x86/kprobes: Fix crash when probing CS CALL instructionsgrafted

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) <mhiramat@kernel.org>
Signed-off-by: Jinke Han <jinkehan@didiglobal.com>
Signed-off-by: Ingo Molnar <mingo@kernel.org>
Reviewed-by: Masami Hiramatsu (Google) <mhiramat@kernel.org>
Acked-by: Yafang Shao <laoar.shao@gmail.com>
Acked-by: Borislav Petkov <bp@alien8.de>
Cc: Peter Zijlstra <peterz@infradead.org>
Link: https://patch.msgid.link/20260908073742.GA10517@didi-ThinkCentre-M920t-N000

1 files changed, 248 insertions, 0 deletions

diff --git a/net/mac80211/ocb.c b/net/mac80211/ocb.c
new file mode 100644
index 000000000..e9bf4ba3e
--- /dev/null
+++ b/net/mac80211/ocb.c
@@ -0,0 +1,248 @@
+// SPDX-License-Identifier: GPL-2.0-only
+/*
+ * OCB mode implementation
+ *
+ * Copyright: (c) 2014 Czech Technical University in Prague
+ *            (c) 2014 Volkswagen Group Research
+ * Copyright (C) 2022 - 2024, 2026 Intel Corporation
+ * Author:    Rostislav Lisovy <rostislav.lisovy@fel.cvut.cz>
+ * Funded by: Volkswagen Group Research
+ */
+
+#include <linux/delay.h>
+#include <linux/if_ether.h>
+#include <linux/skbuff.h>
+#include <linux/if_arp.h>
+#include <linux/etherdevice.h>
+#include <linux/rtnetlink.h>
+#include <net/mac80211.h>
+#include <linux/unaligned.h>
+
+#include "ieee80211_i.h"
+#include "driver-ops.h"
+#include "rate.h"
+
+#define IEEE80211_OCB_HOUSEKEEPING_INTERVAL		(60 * HZ)
+#define IEEE80211_OCB_PEER_INACTIVITY_LIMIT		(240 * HZ)
+#define IEEE80211_OCB_MAX_STA_ENTRIES			128
+
+/**
+ * enum ocb_deferred_task_flags - mac80211 OCB deferred tasks
+ * @OCB_WORK_HOUSEKEEPING: run the periodic OCB housekeeping tasks
+ *
+ * These flags are used in @wrkq_flags field of &struct ieee80211_if_ocb
+ */
+enum ocb_deferred_task_flags {
+	OCB_WORK_HOUSEKEEPING,
+};
+
+void ieee80211_ocb_rx_no_sta(struct ieee80211_sub_if_data *sdata,
+			     const u8 *bssid, const u8 *addr,
+			     u32 supp_rates)
+{
+	struct ieee80211_if_ocb *ifocb = &sdata->u.ocb;
+	struct ieee80211_local *local = sdata->local;
+	struct ieee80211_chanctx_conf *chanctx_conf;
+	struct ieee80211_supported_band *sband;
+	struct sta_info *sta;
+	int band;
+
+	if (!ifocb->joined)
+		return;
+
+	/* XXX: Consider removing the least recently used entry and
+	 *      allow new one to be added.
+	 */
+	if (local->num_sta >= IEEE80211_OCB_MAX_STA_ENTRIES) {
+		net_info_ratelimited("%s: No room for a new OCB STA entry %pM\n",
+				     sdata->name, addr);
+		return;
+	}
+
+	ocb_dbg(sdata, "Adding new OCB station %pM\n", addr);
+
+	rcu_read_lock();
+	chanctx_conf = rcu_dereference(sdata->vif.bss_conf.chanctx_conf);
+	if (WARN_ON_ONCE(!chanctx_conf)) {
+		rcu_read_unlock();
+		return;
+	}
+	band = chanctx_conf->def.chan->band;
+	rcu_read_unlock();
+
+	sta = sta_info_alloc(sdata, addr, GFP_ATOMIC);
+	if (!sta)
+		return;
+
+	/* Add only mandatory rates for now */
+	sband = local->hw.wiphy->bands[band];
+	sta->sta.deflink.supp_rates[band] = ieee80211_mandatory_rates(sband);
+
+	spin_lock(&ifocb->incomplete_lock);
+	list_add(&sta->list, &ifocb->incomplete_stations);
+	spin_unlock(&ifocb->incomplete_lock);
+	wiphy_work_queue(local->hw.wiphy, &sdata->work);
+}
+
+static struct sta_info *ieee80211_ocb_finish_sta(struct sta_info *sta)
+	__acquires(RCU)
+{
+	struct ieee80211_sub_if_data *sdata = sta->sdata;
+	u8 addr[ETH_ALEN];
+
+	memcpy(addr, sta->sta.addr, ETH_ALEN);
+
+	ieee80211_sta_init_nss_bw_capa(&sta->deflink,
+				       &sdata->deflink.conf->chanreq.oper);
+
+	ocb_dbg(sdata, "Adding new IBSS station %pM (dev=%s)\n",
+		addr, sdata->name);
+
+	sta_info_move_state(sta, IEEE80211_STA_AUTH);
+	sta_info_move_state(sta, IEEE80211_STA_ASSOC);
+	sta_info_move_state(sta, IEEE80211_STA_AUTHORIZED);
+
+	rate_control_rate_init(&sta->deflink);
+
+	/* If it fails, maybe we raced another insertion? */
+	if (sta_info_insert_rcu(sta))
+		return sta_info_get(sdata, addr);
+	return sta;
+}
+
+static void ieee80211_ocb_housekeeping(struct ieee80211_sub_if_data *sdata)
+{
+	struct ieee80211_if_ocb *ifocb = &sdata->u.ocb;
+
+	ocb_dbg(sdata, "Running ocb housekeeping\n");
+
+	ieee80211_sta_expire(sdata, IEEE80211_OCB_PEER_INACTIVITY_LIMIT);
+
+	mod_timer(&ifocb->housekeeping_timer,
+		  round_jiffies(jiffies + IEEE80211_OCB_HOUSEKEEPING_INTERVAL));
+}
+
+void ieee80211_ocb_work(struct ieee80211_sub_if_data *sdata)
+{
+	struct ieee80211_if_ocb *ifocb = &sdata->u.ocb;
+	struct sta_info *sta;
+
+	lockdep_assert_wiphy(sdata->local->hw.wiphy);
+
+	if (ifocb->joined != true)
+		return;
+
+	spin_lock_bh(&ifocb->incomplete_lock);
+	while (!list_empty(&ifocb->incomplete_stations)) {
+		sta = list_first_entry(&ifocb->incomplete_stations,
+				       struct sta_info, list);
+		list_del(&sta->list);
+		spin_unlock_bh(&ifocb->incomplete_lock);
+
+		ieee80211_ocb_finish_sta(sta);
+		rcu_read_unlock();
+		spin_lock_bh(&ifocb->incomplete_lock);
+	}
+	spin_unlock_bh(&ifocb->incomplete_lock);
+
+	if (test_and_clear_bit(OCB_WORK_HOUSEKEEPING, &ifocb->wrkq_flags))
+		ieee80211_ocb_housekeeping(sdata);
+}
+
+static void ieee80211_ocb_housekeeping_timer(struct timer_list *t)
+{
+	struct ieee80211_sub_if_data *sdata =
+		timer_container_of(sdata, t, u.ocb.housekeeping_timer);
+	struct ieee80211_local *local = sdata->local;
+	struct ieee80211_if_ocb *ifocb = &sdata->u.ocb;
+
+	set_bit(OCB_WORK_HOUSEKEEPING, &ifocb->wrkq_flags);
+
+	wiphy_work_queue(local->hw.wiphy, &sdata->work);
+}
+
+void ieee80211_ocb_setup_sdata(struct ieee80211_sub_if_data *sdata)
+{
+	struct ieee80211_if_ocb *ifocb = &sdata->u.ocb;
+
+	timer_setup(&ifocb->housekeeping_timer,
+		    ieee80211_ocb_housekeeping_timer, 0);
+	INIT_LIST_HEAD(&ifocb->incomplete_stations);
+	spin_lock_init(&ifocb->incomplete_lock);
+}
+
+int ieee80211_ocb_join(struct ieee80211_sub_if_data *sdata,
+		       struct ocb_setup *setup)
+{
+	struct ieee80211_chan_req chanreq = { .oper = setup->chandef };
+	struct ieee80211_local *local = sdata->local;
+	struct ieee80211_if_ocb *ifocb = &sdata->u.ocb;
+	u64 changed = BSS_CHANGED_OCB | BSS_CHANGED_BSSID;
+	int err;
+
+	lockdep_assert_wiphy(sdata->local->hw.wiphy);
+
+	if (ifocb->joined == true)
+		return -EINVAL;
+
+	sdata->deflink.operating_11g_mode = true;
+	sdata->deflink.smps_mode = IEEE80211_SMPS_OFF;
+	sdata->deflink.needed_rx_chains = sdata->local->rx_chains;
+
+	err = ieee80211_link_use_channel(&sdata->deflink, &chanreq,
+					 IEEE80211_CHANCTX_SHARED);
+	if (err)
+		return err;
+
+	ieee80211_bss_info_change_notify(sdata, changed);
+
+	ifocb->joined = true;
+
+	set_bit(OCB_WORK_HOUSEKEEPING, &ifocb->wrkq_flags);
+	wiphy_work_queue(local->hw.wiphy, &sdata->work);
+
+	netif_carrier_on(sdata->dev);
+	return 0;
+}
+
+int ieee80211_ocb_leave(struct ieee80211_sub_if_data *sdata)
+{
+	struct ieee80211_if_ocb *ifocb = &sdata->u.ocb;
+	struct ieee80211_local *local = sdata->local;
+	struct sta_info *sta;
+
+	lockdep_assert_wiphy(sdata->local->hw.wiphy);
+
+	ifocb->joined = false;
+	sta_info_flush(sdata, -1);
+
+	spin_lock_bh(&ifocb->incomplete_lock);
+	while (!list_empty(&ifocb->incomplete_stations)) {
+		sta = list_first_entry(&ifocb->incomplete_stations,
+				       struct sta_info, list);
+		list_del(&sta->list);
+		spin_unlock_bh(&ifocb->incomplete_lock);
+
+		sta_info_free(local, sta);
+		spin_lock_bh(&ifocb->incomplete_lock);
+	}
+	spin_unlock_bh(&ifocb->incomplete_lock);
+
+	netif_carrier_off(sdata->dev);
+	clear_bit(SDATA_STATE_OFFCHANNEL, &sdata->state);
+	ieee80211_bss_info_change_notify(sdata, BSS_CHANGED_OCB);
+
+	ieee80211_link_release_channel(&sdata->deflink);
+
+	skb_queue_purge(&sdata->skb_queue);
+
+	timer_delete_sync(&sdata->u.ocb.housekeeping_timer);
+	/* If the timer fired while we waited for it, it will have
+	 * requeued the work. Now the work will be running again
+	 * but will not rearm the timer again because it checks
+	 * whether we are connected to the network or not -- at this
+	 * point we shouldn't be anymore.
+	 */
+
+	return 0;
+}