Re: kernel-handbook: How to rebuild official Debian Linux kernel package with debug info disabled?

[Thomas Schwinge]

Hi!

On 2020-03-05T09:35:16+0100, I wrote:
> To test something, I need to rebuild the official Debian Linux kernel
> package, so I'm following
> 
> "4.2. Rebuilding official Debian kernel packages".
>
> # apt-get --purge install build-essential fakeroot
> # apt-get --purge build-dep linux/testing
> $ apt-get source linux/testing
> $ cd linux-5.4.19/
>
> As disk space is limited, I'd like to build with debug info disabled, so
> per the instructions given I "disable debug info by changing the value of
> 'debug-info' to 'false' in 'debian/config/defines'".  I ignored the "(or
> 'debian/config/*arch*/defines' in older package versions)" remark, as I
> didn't think I was building an "older package version".
>
> $ PATH=/usr/lib/ccache:$PATH
> $ fakeroot debian/rules source
> $ fakeroot make -f debian/rules.gen binary-arch_i386_none_686-pae
>
> After a few hours that exploded with "disk full", and indeed
> 'CONFIG_DEBUG_INFO=y' is set in
> 'debian/build/build_i386_none_686-pae/.config' (the file matches
> '/boot/config-5.4.0-4-686-pae').
>
> I now see:
>
> $ grep -r -B1 debug-info debian/config/
> debian/config/amd64/defines-[build]
> debian/config/amd64/defines:debug-info: true
> --
> debian/config/s390x/defines-[s390x_build]
> debian/config/s390x/defines:debug-info: true
> --
> debian/config/hppa/defines-# temporarily disable debug info: Required 
> disk size exceeds our hppa buildd machines.
> debian/config/hppa/defines:debug-info: false
> --
> debian/config/defines-[build]
> debian/config/defines:debug-info: true
> --
> debian/config/arm64/defines-[build]
> debian/config/arm64/defines:debug-info: true
> --
> debian/config/i386/defines-[686-pae_build]
> debian/config/i386/defines:debug-info: true
>
> ..., so I'll try again with setting 'debug-info' to 'false' both in
> 'debian/config/defines' and 'debian/config/i386/defines'.

Eh, same result/problem...  ;-/

I'll (later) try to figure out the right edits or commands to use, but
would also appreciate working instructions, if anyone knows off-hand?
Probably separate '.config' generation and alternation
('CONFIG_DEBUG_INFO=n', is that enough?) from the actual build?

> Why do some of these configuration files override the generic one
> (without giving rationale), and/or do the kernel-handbook instructions
> need to be updated?


Grüße
 Thomas
-
Mentor Graphics (Deutschland) GmbH, Arnulfstraße 201, 80634 München / Germany
Registergericht München HRB 106955, Geschäftsführer: Thomas Heurung, Alexander 
Walter

kernel-handbook: How to rebuild official Debian Linux kernel package with debug info disabled?

[Thomas Schwinge]

-
Mentor Graphics (Deutschland) GmbH, Arnulfstraße 201, 80634 München / Germany
Registergericht München HRB 106955, Geschäftsführer: Thomas Heurung, Alexander 
Walter
--- Begin Message ---
Hi!

To test something, I need to rebuild the official Debian Linux kernel
package, so I'm following

"4.2. Rebuilding official Debian kernel packages".

# apt-get --purge install build-essential fakeroot
# apt-get --purge build-dep linux/testing
$ apt-get source linux/testing
$ cd linux-5.4.19/

As disk space is limited, I'd like to build with debug info disabled, so
per the instructions given I "disable debug info by changing the value of
'debug-info' to 'false' in 'debian/config/defines'".  I ignored the "(or
'debian/config/*arch*/defines' in older package versions)" remark, as I
didn't think I was building an "older package version".

$ PATH=/usr/lib/ccache:$PATH
$ fakeroot debian/rules source
$ fakeroot make -f debian/rules.gen binary-arch_i386_none_686-pae

After a few hours that exploded with "disk full", and indeed
'CONFIG_DEBUG_INFO=y' is set in
'debian/build/build_i386_none_686-pae/.config' (the file matches
'/boot/config-5.4.0-4-686-pae').

I now see:

$ grep -r -B1 debug-info debian/config/
debian/config/amd64/defines-[build]
debian/config/amd64/defines:debug-info: true
--
debian/config/s390x/defines-[s390x_build]
debian/config/s390x/defines:debug-info: true
--
debian/config/hppa/defines-# temporarily disable debug info: Required disk 
size exceeds our hppa buildd machines.
debian/config/hppa/defines:debug-info: false
--
debian/config/defines-[build]
debian/config/defines:debug-info: true
--
debian/config/arm64/defines-[build]
debian/config/arm64/defines:debug-info: true
--
debian/config/i386/defines-[686-pae_build]
debian/config/i386/defines:debug-info: true

..., so I'll try again with setting 'debug-info' to 'false' both in
'debian/config/defines' and 'debian/config/i386/defines'.

Why do some of these configuration files override the generic one
(without giving rationale), and/or do the kernel-handbook instructions
need to be updated?


Grüße
 Thomas


signature.asc
Description: PGP signature
--- End Message ---

Re: Linux kernel crash (3.2.0-2-amd64) when trying to play audio CD

[Thomas Schwinge]

Hi!

On Sat, 22 Sep 2012 13:53:20 +0200, I wrote:
 On Mon, 17 Sep 2012 19:23:45 +0200, Borislav Petkov b...@amd64.org wrote:
  On Thu, Sep 13, 2012 at 02:58:25PM +0200, Borislav Petkov wrote:
   Here's the script, let me know how it goes:
  
  Here's a more correct version. [...]

 However, putting in that one audio CD still crashes the machine: [...]

The problem solved itself: the machine recently died of hardware
failure.  ;-|


Grüße,
 Thomas


pgppEui8H6z5N.pgp
Description: PGP signature

Re: Linux kernel crash (3.2.0-2-amd64) when trying to play audio CD

[Thomas Schwinge]

Hi Borislav!

On Mon, 16 Jun 2014 11:43:24 +0200, Borislav Petkov b...@alien8.de wrote:
 On Mon, Jun 16, 2014 at 09:13:20AM +0200, Thomas Schwinge wrote:
  The problem solved itself: the machine recently died of hardware
  failure. ;-|
 
 How convenient. :-P

;-P

 What happened?

The power supply and/or mainboard didn't wake up after a recent power
cycle.  As both these aren't standard form factors (and the
machine/architecture had some issues, as discussed, and see also
http://news.gmane.org/find-root.php?message_id=%3C87r45fajun.fsf%40schwinge.name%3E
for another issue), I didn't bother looking for spare parts.  Want me to
ship it to you?  ;-)


Grüße,
 Thomas


pgpVNmD8lhCzz.pgp
Description: PGP signature

Re: Linux kernel crash (3.2.0-2-amd64) when trying to play audio CD

[Thomas Schwinge]

Hi Borislav!

On Mon, 17 Sep 2012 19:23:45 +0200, Borislav Petkov b...@amd64.org wrote:
 On Thu, Sep 13, 2012 at 02:58:25PM +0200, Borislav Petkov wrote:
  Here's the script, let me know how it goes:
 
 Here's a more correct version. I'm not saying yours won't work based
 on the rdmsr and setpci output on your box, but here's a version which
 should pay attention to one other workaround and work on all K8s.

Thanks for that.  I invoke it from /etc/rc.local, and then get:

$ sudo setpci -s 18.0 0x68.l
0f20cc0f
$ sudo rdmsr 0xc001001f
10008

Previously that was:

$ sudo setpci -s 18.0 0x68.l
0f00cc0f
$ sudo rdmsr 0xc001001f
8

However, putting in that one audio CD still crashes the machine:
http://schwinge.homeip.net/~thomas/tmp/linux-crash-audio-cd-2.jpeg.
Original (identical, from a quick glance) MCE is at
http://schwinge.homeip.net/~thomas/tmp/linux-crash-audio-cd.jpeg.


Grüße,
 Thomas


pgpfX9TohlnFr.pgp
Description: PGP signature

Re: Linux kernel crash (3.2.0-2-amd64) when trying to play audio CD

[Thomas Schwinge]

Hi Borislav!

I'm sorry for the late answer.


On Mon, 18 Jun 2012 17:51:41 +0200, Borislav Petkov b...@amd64.org wrote:
 On Mon, Jun 18, 2012 at 04:52:36PM +0200, Thomas Schwinge wrote:
   Ok, can you do the following:
   
   $ setpci -s 18.0 0x68.l
  
  $ sudo setpci -s 18.0 0x68.l
  0f00cc0f
  
   $ rdmsr 0xc001001f
  
  $ sudo modprobe msr
  $ sudo rdmsr 0xc001001f
  8
 
 Ok, this is a BIOS bug - we have a recommendation about what the BIOS
 should do but yours doesn't do it. Are you sure you're using the latest
 BIOS for your motherboard?
 
 Can you check for updates and if you get a newer version, can you update
 your BIOS and do the two commands above?

I have now finally been able to check this, and yes, I'm already using
the latest version of the BIOS, which is Phoenix cME FirstBIOS Desktop
Pro version 5.00 R1.07.2264.A1 (a.k.a.  5.00.1.07, 25.04.2006) for
Fujitsu Siemens Esprimo E5600, mainboard D2264-A1, CPU AMD Sempron 3000+.

 If there's no newer BIOS, we can try applying the workaround at boot in
 an init script to fix your issue.

How'd that be done?


Grüße,
 Thomas


pgpFrsTTT5BUI.pgp
Description: PGP signature

Re: Linux kernel crash (3.2.0-2-amd64) when trying to play audio CD

[Thomas Schwinge]

Hi Borislav!

First, thanks for helping!

On Mon, 18 Jun 2012 16:29:00 +0200, Borislav Petkov b...@amd64.org wrote:
 On Mon, Jun 18, 2012 at 07:28:07AM +0200, Thomas Schwinge wrote:
   Let me guess: AMD K8?
  
  Correct.
  
   Also, can I get /proc/cpuinfo please?
  
  $ cat  /proc/cpuinfo
  processor   : 0
 
 Ok, a single core.
 
  vendor_id   : AuthenticAMD
  cpu family  : 15
  model   : 47
  model name  : AMD Sempron(tm) Processor 3000+
  stepping: 2
  cpu MHz : 1000.000
  cache size  : 128 KB
  fpu : yes
  fpu_exception   : yes
  cpuid level : 1
  wp  : yes
  flags   : fpu vme de pse tsc msr pae mce cx8 apic sep mtrr pge 
  mca cmov pat pse36 clflush mmx fxsr sse sse2 syscall nx mmxext fxsr_opt lm 
  3dnowext 3dnow up rep_good nopl pni lahf_lm
  bogomips: 2000.16
  TLB size: 1024 4K pages
  clflush size: 64
  cache_alignment : 64
  address sizes   : 40 bits physical, 48 bits virtual
  power management: ts fid vid ttp tm stc
 
 Ok, can you do the following:
 
 $ setpci -s 18.0 0x68.l

$ sudo setpci -s 18.0 0x68.l
0f00cc0f

 $ rdmsr 0xc001001f

$ sudo modprobe msr
$ sudo rdmsr 0xc001001f
8


Grüße,
 Thomas


pgp76lV6tsmbc.pgp
Description: PGP signature

Re: Linux kernel crash (3.2.0-2-amd64) when trying to play audio CD

[Thomas Schwinge]

Hi!

On Sat, 16 Jun 2012 20:10:08 +0200, Borislav Petkov b...@alien8.de wrote:
 On Sat, Jun 16, 2012 at 12:14:14PM +0200, Thomas Schwinge wrote:
  After many others that played with no problems, I just tried playing
  Guano Apes' audio CD »Proud Like a God« (1997) in Rhythmbox, on my mostly
  up-to-date Debian GNU/Linux testing system.
  
  $ uname -a
  Linux boole 3.2.0-2-amd64 #1 SMP Mon May 21 17:45:41 UTC 2012 x86_64 
  GNU/Linux
  
  ... which specifically is »linux-image-3.2.0-2-amd64 3.2.18-1«.
  
  While loading the CD in Rhythmbox, after a few seconds the kernel
  crashed.  I rebooted, tried again, and it crashed again.  At
  http://schwinge.homeip.net/~thomas/tmp/linux-crash-audio-cd.jpeg you
  can find the panic message (low resolution, sorry), pointing to the
  SCSI/ATA code.
 
 Let me guess: AMD K8?

Correct.

 Also, can I get /proc/cpuinfo please?

$ cat  /proc/cpuinfo
processor   : 0
vendor_id   : AuthenticAMD
cpu family  : 15
model   : 47
model name  : AMD Sempron(tm) Processor 3000+
stepping: 2
cpu MHz : 1000.000
cache size  : 128 KB
fpu : yes
fpu_exception   : yes
cpuid level : 1
wp  : yes
flags   : fpu vme de pse tsc msr pae mce cx8 apic sep mtrr pge mca 
cmov pat pse36 clflush mmx fxsr sse sse2 syscall nx mmxext fxsr_opt lm 3dnowext 
3dnow up rep_good nopl pni lahf_lm
bogomips: 2000.16
TLB size: 1024 4K pages
clflush size: 64
cache_alignment : 64
address sizes   : 40 bits physical, 48 bits virtual
power management: ts fid vid ttp tm stc


Grüße,
 Thomas


pgpIuKAfGVXGP.pgp
Description: PGP signature

Re: Linux kernel crash (3.2.0-2-amd64) when trying to play audio CD

[Thomas Schwinge]

Hi!

On Sat, 16 Jun 2012 14:40:34 +0100, Ben Hutchings b...@decadent.org.uk wrote:
 On Sat, 2012-06-16 at 12:14 +0200, Thomas Schwinge wrote:
  After many others that played with no problems, I just tried playing
  Guano Apes' audio CD »Proud Like a God« (1997) in Rhythmbox, on my mostly
  up-to-date Debian GNU/Linux testing system.
  
  $ uname -a
  Linux boole 3.2.0-2-amd64 #1 SMP Mon May 21 17:45:41 UTC 2012 x86_64 
  GNU/Linux
  
  ... which specifically is »linux-image-3.2.0-2-amd64 3.2.18-1«.
  
  While loading the CD in Rhythmbox, after a few seconds the kernel
  crashed.  I rebooted, tried again, and it crashed again.  At
  http://schwinge.homeip.net/~thomas/tmp/linux-crash-audio-cd.jpeg you
  can find the panic message (low resolution, sorry), pointing to the
  SCSI/ATA code.
  
  Are there any such known issues?  Does the panic message give any clue
  already, or would you want me to reproduce that with debugging symbols,
  etc.?
 
 Your computer is broken; Machine Check Exception indicates a hardware
 fault.

Hmm, I see, thanks.  Nothing the kernel could do to catch that, I
suppose?  (Scary that simply putting a regular audio CD into the drive
can crash the kernel.)

 The other text at the top of the screen may help to identify
 which component is at fault.


Grüße,
 Thomas


pgpXm0va8eIAE.pgp
Description: PGP signature

Linux kernel crash (3.2.0-2-amd64) when trying to play audio CD

[Thomas Schwinge]

Hi!

After many others that played with no problems, I just tried playing
Guano Apes' audio CD »Proud Like a God« (1997) in Rhythmbox, on my mostly
up-to-date Debian GNU/Linux testing system.

$ uname -a
Linux boole 3.2.0-2-amd64 #1 SMP Mon May 21 17:45:41 UTC 2012 x86_64 
GNU/Linux

... which specifically is »linux-image-3.2.0-2-amd64 3.2.18-1«.

While loading the CD in Rhythmbox, after a few seconds the kernel
crashed.  I rebooted, tried again, and it crashed again.  At
http://schwinge.homeip.net/~thomas/tmp/linux-crash-audio-cd.jpeg you
can find the panic message (low resolution, sorry), pointing to the
SCSI/ATA code.

Are there any such known issues?  Does the panic message give any clue
already, or would you want me to reproduce that with debugging symbols,
etc.?


Grüße,
 Thomas


pgpt7m0FZdY6X.pgp
Description: PGP signature

Bug#588509: hang sometimes occurs earlier

[Thomas Schwinge]

Hello!

FWIW, I'm seeing these hangs, too.  Both when booting as a domU and dom0.


On Wed, Jul 14, 2010 at 09:40:42AM +0200, Rik Theys wrote:
 The dom0 runs Xen 3.2

Wow, I don't think that's supported anymore.  Isn't switching to the 4.0
packages an option for you?


Regards,
 Thomas


signature.asc
Description: Digital signature

Bug#576877: linux-image-2.6.32-4-xen-amd64: doesn't boot with 4 GiB; resets immediatelly, no log messages

[Thomas Schwinge]

Hello!

On Thu, Apr 08, 2010 at 12:52:25AM +0200, I wrote:
 2.6.32-11: resets the machine immediatelly, no log messages visible
 (don't have serial console set-up).

This has been fixed with commit 28c3e64a77c36a6e52a50f713e7caf268e4b9b1f
in jeremy/xen/stable-2.6.32.x.


Regards,
 Thomas


signature.asc
Description: Digital signature

Bug#576877: linux-image-2.6.32-4-xen-amd64: doesn't boot with 4 GiB; resets immediatelly, no log messages

[Thomas Schwinge]

Package: linux-2.6
Version: 2.6.32-11
Severity: important

Hello!

To get the 2.6.32-11 kernel to boot, I need to supply mem=4G.  This was
not necessary with the 2.6.32-10 kernel.

2.6.32-10:
(XEN) Xen version 3.4 (Debian 3.4.3~rc3-1) (wa...@debian.org) (gcc version 
4.4.3 (Debian 4.4.3-2) ) Mon Mar  1 21:16:18 UTC 2010
(XEN) Command line: 
(XEN) Video information:
(XEN)  VGA is text mode 80x25, font 8x16
(XEN)  VBE/DDC methods: V2; EDID transfer time: 1 seconds
(XEN) Disc information:
(XEN)  Found 1 MBR signatures
(XEN)  Found 1 EDD information structures
(XEN) Xen-e820 RAM map:
(XEN)   - 0009e000 (usable)
(XEN)  0009e000 - 000a (reserved)
(XEN)  000e4000 - 0010 (reserved)
(XEN)  0010 - ddee (usable)
(XEN)  ddee - ddeef000 (ACPI data)
(XEN)  ddeef000 - ddf0 (ACPI NVS)
(XEN)  ddf0 - e000 (reserved)
(XEN)  fec0 - fec1 (reserved)
(XEN)  fee0 - fee01000 (reserved)
(XEN)  fff0 - 0001 (reserved)
(XEN)  0001 - 00012000 (usable)
(XEN) System RAM: 4062MB (4159992kB)

2.6.32-11: resets the machine immediatelly, no log messages visible
(don't have serial console set-up).

2.6.32-11, with mem=4G:
(XEN) Xen version 3.4 (Debian 3.4.3~rc3-1) (wa...@debian.org) (gcc version 
4.4.3 (Debian 4.4.3-2) ) Mon Mar  1 21:16:18 UTC 2010
(XEN) Command line: mem=4G
(XEN) Video information:
(XEN)  VGA is text mode 80x25, font 8x16
(XEN)  VBE/DDC methods: V2; EDID transfer time: 1 seconds
(XEN) Disc information:
(XEN)  Found 1 MBR signatures
(XEN)  Found 1 EDD information structures
(XEN) Truncating RAM from 4718592kB to 4194304kB
(XEN) Xen-e820 RAM map:
(XEN)   - 0009e000 (usable)
(XEN)  0009e000 - 000a (reserved)
(XEN)  000e4000 - 0010 (reserved)
(XEN)  0010 - ddee (usable)
(XEN)  ddee - ddeef000 (ACPI data)
(XEN)  ddeef000 - ddf0 (ACPI NVS)
(XEN)  ddf0 - e000 (reserved)
(XEN)  fec0 - fec1 (reserved)
(XEN)  fee0 - fee01000 (reserved)
(XEN)  fff0 - 0001 (reserved)
(XEN) System RAM: 3550MB (3635704kB)


/proc/cpuinfo 
processor   : 0
vendor_id   : AuthenticAMD
cpu family  : 15
model   : 47
model name  : AMD Sempron(tm) Processor 3000+
stepping: 2
cpu MHz : 1800.085
cache size  : 128 KB
fpu : yes
fpu_exception   : yes
cpuid level : 1
wp  : yes
flags   : fpu de tsc msr pae cx8 apic mtrr cmov clflush mmx fxsr 
sse sse2 syscall nx mmxext fxsr_opt lm 3dnowext 3dnow up rep_good pni 
hypervisor lahf_lm
bogomips: 3600.17
TLB size: 1024 4K pages
clflush size: 64
cache_alignment : 64
address sizes   : 40 bits physical, 48 bits virtual
power management: ts fid vid ttp tm stc


-- Package-specific info:
** Version:
Linux version 2.6.32-4-xen-amd64 (Debian 2.6.32-11) (m...@debian.org) (gcc 
version 4.3.4 (Debian 4.3.4-8) ) #1 SMP Mon Apr 5 23:19:15 UTC 2010

** Command line:
root=/dev/mapper/vg0-boole--root ro console=tty0

** Tainted: W (512)
 * Taint on warning.

** Kernel log:
[3.594644] sr 0:0:0:0: Attached scsi CD-ROM sr0
[3.595448] usbcore: registered new interface driver hiddev
[3.599715] sd 2:0:0:0: [sda] 156301488 512-byte logical blocks: (80.0 
GB/74.5 GiB)
[3.599855] sd 2:0:0:0: [sda] Write Protect is off
[3.599893] sd 2:0:0:0: [sda] Mode Sense: 00 3a 00 00
[3.599938] sd 2:0:0:0: [sda] Write cache: enabled, read cache: enabled, 
doesn't support DPO or FUA
[3.600295]  sda:
[3.603437] input: Logitech USB-PS/2 Optical Mouse as 
/devices/pci:00/:00:03.0/usb2/2-3/2-3:1.0/input/input4
[3.603627] generic-usb 0003:046D:C00E.0001: input,hidraw0: USB HID v1.10 
Mouse [Logitech USB-PS/2 Optical Mouse] on usb-:00:03.0-3/input0
[3.603717] usbcore: registered new interface driver usbhid
[3.603755] usbhid: v2.6:USB HID core driver
[3.613586] sr 0:0:0:0: Attached scsi generic sg0 type 5
[3.614071] sd 2:0:0:0: Attached scsi generic sg1 type 0
[3.625464]  sda1
[3.626452] sd 2:0:0:0: [sda] Attached SCSI disk
[5.544048] floppy0: no floppy controllers found
[5.611207] device-mapper: uevent: version 1.0.3
[5.612270] device-mapper: ioctl: 4.15.0-ioctl (2009-04-01) initialised: 
dm-de...@redhat.com
[5.876201] PM: Starting manual resume from disk
[5.876242] PM: Resume from partition 254:1
[5.876244] PM: Checking hibernation image.
[5.876535] PM: Error -22 checking image 

Re: Xen dom0 2.6.32 stable branch

[Thomas Schwinge]

Hello!

On Wed, 24 Feb 2010 22:07:42 +, Ian Campbell wrote:
 On Wed, 2010-02-24 at 18:29 +0100, Bastian Blank wrote:
  On Wed, Feb 24, 2010 at 03:42:32PM +0100, Bastian Blank wrote:
   http://hermes.jura.uni-tuebingen.de/~blank/debian/xen-test/
  
  It seems to run. 
 
 I get a panic very early on No available IRQ to bind to: increase
 nr_irqs!. Will investigate tomorrow.

Same for me.  Has this been resolved already?

I'm trying to use the 2.6.32-10~xen.1 amd64 packages from
http://hermes.jura.uni-tuebingen.de/~blank/debian/xen-test/ together
with the 3.4.3~rc3-1 amd64 hypervisor.


Regards,
 Thomas


signature.asc
Description: Digital signature

Re: [Pkg-xen-devel] Success with linux-image-2.6.31-1-xen-amd64_2.6.31-2_amd64.deb

[Thomas Schwinge]

Hello!

On Wed, Nov 18, 2009 at 11:54:49PM +0100, I wrote:
 I finally got back to spending a few hours on working on xenifying this
 pre-hardware-virtualization AMD64 machine.  I re-installed the system.
 Basing it on stable (Lenny) this time.  The ol' combo of
 xen-hypervisor-3.2-1-amd64 / linux-image-2.6.26-2-xen-amd64 boots fine.
 Likewise when switching to xen-hypervisor-3.4-amd64.  And then the goal,
 http://hermes.jura.uni-tuebingen.de/~blank/debian/xen-test/linux-image-2.6.31-1-xen-amd64_2.6.31-2_amd64.deb
 [...]

 The plan is then to gradually update the system to Debian testing and see
 whether the booting / lvm / initramfs problems come back.

Here we go: http://bugs.debian.org/557645


Regards,
 Thomas


signature.asc
Description: Digital signature

Success with linux-image-2.6.31-1-xen-amd64_2.6.31-2_amd64.deb

[Thomas Schwinge]

Hello!

I finally got back to spending a few hours on working on xenifying this
pre-hardware-virtualization AMD64 machine.  I re-installed the system.
Basing it on stable (Lenny) this time.  The ol' combo of
xen-hypervisor-3.2-1-amd64 / linux-image-2.6.26-2-xen-amd64 boots fine.
Likewise when switching to xen-hypervisor-3.4-amd64.  And then the goal,
http://hermes.jura.uni-tuebingen.de/~blank/debian/xen-test/linux-image-2.6.31-1-xen-amd64_2.6.31-2_amd64.deb
-- which finally, after rising dom0_mem=15 (was 10 before) the
system actually boots instead of crashing / rebooting during the handover
from initial Xen to Linux (I suppose).  Please tell if you're interested
in any logs etc.

What I didn't try yet is creating domUs.  But I'm now confident that this
will work as well.  :-)

The plan is then to gradually update the system to Debian testing and see
whether the booting / lvm / initramfs problems come back.


Regards,
 Thomas


signature.asc
Description: Digital signature

Re: [Pkg-xen-devel] oldschool xen kernel on sid

[Thomas Schwinge]

Hello!

On Wed, Oct 28, 2009 at 12:43:00AM +0100, I wrote:
 10.jpg: fstype vg0-boole-root (ext3); mount /dev/mapper/vg0-boole--root
 /root (failed -- no such file or directory?!).

I just noticed in linux-image-2.6.31-1-xen-amd64_2.6.31-1_amd64.deb's
/boot/config-2.6.31-1-xen-amd64 file: ``CONFIG_BLK_DEV_DM=m''.  Tomorrow,
I'll set that to ``y'' and redo the package.  Hopefully that helps.


Regards,
 Thomas


signature.asc
Description: Digital signature

Re: oldschool xen kernel on sid

[Thomas Schwinge]

Hello!

In an attempt to finally get a recent Xen dom0 kernel booted on this
pesky amd64 system (see
http://lists.xensource.com/archives/html/xen-users/2009-10/msg00485.html
for the (short) story), I wanted to give this one one a try:

Marco Nenciarini mnen...@prato.linux.it wrote:
 I've applied latest forward ported patches from
 
 http://code.google.com/p/gentoo-xen-kernel/downloads/list
 
 to the 2.6.31 experimental package.
 The resulting kernel works well for me (at least on amd64 arch).
 
 Fell free to reuse my little work
 http://www.prato.linux.it/~mnencia/xen/oldschool-xen-2.6.31-5.patch.bz2

I applied the patch, built the kernel, but it'll crash / reset the system
in the middle of the booting process.  The last thing I can see are some
``NET:'' messages.  How to help debugging this?  Marco, should I perhaps
give your binary packages a try?


Bastian, is any of your Xen dom0 work available somewhere?


Regards,
 Thomas


signature.asc
Description: Digital signature

Re: oldschool xen kernel on sid

[Thomas Schwinge]

Hello!

On Tue, Oct 27, 2009 at 10:32:52PM +0100, Bastian Blank wrote:
 On Tue, Oct 27, 2009 at 10:05:45PM +0100, Thomas Schwinge wrote:
  Bastian, is any of your Xen dom0 work available somewhere?
 
 Yes. http://hermes.jura.uni-tuebingen.de/~blank/debian/xen-test.

Thanks!  That one boots (don't specify console=...) -- but then also gets
dropped into the BusyBox shell due to ``No init found.'' -- the same
which I also saw with another kernel, and which independently was
reported at
http://groups.google.com/group/linux.debian.user/msg/04310801bbce6417.

At this point, nothing is mounted on /root -- it's supposed to be an ext3
filesystem based on a LVM LV.  And the very strange thing is that
/dev/mapper/ does contain only one LV device, and that is the swap LV
(254,0), but not the root filessystem one.  If I manually run ``lvm
vgchange -ay'' the missing LVs appear, root (254,1) and data (254,2).
How would I manually resume the booting process from here?

The exactly same arrangement boots without any problems with a non-Xen
kernel.  This is a up-to-date Debian testing, by the way.  Help?


Regards,
 Thomas


signature.asc
Description: Digital signature

Re: [Pkg-xen-devel] oldschool xen kernel on sid

[Thomas Schwinge]

Hello!

Low-res pictures are on http://www.thomas.schwinge.homeip.net/tmp/xen/.
Sorry, I don't have anything better at the moment, but I'll explain them
below.

On Tue, Oct 27, 2009 at 11:55:33PM +0100, Bastian Blank wrote:
 On Tue, Oct 27, 2009 at 11:15:13PM +0100, Thomas Schwinge wrote:
  At this point, nothing is mounted on /root -- it's supposed to be an ext3
  filesystem based on a LVM LV.  And the very strange thing is that
  /dev/mapper/ does contain only one LV device, and that is the swap LV
  (254,0), but not the root filessystem one.
 
 Please provide the complete log and the bootloader config.

GRUB2:

menuentry Xen 3.4, Debian GNU/Linux, Linux 2.6.31-1-xen-amd64 {
insmod lvm
insmod ext2
set root=(vg0-boole-root)
search --no-floppy --fs-uuid --set 54ccee0d-3455-43dc-af2e-e2bb3940776a
multiboot /boot/xen-3.4-amd64.gz noreboot
module /boot/vmlinuz-2.6.31-1-xen-amd64 
root=/dev/mapper/vg0-boole--root ro
module /boot/initrd.img-2.6.31-1-xen-amd64
}

03.jpg, 07.jpg: Dropped into the BusyBox shell.

08.jpg, 09.jpg: mount output (nothing on /root); ls -l /dev/mapper/ (only
control and vg0-boole--swap (254,0)); lvm vgchange -ay; ls -l
/dev/mapper/ (control, vg0-boole--data (254,2), vg0-boole--root (254,1),
vg0-boole--swap (254,0).

  If I manually run ``lvm
  vgchange -ay'' the missing LVs appear, root (254,1) and data (254,2).
  How would I manually resume the booting process from here?
 
 mount $dev /root
 CTRL-D
 
 should do it.

10.jpg: fstype vg0-boole-root (ext3); mount /dev/mapper/vg0-boole--root
/root (failed -- no such file or directory?!).

  The exactly same arrangement boots without any problems with a non-Xen
  kernel.  This is a up-to-date Debian testing, by the way.  Help?
 
 I doubt that.

11.jpg: successful boot without Xen; here the device number are
different: vg0-boole--data (254,2), vg0-boole--root (254,0),
vg0-boole--swap (254,1) -- root has minor zero instead of one, and swap
vice versa?!


Regards,
 Thomas


signature.asc
Description: Digital signature