tree: https://git.kernel.org/pub/scm/linux/kernel/git/tip/tip.git x86/entry head: 095b7a3e7745e6fb7cf0a1c09967c4f43e76f8f4 commit: aa9712e07f82a5458f2f16c100c491d736240d60 [23/80] x86/entry/common: Protect against instrumentation config: x86_64-randconfig-r031-20200520 (attached as .config) compiler: clang version 11.0.0 (https://github.com/llvm/llvm-project e6658079aca6d971b4e9d7137a3a2ecbc9c34aec) reproduce: wget https://raw.githubusercontent.com/intel/lkp-tests/master/sbin/make.cross -O ~/bin/make.cross chmod +x ~/bin/make.cross # install x86_64 cross compiling tool for clang build # apt-get install binutils-x86-64-linux-gnu git checkout aa9712e07f82a5458f2f16c100c491d736240d60 # save the attached .config to linux build tree COMPILER_INSTALL_PATH=$HOME/0day COMPILER=clang make.cross ARCH=x86_64 If you fix the issue, kindly add following tag as appropriate Reported-by: kbuild test robot <lkp(a)intel.com&gt; All warnings (new ones prefixed by >>, old ones prefixed by <<): __visible noinstr void enter_from_user_mode(void) ^ arch/x86/entry/common.c:44:19: note: declare 'static' if the function is not intended to be used outside of this translation unit __visible noinstr void enter_from_user_mode(void) ^ static __visible noinstr void prepare_exit_to_usermode(struct pt_regs *regs) ^ arch/x86/entry/common.c:234:19: note: declare 'static' if the function is not intended to be used outside of this translation unit __visible noinstr void prepare_exit_to_usermode(struct pt_regs *regs) ^ static __visible noinstr void syscall_return_slowpath(struct pt_regs *regs) ^ arch/x86/entry/common.c:296:19: note: declare 'static' if the function is not intended to be used outside of this translation unit __visible noinstr void syscall_return_slowpath(struct pt_regs *regs) ^ static 3 warnings generated. vim +/enter_from_user_mode +44 arch/x86/entry/common.c 41 42 #ifdef CONFIG_CONTEXT_TRACKING 43 /* Called on entry from user mode with IRQs off. */ 45 { 46 enum ctx_state state = ct_state(); 47 48 user_exit_irqoff(); 49 50 instrumentation_begin(); 51 CT_WARN_ON(state != CONTEXT_USER); 52 instrumentation_end(); 53 } 54 #else 55 static inline void enter_from_user_mode(void) {} 56 #endif 57 58 static noinstr void exit_to_user_mode(void) 59 { 60 user_enter_irqoff(); 61 mds_user_clear_cpu_buffers(); 62 } 63 64 static void do_audit_syscall_entry(struct pt_regs *regs, u32 arch) 65 { 66 #ifdef CONFIG_X86_64 67 if (arch == AUDIT_ARCH_X86_64) { 68 audit_syscall_entry(regs->orig_ax, regs->di, 69 regs->si, regs->dx, regs->r10); 70 } else 71 #endif 72 { 73 audit_syscall_entry(regs->orig_ax, regs->bx, 74 regs->cx, regs->dx, regs->si); 75 } 76 } 77 78 /* 79 * Returns the syscall nr to run (which should match regs->orig_ax) or -1 80 * to skip the syscall. 81 */ 82 static long syscall_trace_enter(struct pt_regs *regs) 83 { 84 u32 arch = in_ia32_syscall() ? AUDIT_ARCH_I386 : AUDIT_ARCH_X86_64; 85 86 struct thread_info *ti = current_thread_info(); 87 unsigned long ret = 0; 88 u32 work; 89 90 if (IS_ENABLED(CONFIG_DEBUG_ENTRY)) 91 BUG_ON(regs != task_pt_regs(current)); 92 93 work = READ_ONCE(ti->flags); 94 95 if (work & (_TIF_SYSCALL_TRACE | _TIF_SYSCALL_EMU)) { 96 ret = tracehook_report_syscall_entry(regs); 97 if (ret || (work & _TIF_SYSCALL_EMU)) 98 return -1L; 99 } 100 101 #ifdef CONFIG_SECCOMP 102 /* 103 * Do seccomp after ptrace, to catch any tracer changes. 104 */ 105 if (work & _TIF_SECCOMP) { 106 struct seccomp_data sd; 107 108 sd.arch = arch; 109 sd.nr = regs->orig_ax; 110 sd.instruction_pointer = regs->ip; 111 #ifdef CONFIG_X86_64 112 if (arch == AUDIT_ARCH_X86_64) { 113 sd.args[0] = regs->di; 114 sd.args[1] = regs->si; 115 sd.args[2] = regs->dx; 116 sd.args[3] = regs->r10; 117 sd.args[4] = regs->r8; 118 sd.args[5] = regs->r9; 119 } else 120 #endif 121 { 122 sd.args[0] = regs->bx; 123 sd.args[1] = regs->cx; 124 sd.args[2] = regs->dx; 125 sd.args[3] = regs->si; 126 sd.args[4] = regs->di; 127 sd.args[5] = regs->bp; 128 } 129 130 ret = __secure_computing(&sd); 131 if (ret == -1) 132 return ret; 133 } 134 #endif 135 136 if (unlikely(test_thread_flag(TIF_SYSCALL_TRACEPOINT))) 137 trace_sys_enter(regs, regs->orig_ax); 138 139 do_audit_syscall_entry(regs, arch); 140 141 return ret ?: regs->orig_ax; 142 } 143 144 #define EXIT_TO_USERMODE_LOOP_FLAGS \ 145 (_TIF_SIGPENDING | _TIF_NOTIFY_RESUME | _TIF_UPROBE | \ 146 _TIF_NEED_RESCHED | _TIF_USER_RETURN_NOTIFY | _TIF_PATCH_PENDING) 147 148 static void exit_to_usermode_loop(struct pt_regs *regs, u32 cached_flags) 149 { 150 /* 151 * In order to return to user mode, we need to have IRQs off with 152 * none of EXIT_TO_USERMODE_LOOP_FLAGS set. Several of these flags 153 * can be set at any time on preemptible kernels if we have IRQs on, 154 * so we need to loop. Disabling preemption wouldn't help: doing the 155 * work to clear some of the flags can sleep. 156 */ 157 while (true) { 158 /* We have work to do. */ 159 local_irq_enable(); 160 161 if (cached_flags & _TIF_NEED_RESCHED) 162 schedule(); 163 164 if (cached_flags & _TIF_UPROBE) 165 uprobe_notify_resume(regs); 166 167 if (cached_flags & _TIF_PATCH_PENDING) 168 klp_update_patch_state(current); 169 170 /* deal with pending signal delivery */ 171 if (cached_flags & _TIF_SIGPENDING) 172 do_signal(regs); 173 174 if (cached_flags & _TIF_NOTIFY_RESUME) { 175 clear_thread_flag(TIF_NOTIFY_RESUME); 176 tracehook_notify_resume(regs); 177 rseq_handle_notify_resume(NULL, regs); 178 } 179 180 if (cached_flags & _TIF_USER_RETURN_NOTIFY) 181 fire_user_return_notifiers(); 182 183 /* Disable IRQs and retry */ 184 local_irq_disable(); 185 186 cached_flags = READ_ONCE(current_thread_info()->flags); 187 188 if (!(cached_flags & EXIT_TO_USERMODE_LOOP_FLAGS)) 189 break; 190 } 191 } 192 193 static void __prepare_exit_to_usermode(struct pt_regs *regs) 194 { 195 struct thread_info *ti = current_thread_info(); 196 u32 cached_flags; 197 198 addr_limit_user_check(); 199 200 lockdep_assert_irqs_disabled(); 201 lockdep_sys_exit(); 202 203 cached_flags = READ_ONCE(ti->flags); 204 205 if (unlikely(cached_flags & EXIT_TO_USERMODE_LOOP_FLAGS)) 206 exit_to_usermode_loop(regs, cached_flags); 207 208 /* Reload ti->flags; we may have rescheduled above. */ 209 cached_flags = READ_ONCE(ti->flags); 210 211 if (unlikely(cached_flags & _TIF_IO_BITMAP)) 212 tss_update_io_bitmap(); 213 214 fpregs_assert_state_consistent(); 215 if (unlikely(cached_flags & _TIF_NEED_FPU_LOAD)) 216 switch_fpu_return(); 217 218 #ifdef CONFIG_COMPAT 219 /* 220 * Compat syscalls set TS_COMPAT. Make sure we clear it before 221 * returning to user mode. We need to clear it *after* signal 222 * handling, because syscall restart has a fixup for compat 223 * syscalls. The fixup is exercised by the ptrace_syscall_32 224 * selftest. 225 * 226 * We also need to clear TS_REGS_POKED_I386: the 32-bit tracer 227 * special case only applies after poking regs and before the 228 * very next return to user mode. 229 */ 230 ti->status &= ~(TS_COMPAT|TS_I386_REGS_POKED); 231 #endif 232 } 233 235 { 236 instrumentation_begin(); 237 __prepare_exit_to_usermode(regs); 238 instrumentation_end(); 239 exit_to_user_mode(); 240 } 241 --- 0-DAY CI Kernel Test Service, Intel Corporation https://lists.01.org/hyperkitty/list/kbuild-all@lists.01.org