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/android/binder/context.rs | 175 ++++++++++++++++++++++++++++++++++++++ 1 file changed, 175 insertions(+) create mode 100644 drivers/android/binder/context.rs (limited to 'drivers/android/binder/context.rs') diff --git a/drivers/android/binder/context.rs b/drivers/android/binder/context.rs new file mode 100644 index 000000000..ddddb66b3 --- /dev/null +++ b/drivers/android/binder/context.rs @@ -0,0 +1,175 @@ +// SPDX-License-Identifier: GPL-2.0 + +// Copyright (C) 2025 Google LLC. + +use kernel::{ + alloc::kvec::KVVec, + error::code::*, + prelude::*, + security, + str::{CStr, CString}, + sync::{Arc, Mutex}, + task::Kuid, +}; + +use crate::{error::BinderError, node::NodeRef, process::Process}; + +kernel::sync::global_lock! { + // SAFETY: We call `init` in the module initializer, so it's initialized before first use. + pub(crate) unsafe(uninit) static CONTEXTS: Mutex = ContextList { + contexts: KVVec::new(), + }; +} + +pub(crate) struct ContextList { + contexts: KVVec>, +} + +pub(crate) fn get_all_contexts() -> Result>> { + let lock = CONTEXTS.lock(); + let mut ctxs = KVVec::with_capacity(lock.contexts.len(), GFP_KERNEL)?; + for ctx in lock.contexts.iter() { + ctxs.push(ctx.clone(), GFP_KERNEL)?; + } + Ok(ctxs) +} + +/// This struct keeps track of the processes using this context, and which process is the context +/// manager. +struct Manager { + node: Option, + uid: Option, + all_procs: KVVec>, +} + +/// There is one context per binder file (/dev/binder, /dev/hwbinder, etc) +#[pin_data] +pub(crate) struct Context { + #[pin] + manager: Mutex, + pub(crate) name: CString, +} + +impl Context { + pub(crate) fn new(name: &CStr) -> Result> { + let name = CString::try_from(name)?; + let ctx = Arc::pin_init( + try_pin_init!(Context { + name, + manager <- kernel::new_mutex!(Manager { + all_procs: KVVec::new(), + node: None, + uid: None, + }, "Context::manager"), + }), + GFP_KERNEL, + )?; + + CONTEXTS.lock().contexts.push(ctx.clone(), GFP_KERNEL)?; + + Ok(ctx) + } + + /// Called when the file for this context is unlinked. + /// + /// No-op if called twice. + pub(crate) fn deregister(self: &Arc) { + // Safe removal using retain + CONTEXTS.lock().contexts.retain(|c| !Arc::ptr_eq(c, self)); + } + + pub(crate) fn register_process(self: &Arc, proc: Arc) -> Result { + if !Arc::ptr_eq(self, &proc.ctx) { + pr_err!("Context::register_process called on the wrong context."); + return Err(EINVAL); + } + self.manager.lock().all_procs.push(proc, GFP_KERNEL)?; + Ok(()) + } + + pub(crate) fn deregister_process(self: &Arc, proc: &Arc) { + if !Arc::ptr_eq(self, &proc.ctx) { + pr_err!("Context::deregister_process called on the wrong context."); + return; + } + let mut manager = self.manager.lock(); + manager.all_procs.retain(|p| !Arc::ptr_eq(p, proc)); + + // Shrink the vector if it has significant unused capacity to avoid memory waste, + // but use a conservative strategy to prevent shrink-then-regrow oscillation. + // Only shrink when length drops below 1/4 of capacity, and shrink to twice the length. + let len = manager.all_procs.len(); + let cap = manager.all_procs.capacity(); + if len < cap / 4 { + // Shrink to twice the current length. Ignore allocation failures since this + // is just an optimization; the vector remains valid even if shrinking fails. + let _ = manager.all_procs.shrink_to(len * 2, GFP_KERNEL); + } + } + + pub(crate) fn set_manager_node(&self, node_ref: NodeRef) -> Result { + let mut manager = self.manager.lock(); + if manager.node.is_some() { + pr_warn!("BINDER_SET_CONTEXT_MGR already set"); + return Err(EBUSY); + } + security::binder_set_context_mgr(&node_ref.node.owner.cred)?; + + // If the context manager has been set before, ensure that we use the same euid. + let caller_uid = Kuid::current_euid(); + if let Some(ref uid) = manager.uid { + if *uid != caller_uid { + return Err(EPERM); + } + } + + manager.node = Some(node_ref); + manager.uid = Some(caller_uid); + Ok(()) + } + + pub(crate) fn unset_manager_node(&self) { + let node_ref = self.manager.lock().node.take(); + drop(node_ref); + } + + pub(crate) fn get_manager_node(&self, strong: bool) -> Result { + self.manager + .lock() + .node + .as_ref() + .ok_or_else(BinderError::new_dead)? + .clone(strong) + .map_err(BinderError::from) + } + + pub(crate) fn for_each_proc(&self, mut func: F) + where + F: FnMut(&Process), + { + let lock = self.manager.lock(); + for proc in &lock.all_procs { + func(proc); + } + } + + pub(crate) fn get_all_procs(&self) -> Result>> { + let lock = self.manager.lock(); + let mut procs = KVVec::with_capacity(lock.all_procs.len(), GFP_KERNEL)?; + for proc in lock.all_procs.iter() { + procs.push(Arc::clone(proc), GFP_KERNEL)?; + } + Ok(procs) + } + + pub(crate) fn get_procs_with_pid(&self, pid: i32) -> Result>> { + let lock = self.manager.lock(); + let mut matching_procs = KVVec::new(); + for proc in lock.all_procs.iter() { + if proc.task.pid() == pid { + matching_procs.push(Arc::clone(proc), GFP_KERNEL)?; + } + } + Ok(matching_procs) + } +} -- cgit v1.3.1