Re: [PATCH 2/2] kprobes/x86: Use 16 bytes for each instruction slot again

[Ingo Molnar <mingo@kernel.org>]

* Eugene Shatokhin <eugene.shatokhin@rosalab.ru> wrote:

> Commit 91e5ed49fca0 ("x86/asm/decoder: Fix and enforce max instruction
> size in the insn decoder") has changed MAX_INSN_SIZE from 16 to 15 bytes
> on x86.
> 
> As a side effect, the slots Kprobes use to store the instructions became
> 1 byte shorter. This is unfortunate because, for example, the Kprobes'
> "boost" feature can not be used now for the instructions of length 11,
> like a quite common kind of MOV:
> * movq $0xffffffffffffffff,-0x3fe8(%rax) (48 c7 80 18 c0 ff ff ff ff ff ff)
> * movq $0x0,0x88(%rdi)                   (48 c7 87 88 00 00 00 00 00 00 00)
> and so on.
> 
> This patch makes the insn slots 16 bytes long, like they were before while
> keeping MAX_INSN_SIZE intact.
> 
> Other tools may benefit from this change as well.
> 
> Signed-off-by: Eugene Shatokhin <eugene.shatokhin@rosalab.ru>
> ---
>  arch/x86/include/asm/kprobes.h | 1 +
>  arch/x86/kernel/kprobes/core.c | 2 +-
>  kernel/kprobes.c               | 8 ++++++--
>  3 files changed, 8 insertions(+), 3 deletions(-)
> 
> diff --git a/arch/x86/include/asm/kprobes.h b/arch/x86/include/asm/kprobes.h
> index 4421b5d..f3f0b4e 100644
> --- a/arch/x86/include/asm/kprobes.h
> +++ b/arch/x86/include/asm/kprobes.h
> @@ -27,6 +27,7 @@
>  #include <asm/insn.h>
>  
>  #define  __ARCH_WANT_KPROBES_INSN_SLOT
> +#define KPROBE_INSN_SLOT_SIZE 16

Naming aside, it's totally unacceptable to add a define like this with no 
explanation in the code at all.

> --- a/kernel/kprobes.c
> +++ b/kernel/kprobes.c
> @@ -90,6 +89,11 @@ static raw_spinlock_t *kretprobe_table_lock_ptr(unsigned long hash)
>  static LIST_HEAD(kprobe_blacklist);
>  
>  #ifdef __ARCH_WANT_KPROBES_INSN_SLOT
> +
> +#ifndef KPROBE_INSN_SLOT_SIZE
> +#define KPROBE_INSN_SLOT_SIZE MAX_INSN_SIZE
> +#endif

Ditto. There's no explanation, no way for an arch maintainer to find out whether 
to redefine this or not.

This adds some magic logic that is just as likely to break in the future as the 
old code did. This series needs to fix the primary cause (==poorly documented, 
illogical code), not just the symptom/bug.

Thanks,

	Ingo