Lines Matching full:int3
217 * by int3. And the instruction at kp->addr is not modified by kprobes in __recover_probed_insn()
417 /* Prepare reljump or int3 right after instruction */
435 /* Otherwise, put an int3 for trapping singlestep */ in prepare_singlestep()
603 * int3 single stepping. in prepare_emulation()
616 case 0xcc: /* int3 */ in prepare_emulation()
725 /* Add int3 for single-step or booster jmp */ in arch_copy_kprobe()
770 u8 int3 = INT3_INSN_OPCODE; in arch_arm_kprobe() local
772 text_poke(p->addr, &int3, 1); in arch_arm_kprobe()
774 perf_event_text_poke(p->addr, &p->opcode, 1, &int3, 1); in arch_arm_kprobe()
779 u8 int3 = INT3_INSN_OPCODE; in arch_disarm_kprobe() local
781 perf_event_text_poke(p->addr, &int3, 1, &p->opcode, 1); in arch_disarm_kprobe()
885 * instruction whose first byte has been replaced by the "int3"
889 * copy is p->ainsn.insn. We also doesn't use trap, but "int3" again
891 * Different from the trap single-step, "int3" single-step can not
896 * using "int3".
898 * When the 2nd "int3" handled, the regs->ip and regs->flags needs to
909 /* Note that regs->ip is executed int3 so must be a step back */ in resume_singlestep()
971 * processing. Since int3 and debug trap disables irqs and we clear in kprobe_int3_handler()
1003 /* Most provably this is the second int3 for singlestep */ in kprobe_int3_handler()