[Xen-ia64-devel] RE: Réf. : RE: How is the basic tree coming? SMP support on Itanium

["Magenheimer, Dan (HP Labs Fort Collins)" <dan.magenheimer@xxxxxx>]

Tres bien!

Please do post all progress to the list.  I know there are several people

(including myself) that are eager to contribute to SMP and it would be

good to avoid duplicating effort.

So don't wait until SMP is solid to post a patch.  Messages like the

following would be great:

 

"I got the AP (application processor) to receive an IPI and it started

booting but died at xxx."

Thanks,

Dan

---

From: jean-paul.pigache@xxxxxxxx [mailto:jean-paul.pigache@xxxxxxxx]
Sent: Monday, September 05, 2005 9:50 AM
To: Magenheimer, Dan (HP Labs Fort Collins)
Cc: Tian, Kevin; xen-ia64-devel@xxxxxxxxxxxxxxxxxxx; xen-ia64-devel-bounces@xxxxxxxxxxxxxxxxxxx
Subject: Réf. : RE: How is the basic tree coming? SMP support on Itanium

Dan,

I want to let you know that Bull is interested by the SMP support for Itanium platforms.
We have started to work on the support of "SMP host" on Itanium VT multiprocessor systems.
We will of course communicate with you (the Xen IA64 man) through this mailing list for a good synchronization with other developments, and mainly with the SMP support in the paravirtualized Linux and the support of SMP guest.

Regards
Jean-Paul

"Magenheimer, Dan (HP Labs Fort Collins)" <dan.magenheimer@xxxxxx> Envoyé par : xen-ia64-devel-bounces@xxxxxxxxxxxxxxxxxxx 01/09/2005 02:47

Pour :        "Tian, Kevin" <kevin.tian@xxxxxxxxx>         cc :        xen-ia64-devel@xxxxxxxxxxxxxxxxxxx         Objet :        [Xen-ia64-devel] RE: How is the basic tree coming?

Sorry... 99% of the changes I've committed have been
maintenance, namely getting Xen/ia64 to compile with
CONFIG_SMP (which can't even be turned off on Xen/x86)
and catching the code up to the latest linux release.
I wasn't intending to add new functionality or impact
the multi-domain effort.  I did get curious to see how
far the SMP boot was getting and added some printk's
and missing code (that was already in Xen/x86)
but gave up on that rather quickly due to needing
to do other maintenance activities (2.6.13).  Rather
than throw away useful debugging code, I checked it in.

Last, I've been cleaning up some linux files that
were changed (without ifdef's) and doing some minor
restructuring following Arun's removal of the mkbuildtree
patch process.

I still have some more 2.6.13 work, but it should
be done tomorrow.

If any of this maintenance interfered with the multi-domain
repair, I apologize.  But it seemed like it was orthogonal
so a good time to do it.  (And if anyone else wants to
volunteer to do my boring maintenance tasks, please
let me know :-)

Dan

> -----Original Message-----
> From: Tian, Kevin [mailto:kevin.tian@xxxxxxxxx]
> Sent: Wednesday, August 31, 2005 6:31 PM
> To: Magenheimer, Dan (HP Labs Fort Collins)
> Cc: xen-ia64-devel@xxxxxxxxxxxxxxxxxxx
> Subject: How is the basic tree coming?
>
> Hi, Dan,
>                  Per your previous mail, you said:
> >>>> I would like to solicit (or maybe beg!) for your help
> >>>> in bug-hunting and bug-fixing to bring xen-ia64 back
> >>>> to the previous functionality (e.g. multiple domains
> >>>> working with grant tables, console and blk drivers).
> >>>> Some of you may have participated in implementation
> >>>> and debugging of the equivalent driver code on the x86 side
> >>>> and your help especially would be greatly appreciated.
> >>>> I see this as a community effort...
> ################################
> >>>>I do not expect to
> >>>> accept new functionality patches until this is all
> >>>> working again.
> ###############################
> >>>>
> >>>> Thanks,
> >>>> Dan
>
>                  While xen-ia64-unstable
>
> 15 hours ago: Baby step toward host SMP support  
> changeset 6453: 44316ce83277
> parent 6452: f5c4042212b0
> tag: tip
> author: djm@xxxxxxxxxxxxxxx
> date: Tue Aug 30 23:51:51 2005
> files: xen/arch/ia64/linux-xen/mm_contig.c
> xen/arch/ia64/linux-xen/smp.c
> xen/arch/ia64/linux-xen/smpboot.c xen/arch/ia64/vcpu.c
> xen/arch/ia64/xenmisc.c xen/arch/ia64/xensetup.c  
> 20 hours ago: Enable CONFIG_SMP compile and link  
> changeset 6452: f5c4042212b0
> parent 6451: 2b95125015a5
> author: djm@xxxxxxxxxxxxxxx
> date: Tue Aug 30 18:41:54 2005
> files: xen/arch/ia64/hyperprivop.S xen/arch/ia64/irq.c
> xen/arch/ia64/linux-xen/irq_ia64.c xen/arch/ia64/linux-xen/mm_contig.c
> xen/arch/ia64/linux-xen/sal.c xen/arch/ia64/linux-xen/smp.c
> xen/arch/ia64/linux-xen/smpboot.c xen/arch/ia64/process.c
> xen/arch/ia64/xensetup.c xen/arch/ia64/xentime.c ...
>
>                  Self-contradictory? Why you are debugging MP and check in code
> that is to make our basic tree effort harder?
>
> Thanks,
> Kevin
>

_______________________________________________
Xen-ia64-devel mailing list
Xen-ia64-devel@xxxxxxxxxxxxxxxxxxx
http://lists.xensource.com/xen-ia64-devel

_______________________________________________
Xen-ia64-devel mailing list
Xen-ia64-devel@xxxxxxxxxxxxxxxxxxx
http://lists.xensource.com/xen-ia64-devel