Hi,
 I just tried building 2.6.24 with vz from git, and it builds nicely (when
I disable CPT), but trying to boot it - it fails when 'initiating processor
0' ( and machine restarts ). 
 Are such failures interesting and should be reported, or do you prefer
reports from your own testing team and external reports should be only
sent for official releases?

-- 
Key fingerprint = 40D0 9FFB 9939 7320 8294  05E0 BCC7 02C4 75CC 50D9
 Total Existance Failure

Your reports are needed very much and we very appreciate any feedback from users.

Can you please provide some more info such as .config file you used to build kernel?
Maybe something else specific you have done/configured?


Dariush Pietrzak wrote:
> Hi,
>  I just tried building 2.6.24 with vz from git, and it builds nicely (when
> I disable CPT), but trying to boot it - it fails when 'initiating processor
> 0' ( and machine restarts ). 
>  Are such failures interesting and should be reported, or do you prefer
> reports from your own testing team and external reports should be only
> sent for official releases?
>

> Can you please provide some more info such as .config file you used to build kernel?
attached and also available at http://kuszelas.eu/~eyck/config-2.6.24

> Maybe something else specific you have done/configured?
 can't recall anything exotic, with the same config and clean 2.6.24
machine boots OK.
 At first I though that the problem was with 'video=791',
but after changing this it looks like it restarts right after trying to
init processor 0. I ain't got rs232 in this machine, and I don't know what
other way could catch that boot sequence (I think netconsole initializes
way later)
 
-- 
Key fingerprint = 40D0 9FFB 9939 7320 8294  05E0 BCC7 02C4 75CC 50D9
 Total Existance Failure

[Box reboots with 2.6.24-openvz from git after
 Initiating processor 0]

Doh! It was indeed descriptors moving in GDT.

The following should fix it:

git diff v2.6.24.. -- arch/x86/boot/compressed/head_64.S arch/x86/kernel/acpi/wakeup_64.S arch/x86/kernel/head_64.S include/asm-x86/segment_64.h  >../1.patch
patch -p1 -R <../1.patch

>  Initiating processor 0]
> Doh! It was indeed descriptors moving in GDT.
> The following should fix it:
> git diff v2.6.24.. -- arch/x86/boot/compressed/head_64.S arch/x86/kernel/acpi/wakeup_64.S arch/x86/kernel/head_64.S include/asm-x86/segment_64.h  >../1.patch
> patch -p1 -R <../1.patch
 Yes, thanks, now it boots, and starts spewing:


Driver 'sr' needs updating - please use bus_type methods
sysfs: duplicate filename ':0000288' can not be created
WARNING: at fs/sysfs/dir.c:424 sysfs_add_one()
Pid: 9, comm: events/0 Not tainted 2.6.24 #3

Call Trace:
 [<ffffffff802e98a0>] sysfs_add_one+0x54/0xbd
 [<ffffffff802e9ddb>] create_dir+0x4f/0x87
 [<ffffffff802e9f95>] sysfs_create_dir+0x47/0x5c
 [<ffffffff803069f1>] kobject_get+0x12/0x17
 [<ffffffff80306f2b>] kobject_add+0xf3/0x1a7
 [<ffffffff802a0d81>] sysfs_slab_add+0x122/0x16c
 [<ffffffff802a0dcb>] sysfs_add_func+0x0/0x59
 [<ffffffff802a0dfc>] sysfs_add_func+0x31/0x59
 [<ffffffff80247141>] run_workqueue+0x7f/0x10b
 [<ffffffff80247a5d>] worker_thread+0x0/0xe4
sr0: scsi3-mmc drive: 24x/24x writer dvd-ram cd/rw xa/form2 cdda tray
Uniform CD-ROM driver Revision: 3.20
 [<ffffffff80247b37>] worker_thread+0xda/0xe4
 [<ffffffff8024b4fa>] autoremove_wake_function+0x0/0x2e
 [<ffffffff8024b3da>] kthread+0x47/0x75
 [<ffffffff8020cd78>] child_rip+0xa/0x12
 [<ffffffff8024b393>] kthread+0x0/0x75
 [<ffffffff8020cd6e>] child_rip+0x0/0x12

kobject_add failed for :0000288 with -EEXIST, don't try to register things with the same name in the same directory.
sr 3:0:0:0: Attached scsi CD-ROM sr0
Pid: 9, comm: events/0 Not tainted 2.6.24 #3

Call Trace:
 [<ffffffff80306faa>] kobject_add+0x172/0x1a7
 [<ffffffff802a0d81>] sysfs_slab_add+0x122/0x16c
 [<ffffffff802a0dcb>] sysfs_add_func+0x0/0x59
 [<ffffffff802a0dfc>] sysfs_add_func+0x31/0x59
 [<ffffffff80247141>] run_workqueue+0x7f/0x10b
 [<ffffffff80247a5d>] worker_thread+0x0/0xe4
 [<ffffffff80247b37>] worker_thread+0xda/0xe4
 [<ffffffff8024b4fa>] autoremove_wake_function+0x0/0x2e
 [<ffffffff8024b3da>] kthread+0x47/0x75
 [<ffffffff8020cd78>] child_rip+0xa/0x12
 [<ffffffff8024b393>] kthread+0x0/0x75
 [<ffffffff8020cd6e>] child_rip+0x0/0x12

sysfs: duplicate filename ':0004096' can not be created
WARNING: at fs/sysfs/dir.c:424 sysfs_add_one()
Pid: 9, comm: events/0 Not tainted 2.6.24 #3

....

 [<ffffffff802e98a0>] sysfs_add_one+0x54/0xbd
 [<ffffffff802e9ddb>] create_dir+0x4f/0x87
 [<ffffffff802e9f95>] sysfs_create_dir+0x47/0x5c
 [<ffffffff803069f1>] kobject_get+0x12/0x17
 [<ffffffff80306f2b>] kobject_add+0xf3/0x1a7
 [<ffffffff802a0d81>] sysfs_slab_add+0x122/0x16c
 [<ffffffff802a0dcb>] sysfs_add_func+0x0/0x59
 [<ffffffff802a0dfc>] sysfs_add_func+0x31/0x59
 [<ffffffff80247141>] run_workqueue+0x7f/0x10b
 [<ffffffff80247a5d>] worker_thread+0x0/0xe4
 [<ffffffff80247b37>] worker_thread+0xda/0xe4
 [<ffffffff8024b4fa>] autoremove_wake_function+0x0/0x2e
 [<ffffffff8024b3da>] kthread+0x47/0x75
 [<ffffffff8020cd78>] child_rip+0xa/0x12
 [<ffffffff8024b393>] kthread+0x0/0x75
 [<ffffffff8020cd6e>] child_rip+0x0/0x12

kobject_add failed for :0000016 with -EEXIST, don't try to register things with the same name in the same directory.
Pid: 9, comm: events/0 Not tainted 2.6.24 #3

Call Trace:
 [<ffffffff80306faa>] kobject_add+0x172/0x1a7
 [<ffffffff802a0d81>] sysfs_slab_add+0x122/0x16c
 [<ffffffff802a0dcb>] sysfs_add_func+0x0/0x59
 [<ffffffff802a0dfc>] sysfs_add_func+0x31/0x59
 [<ffffffff80247141>] run_workqueue+0x7f/0x10b
 [<ffffffff80247a5d>] worker_thread+0x0/0xe4
 [<ffffffff80247b37>] worker_thread+0xda/0xe4
 [<ffffffff8024b4fa>] autoremove_wake_function+0x0/0x2e
 [<ffffffff8024b3da>] kthread+0x47/0x75
 [<ffffffff8020cd78>] child_rip+0xa/0x12
 [<ffffffff8024b393>] kthread+0x0/0x75
 [<ffffffff8020cd6e>] child_rip+0x0/0x12

usb 6-4: configuration #1 chosen from 1 choice


but other then this sysfs issue everything else seems to work fine...

-- 
Key fingerprint = 40D0 9FFB 9939 7320 8294  05E0 BCC7 02C4 75CC 50D9
 Total Existance Failure

On Thu, Jan 31, 2008 at 04:31:20PM +0100, Dariush Pietrzak wrote:
> >  Initiating processor 0]
> > Doh! It was indeed descriptors moving in GDT.
> > The following should fix it:
> > git diff v2.6.24.. -- arch/x86/boot/compressed/head_64.S arch/x86/kernel/acpi/wakeup_64.S arch/x86/kernel/head_64.S include/asm-x86/segment_64.h  >../1.patch
> > patch -p1 -R <../1.patch
>  Yes, thanks, now it boots, and starts spewing:
> 
> 
> Driver 'sr' needs updating - please use bus_type methods

This is fine (as in not a problem at all, just message from sysfs
developers to scsi developers).

> sysfs: duplicate filename ':0000288' can not be created
> WARNING: at fs/sysfs/dir.c:424 sysfs_add_one()

OK, I haven't seen this.

On 1/31/08, Dariush Pietrzak <ml-openvz-eyck@ddiary.eu> wrote:
> > Can you please provide some more info such as .config file you used to
> build kernel?
> attached and also available at http://kuszelas.eu/~eyck/config-2.6.24
>
> > Maybe something else specific you have done/configured?
>  can't recall anything exotic, with the same config and clean 2.6.24
> machine boots OK.

I must have messed up KERNEL_CS/DS descriptors during forward porting.
Let me recheck.

>  At first I though that the problem was with 'video=791',
> but after changing this it looks like it restarts right after trying to
> init processor 0. I ain't got rs232 in this machine, and I don't know what
> other way could catch that boot sequence (I think netconsole initializes
> way later)

Chances that you'll get anything on console in such situation are pretty slim.