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/xen/efi.c | 355 ++++++++++++++++++++++++++++++++++++++++++++++++++++++ 1 file changed, 355 insertions(+) create mode 100644 drivers/xen/efi.c (limited to 'drivers/xen/efi.c') diff --git a/drivers/xen/efi.c b/drivers/xen/efi.c new file mode 100644 index 000000000..fb321cd64 --- /dev/null +++ b/drivers/xen/efi.c @@ -0,0 +1,355 @@ +// SPDX-License-Identifier: GPL-2.0-only +/* + * EFI support for Xen. + * + * Copyright (C) 1999 VA Linux Systems + * Copyright (C) 1999 Walt Drummond + * Copyright (C) 1999-2002 Hewlett-Packard Co. + * David Mosberger-Tang + * Stephane Eranian + * Copyright (C) 2005-2008 Intel Co. + * Fenghua Yu + * Bibo Mao + * Chandramouli Narayanan + * Huang Ying + * Copyright (C) 2011 Novell Co. + * Jan Beulich + * Copyright (C) 2011-2012 Oracle Co. + * Liang Tang + * Copyright (c) 2014 Oracle Co., Daniel Kiper + */ + +#include +#include +#include +#include + +#include +#include +#include +#include +#include + +#include + +#include + +#define INIT_EFI_OP(name) \ + {.cmd = XENPF_efi_runtime_call, \ + .u.efi_runtime_call.function = XEN_EFI_##name, \ + .u.efi_runtime_call.misc = 0} + +#define efi_data(op) (op.u.efi_runtime_call) + +static efi_status_t xen_efi_get_time(efi_time_t *tm, efi_time_cap_t *tc) +{ + struct xen_platform_op op = INIT_EFI_OP(get_time); + + if (HYPERVISOR_platform_op(&op) < 0) + return EFI_UNSUPPORTED; + + if (tm) { + BUILD_BUG_ON(sizeof(*tm) != sizeof(efi_data(op).u.get_time.time)); + memcpy(tm, &efi_data(op).u.get_time.time, sizeof(*tm)); + } + + if (tc) { + tc->resolution = efi_data(op).u.get_time.resolution; + tc->accuracy = efi_data(op).u.get_time.accuracy; + tc->sets_to_zero = !!(efi_data(op).misc & + XEN_EFI_GET_TIME_SET_CLEARS_NS); + } + + return efi_data(op).status; +} + +static efi_status_t xen_efi_set_time(efi_time_t *tm) +{ + struct xen_platform_op op = INIT_EFI_OP(set_time); + + BUILD_BUG_ON(sizeof(*tm) != sizeof(efi_data(op).u.set_time)); + memcpy(&efi_data(op).u.set_time, tm, sizeof(*tm)); + + if (HYPERVISOR_platform_op(&op) < 0) + return EFI_UNSUPPORTED; + + return efi_data(op).status; +} + +static efi_status_t xen_efi_get_wakeup_time(efi_bool_t *enabled, + efi_bool_t *pending, + efi_time_t *tm) +{ + struct xen_platform_op op = INIT_EFI_OP(get_wakeup_time); + + if (HYPERVISOR_platform_op(&op) < 0) + return EFI_UNSUPPORTED; + + if (tm) { + BUILD_BUG_ON(sizeof(*tm) != sizeof(efi_data(op).u.get_wakeup_time)); + memcpy(tm, &efi_data(op).u.get_wakeup_time, sizeof(*tm)); + } + + if (enabled) + *enabled = !!(efi_data(op).misc & XEN_EFI_GET_WAKEUP_TIME_ENABLED); + + if (pending) + *pending = !!(efi_data(op).misc & XEN_EFI_GET_WAKEUP_TIME_PENDING); + + return efi_data(op).status; +} + +static efi_status_t xen_efi_set_wakeup_time(efi_bool_t enabled, efi_time_t *tm) +{ + struct xen_platform_op op = INIT_EFI_OP(set_wakeup_time); + + BUILD_BUG_ON(sizeof(*tm) != sizeof(efi_data(op).u.set_wakeup_time)); + if (enabled) + efi_data(op).misc = XEN_EFI_SET_WAKEUP_TIME_ENABLE; + if (tm) + memcpy(&efi_data(op).u.set_wakeup_time, tm, sizeof(*tm)); + else + efi_data(op).misc |= XEN_EFI_SET_WAKEUP_TIME_ENABLE_ONLY; + + if (HYPERVISOR_platform_op(&op) < 0) + return EFI_UNSUPPORTED; + + return efi_data(op).status; +} + +static efi_status_t xen_efi_get_variable(efi_char16_t *name, efi_guid_t *vendor, + u32 *attr, unsigned long *data_size, + void *data) +{ + struct xen_platform_op op = INIT_EFI_OP(get_variable); + + set_xen_guest_handle(efi_data(op).u.get_variable.name, name); + BUILD_BUG_ON(sizeof(*vendor) != + sizeof(efi_data(op).u.get_variable.vendor_guid)); + memcpy(&efi_data(op).u.get_variable.vendor_guid, vendor, sizeof(*vendor)); + efi_data(op).u.get_variable.size = *data_size; + set_xen_guest_handle(efi_data(op).u.get_variable.data, data); + + if (HYPERVISOR_platform_op(&op) < 0) + return EFI_UNSUPPORTED; + + *data_size = efi_data(op).u.get_variable.size; + if (attr) + *attr = efi_data(op).misc; + + return efi_data(op).status; +} + +static efi_status_t xen_efi_get_next_variable(unsigned long *name_size, + efi_char16_t *name, + efi_guid_t *vendor) +{ + struct xen_platform_op op = INIT_EFI_OP(get_next_variable_name); + + efi_data(op).u.get_next_variable_name.size = *name_size; + set_xen_guest_handle(efi_data(op).u.get_next_variable_name.name, name); + BUILD_BUG_ON(sizeof(*vendor) != + sizeof(efi_data(op).u.get_next_variable_name.vendor_guid)); + memcpy(&efi_data(op).u.get_next_variable_name.vendor_guid, vendor, + sizeof(*vendor)); + + if (HYPERVISOR_platform_op(&op) < 0) + return EFI_UNSUPPORTED; + + *name_size = efi_data(op).u.get_next_variable_name.size; + memcpy(vendor, &efi_data(op).u.get_next_variable_name.vendor_guid, + sizeof(*vendor)); + + return efi_data(op).status; +} + +static efi_status_t xen_efi_set_variable(efi_char16_t *name, efi_guid_t *vendor, + u32 attr, unsigned long data_size, + void *data) +{ + struct xen_platform_op op = INIT_EFI_OP(set_variable); + + set_xen_guest_handle(efi_data(op).u.set_variable.name, name); + efi_data(op).misc = attr; + BUILD_BUG_ON(sizeof(*vendor) != + sizeof(efi_data(op).u.set_variable.vendor_guid)); + memcpy(&efi_data(op).u.set_variable.vendor_guid, vendor, sizeof(*vendor)); + efi_data(op).u.set_variable.size = data_size; + set_xen_guest_handle(efi_data(op).u.set_variable.data, data); + + if (HYPERVISOR_platform_op(&op) < 0) + return EFI_UNSUPPORTED; + + return efi_data(op).status; +} + +static efi_status_t xen_efi_query_variable_info(u32 attr, u64 *storage_space, + u64 *remaining_space, + u64 *max_variable_size) +{ + struct xen_platform_op op = INIT_EFI_OP(query_variable_info); + + if (efi.runtime_version < EFI_2_00_SYSTEM_TABLE_REVISION) + return EFI_UNSUPPORTED; + + efi_data(op).u.query_variable_info.attr = attr; + + if (HYPERVISOR_platform_op(&op) < 0) + return EFI_UNSUPPORTED; + + *storage_space = efi_data(op).u.query_variable_info.max_store_size; + *remaining_space = efi_data(op).u.query_variable_info.remain_store_size; + *max_variable_size = efi_data(op).u.query_variable_info.max_size; + + return efi_data(op).status; +} + +static efi_status_t xen_efi_get_next_high_mono_count(u32 *count) +{ + struct xen_platform_op op = INIT_EFI_OP(get_next_high_monotonic_count); + + if (HYPERVISOR_platform_op(&op) < 0) + return EFI_UNSUPPORTED; + + *count = efi_data(op).misc; + + return efi_data(op).status; +} + +static efi_status_t xen_efi_update_capsule(efi_capsule_header_t **capsules, + unsigned long count, unsigned long sg_list) +{ + struct xen_platform_op op = INIT_EFI_OP(update_capsule); + + if (efi.runtime_version < EFI_2_00_SYSTEM_TABLE_REVISION) + return EFI_UNSUPPORTED; + + set_xen_guest_handle(efi_data(op).u.update_capsule.capsule_header_array, + capsules); + efi_data(op).u.update_capsule.capsule_count = count; + efi_data(op).u.update_capsule.sg_list = sg_list; + + if (HYPERVISOR_platform_op(&op) < 0) + return EFI_UNSUPPORTED; + + return efi_data(op).status; +} + +static efi_status_t xen_efi_query_capsule_caps(efi_capsule_header_t **capsules, + unsigned long count, u64 *max_size, int *reset_type) +{ + struct xen_platform_op op = INIT_EFI_OP(query_capsule_capabilities); + + if (efi.runtime_version < EFI_2_00_SYSTEM_TABLE_REVISION) + return EFI_UNSUPPORTED; + + set_xen_guest_handle(efi_data(op).u.query_capsule_capabilities.capsule_header_array, + capsules); + efi_data(op).u.query_capsule_capabilities.capsule_count = count; + + if (HYPERVISOR_platform_op(&op) < 0) + return EFI_UNSUPPORTED; + + *max_size = efi_data(op).u.query_capsule_capabilities.max_capsule_size; + *reset_type = efi_data(op).u.query_capsule_capabilities.reset_type; + + return efi_data(op).status; +} + +static void xen_efi_reset_system(int reset_type, efi_status_t status, + unsigned long data_size, efi_char16_t *data) +{ + switch (reset_type) { + case EFI_RESET_COLD: + case EFI_RESET_WARM: + xen_reboot(SHUTDOWN_reboot); + break; + case EFI_RESET_SHUTDOWN: + xen_reboot(SHUTDOWN_poweroff); + break; + default: + BUG(); + } +} + +/* + * Set XEN EFI runtime services function pointers. Other fields of struct efi, + * e.g. efi.systab, will be set like normal EFI. + */ +void __init xen_efi_runtime_setup(void) +{ + efi.get_time = xen_efi_get_time; + efi.set_time = xen_efi_set_time; + efi.get_wakeup_time = xen_efi_get_wakeup_time; + efi.set_wakeup_time = xen_efi_set_wakeup_time; + efi.get_variable = xen_efi_get_variable; + efi.get_next_variable = xen_efi_get_next_variable; + efi.set_variable = xen_efi_set_variable; + efi.set_variable_nonblocking = xen_efi_set_variable; + efi.query_variable_info = xen_efi_query_variable_info; + efi.query_variable_info_nonblocking = xen_efi_query_variable_info; + efi.update_capsule = xen_efi_update_capsule; + efi.query_capsule_caps = xen_efi_query_capsule_caps; + efi.get_next_high_mono_count = xen_efi_get_next_high_mono_count; + efi.reset_system = xen_efi_reset_system; +} + +int efi_mem_desc_lookup(u64 phys_addr, efi_memory_desc_t *out_md) +{ + static_assert(XEN_PAGE_SHIFT == EFI_PAGE_SHIFT, + "Mismatch between EFI_PAGE_SHIFT and XEN_PAGE_SHIFT"); + struct xen_platform_op op; + union xenpf_efi_info *info = &op.u.firmware_info.u.efi_info; + int rc; + + if (!efi_enabled(EFI_PARAVIRT) || efi_enabled(EFI_MEMMAP)) + return __efi_mem_desc_lookup(phys_addr, out_md); + phys_addr &= ~(u64)(EFI_PAGE_SIZE - 1); + op = (struct xen_platform_op) { + .cmd = XENPF_firmware_info, + .u.firmware_info = { + .type = XEN_FW_EFI_INFO, + .index = XEN_FW_EFI_MEM_INFO, + .u.efi_info.mem.addr = phys_addr, + .u.efi_info.mem.size = U64_MAX - phys_addr, + }, + }; + + rc = HYPERVISOR_platform_op(&op); + if (rc) { + pr_warn("Failed to lookup header 0x%llx in Xen memory map: error %d\n", + phys_addr, rc); + } + + out_md->phys_addr = info->mem.addr; + out_md->num_pages = info->mem.size >> EFI_PAGE_SHIFT; + out_md->type = info->mem.type; + out_md->attribute = info->mem.attr; + + return 0; +} + +bool __init xen_efi_config_table_is_usable(const efi_guid_t *guid, + unsigned long table) +{ + efi_memory_desc_t md; + int rc; + + if (!efi_enabled(EFI_PARAVIRT)) + return true; + + rc = efi_mem_desc_lookup(table, &md); + if (rc) + return false; + + switch (md.type) { + case EFI_RUNTIME_SERVICES_CODE: + case EFI_RUNTIME_SERVICES_DATA: + case EFI_ACPI_RECLAIM_MEMORY: + case EFI_ACPI_MEMORY_NVS: + case EFI_RESERVED_TYPE: + return true; + default: + return false; + } +} -- cgit v1.3.1