2.6.25.4-rt4

2.6.25.4-rt4

[Steven Rostedt]

We are pleased to announce the 2.6.25.4-rt4 tree, which can be
downloaded from the location:

  http://rt.et.redhat.com/download/

Information on the RT patch can be found at:

  http://rt.wiki.kernel.org/index.php/Main_Page

Changes since 2.6.25.4-rt3

(Mostly stuff that has been pushed to 24-rt)

 - Upgraded new ftrace (Steven Rostedt)

  - RT downgrade_write (Steven Rostedt)

  - fix RT rwlocks typecasts (Steven Rostedt)

  - default sched_nr_migrate to 8 (Steven Rostedt)

  - ARM compile fix (Remy Bohmer)

  - vdso vsyscall generic (Luis Goncalves)

  - git-ignore update (Hiroshi Shimamoto)

  - lockstat updates (Peter Zijlstra)

  - Adjust pi_lock in wakeup waiters (Peter Morreale)

  - optimize rt_spin wakeup (Peter Morreale)

  - all archs use task->oncpu (Gregory Haskins)
     [modified use in header by Steven Rostedt]

  - remove redundant try_to_take_lock (Peter Morreale)

  - fastpath try_to_take_lock on steal (Steven Rostedt)

  - preempt __delay fix (Steven Rostedt)

  - ftrace minor fixes (Steven Rostedt, Clark Williams,
                     and Abhishek Sagar)

  - disable ftrace daemon (Steven Rostedt)

  - sched_rt migration updates (Gregory Haskins)

  - nmi show regs fix (Hiroshi Shimamoto)

  - Trace histogram divide by zero fix (Carsten Emde)

to build a 2.6.25.4-rt4 tree, the following patches should be applied:

  http://www.kernel.org/pub/linux/kernel/v2.6/linux-2.6.25.tar.bz2
  http://kernel.org/pub/linux/kernel/v2.6/patch-2.6.25.4.bz2
  http://rt.et.redhat.com/download/patch-2.6.25.4-rt4.bz2



And like always, my RT version of Matt Mackall's ketchup will get this
for you nicely:

  http://people.redhat.com/srostedt/rt/tools/ketchup-0.9.8-rt3


The broken out patches are also available.


-- Steve

Re: 2.6.25.4-rt4

[Tarkan Erimer]

Steven Rostedt wrote:
> We are pleased to announce the 2.6.25.4-rt4 tree, which can be
> downloaded from the location:
>
>   

Hi,

I just downloaded and tried to compile it. But I got the following error 
message at compile stage. I use Kubuntu-8.04 with linux-2.6.25.4-rt3 
without problems.

kernel/sched.c: In function 'sched_init':
kernel/sched.c:7677: error: implicit declaration of function 
'global_rt_runtime'
make[2]: *** [kernel/sched.o] Error 1
make[1]: *** [kernel] Error 2
make[1]: Leaving directory `/usr/src/linux-2.6.25.4'
make: *** [debian/stamp-build-kernel] Error 2

Re: 2.6.25.4-rt4

[Jan Engelhardt]

On Thursday 2008-05-29 17:59, Steven Rostedt wrote:

>We are pleased to announce the 2.6.25.4-rt4 tree,[...]

I am getting tons of the following two stack traces (see below)
during normal operation. 2.6.25.4-rt4, compiled for i586+SMP,
running on i686+UP. syslogd even gets so far as to say "suppressed 134
messages", so it's really a lot. I am tempted to just disable
CONFIG_DEBUG_PREEMPT because otherwise it seems to work.


kernel: BUG: using smp_processor_id() in preemptible [00000000] code: xfssyncd/491
l: caller is __rcu_preempt_unboost+0x20/0xd0
l: Pid: 491, comm: xfssyncd Tainted: G        N 2.6.25.4-jen67-rt #1
l:  [<c0214b4a>] debug_smp_processor_id+0xba/0xc0
l:  [<c016a7b0>] __rcu_preempt_unboost+0x20/0xd0
l:  [<c020dd3c>] ? radix_tree_lookup_slot+0x3c/0xf0
l:  [<c0169e42>] __rcu_read_unlock+0x92/0xa0
l:  [<c016e8d7>] find_get_page+0x87/0x120
l:  [<c016ecae>] find_lock_page+0x1e/0x70
l:  [<c01710d4>] find_or_create_page+0x34/0xc0
l:  [<f0970352>] _xfs_buf_lookup_pages+0x142/0x290 [xfs]
l:  [<c01411e5>] ? init_waitqueue_head+0x15/0x30
l:  [<f097172a>] xfs_buf_get_flags+0x6a/0x140 [xfs]
l:  [<f097181d>] xfs_buf_read_flags+0x1d/0x90 [xfs]
l:  [<f0963f1b>] xfs_trans_read_buf+0x4b/0x340 [xfs]
l:  [<f094b680>] xfs_itobp+0x70/0x1e0 [xfs]
l:  [<f094e2d1>] xfs_iflush+0xe1/0x480 [xfs]
l:  [<f0967801>] xfs_finish_reclaim+0x101/0x190 [xfs]
l:  [<f0967949>] xfs_finish_reclaim_all+0xb9/0xd0 [xfs]
l:  [<f0965aaf>] xfs_syncsub+0x5f/0x2f0 [xfs]
l:  [<f0965d88>] xfs_sync+0x48/0x70 [xfs]
l:  [<f0976ef7>] xfs_sync_worker+0x27/0x50 [xfs]
l:  [<f0977bc9>] xfssyncd+0x129/0x1a0 [xfs]
l:  [<f0977aa0>] ? xfssyncd+0x0/0x1a0 [xfs]
l:  [<c0140ce2>] kthread+0x42/0x70
l:  [<c0140ca0>] ? kthread+0x0/0x70
l:  [<c0107257>] kernel_thread_helper+0x7/0x10


kernel: BUG: using smp_processor_id() in preemptible [00000000] code: IRQ-1/30
l: caller is __rcu_preempt_unboost+0x20/0xd0
l: Pid: 30, comm: IRQ-1 Tainted: G        N 2.6.25.4-jen67-rt #1
l:  [<c0214b4a>] debug_smp_processor_id+0xba/0xc0
l:  [<c016a7b0>] __rcu_preempt_unboost+0x20/0xd0
l:  [<c0169e42>] __rcu_read_unlock+0x92/0xa0
l:  [<c0294960>] ? evdev_event+0x0/0xc0
l:  [<c029022e>] input_pass_event+0x3e/0xb0
l:  [<c0295940>] ? atkbd_event+0x0/0x50
l:  [<c0290a45>] input_handle_event+0x135/0x3f0
l:  [<c0291fcc>] input_event+0x5c/0x70
l:  [<c0296441>] atkbd_interrupt+0x151/0x5c0
l:  [<c0330881>] ? add_preempt_count+0x11/0xa0
l:  [<c032d9e5>] ? __rt_spin_lock+0x25/0x70
l:  [<c028ddb8>] serio_interrupt+0x38/0x80
l:  [<c028ea5a>] i8042_interrupt+0x10a/0x240
l:  [<c0166b00>] handle_IRQ_event+0x60/0xf0
l:  [<c0330881>] ? add_preempt_count+0x11/0xa0
l:  [<c01678fb>] do_irqd+0x1cb/0x290
l:  [<c0167730>] ? do_irqd+0x0/0x290
l:  [<c0140ce2>] kthread+0x42/0x70
l:  [<c0140ca0>] ? kthread+0x0/0x70

Re: 2.6.25.4-rt4

[Steven Rostedt]

On Sun, 1 Jun 2008, Jan Engelhardt wrote:

>
>
> On Thursday 2008-05-29 17:59, Steven Rostedt wrote:
>
> >We are pleased to announce the 2.6.25.4-rt4 tree,[...]
>
> I am getting tons of the following two stack traces (see below)
> during normal operation. 2.6.25.4-rt4, compiled for i586+SMP,
> running on i686+UP. syslogd even gets so far as to say "suppressed 134
> messages", so it's really a lot. I am tempted to just disable
> CONFIG_DEBUG_PREEMPT because otherwise it seems to work.

Yeah, this has been reported. I'll work on getting a -rt5 out on Monday
with more fixes.

Thanks,

-- Steve