LinuxLists.cc - [000/474] 2.6.33.8-longterm review

2011-03-17 21:39:16

Subject: [000/474] 2.6.33.8-longterm review

2011-03-17 21:46:36

by Greg KH

[permalink] [raw]

Subject: [421/474] powerpc/kexec: Fix orphaned offline CPUs across kexec

2.6.33-longterm review patch. If anyone has any objections, please let us know.

------------------
Content-Length: 2566
Lines: 74

From: Matt Evans <[email protected]>

Commit: e8e5c2155b0035b6e04f29be67f6444bc914005b upstream

When CPU hotplug is used, some CPUs may be offline at the time a kexec is
performed. The subsequent kernel may expect these CPUs to be already running,
and will declare them stuck. On pseries, there's also a soft-offline (cede)
state that CPUs may be in; this can also cause problems as the kexeced kernel
may ask RTAS if they're online -- and RTAS would say they are. The CPU will
either appear stuck, or will cause a crash as we replace its cede loop beneath
it.

This patch kicks each present offline CPU awake before the kexec, so that
none are forever lost to these assumptions in the subsequent kernel.

Now, the behaviour is that all available CPUs that were offlined are now
online & usable after the kexec. This mimics the behaviour of a full reboot
(on which all CPUs will be restarted).

Signed-off-by: Matt Evans <[email protected]>
Signed-off-by: Benjamin Herrenschmidt <[email protected]>
Signed-off-by: Kamalesh babulal <[email protected]>
cc: Anton Blanchard <[email protected]>
Signed-off-by: Greg Kroah-Hartman <[email protected]>
---
arch/powerpc/kernel/machine_kexec_64.c | 25 +++++++++++++++++++++++++
1 file changed, 25 insertions(+)

--- a/arch/powerpc/kernel/machine_kexec_64.c
+++ b/arch/powerpc/kernel/machine_kexec_64.c
@@ -15,6 +15,7 @@
#include <linux/thread_info.h>
#include <linux/init_task.h>
#include <linux/errno.h>
+#include <linux/cpu.h>

#include <asm/page.h>
#include <asm/current.h>
@@ -169,10 +170,34 @@ static void kexec_smp_down(void *arg)
/* NOTREACHED */
}

+/*
+ * We need to make sure each present CPU is online. The next kernel will scan
+ * the device tree and assume primary threads are online and query secondary
+ * threads via RTAS to online them if required. If we don't online primary
+ * threads, they will be stuck. However, we also online secondary threads as we
+ * may be using 'cede offline'. In this case RTAS doesn't see the secondary
+ * threads as offline -- and again, these CPUs will be stuck.
+ *
+ * So, we online all CPUs that should be running, including secondary threads.
+ */
+static void wake_offline_cpus(void)
+{
+ int cpu = 0;
+
+ for_each_present_cpu(cpu) {
+ if (!cpu_online(cpu)) {
+ printk(KERN_INFO "kexec: Waking offline cpu %d.\n",
+ cpu);
+ cpu_up(cpu);
+ }
+ }
+}
+
static void kexec_prepare_cpus(void)
{
int my_cpu, i, notified=-1;

+ wake_offline_cpus();
smp_call_function(kexec_smp_down, NULL, /* wait */0);
my_cpu = get_cpu();

2011-03-18 05:26:16

by Mike Galbraith

[permalink] [raw]

Subject: Re: [000/474] 2.6.33.8-longterm review

On Thu, 2011-03-17 at 14:38 -0700, Greg KH wrote:
> Yeah, I know what you are thinking, "What, 2.6.33, are you crazy?"

Nah, dedicated. Thanks for the effort.

> So, if you are using .33, please give this patch a test and let me know
> if I've messed up anything (hint, I think I did, it's up to you to
> figure it out, consider it a test to see if anyone actually cares about
> this kernel version.)

I've cloned it, donning hip-waders.

-Mike

2011-03-18 08:48:07

by Thomas Gleixner

[permalink] [raw]

Subject: Re: [000/474] 2.6.33.8-longterm review

On Thu, 17 Mar 2011, Greg KH wrote:

> Yeah, I know what you are thinking, "What, 2.6.33, are you crazy?"

Aren't we all crazy ?

> I'd prefer it if you didn't answer that question...

Too late. You should have put this before the question.

> So, if you are using .33, please give this patch a test and let me know
> if I've messed up anything (hint, I think I did, it's up to you to
> figure it out, consider it a test to see if anyone actually cares about
> this kernel version.)

We'll pull it into rt and figure out what breaks.

Thanks for the effort,

tglx

2011-03-18 09:21:59

by John Kacur

[permalink] [raw]

Subject: Re: [000/474] 2.6.33.8-longterm review

On Thu, Mar 17, 2011 at 10:38 PM, Greg KH <[email protected]> wrote:
> Yeah, I know what you are thinking, "What, 2.6.33, are you crazy?"
>
> I'd prefer it if you didn't answer that question...
>
> Anyway, yes, a .33 kernel release, strange isn't it.
>
> Turns out there are a lot of people "stuck" on the 2.6.33 kernel series
> for a variety of valid reasons, not the least being the Real Time
> patchset. ?So, in the copious spare time I have, I've been trying to
> keep up with the patches that have been applied to the different
> longterm and stable kernel trees and have been queueing them up for .33
> as well.
>
> Here's the result of that. ?A 474 patch series that is bigger than most
> people ever want to review, or see in their inboxes. ?Because of that,
> it is ONLY being sent to the [email protected] mailing list. ?If
> you want to see the individual patches, look at the git tree of the
> patch queue, or look at that mailing list.
>
> So, if you are using .33, please give this patch a test and let me know
> if I've messed up anything (hint, I think I did, it's up to you to
> figure it out, consider it a test to see if anyone actually cares about
> this kernel version.)
>
> If I missed anything that I should have in here, please let me know.
>
> Responses should be made by Saturday, March 19, 2011, 21:00:00 UTC.
> Anything received after that time might be too late.
>
> The whole patch series can be found in one patch at:
> ? ? ? ?kernel.org/pub/linux/kernel/v2.6/longterm-review/patch-2.6.33.8-rc1.gz
> and the diffstat can be found below.
>

There is no rc1, so I assume
http://kernel.org/pub/linux/kernel/v2.6/longterm-review/patch-2.6.33.8.bz2
is correct?
Which git tree do I want to use?

Thanks
John Kacur

2011-03-18 19:47:39

by Greg KH

[permalink] [raw]

Subject: Re: [000/474] 2.6.33.8-longterm review

On Fri, Mar 18, 2011 at 10:21:54AM +0100, John Kacur wrote:
> On Thu, Mar 17, 2011 at 10:38 PM, Greg KH <[email protected]> wrote:
> > Yeah, I know what you are thinking, "What, 2.6.33, are you crazy?"
> >
> > I'd prefer it if you didn't answer that question...
> >
> > Anyway, yes, a .33 kernel release, strange isn't it.
> >
> > Turns out there are a lot of people "stuck" on the 2.6.33 kernel series
> > for a variety of valid reasons, not the least being the Real Time
> > patchset. ?So, in the copious spare time I have, I've been trying to
> > keep up with the patches that have been applied to the different
> > longterm and stable kernel trees and have been queueing them up for .33
> > as well.
> >
> > Here's the result of that. ?A 474 patch series that is bigger than most
> > people ever want to review, or see in their inboxes. ?Because of that,
> > it is ONLY being sent to the [email protected] mailing list. ?If
> > you want to see the individual patches, look at the git tree of the
> > patch queue, or look at that mailing list.
> >
> > So, if you are using .33, please give this patch a test and let me know
> > if I've messed up anything (hint, I think I did, it's up to you to
> > figure it out, consider it a test to see if anyone actually cares about
> > this kernel version.)
> >
> > If I missed anything that I should have in here, please let me know.
> >
> > Responses should be made by Saturday, March 19, 2011, 21:00:00 UTC.
> > Anything received after that time might be too late.
> >
> > The whole patch series can be found in one patch at:
> > ? ? ? ?kernel.org/pub/linux/kernel/v2.6/longterm-review/patch-2.6.33.8-rc1.gz
> > and the diffstat can be found below.
> >
>
> There is no rc1, so I assume
> http://kernel.org/pub/linux/kernel/v2.6/longterm-review/patch-2.6.33.8.bz2
> is correct?

Doh, yes, that is correct, I forgot to put -rc in the Makefile, and my
scripts just picked it up and went from there. My fault.

> Which git tree do I want to use?

There is no git tree of the expanded tree of patches, just use that
patch on top of 2.6.33.7 please.

Or you can apply the quilt patchs, all 474 of them, which are included
in the longterm-2.6.33 queue on git.kernel.org

thanks,

greg k-h

2011-03-19 10:08:29

by John Kacur

[permalink] [raw]

Subject: Re: [stable] [000/474] 2.6.33.8-longterm review

On Fri, Mar 18, 2011 at 8:36 PM, Greg KH <[email protected]> wrote:
> On Fri, Mar 18, 2011 at 10:21:54AM +0100, John Kacur wrote:
>> On Thu, Mar 17, 2011 at 10:38 PM, Greg KH <[email protected]> wrote:
>> > Yeah, I know what you are thinking, "What, 2.6.33, are you crazy?"
>> >
>> > I'd prefer it if you didn't answer that question...
>> >
>> > Anyway, yes, a .33 kernel release, strange isn't it.
>> >
>> > Turns out there are a lot of people "stuck" on the 2.6.33 kernel series
>> > for a variety of valid reasons, not the least being the Real Time
>> > patchset. ?So, in the copious spare time I have, I've been trying to
>> > keep up with the patches that have been applied to the different
>> > longterm and stable kernel trees and have been queueing them up for .33
>> > as well.
>> >
>> > Here's the result of that. ?A 474 patch series that is bigger than most
>> > people ever want to review, or see in their inboxes. ?Because of that,
>> > it is ONLY being sent to the [email protected] mailing list. ?If
>> > you want to see the individual patches, look at the git tree of the
>> > patch queue, or look at that mailing list.
>> >
>> > So, if you are using .33, please give this patch a test and let me know
>> > if I've messed up anything (hint, I think I did, it's up to you to
>> > figure it out, consider it a test to see if anyone actually cares about
>> > this kernel version.)
>> >
>> > If I missed anything that I should have in here, please let me know.
>> >
>> > Responses should be made by Saturday, March 19, 2011, 21:00:00 UTC.
>> > Anything received after that time might be too late.
>> >
>> > The whole patch series can be found in one patch at:
>> > ? ? ? ?kernel.org/pub/linux/kernel/v2.6/longterm-review/patch-2.6.33.8-rc1.gz
>> > and the diffstat can be found below.
>> >
>>
>> There is no rc1, so I assume
>> http://kernel.org/pub/linux/kernel/v2.6/longterm-review/patch-2.6.33.8.bz2
>> is correct?
>
> Doh, yes, that is correct, I forgot to put -rc in the Makefile, and my
> scripts just picked it up and went from there. ?My fault.
>
>> Which git tree do I want to use?
>
> There is no git tree of the expanded tree of patches, just use that
> patch on top of 2.6.33.7 please.
>
> Or you can apply the quilt patchs, all 474 of them, which are included
> in the longterm-2.6.33 queue on git.kernel.org
>

Thanks, actually that was what I was looking for, the git tree of the
quilt patches. Which are easy to script as git commits which makes
techniques like bisection easy.

Thanks

2011-03-19 14:47:11

by Remy Bohmer

[permalink] [raw]

Subject: Re: [000/474] 2.6.33.8-longterm review

Hi,

2011/3/18 Thomas Gleixner <[email protected]>:
> On Thu, 17 Mar 2011, Greg KH wrote:
>
>> Yeah, I know what you are thinking, "What, 2.6.33, are you crazy?"
>
> Aren't we all crazy ?
>
>> I'd prefer it if you didn't answer that question...
>
> Too late. You should have put this before the question.
>
>> So, if you are using .33, please give this patch a test and let me know
>> if I've messed up anything (hint, I think I did, it's up to you to
>> figure it out, consider it a test to see if anyone actually cares about
>> this kernel version.)
>
> We'll pull it into rt and figure out what breaks.

Can you please elaborate somewhat on the status of the RT-patchset
development for 2.6.38 and on?
It is silent about it for quite a while now, and actually the gap
between mainstream and the latest RT has never been this big before...
Are there major complications, lack on time or hands? Can I help ?

Thanks for the all the effort,

Kind regards,

Remy

2011-03-20 12:53:01

by Thomas Gleixner

[permalink] [raw]

Subject: Re: [000/474] 2.6.33.8-longterm review

On Sat, 19 Mar 2011, Remy Bohmer wrote:

> Hi,
>
> 2011/3/18 Thomas Gleixner <[email protected]>:
> > On Thu, 17 Mar 2011, Greg KH wrote:
> >
> >> Yeah, I know what you are thinking, "What, 2.6.33, are you crazy?"
> >
> > Aren't we all crazy ?
> >
> >> I'd prefer it if you didn't answer that question...
> >
> > Too late. You should have put this before the question.
> >
> >> So, if you are using .33, please give this patch a test and let me know
> >> if I've messed up anything (hint, I think I did, it's up to you to
> >> figure it out, consider it a test to see if anyone actually cares about
> >> this kernel version.)
> >
> > We'll pull it into rt and figure out what breaks.
>
> Can you please elaborate somewhat on the status of the RT-patchset
> development for 2.6.38 and on?

Working on it.

> It is silent about it for quite a while now, and actually the gap
> between mainstream and the latest RT has never been this big before...
> Are there major complications, lack on time or hands? Can I help ?

I concentrated on getting stuff upstream and reworking to make it more
palatable and maintainable.

Thanks,

tglx

2011-03-20 14:06:49

by Niccolò Belli

[permalink] [raw]

Subject: Re: [000/474] 2.6.33.8-longterm review

Il 20/03/2011 13:52, Thomas Gleixner ha scritto:
> On Sat, 19 Mar 2011, Remy Bohmer wrote:
>> Can you please elaborate somewhat on the status of the RT-patchset
>> development for 2.6.38 and on?
>
> Working on it.
>
>> It is silent about it for quite a while now, and actually the gap
>> between mainstream and the latest RT has never been this big before...
>> Are there major complications, lack on time or hands? Can I help ?
>
> I concentrated on getting stuff upstream and reworking to make it more
> palatable and maintainable.

Is there a working (where working means working for testing purposes)
snapshot somewhere?

Thanks,
Darkbasic