[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]
Message-ID: <20200309142223.GD9615@lenoir>
Date: Mon, 9 Mar 2020 15:22:24 +0100
From: Frederic Weisbecker <frederic@...nel.org>
To: Thomas Gleixner <tglx@...utronix.de>
Cc: LKML <linux-kernel@...r.kernel.org>, x86@...nel.org,
Steven Rostedt <rostedt@...dmis.org>,
Brian Gerst <brgerst@...il.com>,
Juergen Gross <jgross@...e.com>,
Alexandre Chartre <alexandre.chartre@...cle.com>
Subject: Re: [patch part-II V2 01/13] context_tracking: Ensure that the
critical path cannot be instrumented
On Sun, Mar 08, 2020 at 11:24:00PM +0100, Thomas Gleixner wrote:
> context tracking lacks a few protection mechanisms against instrumentation:
>
> - While the core functions are marked NOKPROBE they lack protection
> against function tracing which is required as the function entry/exit
> points can be utilized by BPF.
Just to clarify things up: IIUC, BPF scripts can be called from the
function graph tracer hooks, and that BPF code uses RCU, right?
>
> - static functions invoked from the protected functions need to be marked
> as well as they can be instrumented otherwise.
>
> - using plain inline allows the compiler to emit traceable and probable
> functions.
>
> Fix this by adding the missing notrace/NOKPROBE annotations and converting
> the plain inlines to __always_inline.
>
> Signed-off-by: Thomas Gleixner <tglx@...utronix.de>
Acked-by: Frederic Weisbecker <frederic@...nel.org>
Powered by blists - more mailing lists