headers
Mike Galbraith | 19 Jun 2013 16:26
Picon

[rfc patch] x86,uv: -rt conversions


The below is what I use to run -rt on UV boxen, diff generated against
3.8-rt.  Should the lock conversions perhaps wander to mainline?

Signed-off-by: Mike Galbraith <bitbucket <at> online.de>

---
 arch/x86/include/asm/uv/uv_bau.h   |   14 +++++++-------
 arch/x86/kernel/apic/x2apic_uv_x.c |    6 +++---
 arch/x86/platform/uv/tlb_uv.c      |   26 +++++++++++++-------------
 arch/x86/platform/uv/uv_time.c     |   21 +++++++++++++--------
 4 files changed, 36 insertions(+), 31 deletions(-)

Index: linux-2.6/arch/x86/include/asm/uv/uv_bau.h
===================================================================
--- linux-2.6.orig/arch/x86/include/asm/uv/uv_bau.h
+++ linux-2.6/arch/x86/include/asm/uv/uv_bau.h
 <at>  <at>  -611,9 +611,9  <at>  <at>  struct bau_control {
 	cycles_t		send_message;
 	cycles_t		period_end;
 	cycles_t		period_time;
-	spinlock_t		uvhub_lock;
-	spinlock_t		queue_lock;
-	spinlock_t		disable_lock;
+	raw_spinlock_t		uvhub_lock;
+	raw_spinlock_t		queue_lock;
+	raw_spinlock_t		disable_lock;
 	/* tunables */
 	int			max_concurr;
 	int			max_concurr_const;
(Continue reading)

Permalink | Reply | 0 comments |
headers
Nicholas Mc Guire | 18 Jun 2013 17:04
Picon

15th Real Time Linux Workshop - Call for Papers


                      15th Real Time Linux Workshop
                            1st Call for Papers

                         October 28 - 31, 2013
                    Dipartimento tecnologie innovative
        Scuola universitaria professionale della Svizzera italiana
                               Lugano-Manno
                               Switzerland 

Following the meetings of academics, developers and users of real-time
and embedded Linux at the previous 14 Real Time Linux Workshops held
world-wide (Vienna, Orlando, Milano, Boston, Valencia, Singapore, Lille,
Lanzhou, Linz, Guadalajara, Dresden, Nairobi, Prague and Chapel Hill) - 
the 2013 Real Time Linux Workshop will come to the Scuola Universitaria 
Professionale della Svizzera Italiana in Lugano-Manno, Switzerland. It will 
be held from October 28 to October 31, 2013.

Rationale

Real-time systems have evolved over the past decades in a relatively
calm manner - performance has increased, one can say dramatically, but
the main paradigms were pretty stable since the mid 80s. With the 
considerable increase of computing power, notably with multicore systems
becoming mainstream in embedded devices, the requirements on RTOS are
shifting and the tools are again trying to catch up.

In stark contrast to 1999 (when the first RTLWS took place), Linux has a 
strong position in all kind of embedded systems, ranging from consumer
electronics to network devices and a broad range of industrial applications
(Continue reading)

Permalink | Reply | 0 comments |
headers
Sebastian Andrzej Siewior | 14 Jun 2013 21:30
Picon

[ANNOUNCE] 3.8.13-rt11

Dear RT Folks,

I'm pleased to announce the 3.8.13-rt11 release.

changes since v3.8.13-rt10:
- use wakeup_timer_waiters() in wake_up() so we do nothing on nort
  kernel. Sent by Zhao Hongjiang
- a fix for a cpu down problem. If kthread is pinned to the same CPU
  which is going down we will spin for ever and wait until kthread
  leaves the CPU. This does not trigger on v3.6-rt because the workqueue
  code there does not create a new process in the notfier callback.
  Reported by Qiang Huang.
- a check if we lose PF_THREAD_BOUND in the workqueue code. Shouldn't
  happen yet it seems it happens from time to time.
- save the cpu mask of the application which disables a CPU. Prior this
  change the application which put a CPU down was allowed to run on any
  CPU if it was restricted to a specifc one. Reported by Zhao Chenhui.

- the SLxB PowerPC, e500 problem is removed from the list without a
  change. The problem triggers even on a v3.6 non-RT kernel after 1-2
  days of runtime on my MPC8572DS. I don't see any problem so far on
  MPC8536 which is mostly the same HW except it is UP but MPC8572DS
  crashes also in UP mode so I belive it is a HW problem.

Known issues:

    - Steven reported a missing acpi from the v3.6 release.
    - a "fix" for i915 leads to high latencies due to wbinvd(). Not sure
      what is the best thing to do here.
(Continue reading)

Permalink | Reply | 0 comments |
headers
Lars Segerlund | 14 Jun 2013 07:18
Picon

pagefaults ...

 Hi,

 I have an application which should be realtime but apparently
generates page faults from a realtime context, I am trying to find a
way to get a stacktrace when this occurs, does anybody have any ideas
?

 I verified with some other testcases that the system actually looks
fine. And I would prefer to know where to look before I dive into the
code.

 I looked at ftrace and perf but haven't figured out how to get a
trace from the app, but I figured this must have occured before so I
better ask first.

 Any hints would be most appreciated !

 / regards, Lars Segerlund.
--
To unsubscribe from this list: send the line "unsubscribe linux-rt-users" in
the body of a message to majordomo <at> vger.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html
Permalink | Reply | 2 comments |
headers
Paul Gortmaker | 10 Jun 2013 23:36
Favicon

[PATCH 0/2] Avoid more bit_spin_lock usage on RT kernels

The bit_spin_locks have been problematic on RT for a long time; dating
at least back to 2.6.11 and their use in the journalling code[1].  We
still have patches today that clobber them for cgroups and jbd/jbd2
code on RT[2].  But there have been some newer users added.

In commit 4e35e6070b1c [2.6.38] ("kernel: add bl_list") we got
the list heads with the zero'th bit reserved for locking.

It was shortly followed with ceb5bdc2d24 ("fs: dcache per-bucket
dcache hash locking") that made it clear the bit was now being used
in a bit_spin_lock context (e.g. in fs/dcache.c).

As of commit 1879fd6a265 [2.6.39] ("add hlist_bl_lock/unlock helpers")
we got helper functions that combined the open coded bit locks into
one place.  At the same time, it makes it more clear that bit_spin_lock
is being used, and where.

Assuming that we still can not use the bit_spin_lock safely on RT,
then users of these helpers will also result in unsafe usage.  Following
the style of "fix" used for jbd code[2], I've done a similar thing here
and introduce a stand-alone lock for the list head.  This may be less
than ideal from a performance standpoint -- currently unclear to me.

I can't pin an actual failing on not having these patches present; I
came by it simply by inspecting the jbd2 code while trying to diagnose
another problem (one which these patches unfortunately don't fix) and
ended up searching for users of bit_spin.

Noting the above, there is also another use case which may be
undesireable for RT -- for the RT trees which now support SLUB,
(Continue reading)

Permalink | Reply | 3 comments |
headers
Steven Rostedt | 7 Jun 2013 22:34
Gravatar

Re: [ANNOUNCE] 3.6.11.4-rt36

On Mon, 2013-05-27 at 09:34 +0200, Christoph Mathys wrote:
> Just did a quick "smoketest" with cyclictest. This release spikes to
> over 600us when opening other gnome-terminals or switching to a VTY
> etc. I checked with 3.6.11.3-rt35, and the problem does not occur
> there.

I'm not able to reproduce this. Can you send me your config.

Thanks!

-- Steve

--
To unsubscribe from this list: send the line "unsubscribe linux-rt-users" in
the body of a message to majordomo <at> vger.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html
Permalink | Reply | 7 comments |
headers
Steven Rostedt | 7 Jun 2013 19:30
Gravatar

[ANNOUNCE] 3.2.46-rt67


Dear RT Folks,

I'm pleased to announce the 3.2.46-rt67 stable release.

This release is just an update to the new stable 3.2.46 version
and no RT specific changes have been made.

You can get this release via the git tree at:

  git://git.kernel.org/pub/scm/linux/kernel/git/rt/linux-stable-rt.git

  branch: v3.2-rt
  Head SHA1: cf4f969e9d468ae99a2897023ad1d343266b0362

Or to build 3.2.46-rt67 directly, the following patches should be applied:

  http://www.kernel.org/pub/linux/kernel/v3.x/linux-3.2.tar.xz

  http://www.kernel.org/pub/linux/kernel/v3.x/patch-3.2.46.xz

  http://www.kernel.org/pub/linux/kernel/projects/rt/3.2/patch-3.2.46-rt67.patch.xz

Enjoy,

-- Steve

--
To unsubscribe from this list: send the line "unsubscribe linux-rt-users" in
the body of a message to majordomo <at> vger.kernel.org
(Continue reading)

Permalink | Reply | 0 comments |
headers
Steven Rostedt | 7 Jun 2013 03:40
Gravatar

[ANNOUNCE] 3.6.11.5-rt37


Dear RT Folks,

I'm pleased to announce the 3.6.11.5-rt37 stable release.

This release is just an update to the new stable 3.6.11.5 version
and no RT specific changes have been made.

You can get this release via the git tree at:

  git://git.kernel.org/pub/scm/linux/kernel/git/rt/linux-stable-rt.git

  branch: v3.6-rt
  Head SHA1: ef774ab94415f71f2740a73b8880abe58ffa0a0a

Or to build 3.6.11.5-rt37 directly, the following patches should be applied:

  http://www.kernel.org/pub/linux/kernel/v3.x/linux-3.6.tar.xz

  http://www.kernel.org/pub/linux/kernel/v3.x/patch-3.6.11.xz

  http://www.kernel.org/pub/linux/kernel/projects/rt/3.6/stable/patch-3.6.11.5.xz

  http://www.kernel.org/pub/linux/kernel/projects/rt/3.6/patch-3.6.11.5-rt37.patch.xz

Enjoy,

-- Steve

--
(Continue reading)

Permalink | Reply | 0 comments |
headers
Kirill Tkhai | 7 Jun 2013 00:31
Picon
Favicon

[PATCH]sched/rt: Simplify pull_rt_task() logic and remove .leaf_rt_rq_list

Based on the patch by Steven Rostedt from previous year:

https://lkml.org/lkml/2012/4/18/517

1)Simplify pull_rt_task() logic: search in pushable tasks of dest runqueue.
The only pullable tasks are the tasks which are pushable in their local rq,
and no others.

2)Remove .leaf_rt_rq_list member of struct rt_rq and functions connected
with it: nobody uses it since now.

Signed-off-by: Kirill Tkhai <tkhai <at> yandex.ru>
CC: Steven Rostedt <rostedt <at> goodmis.org>
CC: Ingo Molnar <mingo <at> redhat.com>
CC: Peter Zijlstra <peterz <at> infradead.org>
---
 kernel/sched/rt.c    |   82 ++++++++++----------------------------------------
 kernel/sched/sched.h |    1 -
 2 files changed, 16 insertions(+), 67 deletions(-)
diff --git a/kernel/sched/rt.c b/kernel/sched/rt.c
index 127a2c4..75286d6 100644
--- a/kernel/sched/rt.c
+++ b/kernel/sched/rt.c
 <at>  <at>  -399,20 +399,6  <at>  <at>  static inline struct task_group *next_task_group(struct task_group *tg)
 		(iter = next_task_group(iter)) &&			\
 		(rt_rq = iter->rt_rq[cpu_of(rq)]);)

-static inline void list_add_leaf_rt_rq(struct rt_rq *rt_rq)
-{
-	list_add_rcu(&rt_rq->leaf_rt_rq_list,
(Continue reading)

Permalink | Reply | 1 comment |
headers
Carsten Emde | 4 Jun 2013 23:12
Favicon

[PATCH 0/1] Fix a longstanding ARM RT bug

The reward of nearly six months of debugging: At least one
of the ARM test systems no longer crashes.

	-Carsten.

--
To unsubscribe from this list: send the line "unsubscribe linux-rt-users" in
the body of a message to majordomo <at> vger.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html
Permalink | Reply | 3 comments |
headers
Lee K Leung | 3 Jun 2013 14:42
Picon
Favicon

RE: Reply

I am Lee Leung, an account manager in China Trust Bank and i have an urgent and confidential business
proposition for you.
Awaiting your urgent response : lkamleung0 <at> yahoo.com.hk
Sincerely,Lee Leung.
--
To unsubscribe from this list: send the line "unsubscribe linux-rt-users" in
the body of a message to majordomo <at> vger.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html
Permalink | Reply | 0 comments |

Gmane