kexec and kdump on Xen？

kexec and kdump on Xen？

[Zhang, Yang Z]

Hi Daniel

Are kexec/kdump supported by latest Xen + upstream Linux? I remember it only works with xen+2.6.18 linux. I know you are working on enable it. So I wonder to know what the current status is. Thanks.	

best regards
yang

Re: kexec and kdump on Xen??

[Daniel Kiper]

Hi,

On Mon, Apr 14, 2014 at 03:38:26AM +0000, Zhang, Yang Z wrote:
> Hi Daniel
>
> Are kexec/kdump supported by latest Xen + upstream Linux?
> I remember it only works with xen+2.6.18 linux. I know you
> are working on enable it. So I wonder to know what the
> current status is. Thanks.

Xen 4.4 has new kexec implementation written by David Vrabel.
It does not require any specific Linux Kernel version because
this implementation is self contained. However, please do not
forget to use latest stable version of kexec-tools too.

Daniel

Re: kexec and kdump on Xen??

[Zhang, Yang Z]

Daniel Kiper wrote on 2014-04-14:
> Hi,
> 
> On Mon, Apr 14, 2014 at 03:38:26AM +0000, Zhang, Yang Z wrote:
>> Hi Daniel
>> 
>> Are kexec/kdump supported by latest Xen + upstream Linux?
>> I remember it only works with xen+2.6.18 linux. I know you are
>> working on enable it. So I wonder to know what the current status is. Thanks.
> 
> Xen 4.4 has new kexec implementation written by David Vrabel.
> It does not require any specific Linux Kernel version because this
> implementation is self contained. However, please do not forget to use

Cool..

> latest stable version of kexec-tools too.

Is there any BKM to use it? I saw an old wiki page but not sure whether it is still suitable.

> 
> Daniel


Best regards,
Yang

Re: kexec and kdump on Xen??

[Daniel Kiper]

Hey,

On Mon, Apr 14, 2014 at 08:15:35AM +0000, Zhang, Yang Z wrote:
> Daniel Kiper wrote on 2014-04-14:
> > Hi,
> >
> > On Mon, Apr 14, 2014 at 03:38:26AM +0000, Zhang, Yang Z wrote:
> >> Hi Daniel
> >>
> >> Are kexec/kdump supported by latest Xen + upstream Linux?
> >> I remember it only works with xen+2.6.18 linux. I know you are
> >> working on enable it. So I wonder to know what the current status is. Thanks.
> >
> > Xen 4.4 has new kexec implementation written by David Vrabel.
> > It does not require any specific Linux Kernel version because this
> > implementation is self contained. However, please do not forget to use
>
> Cool..
>
> > latest stable version of kexec-tools too.
>
> Is there any BKM to use it? I saw an old wiki page but
> not sure whether it is still suitable.

I assume that you are asking about kexec/kdump support in dom0.

Just build kexec-tools with Xen support enabled (you
must provide relevant Xen header files and libraries).
Then add crashkernel argument to Xen (e.g. crashkernel=512m@64m)
and reboot machine. Use kexec-tools like on bare metal machines.

Daniel

Re: kexec and kdump on Xen??

[Zhang, Yang Z]

Daniel Kiper wrote on 2014-04-14:
> Hey,
> 
> On Mon, Apr 14, 2014 at 08:15:35AM +0000, Zhang, Yang Z wrote:
>> Daniel Kiper wrote on 2014-04-14:
>>> Hi,
>>> 
>>> On Mon, Apr 14, 2014 at 03:38:26AM +0000, Zhang, Yang Z wrote:
>>>> Hi Daniel
>>>> 
>>>> Are kexec/kdump supported by latest Xen + upstream Linux?
>>>> I remember it only works with xen+2.6.18 linux. I know you are
>>>> working on enable it. So I wonder to know what the current status is.
> Thanks.
>>> 
>>> Xen 4.4 has new kexec implementation written by David Vrabel.
>>> It does not require any specific Linux Kernel version because this
>>> implementation is self contained. However, please do not forget to
>>> use
>> 
>> Cool..
>> 
>>> latest stable version of kexec-tools too.
>> 
>> Is there any BKM to use it? I saw an old wiki page but not sure
>> whether it is still suitable.
> 
> I assume that you are asking about kexec/kdump support in dom0.
> 
> Just build kexec-tools with Xen support enabled (you must provide
> relevant Xen header files and libraries).
> Then add crashkernel argument to Xen (e.g. crashkernel=512m@64m) and
> reboot machine. Use kexec-tools like on bare metal machines.

I am trying to kexec from Xen to Linux. But it fails to work.
./kexec -l /boot/vmlinuz-2.6.32-358.el6.x86_64 

kexec_load failed: Invalid argument
entry       = 0x7fff6760 flags = 0x3e0000
nr_segments = 3
segment[0].buf   = 0x138f660
segment[0].bufsz = 0x3641
segment[0].mem   = 0x3000
segment[0].memsz = 0x4000
segment[1].buf   = 0x7fc68badf610
segment[1].bufsz = 0x3d8110
segment[1].mem   = 0x1000000
segment[1].memsz = 0x3d9000
segment[2].buf   = 0x1387540
segment[2].bufsz = 0x80e0
segment[2].mem   = 0x7fff6000
segment[2].memsz = 0xa000


Here is my environment:
Xen version: latest Xen
Xen boot command:
	kernel (hd0,0)/boot/xen.gz dom0_mem=4096M loglvl=all guest_loglvl=all conring_size=4M crashkernel=128M@64M
    module (hd0,0)/boot/vmlinuz-xen o root=UUID=03a5e7c4-b8d0-45b4-a090-c34b06821cea pci=noaer max_loop=64 3 debug
    module (hd0,0)/boot/initrd-xen.img
kexec version with Xen enabled:  kexec-tools 2.0.6 released 18 April 2014

Also, it shows error when trying to kexec (Xen -> Xen)
./kexec -l /boot/xen.gz
Segmentation fault (core dumped)

BTW: kexec from Linux to Linux works well.

Best regards,
Yang

Re: kexec and kdump on Xen??

[Daniel Kiper]

On Tue, Apr 15, 2014 at 02:13:56AM +0000, Zhang, Yang Z wrote:
> Daniel Kiper wrote on 2014-04-14:
> > Hey,
> >
> > On Mon, Apr 14, 2014 at 08:15:35AM +0000, Zhang, Yang Z wrote:
> >> Daniel Kiper wrote on 2014-04-14:
> >>> Hi,
> >>>
> >>> On Mon, Apr 14, 2014 at 03:38:26AM +0000, Zhang, Yang Z wrote:
> >>>> Hi Daniel
> >>>>
> >>>> Are kexec/kdump supported by latest Xen + upstream Linux?
> >>>> I remember it only works with xen+2.6.18 linux. I know you are
> >>>> working on enable it. So I wonder to know what the current status is.
> > Thanks.
> >>>
> >>> Xen 4.4 has new kexec implementation written by David Vrabel.
> >>> It does not require any specific Linux Kernel version because this
> >>> implementation is self contained. However, please do not forget to
> >>> use
> >>
> >> Cool..
> >>
> >>> latest stable version of kexec-tools too.
> >>
> >> Is there any BKM to use it? I saw an old wiki page but not sure
> >> whether it is still suitable.
> >
> > I assume that you are asking about kexec/kdump support in dom0.
> >
> > Just build kexec-tools with Xen support enabled (you must provide
> > relevant Xen header files and libraries).
> > Then add crashkernel argument to Xen (e.g. crashkernel=512m@64m) and
> > reboot machine. Use kexec-tools like on bare metal machines.
>
> I am trying to kexec from Xen to Linux. But it fails to work.
> ./kexec -l /boot/vmlinuz-2.6.32-358.el6.x86_64

What about --append or --command-line?

> kexec_load failed: Invalid argument
> entry       = 0x7fff6760 flags = 0x3e0000
> nr_segments = 3
> segment[0].buf   = 0x138f660
> segment[0].bufsz = 0x3641
> segment[0].mem   = 0x3000
> segment[0].memsz = 0x4000
> segment[1].buf   = 0x7fc68badf610
> segment[1].bufsz = 0x3d8110
> segment[1].mem   = 0x1000000
> segment[1].memsz = 0x3d9000
> segment[2].buf   = 0x1387540
> segment[2].bufsz = 0x80e0
> segment[2].mem   = 0x7fff6000
> segment[2].memsz = 0xa000

Could you configure kexec-tools in following way:

./configure CFLAGS=-DDEBUG

Then build it as usual.

> Here is my environment:
> Xen version: latest Xen
> Xen boot command:
> 	kernel (hd0,0)/boot/xen.gz dom0_mem=4096M loglvl=all guest_loglvl=all conring_size=4M crashkernel=128M@64M
>     module (hd0,0)/boot/vmlinuz-xen o root=UUID=03a5e7c4-b8d0-45b4-a090-c34b06821cea pci=noaer max_loop=64 3 debug
>     module (hd0,0)/boot/initrd-xen.img
> kexec version with Xen enabled:  kexec-tools 2.0.6 released 18 April 2014
>
> Also, it shows error when trying to kexec (Xen -> Xen)
> ./kexec -l /boot/xen.gz
> Segmentation fault (core dumped)

We were focusing more on kdump than on other scenarios.
I suppose that kexec-tools assume that this compressed
ELF file is Linux Kernel and tries load it as usual which
of course is completely wrong. You should load Xen with
at least "-t multiboot-x86" argument (read help printed
by kexec). However, I am not sure that this loader
is actively maintained and works.

Daniel

Re: kexec and kdump on Xen??

[Zhang, Yang Z]

Daniel Kiper wrote on 2014-04-16:
> On Tue, Apr 15, 2014 at 02:13:56AM +0000, Zhang, Yang Z wrote:
>> Daniel Kiper wrote on 2014-04-14:
>>> Hey,
>>> 
>>> On Mon, Apr 14, 2014 at 08:15:35AM +0000, Zhang, Yang Z wrote:
>>>> Daniel Kiper wrote on 2014-04-14:
>>>>> Hi,
>>>>> 
>>>>> On Mon, Apr 14, 2014 at 03:38:26AM +0000, Zhang, Yang Z wrote:
>>>>>> Hi Daniel
>>>>>> 
>>>>>> Are kexec/kdump supported by latest Xen + upstream Linux?
>>>>>> I remember it only works with xen+2.6.18 linux. I know you are
>>>>>> working on enable it. So I wonder to know what the current status is.
>>> Thanks.
>>>>> 
>>>>> Xen 4.4 has new kexec implementation written by David Vrabel.
>>>>> It does not require any specific Linux Kernel version because
>>>>> this implementation is self contained. However, please do not
>>>>> forget to use
>>>> 
>>>> Cool..
>>>> 
>>>>> latest stable version of kexec-tools too.
>>>> 
>>>> Is there any BKM to use it? I saw an old wiki page but not sure
>>>> whether it is still suitable.
>>> 
>>> I assume that you are asking about kexec/kdump support in dom0.
>>> 
>>> Just build kexec-tools with Xen support enabled (you must provide
>>> relevant Xen header files and libraries).
>>> Then add crashkernel argument to Xen (e.g. crashkernel=512m@64m)
>>> and reboot machine. Use kexec-tools like on bare metal machines.
>> 
>> I am trying to kexec from Xen to Linux. But it fails to work.
>> ./kexec -l /boot/vmlinuz-2.6.32-358.el6.x86_64
> 
> What about --append or --command-line?
> 
>> kexec_load failed: Invalid argument
>> entry       = 0x7fff6760 flags = 0x3e0000
>> nr_segments = 3
>> segment[0].buf   = 0x138f660
>> segment[0].bufsz = 0x3641
>> segment[0].mem   = 0x3000
>> segment[0].memsz = 0x4000
>> segment[1].buf   = 0x7fc68badf610
>> segment[1].bufsz = 0x3d8110
>> segment[1].mem   = 0x1000000
>> segment[1].memsz = 0x3d9000
>> segment[2].buf   = 0x1387540
>> segment[2].bufsz = 0x80e0
>> segment[2].mem   = 0x7fff6000
>> segment[2].memsz = 0xa000
> 
> Could you configure kexec-tools in following way:
> 
> ./configure CFLAGS=-DDEBUG
> 
> Then build it as usual.
> 
>> Here is my environment:
>> Xen version: latest Xen
>> Xen boot command:
>> 	kernel (hd0,0)/boot/xen.gz dom0_mem=4096M loglvl=all
>> guest_loglvl=all
> conring_size=4M crashkernel=128M@64M
>>     module (hd0,0)/boot/vmlinuz-xen o
> root=UUID=03a5e7c4-b8d0-45b4-a090-c34b06821cea pci=noaer max_loop=64
> 3 debug
>>     module (hd0,0)/boot/initrd-xen.img kexec version with Xen enabled:
>> kexec-tools 2.0.6 released 18 April 2014
>> 
>> Also, it shows error when trying to kexec (Xen -> Xen) ./kexec -l
>> /boot/xen.gz Segmentation fault (core dumped)
> 
> We were focusing more on kdump than on other scenarios.
> I suppose that kexec-tools assume that this compressed ELF file is
> Linux Kernel and tries load it as usual which of course is completely
> wrong. You should load Xen with at least "-t multiboot-x86" argument (read help printed by kexec).
> However, I am not sure that this loader is actively maintained and works.
> 
> Daniel

Thanks for your comments.

After investigation, I found some bugs (One is the segment overlap check failure and another one is related to x2apic). With some workarounds, now I am able to kexec from Xen into Linux.
But it still fails to run kdump(Xen->linux). I add some debug messages in the key code path. It shows kexec_reloc() will cause machine reboot. Is it a known issue?

BTW: Have you ever tried kdump(Xen->linux) successfully? If yes, can you share some BKMs? Thanks.

Best regards,
Yang

Re: kexec and kdump on Xen??

[Andrew Cooper]

On 17/04/14 02:19, Zhang, Yang Z wrote:
>
>>> Here is my environment:
>>> Xen version: latest Xen
>>> Xen boot command:
>>> 	kernel (hd0,0)/boot/xen.gz dom0_mem=4096M loglvl=all
>>> guest_loglvl=all
>> conring_size=4M crashkernel=128M@64M
>>>     module (hd0,0)/boot/vmlinuz-xen o
>> root=UUID=03a5e7c4-b8d0-45b4-a090-c34b06821cea pci=noaer max_loop=64
>> 3 debug
>>>     module (hd0,0)/boot/initrd-xen.img kexec version with Xen enabled:
>>> kexec-tools 2.0.6 released 18 April 2014
>>>
>>> Also, it shows error when trying to kexec (Xen -> Xen) ./kexec -l
>>> /boot/xen.gz Segmentation fault (core dumped)
>> We were focusing more on kdump than on other scenarios.
>> I suppose that kexec-tools assume that this compressed ELF file is
>> Linux Kernel and tries load it as usual which of course is completely
>> wrong. You should load Xen with at least "-t multiboot-x86" argument (read help printed by kexec).
>> However, I am not sure that this loader is actively maintained and works.
>>
>> Daniel
> Thanks for your comments.
>
> After investigation, I found some bugs (One is the segment overlap check failure and another one is related to x2apic). With some workarounds, now I am able to kexec from Xen into Linux.
> But it still fails to run kdump(Xen->linux). I add some debug messages in the key code path. It shows kexec_reloc() will cause machine reboot. Is it a known issue?
>
> BTW: Have you ever tried kdump(Xen->linux) successfully? If yes, can you share some BKMs? Thanks.
>
> Best regards,
> Yang

You are going to have to be more descriptive than this with setups and
problems.

As for real examples,  XenServer unconditionally sets up a native crash
kernel and does use it in the case of a Xen/dom0 crash.  It does work
reliably and repeatedly[1] and is routinely tested.

~Andrew

[1] We are aware of one hardware specific bug whereby the crash kernel
fails to set up a timer interrupt.  Investigating it is on my todo list.

Re: kexec and kdump on Xen??

[Zhang, Yang Z]

[-- Attachment #1: Type: text/plain, Size: 2501 bytes --]

Andrew Cooper wrote on 2014-04-17:
> On 17/04/14 02:19, Zhang, Yang Z wrote:
>> 
>>>> Here is my environment:
>>>> Xen version: latest Xen
>>>> Xen boot command:
>>>> 	kernel (hd0,0)/boot/xen.gz dom0_mem=4096M loglvl=all
>>>> guest_loglvl=all
>>> conring_size=4M crashkernel=128M@64M
>>>>     module (hd0,0)/boot/vmlinuz-xen o
>>> root=UUID=03a5e7c4-b8d0-45b4-a090-c34b06821cea pci=noaer max_loop=64 3
>>> debug
>>>>     module (hd0,0)/boot/initrd-xen.img kexec version with Xen enabled:
>>>> kexec-tools 2.0.6 released 18 April 2014
>>>> 
>>>> Also, it shows error when trying to kexec (Xen -> Xen) ./kexec -l
>>>> /boot/xen.gz Segmentation fault (core dumped)
>>> We were focusing more on kdump than on other scenarios. I suppose that
>>> kexec-tools assume that this compressed ELF file is Linux Kernel and
>>> tries load it as usual which of course is completely wrong. You should
>>> load Xen with at least "-t multiboot-x86" argument (read help printed
>>> by kexec). However, I am not sure that this loader is actively
>>> maintained and works.
>>> 
>>> Daniel
>> Thanks for your comments.
>> 
>> After investigation, I found some bugs (One is the segment overlap
>> check
> failure and another one is related to x2apic). With some workarounds,
> now I am able to kexec from Xen into Linux.
>> But it still fails to run kdump(Xen->linux). I add some debug
>> messages in the
> key code path. It shows kexec_reloc() will cause machine reboot. Is it
> a known issue?
>> 
>> BTW: Have you ever tried kdump(Xen->linux) successfully? If yes, can
>> you share some BKMs? Thanks.
>> 
>> Best regards,
>> Yang
> 
> You are going to have to be more descriptive than this with setups and
> problems.

The previous mail(attached) has the details.

> 
> As for real examples,  XenServer unconditionally sets up a native
> crash kernel and does use it in the case of a Xen/dom0 crash.  It does
> work reliably and repeatedly[1] and is routinely tested.

I think XenServer uses the old approach to do the kexec. What I am trying is the new kexec/kdump feature which enabled in Xen4.4. 
One question, I remember kexec support is missing in 2.6.32-pvops kernel. If XenServer works well, do you know where can I find the patch to enable it for 2.6.32 kernel?

> 
> ~Andrew
> 
> [1] We are aware of one hardware specific bug whereby the crash kernel
> fails to set up a timer interrupt.  Investigating it is on my todo list.


Best regards,
Yang


[-- Attachment #2: Type: message/rfc822, Size: 3257 bytes --]

From: "Zhang, Yang Z" <yang.z.zhang@intel.com>
To: Daniel Kiper <dkiper@net-space.pl>
Cc: "Konrad Rzeszutek Wilk (konrad.wilk@oracle.com)" <konrad.wilk@oracle.com>, "xen-devel@lists.xen.org" <xen-devel@lists.xen.org>
Subject: RE: kexec and kdump  on Xen??
Date: Tue, 15 Apr 2014 02:13:55 +0000

Daniel Kiper wrote on 2014-04-14:
> Hey,
>
> On Mon, Apr 14, 2014 at 08:15:35AM +0000, Zhang, Yang Z wrote:
>> Daniel Kiper wrote on 2014-04-14:
>>> Hi,
>>>
>>> On Mon, Apr 14, 2014 at 03:38:26AM +0000, Zhang, Yang Z wrote:
>>>> Hi Daniel
>>>>
>>>> Are kexec/kdump supported by latest Xen + upstream Linux?
>>>> I remember it only works with xen+2.6.18 linux. I know you are
>>>> working on enable it. So I wonder to know what the current status is.
> Thanks.
>>>
>>> Xen 4.4 has new kexec implementation written by David Vrabel.
>>> It does not require any specific Linux Kernel version because this
>>> implementation is self contained. However, please do not forget to
>>> use
>>
>> Cool..
>>
>>> latest stable version of kexec-tools too.
>>
>> Is there any BKM to use it? I saw an old wiki page but not sure
>> whether it is still suitable.
>
> I assume that you are asking about kexec/kdump support in dom0.
>
> Just build kexec-tools with Xen support enabled (you must provide
> relevant Xen header files and libraries).
> Then add crashkernel argument to Xen (e.g. crashkernel=512m@64m) and
> reboot machine. Use kexec-tools like on bare metal machines.

I am trying to kexec from Xen to Linux. But it fails to work.
./kexec -l /boot/vmlinuz-2.6.32-358.el6.x86_64

kexec_load failed: Invalid argument
entry       = 0x7fff6760 flags = 0x3e0000
nr_segments = 3
segment[0].buf   = 0x138f660
segment[0].bufsz = 0x3641
segment[0].mem   = 0x3000
segment[0].memsz = 0x4000
segment[1].buf   = 0x7fc68badf610
segment[1].bufsz = 0x3d8110
segment[1].mem   = 0x1000000
segment[1].memsz = 0x3d9000
segment[2].buf   = 0x1387540
segment[2].bufsz = 0x80e0
segment[2].mem   = 0x7fff6000
segment[2].memsz = 0xa000


Here is my environment:
Xen version: latest Xen
Xen boot command:
        kernel (hd0,0)/boot/xen.gz dom0_mem=4096M loglvl=all guest_loglvl=all conring_size=4M crashkernel=128M@64M
    module (hd0,0)/boot/vmlinuz-xen o root=UUID=03a5e7c4-b8d0-45b4-a090-c34b06821cea pci=noaer max_loop=64 3 debug
    module (hd0,0)/boot/initrd-xen.img
kexec version with Xen enabled:  kexec-tools 2.0.6 released 18 April 2014

Also, it shows error when trying to kexec (Xen -> Xen)
./kexec -l /boot/xen.gz
Segmentation fault (core dumped)

BTW: kexec from Linux to Linux works well.

Best regards,
Yang



[-- Attachment #3: Type: text/plain, Size: 126 bytes --]

_______________________________________________
Xen-devel mailing list
Xen-devel@lists.xen.org
http://lists.xen.org/xen-devel

Re: kexec and kdump on Xen??

[Don Slutz]

On 04/17/14 21:45, Zhang, Yang Z wrote:
> Andrew Cooper wrote on 2014-04-17:
>> On 17/04/14 02:19, Zhang, Yang Z wrote:
>> As for real examples,  XenServer unconditionally sets up a native
>> crash kernel and does use it in the case of a Xen/dom0 crash.  It does
>> work reliably and repeatedly[1] and is routinely tested.
> I think XenServer uses the old approach to do the kexec. What I am trying is the new kexec/kdump feature which enabled in Xen4.4.
> One question, I remember kexec support is missing in 2.6.32-pvops kernel. If XenServer works well, do you know where can I find the patch to enable it for 2.6.32 kernel?
>

Verizon is using the new 4.4 way to do kdump.  So far all Xen/dom0
crashes have produced dumps.  This is using an unchanged kernel
3.8.11-100.fc17.x86_64.

     -Don Slutz

>> ~Andrew
>>
>> [1] We are aware of one hardware specific bug whereby the crash kernel
>> fails to set up a timer interrupt.  Investigating it is on my todo list.
>
> Best regards,
> Yang
>
>
>
> _______________________________________________
> Xen-devel mailing list
> Xen-devel@lists.xen.org
> http://lists.xen.org/xen-devel

Re: kexec and kdump on Xen??

[Andrew Cooper]

On 18/04/14 02:45, Zhang, Yang Z wrote:
> Andrew Cooper wrote on 2014-04-17:
>> On 17/04/14 02:19, Zhang, Yang Z wrote:
>> As for real examples,  XenServer unconditionally sets up a native
>> crash kernel and does use it in the case of a Xen/dom0 crash.  It does
>> work reliably and repeatedly[1] and is routinely tested.
> I think XenServer uses the old approach to do the kexec.

Older versions of XenServer did.  While investigating an issue, David
discovered quite how broken^W fragile it was and reimplemented the Xen
side, which is the code present in Xen 4.4

We then backported the Xen side fixes to XenServer 6.2, which otherwise
uses Xen 4.1.5 and 2.6.32-classic dom0.

Since doing so, I am not aware of a single customer crash report which
has arrived without proper set of crash logs, taken from the crash
environment.

> What I am trying is the new kexec/kdump feature which enabled in Xen4.4. 
> One question, I remember kexec support is missing in 2.6.32-pvops kernel. If XenServer works well, do you know where can I find the patch to enable it for 2.6.32 kernel?

It is, but that is the point.  The new Xen 4.4 code bypasses the dom0
kernel entirely, and /sbin/kexec uses a libxc interface for loading the
hypervisor crash/load images.

~Andrew

Re: kexec and kdump on Xen??

[David Vrabel]

On 17/04/14 02:19, Zhang, Yang Z wrote:
> 
> After investigation, I found some bugs (One is the segment overlap
> check failure and another one is related to x2apic). With some
> workarounds, now I am able to kexec from Xen into Linux. But it still
> fails to run kdump(Xen->linux). I add some debug messages in the key
> code path. It shows kexec_reloc() will cause machine reboot. Is it a
> known issue?

Rebooting after the kexec_reloc() call could be triggered from Xen,
purgatory or the kernel.

Can you turn on purgatory debug output (--console-serial
--serial-baud=115200 options to kexec for serial) and earlyprintk in the
kernel?

You may also want to specify a crashregion location.  e.g., we use the
Xen option "crashkernel=128M@32M".

David

Re: kexec and kdump on Xen??

[David Vrabel]

On 17/04/14 02:19, Zhang, Yang Z wrote:
> 
> 
> After investigation, I found some bugs (One is the segment overlap
> check failure and another one is related to x2apic). With some
> workarounds, now I am able to kexec from Xen into Linux. But it still
> fails to run kdump(Xen->linux). I add some debug messages in the key
> code path. It shows kexec_reloc() will cause machine reboot. Is it a
> known issue?

How did you fix these bugs?  Perhaps you inadvertently broke something else?

David

Re: kexec and kdump on Xen??

[Zhang, Yang Z]

Zhang, Yang Z wrote on 2014-04-15:
> Daniel Kiper wrote on 2014-04-14:
>> Hey,
>> 
>> On Mon, Apr 14, 2014 at 08:15:35AM +0000, Zhang, Yang Z wrote:
>>> Daniel Kiper wrote on 2014-04-14:
>>>> Hi,
>>>> 
>>>> On Mon, Apr 14, 2014 at 03:38:26AM +0000, Zhang, Yang Z wrote:
>>>>> Hi Daniel
>>>>> 
>>>>> Are kexec/kdump supported by latest Xen + upstream Linux?
>>>>> I remember it only works with xen+2.6.18 linux. I know you are
>>>>> working on enable it. So I wonder to know what the current status is.
>> Thanks.
>>>> 
>>>> Xen 4.4 has new kexec implementation written by David Vrabel.
>>>> It does not require any specific Linux Kernel version because this
>>>> implementation is self contained. However, please do not forget to
>>>> use
>>> 
>>> Cool..
>>> 
>>>> latest stable version of kexec-tools too.
>>> 
>>> Is there any BKM to use it? I saw an old wiki page but not sure
>>> whether it is still suitable.
>> 
>> I assume that you are asking about kexec/kdump support in dom0.
>> 
>> Just build kexec-tools with Xen support enabled (you must provide
>> relevant Xen header files and libraries).
>> Then add crashkernel argument to Xen (e.g. crashkernel=512m@64m) and
>> reboot machine. Use kexec-tools like on bare metal machines.
> 
> I am trying to kexec from Xen to Linux. But it fails to work.
> ./kexec -l /boot/vmlinuz-2.6.32-358.el6.x86_64
> 
> kexec_load failed: Invalid argument
> entry       = 0x7fff6760 flags = 0x3e0000
> nr_segments = 3
> segment[0].buf   = 0x138f660
> segment[0].bufsz = 0x3641
> segment[0].mem   = 0x3000
> segment[0].memsz = 0x4000
> segment[1].buf   = 0x7fc68badf610
> segment[1].bufsz = 0x3d8110
> segment[1].mem   = 0x1000000
> segment[1].memsz = 0x3d9000
> segment[2].buf   = 0x1387540
> segment[2].bufsz = 0x80e0
> segment[2].mem   = 0x7fff6000
> segment[2].memsz = 0xa000
> 
> 
> Here is my environment:
> Xen version: latest Xen
> Xen boot command:
> 	kernel (hd0,0)/boot/xen.gz dom0_mem=4096M loglvl=all guest_loglvl=all
> conring_size=4M crashkernel=128M@64M
>     module (hd0,0)/boot/vmlinuz-xen o
> root=UUID=03a5e7c4-b8d0-45b4-a090-c34b06821cea pci=noaer max_loop=64
> 3 debug
>     module (hd0,0)/boot/initrd-xen.img kexec version with Xen enabled:
> kexec-tools 2.0.6 released 18 April 2014
> 
> Also, it shows error when trying to kexec (Xen -> Xen) ./kexec -l
> /boot/xen.gz Segmentation fault (core dumped)
> 
> BTW: kexec from Linux to Linux works well.

The reason is that the first segment in my box (Actually I am using nested Xen to try) starts from 0x3000 to 0x7000, but kexec will map 0 - 1MiB unconditionally:
    /*  
     * Ensure 0 - 1 MiB is mapped and accessible by the image.
     *
     * This allows access to the VGA memory and the region
     * purgatory copies in the crash case.
     */
    set_xen_guest_handle(xen_segs[s].buf.h, HYPERCALL_BUFFER_NULL);
    xen_segs[s].buf_size = 0;
    xen_segs[s].dest_maddr = 0;
    xen_segs[s].dest_size = 1 * 1024 * 1024;

Then overlap happens. 

Any idea?

Best regards,
Yang

Re: kexec and kdump on Xen？

[David Vrabel]

On 14/04/14 04:38, Zhang, Yang Z wrote:
> Hi Daniel
> 
> Are kexec/kdump supported by latest Xen + upstream Linux? I remember
> it only works with xen+2.6.18 linux. I know you are working on enable
> it. So I wonder to know what the current status is. Thanks.

It depends on what you're asking for.

Xen kexec is available in Xen 4.4 using the latest kexec-tools.  The
kexec tool works the same as for bare metal.

In guest kexec works for PVHVM/HVM guests but does not work for PV guests.

David