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/devfreq/imx-bus.c | 166 ++++++++++++++++++++++++++++++++++++++++++++++ 1 file changed, 166 insertions(+) create mode 100644 drivers/devfreq/imx-bus.c (limited to 'drivers/devfreq/imx-bus.c') diff --git a/drivers/devfreq/imx-bus.c b/drivers/devfreq/imx-bus.c new file mode 100644 index 000000000..49798f542 --- /dev/null +++ b/drivers/devfreq/imx-bus.c @@ -0,0 +1,166 @@ +// SPDX-License-Identifier: GPL-2.0 +/* + * Copyright 2019 NXP + */ + +#include +#include +#include +#include +#include +#include +#include +#include + +struct imx_bus { + struct devfreq_dev_profile profile; + struct devfreq *devfreq; + struct clk *clk; + struct platform_device *icc_pdev; +}; + +static int imx_bus_target(struct device *dev, + unsigned long *freq, u32 flags) +{ + struct dev_pm_opp *new_opp; + int ret; + + new_opp = devfreq_recommended_opp(dev, freq, flags); + if (IS_ERR(new_opp)) { + ret = PTR_ERR(new_opp); + dev_err(dev, "failed to get recommended opp: %d\n", ret); + return ret; + } + dev_pm_opp_put(new_opp); + + return dev_pm_opp_set_rate(dev, *freq); +} + +static int imx_bus_get_cur_freq(struct device *dev, unsigned long *freq) +{ + struct imx_bus *priv = dev_get_drvdata(dev); + + *freq = clk_get_rate(priv->clk); + + return 0; +} + +static void imx_bus_exit(struct device *dev) +{ + struct imx_bus *priv = dev_get_drvdata(dev); + + dev_pm_opp_of_remove_table(dev); + platform_device_unregister(priv->icc_pdev); +} + +/* imx_bus_init_icc() - register matching icc provider if required */ +static int imx_bus_init_icc(struct device *dev) +{ + struct imx_bus *priv = dev_get_drvdata(dev); + const char *icc_driver_name; + + if (!of_property_present(dev->of_node, "#interconnect-cells")) + return 0; + if (!IS_ENABLED(CONFIG_INTERCONNECT_IMX)) { + dev_warn(dev, "imx interconnect drivers disabled\n"); + return 0; + } + + icc_driver_name = of_device_get_match_data(dev); + if (!icc_driver_name) { + dev_err(dev, "unknown interconnect driver\n"); + return 0; + } + + priv->icc_pdev = platform_device_register_data( + dev, icc_driver_name, -1, NULL, 0); + if (IS_ERR(priv->icc_pdev)) { + dev_err(dev, "failed to register icc provider %s: %ld\n", + icc_driver_name, PTR_ERR(priv->icc_pdev)); + return PTR_ERR(priv->icc_pdev); + } + + return 0; +} + +static int imx_bus_probe(struct platform_device *pdev) +{ + struct device *dev = &pdev->dev; + struct imx_bus *priv; + const char *gov = DEVFREQ_GOV_USERSPACE; + int ret; + + priv = devm_kzalloc(dev, sizeof(*priv), GFP_KERNEL); + if (!priv) + return -ENOMEM; + + /* + * Fetch the clock to adjust but don't explicitly enable. + * + * For imx bus clock clk_set_rate is safe no matter if the clock is on + * or off and some peripheral side-buses might be off unless enabled by + * drivers for devices on those specific buses. + * + * Rate adjustment on a disabled bus clock just takes effect later. + */ + priv->clk = devm_clk_get(dev, NULL); + if (IS_ERR(priv->clk)) { + ret = PTR_ERR(priv->clk); + dev_err(dev, "failed to fetch clk: %d\n", ret); + return ret; + } + platform_set_drvdata(pdev, priv); + + ret = dev_pm_opp_of_add_table(dev); + if (ret < 0) { + dev_err(dev, "failed to get OPP table\n"); + return ret; + } + + priv->profile.target = imx_bus_target; + priv->profile.exit = imx_bus_exit; + priv->profile.get_cur_freq = imx_bus_get_cur_freq; + priv->profile.initial_freq = clk_get_rate(priv->clk); + + priv->devfreq = devm_devfreq_add_device(dev, &priv->profile, + gov, NULL); + if (IS_ERR(priv->devfreq)) { + ret = PTR_ERR(priv->devfreq); + dev_err(dev, "failed to add devfreq device: %d\n", ret); + goto err; + } + + ret = imx_bus_init_icc(dev); + if (ret) + goto err; + + return 0; + +err: + dev_pm_opp_of_remove_table(dev); + return ret; +} + +static const struct of_device_id imx_bus_of_match[] = { + { .compatible = "fsl,imx8mq-noc", .data = "imx8mq-interconnect", }, + { .compatible = "fsl,imx8mm-noc", .data = "imx8mm-interconnect", }, + { .compatible = "fsl,imx8mn-noc", .data = "imx8mn-interconnect", }, + { .compatible = "fsl,imx8mp-noc", .data = "imx8mp-interconnect", }, + { .compatible = "fsl,imx8m-noc", }, + { .compatible = "fsl,imx8m-nic", }, + { /* sentinel */ }, +}; +MODULE_DEVICE_TABLE(of, imx_bus_of_match); + +static struct platform_driver imx_bus_platdrv = { + .probe = imx_bus_probe, + .driver = { + .name = "imx-bus-devfreq", + .of_match_table = imx_bus_of_match, + }, +}; +module_platform_driver(imx_bus_platdrv); + +MODULE_DESCRIPTION("Generic i.MX bus frequency scaling driver"); +MODULE_AUTHOR("Leonard Crestez "); +MODULE_LICENSE("GPL v2"); -- cgit v1.3.1