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/clocksource/timer-milbeaut.c | 189 +++++++++++++++++++++++++++++++++++ 1 file changed, 189 insertions(+) create mode 100644 drivers/clocksource/timer-milbeaut.c (limited to 'drivers/clocksource/timer-milbeaut.c') diff --git a/drivers/clocksource/timer-milbeaut.c b/drivers/clocksource/timer-milbeaut.c new file mode 100644 index 000000000..fa9fb4eac --- /dev/null +++ b/drivers/clocksource/timer-milbeaut.c @@ -0,0 +1,189 @@ +// SPDX-License-Identifier: GPL-2.0 +/* + * Copyright (C) 2018 Socionext Inc. + */ + +#include +#include +#include +#include +#include +#include "timer-of.h" + +#define MLB_TMR_TMCSR_OFS 0x0 +#define MLB_TMR_TMR_OFS 0x4 +#define MLB_TMR_TMRLR1_OFS 0x8 +#define MLB_TMR_TMRLR2_OFS 0xc +#define MLB_TMR_REGSZPCH 0x10 + +#define MLB_TMR_TMCSR_OUTL BIT(5) +#define MLB_TMR_TMCSR_RELD BIT(4) +#define MLB_TMR_TMCSR_INTE BIT(3) +#define MLB_TMR_TMCSR_UF BIT(2) +#define MLB_TMR_TMCSR_CNTE BIT(1) +#define MLB_TMR_TMCSR_TRG BIT(0) + +#define MLB_TMR_TMCSR_CSL_DIV2 0 +#define MLB_TMR_DIV_CNT 2 + +#define MLB_TMR_SRC_CH 1 +#define MLB_TMR_EVT_CH 0 + +#define MLB_TMR_SRC_CH_OFS (MLB_TMR_REGSZPCH * MLB_TMR_SRC_CH) +#define MLB_TMR_EVT_CH_OFS (MLB_TMR_REGSZPCH * MLB_TMR_EVT_CH) + +#define MLB_TMR_SRC_TMCSR_OFS (MLB_TMR_SRC_CH_OFS + MLB_TMR_TMCSR_OFS) +#define MLB_TMR_SRC_TMR_OFS (MLB_TMR_SRC_CH_OFS + MLB_TMR_TMR_OFS) +#define MLB_TMR_SRC_TMRLR1_OFS (MLB_TMR_SRC_CH_OFS + MLB_TMR_TMRLR1_OFS) +#define MLB_TMR_SRC_TMRLR2_OFS (MLB_TMR_SRC_CH_OFS + MLB_TMR_TMRLR2_OFS) + +#define MLB_TMR_EVT_TMCSR_OFS (MLB_TMR_EVT_CH_OFS + MLB_TMR_TMCSR_OFS) +#define MLB_TMR_EVT_TMR_OFS (MLB_TMR_EVT_CH_OFS + MLB_TMR_TMR_OFS) +#define MLB_TMR_EVT_TMRLR1_OFS (MLB_TMR_EVT_CH_OFS + MLB_TMR_TMRLR1_OFS) +#define MLB_TMR_EVT_TMRLR2_OFS (MLB_TMR_EVT_CH_OFS + MLB_TMR_TMRLR2_OFS) + +#define MLB_TIMER_RATING 500 +#define MLB_TIMER_ONESHOT 0 +#define MLB_TIMER_PERIODIC 1 + +static irqreturn_t mlb_timer_interrupt(int irq, void *dev_id) +{ + struct clock_event_device *clk = dev_id; + struct timer_of *to = to_timer_of(clk); + u32 val; + + val = readl_relaxed(timer_of_base(to) + MLB_TMR_EVT_TMCSR_OFS); + val &= ~MLB_TMR_TMCSR_UF; + writel_relaxed(val, timer_of_base(to) + MLB_TMR_EVT_TMCSR_OFS); + + clk->event_handler(clk); + + return IRQ_HANDLED; +} + +static void mlb_evt_timer_start(struct timer_of *to, bool periodic) +{ + u32 val = MLB_TMR_TMCSR_CSL_DIV2; + + val |= MLB_TMR_TMCSR_CNTE | MLB_TMR_TMCSR_TRG | MLB_TMR_TMCSR_INTE; + if (periodic) + val |= MLB_TMR_TMCSR_RELD; + writel_relaxed(val, timer_of_base(to) + MLB_TMR_EVT_TMCSR_OFS); +} + +static void mlb_evt_timer_stop(struct timer_of *to) +{ + u32 val = readl_relaxed(timer_of_base(to) + MLB_TMR_EVT_TMCSR_OFS); + + val &= ~MLB_TMR_TMCSR_CNTE; + writel_relaxed(val, timer_of_base(to) + MLB_TMR_EVT_TMCSR_OFS); +} + +static void mlb_evt_timer_register_count(struct timer_of *to, unsigned long cnt) +{ + writel_relaxed(cnt, timer_of_base(to) + MLB_TMR_EVT_TMRLR1_OFS); +} + +static int mlb_set_state_periodic(struct clock_event_device *clk) +{ + struct timer_of *to = to_timer_of(clk); + + mlb_evt_timer_stop(to); + mlb_evt_timer_register_count(to, to->of_clk.period); + mlb_evt_timer_start(to, MLB_TIMER_PERIODIC); + return 0; +} + +static int mlb_set_state_oneshot(struct clock_event_device *clk) +{ + struct timer_of *to = to_timer_of(clk); + + mlb_evt_timer_stop(to); + mlb_evt_timer_start(to, MLB_TIMER_ONESHOT); + return 0; +} + +static int mlb_set_state_shutdown(struct clock_event_device *clk) +{ + struct timer_of *to = to_timer_of(clk); + + mlb_evt_timer_stop(to); + return 0; +} + +static int mlb_clkevt_next_event(unsigned long event, + struct clock_event_device *clk) +{ + struct timer_of *to = to_timer_of(clk); + + mlb_evt_timer_stop(to); + mlb_evt_timer_register_count(to, event); + mlb_evt_timer_start(to, MLB_TIMER_ONESHOT); + return 0; +} + +static int mlb_config_clock_source(struct timer_of *to) +{ + u32 val = MLB_TMR_TMCSR_CSL_DIV2; + + writel_relaxed(val, timer_of_base(to) + MLB_TMR_SRC_TMCSR_OFS); + writel_relaxed(~0, timer_of_base(to) + MLB_TMR_SRC_TMRLR1_OFS); + writel_relaxed(~0, timer_of_base(to) + MLB_TMR_SRC_TMRLR2_OFS); + val |= MLB_TMR_TMCSR_RELD | MLB_TMR_TMCSR_CNTE | MLB_TMR_TMCSR_TRG; + writel_relaxed(val, timer_of_base(to) + MLB_TMR_SRC_TMCSR_OFS); + return 0; +} + +static int mlb_config_clock_event(struct timer_of *to) +{ + writel_relaxed(0, timer_of_base(to) + MLB_TMR_EVT_TMCSR_OFS); + return 0; +} + +static struct timer_of to = { + .flags = TIMER_OF_IRQ | TIMER_OF_BASE | TIMER_OF_CLOCK, + + .clkevt = { + .name = "mlb-clkevt", + .rating = MLB_TIMER_RATING, + .cpumask = cpu_possible_mask, + .features = CLOCK_EVT_FEAT_DYNIRQ | CLOCK_EVT_FEAT_ONESHOT, + .set_state_oneshot = mlb_set_state_oneshot, + .set_state_periodic = mlb_set_state_periodic, + .set_state_shutdown = mlb_set_state_shutdown, + .set_next_event = mlb_clkevt_next_event, + }, + + .of_irq = { + .flags = IRQF_TIMER | IRQF_IRQPOLL, + .handler = mlb_timer_interrupt, + }, +}; + +static u64 notrace mlb_timer_sched_read(void) +{ + return ~readl_relaxed(timer_of_base(&to) + MLB_TMR_SRC_TMR_OFS); +} + +static int __init mlb_timer_init(struct device_node *node) +{ + int ret; + unsigned long rate; + + ret = timer_of_init(node, &to); + if (ret) + return ret; + + rate = timer_of_rate(&to) / MLB_TMR_DIV_CNT; + mlb_config_clock_source(&to); + clocksource_mmio_init(timer_of_base(&to) + MLB_TMR_SRC_TMR_OFS, + node->name, rate, MLB_TIMER_RATING, 32, + clocksource_mmio_readl_down); + sched_clock_register(mlb_timer_sched_read, 32, rate); + mlb_config_clock_event(&to); + clockevents_config_and_register(&to.clkevt, timer_of_rate(&to), 15, + 0xffffffff); + return 0; +} +TIMER_OF_DECLARE(mlb_peritimer, "socionext,milbeaut-timer", + mlb_timer_init); -- cgit v1.3.1