Date: Thu, 30 Apr 2015 10:48:02 -0400
From: Steven Rostedt <rostedt@goodmis.org>
To: Jerry Snitselaar <jsnitsel@redhat.com>
Cc: linux-kernel@vger.kernel.org, Ingo Molnar <mingo@redhat.com>
Subject: Re: [PATCH] tracing: Export tracing clock functions
Message-ID: <20150430104802.239c8943@gandalf.local.home>
In-Reply-To: <1430404600-15141-1-git-send-email-jsnitsel@redhat.com>
References: <1430404600-15141-1-git-send-email-jsnitsel@redhat.com>
MIME-Version: 1.0
Content-Type: text/plain; charset=US-ASCII
Content-Transfer-Encoding: 7bit
Sender: linux-kernel-owner@vger.kernel.org
Content-Length: 1270
Lines: 34

On Thu, 30 Apr 2015 07:36:40 -0700
Jerry Snitselaar <jsnitsel@redhat.com> wrote:

> Critical tracepoint hooks shoud never call anything that takes a lock,
> so they are unable to call getrawmonotonic() or ktime_get().
> 
> Export the rest of the tracing clock functions so can be used in
> tracepoint hooks.

A little background needs to be explained here.

A customer adds their own module to do some analysis and hooks to
tracepoints to do so. They were using ktime_get() as their time source,
but as that grabs a seq lock, it was causing deadlocks.

The tracing clocks were made for this purpose, and I find nothing wrong
with letting users add their own modules (GPL of course) and adding
their own hooks to the tracepoint code. I created it for that purpose.
But if they need to do any timings, they must use lockless clocks,
which the trace clocks are good for.

-- Steve


> 
> Cc: Steven Rostedt <rostedt@goodmis.org>
> Cc: Ingo Molnar <mingo@redhat.com>
> Signed-off-by: Jerry Snitselaar <jsnitsel@redhat.com>
> ---
--
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@vger.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html
Please read the FAQ at  http://www.tux.org/lkml/