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, 154 insertions, 0 deletions

diff --git a/arch/nios2/mm/init.c b/arch/nios2/mm/init.c
new file mode 100644
index 000000000..6b22f1995
--- /dev/null
+++ b/arch/nios2/mm/init.c
@@ -0,0 +1,154 @@
+/*
+ * Copyright (C) 2013 Altera Corporation
+ * Copyright (C) 2010 Tobias Klauser <tklauser@distanz.ch>
+ * Copyright (C) 2009 Wind River Systems Inc
+ *   Implemented by fredrik.markstrom@gmail.com and ivarholmqvist@gmail.com
+ * Copyright (C) 2004 Microtronix Datacom Ltd
+ *
+ * based on arch/m68k/mm/init.c
+ *
+ * This file is subject to the terms and conditions of the GNU General Public
+ * License. See the file "COPYING" in the main directory of this archive
+ * for more details.
+ */
+
+#include <linux/signal.h>
+#include <linux/sched.h>
+#include <linux/kernel.h>
+#include <linux/errno.h>
+#include <linux/string.h>
+#include <linux/types.h>
+#include <linux/ptrace.h>
+#include <linux/mman.h>
+#include <linux/mm.h>
+#include <linux/init.h>
+#include <linux/pagemap.h>
+#include <linux/memblock.h>
+#include <linux/slab.h>
+#include <linux/binfmts.h>
+#include <linux/execmem.h>
+
+#include <asm/setup.h>
+#include <asm/page.h>
+#include <asm/sections.h>
+#include <asm/tlb.h>
+#include <asm/mmu_context.h>
+#include <asm/cpuinfo.h>
+#include <asm/processor.h>
+
+pgd_t *pgd_current;
+
+void __init arch_zone_limits_init(unsigned long *max_zone_pfns)
+{
+	max_zone_pfns[ZONE_NORMAL] = max_low_pfn;
+}
+
+/*
+ * paging_init() continues the virtual memory environment setup which
+ * was begun by the code in arch/head.S.
+ * The parameters are pointers to where to stick the starting and ending
+ * addresses of available kernel virtual memory.
+ */
+void __init paging_init(void)
+{
+	pagetable_init();
+	pgd_current = swapper_pg_dir;
+
+	flush_dcache_range((unsigned long)empty_zero_page,
+			(unsigned long)empty_zero_page + PAGE_SIZE);
+}
+
+void __init mmu_init(void)
+{
+	flush_tlb_all();
+}
+
+pgd_t swapper_pg_dir[PTRS_PER_PGD] __aligned(PAGE_SIZE);
+pte_t invalid_pte_table[PTRS_PER_PTE] __aligned(PAGE_SIZE);
+static struct page *kuser_page[1];
+static struct vm_special_mapping vdso_mapping = {
+	.name = "[vdso]",
+	.pages = kuser_page,
+};
+
+static int alloc_kuser_page(void)
+{
+	extern char __kuser_helper_start[], __kuser_helper_end[];
+	int kuser_sz = __kuser_helper_end - __kuser_helper_start;
+	unsigned long vpage;
+
+	vpage = get_zeroed_page(GFP_ATOMIC);
+	if (!vpage)
+		return -ENOMEM;
+
+	/* Copy kuser helpers */
+	memcpy((void *)vpage, __kuser_helper_start, kuser_sz);
+
+	flush_icache_range(vpage, vpage + KUSER_SIZE);
+	kuser_page[0] = virt_to_page(vpage);
+
+	return 0;
+}
+arch_initcall(alloc_kuser_page);
+
+int arch_setup_additional_pages(struct linux_binprm *bprm, int uses_interp)
+{
+	struct mm_struct *mm = current->mm;
+	struct vm_area_struct *vma;
+
+	mmap_write_lock(mm);
+
+	/* Map kuser helpers to user space address */
+	vma = _install_special_mapping(mm, KUSER_BASE, KUSER_SIZE,
+				      VM_READ | VM_EXEC | VM_MAYREAD |
+				      VM_MAYEXEC, &vdso_mapping);
+
+	mmap_write_unlock(mm);
+
+	return IS_ERR(vma) ? PTR_ERR(vma) : 0;
+}
+
+const char *arch_vma_name(struct vm_area_struct *vma)
+{
+	return (vma->vm_start == KUSER_BASE) ? "[kuser]" : NULL;
+}
+
+static const pgprot_t protection_map[16] = {
+	[VM_NONE]					= MKP(0, 0, 0),
+	[VM_READ]					= MKP(0, 0, 1),
+	[VM_WRITE]					= MKP(0, 0, 0),
+	[VM_WRITE | VM_READ]				= MKP(0, 0, 1),
+	[VM_EXEC]					= MKP(1, 0, 0),
+	[VM_EXEC | VM_READ]				= MKP(1, 0, 1),
+	[VM_EXEC | VM_WRITE]				= MKP(1, 0, 0),
+	[VM_EXEC | VM_WRITE | VM_READ]			= MKP(1, 0, 1),
+	[VM_SHARED]					= MKP(0, 0, 0),
+	[VM_SHARED | VM_READ]				= MKP(0, 0, 1),
+	[VM_SHARED | VM_WRITE]				= MKP(0, 1, 0),
+	[VM_SHARED | VM_WRITE | VM_READ]		= MKP(0, 1, 1),
+	[VM_SHARED | VM_EXEC]				= MKP(1, 0, 0),
+	[VM_SHARED | VM_EXEC | VM_READ]			= MKP(1, 0, 1),
+	[VM_SHARED | VM_EXEC | VM_WRITE]		= MKP(1, 1, 0),
+	[VM_SHARED | VM_EXEC | VM_WRITE | VM_READ]	= MKP(1, 1, 1)
+};
+DECLARE_VM_GET_PAGE_PROT
+
+#ifdef CONFIG_EXECMEM
+static struct execmem_info execmem_info __ro_after_init;
+
+struct execmem_info __init *execmem_arch_setup(void)
+{
+	execmem_info = (struct execmem_info){
+		.ranges = {
+			[EXECMEM_DEFAULT] = {
+				.start	= MODULES_VADDR,
+				.end	= MODULES_END,
+				.pgprot	= PAGE_KERNEL_EXEC,
+				.alignment = 1,
+			},
+		},
+	};
+
+	return &execmem_info;
+}
+#endif /* CONFIG_EXECMEM */