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/rxrpc/txbuf.c | 96 +++++++++++++++++++++++++++++++++++++++++++++++++++++++ 1 file changed, 96 insertions(+) create mode 100644 net/rxrpc/txbuf.c (limited to 'net/rxrpc/txbuf.c') diff --git a/net/rxrpc/txbuf.c b/net/rxrpc/txbuf.c new file mode 100644 index 000000000..55ef7a048 --- /dev/null +++ b/net/rxrpc/txbuf.c @@ -0,0 +1,96 @@ +// SPDX-License-Identifier: GPL-2.0-or-later +/* RxRPC Tx data buffering. + * + * Copyright (C) 2022 Red Hat, Inc. All Rights Reserved. + * Written by David Howells (dhowells@redhat.com) + */ + +#define pr_fmt(fmt) KBUILD_MODNAME ": " fmt + +#include +#include "ar-internal.h" + +static atomic_t rxrpc_txbuf_debug_ids; +atomic_t rxrpc_nr_txbuf; + +/* + * Allocate and partially initialise a data transmission buffer. + */ +struct rxrpc_txbuf *rxrpc_alloc_data_txbuf(struct rxrpc_call *call, size_t data_size, + size_t data_align, gfp_t gfp) +{ + struct rxrpc_txbuf *txb; + size_t total, doff, jsize = sizeof(struct rxrpc_jumbo_header); + void *buf; + + txb = kzalloc_obj(*txb, gfp); + if (!txb) + return NULL; + + /* We put a jumbo header in the buffer, but not a full wire header to + * avoid delayed-corruption problems with zerocopy. + */ + doff = round_up(jsize, data_align); + total = doff + data_size; + + data_align = umax(data_align, L1_CACHE_BYTES); + mutex_lock(&call->conn->tx_data_alloc_lock); + buf = page_frag_alloc_align(&call->conn->tx_data_alloc, total, gfp, + data_align); + mutex_unlock(&call->conn->tx_data_alloc_lock); + if (!buf) { + kfree(txb); + return NULL; + } + + refcount_set(&txb->ref, 1); + txb->call_debug_id = call->debug_id; + txb->debug_id = atomic_inc_return(&rxrpc_txbuf_debug_ids); + txb->alloc_size = data_size; + txb->space = data_size; + txb->offset = 0; + txb->flags = call->conn->out_clientflag; + txb->seq = call->send_top + 1; + txb->data = buf + doff; + + trace_rxrpc_txbuf(txb->debug_id, txb->call_debug_id, txb->seq, 1, + rxrpc_txbuf_alloc_data); + + atomic_inc(&rxrpc_nr_txbuf); + return txb; +} + +void rxrpc_see_txbuf(struct rxrpc_txbuf *txb, enum rxrpc_txbuf_trace what) +{ + int r = refcount_read(&txb->ref); + + trace_rxrpc_txbuf(txb->debug_id, txb->call_debug_id, txb->seq, r, what); +} + +static void rxrpc_free_txbuf(struct rxrpc_txbuf *txb) +{ + trace_rxrpc_txbuf(txb->debug_id, txb->call_debug_id, txb->seq, 0, + rxrpc_txbuf_free); + if (txb->data) + page_frag_free(txb->data); + kfree(txb); + atomic_dec(&rxrpc_nr_txbuf); +} + +void rxrpc_put_txbuf(struct rxrpc_txbuf *txb, enum rxrpc_txbuf_trace what) +{ + unsigned int debug_id, call_debug_id; + rxrpc_seq_t seq; + bool dead; + int r; + + if (txb) { + debug_id = txb->debug_id; + call_debug_id = txb->call_debug_id; + seq = txb->seq; + dead = __refcount_dec_and_test(&txb->ref, &r); + trace_rxrpc_txbuf(debug_id, call_debug_id, seq, r - 1, what); + if (dead) + rxrpc_free_txbuf(txb); + } +} -- cgit v1.3.1