Second release candidate for Xen 4.0.0

Second release candidate for Xen 4.0.0

[Keir Fraser]

Folks,

RC2 is tagged:
http://xenbits.xensource.com/xen-unstable.hg tagged 4.0.0-rc2

A fair few issues reported on RC1 are fixed. Please test!

 -- Keir

Re: Second release candidate for Xen 4.0.0

[Gilberto Nunes]

[-- Attachment #1.1: Type: text/plain, Size: 579 bytes --]

Hi Keir

Any updte to remus code?

Thanks
Em Ter 26 Jan 2010, às 12:27:39, Keir Fraser escreveu:
> Folks,
> 
> RC2 is tagged:
> http://xenbits.xensource.com/xen-unstable.hg tagged 4.0.0-rc2
> 
> A fair few issues reported on RC1 are fixed. Please test!
> 
>  -- Keir
> 
> 
> 
> _______________________________________________
> Xen-devel mailing list
> Xen-devel@lists.xensource.com
> http://lists.xensource.com/xen-devel
> 

-- 

Gilberto Nunes Ferreira

Selbetti Gestão de Documentos
Suporte TI

Telefone: (47) 3441-6004
Celular: (47) 8861-6672

[-- Attachment #1.2: Type: text/html, Size: 5815 bytes --]

[-- Attachment #2: Type: text/plain, Size: 138 bytes --]

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

Re: Second release candidate for Xen 4.0.0

[Keir Fraser]

No. I think the remus maintainer tried to repro your problems and failed to?
That would make it unlikely that any issues will be resolved in your setup
by this new release candidate.

 -- Keir

On 26/01/2010 15:20, "Gilberto Nunes" <gilberto.nunes@selbetti.com.br>
wrote:

> Hi Keir
> 
> Any updte to remus code?
> 
> Thanks
> 
> Em Ter 26 Jan 2010, às 12:27:39, Keir Fraser escreveu:
> 
>> Folks,
> 
>> 
> 
>> RC2 is tagged:
> 
>> http://xenbits.xensource.com/xen-unstable.hg tagged 4.0.0-rc2
> 
>> 
> 
>> A fair few issues reported on RC1 are fixed. Please test!
> 
>> 
> 
>>  -- Keir
> 
>> 
> 
>> 
> 
>> 
> 
>> _______________________________________________
> 
>> Xen-devel mailing list
> 
>> Xen-devel@lists.xensource.com
> 
>> http://lists.xensource.com/xen-devel
> 
>> 
> 

Re: Second release candidate for Xen 4.0.0

[Ferenc Wagner]

Keir Fraser <keir.fraser@eu.citrix.com> writes:

> A fair few issues reported on RC1 are fixed. Please test!

I'd like to...

git clone git://git.kernel.org/pub/scm/linux/kernel/git/jeremy/xen.git linux-2.6-pvops.git.tmp
Initialized empty Git repository in /home/wferi/xen/xen-unstable.hg/linux-2.6-pvops.git.tmp/.git/
remote: error: refs/remotes/tip/core/header-fixes does not point to a valid object!
remote: error: refs/remotes/tip/header-fixes-for-linus does not point to a valid object!
remote: error: refs/remotes/tip/master does not point to a valid object!
remote: error: refs/remotes/tip/sched-fixes-for-linus does not point to a valid object!
remote: error: refs/remotes/tip/sched/urgent does not point to a valid object!
remote: error: Could not read 9e02a0dd1311b2126c55f1f5487468d38d05f540
remote: fatal: Failed to traverse parents of commit 06adbef163e61b0ec4c5aremote: 96534120eb3b59948f7
remote: aborting due to possible repository corruption on the remote side.
fatal: early EOF
fatal: index-pack failed

Any idea?
-- 
Thanks,
Feri.

Re: Re: Second release candidate for Xen 4.0.0

[Keir Fraser]

On 27/01/2010 13:31, "Ferenc Wagner" <wferi@niif.hu> wrote:

> Keir Fraser <keir.fraser@eu.citrix.com> writes:
>
> git clone git://git.kernel.org/pub/scm/linux/kernel/git/jeremy/xen.git
> linux-2.6-pvops.git.tmp
> ...
> fatal: early EOF
> fatal: index-pack failed
> 
> Any idea?

Hm, yes it appears to be down or broken. It's probably temporary, but maybe
Jeremy can advise. It mighrt be worthwhile moving the tree onto our own
xenbits server, which is where all our other build prerequisites live.

For now, you can actually use your old kernel: there's no need to build a
new one as it's not part of 4.0.0-rc2 anyway. Instead of 'make world' you
can do 'cd xen; make install; cd ../tools; make install' to
rebuild/reinstall only hypervisor and dom0 tools.

 -- Keir

Re: Second release candidate for Xen 4.0.0

[Ferenc Wagner]

Keir Fraser <keir.fraser@eu.citrix.com> writes:

> On 27/01/2010 13:31, "Ferenc Wagner" <wferi@niif.hu> wrote:
>
>> Keir Fraser <keir.fraser@eu.citrix.com> writes:
>>
>> git clone git://git.kernel.org/pub/scm/linux/kernel/git/jeremy/xen.git
>> linux-2.6-pvops.git.tmp
>> ...
>> fatal: early EOF
>> fatal: index-pack failed
>> 
>> Any idea?
>
> Hm, yes it appears to be down or broken. It's probably temporary, but maybe
> Jeremy can advise.

It's indeed fixed now, I get further, but still not quite through:

make[5]: Entering directory `/home/wferi/xen/xen-unstable.hg/tools/ioemu-remote/i386-dm'
../xen-hooks.mak:56: === pciutils-dev package not found - missing /usr/include/pci
../xen-hooks.mak:57: === PCI passthrough capability has been disabled
../xen-hooks.mak:56: === pciutils-dev package not found - missing /usr/include/pci
../xen-hooks.mak:57: === PCI passthrough capability has been disabled
  CC    i386-dm/vl.o
  CC    i386-dm/monitor.o
[...]
  CC    i386-dm/pci_emulation.o
  CC    i386-dm/helper2.o
helper2.c: In function ‘cpu_x86_init’:
helper2.c:142: error: ‘shared_iopage_t’ has no member named ‘vcpu_iodata’
helper2.c: In function ‘sp_info’:
helper2.c:219: error: ‘shared_iopage_t’ has no member named ‘vcpu_iodata’
helper2.c:224: warning: format ‘%llx’ expects type ‘long long unsigned int’, but argument 6 has type ‘uint32_t’
helper2.c:224: warning: format ‘%llx’ expects type ‘long long unsigned int’, but argument 7 has type ‘uint32_t’
helper2.c:226: error: ‘ioreq_t’ has no member named ‘io_count’
helper2.c: In function ‘__cpu_get_ioreq’:
helper2.c:235: error: ‘shared_iopage_t’ has no member named ‘vcpu_iodata’
helper2.c:242: warning: format ‘%llx’ expects type ‘long long unsigned int’, but argument 7 has type ‘uint32_t’
helper2.c:242: warning: format ‘%llx’ expects type ‘long long unsigned int’, but argument 8 has type ‘uint32_t’
helper2.c: In function ‘cpu_handle_ioreq’:
helper2.c:517: warning: format ‘%llx’ expects type ‘long long unsigned int’, but argument 7 has type ‘uint32_t’
helper2.c:517: warning: format ‘%llx’ expects type ‘long long unsigned int’, but argument 8 has type ‘uint32_t’
make[5]: *** [helper2.o] Error 1

Installing the libpci-dev does not help, as expected.
I'd be grateful for further hints, I'm at 20858:2636e5619708.
-- 
Thanks,
Feri.

Re: Second release candidate for Xen 4.0.0

[Ferenc Wagner]

Ferenc Wagner <wferi@niif.hu> writes:

> Keir Fraser <keir.fraser@eu.citrix.com> writes:
>
>> On 27/01/2010 13:31, "Ferenc Wagner" <wferi@niif.hu> wrote:
>>
>>> Keir Fraser <keir.fraser@eu.citrix.com> writes:
>>>
>>> git clone git://git.kernel.org/pub/scm/linux/kernel/git/jeremy/xen.git
>>> linux-2.6-pvops.git.tmp
>>> ...
>>> fatal: early EOF
>>> fatal: index-pack failed
>>> 
>>> Any idea?
>>
>> Hm, yes it appears to be down or broken. It's probably temporary, but maybe
>> Jeremy can advise.
>
> It's indeed fixed now, I get further, but still not quite through:

Btw. is it intentional that the kernel configuration phase is
interactive?  After make world, I've got to answer the following
questions:

  Lguest guest support (LGUEST_GUEST) [N/y/?] (NEW) 
    Paravirtualization layer for spinlocks (PARAVIRT_SPINLOCKS) [N/y/?] (NEW) y
Xen PCI Frontend (XEN_PCIDEV_FRONTEND) [N/m/y/?] (NEW) m
    SCSI tape drive support for Smart Array 5xxx (CISS_SCSI_TAPE) [N/y/?] (NEW) 
  Fusion MPT ScsiHost drivers for FC (FUSION_FC) [N/m/y/?] (NEW) m
  Maximum number of scatter gather entries (16 - 128) (FUSION_MAX_SGE) [128] (NEW) 
  Fusion MPT misc device (ioctl) driver (FUSION_CTL) [N/m/?] (NEW) m
  Fusion MPT logging facility (FUSION_LOGGING) [N/y/?] (NEW) 
  Net channel 2 support (XEN_NETCHANNEL2) [N/m/y/?] (NEW) y
    Net channel 2 backend support (XEN_NETDEV2_BACKEND) [Y/n] (NEW) 
    Net channel 2 frontend support (XEN_NETDEV2_FRONTEND) [Y/n] (NEW) 
  Net channel 2 bypassee support (XEN_NETDEV2_BYPASSABLE) [Y/n/?] (NEW) 
  Net channel 2 bypass endpoint support (XEN_NETDEV2_BYPASS_ENDPOINT) [Y/n/?] (NEW) 
    Automatically manage netchannel2 bypasses (XEN_NETDEV2_AUTOMATIC_BYPASS) [Y/n/?] (NEW) 
  Block-device tap backend driver (XEN_BLKDEV_TAP) [N/m/y/?] (NEW) y
  PCI-device backend driver (XEN_PCIDEV_BACKEND) [Y/n/m/?] (NEW) 
    PCI Backend Mode
    > 1. Virtual PCI (XEN_PCIDEV_BACKEND_VPCI) (NEW)
      2. Passthrough (XEN_PCIDEV_BACKEND_PASS) (NEW)
      3. Slot (XEN_PCIDEV_BACKEND_SLOT) (NEW)
    choice[1-3]: 1
-- 
Thanks,
Feri.

Re: Re: Second release candidate for Xen 4.0.0

[Ian Campbell]

On Wed, 2010-02-03 at 22:32 +0000, Ferenc Wagner wrote:
> Btw. is it intentional that the kernel configuration phase is
> interactive?  After make world, I've got to answer the following
> questions: 

The pvops kernel build integration into xen-unstable.hg uses the
upstream kernel's defconfig (which may not always be complete to start
with) tweaked to enable Xen. If new config options have been added then
they need to be added to xen-unstable.hg/buildconfigs/enable-xen-config

Ian.

Re: Re: Second release candidate for Xen 4.0.0

[Keir Fraser]

I'd say your qemu is out of date. Find the qemu source tree (e.g.,
../qemu-xen-unstable.git), cd into it, and 'git pull -a' to update it to
latest.

 -- Keir

On 03/02/2010 18:43, "Ferenc Wagner" <wferi@niif.hu> wrote:

> helper2.c: In function Œcpu_x86_init¹:
> helper2.c:142: error: Œshared_iopage_t¹ has no member named Œvcpu_iodata¹
> helper2.c: In function Œsp_info¹:
> helper2.c:219: error: Œshared_iopage_t¹ has no member named Œvcpu_iodata¹
> helper2.c:224: warning: format Œ%llx¹ expects type Œlong long unsigned int¹,
> but argument 6 has type Œuint32_t¹
> helper2.c:224: warning: format Œ%llx¹ expects type Œlong long unsigned int¹,
> but argument 7 has type Œuint32_t¹
> helper2.c:226: error: Œioreq_t¹ has no member named Œio_count¹
> helper2.c: In function Œ__cpu_get_ioreq¹:
> helper2.c:235: error: Œshared_iopage_t¹ has no member named Œvcpu_iodata¹
> helper2.c:242: warning: format Œ%llx¹ expects type Œlong long unsigned int¹,
> but argument 7 has type Œuint32_t¹
> helper2.c:242: warning: format Œ%llx¹ expects type Œlong long unsigned int¹,
> but argument 8 has type Œuint32_t¹
> helper2.c: In function Œcpu_handle_ioreq¹:
> helper2.c:517: warning: format Œ%llx¹ expects type Œlong long unsigned int¹,
> but argument 7 has type Œuint32_t¹
> helper2.c:517: warning: format Œ%llx¹ expects type Œlong long unsigned int¹,
> but argument 8 has type Œuint32_t¹
> make[5]: *** [helper2.o] Error 1
> 
> Installing the libpci-dev does not help, as expected.
> I'd be grateful for further hints, I'm at 20858:2636e5619708.

Re: Second release candidate for Xen 4.0.0

[Ferenc Wagner]

Keir Fraser <keir.fraser@eu.citrix.com> writes:

> On 03/02/2010 18:43, "Ferenc Wagner" <wferi@niif.hu> wrote:
>
>> helper2.c: In function Œcpu_x86_init¹:
>> helper2.c:142: error: Œshared_iopage_t¹ has no member named Œvcpu_iodata¹
>> [...]
>> make[5]: *** [helper2.o] Error 1
>
> I'd say your qemu is out of date. Find the qemu source tree (e.g.,
> ../qemu-xen-unstable.git), cd into it, and 'git pull -a' to update it to
> latest.

Ok.  I hg updated, then git pulled (forgot the -a, hope it won't strike
back) and that let the build finish successfully!  I'm very surprised
that the resulting 2.6.31.6 kernel has almost everything built-in, but
that's a different story, looks like I can test this stuff soon. :)
-- 
Thanks for the support!
Feri.