diff --git a/arch/x86/include/asm/vsyscall.h b/arch/x86/include/asm/vsyscall.h index ab60a71a8dcb98..472f0263dbc612 100644 --- a/arch/x86/include/asm/vsyscall.h +++ b/arch/x86/include/asm/vsyscall.h @@ -4,6 +4,7 @@ #include #include +#include #ifdef CONFIG_X86_VSYSCALL_EMULATION extern void map_vsyscall(void); @@ -24,4 +25,13 @@ static inline bool emulate_vsyscall(unsigned long error_code, } #endif +/* + * The (legacy) vsyscall page is the long page in the kernel portion + * of the address space that has user-accessible permissions. + */ +static inline bool is_vsyscall_vaddr(unsigned long vaddr) +{ + return unlikely((vaddr & PAGE_MASK) == VSYSCALL_ADDR); +} + #endif /* _ASM_X86_VSYSCALL_H */ diff --git a/arch/x86/mm/fault.c b/arch/x86/mm/fault.c index 679b09cfe241c7..d6375b3c633bc4 100644 --- a/arch/x86/mm/fault.c +++ b/arch/x86/mm/fault.c @@ -798,15 +798,6 @@ show_signal_msg(struct pt_regs *regs, unsigned long error_code, show_opcodes(regs, loglvl); } -/* - * The (legacy) vsyscall page is the long page in the kernel portion - * of the address space that has user-accessible permissions. - */ -static bool is_vsyscall_vaddr(unsigned long vaddr) -{ - return unlikely((vaddr & PAGE_MASK) == VSYSCALL_ADDR); -} - static void __bad_area_nosemaphore(struct pt_regs *regs, unsigned long error_code, unsigned long address, u32 pkey, int si_code) diff --git a/arch/x86/mm/maccess.c b/arch/x86/mm/maccess.c index 6993f026adec9d..42115ac079cfe6 100644 --- a/arch/x86/mm/maccess.c +++ b/arch/x86/mm/maccess.c @@ -3,6 +3,8 @@ #include #include +#include + #ifdef CONFIG_X86_64 bool copy_from_kernel_nofault_allowed(const void *unsafe_src, size_t size) { @@ -15,6 +17,14 @@ bool copy_from_kernel_nofault_allowed(const void *unsafe_src, size_t size) if (vaddr < TASK_SIZE_MAX + PAGE_SIZE) return false; + /* + * Reading from the vsyscall page may cause an unhandled fault in + * certain cases. Though it is at an address above TASK_SIZE_MAX, it is + * usually considered as a user space address. + */ + if (is_vsyscall_vaddr(vaddr)) + return false; + /* * Allow everything during early boot before 'x86_virt_bits' * is initialized. Needed for instruction decoding in early diff --git a/tools/testing/selftests/bpf/prog_tests/read_vsyscall.c b/tools/testing/selftests/bpf/prog_tests/read_vsyscall.c new file mode 100644 index 00000000000000..3405923fe4e65b --- /dev/null +++ b/tools/testing/selftests/bpf/prog_tests/read_vsyscall.c @@ -0,0 +1,57 @@ +// SPDX-License-Identifier: GPL-2.0 +/* Copyright (C) 2024. Huawei Technologies Co., Ltd */ +#include "test_progs.h" +#include "read_vsyscall.skel.h" + +#if defined(__x86_64__) +/* For VSYSCALL_ADDR */ +#include +#else +/* To prevent build failure on non-x86 arch */ +#define VSYSCALL_ADDR 0UL +#endif + +struct read_ret_desc { + const char *name; + int ret; +} all_read[] = { + { .name = "probe_read_kernel", .ret = -ERANGE }, + { .name = "probe_read_kernel_str", .ret = -ERANGE }, + { .name = "probe_read", .ret = -ERANGE }, + { .name = "probe_read_str", .ret = -ERANGE }, + { .name = "probe_read_user", .ret = -EFAULT }, + { .name = "probe_read_user_str", .ret = -EFAULT }, + { .name = "copy_from_user", .ret = -EFAULT }, + { .name = "copy_from_user_task", .ret = -EFAULT }, +}; + +void test_read_vsyscall(void) +{ + struct read_vsyscall *skel; + unsigned int i; + int err; + +#if !defined(__x86_64__) + test__skip(); + return; +#endif + skel = read_vsyscall__open_and_load(); + if (!ASSERT_OK_PTR(skel, "read_vsyscall open_load")) + return; + + skel->bss->target_pid = getpid(); + err = read_vsyscall__attach(skel); + if (!ASSERT_EQ(err, 0, "read_vsyscall attach")) + goto out; + + /* userspace may don't have vsyscall page due to LEGACY_VSYSCALL_NONE, + * but it doesn't affect the returned error codes. + */ + skel->bss->user_ptr = (void *)VSYSCALL_ADDR; + usleep(1); + + for (i = 0; i < ARRAY_SIZE(all_read); i++) + ASSERT_EQ(skel->bss->read_ret[i], all_read[i].ret, all_read[i].name); +out: + read_vsyscall__destroy(skel); +} diff --git a/tools/testing/selftests/bpf/progs/read_vsyscall.c b/tools/testing/selftests/bpf/progs/read_vsyscall.c new file mode 100644 index 00000000000000..986f96687ae152 --- /dev/null +++ b/tools/testing/selftests/bpf/progs/read_vsyscall.c @@ -0,0 +1,45 @@ +// SPDX-License-Identifier: GPL-2.0 +/* Copyright (C) 2024. Huawei Technologies Co., Ltd */ +#include +#include + +#include "bpf_misc.h" + +int target_pid = 0; +void *user_ptr = 0; +int read_ret[8]; + +char _license[] SEC("license") = "GPL"; + +SEC("fentry/" SYS_PREFIX "sys_nanosleep") +int do_probe_read(void *ctx) +{ + char buf[8]; + + if ((bpf_get_current_pid_tgid() >> 32) != target_pid) + return 0; + + read_ret[0] = bpf_probe_read_kernel(buf, sizeof(buf), user_ptr); + read_ret[1] = bpf_probe_read_kernel_str(buf, sizeof(buf), user_ptr); + read_ret[2] = bpf_probe_read(buf, sizeof(buf), user_ptr); + read_ret[3] = bpf_probe_read_str(buf, sizeof(buf), user_ptr); + read_ret[4] = bpf_probe_read_user(buf, sizeof(buf), user_ptr); + read_ret[5] = bpf_probe_read_user_str(buf, sizeof(buf), user_ptr); + + return 0; +} + +SEC("fentry.s/" SYS_PREFIX "sys_nanosleep") +int do_copy_from_user(void *ctx) +{ + char buf[8]; + + if ((bpf_get_current_pid_tgid() >> 32) != target_pid) + return 0; + + read_ret[6] = bpf_copy_from_user(buf, sizeof(buf), user_ptr); + read_ret[7] = bpf_copy_from_user_task(buf, sizeof(buf), user_ptr, + bpf_get_current_task_btf(), 0); + + return 0; +}