[PATCH RFC 3/6] arm64: Kprobes instruction simulation support

[Sandeepa Prabhu]

On 8 November 2013 22:33, Will Deacon <will.deacon at arm.com> wrote:
> On Thu, Oct 17, 2013 at 12:17:48PM +0100, Sandeepa Prabhu wrote:
>> Add support for AArch64 instruction simulation in kprobes.
>>
>> Kprobes need simulation of instructions that cannot be stepped
>> right-away from different memory location. i.e. those instructions
>> that uses PC-relative addressing. In simulation, the behaviour
>> of the instruction is implemented using copy of pt_regs.
>>
>> Following instruction catagories are simulated:
>>  - All branching instructions(conditional, register, and immediate)
>>  - Literal access instructions(load-literal, adr/adrp)
>>
>> conditional execution are limited to branching instructions in
>> ARM v8. If conditions at PSTATE does not match the condition fields
>> of opcode, the instruction is effectively NOP. Kprobes consider
>> this case as 'miss'.
>
> [...]
>
>> diff --git a/arch/arm64/kernel/kprobes-arm64.c b/arch/arm64/kernel/kprobes-arm64.c
>> index 30d1c14..c690be3 100644
>> --- a/arch/arm64/kernel/kprobes-arm64.c
>> +++ b/arch/arm64/kernel/kprobes-arm64.c
>> @@ -20,6 +20,101 @@
>>
>>  #include "probes-decode.h"
>>  #include "kprobes-arm64.h"
>> +#include "simulate-insn.h"
>> +
>> +/*
>> + * condition check functions for kprobes simulation
>> + */
>> +static unsigned long __kprobes
>> +__check_pstate(struct kprobe *p, struct pt_regs *regs)
>> +{
>> +       struct arch_specific_insn *asi = &p->ainsn;
>> +       unsigned long pstate = regs->pstate & 0xffffffff;
>> +
>> +       return asi->pstate_cc(pstate);
>> +}
>> +
>> +static unsigned long __kprobes
>> +__check_cbz(struct kprobe *p, struct pt_regs *regs)
>> +{
>> +       return check_cbz((u32)p->opcode, regs);
>
> Isn't p->opcode already a u32? (by your definition of kprobe_opcode_t).
Yup, can avoid typecasting.
>
>> diff --git a/arch/arm64/kernel/simulate-insn.c b/arch/arm64/kernel/simulate-insn.c
>> new file mode 100644
>> index 0000000..10173cf
>> --- /dev/null
>> +++ b/arch/arm64/kernel/simulate-insn.c
>> @@ -0,0 +1,184 @@
>> +/*
>> + * arch/arm64/kernel/simulate-insn.c
>> + *
>> + * Copyright (C) 2013 Linaro Limited.
>> + *
>> + * This program is free software; you can redistribute it and/or modify
>> + * it under the terms of the GNU General Public License version 2 as
>> + * published by the Free Software Foundation.
>> + *
>> + * This program is distributed in the hope that it will be useful,
>> + * but WITHOUT ANY WARRANTY; without even the implied warranty of
>> + * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the GNU
>> + * General Public License for more details.
>> + */
>> +
>> +#include <linux/kernel.h>
>> +#include <linux/kprobes.h>
>> +#include <linux/module.h>
>> +
>> +#include "simulate-insn.h"
>> +
>> +#define sign_extend(x, signbit)                \
>> +       ((x) | (0 - ((x) & (1 << (signbit)))))
>> +
>> +#define bbl_displacement(insn)         \
>> +       sign_extend(((insn) & 0x3ffffff) << 2, 27)
>> +
>> +#define bcond_displacement(insn)       \
>> +       sign_extend(((insn >> 5) & 0xfffff) << 2, 21)
>> +
>> +#define cbz_displacement(insn) \
>> +       sign_extend(((insn >> 5) & 0xfffff) << 2, 21)
>> +
>> +#define tbz_displacement(insn) \
>> +       sign_extend(((insn >> 5) & 0x3fff) << 2, 15)
>> +
>> +#define ldr_displacement(insn) \
>> +       sign_extend(((insn >> 5) & 0xfffff) << 2, 21)
>
> The mask, shift and signbit position are all related here, so you could
> rework the definition of sign_extend to avoid having three magic numbers.
Hmm, mask and signbit are related, shift is based on instruction type
(conditional instructions need >> 5 for extracting immediate offset).
I will refine these macros in next version.

>
> Will
> --
> To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
> the body of a message to majordomo at vger.kernel.org
> More majordomo info at  http://vger.kernel.org/majordomo-info.html
> Please read the FAQ at  http://www.tux.org/lkml/