Re: [PATCH 2/3] tracing/probes: Fix to exit fetching if an error is detected

[Google]

On Sun,  2 Jul 2023 23:47:26 +0900
"Masami Hiramatsu (Google)" <mhiramat@xxxxxxxxxx> wrote:

> From: Masami Hiramatsu (Google) <mhiramat@xxxxxxxxxx>
> 
> Fix to exit fetching arguments if an error is detected when storing
> strings. Without this fix, if an array is specified with string types
> it may store the data at the wrong address.
> 
> Reported-by: Dan Carpenter <dan.carpenter@xxxxxxxxxx>
> Closes: https://lore.kernel.org/all/8819b154-2ba1-43c3-98a2-cbde20892023@moroto.mountain/
> Fixes: 9b960a38835f ("tracing: probeevent: Unify fetch_insn processing common part")
> Cc: stable@xxxxxxxxxxxxxxx
> Signed-off-by: Masami Hiramatsu (Google) <mhiramat@xxxxxxxxxx>
> Reviewed-by: Steven Rostedt (Google) <rostedt@xxxxxxxxxxx>
> ---
>  kernel/trace/trace_probe_tmpl.h |    2 ++
>  1 file changed, 2 insertions(+)
> 
> diff --git a/kernel/trace/trace_probe_tmpl.h b/kernel/trace/trace_probe_tmpl.h
> index 4735c5cb76fa..d6f2bf69f9bc 100644
> --- a/kernel/trace/trace_probe_tmpl.h
> +++ b/kernel/trace/trace_probe_tmpl.h
> @@ -193,6 +193,8 @@ process_fetch_insn_bottom(struct fetch_insn *code, unsigned long val,
>  	default:
>  		return -EILSEQ;
>  	}
> +	if (ret < 0)
> +		return ret;

I found this will leave a garbage data on the trace data if we are in the array.
Let me fix this issue.
(-EILSEQ case has another issue. I think it should not be recorded)

Thank you,

>  	code++;
>  
>  	/* 4th stage: modify stored value if needed */
> 


-- 
Masami Hiramatsu (Google) <mhiramat@xxxxxxxxxx>