KCSAN: data-race in percpu_counter_add_batch / try_to_unmap_one

Status: moderation: reported on 2026/11/18 06:16
Subsystems: mm
Labels: prio:low race:benign
[Documentation on labels]
Reported-by: syzbot+9194be7e22e1abaaa239@syzkaller.appspotmail.com
First crash: 22h09m, last: 8h25m
✨ AI-generated root cause summary:

A KCSAN data race occurs between a marked lockless read in percpu_counter_read_positive() and an unmarked locked update in percpu_counter_add_batch(). When update_hiwater_rss() reads mm->rss_stat via percpu_counter_read_positive(), it executes READ_ONCE(fbc->count) without holding fbc->lock. Concurrently, a page fault slow path in percpu_counter_add_batch() updates fbc->count via a plain read-modify-write (fbc->count += count + amount) under fbc->lock. Under the Linux Kernel Memory Model and KCSAN (with CONFIG_KCSAN_ASSUME_PLAIN_WRITES_ATOMIC disabled), pairing an unannotated lockless read with a plain locked write constitutes a data race.

▶ ▼ ✨ Full AI root cause analysis
Based on kernel commit 63fdb511c27a:

Root Cause Analysis

1. Functions, Structs, and Fields Involved

  • Data Object: mm->rss_stat[MM_FILEPAGES].count (s64 count within struct percpu_counter, defined in include/linux/percpu_counter.h:24 and embedded in struct mm_struct at include/linux/mm_types.h:1353).
  • Writer: percpu_counter_add_batch() (lib/percpu_counter.c:107), called via:
    filemap_map_pages() mm/filemap.c:4022
      -> add_mm_counter() include/linux/mm.h:3687
        -> percpu_counter_add() include/linux/percpu_counter.h:71
          -> percpu_counter_add_batch() lib/percpu_counter.c:107
    
  • Reader: percpu_counter_read_positive() (include/linux/percpu_counter.h:121), called via:
    try_to_unmap_one() mm/rmap.c:2360
      -> update_hiwater_rss() include/linux/mm.h:3747
        -> get_mm_rss() include/linux/mm.h:3723
          -> get_mm_counter() include/linux/mm.h:3675
            -> percpu_counter_read_positive() include/linux/percpu_counter.h:121
    

2. Trigger Mechanism and Concurrency Violation

  1. Slow-path counter update in percpu_counter_add_batch(): During fault-around in filemap_map_pages(), task 3805 updates the process RSS counter via add_mm_counter(). When the per-CPU batch overflows (abs(count + amount) >= batch), it enters the slow path:

    raw_spin_lock_irqsave(&fbc->lock, flags);
    count = __this_cpu_read(*fbc->counters);
    fbc->count += count + amount;
    __this_cpu_sub(*fbc->counters, count);
    raw_spin_unlock_irqrestore(&fbc->lock, flags);
    

    While this read-modify-write is serialized against other writers by fbc->lock, the assignment to fbc->count is a plain, unmarked C write (fbc->count += count + amount;).

  2. Lockless statistical read in percpu_counter_read_positive(): Concurrently, task 3804 is running folio unmapping / page reclaim in try_to_unmap_one(). Before decrementing RSS, it calls update_hiwater_rss(), which samples the current RSS usage via get_mm_rss() -> percpu_counter_read_positive(&mm->rss_stat[member]). percpu_counter_read_positive() is designed for fast, lockless approximate reads and does not acquire fbc->lock:

    static inline s64 percpu_counter_read_positive(struct percpu_counter *fbc)
    {
        /* Prevent reloads of fbc->count */
        s64 ret = READ_ONCE(fbc->count);
    
        if (ret >= 0)
            return ret;
        return 0;
    }
    
  3. Violation of LKMM / KCSAN rules: Under the Linux Kernel Memory Model (LKMM) and with # CONFIG_KCSAN_ASSUME_PLAIN_WRITES_ATOMIC is not set, KCSAN sets watchpoints on plain writes. Pairing an unannotated lockless reader (READ_ONCE(fbc->count)) with a plain write (fbc->count += ...) triggers a data-race report because compilers are theoretically allowed to tear or refetch plain memory accesses.

    While commit 8535c259c8fb previously addressed this identical scenario in __percpu_counter_limited_add() by annotating the lockless read as data_race(READ_ONCE(fbc->count)) (noting that READ_ONCE() alone still triggers KCSAN against plain writes), percpu_counter_read_positive() in include/linux/percpu_counter.h still used plain READ_ONCE(fbc->count) without data_race(...). Conversely, the locked updates to fbc->count in lib/percpu_counter.c remain plain writes rather than WRITE_ONCE() or data_race().

▶ ▼ ✨ AI Jobs (3)
ID Workflow Result Correct Bug Created Started Finished Revision
948dd161 bisect-cause 🏃 KCSAN: data-race in percpu_counter_add_batch / try_to_unmap_one 2026/11/18 06:26 2026/11/18 06:26 eb4895d7
906a9a83 assessment-kcsan Benign: ✅ ❓ KCSAN: data-race in percpu_counter_add_batch / try_to_unmap_one 2026/11/18 06:16 2026/11/18 06:16 2026/11/18 06:30 eb4895d7
98cd1c6d assessment-security DenialOfService: ❌ Exploitable: ❌ FilesystemTrigger: ❌ FuzzingArtifact: ❌ MemoryCorruption: ❌ NetworkNeedsSetup: ❌ NetworkTrigger: ❌ PeripheralTrigger: ❌ RemoteTrigger: ❌ Unprivileged: ✅ UserNamespace: ✅ VMGuestTrigger: ❌ VMHostTrigger: ❌ ❓ KCSAN: data-race in percpu_counter_add_batch / try_to_unmap_one 2026/11/18 06:13 2026/11/18 06:13 2026/11/18 06:16 eb4895d7

Sample crash report:
==================================================================
BUG: KCSAN: data-race in percpu_counter_add_batch / try_to_unmap_one

read (marked) to 0xffff8881048d3ac0 of 8 bytes by task 3804 on cpu 1:
 percpu_counter_read_positive include/linux/percpu_counter.h:121 [inline]
 get_mm_counter include/linux/mm.h:3675 [inline]
 get_mm_rss include/linux/mm.h:3723 [inline]
 update_hiwater_rss include/linux/mm.h:3747 [inline]
 try_to_unmap_one+0xcc8/0x1bb0 mm/rmap.c:2360
 __rmap_walk_file+0x1c9/0x280 mm/rmap.c:3121
 try_to_unmap+0x245/0x270 mm/rmap.c:-1
 shrink_folio_list+0xf84/0x2d70 mm/vmscan.c:1457
 evict_folios+0x1fa4/0x2990 mm/vmscan.c:5097
 try_to_shrink_lruvec+0x9fb/0xd60 mm/vmscan.c:5271
 lru_gen_shrink_lruvec mm/vmscan.c:5422 [inline]
 shrink_lruvec mm/vmscan.c:6199 [inline]
 shrink_node_memcgs mm/vmscan.c:6438 [inline]
 shrink_node+0x97e/0x4600 mm/vmscan.c:6482
 shrink_zones mm/vmscan.c:6701 [inline]
 do_try_to_free_pages+0x310/0xb80 mm/vmscan.c:6763
 try_to_free_mem_cgroup_pages+0x201/0x450 mm/vmscan.c:7044
 mem_cgroup_resize_max+0x161/0x310 mm/memcontrol-v1.c:1466
 mem_cgroup_write+0x132/0x1c0 mm/memcontrol-v1.c:-1
 cgroup_file_write+0x197/0x350 kernel/cgroup/cgroup.c:4426
 kernfs_fop_write_iter+0x1d2/0x2e0 fs/kernfs/file.c:345
 new_sync_write fs/read_write.c:575 [inline]
 vfs_write+0x57a/0x9b0 fs/read_write.c:667
 ksys_write+0xdc/0x1a0 fs/read_write.c:719
 __do_sys_write fs/read_write.c:730 [inline]
 __se_sys_write fs/read_write.c:727 [inline]
 __x64_sys_write+0x40/0x50 fs/read_write.c:727
 x64_sys_call+0x154e/0x2550 arch/x86/include/generated/asm/syscalls_64.h:2
 do_syscall_x64 arch/x86/entry/syscall_64.c:61 [inline]
 do_syscall_64+0x11e/0x3a0 arch/x86/entry/syscall_64.c:89
 entry_SYSCALL_64_after_hwframe+0x77/0x7f

read-write to 0xffff8881048d3ac0 of 8 bytes by task 3805 on cpu 0:
 percpu_counter_add_batch+0x101/0x120 lib/percpu_counter.c:107
 percpu_counter_add include/linux/percpu_counter.h:71 [inline]
 add_mm_counter include/linux/mm.h:3687 [inline]
 filemap_map_pages+0xd54/0xed0 mm/filemap.c:4022
 do_fault_around mm/memory.c:6094 [inline]
 do_read_fault mm/memory.c:6127 [inline]
 do_fault mm/memory.c:6270 [inline]
 do_pte_missing mm/memory.c:4797 [inline]
 handle_pte_fault mm/memory.c:6689 [inline]
 __handle_mm_fault mm/memory.c:6841 [inline]
 handle_mm_fault+0x1f41/0x3880 mm/memory.c:7008
 faultin_page mm/gup.c:1140 [inline]
 __get_user_pages+0xf69/0x19e0 mm/gup.c:1442
 __get_user_pages_locked mm/gup.c:1706 [inline]
 get_dump_page+0xc9/0x250 mm/gup.c:2221
 dump_user_range+0x15c/0x950 fs/coredump.c:1595
 elf_core_dump+0x2183/0x2310 fs/binfmt_elf.c:2106
 coredump_write fs/coredump.c:1150 [inline]
 do_coredump+0x3abb/0x4040 fs/coredump.c:1237
 vfs_coredump+0x757/0x990 fs/coredump.c:1312
 get_signal+0xb41/0xd30 kernel/signal.c:3098
 arch_do_signal_or_restart+0x96/0x480 arch/x86/kernel/signal.c:337
 __exit_to_user_mode_loop kernel/entry/common.c:66 [inline]
 exit_to_user_mode_loop kernel/entry/common.c:101 [inline]
 __exit_to_user_mode_prepare include/linux/irq-entry-common.h:207 [inline]
 irqentry_exit_to_user_mode_prepare include/linux/irq-entry-common.h:244 [inline]
 irqentry_exit_to_user_mode include/linux/irq-entry-common.h:315 [inline]
 irqentry_exit+0x15f/0x640 kernel/entry/common.c:152
 asm_exc_page_fault+0x26/0x30 arch/x86/include/asm/idtentry.h:595

Reported by Kernel Concurrency Sanitizer on:
CPU: 0 UID: 0 PID: 3805 Comm: syz.3.67 Not tainted syzkaller #0 PREEMPT(lazy) 
Hardware name: Google Google Compute Engine/Google Compute Engine, BIOS Google 11/06/2026
==================================================================
syz.3.67 (3805) used greatest stack depth: 9784 bytes left

Crashes (3):
Time Kernel Commit Syzkaller Config Log Report Syz repro C repro VM info Assets (help?) Manager Title
2026/11/17 22:34 upstream 63fdb511c27a 122f94d5 .config console log report info [disk image] [vmlinux] [kernel image] ci2-upstream-kcsan-gce KCSAN: data-race in percpu_counter_add_batch / try_to_unmap_one
2026/11/17 21:01 upstream 666539eecdf4 122f94d5 .config console log report info [disk image] [vmlinux] [kernel image] ci2-upstream-kcsan-gce KCSAN: data-race in percpu_counter_add_batch / try_to_unmap_one
2026/11/17 08:50 upstream 666539eecdf4 a3e64cfc .config console log report info [disk image] [vmlinux] [kernel image] ci2-upstream-kcsan-gce KCSAN: data-race in percpu_counter_add_batch / try_to_unmap_one
* Struck through repros no longer work on HEAD.