We are pleased to announce the next update to our new preempt-rt
series.
Changes from 2.6.33.3-rt19 to 2.6.33.4-rt20:
5b6c7b0: net: Make [dis/en]able_irq_*_lockdep() RT safe
d6434af: fs: Resolve mntput_no_expire issues.
56e4c7f: fs: Fix mnt_count typo
1540c84: Merge branch '2.6.33.4' into rt/2.6.33
1a3a403: autofs4: Remove another autofs4_lock deadlock
6b568a0: autofs: Remove deadlock
c54c071: fs: Add missing parantheses
Changes from 2.6.33.4-rt20 to 2.6.33.4-rt21:
cc4a826: sched: Fix wake_affine() vs RT tasks
ea17fed: printk: Fix missing klogd wakeup
9846da2: fix undefined references to kernel_sem
e868252: fs: namespace: Fix fuse mount fallout
44b341a: net: ehea: make rx irq handler non-threaded (IRQF_NODELAY)
Changes from 2.6.33.4-rt21 to 2.6.33.5-rt22:
ef4afcc: Merge stable/linux-2.6.33.y into rt/2.6.33
Changes from 2.6.33.5-rt22 to 2.6.33.5-rt23:
ab774e8: drivers/hwmon/coretemp.c: get TjMax value from MSR
e8bfcca: drivers/hwmon/coretemp.c: detect the thermal sensors by CPUID
a06a867: dcache: Prevent d_genocide() from decrementing d_count more than once
bdeadaf: nfs: Avoid igrab deadlock.
6976846: dcache: Fix select_parent dentry traversal locking
1f0f6ec: powerpc: chrp/time.c fix preprocessor conditional
Bulk of changes comes from stable tree. The -rt related changes are
mostly fixes for the vfs fallout the rest are fixlets all over the
place and hardware enablement.
Download locations:
http://www.kernel.org/pub/linux/kernel/projects/rt/
Git release branch:
git://git.kernel.org/pub/scm/linux/kernel/git/tip/linux-2.6-tip.git rt/2.6.33
Gitweb:
http://git.kernel.org/?p=linux/kernel/git/tip/linux-2.6-tip.git;a=shortlog;h=rt/2.6.33
Information on the RT patch can be found at:
http://rt.wiki.kernel.org/index.php/Main_Page
To build the 2.6.33.5-rt23 tree, the following patches should be
applied:
http://www.kernel.org/pub/linux/kernel/v2.6/linux-2.6.33.5.tar.bz2
http://www.kernel.org/pub/linux/kernel/projects/rt/patch-2.6.33.5-rt23.bz2
Enjoy !
tglx
On Wed, 2010-06-09 at 18:44 +0200, Thomas Gleixner wrote:
> We are pleased to announce the next update to our new preempt-rt
> series.
>
<snip>
> Changes from 2.6.33.4-rt21 to 2.6.33.5-rt22:
>
> ef4afcc: Merge stable/linux-2.6.33.y into rt/2.6.33
>
> Changes from 2.6.33.5-rt22 to 2.6.33.5-rt23:
>
> ab774e8: drivers/hwmon/coretemp.c: get TjMax value from MSR
> e8bfcca: drivers/hwmon/coretemp.c: detect the thermal sensors by CPUID
> a06a867: dcache: Prevent d_genocide() from decrementing d_count more than once
> bdeadaf: nfs: Avoid igrab deadlock.
> 6976846: dcache: Fix select_parent dentry traversal locking
> 1f0f6ec: powerpc: chrp/time.c fix preprocessor conditional
I'm still on 2.6.33.5-rt22 on powerpc, and have run into a couple
"sleeping function called from invalid context..." BUGs. The first was
during system boot, second during a 'make'.
relevant code paths don't seem to have changed between -rt22 and -rt23,
so I think still valid.. :-)
This is on a POWER7 system, which may have opened up some timing
windows, wrt previous runs on POWER6 systems.
BUG: sleeping function called from invalid context at
kernel/rtmutex.c:684
pcnt: 1 0 in_atomic(): 1, irqs_disabled(): 1, pid: 3209, name: ip
Call Trace:
[c000000078cf28e0] [c000000000015550] .show_stack+0xd8/0x218
(unreliable)
[c000000078cf29c0] [c0000000006a99c8] .dump_stack+0x28/0x3c
[c000000078cf2a40] [c00000000006cea8] .__might_sleep+0x134/0x15c
[c000000078cf2ae0] [c0000000006a2f78] .rt_spin_lock+0x4c/0x9c
[c000000078cf2b70] [c0000000001799e8] ._slab_irq_disable+0x70/0xd8
[c000000078cf2c20] [c00000000017d1b4] .kmem_cache_alloc_node_notrace
+0x78/0x234
[c000000078cf2cf0] [c000000000131a04] .irq_to_desc_alloc_node+0xdc/0x2b8
[c000000078cf2db0] [c00000000000f238] .irq_setup_virq+0x44/0xe4
[c000000078cf2e50] [c00000000000f3f8] .irq_create_mapping+0x120/0x188
[c000000078cf2ef0] [c00000000002edfc] .ibmebus_request_irq+0x50/0xac
[c000000078cf2fa0] [c0000000004f3bd0] .ehea_up+0x314/0x69c
[c000000078cf30b0] [c0000000004f3fc8] .ehea_open+0x70/0x128
[c000000078cf3150] [c0000000005e504c] .dev_open+0x120/0x198
[c000000078cf31f0] [c0000000005e3ee8] .dev_change_flags+0x11c/0x21c
[c000000078cf32a0] [c0000000005f10ec] .do_setlink+0x2e0/0x400
[c000000078cf3380] [c0000000005f23a0] .rtnl_newlink+0x380/0x5b8
[c000000078cf3580] [c0000000005f1fd4] .rtnetlink_rcv_msg+0x258/0x2a4
[c000000078cf3630] [c000000000604094] .netlink_rcv_skb+0x84/0x120
[c000000078cf36d0] [c0000000005f1d58] .rtnetlink_rcv+0x38/0x5c
[c000000078cf3760] [c000000000603ba0] .netlink_unicast+0x31c/0x3fc
[c000000078cf3840] [c000000000604838] .netlink_sendmsg+0x2c8/0x30c
[c000000078cf3930] [c0000000005c8bdc] .sock_sendmsg+0xbc/0xfc
[c000000078cf3b10] [c0000000005c92fc] .SyS_sendmsg+0x20c/0x2a4
[c000000078cf3d50] [c0000000005c7f6c] .SyS_socketcall+0x238/0x2b0
[c000000078cf3e30] [c00000000000852c] syscall_exit+0x0/0x40
BUG: sleeping function called from invalid context at kernel/rtmutex.c:684
pcnt: 0 0 in_atomic(): 0, irqs_disabled(): 1, pid: 4760, name: make
Call Trace:
[c0000000792978b0] [c000000000015550] .show_stack+0xd8/0x218 (unreliable)
[c000000079297990] [c0000000006a99c8] .dump_stack+0x28/0x3c
[c000000079297a10] [c00000000006cea8] .__might_sleep+0x134/0x15c
[c000000079297ab0] [c0000000006a2f78] .rt_spin_lock+0x4c/0x9c
[c000000079297b40] [c0000000000a1ff0] .get_signal_to_deliver+0xcc/0x4ec
[c000000079297c50] [c000000000018604] .do_signal_pending.clone.0+0x6c/0x2e0
[c000000079297d90] [c0000000000188b0] .do_signal+0x38/0x74
[c000000079297e30] [c000000000008b40] user_work+0x24/0x28
>
> Bulk of changes comes from stable tree. The -rt related changes are
> mostly fixes for the vfs fallout the rest are fixlets all over the
> place and hardware enablement.
>
> Download locations:
>
> http://www.kernel.org/pub/linux/kernel/projects/rt/
>
> Git release branch:
> git://git.kernel.org/pub/scm/linux/kernel/git/tip/linux-2.6-tip.git rt/2.6.33
>
> Gitweb:
> http://git.kernel.org/?p=linux/kernel/git/tip/linux-2.6-tip.git;a=shortlog;h=rt/2.6.33
>
> Information on the RT patch can be found at:
>
> http://rt.wiki.kernel.org/index.php/Main_Page
>
> To build the 2.6.33.5-rt23 tree, the following patches should be
> applied:
>
> http://www.kernel.org/pub/linux/kernel/v2.6/linux-2.6.33.5.tar.bz2
> http://www.kernel.org/pub/linux/kernel/projects/rt/patch-2.6.33.5-rt23.bz2
>
> Enjoy !
>
> tglx
>
> --
> To unsubscribe from this list: send the line "unsubscribe linux-rt-users" in
> the body of a message to [email protected]
> More majordomo info at http://vger.kernel.org/majordomo-info.html
On Wed, 9 Jun 2010, Will Schmidt wrote:
> On Wed, 2010-06-09 at 18:44 +0200, Thomas Gleixner wrote:
> I'm still on 2.6.33.5-rt22 on powerpc, and have run into a couple
> "sleeping function called from invalid context..." BUGs. The first was
> during system boot, second during a 'make'.
>
> relevant code paths don't seem to have changed between -rt22 and -rt23,
> so I think still valid.. :-)
>
> This is on a POWER7 system, which may have opened up some timing
> windows, wrt previous runs on POWER6 systems.
>
>
> BUG: sleeping function called from invalid context at
> kernel/rtmutex.c:684
> [c000000078cf2cf0] [c000000000131a04] .irq_to_desc_alloc_node+0xdc/0x2b8
Grr. I thought I had the SPARSE_IRQ stuff disabled on -rt, but seems
it came back somehow. Can you disable that, or does you machine depend
on it ?
> BUG: sleeping function called from invalid context at kernel/rtmutex.c:684
> pcnt: 0 0 in_atomic(): 0, irqs_disabled(): 1, pid: 4760, name: make
> Call Trace:
> [c0000000792978b0] [c000000000015550] .show_stack+0xd8/0x218 (unreliable)
> [c000000079297990] [c0000000006a99c8] .dump_stack+0x28/0x3c
> [c000000079297a10] [c00000000006cea8] .__might_sleep+0x134/0x15c
> [c000000079297ab0] [c0000000006a2f78] .rt_spin_lock+0x4c/0x9c
> [c000000079297b40] [c0000000000a1ff0] .get_signal_to_deliver+0xcc/0x4ec
> [c000000079297c50] [c000000000018604] .do_signal_pending.clone.0+0x6c/0x2e0
> [c000000079297d90] [c0000000000188b0] .do_signal+0x38/0x74
> [c000000079297e30] [c000000000008b40] user_work+0x24/0x28
Does the patch below fix it ?
Thanks,
tglx
-----
diff --git a/arch/powerpc/kernel/signal.c b/arch/powerpc/kernel/signal.c
index 00b5078..fa580f9 100644
--- a/arch/powerpc/kernel/signal.c
+++ b/arch/powerpc/kernel/signal.c
@@ -120,6 +120,14 @@ static int do_signal_pending(sigset_t *oldset, struct pt_regs *regs)
int ret;
int is32 = is_32bit_task();
+#ifdef CONFIG_PREEMPT_RT
+ /*
+ * Fully-preemptible kernel does not need interrupts disabled:
+ */
+ local_irq_enable();
+ preempt_check_resched();
+#endif
+
if (current_thread_info()->local_flags & _TLF_RESTORE_SIGMASK)
oldset = ¤t->saved_sigmask;
else if (!oldset)
On Thu, 2010-06-10 at 08:07 +0200, Thomas Gleixner wrote:
> On Wed, 9 Jun 2010, Will Schmidt wrote:
>
> > On Wed, 2010-06-09 at 18:44 +0200, Thomas Gleixner wrote:
> > I'm still on 2.6.33.5-rt22 on powerpc, and have run into a couple
> > "sleeping function called from invalid context..." BUGs. The first was
> > during system boot, second during a 'make'.
> >
> > relevant code paths don't seem to have changed between -rt22 and -rt23,
> > so I think still valid.. :-)
> >
> > This is on a POWER7 system, which may have opened up some timing
> > windows, wrt previous runs on POWER6 systems.
> >
> >
> > BUG: sleeping function called from invalid context at
> > kernel/rtmutex.c:684
>
> > [c000000078cf2cf0] [c000000000131a04] .irq_to_desc_alloc_node+0xdc/0x2b8
>
> Grr. I thought I had the SPARSE_IRQ stuff disabled on -rt, but seems
> it came back somehow. Can you disable that, or does you machine depend
> on it ?
I'm not sure whether it is required by the architecture. I do see that
SPARSE_IRQ is disabled in arch/x86/Kconfig with a !PREEMPT_RT, but I
don't see an equivalent 'disable' in arch/powerpc/Kconfig.
> > BUG: sleeping function called from invalid context at kernel/rtmutex.c:684
> > pcnt: 0 0 in_atomic(): 0, irqs_disabled(): 1, pid: 4760, name: make
> > Call Trace:
> > [c0000000792978b0] [c000000000015550] .show_stack+0xd8/0x218 (unreliable)
> > [c000000079297990] [c0000000006a99c8] .dump_stack+0x28/0x3c
> > [c000000079297a10] [c00000000006cea8] .__might_sleep+0x134/0x15c
> > [c000000079297ab0] [c0000000006a2f78] .rt_spin_lock+0x4c/0x9c
> > [c000000079297b40] [c0000000000a1ff0] .get_signal_to_deliver+0xcc/0x4ec
> > [c000000079297c50] [c000000000018604] .do_signal_pending.clone.0+0x6c/0x2e0
> > [c000000079297d90] [c0000000000188b0] .do_signal+0x38/0x74
> > [c000000079297e30] [c000000000008b40] user_work+0x24/0x28
>
> Does the patch below fix it ?
Probably.. :-) I'm running with the patch now, and no repeats of that
BUG yet.
Thanks,
-Will
>
> Thanks,
>
> tglx
> -----
> diff --git a/arch/powerpc/kernel/signal.c b/arch/powerpc/kernel/signal.c
> index 00b5078..fa580f9 100644
> --- a/arch/powerpc/kernel/signal.c
> +++ b/arch/powerpc/kernel/signal.c
> @@ -120,6 +120,14 @@ static int do_signal_pending(sigset_t *oldset, struct pt_regs *regs)
> int ret;
> int is32 = is_32bit_task();
>
> +#ifdef CONFIG_PREEMPT_RT
> + /*
> + * Fully-preemptible kernel does not need interrupts disabled:
> + */
> + local_irq_enable();
> + preempt_check_resched();
> +#endif
> +
> if (current_thread_info()->local_flags & _TLF_RESTORE_SIGMASK)
> oldset = ¤t->saved_sigmask;
> else if (!oldset)
On Thu, 2010-06-10 at 08:56 -0700, Will Schmidt wrote:
> On Thu, 2010-06-10 at 08:07 +0200, Thomas Gleixner wrote:
> > On Wed, 9 Jun 2010, Will Schmidt wrote:
> >
> > > On Wed, 2010-06-09 at 18:44 +0200, Thomas Gleixner wrote:
> > > I'm still on 2.6.33.5-rt22 on powerpc, and have run into a couple
> > > "sleeping function called from invalid context..." BUGs. The
> first was
> > > during system boot, second during a 'make'.
> > >
> > > relevant code paths don't seem to have changed between -rt22 and
> -rt23,
> > > so I think still valid.. :-)
> > >
> > > This is on a POWER7 system, which may have opened up some timing
> > > windows, wrt previous runs on POWER6 systems.
> > >
> > >
> > > BUG: sleeping function called from invalid context at
> > > kernel/rtmutex.c:684
> >
> > > [c000000078cf2cf0] [c000000000131a04] .irq_to_desc_alloc_node
> +0xdc/0x2b8
> >
> > Grr. I thought I had the SPARSE_IRQ stuff disabled on -rt, but seems
> > it came back somehow. Can you disable that, or does you machine
> depend
> > on it ?
>
> I'm not sure whether it is required by the architecture. I do see
> that
> SPARSE_IRQ is disabled in arch/x86/Kconfig with a !PREEMPT_RT, but I
> don't see an equivalent 'disable' in arch/powerpc/Kconfig.
>
I see it as "y" in my .config for a powerpc 405 build. This appears to
run on my system (although I haven't loaded up the rt23 build on the
target yet - last one loaded up was rt22).
What should I expect to see different if SPARSE_IRQ is turned off?
>
> > > BUG: sleeping function called from invalid context at
> kernel/rtmutex.c:684
> > > pcnt: 0 0 in_atomic(): 0, irqs_disabled(): 1, pid: 4760, name:
> make
> > > Call Trace:
> > > [c0000000792978b0] [c000000000015550] .show_stack+0xd8/0x218
> (unreliable)
> > > [c000000079297990] [c0000000006a99c8] .dump_stack+0x28/0x3c
> > > [c000000079297a10] [c00000000006cea8] .__might_sleep+0x134/0x15c
> > > [c000000079297ab0] [c0000000006a2f78] .rt_spin_lock+0x4c/0x9c
> > > [c000000079297b40] [c0000000000a1ff0] .get_signal_to_deliver
> +0xcc/0x4ec
> > > [c000000079297c50] [c000000000018604] .do_signal_pending.clone.0
> +0x6c/0x2e0
> > > [c000000079297d90] [c0000000000188b0] .do_signal+0x38/0x74
> > > [c000000079297e30] [c000000000008b40] user_work+0x24/0x28
> >
> > Does the patch below fix it ?
>
> Probably.. :-) I'm running with the patch now, and no repeats of
> that
> BUG yet.
>
> Thanks,
> -Will
>
>
> >
> > Thanks,
> >
> > tglx
> > -----
> > diff --git a/arch/powerpc/kernel/signal.c
> b/arch/powerpc/kernel/signal.c
> > index 00b5078..fa580f9 100644
> > --- a/arch/powerpc/kernel/signal.c
> > +++ b/arch/powerpc/kernel/signal.c
> > @@ -120,6 +120,14 @@ static int do_signal_pending(sigset_t *oldset,
> struct pt_regs *regs)
> > int ret;
> > int is32 = is_32bit_task();
> >
> > +#ifdef CONFIG_PREEMPT_RT
> > + /*
> > + * Fully-preemptible kernel does not need interrupts disabled:
> > + */
> > + local_irq_enable();
> > + preempt_check_resched();
> > +#endif
> > +
> > if (current_thread_info()->local_flags & _TLF_RESTORE_SIGMASK)
> > oldset = ¤t->saved_sigmask;
> > else if (!oldset)
>
>
> --
> To unsubscribe from this list: send the line "unsubscribe
> linux-rt-users" in
> the body of a message to [email protected]
> More majordomo info at http://vger.kernel.org/majordomo-info.html
>
>
On Thu, 10 Jun 2010, Darcy L. Watkins wrote:
>
>
> On Thu, 2010-06-10 at 08:56 -0700, Will Schmidt wrote:
> > On Thu, 2010-06-10 at 08:07 +0200, Thomas Gleixner wrote:
> > > On Wed, 9 Jun 2010, Will Schmidt wrote:
> > >
> > > > On Wed, 2010-06-09 at 18:44 +0200, Thomas Gleixner wrote:
> > > > I'm still on 2.6.33.5-rt22 on powerpc, and have run into a couple
> > > > "sleeping function called from invalid context..." BUGs. The
> > first was
> > > > during system boot, second during a 'make'.
> > > >
> > > > relevant code paths don't seem to have changed between -rt22 and
> > -rt23,
> > > > so I think still valid.. :-)
> > > >
> > > > This is on a POWER7 system, which may have opened up some timing
> > > > windows, wrt previous runs on POWER6 systems.
> > > >
> > > >
> > > > BUG: sleeping function called from invalid context at
> > > > kernel/rtmutex.c:684
> > >
> > > > [c000000078cf2cf0] [c000000000131a04] .irq_to_desc_alloc_node
> > +0xdc/0x2b8
> > >
> > > Grr. I thought I had the SPARSE_IRQ stuff disabled on -rt, but seems
> > > it came back somehow. Can you disable that, or does you machine
> > depend
> > > on it ?
> >
> > I'm not sure whether it is required by the architecture. I do see
> > that
> > SPARSE_IRQ is disabled in arch/x86/Kconfig with a !PREEMPT_RT, but I
> > don't see an equivalent 'disable' in arch/powerpc/Kconfig.
> >
>
> I see it as "y" in my .config for a powerpc 405 build. This appears to
> run on my system (although I haven't loaded up the rt23 build on the
> target yet - last one loaded up was rt22).
>
> What should I expect to see different if SPARSE_IRQ is turned off?
The difference is that you can trigger "scheduling while atomic". The
warning Will has seen depends on debug options.
Thanks,
tglx
On Thu, 10 Jun 2010, Will Schmidt wrote:
> On Thu, 2010-06-10 at 08:07 +0200, Thomas Gleixner wrote:
> > On Wed, 9 Jun 2010, Will Schmidt wrote:
> >
> > > On Wed, 2010-06-09 at 18:44 +0200, Thomas Gleixner wrote:
> > > I'm still on 2.6.33.5-rt22 on powerpc, and have run into a couple
> > > "sleeping function called from invalid context..." BUGs. The first was
> > > during system boot, second during a 'make'.
> > >
> > > relevant code paths don't seem to have changed between -rt22 and -rt23,
> > > so I think still valid.. :-)
> > >
> > > This is on a POWER7 system, which may have opened up some timing
> > > windows, wrt previous runs on POWER6 systems.
> > >
> > >
> > > BUG: sleeping function called from invalid context at
> > > kernel/rtmutex.c:684
> >
> > > [c000000078cf2cf0] [c000000000131a04] .irq_to_desc_alloc_node+0xdc/0x2b8
> >
> > Grr. I thought I had the SPARSE_IRQ stuff disabled on -rt, but seems
> > it came back somehow. Can you disable that, or does you machine depend
> > on it ?
>
> I'm not sure whether it is required by the architecture. I do see that
> SPARSE_IRQ is disabled in arch/x86/Kconfig with a !PREEMPT_RT, but I
> don't see an equivalent 'disable' in arch/powerpc/Kconfig.
Can you disable it ?
Thanks,
tglx
On Mon, 2010-06-14 at 15:34 +0200, Thomas Gleixner wrote:
> On Thu, 10 Jun 2010, Will Schmidt wrote:
>
> > On Thu, 2010-06-10 at 08:07 +0200, Thomas Gleixner wrote:
> > > On Wed, 9 Jun 2010, Will Schmidt wrote:
> > >
> > > > On Wed, 2010-06-09 at 18:44 +0200, Thomas Gleixner wrote:
> > > > I'm still on 2.6.33.5-rt22 on powerpc, and have run into a couple
> > > > "sleeping function called from invalid context..." BUGs. The first was
> > > > during system boot, second during a 'make'.
> > > >
> > > > relevant code paths don't seem to have changed between -rt22 and -rt23,
> > > > so I think still valid.. :-)
> > > >
> > > > This is on a POWER7 system, which may have opened up some timing
> > > > windows, wrt previous runs on POWER6 systems.
> > > >
> > > >
> > > > BUG: sleeping function called from invalid context at
> > > > kernel/rtmutex.c:684
> > >
> > > > [c000000078cf2cf0] [c000000000131a04] .irq_to_desc_alloc_node+0xdc/0x2b8
> > >
> > > Grr. I thought I had the SPARSE_IRQ stuff disabled on -rt, but seems
> > > it came back somehow. Can you disable that, or does you machine depend
> > > on it ?
> >
> > I'm not sure whether it is required by the architecture. I do see that
> > SPARSE_IRQ is disabled in arch/x86/Kconfig with a !PREEMPT_RT, but I
> > don't see an equivalent 'disable' in arch/powerpc/Kconfig.
>
> Can you disable it ?
Yeah.
[root@foo ~]# zcat /proc/config.gz | grep SPARSE
# CONFIG_SPARSE_IRQ is not set
...
Thats a sample size of one (POWER7) system, but seems to have worked.
Thanks,
-Will
>
> Thanks,
>
> tglx
We are pleased to announce the next update to our new preempt-rt
series.
Changes from 2.6.33.5-rt23 to 2.6.33.5-rt24:
42edbcb: gianfar: Fix TX ring processing on SMP machines
faf3581: timer_stats: Convert table_lock to raw_spin_lock
654e404: hvc_console: Fix race between hvc_close and hvc_remove
09fd768: hvc_console: Fix race between hvc_close and hvc_remove
ea2d2a0: perf: Fix errors path in perf_output_begin()
60bd946: KEYS: find_keyring_by_name() can gain access to a freed keyring
b5ac516: tracing: Update the comm field in the right variable in update_max_tr
77cb7c1: lockdep: Reduce stack_trace usage
cee9ee0: powerpc: Disable sparse irq for RT
c7dec16: powerpc: Enable interrupts in do_signal()
Changes from 2.6.33.5-rt24 to 2.6.33.5-rt25:
ec646ea: vfs: Revert the scalability patches
Changes from 2.6.33.5-rt26 to 2.6.33.6-rt26:
7e1af11: Merge git://git.kernel.org/pub/scm/linux/kernel/git/stable/linux-2.6.33.y
Bulk of changes comes from the stable tree and from the big revert of
the VFS scalability patches. They looked promising, but the nasty
details bit us hard. We'll revisit them later.
There are some pending issues:
- rtc_lock suspend/resume (working on a patch)
- futex requeue_pi (patch available, needs testing/confirmation)
- powerpc64 oddities (patch available, needs rework)
Download locations:
http://www.kernel.org/pub/linux/kernel/projects/rt/
Git release branch:
git://git.kernel.org/pub/scm/linux/kernel/git/tip/linux-2.6-tip.git rt/2.6.33
Gitweb:
http://git.kernel.org/?p=linux/kernel/git/tip/linux-2.6-tip.git;a=shortlog;h=rt/2.6.33
Information on the RT patch can be found at:
http://rt.wiki.kernel.org/index.php/Main_Page
To build the 2.6.33.6-rt26 tree, the following patches should be
applied:
http://www.kernel.org/pub/linux/kernel/v2.6/linux-2.6.33.6.tar.bz2
http://www.kernel.org/pub/linux/kernel/projects/rt/patch-2.6.33.6-rt26.bz2
Enjoy !
tglx
Hi,
> Enjoy !
>
I hit this with rt26. (Quite possible is on vanilla as well, will
build a kernel soemtime soon and try)
=======================================================
[ INFO: possible circular locking dependency detected ]
2.6.33.6-rt26 #36
-------------------------------------------------------
iperf/1629 is trying to acquire lock:
(slock-AF_INET){+.+...}, at: [<ffffffff8135a4fa>] release_sock+0x30/0x108
but task is already holding lock:
(&lock->lock){+.+...}, at: [<ffffffff813cd6cc>] ipt_do_table+0x118/0x650
which lock already depends on the new lock.
the existing dependency chain (in reverse order) is:
-> #1 (&lock->lock){+.+...}:
[<ffffffff810697e7>] lock_acquire+0xd6/0xfa
[<ffffffff813fcfaa>] rt_spin_lock+0x31/0x38
[<ffffffffa00ba8ef>] nf_nat_get_offset+0x42/0x6b [nf_nat]
[<ffffffff8138b2ba>] tcp_packet+0x659/0xd5b
[<ffffffff81387eb9>] nf_conntrack_in+0x445/0x4fc
[<ffffffff813ca933>] ipv4_conntrack_local+0x3b/0x44
[<ffffffff81384c22>] nf_iterate+0x41/0x84
[<ffffffff81384cd8>] nf_hook_slow+0x73/0xdb
[<ffffffff813970b4>] nf_hook_thresh.clone.0+0x3c/0x45
[<ffffffff81397136>] __ip_local_out+0x79/0x7b
[<ffffffff81397149>] ip_local_out+0x11/0x22
[<ffffffff81397407>] ip_push_pending_frames+0x2ad/0x30b
[<ffffffff81398a41>] ip_send_reply+0x21a/0x259
[<ffffffff813acdf1>] tcp_v4_send_reset+0x1c8/0x1eb
[<ffffffff813b0367>] tcp_v4_rcv+0x500/0x661
[<ffffffff8139400a>] ip_local_deliver_finish+0x14d/0x1c3
[<ffffffff813940f2>] ip_local_deliver+0x72/0x7c
[<ffffffff81393ac3>] ip_rcv_finish+0x31a/0x346
[<ffffffff81393d29>] ip_rcv+0x23a/0x268
[<ffffffff81366862>] netif_receive_skb+0x3fa/0x424
[<ffffffffa023fb41>] br_handle_frame_finish+0x105/0x13d [bridge]
[<ffffffffa023fd08>] br_handle_frame+0x18f/0x1a7 [bridge]
[<ffffffff81366792>] netif_receive_skb+0x32a/0x424
[<ffffffff813669d4>] napi_skb_finish+0x24/0x3b
[<ffffffff81366e83>] napi_gro_receive+0x2a/0x2f
[<ffffffff812d03b9>] sky2_poll+0x86e/0xaef
[<ffffffff81366fcc>] net_rx_action+0xa3/0x1f1
[<ffffffff81044a02>] run_ksoftirqd+0x14e/0x254
[<ffffffff8105975f>] kthread+0x88/0x90
[<ffffffff81009844>] kernel_thread_helper+0x4/0x10
-> #0 (slock-AF_INET){+.+...}:
[<ffffffff8106941a>] __lock_acquire+0x9f1/0xce8
[<ffffffff810697e7>] lock_acquire+0xd6/0xfa
[<ffffffff813fcfaa>] rt_spin_lock+0x31/0x38
[<ffffffff8135a4fa>] release_sock+0x30/0x108
[<ffffffff813a012a>] tcp_recvmsg+0x8b8/0x9a3
[<ffffffff81359864>] sock_common_recvmsg+0x37/0x48
[<ffffffff81357178>] __sock_recvmsg+0x6d/0x78
(Just run iperf as server and test it from another system)
thanks,
Dhaval
Hi Thomas,
Just want to know when there will be released RT patch for 2.6.34? We
have some tasks to patch RT to our local 2.6.34 tree.
Thanks
Barry
On Tue, Jul 13, 2010 at 11:54 PM, Thomas Gleixner <[email protected]> wrote:
> We are pleased to announce the next update to our new preempt-rt
> series.
>
> Changes from 2.6.33.5-rt23 to 2.6.33.5-rt24:
>
> 42edbcb: gianfar: Fix TX ring processing on SMP machines
> faf3581: timer_stats: Convert table_lock to raw_spin_lock
> 654e404: hvc_console: Fix race between hvc_close and hvc_remove
> 09fd768: hvc_console: Fix race between hvc_close and hvc_remove
> ea2d2a0: perf: Fix errors path in perf_output_begin()
> 60bd946: KEYS: find_keyring_by_name() can gain access to a freed keyring
> b5ac516: tracing: Update the comm field in the right variable in update_max_tr
> 77cb7c1: lockdep: Reduce stack_trace usage
> cee9ee0: powerpc: Disable sparse irq for RT
> c7dec16: powerpc: Enable interrupts in do_signal()
>
> Changes from 2.6.33.5-rt24 to 2.6.33.5-rt25:
>
> ec646ea: vfs: Revert the scalability patches
>
> Changes from 2.6.33.5-rt26 to 2.6.33.6-rt26:
>
> 7e1af11: Merge git://git.kernel.org/pub/scm/linux/kernel/git/stable/linux-2.6.33.y
>
> Bulk of changes comes from the stable tree and from the big revert of
> the VFS scalability patches. They looked promising, but the nasty
> details bit us hard. We'll revisit them later.
>
> There are some pending issues:
> - rtc_lock suspend/resume (working on a patch)
> - futex requeue_pi (patch available, needs testing/confirmation)
> - powerpc64 oddities (patch available, needs rework)
>
> Download locations:
>
> http://www.kernel.org/pub/linux/kernel/projects/rt/
>
> Git release branch:
> git://git.kernel.org/pub/scm/linux/kernel/git/tip/linux-2.6-tip.git rt/2.6.33
>
> Gitweb:
> http://git.kernel.org/?p=linux/kernel/git/tip/linux-2.6-tip.git;a=shortlog;h=rt/2.6.33
>
> Information on the RT patch can be found at:
>
> http://rt.wiki.kernel.org/index.php/Main_Page
>
> To build the 2.6.33.6-rt26 tree, the following patches should be
> applied:
>
> http://www.kernel.org/pub/linux/kernel/v2.6/linux-2.6.33.6.tar.bz2
> http://www.kernel.org/pub/linux/kernel/projects/rt/patch-2.6.33.6-rt26.bz2
>
> Enjoy !
>
> tglx
> --
> To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
> the body of a message to [email protected]
> More majordomo info at http://vger.kernel.org/majordomo-info.html
> Please read the FAQ at http://www.tux.org/lkml/
>
We are pleased to announce the next update to our new preempt-rt
series.
Changes from 2.6.33.6-rt26 to 2.6.33.6-rt27:
d2decbf: v2.6.33.6-rt27
d972e60: UIO: Add a driver for Hilscher netX-based fieldbus cards
911d28c: dca: Fix fallout from raw_spinlock conversion
59fdcc0: powerpc: cpu-hotplug: Prevent softirq wakeup on wrong CPU
5805f5a: drivers/dca: Convert dca_lock to a raw spinlock
96c5397: cpu-hotplug: Prevent softirq wakeup on wrong CPU
8e7dbdf: cpu-hotplug: Don't wake up the desched thread from idle_task_exit()
5706af4: x86: mce: Convert cmci_discover_lock to raw_spinlock
5bbbedc: net: iptables: Fix xt_info locking
60f778f: suspend: Prevent might sleep splats
19491de: futex: Protect against pi_blocked_on corruption during requeue PI
d69f558: sched: Fix TASK_WAKING vs fork deadlock
111abc8: sched: Make select_fallback_rq() cpuset friendly
bb8eef7: sched: _cpu_down(): Don't play with current->cpus_allowed
a325588: sched: sched_exec(): Remove the select_fallback_rq() logic
2d5a7c7: sched: move_task_off_dead_cpu(): Remove retry logic
ae35e6f: sched: move_task_off_dead_cpu(): Take rq->lock around select_fallback_rq()
a1e0f4a: sched: Kill the broken and deadlockable cpuset_lock/cpuset_cpus_allowed_locked code
dd56c23: sched: set_cpus_allowed_ptr(): Don't use rq->migration_thread after unlock
Download locations:
http://www.kernel.org/pub/linux/kernel/projects/rt/
Git release branch:
git://git.kernel.org/pub/scm/linux/kernel/git/tip/linux-2.6-tip.git rt/2.6.33
Gitweb:
http://git.kernel.org/?p=linux/kernel/git/tip/linux-2.6-tip.git;a=shortlog;h=rt/2.6.33
Information on the RT patch can be found at:
http://rt.wiki.kernel.org/index.php/Main_Page
To build the 2.6.33.6-rt27 tree, the following patches should be
applied:
http://www.kernel.org/pub/linux/kernel/v2.6/linux-2.6.33.6.tar.bz2
http://www.kernel.org/pub/linux/kernel/projects/rt/patch-2.6.33.6-rt27.bz2
Enjoy !
tglx
We are pleased to announce the next update to our new preempt-rt
series.
Changes from 2.6.33.6-rt28 to 2.6.33.7-rt29:
c88417f: Merge stable update
Changes from 2.6.33.6-rt27 to 2.6.33.6-rt28:
8a990ab: printk: Restore irqs before calling release_console_mutex()
Download locations:
http://www.kernel.org/pub/linux/kernel/projects/rt/
Git release branch:
git://git.kernel.org/pub/scm/linux/kernel/git/tip/linux-2.6-tip.git rt/2.6.33
Gitweb:
http://git.kernel.org/?p=linux/kernel/git/tip/linux-2.6-tip.git;a=shortlog;h=rt/2.6.33
Information on the RT patch can be found at:
http://rt.wiki.kernel.org/index.php/Main_Page
To build the 2.6.33.7-rt29 tree, the following patches should be
applied:
http://www.kernel.org/pub/linux/kernel/v2.6/linux-2.6.33.7.tar.bz2
http://www.kernel.org/pub/linux/kernel/projects/rt/patch-2.6.33.7-rt29.bz2
Enjoy !
tglx
We are pleased to announce the next update to our new preempt-rt
series.
Changes from 2.6.33.7-rt29 to 2.6.33.7.2-rt30:
5fed714: x86: stackprotector: Avoid random pool on rt
76bdf0c: x86: Convert mce timer to hrtimer
a1026af: net: ebtables: Work around per cpu assumptions
cfaca74: tracing: Add combined timer and wakeup latency histogram
e94445b: rt: mm: fix kthreadd livelock in drain_local_pages_work
a09b02c: powerpc/fsl-booke: Move loadcam_entry back to asm code to fix SMP ftrace
a8b8812: AT91: SAM9G45 - add a separate clock entry for every single TC block
3cf7bf5: mm: make stack guard page logic use vm_prev pointer
901f364: mm: make the mlock() stack guard page checks stricter
498f700: mm: make the vma list be doubly linked
e276f93c: mm: fix up some user-visible effects of the stack guard page
a487cf3: mm: fix page table unmap for stack guard page properly
27ed5f7: mm: fix missing page table unmap for stack guard page failure case
ed313b5: mm: keep a guard page below a grow-down stack segment
bf18be0: drm: block userspace under allocating buffer and having drivers overwrite it (v2)
3609ee7: x86-64, compat: Retruncate rax after ia32 syscall entry tracing
b2104d5: x86-64, compat: Test %rax for the syscall number, not %eax
c2841fe: compat: Make compat_alloc_user_space() incorporate the access_ok()
85b5abf: ALSA: seq/oss - Fix double-free at error path of snd_seq_oss_open()
5741fe4: wireless extensions: fix kernel heap content leak
49aa1d9: net sched: fix some kernel memory leaks
cd92f99: act_nat: use stack variable
dc7dde7: tracing: Do not allow llseek to set_ftrace_filter
2da92b6: x86: Limit 32bit PAE memory to 16G
1389ed5: niu: Fix kernel buffer overflow for ETHTOOL_GRXCLSRLALL
4d4ea8b: KEYS: Fix bug in keyctl_session_to_parent() if parent has no session keyring
300f0c0: KEYS: Fix RCU no-lock warning in keyctl_session_to_parent()
300bc74: can: add limit for nframes and clean up signed/unsigned variables
88df886: NFS: Fix an Oops in the NFSv4 atomic open code
00b073d: staging/comedi: Fix type in semaphore conversion
327735e: perf: Provide kmap_atomic_direct for !CONFIG_HIGHMEM
ddeca62: perf: Use kmap_atomic_direct
d07d5cf: v2.6.33.7.1-rt29
e6c0211: ssb: Handle alternate SSPROM location
0bc60c2: ssb: fix NULL ptr deref when pcihost_wrapper is used
9ab4cde: ssb: Look for SPROM at different offset on higher rev CC
314beea: ssb: do not read SPROM if it does not exist
4ed85ce: Revert "ssb: Handle Netbook devices where the SPROM address is changed"
d4202e4: drm/i915: Check overlay stride errata for i830 and i845
60853cb: KVM: MMU: flush remote tlbs when overwriting spte with different pfn
fbbddb2: mac80211: avoid scheduling while atomic in mesh_rx_plink_frame
8d5c466: xen: drop xen_sched_clock in favour of using plain wallclock time
e4de451: CIFS: Fix compile error with __init in cifs_init_dns_resolver() definition
a9229e0: CIFS: Remove __exit mark from cifs_exit_dns_resolver()
ab006e3: slow-work: use get_ref wrapper instead of directly calling get_ref
7485c1e: GFS2: rename causes kernel Oops
432cfce: SCSI: enclosure: fix error path - actually return ERR_PTR() on error
127524e: xfs: prevent swapext from operating on write-only files
ab1b2be: cfg80211: don't get expired BSSes
d377614: cfg80211: ignore spurious deauth
efce9f4: iwlwifi: fix scan abort
8dc20bd: ath9k: fix yet another buffer leak in the tx aggregation code
9fd8aa6: ath9k: fix TSF after reset on AR913x
2089c7f: ath9k_hw: fix an off-by-one error in the PDADC boundaries calculation
e10e4ec: ath9k: enable serialize_regmode for non-PCIE AR9160
cad9c43: 9p: strlen() doesn't count the terminator
d2f1f93: e1000e: 82577/82578 PHY register access issues
8ed2a49: e1000e: don't inadvertently re-set INTX_DISABLE
257a4e8: amd64_edac: Correct scrub rate setting
3f5500c: amd64_edac: Fix DCT base address selector
8c0b0d6: parisc: pass through '\t' to early (iodc) console
d7e94ae: arm/imx/gpio: add spinlock protection
42aa826: PARISC: led.c - fix potential stack overflow in led_proc_write()
49337d1: NFS: kswapd must not block in nfs_release_page
7486bb3: comedi: Uncripple 8255-based DIO subdevices
6388252: mm: fix ia64 crash when gcore reads gate area
This is a rather large update with cherry-picks from the stable tree
and quite a few fixes for long standing problems in the rt patch.
Download locations:
http://www.kernel.org/pub/linux/kernel/projects/rt/
Git release branch:
git://git.kernel.org/pub/scm/linux/kernel/git/tip/linux-2.6-tip.git rt/2.6.33
Gitweb:
http://git.kernel.org/?p=linux/kernel/git/tip/linux-2.6-tip.git;a=shortlog;h=rt/2.6.33
Information on the RT patch can be found at:
http://rt.wiki.kernel.org/index.php/Main_Page
To build the 2.6.33.7.2-rt30 tree, the following patches should be
applied:
http://www.kernel.org/pub/linux/kernel/v2.6/linux-2.6.33.7.tar.bz2
http://www.kernel.org/pub/linux/kernel/projects/rt/patch-2.6.33.7.2-rt30.bz2
Enjoy !
This is probably the last update to the 33-rt series.
The next release will be based on 2.6.37, but I'm still in the middle
of a major overhaul of key compononents of the patch. This overhaul is
necessary to make progress in mainlining the missing bits and
pieces. I hope to have it ready not too long after the 2.6.37 mainline
release, but no promise.
At this point I want to say thanks to everyone who supported the
development of the -rt patches with testing, bugfixes, stable
backports etc. in the last years. I hope that I can rely on your help
for the upcoming 37 release and the ongoing effort to merge the patch
fully into the mainline.
I wish you all a Merry Christmas and a Happy New Year!
Thomas
@Thomas et al.:
Thank you so much for your work! It's really a amazing.
Merry Christmas!
Dennis
Thomas Gleixner schrieb:
> We are pleased to announce the next update to our new preempt-rt
> series.
>
> Changes from 2.6.33.7-rt29 to 2.6.33.7.2-rt30:
>
> 5fed714: x86: stackprotector: Avoid random pool on rt
> 76bdf0c: x86: Convert mce timer to hrtimer
> a1026af: net: ebtables: Work around per cpu assumptions
> cfaca74: tracing: Add combined timer and wakeup latency histogram
> e94445b: rt: mm: fix kthreadd livelock in drain_local_pages_work
> a09b02c: powerpc/fsl-booke: Move loadcam_entry back to asm code to fix SMP ftrace
> a8b8812: AT91: SAM9G45 - add a separate clock entry for every single TC block
> 3cf7bf5: mm: make stack guard page logic use vm_prev pointer
> 901f364: mm: make the mlock() stack guard page checks stricter
> 498f700: mm: make the vma list be doubly linked
> e276f93c: mm: fix up some user-visible effects of the stack guard page
> a487cf3: mm: fix page table unmap for stack guard page properly
> 27ed5f7: mm: fix missing page table unmap for stack guard page failure case
> ed313b5: mm: keep a guard page below a grow-down stack segment
> bf18be0: drm: block userspace under allocating buffer and having drivers overwrite it (v2)
> 3609ee7: x86-64, compat: Retruncate rax after ia32 syscall entry tracing
> b2104d5: x86-64, compat: Test %rax for the syscall number, not %eax
> c2841fe: compat: Make compat_alloc_user_space() incorporate the access_ok()
> 85b5abf: ALSA: seq/oss - Fix double-free at error path of snd_seq_oss_open()
> 5741fe4: wireless extensions: fix kernel heap content leak
> 49aa1d9: net sched: fix some kernel memory leaks
> cd92f99: act_nat: use stack variable
> dc7dde7: tracing: Do not allow llseek to set_ftrace_filter
> 2da92b6: x86: Limit 32bit PAE memory to 16G
> 1389ed5: niu: Fix kernel buffer overflow for ETHTOOL_GRXCLSRLALL
> 4d4ea8b: KEYS: Fix bug in keyctl_session_to_parent() if parent has no session keyring
> 300f0c0: KEYS: Fix RCU no-lock warning in keyctl_session_to_parent()
> 300bc74: can: add limit for nframes and clean up signed/unsigned variables
> 88df886: NFS: Fix an Oops in the NFSv4 atomic open code
> 00b073d: staging/comedi: Fix type in semaphore conversion
> 327735e: perf: Provide kmap_atomic_direct for !CONFIG_HIGHMEM
> ddeca62: perf: Use kmap_atomic_direct
> d07d5cf: v2.6.33.7.1-rt29
> e6c0211: ssb: Handle alternate SSPROM location
> 0bc60c2: ssb: fix NULL ptr deref when pcihost_wrapper is used
> 9ab4cde: ssb: Look for SPROM at different offset on higher rev CC
> 314beea: ssb: do not read SPROM if it does not exist
> 4ed85ce: Revert "ssb: Handle Netbook devices where the SPROM address is changed"
> d4202e4: drm/i915: Check overlay stride errata for i830 and i845
> 60853cb: KVM: MMU: flush remote tlbs when overwriting spte with different pfn
> fbbddb2: mac80211: avoid scheduling while atomic in mesh_rx_plink_frame
> 8d5c466: xen: drop xen_sched_clock in favour of using plain wallclock time
> e4de451: CIFS: Fix compile error with __init in cifs_init_dns_resolver() definition
> a9229e0: CIFS: Remove __exit mark from cifs_exit_dns_resolver()
> ab006e3: slow-work: use get_ref wrapper instead of directly calling get_ref
> 7485c1e: GFS2: rename causes kernel Oops
> 432cfce: SCSI: enclosure: fix error path - actually return ERR_PTR() on error
> 127524e: xfs: prevent swapext from operating on write-only files
> ab1b2be: cfg80211: don't get expired BSSes
> d377614: cfg80211: ignore spurious deauth
> efce9f4: iwlwifi: fix scan abort
> 8dc20bd: ath9k: fix yet another buffer leak in the tx aggregation code
> 9fd8aa6: ath9k: fix TSF after reset on AR913x
> 2089c7f: ath9k_hw: fix an off-by-one error in the PDADC boundaries calculation
> e10e4ec: ath9k: enable serialize_regmode for non-PCIE AR9160
> cad9c43: 9p: strlen() doesn't count the terminator
> d2f1f93: e1000e: 82577/82578 PHY register access issues
> 8ed2a49: e1000e: don't inadvertently re-set INTX_DISABLE
> 257a4e8: amd64_edac: Correct scrub rate setting
> 3f5500c: amd64_edac: Fix DCT base address selector
> 8c0b0d6: parisc: pass through '\t' to early (iodc) console
> d7e94ae: arm/imx/gpio: add spinlock protection
> 42aa826: PARISC: led.c - fix potential stack overflow in led_proc_write()
> 49337d1: NFS: kswapd must not block in nfs_release_page
> 7486bb3: comedi: Uncripple 8255-based DIO subdevices
> 6388252: mm: fix ia64 crash when gcore reads gate area
>
> This is a rather large update with cherry-picks from the stable tree
> and quite a few fixes for long standing problems in the rt patch.
>
> Download locations:
>
> http://www.kernel.org/pub/linux/kernel/projects/rt/
>
> Git release branch:
> git://git.kernel.org/pub/scm/linux/kernel/git/tip/linux-2.6-tip.git rt/2.6.33
>
> Gitweb:
> http://git.kernel.org/?p=linux/kernel/git/tip/linux-2.6-tip.git;a=shortlog;h=rt/2.6.33
>
> Information on the RT patch can be found at:
>
> http://rt.wiki.kernel.org/index.php/Main_Page
>
> To build the 2.6.33.7.2-rt30 tree, the following patches should be
> applied:
>
> http://www.kernel.org/pub/linux/kernel/v2.6/linux-2.6.33.7.tar.bz2
> http://www.kernel.org/pub/linux/kernel/projects/rt/patch-2.6.33.7.2-rt30.bz2
>
> Enjoy !
>
> This is probably the last update to the 33-rt series.
>
> The next release will be based on 2.6.37, but I'm still in the middle
> of a major overhaul of key compononents of the patch. This overhaul is
> necessary to make progress in mainlining the missing bits and
> pieces. I hope to have it ready not too long after the 2.6.37 mainline
> release, but no promise.
>
> At this point I want to say thanks to everyone who supported the
> development of the -rt patches with testing, bugfixes, stable
> backports etc. in the last years. I hope that I can rely on your help
> for the upcoming 37 release and the ongoing effort to merge the patch
> fully into the mainline.
>
> I wish you all a Merry Christmas and a Happy New Year!
>
> Thomas
> --
> To unsubscribe from this list: send the line "unsubscribe linux-rt-users" in
> the body of a message to [email protected]
> More majordomo info at http://vger.kernel.org/majordomo-info.html
>
Hi Thomas,
2010/12/21 Thomas Gleixner <[email protected]>:
>
> At this point I want to say thanks to everyone who supported the
> development of the -rt patches with testing, bugfixes, stable
> backports etc. in the last years.
First of all: Thank YOU for maintaining the RT-patch and the huge
amount of effort you put into it!
> I hope that I can rely on your help
> for the upcoming 37 release and the ongoing effort to merge the patch
> fully into the mainline.
>From my side: I will test this new patch next Friday on ARM/at91, and
when 2.6.37-rt becomes available: I am happy to test that one also.
> I wish you all a Merry Christmas and a Happy New Year!
Merry Christmas and a Happy New Year to you and anyone else too!
Kind regards,
Remy
On 12/21/2010 05:52 AM, Thomas Gleixner wrote:
> We are pleased to announce the next update to our new preempt-rt
> series.
> ...
> Information on the RT patch can be found at:
>
> http://rt.wiki.kernel.org/index.php/Main_Page
>
> To build the 2.6.33.7.2-rt30 tree, the following patches should be
> applied:
>
> http://www.kernel.org/pub/linux/kernel/v2.6/linux-2.6.33.7.tar.bz2
> http://www.kernel.org/pub/linux/kernel/projects/rt/patch-2.6.33.7.2-rt30.bz2
>
> Enjoy !
I'm testing it right now and it does seem to have fixed some stuff, for
example I don't see any more errors when waking up from a sleep state.
Anyway, I'll let you know of any problems...
Thanks a lot! (and also good to know about the plans for 2.6.37.x rt!)
-- Fernando
On 12/24/2010 02:36 PM, Fernando Lopez-Lezcano wrote:
> On 12/21/2010 05:52 AM, Thomas Gleixner wrote:
>> We are pleased to announce the next update to our new preempt-rt
>> series.
>> ...
>> Information on the RT patch can be found at:
>>
>> http://rt.wiki.kernel.org/index.php/Main_Page
>>
>> To build the 2.6.33.7.2-rt30 tree, the following patches should be
>> applied:
>>
>> http://www.kernel.org/pub/linux/kernel/v2.6/linux-2.6.33.7.tar.bz2
>> http://www.kernel.org/pub/linux/kernel/projects/rt/patch-2.6.33.7.2-rt30.bz2
>>
>>
>> Enjoy !
>
> I'm testing it right now and it does seem to have fixed some stuff, for
> example I don't see any more errors when waking up from a sleep state.
> Anyway, I'll let you know of any problems...
So far so good...
> Thanks a lot! (and also good to know about the plans for 2.6.37.x rt!)
A question: I need to find out the pid of the irq process that handles a
particular audio device from the information available from udev (to
dynamically change its rt priority).
AFAICT that is not really possible. I can find the irq number easily but
if there is more than one device using that number then I'm sort of
stuck. The naming of the irq processes for each card appear to be
hardwired strings in the driver code that do not necessarily correspond
to any other name in the driver itself.
Is there a connection between the two somewhere in /sys, /proc or
anywhere else that I'm missing? (ie: which interrupt process corresponds
to which driver?)
-- Fernando