linux-kernel - Re: [PATCH 02/13] tracing: Extract calls to trace_define_common

lists.openwall.net		lists / announce owl-users owl-dev john-users john-dev passwdqc-users yescrypt popa3d-users / oss-security kernel-hardening musl sabotage tlsify passwords / crypt-dev xvendor / Bugtraq Full-Disclosure linux-kernel linux-netdev linux-ext4 linux-hardening linux-cve-announce PHC
Open Source and information security mailing list archives

Hash Suite for Android: free password hash cracker in your pocket

[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]

Date:	Tue, 08 Dec 2009 13:36:20 +0800
From:	Lai Jiangshan <laijs@...fujitsu.com>
To:	Ingo Molnar <mingo@...e.hu>
CC:	Li Zefan <lizf@...fujitsu.com>,
	Steven Rostedt <rostedt@...dmis.org>,
	Frederic Weisbecker <fweisbec@...il.com>,
	LKML <linux-kernel@...r.kernel.org>,
	Masami Hiramatsu <mhiramat@...hat.com>
Subject: Re: [PATCH 02/13] tracing: Extract calls to trace_define_common_fields()

Li Zefan wrote:
> Call trace_define_common_fields() in event_create_dir() only.
> 
>    text    data     bss     dec     hex filename
> 5346802 1961864 7103260 14411926         dbe896 vmlinux.o.old
> 5345151 1961864 7103260 14410275         dbe223 vmlinux.o
> 
> Signed-off-by: Li Zefan <lizf@...fujitsu.com>
> Acked-by: Frederic Weisbecker <fweisbec@...il.com>

I think this is a urgent patch for mainline.

I have found a regression at trace_kprobe, and I'm lucky to find that this
patch can also fix this regression.


> --- a/kernel/trace/trace_kprobe.c
> +++ b/kernel/trace/trace_kprobe.c
> @@ -1113,10 +1113,6 @@ static int kprobe_event_define_fields(struct ftrace_event_call *event_call)
>  	struct kprobe_trace_entry field;
>  	struct trace_probe *tp = (struct trace_probe *)event_call->data;
>  
> -	ret = trace_define_common_fields(event_call);
> -	if (!ret)
> -		return ret;
> -
>  	DEFINE_FIELD(unsigned long, ip, FIELD_STRING_IP, 0);
>  	DEFINE_FIELD(int, nargs, FIELD_STRING_NARGS, 1);
>  	/* Set argument names as fields */
> @@ -1131,10 +1127,6 @@ static int kretprobe_event_define_fields(struct ftrace_event_call *event_call)
>  	struct kretprobe_trace_entry field;
>  	struct trace_probe *tp = (struct trace_probe *)event_call->data;
>  
> -	ret = trace_define_common_fields(event_call);
> -	if (!ret)
> -		return ret;
> -
>  	DEFINE_FIELD(unsigned long, func, FIELD_STRING_FUNC, 0);
>  	DEFINE_FIELD(unsigned long, ret_ip, FIELD_STRING_RETIP, 0);
>  	DEFINE_FIELD(int, nargs, FIELD_STRING_NARGS, 1);

"if (!ret)" is wrong. trace_define_common_fields() returns zero when success.
So "unsigned long, func", "unsigned long, ret_ip" ...etc are NOT "defined"
and filters can not be applied for trace_kprobe.

Reviewed-by: Lai Jiangshan <laijs@...fujitsu.com>

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