm/rmap.c?

The order should be mapping->i_mmap_rwsem -> folio_lock instead of
the other way around, or it should use trylock variant instead of folio_lock()?

--
Cheers,
Harry / Hyeonggon

[syzbot]

For archival purposes, forwarding an incoming command email to
syzkall...@googlegroups.com.

***

Subject: [PATCH] try fix
Author: wangjin...@gmail.com

#syz test

---
fs/hugetlbfs/inode.c | 33 ++++++++++++++++++++++-----------
include/linux/hugetlb.h | 2 +-
include/linux/pagemap.h | 6 ++++++
3 files changed, 29 insertions(+), 12 deletions(-)

diff --git a/fs/hugetlbfs/inode.c b/fs/hugetlbfs/inode.c
index 9c94ed8c3ab0..a8f33b72a8a6 100644
--- a/fs/hugetlbfs/inode.c
+++ b/fs/hugetlbfs/inode.c
@@ -643,17 +643,16 @@ static void hugetlb_vmtruncate(struct inode *inode, loff_t offset)
remove_inode_hugepages(inode, offset, LLONG_MAX);
}

-static void hugetlbfs_zero_partial_page(struct hstate *h,
- struct address_space *mapping,
- loff_t start,
- loff_t end)
+static int hugetlbfs_zero_partial_page(struct hstate *h,
+ struct address_space *mapping,
+ loff_t start, loff_t end)
{
pgoff_t idx = start >> huge_page_shift(h);
struct folio *folio;

folio = filemap_lock_hugetlb_folio(h, mapping, idx);
if (IS_ERR(folio))
- return;
+ return PTR_ERR(folio);

start = start & ~huge_page_mask(h);
end = end & ~huge_page_mask(h);
@@ -673,6 +672,7 @@ static long hugetlbfs_punch_hole(struct inode *inode, loff_t offset, loff_t len)
struct hstate *h = hstate_inode(inode);
loff_t hpage_size = huge_page_size(h);
loff_t hole_start, hole_end;
+ int rc;

/*
* hole_start and hole_end indicate the full pages within the hole.
@@ -688,12 +688,18 @@ static long hugetlbfs_punch_hole(struct inode *inode, loff_t offset, loff_t len)
return -EPERM;
}

+repeat:
i_mmap_lock_write(mapping);

/* If range starts before first full page, zero partial page. */
- if (offset < hole_start)
- hugetlbfs_zero_partial_page(h, mapping,
- offset, min(offset + len, hole_start));
+ if (offset < hole_start) {
+ rc = hugetlbfs_zero_partial_page(h, mapping, offset,
+ min(offset + len, hole_start));
+ if (rc == -EAGAIN) {
+ i_mmap_unlock_write(mapping);
+ goto repeat;
+ }
+ }

/* Unmap users of full pages in the hole. */
if (hole_end > hole_start) {
@@ -704,9 +710,14 @@ static long hugetlbfs_punch_hole(struct inode *inode, loff_t offset, loff_t len)
}

/* If range extends beyond last full page, zero partial page. */
- if ((offset + len) > hole_end && (offset + len) > hole_start)
- hugetlbfs_zero_partial_page(h, mapping,
- hole_end, offset + len);
+ if ((offset + len) > hole_end && (offset + len) > hole_start) {
+ rc = hugetlbfs_zero_partial_page(h, mapping, hole_end,
+ offset + len);
+ if (rc == -EAGAIN) {
+ i_mmap_unlock_write(mapping);
+ goto repeat;
+ }
+ }

i_mmap_unlock_write(mapping);

diff --git a/include/linux/hugetlb.h b/include/linux/hugetlb.h
index 8e63e46b8e1f..e7d99183361a 100644
--- a/include/linux/hugetlb.h
+++ b/include/linux/hugetlb.h
@@ -811,7 +811,7 @@ static inline unsigned int blocks_per_huge_page(struct hstate *h)
static inline struct folio *filemap_lock_hugetlb_folio(struct hstate *h,
struct address_space *mapping, pgoff_t idx)
{
- return filemap_lock_folio(mapping, idx << huge_page_order(h));
+ return filemap_lock_folio_nowait(mapping, idx << huge_page_order(h));
}

#include <asm/hugetlb.h>
diff --git a/include/linux/pagemap.h b/include/linux/pagemap.h
index 09b581c1d878..94f31035fe1d 100644
--- a/include/linux/pagemap.h
+++ b/include/linux/pagemap.h
@@ -820,6 +820,12 @@ static inline struct folio *filemap_lock_folio(struct address_space *mapping,
return __filemap_get_folio(mapping, index, FGP_LOCK, 0);
}

+static inline struct folio *filemap_lock_folio_nowait(struct address_space *mapping,
+ pgoff_t index)
+{
+ return __filemap_get_folio(mapping, index, FGP_LOCK | FGP_NOWAIT, 0);
+}
+
/**
* filemap_grab_folio - grab a folio from the page cache
* @mapping: The address space to search
--
2.43.0

[syzbot]

For archival purposes, forwarding an incoming command email to

linux-...@vger.kernel.org, syzkall...@googlegroups.com.

***

Subject: [PATCH] try fix
Author: wangjin...@gmail.com

#syz test

Signed-off-by: Jinchao Wang <wangjin...@gmail.com>
---
fs/hugetlbfs/inode.c | 34 +++++++++++++++++++++++-----------

include/linux/hugetlb.h | 2 +-
include/linux/pagemap.h | 6 ++++++

3 files changed, 30 insertions(+), 12 deletions(-)

diff --git a/fs/hugetlbfs/inode.c b/fs/hugetlbfs/inode.c
index 3b4c152c5c73..e903344aa0ec 100644
--- a/fs/hugetlbfs/inode.c
+++ b/fs/hugetlbfs/inode.c
@@ -653,17 +653,16 @@ static void hugetlb_vmtruncate(struct inode *inode, loff_t offset)

remove_inode_hugepages(inode, offset, LLONG_MAX);
}

-static void hugetlbfs_zero_partial_page(struct hstate *h,
- struct address_space *mapping,
- loff_t start,
- loff_t end)
+static int hugetlbfs_zero_partial_page(struct hstate *h,
+ struct address_space *mapping,
+ loff_t start, loff_t end)
{
pgoff_t idx = start >> huge_page_shift(h);
struct folio *folio;

folio = filemap_lock_hugetlb_folio(h, mapping, idx);
if (IS_ERR(folio))
- return;
+ return PTR_ERR(folio);

start = start & ~huge_page_mask(h);
end = end & ~huge_page_mask(h);

@@ -674,6 +673,7 @@ static void hugetlbfs_zero_partial_page(struct hstate *h,

folio_unlock(folio);
folio_put(folio);
+ return 0;

}

static long hugetlbfs_punch_hole(struct inode *inode, loff_t offset, loff_t len)

@@ -683,6 +683,7 @@ static long hugetlbfs_punch_hole(struct inode *inode, loff_t offset, loff_t len)

struct hstate *h = hstate_inode(inode);
loff_t hpage_size = huge_page_size(h);
loff_t hole_start, hole_end;
+ int rc;

/*
* hole_start and hole_end indicate the full pages within the hole.

@@ -698,12 +699,18 @@ static long hugetlbfs_punch_hole(struct inode *inode, loff_t offset, loff_t len)

return -EPERM;
}

+repeat:
i_mmap_lock_write(mapping);

/* If range starts before first full page, zero partial page. */
- if (offset < hole_start)
- hugetlbfs_zero_partial_page(h, mapping,
- offset, min(offset + len, hole_start));
+ if (offset < hole_start) {
+ rc = hugetlbfs_zero_partial_page(h, mapping, offset,
+ min(offset + len, hole_start));
+ if (rc == -EAGAIN) {
+ i_mmap_unlock_write(mapping);
+ goto repeat;
+ }
+ }

/* Unmap users of full pages in the hole. */
if (hole_end > hole_start) {

@@ -714,9 +721,14 @@ static long hugetlbfs_punch_hole(struct inode *inode, loff_t offset, loff_t len)

}

/* If range extends beyond last full page, zero partial page. */
- if ((offset + len) > hole_end && (offset + len) > hole_start)
- hugetlbfs_zero_partial_page(h, mapping,
- hole_end, offset + len);
+ if ((offset + len) > hole_end && (offset + len) > hole_start) {
+ rc = hugetlbfs_zero_partial_page(h, mapping, hole_end,
+ offset + len);
+ if (rc == -EAGAIN) {
+ i_mmap_unlock_write(mapping);
+ goto repeat;
+ }
+ }

i_mmap_unlock_write(mapping);

diff --git a/include/linux/hugetlb.h b/include/linux/hugetlb.h

index 019a1c5281e4..ad55b9dada0a 100644
--- a/include/linux/hugetlb.h
+++ b/include/linux/hugetlb.h
@@ -814,7 +814,7 @@ static inline unsigned int blocks_per_huge_page(struct hstate *h)

static inline struct folio *filemap_lock_hugetlb_folio(struct hstate *h,
struct address_space *mapping, pgoff_t idx)
{
- return filemap_lock_folio(mapping, idx << huge_page_order(h));
+ return filemap_lock_folio_nowait(mapping, idx << huge_page_order(h));
}

#include <asm/hugetlb.h>
diff --git a/include/linux/pagemap.h b/include/linux/pagemap.h

index 31a848485ad9..b9d818a9409b 100644
--- a/include/linux/pagemap.h
+++ b/include/linux/pagemap.h
@@ -825,6 +825,12 @@ static inline struct folio *filemap_lock_folio(struct address_space *mapping,

[syzbot]

For archival purposes, forwarding an incoming command email to
linux-...@vger.kernel.org, syzkall...@googlegroups.com.

***

Subject: test upstream
Author: wangjin...@gmail.com

#syz test

[Matthew Wilcox]

On Fri, Oct 10, 2025 at 01:49:30PM -0700, syzbot wrote:
> syz repro: https://syzkaller.appspot.com/x/repro.syz?x=133e89e2580000
> C reproducer: https://syzkaller.appspot.com/x/repro.c?x=17f3ba7c580000

#syz test

diff --git a/mm/migrate.c b/mm/migrate.c
index 5169f9717f60..4688b9e38cd2 100644
--- a/mm/migrate.c
+++ b/mm/migrate.c
@@ -1458,6 +1458,7 @@ static int unmap_and_move_huge_page(new_folio_t get_new_folio,
int page_was_mapped = 0;
struct anon_vma *anon_vma = NULL;
struct address_space *mapping = NULL;
+ enum ttu_flags ttu = 0;

if (folio_ref_count(src) == 1) {
/* page was freed from under us. So we are done. */
@@ -1498,8 +1499,6 @@ static int unmap_and_move_huge_page(new_folio_t get_new_folio,
goto put_anon;

if (folio_mapped(src)) {
- enum ttu_flags ttu = 0;
-
if (!folio_test_anon(src)) {
/*
* In shared mappings, try_to_unmap could potentially
@@ -1516,16 +1515,17 @@ static int unmap_and_move_huge_page(new_folio_t get_new_folio,

try_to_migrate(src, ttu);
page_was_mapped = 1;
-
- if (ttu & TTU_RMAP_LOCKED)
- i_mmap_unlock_write(mapping);
}

if (!folio_mapped(src))
rc = move_to_new_folio(dst, src, mode);

if (page_was_mapped)
- remove_migration_ptes(src, !rc ? dst : src, 0);
+ remove_migration_ptes(src, !rc ? dst : src,
+ ttu ? RMP_LOCKED : 0);
+
+ if (ttu & TTU_RMAP_LOCKED)
+ i_mmap_unlock_write(mapping);

unlock_put_anon:
folio_unlock(dst);

[syzbot]

Hello,

syzbot has tested the proposed patch and the reproducer did not trigger any issue:

Reported-by: syzbot+2d9c96...@syzkaller.appspotmail.com
Tested-by: syzbot+2d9c96...@syzkaller.appspotmail.com

Tested on:

commit: f0b9d8eb Merge tag 'nfsd-6.19-3' of git://git.kernel.o..
git tree: upstream
console output: https://syzkaller.appspot.com/x/log.txt?x=13df11fc580000
kernel config: https://syzkaller.appspot.com/x/.config?x=e5a01f6352682866

dashboard link: https://syzkaller.appspot.com/bug?extid=2d9c96466c978346b55f
compiler: Debian clang version 20.1.8 (++20250708063551+0c9f909b7976-1~exp1~20250708183702.136), Debian LLD 20.1.8

patch: https://syzkaller.appspot.com/x/patch.diff?x=159dc19a580000

Note: testing is done by a robot and is best-effort only.