^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1) /*
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 2) * Copyright (C) 1991, 1992 Linus Torvalds
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 3) * Copyright (C) 2000, 2001, 2002 Andi Kleen, SuSE Labs
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 4) *
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 5) * Pentium III FXSR, SSE support
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 6) * Gareth Hughes <gareth@valinux.com>, May 2000
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 7) */
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 8)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 9) /*
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 10) * Handle hardware traps and faults.
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 11) */
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 12)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 13) #define pr_fmt(fmt) KBUILD_MODNAME ": " fmt
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 14)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 15) #include <linux/context_tracking.h>
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 16) #include <linux/interrupt.h>
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 17) #include <linux/kallsyms.h>
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 18) #include <linux/spinlock.h>
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 19) #include <linux/kprobes.h>
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 20) #include <linux/uaccess.h>
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 21) #include <linux/kdebug.h>
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 22) #include <linux/kgdb.h>
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 23) #include <linux/kernel.h>
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 24) #include <linux/export.h>
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 25) #include <linux/ptrace.h>
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 26) #include <linux/uprobes.h>
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 27) #include <linux/string.h>
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 28) #include <linux/delay.h>
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 29) #include <linux/errno.h>
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 30) #include <linux/kexec.h>
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 31) #include <linux/sched.h>
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 32) #include <linux/sched/task_stack.h>
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 33) #include <linux/timer.h>
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 34) #include <linux/init.h>
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 35) #include <linux/bug.h>
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 36) #include <linux/nmi.h>
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 37) #include <linux/mm.h>
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 38) #include <linux/smp.h>
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 39) #include <linux/io.h>
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 40) #include <linux/hardirq.h>
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 41) #include <linux/atomic.h>
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 42)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 43) #include <asm/stacktrace.h>
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 44) #include <asm/processor.h>
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 45) #include <asm/debugreg.h>
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 46) #include <asm/realmode.h>
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 47) #include <asm/text-patching.h>
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 48) #include <asm/ftrace.h>
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 49) #include <asm/traps.h>
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 50) #include <asm/desc.h>
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 51) #include <asm/fpu/internal.h>
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 52) #include <asm/cpu.h>
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 53) #include <asm/cpu_entry_area.h>
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 54) #include <asm/mce.h>
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 55) #include <asm/fixmap.h>
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 56) #include <asm/mach_traps.h>
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 57) #include <asm/alternative.h>
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 58) #include <asm/fpu/xstate.h>
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 59) #include <asm/vm86.h>
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 60) #include <asm/umip.h>
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 61) #include <asm/insn.h>
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 62) #include <asm/insn-eval.h>
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 63)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 64) #ifdef CONFIG_X86_64
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 65) #include <asm/x86_init.h>
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 66) #include <asm/proto.h>
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 67) #else
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 68) #include <asm/processor-flags.h>
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 69) #include <asm/setup.h>
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 70) #include <asm/proto.h>
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 71) #endif
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 72)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 73) DECLARE_BITMAP(system_vectors, NR_VECTORS);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 74)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 75) static inline void cond_local_irq_enable(struct pt_regs *regs)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 76) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 77) if (regs->flags & X86_EFLAGS_IF)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 78) local_irq_enable();
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 79) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 80)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 81) static inline void cond_local_irq_disable(struct pt_regs *regs)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 82) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 83) if (regs->flags & X86_EFLAGS_IF)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 84) local_irq_disable();
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 85) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 86)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 87) __always_inline int is_valid_bugaddr(unsigned long addr)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 88) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 89) if (addr < TASK_SIZE_MAX)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 90) return 0;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 91)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 92) /*
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 93) * We got #UD, if the text isn't readable we'd have gotten
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 94) * a different exception.
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 95) */
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 96) return *(unsigned short *)addr == INSN_UD2;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 97) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 98)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 99) static nokprobe_inline int
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 100) do_trap_no_signal(struct task_struct *tsk, int trapnr, const char *str,
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 101) struct pt_regs *regs, long error_code)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 102) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 103) if (v8086_mode(regs)) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 104) /*
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 105) * Traps 0, 1, 3, 4, and 5 should be forwarded to vm86.
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 106) * On nmi (interrupt 2), do_trap should not be called.
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 107) */
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 108) if (trapnr < X86_TRAP_UD) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 109) if (!handle_vm86_trap((struct kernel_vm86_regs *) regs,
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 110) error_code, trapnr))
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 111) return 0;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 112) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 113) } else if (!user_mode(regs)) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 114) if (fixup_exception(regs, trapnr, error_code, 0))
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 115) return 0;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 116)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 117) tsk->thread.error_code = error_code;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 118) tsk->thread.trap_nr = trapnr;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 119) die(str, regs, error_code);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 120) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 121)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 122) /*
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 123) * We want error_code and trap_nr set for userspace faults and
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 124) * kernelspace faults which result in die(), but not
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 125) * kernelspace faults which are fixed up. die() gives the
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 126) * process no chance to handle the signal and notice the
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 127) * kernel fault information, so that won't result in polluting
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 128) * the information about previously queued, but not yet
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 129) * delivered, faults. See also exc_general_protection below.
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 130) */
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 131) tsk->thread.error_code = error_code;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 132) tsk->thread.trap_nr = trapnr;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 133)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 134) return -1;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 135) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 136)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 137) static void show_signal(struct task_struct *tsk, int signr,
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 138) const char *type, const char *desc,
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 139) struct pt_regs *regs, long error_code)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 140) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 141) if (show_unhandled_signals && unhandled_signal(tsk, signr) &&
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 142) printk_ratelimit()) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 143) pr_info("%s[%d] %s%s ip:%lx sp:%lx error:%lx",
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 144) tsk->comm, task_pid_nr(tsk), type, desc,
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 145) regs->ip, regs->sp, error_code);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 146) print_vma_addr(KERN_CONT " in ", regs->ip);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 147) pr_cont("\n");
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 148) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 149) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 150)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 151) static void
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 152) do_trap(int trapnr, int signr, char *str, struct pt_regs *regs,
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 153) long error_code, int sicode, void __user *addr)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 154) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 155) struct task_struct *tsk = current;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 156)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 157) if (!do_trap_no_signal(tsk, trapnr, str, regs, error_code))
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 158) return;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 159)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 160) show_signal(tsk, signr, "trap ", str, regs, error_code);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 161)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 162) if (!sicode)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 163) force_sig(signr);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 164) else
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 165) force_sig_fault(signr, sicode, addr);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 166) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 167) NOKPROBE_SYMBOL(do_trap);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 168)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 169) static void do_error_trap(struct pt_regs *regs, long error_code, char *str,
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 170) unsigned long trapnr, int signr, int sicode, void __user *addr)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 171) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 172) RCU_LOCKDEP_WARN(!rcu_is_watching(), "entry code didn't wake RCU");
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 173)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 174) if (notify_die(DIE_TRAP, str, regs, error_code, trapnr, signr) !=
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 175) NOTIFY_STOP) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 176) cond_local_irq_enable(regs);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 177) do_trap(trapnr, signr, str, regs, error_code, sicode, addr);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 178) cond_local_irq_disable(regs);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 179) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 180) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 181)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 182) /*
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 183) * Posix requires to provide the address of the faulting instruction for
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 184) * SIGILL (#UD) and SIGFPE (#DE) in the si_addr member of siginfo_t.
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 185) *
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 186) * This address is usually regs->ip, but when an uprobe moved the code out
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 187) * of line then regs->ip points to the XOL code which would confuse
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 188) * anything which analyzes the fault address vs. the unmodified binary. If
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 189) * a trap happened in XOL code then uprobe maps regs->ip back to the
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 190) * original instruction address.
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 191) */
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 192) static __always_inline void __user *error_get_trap_addr(struct pt_regs *regs)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 193) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 194) return (void __user *)uprobe_get_trap_addr(regs);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 195) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 196)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 197) DEFINE_IDTENTRY(exc_divide_error)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 198) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 199) do_error_trap(regs, 0, "divide error", X86_TRAP_DE, SIGFPE,
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 200) FPE_INTDIV, error_get_trap_addr(regs));
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 201) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 202)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 203) DEFINE_IDTENTRY(exc_overflow)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 204) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 205) do_error_trap(regs, 0, "overflow", X86_TRAP_OF, SIGSEGV, 0, NULL);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 206) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 207)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 208) #ifdef CONFIG_X86_F00F_BUG
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 209) void handle_invalid_op(struct pt_regs *regs)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 210) #else
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 211) static inline void handle_invalid_op(struct pt_regs *regs)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 212) #endif
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 213) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 214) do_error_trap(regs, 0, "invalid opcode", X86_TRAP_UD, SIGILL,
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 215) ILL_ILLOPN, error_get_trap_addr(regs));
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 216) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 217)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 218) static noinstr bool handle_bug(struct pt_regs *regs)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 219) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 220) bool handled = false;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 221)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 222) if (!is_valid_bugaddr(regs->ip))
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 223) return handled;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 224)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 225) /*
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 226) * All lies, just get the WARN/BUG out.
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 227) */
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 228) instrumentation_begin();
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 229) /*
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 230) * Since we're emulating a CALL with exceptions, restore the interrupt
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 231) * state to what it was at the exception site.
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 232) */
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 233) if (regs->flags & X86_EFLAGS_IF)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 234) raw_local_irq_enable();
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 235) if (report_bug(regs->ip, regs) == BUG_TRAP_TYPE_WARN) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 236) regs->ip += LEN_UD2;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 237) handled = true;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 238) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 239) if (regs->flags & X86_EFLAGS_IF)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 240) raw_local_irq_disable();
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 241) instrumentation_end();
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 242)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 243) return handled;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 244) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 245)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 246) DEFINE_IDTENTRY_RAW(exc_invalid_op)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 247) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 248) irqentry_state_t state;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 249)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 250) /*
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 251) * We use UD2 as a short encoding for 'CALL __WARN', as such
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 252) * handle it before exception entry to avoid recursive WARN
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 253) * in case exception entry is the one triggering WARNs.
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 254) */
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 255) if (!user_mode(regs) && handle_bug(regs))
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 256) return;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 257)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 258) state = irqentry_enter(regs);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 259) instrumentation_begin();
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 260) handle_invalid_op(regs);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 261) instrumentation_end();
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 262) irqentry_exit(regs, state);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 263) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 264)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 265) DEFINE_IDTENTRY(exc_coproc_segment_overrun)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 266) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 267) do_error_trap(regs, 0, "coprocessor segment overrun",
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 268) X86_TRAP_OLD_MF, SIGFPE, 0, NULL);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 269) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 270)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 271) DEFINE_IDTENTRY_ERRORCODE(exc_invalid_tss)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 272) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 273) do_error_trap(regs, error_code, "invalid TSS", X86_TRAP_TS, SIGSEGV,
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 274) 0, NULL);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 275) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 276)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 277) DEFINE_IDTENTRY_ERRORCODE(exc_segment_not_present)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 278) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 279) do_error_trap(regs, error_code, "segment not present", X86_TRAP_NP,
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 280) SIGBUS, 0, NULL);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 281) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 282)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 283) DEFINE_IDTENTRY_ERRORCODE(exc_stack_segment)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 284) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 285) do_error_trap(regs, error_code, "stack segment", X86_TRAP_SS, SIGBUS,
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 286) 0, NULL);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 287) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 288)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 289) DEFINE_IDTENTRY_ERRORCODE(exc_alignment_check)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 290) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 291) char *str = "alignment check";
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 292)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 293) if (notify_die(DIE_TRAP, str, regs, error_code, X86_TRAP_AC, SIGBUS) == NOTIFY_STOP)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 294) return;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 295)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 296) if (!user_mode(regs))
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 297) die("Split lock detected\n", regs, error_code);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 298)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 299) local_irq_enable();
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 300)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 301) if (handle_user_split_lock(regs, error_code))
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 302) goto out;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 303)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 304) do_trap(X86_TRAP_AC, SIGBUS, "alignment check", regs,
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 305) error_code, BUS_ADRALN, NULL);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 306)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 307) out:
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 308) local_irq_disable();
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 309) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 310)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 311) #ifdef CONFIG_VMAP_STACK
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 312) __visible void __noreturn handle_stack_overflow(const char *message,
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 313) struct pt_regs *regs,
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 314) unsigned long fault_address)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 315) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 316) printk(KERN_EMERG "BUG: stack guard page was hit at %p (stack is %p..%p)\n",
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 317) (void *)fault_address, current->stack,
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 318) (char *)current->stack + THREAD_SIZE - 1);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 319) die(message, regs, 0);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 320)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 321) /* Be absolutely certain we don't return. */
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 322) panic("%s", message);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 323) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 324) #endif
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 325)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 326) /*
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 327) * Runs on an IST stack for x86_64 and on a special task stack for x86_32.
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 328) *
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 329) * On x86_64, this is more or less a normal kernel entry. Notwithstanding the
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 330) * SDM's warnings about double faults being unrecoverable, returning works as
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 331) * expected. Presumably what the SDM actually means is that the CPU may get
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 332) * the register state wrong on entry, so returning could be a bad idea.
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 333) *
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 334) * Various CPU engineers have promised that double faults due to an IRET fault
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 335) * while the stack is read-only are, in fact, recoverable.
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 336) *
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 337) * On x86_32, this is entered through a task gate, and regs are synthesized
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 338) * from the TSS. Returning is, in principle, okay, but changes to regs will
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 339) * be lost. If, for some reason, we need to return to a context with modified
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 340) * regs, the shim code could be adjusted to synchronize the registers.
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 341) *
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 342) * The 32bit #DF shim provides CR2 already as an argument. On 64bit it needs
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 343) * to be read before doing anything else.
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 344) */
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 345) DEFINE_IDTENTRY_DF(exc_double_fault)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 346) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 347) static const char str[] = "double fault";
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 348) struct task_struct *tsk = current;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 349)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 350) #ifdef CONFIG_VMAP_STACK
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 351) unsigned long address = read_cr2();
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 352) #endif
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 353)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 354) #ifdef CONFIG_X86_ESPFIX64
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 355) extern unsigned char native_irq_return_iret[];
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 356)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 357) /*
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 358) * If IRET takes a non-IST fault on the espfix64 stack, then we
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 359) * end up promoting it to a doublefault. In that case, take
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 360) * advantage of the fact that we're not using the normal (TSS.sp0)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 361) * stack right now. We can write a fake #GP(0) frame at TSS.sp0
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 362) * and then modify our own IRET frame so that, when we return,
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 363) * we land directly at the #GP(0) vector with the stack already
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 364) * set up according to its expectations.
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 365) *
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 366) * The net result is that our #GP handler will think that we
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 367) * entered from usermode with the bad user context.
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 368) *
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 369) * No need for nmi_enter() here because we don't use RCU.
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 370) */
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 371) if (((long)regs->sp >> P4D_SHIFT) == ESPFIX_PGD_ENTRY &&
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 372) regs->cs == __KERNEL_CS &&
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 373) regs->ip == (unsigned long)native_irq_return_iret)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 374) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 375) struct pt_regs *gpregs = (struct pt_regs *)this_cpu_read(cpu_tss_rw.x86_tss.sp0) - 1;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 376) unsigned long *p = (unsigned long *)regs->sp;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 377)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 378) /*
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 379) * regs->sp points to the failing IRET frame on the
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 380) * ESPFIX64 stack. Copy it to the entry stack. This fills
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 381) * in gpregs->ss through gpregs->ip.
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 382) *
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 383) */
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 384) gpregs->ip = p[0];
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 385) gpregs->cs = p[1];
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 386) gpregs->flags = p[2];
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 387) gpregs->sp = p[3];
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 388) gpregs->ss = p[4];
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 389) gpregs->orig_ax = 0; /* Missing (lost) #GP error code */
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 390)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 391) /*
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 392) * Adjust our frame so that we return straight to the #GP
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 393) * vector with the expected RSP value. This is safe because
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 394) * we won't enable interupts or schedule before we invoke
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 395) * general_protection, so nothing will clobber the stack
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 396) * frame we just set up.
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 397) *
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 398) * We will enter general_protection with kernel GSBASE,
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 399) * which is what the stub expects, given that the faulting
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 400) * RIP will be the IRET instruction.
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 401) */
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 402) regs->ip = (unsigned long)asm_exc_general_protection;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 403) regs->sp = (unsigned long)&gpregs->orig_ax;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 404)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 405) return;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 406) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 407) #endif
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 408)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 409) irqentry_nmi_enter(regs);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 410) instrumentation_begin();
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 411) notify_die(DIE_TRAP, str, regs, error_code, X86_TRAP_DF, SIGSEGV);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 412)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 413) tsk->thread.error_code = error_code;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 414) tsk->thread.trap_nr = X86_TRAP_DF;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 415)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 416) #ifdef CONFIG_VMAP_STACK
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 417) /*
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 418) * If we overflow the stack into a guard page, the CPU will fail
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 419) * to deliver #PF and will send #DF instead. Similarly, if we
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 420) * take any non-IST exception while too close to the bottom of
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 421) * the stack, the processor will get a page fault while
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 422) * delivering the exception and will generate a double fault.
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 423) *
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 424) * According to the SDM (footnote in 6.15 under "Interrupt 14 -
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 425) * Page-Fault Exception (#PF):
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 426) *
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 427) * Processors update CR2 whenever a page fault is detected. If a
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 428) * second page fault occurs while an earlier page fault is being
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 429) * delivered, the faulting linear address of the second fault will
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 430) * overwrite the contents of CR2 (replacing the previous
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 431) * address). These updates to CR2 occur even if the page fault
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 432) * results in a double fault or occurs during the delivery of a
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 433) * double fault.
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 434) *
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 435) * The logic below has a small possibility of incorrectly diagnosing
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 436) * some errors as stack overflows. For example, if the IDT or GDT
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 437) * gets corrupted such that #GP delivery fails due to a bad descriptor
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 438) * causing #GP and we hit this condition while CR2 coincidentally
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 439) * points to the stack guard page, we'll think we overflowed the
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 440) * stack. Given that we're going to panic one way or another
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 441) * if this happens, this isn't necessarily worth fixing.
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 442) *
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 443) * If necessary, we could improve the test by only diagnosing
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 444) * a stack overflow if the saved RSP points within 47 bytes of
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 445) * the bottom of the stack: if RSP == tsk_stack + 48 and we
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 446) * take an exception, the stack is already aligned and there
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 447) * will be enough room SS, RSP, RFLAGS, CS, RIP, and a
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 448) * possible error code, so a stack overflow would *not* double
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 449) * fault. With any less space left, exception delivery could
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 450) * fail, and, as a practical matter, we've overflowed the
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 451) * stack even if the actual trigger for the double fault was
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 452) * something else.
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 453) */
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 454) if ((unsigned long)task_stack_page(tsk) - 1 - address < PAGE_SIZE) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 455) handle_stack_overflow("kernel stack overflow (double-fault)",
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 456) regs, address);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 457) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 458) #endif
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 459)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 460) pr_emerg("PANIC: double fault, error_code: 0x%lx\n", error_code);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 461) die("double fault", regs, error_code);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 462) panic("Machine halted.");
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 463) instrumentation_end();
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 464) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 465)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 466) DEFINE_IDTENTRY(exc_bounds)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 467) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 468) if (notify_die(DIE_TRAP, "bounds", regs, 0,
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 469) X86_TRAP_BR, SIGSEGV) == NOTIFY_STOP)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 470) return;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 471) cond_local_irq_enable(regs);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 472)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 473) if (!user_mode(regs))
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 474) die("bounds", regs, 0);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 475)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 476) do_trap(X86_TRAP_BR, SIGSEGV, "bounds", regs, 0, 0, NULL);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 477)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 478) cond_local_irq_disable(regs);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 479) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 480)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 481) enum kernel_gp_hint {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 482) GP_NO_HINT,
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 483) GP_NON_CANONICAL,
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 484) GP_CANONICAL
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 485) };
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 486)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 487) /*
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 488) * When an uncaught #GP occurs, try to determine the memory address accessed by
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 489) * the instruction and return that address to the caller. Also, try to figure
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 490) * out whether any part of the access to that address was non-canonical.
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 491) */
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 492) static enum kernel_gp_hint get_kernel_gp_address(struct pt_regs *regs,
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 493) unsigned long *addr)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 494) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 495) u8 insn_buf[MAX_INSN_SIZE];
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 496) struct insn insn;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 497)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 498) if (copy_from_kernel_nofault(insn_buf, (void *)regs->ip,
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 499) MAX_INSN_SIZE))
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 500) return GP_NO_HINT;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 501)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 502) kernel_insn_init(&insn, insn_buf, MAX_INSN_SIZE);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 503) insn_get_modrm(&insn);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 504) insn_get_sib(&insn);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 505)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 506) *addr = (unsigned long)insn_get_addr_ref(&insn, regs);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 507) if (*addr == -1UL)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 508) return GP_NO_HINT;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 509)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 510) #ifdef CONFIG_X86_64
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 511) /*
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 512) * Check that:
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 513) * - the operand is not in the kernel half
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 514) * - the last byte of the operand is not in the user canonical half
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 515) */
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 516) if (*addr < ~__VIRTUAL_MASK &&
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 517) *addr + insn.opnd_bytes - 1 > __VIRTUAL_MASK)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 518) return GP_NON_CANONICAL;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 519) #endif
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 520)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 521) return GP_CANONICAL;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 522) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 523)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 524) #define GPFSTR "general protection fault"
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 525)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 526) static bool fixup_iopl_exception(struct pt_regs *regs)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 527) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 528) struct thread_struct *t = ¤t->thread;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 529) unsigned char byte;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 530) unsigned long ip;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 531)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 532) if (!IS_ENABLED(CONFIG_X86_IOPL_IOPERM) || t->iopl_emul != 3)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 533) return false;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 534)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 535) ip = insn_get_effective_ip(regs);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 536) if (!ip)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 537) return false;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 538)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 539) if (get_user(byte, (const char __user *)ip))
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 540) return false;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 541)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 542) if (byte != 0xfa && byte != 0xfb)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 543) return false;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 544)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 545) if (!t->iopl_warn && printk_ratelimit()) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 546) pr_err("%s[%d] attempts to use CLI/STI, pretending it's a NOP, ip:%lx",
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 547) current->comm, task_pid_nr(current), ip);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 548) print_vma_addr(KERN_CONT " in ", ip);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 549) pr_cont("\n");
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 550) t->iopl_warn = 1;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 551) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 552)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 553) regs->ip += 1;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 554) return true;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 555) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 556)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 557) DEFINE_IDTENTRY_ERRORCODE(exc_general_protection)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 558) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 559) char desc[sizeof(GPFSTR) + 50 + 2*sizeof(unsigned long) + 1] = GPFSTR;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 560) enum kernel_gp_hint hint = GP_NO_HINT;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 561) struct task_struct *tsk;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 562) unsigned long gp_addr;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 563) int ret;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 564)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 565) cond_local_irq_enable(regs);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 566)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 567) if (static_cpu_has(X86_FEATURE_UMIP)) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 568) if (user_mode(regs) && fixup_umip_exception(regs))
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 569) goto exit;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 570) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 571)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 572) if (v8086_mode(regs)) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 573) local_irq_enable();
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 574) handle_vm86_fault((struct kernel_vm86_regs *) regs, error_code);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 575) local_irq_disable();
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 576) return;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 577) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 578)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 579) tsk = current;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 580)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 581) if (user_mode(regs)) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 582) if (fixup_iopl_exception(regs))
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 583) goto exit;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 584)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 585) tsk->thread.error_code = error_code;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 586) tsk->thread.trap_nr = X86_TRAP_GP;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 587)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 588) show_signal(tsk, SIGSEGV, "", desc, regs, error_code);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 589) force_sig(SIGSEGV);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 590) goto exit;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 591) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 592)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 593) if (fixup_exception(regs, X86_TRAP_GP, error_code, 0))
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 594) goto exit;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 595)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 596) tsk->thread.error_code = error_code;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 597) tsk->thread.trap_nr = X86_TRAP_GP;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 598)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 599) /*
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 600) * To be potentially processing a kprobe fault and to trust the result
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 601) * from kprobe_running(), we have to be non-preemptible.
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 602) */
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 603) if (!preemptible() &&
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 604) kprobe_running() &&
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 605) kprobe_fault_handler(regs, X86_TRAP_GP))
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 606) goto exit;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 607)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 608) ret = notify_die(DIE_GPF, desc, regs, error_code, X86_TRAP_GP, SIGSEGV);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 609) if (ret == NOTIFY_STOP)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 610) goto exit;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 611)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 612) if (error_code)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 613) snprintf(desc, sizeof(desc), "segment-related " GPFSTR);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 614) else
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 615) hint = get_kernel_gp_address(regs, &gp_addr);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 616)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 617) if (hint != GP_NO_HINT)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 618) snprintf(desc, sizeof(desc), GPFSTR ", %s 0x%lx",
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 619) (hint == GP_NON_CANONICAL) ? "probably for non-canonical address"
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 620) : "maybe for address",
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 621) gp_addr);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 622)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 623) /*
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 624) * KASAN is interested only in the non-canonical case, clear it
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 625) * otherwise.
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 626) */
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 627) if (hint != GP_NON_CANONICAL)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 628) gp_addr = 0;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 629)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 630) die_addr(desc, regs, error_code, gp_addr);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 631)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 632) exit:
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 633) cond_local_irq_disable(regs);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 634) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 635)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 636) static bool do_int3(struct pt_regs *regs)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 637) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 638) int res;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 639)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 640) #ifdef CONFIG_KGDB_LOW_LEVEL_TRAP
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 641) if (kgdb_ll_trap(DIE_INT3, "int3", regs, 0, X86_TRAP_BP,
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 642) SIGTRAP) == NOTIFY_STOP)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 643) return true;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 644) #endif /* CONFIG_KGDB_LOW_LEVEL_TRAP */
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 645)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 646) #ifdef CONFIG_KPROBES
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 647) if (kprobe_int3_handler(regs))
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 648) return true;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 649) #endif
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 650) res = notify_die(DIE_INT3, "int3", regs, 0, X86_TRAP_BP, SIGTRAP);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 651)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 652) return res == NOTIFY_STOP;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 653) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 654) NOKPROBE_SYMBOL(do_int3);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 655)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 656) static void do_int3_user(struct pt_regs *regs)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 657) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 658) if (do_int3(regs))
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 659) return;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 660)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 661) cond_local_irq_enable(regs);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 662) do_trap(X86_TRAP_BP, SIGTRAP, "int3", regs, 0, 0, NULL);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 663) cond_local_irq_disable(regs);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 664) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 665)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 666) DEFINE_IDTENTRY_RAW(exc_int3)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 667) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 668) /*
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 669) * poke_int3_handler() is completely self contained code; it does (and
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 670) * must) *NOT* call out to anything, lest it hits upon yet another
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 671) * INT3.
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 672) */
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 673) if (poke_int3_handler(regs))
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 674) return;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 675)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 676) /*
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 677) * irqentry_enter_from_user_mode() uses static_branch_{,un}likely()
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 678) * and therefore can trigger INT3, hence poke_int3_handler() must
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 679) * be done before. If the entry came from kernel mode, then use
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 680) * nmi_enter() because the INT3 could have been hit in any context
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 681) * including NMI.
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 682) */
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 683) if (user_mode(regs)) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 684) irqentry_enter_from_user_mode(regs);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 685) instrumentation_begin();
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 686) do_int3_user(regs);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 687) instrumentation_end();
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 688) irqentry_exit_to_user_mode(regs);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 689) } else {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 690) irqentry_state_t irq_state = irqentry_nmi_enter(regs);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 691)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 692) instrumentation_begin();
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 693) if (!do_int3(regs))
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 694) die("int3", regs, 0);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 695) instrumentation_end();
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 696) irqentry_nmi_exit(regs, irq_state);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 697) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 698) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 699)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 700) #ifdef CONFIG_X86_64
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 701) /*
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 702) * Help handler running on a per-cpu (IST or entry trampoline) stack
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 703) * to switch to the normal thread stack if the interrupted code was in
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 704) * user mode. The actual stack switch is done in entry_64.S
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 705) */
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 706) asmlinkage __visible noinstr struct pt_regs *sync_regs(struct pt_regs *eregs)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 707) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 708) struct pt_regs *regs = (struct pt_regs *)this_cpu_read(cpu_current_top_of_stack) - 1;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 709) if (regs != eregs)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 710) *regs = *eregs;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 711) return regs;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 712) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 713)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 714) #ifdef CONFIG_AMD_MEM_ENCRYPT
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 715) asmlinkage __visible noinstr struct pt_regs *vc_switch_off_ist(struct pt_regs *regs)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 716) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 717) unsigned long sp, *stack;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 718) struct stack_info info;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 719) struct pt_regs *regs_ret;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 720)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 721) /*
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 722) * In the SYSCALL entry path the RSP value comes from user-space - don't
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 723) * trust it and switch to the current kernel stack
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 724) */
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 725) if (ip_within_syscall_gap(regs)) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 726) sp = this_cpu_read(cpu_current_top_of_stack);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 727) goto sync;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 728) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 729)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 730) /*
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 731) * From here on the RSP value is trusted. Now check whether entry
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 732) * happened from a safe stack. Not safe are the entry or unknown stacks,
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 733) * use the fall-back stack instead in this case.
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 734) */
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 735) sp = regs->sp;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 736) stack = (unsigned long *)sp;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 737)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 738) if (!get_stack_info_noinstr(stack, current, &info) || info.type == STACK_TYPE_ENTRY ||
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 739) info.type > STACK_TYPE_EXCEPTION_LAST)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 740) sp = __this_cpu_ist_top_va(VC2);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 741)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 742) sync:
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 743) /*
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 744) * Found a safe stack - switch to it as if the entry didn't happen via
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 745) * IST stack. The code below only copies pt_regs, the real switch happens
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 746) * in assembly code.
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 747) */
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 748) sp = ALIGN_DOWN(sp, 8) - sizeof(*regs_ret);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 749)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 750) regs_ret = (struct pt_regs *)sp;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 751) *regs_ret = *regs;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 752)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 753) return regs_ret;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 754) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 755) #endif
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 756)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 757) struct bad_iret_stack {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 758) void *error_entry_ret;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 759) struct pt_regs regs;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 760) };
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 761)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 762) asmlinkage __visible noinstr
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 763) struct bad_iret_stack *fixup_bad_iret(struct bad_iret_stack *s)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 764) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 765) /*
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 766) * This is called from entry_64.S early in handling a fault
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 767) * caused by a bad iret to user mode. To handle the fault
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 768) * correctly, we want to move our stack frame to where it would
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 769) * be had we entered directly on the entry stack (rather than
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 770) * just below the IRET frame) and we want to pretend that the
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 771) * exception came from the IRET target.
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 772) */
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 773) struct bad_iret_stack tmp, *new_stack =
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 774) (struct bad_iret_stack *)__this_cpu_read(cpu_tss_rw.x86_tss.sp0) - 1;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 775)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 776) /* Copy the IRET target to the temporary storage. */
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 777) __memcpy(&tmp.regs.ip, (void *)s->regs.sp, 5*8);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 778)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 779) /* Copy the remainder of the stack from the current stack. */
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 780) __memcpy(&tmp, s, offsetof(struct bad_iret_stack, regs.ip));
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 781)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 782) /* Update the entry stack */
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 783) __memcpy(new_stack, &tmp, sizeof(tmp));
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 784)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 785) BUG_ON(!user_mode(&new_stack->regs));
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 786) return new_stack;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 787) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 788) #endif
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 789)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 790) static bool is_sysenter_singlestep(struct pt_regs *regs)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 791) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 792) /*
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 793) * We don't try for precision here. If we're anywhere in the region of
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 794) * code that can be single-stepped in the SYSENTER entry path, then
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 795) * assume that this is a useless single-step trap due to SYSENTER
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 796) * being invoked with TF set. (We don't know in advance exactly
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 797) * which instructions will be hit because BTF could plausibly
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 798) * be set.)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 799) */
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 800) #ifdef CONFIG_X86_32
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 801) return (regs->ip - (unsigned long)__begin_SYSENTER_singlestep_region) <
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 802) (unsigned long)__end_SYSENTER_singlestep_region -
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 803) (unsigned long)__begin_SYSENTER_singlestep_region;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 804) #elif defined(CONFIG_IA32_EMULATION)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 805) return (regs->ip - (unsigned long)entry_SYSENTER_compat) <
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 806) (unsigned long)__end_entry_SYSENTER_compat -
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 807) (unsigned long)entry_SYSENTER_compat;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 808) #else
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 809) return false;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 810) #endif
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 811) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 812)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 813) static __always_inline unsigned long debug_read_clear_dr6(void)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 814) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 815) unsigned long dr6;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 816)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 817) /*
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 818) * The Intel SDM says:
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 819) *
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 820) * Certain debug exceptions may clear bits 0-3. The remaining
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 821) * contents of the DR6 register are never cleared by the
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 822) * processor. To avoid confusion in identifying debug
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 823) * exceptions, debug handlers should clear the register before
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 824) * returning to the interrupted task.
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 825) *
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 826) * Keep it simple: clear DR6 immediately.
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 827) */
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 828) get_debugreg(dr6, 6);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 829) set_debugreg(DR6_RESERVED, 6);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 830) dr6 ^= DR6_RESERVED; /* Flip to positive polarity */
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 831)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 832) return dr6;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 833) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 834)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 835) /*
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 836) * Our handling of the processor debug registers is non-trivial.
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 837) * We do not clear them on entry and exit from the kernel. Therefore
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 838) * it is possible to get a watchpoint trap here from inside the kernel.
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 839) * However, the code in ./ptrace.c has ensured that the user can
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 840) * only set watchpoints on userspace addresses. Therefore the in-kernel
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 841) * watchpoint trap can only occur in code which is reading/writing
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 842) * from user space. Such code must not hold kernel locks (since it
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 843) * can equally take a page fault), therefore it is safe to call
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 844) * force_sig_info even though that claims and releases locks.
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 845) *
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 846) * Code in ./signal.c ensures that the debug control register
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 847) * is restored before we deliver any signal, and therefore that
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 848) * user code runs with the correct debug control register even though
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 849) * we clear it here.
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 850) *
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 851) * Being careful here means that we don't have to be as careful in a
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 852) * lot of more complicated places (task switching can be a bit lazy
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 853) * about restoring all the debug state, and ptrace doesn't have to
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 854) * find every occurrence of the TF bit that could be saved away even
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 855) * by user code)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 856) *
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 857) * May run on IST stack.
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 858) */
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 859)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 860) static bool notify_debug(struct pt_regs *regs, unsigned long *dr6)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 861) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 862) /*
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 863) * Notifiers will clear bits in @dr6 to indicate the event has been
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 864) * consumed - hw_breakpoint_handler(), single_stop_cont().
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 865) *
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 866) * Notifiers will set bits in @virtual_dr6 to indicate the desire
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 867) * for signals - ptrace_triggered(), kgdb_hw_overflow_handler().
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 868) */
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 869) if (notify_die(DIE_DEBUG, "debug", regs, (long)dr6, 0, SIGTRAP) == NOTIFY_STOP)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 870) return true;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 871)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 872) return false;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 873) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 874)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 875) static __always_inline void exc_debug_kernel(struct pt_regs *regs,
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 876) unsigned long dr6)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 877) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 878) /*
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 879) * Disable breakpoints during exception handling; recursive exceptions
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 880) * are exceedingly 'fun'.
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 881) *
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 882) * Since this function is NOKPROBE, and that also applies to
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 883) * HW_BREAKPOINT_X, we can't hit a breakpoint before this (XXX except a
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 884) * HW_BREAKPOINT_W on our stack)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 885) *
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 886) * Entry text is excluded for HW_BP_X and cpu_entry_area, which
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 887) * includes the entry stack is excluded for everything.
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 888) */
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 889) unsigned long dr7 = local_db_save();
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 890) irqentry_state_t irq_state = irqentry_nmi_enter(regs);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 891) instrumentation_begin();
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 892)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 893) /*
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 894) * If something gets miswired and we end up here for a user mode
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 895) * #DB, we will malfunction.
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 896) */
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 897) WARN_ON_ONCE(user_mode(regs));
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 898)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 899) if (test_thread_flag(TIF_BLOCKSTEP)) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 900) /*
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 901) * The SDM says "The processor clears the BTF flag when it
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 902) * generates a debug exception." but PTRACE_BLOCKSTEP requested
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 903) * it for userspace, but we just took a kernel #DB, so re-set
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 904) * BTF.
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 905) */
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 906) unsigned long debugctl;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 907)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 908) rdmsrl(MSR_IA32_DEBUGCTLMSR, debugctl);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 909) debugctl |= DEBUGCTLMSR_BTF;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 910) wrmsrl(MSR_IA32_DEBUGCTLMSR, debugctl);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 911) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 912)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 913) /*
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 914) * Catch SYSENTER with TF set and clear DR_STEP. If this hit a
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 915) * watchpoint at the same time then that will still be handled.
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 916) */
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 917) if ((dr6 & DR_STEP) && is_sysenter_singlestep(regs))
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 918) dr6 &= ~DR_STEP;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 919)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 920) if (kprobe_debug_handler(regs))
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 921) goto out;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 922)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 923) /*
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 924) * The kernel doesn't use INT1
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 925) */
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 926) if (!dr6)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 927) goto out;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 928)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 929) if (notify_debug(regs, &dr6))
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 930) goto out;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 931)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 932) /*
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 933) * The kernel doesn't use TF single-step outside of:
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 934) *
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 935) * - Kprobes, consumed through kprobe_debug_handler()
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 936) * - KGDB, consumed through notify_debug()
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 937) *
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 938) * So if we get here with DR_STEP set, something is wonky.
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 939) *
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 940) * A known way to trigger this is through QEMU's GDB stub,
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 941) * which leaks #DB into the guest and causes IST recursion.
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 942) */
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 943) if (WARN_ON_ONCE(dr6 & DR_STEP))
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 944) regs->flags &= ~X86_EFLAGS_TF;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 945) out:
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 946) instrumentation_end();
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 947) irqentry_nmi_exit(regs, irq_state);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 948)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 949) local_db_restore(dr7);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 950) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 951)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 952) static __always_inline void exc_debug_user(struct pt_regs *regs,
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 953) unsigned long dr6)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 954) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 955) bool icebp;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 956)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 957) /*
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 958) * If something gets miswired and we end up here for a kernel mode
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 959) * #DB, we will malfunction.
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 960) */
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 961) WARN_ON_ONCE(!user_mode(regs));
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 962)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 963) /*
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 964) * NB: We can't easily clear DR7 here because
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 965) * irqentry_exit_to_usermode() can invoke ptrace, schedule, access
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 966) * user memory, etc. This means that a recursive #DB is possible. If
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 967) * this happens, that #DB will hit exc_debug_kernel() and clear DR7.
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 968) * Since we're not on the IST stack right now, everything will be
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 969) * fine.
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 970) */
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 971)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 972) irqentry_enter_from_user_mode(regs);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 973) instrumentation_begin();
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 974)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 975) /*
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 976) * Start the virtual/ptrace DR6 value with just the DR_STEP mask
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 977) * of the real DR6. ptrace_triggered() will set the DR_TRAPn bits.
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 978) *
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 979) * Userspace expects DR_STEP to be visible in ptrace_get_debugreg(6)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 980) * even if it is not the result of PTRACE_SINGLESTEP.
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 981) */
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 982) current->thread.virtual_dr6 = (dr6 & DR_STEP);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 983)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 984) /*
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 985) * The SDM says "The processor clears the BTF flag when it
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 986) * generates a debug exception." Clear TIF_BLOCKSTEP to keep
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 987) * TIF_BLOCKSTEP in sync with the hardware BTF flag.
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 988) */
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 989) clear_thread_flag(TIF_BLOCKSTEP);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 990)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 991) /*
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 992) * If dr6 has no reason to give us about the origin of this trap,
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 993) * then it's very likely the result of an icebp/int01 trap.
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 994) * User wants a sigtrap for that.
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 995) */
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 996) icebp = !dr6;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 997)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 998) if (notify_debug(regs, &dr6))
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 999) goto out;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1000)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1001) /* It's safe to allow irq's after DR6 has been saved */
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1002) local_irq_enable();
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1003)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1004) if (v8086_mode(regs)) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1005) handle_vm86_trap((struct kernel_vm86_regs *)regs, 0, X86_TRAP_DB);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1006) goto out_irq;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1007) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1008)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1009) /* Add the virtual_dr6 bits for signals. */
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1010) dr6 |= current->thread.virtual_dr6;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1011) if (dr6 & (DR_STEP | DR_TRAP_BITS) || icebp)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1012) send_sigtrap(regs, 0, get_si_code(dr6));
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1013)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1014) out_irq:
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1015) local_irq_disable();
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1016) out:
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1017) instrumentation_end();
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1018) irqentry_exit_to_user_mode(regs);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1019) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1020)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1021) #ifdef CONFIG_X86_64
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1022) /* IST stack entry */
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1023) DEFINE_IDTENTRY_DEBUG(exc_debug)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1024) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1025) exc_debug_kernel(regs, debug_read_clear_dr6());
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1026) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1027)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1028) /* User entry, runs on regular task stack */
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1029) DEFINE_IDTENTRY_DEBUG_USER(exc_debug)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1030) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1031) exc_debug_user(regs, debug_read_clear_dr6());
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1032) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1033) #else
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1034) /* 32 bit does not have separate entry points. */
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1035) DEFINE_IDTENTRY_RAW(exc_debug)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1036) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1037) unsigned long dr6 = debug_read_clear_dr6();
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1038)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1039) if (user_mode(regs))
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1040) exc_debug_user(regs, dr6);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1041) else
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1042) exc_debug_kernel(regs, dr6);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1043) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1044) #endif
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1045)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1046) /*
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1047) * Note that we play around with the 'TS' bit in an attempt to get
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1048) * the correct behaviour even in the presence of the asynchronous
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1049) * IRQ13 behaviour
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1050) */
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1051) static void math_error(struct pt_regs *regs, int trapnr)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1052) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1053) struct task_struct *task = current;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1054) struct fpu *fpu = &task->thread.fpu;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1055) int si_code;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1056) char *str = (trapnr == X86_TRAP_MF) ? "fpu exception" :
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1057) "simd exception";
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1058)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1059) cond_local_irq_enable(regs);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1060)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1061) if (!user_mode(regs)) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1062) if (fixup_exception(regs, trapnr, 0, 0))
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1063) goto exit;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1064)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1065) task->thread.error_code = 0;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1066) task->thread.trap_nr = trapnr;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1067)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1068) if (notify_die(DIE_TRAP, str, regs, 0, trapnr,
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1069) SIGFPE) != NOTIFY_STOP)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1070) die(str, regs, 0);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1071) goto exit;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1072) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1073)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1074) /*
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1075) * Save the info for the exception handler and clear the error.
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1076) */
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1077) fpu__save(fpu);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1078)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1079) task->thread.trap_nr = trapnr;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1080) task->thread.error_code = 0;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1081)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1082) si_code = fpu__exception_code(fpu, trapnr);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1083) /* Retry when we get spurious exceptions: */
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1084) if (!si_code)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1085) goto exit;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1086)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1087) force_sig_fault(SIGFPE, si_code,
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1088) (void __user *)uprobe_get_trap_addr(regs));
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1089) exit:
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1090) cond_local_irq_disable(regs);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1091) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1092)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1093) DEFINE_IDTENTRY(exc_coprocessor_error)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1094) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1095) math_error(regs, X86_TRAP_MF);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1096) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1097)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1098) DEFINE_IDTENTRY(exc_simd_coprocessor_error)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1099) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1100) if (IS_ENABLED(CONFIG_X86_INVD_BUG)) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1101) /* AMD 486 bug: INVD in CPL 0 raises #XF instead of #GP */
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1102) if (!static_cpu_has(X86_FEATURE_XMM)) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1103) __exc_general_protection(regs, 0);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1104) return;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1105) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1106) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1107) math_error(regs, X86_TRAP_XF);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1108) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1109)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1110) DEFINE_IDTENTRY(exc_spurious_interrupt_bug)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1111) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1112) /*
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1113) * This addresses a Pentium Pro Erratum:
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1114) *
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1115) * PROBLEM: If the APIC subsystem is configured in mixed mode with
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1116) * Virtual Wire mode implemented through the local APIC, an
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1117) * interrupt vector of 0Fh (Intel reserved encoding) may be
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1118) * generated by the local APIC (Int 15). This vector may be
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1119) * generated upon receipt of a spurious interrupt (an interrupt
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1120) * which is removed before the system receives the INTA sequence)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1121) * instead of the programmed 8259 spurious interrupt vector.
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1122) *
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1123) * IMPLICATION: The spurious interrupt vector programmed in the
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1124) * 8259 is normally handled by an operating system's spurious
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1125) * interrupt handler. However, a vector of 0Fh is unknown to some
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1126) * operating systems, which would crash if this erratum occurred.
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1127) *
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1128) * In theory this could be limited to 32bit, but the handler is not
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1129) * hurting and who knows which other CPUs suffer from this.
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1130) */
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1131) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1132)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1133) DEFINE_IDTENTRY(exc_device_not_available)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1134) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1135) unsigned long cr0 = read_cr0();
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1136)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1137) #ifdef CONFIG_MATH_EMULATION
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1138) if (!boot_cpu_has(X86_FEATURE_FPU) && (cr0 & X86_CR0_EM)) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1139) struct math_emu_info info = { };
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1140)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1141) cond_local_irq_enable(regs);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1142)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1143) info.regs = regs;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1144) math_emulate(&info);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1145)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1146) cond_local_irq_disable(regs);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1147) return;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1148) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1149) #endif
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1150)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1151) /* This should not happen. */
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1152) if (WARN(cr0 & X86_CR0_TS, "CR0.TS was set")) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1153) /* Try to fix it up and carry on. */
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1154) write_cr0(cr0 & ~X86_CR0_TS);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1155) } else {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1156) /*
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1157) * Something terrible happened, and we're better off trying
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1158) * to kill the task than getting stuck in a never-ending
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1159) * loop of #NM faults.
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1160) */
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1161) die("unexpected #NM exception", regs, 0);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1162) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1163) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1164)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1165) #ifdef CONFIG_X86_32
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1166) DEFINE_IDTENTRY_SW(iret_error)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1167) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1168) local_irq_enable();
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1169) if (notify_die(DIE_TRAP, "iret exception", regs, 0,
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1170) X86_TRAP_IRET, SIGILL) != NOTIFY_STOP) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1171) do_trap(X86_TRAP_IRET, SIGILL, "iret exception", regs, 0,
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1172) ILL_BADSTK, (void __user *)NULL);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1173) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1174) local_irq_disable();
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1175) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1176) #endif
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1177)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1178) void __init trap_init(void)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1179) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1180) /* Init cpu_entry_area before IST entries are set up */
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1181) setup_cpu_entry_areas();
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1182)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1183) /* Init GHCB memory pages when running as an SEV-ES guest */
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1184) sev_es_init_vc_handling();
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1185)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1186) idt_setup_traps();
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1187)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1188) /*
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1189) * Should be a barrier for any external CPU state:
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1190) */
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1191) cpu_init();
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1192)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1193) idt_setup_ist_traps();
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1194) }
Orange Pi5 kernel
Deprecated Linux kernel 5.10.110 for OrangePi 5/5B/5+ boards
3 Commits
0 Branches
0 Tags