From: Joel Fernandes <joel@joelfernandes.org>
To: Steven Rostedt <rostedt@goodmis.org>
Cc: Mathieu Desnoyers <mathieu.desnoyers@efficios.com>,
Peter Zijlstra <peterz@infradead.org>,
linux-kernel <linux-kernel@vger.kernel.org>,
Greg Kroah-Hartman <gregkh@linuxfoundation.org>,
"Gustavo A. R. Silva" <gustavo@embeddedor.com>,
Ingo Molnar <mingo@redhat.com>,
Richard Fontana <rfontana@redhat.com>,
Thomas Gleixner <tglx@linutronix.de>,
paulmck <paulmck@kernel.org>,
Josh Triplett <josh@joshtriplett.org>,
Lai Jiangshan <jiangshanlai@gmail.com>,
Arnaldo Carvalho de Melo <arnaldo.melo@gmail.com>
Subject: Re: [RFC 0/3] Revert SRCU from tracepoint infrastructure
Date: Mon, 10 Feb 2020 14:53:02 -0500 [thread overview]
Message-ID: <20200210195302.GA231192@google.com> (raw)
In-Reply-To: <20200210133045.3beb774e@gandalf.local.home>
On Mon, Feb 10, 2020 at 01:30:45PM -0500, Steven Rostedt wrote:
> On Mon, 10 Feb 2020 12:33:04 -0500 (EST)
> Mathieu Desnoyers <mathieu.desnoyers@efficios.com> wrote:
>
> > The rcu_irq_enter/exit_irqson() does atomic_add_return(), which is even worse
> > than a memory barrier.
>
> As we discussed on IRC, would something like this work (not even
> compiled tested).
>
> -- Steve
>
> diff --git a/include/linux/tracepoint.h b/include/linux/tracepoint.h
> index 1fb11daa5c53..a83fd076a312 100644
> --- a/include/linux/tracepoint.h
> +++ b/include/linux/tracepoint.h
> @@ -179,10 +179,8 @@ static inline struct tracepoint *tracepoint_ptr_deref(tracepoint_ptr_t *p)
> * For rcuidle callers, use srcu since sched-rcu \
> * doesn't work from the idle path. \
> */ \
> - if (rcuidle) { \
> + if (rcuidle) \
> __idx = srcu_read_lock_notrace(&tracepoint_srcu);\
> - rcu_irq_enter_irqson(); \
> - } \
This would still break out-of-tree modules or future code that does
rcu_read_lock() right in a tracepoint callback right?
Or are we saying that rcu_read_lock() in a tracepoint callback is not
allowed? I believe this should then at least be documented somewhere. Also,
what about code in tracepoint callback that calls rcu_read_lock() indirectly
through a path in the kernel, and also code that may expect RCU readers when
doing preempt_disable()?
So basically we are saying with this patch:
1. Don't call in a callback: rcu_read_lock() or preempt_disable() and expect RCU to do
anything for you.
2. Don't call code that does anything that 1. needs.
Is that intended? thanks,
- Joel
> \
> it_func_ptr = rcu_dereference_raw((tp)->funcs); \
> \
> @@ -194,10 +192,8 @@ static inline struct tracepoint *tracepoint_ptr_deref(tracepoint_ptr_t *p)
> } while ((++it_func_ptr)->func); \
> } \
> \
> - if (rcuidle) { \
> - rcu_irq_exit_irqson(); \
> + if (rcuidle) \
> srcu_read_unlock_notrace(&tracepoint_srcu, __idx);\
> - } \
> \
> preempt_enable_notrace(); \
> } while (0)
> diff --git a/include/trace/perf.h b/include/trace/perf.h
> index dbc6c74defc3..86d3b2eb00cd 100644
> --- a/include/trace/perf.h
> +++ b/include/trace/perf.h
> @@ -39,17 +39,27 @@ perf_trace_##call(void *__data, proto) \
> u64 __count = 1; \
> struct task_struct *__task = NULL; \
> struct hlist_head *head; \
> + bool rcu_watching; \
> int __entry_size; \
> int __data_size; \
> int rctx; \
> \
> + rcu_watching = rcu_is_watching(); \
> + \
> + /* Can not use RCU if rcu is not watching and in NMI */ \
> + if (!rcu_watching && in_nmi()) \
> + return; \
> + \
> __data_size = trace_event_get_offsets_##call(&__data_offsets, args); \
> \
> + if (!rcu_watching) \
> + rcu_irq_enter_irqson(); \
> + \
> head = this_cpu_ptr(event_call->perf_events); \
> if (!bpf_prog_array_valid(event_call) && \
> __builtin_constant_p(!__task) && !__task && \
> hlist_empty(head)) \
> - return; \
> + goto out; \
> \
> __entry_size = ALIGN(__data_size + sizeof(*entry) + sizeof(u32),\
> sizeof(u64)); \
> @@ -57,7 +67,7 @@ perf_trace_##call(void *__data, proto) \
> \
> entry = perf_trace_buf_alloc(__entry_size, &__regs, &rctx); \
> if (!entry) \
> - return; \
> + goto out; \
> \
> perf_fetch_caller_regs(__regs); \
> \
> @@ -68,6 +78,9 @@ perf_trace_##call(void *__data, proto) \
> perf_trace_run_bpf_submit(entry, __entry_size, rctx, \
> event_call, __count, __regs, \
> head, __task); \
> +out: \
> + if (!rcu_watching) \
> + rcu_irq_exit_irqson(); \
> }
>
> /*
next prev parent reply other threads:[~2020-02-10 19:53 UTC|newest]
Thread overview: 24+ messages / expand[flat|nested] mbox.gz Atom feed top
2020-02-07 20:56 [RFC 0/3] Revert SRCU from tracepoint infrastructure Joel Fernandes (Google)
2020-02-07 20:56 ` [RFC 1/3] Revert "tracepoint: Use __idx instead of idx in DO_TRACE macro to make it unique" Joel Fernandes (Google)
2020-02-07 21:07 ` Steven Rostedt
2020-02-07 20:56 ` [RFC 2/3] Revert "tracing: Add back in rcu_irq_enter/exit_irqson() for rcuidle tracepoints" Joel Fernandes (Google)
2020-02-07 20:56 ` [RFC 3/3] Revert "tracepoint: Make rcuidle tracepoint callers use SRCU" Joel Fernandes (Google)
2020-02-07 21:24 ` [RFC 0/3] Revert SRCU from tracepoint infrastructure Paul E. McKenney
2020-02-07 21:43 ` Joel Fernandes
2020-02-08 16:39 ` Mathieu Desnoyers
2020-02-08 16:31 ` Mathieu Desnoyers
2020-02-10 9:46 ` Peter Zijlstra
2020-02-10 10:19 ` Peter Zijlstra
2020-02-10 13:36 ` Paul E. McKenney
2020-02-10 13:44 ` Peter Zijlstra
2020-02-10 13:57 ` Paul E. McKenney
2020-02-10 17:17 ` Joel Fernandes
2020-02-10 17:05 ` Steven Rostedt
2020-02-10 17:33 ` Mathieu Desnoyers
2020-02-10 18:30 ` Steven Rostedt
2020-02-10 19:05 ` Mathieu Desnoyers
2020-02-10 19:53 ` Joel Fernandes [this message]
2020-02-10 20:03 ` Steven Rostedt
2020-02-10 20:30 ` Joel Fernandes
2020-02-10 18:07 ` Paul E. McKenney
2020-02-10 16:59 ` Joel Fernandes
Reply instructions:
You may reply publicly to this message via plain-text email
using any one of the following methods:
* Save the following mbox file, import it into your mail client,
and reply-to-all from there: mbox
Avoid top-posting and favor interleaved quoting:
https://en.wikipedia.org/wiki/Posting_style#Interleaved_style
* Reply using the --to, --cc, and --in-reply-to
switches of git-send-email(1):
git send-email \
--in-reply-to=20200210195302.GA231192@google.com \
--to=joel@joelfernandes.org \
--cc=arnaldo.melo@gmail.com \
--cc=gregkh@linuxfoundation.org \
--cc=gustavo@embeddedor.com \
--cc=jiangshanlai@gmail.com \
--cc=josh@joshtriplett.org \
--cc=linux-kernel@vger.kernel.org \
--cc=mathieu.desnoyers@efficios.com \
--cc=mingo@redhat.com \
--cc=paulmck@kernel.org \
--cc=peterz@infradead.org \
--cc=rfontana@redhat.com \
--cc=rostedt@goodmis.org \
--cc=tglx@linutronix.de \
/path/to/YOUR_REPLY
https://kernel.org/pub/software/scm/git/docs/git-send-email.html
* If your mail client supports setting the In-Reply-To header
via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line
before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).