[RFC PATCH v3 15/26] early kprobes: x86: directly modify code.

[Masami Hiramatsu]

(2015/02/13 14:40), Wang Nan wrote:
> When registering early kprobes, SMP should has not been enabled, so
> doesn't require synchronization in text_poke_bp(). Simply memcpy is
> enough.

BTW, we've already have text_poke_early for this purpose.

Thank you,

> 
> Signed-off-by: Wang Nan <wangnan0 at huawei.com>
> ---
>  arch/x86/kernel/kprobes/opt.c | 12 ++++++++++--
>  1 file changed, 10 insertions(+), 2 deletions(-)
> 
> diff --git a/arch/x86/kernel/kprobes/opt.c b/arch/x86/kernel/kprobes/opt.c
> index 0dd8d08..21847ab 100644
> --- a/arch/x86/kernel/kprobes/opt.c
> +++ b/arch/x86/kernel/kprobes/opt.c
> @@ -36,6 +36,7 @@
>  #include <asm/alternative.h>
>  #include <asm/insn.h>
>  #include <asm/debugreg.h>
> +#include <asm/tlbflush.h>
>  
>  #include "common.h"
>  
> @@ -397,8 +398,15 @@ void arch_optimize_kprobes(struct list_head *oplist)
>  		insn_buf[0] = RELATIVEJUMP_OPCODE;
>  		*(s32 *)(&insn_buf[1]) = rel;
>  
> -		text_poke_bp(op->kp.addr, insn_buf, RELATIVEJUMP_SIZE,
> -			     op->optinsn.insn);
> +		if (unlikely(kprobes_is_early())) {
> +			BUG_ON(!(op->kp.flags & KPROBE_FLAG_EARLY));
> +			memcpy(op->kp.addr, insn_buf, RELATIVEJUMP_SIZE);
> +			local_flush_tlb();
> +			sync_core();
> +		} else {
> +			text_poke_bp(op->kp.addr, insn_buf, RELATIVEJUMP_SIZE,
> +				     op->optinsn.insn);
> +		}
>  
>  		list_del_init(&op->list);
>  	}
> 


-- 
Masami HIRAMATSU
Software Platform Research Dept. Linux Technology Research Center
Hitachi, Ltd., Yokohama Research Laboratory
E-mail: masami.hiramatsu.pt at hitachi.com