linux-kernel - Re: [PATCH v2 2/3] tracing: initialize trace

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: Windows password security audit tool. GUI, reports in PDF.

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

Date: Thu, 25 Jan 2024 15:44:32 -0500
From: Steven Rostedt <rostedt@...dmis.org>
To: "Ricardo B. Marliere" <ricardo@...liere.net>
Cc: Masami Hiramatsu <mhiramat@...nel.org>, Mathieu Desnoyers
 <mathieu.desnoyers@...icios.com>, linux-kernel@...r.kernel.org,
 linux-trace-kernel@...r.kernel.org
Subject: Re: [PATCH v2 2/3] tracing: initialize trace_seq buffers

On Thu, 25 Jan 2024 17:16:21 -0300
"Ricardo B. Marliere" <ricardo@...liere.net> wrote:

> Now that trace_seq_reset have been created, correct the places where the
> buffers need to be initialized.

This patch would need to come first. You don't ever want to intentionally
create a broken kernel.

Also, the change log should be:

  In order to extend trace_seq into being dynamic, the struct trace_seq
  will no longer be valid if simply set to zero. Call trace_seq_init() for
  all trace_seq when they are first created.

> 
> Suggested-by: Steven Rostedt <rostedt@...dmis.org>
> Signed-off-by: Ricardo B. Marliere <ricardo@...liere.net>
> ---
>  kernel/trace/trace.c | 14 ++++++++++++++
>  1 file changed, 14 insertions(+)
> 

You also need to initialize iter.seq in ftrace_dump()

-- Steve

> diff --git a/kernel/trace/trace.c b/kernel/trace/trace.c
> index d4c55d3e21c2..9827700d0164 100644
> --- a/kernel/trace/trace.c
> +++ b/kernel/trace/trace.c
> @@ -4889,6 +4889,9 @@ __tracing_open(struct inode *inode, struct file *file, bool snapshot)
>  
>  	mutex_unlock(&trace_types_lock);
>  
> +	trace_seq_init(&iter->seq);
> +	trace_seq_init(&iter->tmp_seq);
> +
>  	return iter;
>  
>   fail:
> @@ -6770,6 +6773,7 @@ static int tracing_open_pipe(struct inode *inode, struct file *filp)
>  	}
>  
>  	trace_seq_init(&iter->seq);
> +	trace_seq_init(&iter->tmp_seq);
>  	iter->trace = tr->current_trace;
>  
>  	if (!alloc_cpumask_var(&iter->started, GFP_KERNEL)) {
> @@ -6947,6 +6951,7 @@ tracing_read_pipe(struct file *filp, char __user *ubuf,
>  	trace_iterator_reset(iter);
>  	cpumask_clear(iter->started);
>  	trace_seq_init(&iter->seq);
> +	trace_seq_init(&iter->tmp_seq);
>  
>  	trace_event_read_lock();
>  	trace_access_lock(iter->cpu_file);
> @@ -8277,6 +8282,9 @@ static int tracing_buffers_open(struct inode *inode, struct file *filp)
>  	if (ret < 0)
>  		trace_array_put(tr);
>  
> +	trace_seq_init(&info->iter.seq);
> +	trace_seq_init(&info->iter.tmp_seq);
> +
>  	return ret;
>  }
>  
> @@ -10300,6 +10308,9 @@ void trace_init_global_iter(struct trace_iterator *iter)
>  	iter->temp_size = STATIC_TEMP_BUF_SIZE;
>  	iter->fmt = static_fmt_buf;
>  	iter->fmt_size = STATIC_FMT_BUF_SIZE;
> +
> +	trace_seq_init(&iter->seq);
> +	trace_seq_init(&iter->tmp_seq);
>  }
>  
>  void ftrace_dump(enum ftrace_dump_mode oops_dump_mode)
> @@ -10712,6 +10723,9 @@ void __init early_trace_init(void)
>  			tracepoint_printk = 0;
>  		else
>  			static_key_enable(&tracepoint_printk_key.key);
> +
> +		trace_seq_init(&tracepoint_print_iter->seq);
> +		trace_seq_init(&tracepoint_print_iter->tmp_seq);
>  	}
>  	tracer_alloc_buffers();
>