From a5f7a5bb3b7f28ba7e4fa246775b29a0e5537255 Mon Sep 17 00:00:00 2001 From: Jinke Han Date: Tue, 8 Sep 2026 15:37:42 +0800 Subject: x86/kprobes: Fix crash when probing CS CALL instructions When using eBPF to probe CS CALL instructions within a function, a crash can be triggered. The eBPF tool probes offset 257 of the __hrtimer_run_queues() function: <__hrtimer_run_queues+249>: nopl 0x0(%rax,%rax,1) <__hrtimer_run_queues+254>: mov %r14,%rdi <__hrtimer_run_queues+257>: cs call <__x86_indirect_thunk_r12> <__hrtimer_run_queues+263>: mov %eax,%r12d <__hrtimer_run_queues+266>: xchg %ax,%ax <__hrtimer_run_queues+268>: mov %r13,%rdi Which triggers this crash: BUG: unable to handle page fault for address: 00000000000f41c9 #PF: supervisor write access in kernel mode #PF: error_code(0x0002) - not-present page PGD 0 P4D 0 Oops: 0002 [#1] SMP NOPTI CPU: 1 PID: 0 Comm: swapper/1 Kdump: loaded Tainted: P RIP: 0010:__hrtimer_run_queues+0x106/0x230 Note that __hrtimer_run_queues+0x106 is __hrtimer_run_queues+262, which is at the 6th byte of the above CS CALL instruction. Since the CS CALL instruction occupies 6 bytes, the exception occurred in the middle of that call instruction. The root cause is that when using eBPF tools to probe in the middle of a function, a kprobe with INT3 is used as the underlying implementation. During single-step emulation of the original CALL instruction, int3_emulate_call() assumes that the probed CALL instruction is 5 bytes long. However, the actual CS-prefixed CALL instruction occupies 6 bytes, so it constructs an incorrect exception return address. When the CPU returns from the kprobe handler, the next instruction to be executed is at the address of the last byte of that CS CALL instruction. Coincidentally, starting from that address, the CPU fetches and decodes a completely different instruction, which ultimately triggers a kernel crash. Fix the issue by using the actual instruction length obtained from the instruction decoder when constructing the exception return address, rather than relying on the hardcoded CALL_INSN_SIZE macro. [ mingo: Refined the changelog ] Fixes: 6256e668b7af ("x86/kprobes: Use int3 instead of debug trap for single-step") Suggested-by: Masami Hiramatsu (Google) Signed-off-by: Jinke Han Signed-off-by: Ingo Molnar Reviewed-by: Masami Hiramatsu (Google) Acked-by: Yafang Shao Acked-by: Borislav Petkov Cc: Peter Zijlstra Link: https://patch.msgid.link/20260908073742.GA10517@didi-ThinkCentre-M920t-N000 --- samples/hung_task/Makefile | 2 + samples/hung_task/hung_task_tests.c | 164 ++++++++++++++++++++++++++++++++++++ 2 files changed, 166 insertions(+) create mode 100644 samples/hung_task/Makefile create mode 100644 samples/hung_task/hung_task_tests.c (limited to 'samples/hung_task') diff --git a/samples/hung_task/Makefile b/samples/hung_task/Makefile new file mode 100644 index 000000000..86036f1a2 --- /dev/null +++ b/samples/hung_task/Makefile @@ -0,0 +1,2 @@ +# SPDX-License-Identifier: GPL-2.0-only +obj-$(CONFIG_SAMPLE_HUNG_TASK) += hung_task_tests.o diff --git a/samples/hung_task/hung_task_tests.c b/samples/hung_task/hung_task_tests.c new file mode 100644 index 000000000..0360ec916 --- /dev/null +++ b/samples/hung_task/hung_task_tests.c @@ -0,0 +1,164 @@ +// SPDX-License-Identifier: GPL-2.0-or-later +/* + * hung_task_tests.c - Sample code for testing hung tasks with mutex, + * semaphore, etc. + * + * Usage: Load this module and read `/hung_task/mutex`, + * `/hung_task/semaphore`, `/hung_task/rw_semaphore_read`, + * `/hung_task/rw_semaphore_write`, etc., with 2 or more processes. + * + * This is for testing kernel hung_task error messages with various locking + * mechanisms (e.g., mutex, semaphore, rw_semaphore_read, rw_semaphore_write, etc.). + * Note that this may freeze your system or cause a panic. Use only for testing purposes. + */ + +#include +#include +#include +#include +#include +#include +#include + +#define HUNG_TASK_DIR "hung_task" +#define HUNG_TASK_MUTEX_FILE "mutex" +#define HUNG_TASK_SEM_FILE "semaphore" +#define HUNG_TASK_RWSEM_READ_FILE "rw_semaphore_read" +#define HUNG_TASK_RWSEM_WRITE_FILE "rw_semaphore_write" +#define SLEEP_SECOND 256 + +static const char dummy_string[] = "This is a dummy string."; +static DEFINE_MUTEX(dummy_mutex); +static DEFINE_SEMAPHORE(dummy_sem, 1); +static DECLARE_RWSEM(dummy_rwsem); +static struct dentry *hung_task_dir; + +/* Mutex-based read function */ +static ssize_t read_dummy_mutex(struct file *file, char __user *user_buf, + size_t count, loff_t *ppos) +{ + /* Check if data is already read */ + if (*ppos >= sizeof(dummy_string)) + return 0; + + /* Second task waits on mutex, entering uninterruptible sleep */ + guard(mutex)(&dummy_mutex); + + /* First task sleeps here, interruptible */ + msleep_interruptible(SLEEP_SECOND * 1000); + + return simple_read_from_buffer(user_buf, count, ppos, dummy_string, + sizeof(dummy_string)); +} + +/* Semaphore-based read function */ +static ssize_t read_dummy_semaphore(struct file *file, char __user *user_buf, + size_t count, loff_t *ppos) +{ + /* Check if data is already read */ + if (*ppos >= sizeof(dummy_string)) + return 0; + + /* Second task waits on semaphore, entering uninterruptible sleep */ + down(&dummy_sem); + + /* First task sleeps here, interruptible */ + msleep_interruptible(SLEEP_SECOND * 1000); + + up(&dummy_sem); + + return simple_read_from_buffer(user_buf, count, ppos, dummy_string, + sizeof(dummy_string)); +} + +/* Read-write semaphore read function */ +static ssize_t read_dummy_rwsem_read(struct file *file, char __user *user_buf, + size_t count, loff_t *ppos) +{ + /* Check if data is already read */ + if (*ppos >= sizeof(dummy_string)) + return 0; + + /* Acquires read lock, allowing concurrent readers but blocks if write lock is held */ + down_read(&dummy_rwsem); + + /* Sleeps here, potentially triggering hung task detection if lock is held too long */ + msleep_interruptible(SLEEP_SECOND * 1000); + + up_read(&dummy_rwsem); + + return simple_read_from_buffer(user_buf, count, ppos, dummy_string, + sizeof(dummy_string)); +} + +/* Read-write semaphore write function */ +static ssize_t read_dummy_rwsem_write(struct file *file, char __user *user_buf, + size_t count, loff_t *ppos) +{ + /* Check if data is already read */ + if (*ppos >= sizeof(dummy_string)) + return 0; + + /* Acquires exclusive write lock, blocking all other readers and writers */ + down_write(&dummy_rwsem); + + /* Sleeps here, potentially triggering hung task detection if lock is held too long */ + msleep_interruptible(SLEEP_SECOND * 1000); + + up_write(&dummy_rwsem); + + return simple_read_from_buffer(user_buf, count, ppos, dummy_string, + sizeof(dummy_string)); +} + +/* File operations for mutex */ +static const struct file_operations hung_task_mutex_fops = { + .read = read_dummy_mutex, +}; + +/* File operations for semaphore */ +static const struct file_operations hung_task_sem_fops = { + .read = read_dummy_semaphore, +}; + +/* File operations for rw_semaphore read */ +static const struct file_operations hung_task_rwsem_read_fops = { + .read = read_dummy_rwsem_read, +}; + +/* File operations for rw_semaphore write */ +static const struct file_operations hung_task_rwsem_write_fops = { + .read = read_dummy_rwsem_write, +}; + +static int __init hung_task_tests_init(void) +{ + hung_task_dir = debugfs_create_dir(HUNG_TASK_DIR, NULL); + if (IS_ERR(hung_task_dir)) + return PTR_ERR(hung_task_dir); + + /* Create debugfs files for mutex and semaphore tests */ + debugfs_create_file(HUNG_TASK_MUTEX_FILE, 0400, hung_task_dir, NULL, + &hung_task_mutex_fops); + debugfs_create_file(HUNG_TASK_SEM_FILE, 0400, hung_task_dir, NULL, + &hung_task_sem_fops); + debugfs_create_file(HUNG_TASK_RWSEM_READ_FILE, 0400, hung_task_dir, NULL, + &hung_task_rwsem_read_fops); + debugfs_create_file(HUNG_TASK_RWSEM_WRITE_FILE, 0400, hung_task_dir, NULL, + &hung_task_rwsem_write_fops); + + return 0; +} + +static void __exit hung_task_tests_exit(void) +{ + debugfs_remove_recursive(hung_task_dir); +} + +module_init(hung_task_tests_init); +module_exit(hung_task_tests_exit); + +MODULE_LICENSE("GPL"); +MODULE_AUTHOR("Masami Hiramatsu "); +MODULE_AUTHOR("Zi Li "); +MODULE_DESCRIPTION("Simple sleep under lock files for testing hung task"); -- cgit v1.3.1