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 --- net/x25/x25_dev.c | 194 ++++++++++++++++++++++++++++++++++++++++++++++++++++++ 1 file changed, 194 insertions(+) create mode 100644 net/x25/x25_dev.c (limited to 'net/x25/x25_dev.c') diff --git a/net/x25/x25_dev.c b/net/x25/x25_dev.c new file mode 100644 index 000000000..fb8ac1aa5 --- /dev/null +++ b/net/x25/x25_dev.c @@ -0,0 +1,194 @@ +// SPDX-License-Identifier: GPL-2.0-or-later +/* + * X.25 Packet Layer release 002 + * + * This is ALPHA test software. This code may break your machine, randomly fail to work with new + * releases, misbehave and/or generally screw up. It might even work. + * + * This code REQUIRES 2.1.15 or higher + * + * History + * X.25 001 Jonathan Naylor Started coding. + * 2000-09-04 Henner Eisen Prevent freeing a dangling skb. + */ + +#define pr_fmt(fmt) "X25: " fmt + +#include +#include +#include +#include +#include +#include +#include +#include + +static int x25_receive_data(struct sk_buff *skb, struct x25_neigh *nb) +{ + struct sock *sk; + unsigned short frametype; + unsigned int lci; + + if (!pskb_may_pull(skb, X25_STD_MIN_LEN)) + return 0; + + frametype = skb->data[2]; + lci = ((skb->data[0] << 8) & 0xF00) + ((skb->data[1] << 0) & 0x0FF); + + /* + * LCI of zero is always for us, and its always a link control + * frame. + */ + if (lci == 0) { + x25_link_control(skb, nb, frametype); + return 0; + } + + /* + * Find an existing socket. + */ + if ((sk = x25_find_socket(lci, nb)) != NULL) { + int queued = 1; + + skb_reset_transport_header(skb); + bh_lock_sock(sk); + if (!sock_owned_by_user(sk)) { + queued = x25_process_rx_frame(sk, skb); + } else { + queued = !sk_add_backlog(sk, skb, READ_ONCE(sk->sk_rcvbuf)); + } + bh_unlock_sock(sk); + sock_put(sk); + return queued; + } + + /* + * Is is a Call Request ? if so process it. + */ + if (frametype == X25_CALL_REQUEST) + return x25_rx_call_request(skb, nb, lci); + + /* + * Its not a Call Request, nor is it a control frame. + * Can we forward it? + */ + + if (x25_forward_data(lci, nb, skb)) { + if (frametype == X25_CLEAR_CONFIRMATION) { + x25_clear_forward_by_lci(lci); + } + kfree_skb(skb); + return 1; + } + +/* + x25_transmit_clear_request(nb, lci, 0x0D); +*/ + + if (frametype != X25_CLEAR_CONFIRMATION) + pr_debug("x25_receive_data(): unknown frame type %2x\n",frametype); + + return 0; +} + +int x25_lapb_receive_frame(struct sk_buff *skb, struct net_device *dev, + struct packet_type *ptype, struct net_device *orig_dev) +{ + struct sk_buff *nskb; + struct x25_neigh *nb; + + if (!net_eq(dev_net(dev), &init_net)) + goto drop; + + nskb = skb_copy(skb, GFP_ATOMIC); + if (!nskb) + goto drop; + kfree_skb(skb); + skb = nskb; + + /* + * Packet received from unrecognised device, throw it away. + */ + nb = x25_get_neigh(dev); + if (!nb) { + pr_debug("unknown neighbour - %s\n", dev->name); + goto drop; + } + + if (!pskb_may_pull(skb, 1)) { + x25_neigh_put(nb); + goto drop; + } + + switch (skb->data[0]) { + + case X25_IFACE_DATA: + skb_pull(skb, 1); + if (x25_receive_data(skb, nb)) { + x25_neigh_put(nb); + goto out; + } + break; + + case X25_IFACE_CONNECT: + x25_link_established(nb); + break; + + case X25_IFACE_DISCONNECT: + x25_link_terminated(nb); + break; + } + x25_neigh_put(nb); +drop: + kfree_skb(skb); +out: + return 0; +} + +void x25_establish_link(struct x25_neigh *nb) +{ + struct sk_buff *skb; + unsigned char *ptr; + + switch (nb->dev->type) { + case ARPHRD_X25: + if ((skb = alloc_skb(1, GFP_ATOMIC)) == NULL) { + pr_err("x25_dev: out of memory\n"); + return; + } + ptr = skb_put(skb, 1); + *ptr = X25_IFACE_CONNECT; + break; + + default: + return; + } + + skb->protocol = htons(ETH_P_X25); + skb->dev = nb->dev; + + dev_queue_xmit(skb); +} + +void x25_send_frame(struct sk_buff *skb, struct x25_neigh *nb) +{ + unsigned char *dptr; + + skb_reset_network_header(skb); + + switch (nb->dev->type) { + case ARPHRD_X25: + dptr = skb_push(skb, 1); + *dptr = X25_IFACE_DATA; + break; + + default: + kfree_skb(skb); + return; + } + + skb->protocol = htons(ETH_P_X25); + skb->dev = nb->dev; + + dev_queue_xmit(skb); +} -- cgit v1.3.1