Xen/IA64 Healthiness Report
A lot of testing cases failed in 11052, including VTI and XenU testing.
Testing Environment:
Platform: Tiger4
Processor: Itanium 2 Processor
Logic Processors number: 8 (2 processors with Due Core)
PAL version: 8.15
Service OS: RHEL4u3 IA64
Alex,
Due to Cset11052 remove PIN processor for domaim0 vcpu0.
That means, VTI-domain and domain may share one LP,
But, VTI-domain uses do_block to emulate halt_ligh, while domain0 uses do_yield.
This makes VTI-domain scheduled less than domain0,
and makes VTI-domain very slow, sometime crach
Le Mardi 22 Août 2006 21:03, Hollis Blanchard a écrit :
I apologize for my mailer line-wrapping the patch as I quote it below.
On Mon, 2006-08-21 at 17:18 +0200, Tristan Gingold wrote:
diff -r b7db009d622c linux-2.6-xen-sparse/drivers/xen/Kconfig
---
Le Mardi 22 Août 2006 17:50, Aron Griffis a écrit :
Alex Williamson wrote: [Tue Aug 22 2006, 10:31:55AM EDT]
What about adopting something more like what grub does and simply
provide vmm option on the vmm= line after the image file? Something
like:
image=xenlinuz
Le Lundi 21 Août 2006 11:41, Keir Fraser a écrit :
On 21/8/06 10:20 am, Tristan Gingold [EMAIL PROTECTED] wrote:
Fine in principle. Specific comments:
* powerpc should be cleaned up at the same time to use the common
infrastructure. I don't want duplicated code hanging around in
The difference from TAKE2
Status
I tested 2xDomU Linux Kernel Compile Test on 4xpCPU w/ credit scheduler.
It passes more than 16h testing.(21h in real)
Changes
1)migrate_timer for hlt_timer_fn is added for context_switch
This makes correct pCPU work for timer.
2)To avoid soft lockup
Hi Atsushi.
On Wed, Aug 23, 2006 at 05:48:15PM +0900, Atsushi SAKAI wrote:
1)migrate_timer for hlt_timer_fn is added for context_switch
This makes correct pCPU work for timer.
Is it necessary to call migrate_timer() every context switch
instead of calling it right before
Hi,
taking into account Hollis comments I now submit this patch.
IA64 specific stuff will be posted to xen-ia64-unstable after merge.
Tristan.
# HG changeset patch
# User [EMAIL PROTECTED]
# Node ID ba96f5f81464dc3db80af0454fb05c73de620637
# Parent cc7a78171c811b9caa5587a0141d8c8477d10ce8
I agree with Yamahata partially,
If I'm right, credit scheduler only schedules runnable vcpus,
So when context_switch is called, the scheduled in vcpu's hlt_timer must be
Stopped, it is unnecessary to call migrate_timer in context_switch.
My suggestion is like following,
@@ -233,7 +233,10 @@
Previously only vcpu0 can sync interrupt from qemu.
Now all vcpus can sync.
Signed-off-by: Anthony Xu [EMAIL PROTECTED]
Thanks,
Anthony
pass_virq.patch
Description: pass_virq.patch
___
Xen-ia64-devel mailing list
Xen-ia64-devel@lists.xensource.com
Hi,
After changing asm constraints, I removed _hypercall_imm*.
Checked by objdump -dr
Tested by running dom0.
Tristan.
# HG changeset patch
# User [EMAIL PROTECTED]
# Node ID 4732e6e0d6cd380ea1de0fb4d08b8efb535ad463
# Parent 91169603a8e8dded9eba6cb6c3421b5d58a85a97
Clean up: change asm
On Wed, 2006-08-23 at 21:35 +0800, Xu, Anthony wrote:
Previously only vcpu0 can sync interrupt from qemu.
Now all vcpus can sync.
Hi Anthony,
I think we should have a spin_lock_init() call in there somewhere.
Should we introduce a struct for vmx specific data in struct
arch_domain? Thanks,
Hi, Xiantao and all
It's not known exactly why, but this problem is occured
by making wrong elf header of gate.so.
The following patch fix this issue.
But this problem is not related to xen code, is related to linux code.
I look like this problem is fixed in linux-2.6.18-rc4.
So I think
I merged us up with xen-unstable.hg. For the most part things still
work well, but there seem to be a few regressions (networking on VTi
domains is broken again and rebooting domains is less reliable). I'd
like to request a pull within the next day or so, therefore please take
a moment to
On Tue, 2006-08-22 at 16:46 +0900, DOI Tsunehisa wrote:
This patch includes:
* catch up `split maddr.h'
- Split page.h into maddr.h
- This patch is like Alex's patch, but it includes preparation
for PV-on-HVM.
* catch up `new Xen event channles'
- DM over to using
On Wed, 2006-08-23 at 00:08 -0400, Aron Griffis wrote:
Remove extraneous verbose output to clean up Fedora boot.
Applied.
--
Alex Williamson HP Open Source Linux Org.
___
Xen-ia64-devel mailing list
On Wed, Aug 23, 2006 at 09:32:24PM +0800, Xu, Anthony wrote:
I agree with Yamahata partially,
If I'm right, credit scheduler only schedules runnable vcpus,
So when context_switch is called, the scheduled in vcpu's hlt_timer must be
Stopped, it is unnecessary to call migrate_timer in
Hi,
Hi, Xiantao and all
Oops. I'm mistake, I wanted to write Hi, YongKang.
I'm sorry to Xiantao and YongKang.
It's not known exactly why, but this problem is occured
by making wrong elf header of gate.so.
The following patch fix this issue.
But this problem is not related to xen code, is
Hi Akio,
Thank you for your information~ :)
I am expecting next Rawhide can boot up Xen0.
Best Regards,
Yongkang (Kangkang) 永康
-Original Message-
From: [EMAIL PROTECTED]
[mailto:[EMAIL PROTECTED] On Behalf Of Akio
Takebe
Sent: 2006年8月24日 11:52
To: ia64 Fedora Core Development
Cc:
19 matches
Mail list logo
