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 --- drivers/vhost/test.c | 379 +++++++++++++++++++++++++++++++++++++++++++++++++++ 1 file changed, 379 insertions(+) create mode 100644 drivers/vhost/test.c (limited to 'drivers/vhost/test.c') diff --git a/drivers/vhost/test.c b/drivers/vhost/test.c new file mode 100644 index 000000000..24514c8fd --- /dev/null +++ b/drivers/vhost/test.c @@ -0,0 +1,379 @@ +// SPDX-License-Identifier: GPL-2.0-only +/* Copyright (C) 2009 Red Hat, Inc. + * Author: Michael S. Tsirkin + * + * test virtio server in host kernel. + */ + +#include +#include +#include +#include +#include +#include +#include +#include +#include + +#include "test.h" +#include "vhost.h" + +/* Max number of bytes transferred before requeueing the job. + * Using this limit prevents one virtqueue from starving others. */ +#define VHOST_TEST_WEIGHT 0x80000 + +/* Max number of packets transferred before requeueing the job. + * Using this limit prevents one virtqueue from starving others with + * pkts. + */ +#define VHOST_TEST_PKT_WEIGHT 256 + +static const int vhost_test_bits[] = { + VHOST_FEATURES +}; + +#define VHOST_TEST_FEATURES VHOST_FEATURES_U64(vhost_test_bits, 0) + +enum { + VHOST_TEST_VQ = 0, + VHOST_TEST_VQ_MAX = 1, +}; + +struct vhost_test { + struct vhost_dev dev; + struct vhost_virtqueue vqs[VHOST_TEST_VQ_MAX]; +}; + +/* Expects to be always run from workqueue - which acts as + * read-size critical section for our kind of RCU. */ +static void handle_vq(struct vhost_test *n) +{ + struct vhost_virtqueue *vq = &n->vqs[VHOST_TEST_VQ]; + unsigned out, in; + int head; + size_t len, total_len = 0; + void *private; + + mutex_lock(&vq->mutex); + private = vhost_vq_get_backend(vq); + if (!private) { + mutex_unlock(&vq->mutex); + return; + } + + vhost_disable_notify(&n->dev, vq); + + for (;;) { + head = vhost_get_vq_desc(vq, vq->iov, + ARRAY_SIZE(vq->iov), + &out, &in, + NULL, NULL); + /* On error, stop handling until the next kick. */ + if (unlikely(head < 0)) + break; + /* Nothing new? Wait for eventfd to tell us they refilled. */ + if (head == vq->num) { + if (unlikely(vhost_enable_notify(&n->dev, vq))) { + vhost_disable_notify(&n->dev, vq); + continue; + } + break; + } + if (in) { + vq_err(vq, "Unexpected descriptor format for TX: " + "out %d, int %d\n", out, in); + break; + } + len = iov_length(vq->iov, out); + /* Sanity check */ + if (!len) { + vq_err(vq, "Unexpected 0 len for TX\n"); + break; + } + vhost_add_used_and_signal(&n->dev, vq, head, 0); + total_len += len; + if (unlikely(vhost_exceeds_weight(vq, 0, total_len))) + break; + } + + mutex_unlock(&vq->mutex); +} + +static void handle_vq_kick(struct vhost_work *work) +{ + struct vhost_virtqueue *vq = container_of(work, struct vhost_virtqueue, + poll.work); + struct vhost_test *n = container_of(vq->dev, struct vhost_test, dev); + + handle_vq(n); +} + +static int vhost_test_open(struct inode *inode, struct file *f) +{ + struct vhost_test *n = kmalloc_obj(*n); + struct vhost_dev *dev; + struct vhost_virtqueue **vqs; + + if (!n) + return -ENOMEM; + vqs = kmalloc_objs(*vqs, VHOST_TEST_VQ_MAX); + if (!vqs) { + kfree(n); + return -ENOMEM; + } + + dev = &n->dev; + vqs[VHOST_TEST_VQ] = &n->vqs[VHOST_TEST_VQ]; + n->vqs[VHOST_TEST_VQ].handle_kick = handle_vq_kick; + vhost_dev_init(dev, vqs, VHOST_TEST_VQ_MAX, UIO_MAXIOV, + VHOST_TEST_PKT_WEIGHT, VHOST_TEST_WEIGHT, true, NULL); + + f->private_data = n; + + return 0; +} + +static void *vhost_test_stop_vq(struct vhost_test *n, + struct vhost_virtqueue *vq) +{ + void *private; + + mutex_lock(&vq->mutex); + private = vhost_vq_get_backend(vq); + vhost_vq_set_backend(vq, NULL); + mutex_unlock(&vq->mutex); + return private; +} + +static void vhost_test_stop(struct vhost_test *n, void **privatep) +{ + *privatep = vhost_test_stop_vq(n, n->vqs + VHOST_TEST_VQ); +} + +static void vhost_test_flush(struct vhost_test *n) +{ + vhost_dev_flush(&n->dev); +} + +static int vhost_test_release(struct inode *inode, struct file *f) +{ + struct vhost_test *n = f->private_data; + void *private; + + vhost_test_stop(n, &private); + vhost_test_flush(n); + vhost_dev_stop(&n->dev); + vhost_dev_cleanup(&n->dev); + kfree(n->dev.vqs); + kfree(n); + return 0; +} + +static long vhost_test_run(struct vhost_test *n, int test) +{ + void *priv, *oldpriv; + struct vhost_virtqueue *vq; + int r, index; + + if (test < 0 || test > 1) + return -EINVAL; + + mutex_lock(&n->dev.mutex); + r = vhost_dev_check_owner(&n->dev); + if (r) + goto err; + + for (index = 0; index < n->dev.nvqs; ++index) { + /* Verify that ring has been setup correctly. */ + if (!vhost_vq_access_ok(&n->vqs[index])) { + r = -EFAULT; + goto err; + } + } + + for (index = 0; index < n->dev.nvqs; ++index) { + vq = n->vqs + index; + mutex_lock(&vq->mutex); + priv = test ? n : NULL; + + /* start polling new socket */ + oldpriv = vhost_vq_get_backend(vq); + vhost_vq_set_backend(vq, priv); + + r = vhost_vq_init_access(&n->vqs[index]); + + mutex_unlock(&vq->mutex); + + if (r) + goto err; + + if (oldpriv) { + vhost_test_flush(n); + } + } + + mutex_unlock(&n->dev.mutex); + return 0; + +err: + mutex_unlock(&n->dev.mutex); + return r; +} + +static long vhost_test_reset_owner(struct vhost_test *n) +{ + void *priv = NULL; + long err; + struct vhost_iotlb *umem; + + mutex_lock(&n->dev.mutex); + err = vhost_dev_check_owner(&n->dev); + if (err) + goto done; + umem = vhost_dev_reset_owner_prepare(); + if (!umem) { + err = -ENOMEM; + goto done; + } + vhost_test_stop(n, &priv); + vhost_test_flush(n); + vhost_dev_stop(&n->dev); + vhost_dev_reset_owner(&n->dev, umem); +done: + mutex_unlock(&n->dev.mutex); + return err; +} + +static int vhost_test_set_features(struct vhost_test *n, u64 features) +{ + struct vhost_virtqueue *vq; + + mutex_lock(&n->dev.mutex); + if ((features & (1 << VHOST_F_LOG_ALL)) && + !vhost_log_access_ok(&n->dev)) { + mutex_unlock(&n->dev.mutex); + return -EFAULT; + } + vq = &n->vqs[VHOST_TEST_VQ]; + mutex_lock(&vq->mutex); + vq->acked_features = features; + mutex_unlock(&vq->mutex); + mutex_unlock(&n->dev.mutex); + return 0; +} + +static long vhost_test_set_backend(struct vhost_test *n, unsigned index, int fd) +{ + static void *backend; + + const bool enable = fd != -1; + struct vhost_virtqueue *vq; + int r; + + mutex_lock(&n->dev.mutex); + r = vhost_dev_check_owner(&n->dev); + if (r) + goto err; + + if (index >= VHOST_TEST_VQ_MAX) { + r = -ENOBUFS; + goto err; + } + vq = &n->vqs[index]; + mutex_lock(&vq->mutex); + + /* Verify that ring has been setup correctly. */ + if (!vhost_vq_access_ok(vq)) { + r = -EFAULT; + goto err_vq; + } + if (!enable) { + vhost_poll_stop(&vq->poll); + backend = vhost_vq_get_backend(vq); + vhost_vq_set_backend(vq, NULL); + } else { + vhost_vq_set_backend(vq, backend); + r = vhost_vq_init_access(vq); + if (r == 0) + r = vhost_poll_start(&vq->poll, vq->kick); + } + + mutex_unlock(&vq->mutex); + + if (enable) { + vhost_test_flush(n); + } + + mutex_unlock(&n->dev.mutex); + return 0; + +err_vq: + mutex_unlock(&vq->mutex); +err: + mutex_unlock(&n->dev.mutex); + return r; +} + +static long vhost_test_ioctl(struct file *f, unsigned int ioctl, + unsigned long arg) +{ + struct vhost_vring_file backend; + struct vhost_test *n = f->private_data; + void __user *argp = (void __user *)arg; + u64 __user *featurep = argp; + int test; + u64 features; + int r; + switch (ioctl) { + case VHOST_TEST_RUN: + if (copy_from_user(&test, argp, sizeof test)) + return -EFAULT; + return vhost_test_run(n, test); + case VHOST_TEST_SET_BACKEND: + if (copy_from_user(&backend, argp, sizeof backend)) + return -EFAULT; + return vhost_test_set_backend(n, backend.index, backend.fd); + case VHOST_GET_FEATURES: + features = VHOST_TEST_FEATURES; + if (copy_to_user(featurep, &features, sizeof features)) + return -EFAULT; + return 0; + case VHOST_SET_FEATURES: + if (copy_from_user(&features, featurep, sizeof features)) + return -EFAULT; + if (features & ~VHOST_TEST_FEATURES) + return -EOPNOTSUPP; + return vhost_test_set_features(n, features); + case VHOST_RESET_OWNER: + return vhost_test_reset_owner(n); + default: + mutex_lock(&n->dev.mutex); + r = vhost_dev_ioctl(&n->dev, ioctl, argp); + if (r == -ENOIOCTLCMD) + r = vhost_vring_ioctl(&n->dev, ioctl, argp); + vhost_test_flush(n); + mutex_unlock(&n->dev.mutex); + return r; + } +} + +static const struct file_operations vhost_test_fops = { + .owner = THIS_MODULE, + .release = vhost_test_release, + .unlocked_ioctl = vhost_test_ioctl, + .compat_ioctl = compat_ptr_ioctl, + .open = vhost_test_open, + .llseek = noop_llseek, +}; + +static struct miscdevice vhost_test_misc = { + MISC_DYNAMIC_MINOR, + "vhost-test", + &vhost_test_fops, +}; +module_misc_device(vhost_test_misc); + +MODULE_VERSION("0.0.1"); +MODULE_LICENSE("GPL v2"); +MODULE_AUTHOR("Michael S. Tsirkin"); +MODULE_DESCRIPTION("Host kernel side for virtio simulator"); -- cgit v1.3.1