Re: [PATCH 2/7] add syscall tracepoints

[Li Zefan <lizf@cn.fujitsu.com>]

Jason Baron wrote:
> Call arch_init_ftrace_syscalls at boot, so we can determine the set of syscalls
> for the syscall trace events.
> 
> 
> Signed-off-by: Jason Baron <jbaron@redhat.com>
> 
> ---
>  arch/x86/kernel/ftrace.c |    8 +++++---
>  1 files changed, 5 insertions(+), 3 deletions(-)
> 
> diff --git a/arch/x86/kernel/ftrace.c b/arch/x86/kernel/ftrace.c
> index 30baa5b..9f2aa83 100644
> --- a/arch/x86/kernel/ftrace.c
> +++ b/arch/x86/kernel/ftrace.c
> @@ -518,7 +518,7 @@ int syscall_name_to_nr(char *name)
>  	return -1;
>  }
>  
> -void arch_init_ftrace_syscalls(void)
> +int arch_init_ftrace_syscalls(void)

1. It's called at boot, so can be __init.

2. This patch breaks bisectability, now the declaration and definition
   of this function don't match.

>  {
>  	int i;
>  	struct syscall_metadata *meta;
> @@ -532,17 +532,19 @@ void arch_init_ftrace_syscalls(void)
>  					FTRACE_SYSCALL_MAX, GFP_KERNEL);
>  	if (!syscalls_metadata) {
>  		WARN_ON(1);
> -		return;
> +		return -ENOMEM;
>  	}
>  
>  	for (i = 0; i < FTRACE_SYSCALL_MAX; i++) {
>  		meta = find_syscall_meta(psys_syscall_table[i]);
>  		syscalls_metadata[i] = meta;
>  	}
> -	return;
> +	return 0;
>  
>  	/* Paranoid: avoid overflow */
>  end:
>  	atomic_dec(&refs);
> +	return 0;
>  }
> +arch_initcall(arch_init_ftrace_syscalls);
>  #endif