On Thursday 06 March 2008 17:41:03 Yang, Sheng wrote:
> On Thursday 06 March 2008 16:43:18 Yang, Sheng wrote:
> > On Thursday 06 March 2008 16:06:51 Avi Kivity wrote:
> > > Yang, Sheng wrote:
> > > > Here is the updated patch. I kept 0xff because I think it's OK for
> > > > understand easily. :)
>
Yang, Sheng wrote:
> Found more complex for KVM. Xen pulled pm timer down to kernel part, and used
> the guest TSC as source. So only adjust TSC is OK for it. But we are still
> using pm timer in QEmu, which using host time as source. So even we pull back
> TSC, the problem still exists, for 2.6
Greg KH wrote:
> and is on my TODO list, slowly getting
> closer to the top...
>
>
Strange. On my TODO list, things slowly get pushed to the bottom.
--
Any sufficiently difficult bug is indistinguishable from a feature.
--
Glauber Costa wrote:
> Avi,
>
> Hope this is better
>
Applied, thanks.
--
Any sufficiently difficult bug is indistinguishable from a feature.
-
This SF.net email is sponsored by: Microsoft
Defy all challenges. Microsoft
Glauber Costa wrote:
> This patch writes 0 (actually, what really matters is that the
> LSB is cleared) to the system time msr before rebooting/shutting down
> the machine.
>
> Without it, we can have a random memory location being written
> when the guest comes back
>
> Signed-off-by: Glauber Cost
On Friday 07 March 2008 16:53:40 Avi Kivity wrote:
> Yang, Sheng wrote:
> > Found more complex for KVM. Xen pulled pm timer down to kernel part, and
> > used the guest TSC as source. So only adjust TSC is OK for it. But we are
> > still using pm timer in QEmu, which using host time as source. So ev
Glauber Costa wrote:
> This patch writes 0 (actually, what really matters is that the
> LSB is cleared) to the system time msr before rebooting/shutting down
> the machine.
>
> Without it, we can have a random memory location being written
> when the guest comes back
> if (!kvm_para_available
Yang, Sheng wrote:
> On Friday 07 March 2008 16:53:40 Avi Kivity wrote:
>
>> Yang, Sheng wrote:
>>
>>> Found more complex for KVM. Xen pulled pm timer down to kernel part, and
>>> used the guest TSC as source. So only adjust TSC is OK for it. But we are
>>> still using pm timer in QEmu, whi
Hello Avi,
in commit fb56dbb31c4738a3918db81fd24da732ce3b4ae6 you changed
include/linux/Kbuild:
snip
KVM: Export include/linux/kvm.h only if $ARCH actually supports KVM
Currently, make headers_check barfs due to , which
includes, not existing. Rather than add a zillion s, export
kvm.h
Hi
Here is the latest in kernel PIT patch. Not much change from last edition.
One known issue is on 2.6.9 pae guest(e.g. RHEL4), you need "clock=pit" kernel
parameter to get the correct time. That's because the kernel is too active
to "fix the lost interrupt" when PIT interrupts pending... We
From 0e5d4fad7a6a917232d89afb58b960f6951990de Mon Sep 17 00:00:00 2001
From: Sheng Yang <[EMAIL PROTECTED]>
Date: Sun, 27 Jan 2008 11:25:29 +0800
Subject: [PATCH] kvm: libkvm: Add Supporting for in-kernel PIT model
Signed-off-by: Sheng Yang <[EMAIL PROTECTED]>
---
kernel/Kbuild |2 +-
From 6ff60b78c3280505d84d0dc2619e95a087b88458 Mon Sep 17 00:00:00 2001
From: Sheng Yang <[EMAIL PROTECTED]>
Date: Fri, 7 Mar 2008 19:03:52 +0800
Subject: [PATCH] kvm: libkvm: Add interface for PIT save/restore supporting
Signed-off-by: Sheng Yang <[EMAIL PROTECTED]>
---
libkvm/libkvm-x86.c | 3
From 98543bb3c3821e5bc9003bb91d7d0c755394ffac Mon Sep 17 00:00:00 2001
From: Sheng Yang <[EMAIL PROTECTED]>
Date: Fri, 7 Mar 2008 14:24:32 +0800
Subject: [PATCH] kvm: qemu: Add option for enable/disable in kernel PIT
Signed-off-by: Sheng Yang <[EMAIL PROTECTED]>
---
libkvm/libkvm-x86.c |9 ++
From 1af4bc979495e9e51b67635d5a9890c559e31078 Mon Sep 17 00:00:00 2001
From: Sheng Yang <[EMAIL PROTECTED]>
Date: Fri, 7 Mar 2008 19:13:06 +0800
Subject: [PATCH] kvm: qemu: Add save/restore support for in kernel PIT
Signed-off-by: Sheng Yang <[EMAIL PROTECTED]>
---
qemu/hw/i8254.c | 73
++
From 19bc8000b0ed1c2021ddb509a3d923e1cd8d53ec Mon Sep 17 00:00:00 2001
From: Sheng Yang <[EMAIL PROTECTED]>
Date: Mon, 28 Jan 2008 05:10:22 +0800
Subject: [PATCH] KVM: In kernel PIT model
The patch moved PIT from userspace to kernel, and increase the timer accuracy
greatly.
Signed-off-by: Sheng
From a16353c6fd51d5057d748198ba24d272be73d86b Mon Sep 17 00:00:00 2001
From: Sheng Yang <[EMAIL PROTECTED]>
Date: Tue, 4 Mar 2008 00:50:59 +0800
Subject: [PATCH] KVM: Add save/restore supporting of in kernel PIT
Signed-off-by: Sheng Yang <[EMAIL PROTECTED]>
---
arch/x86/kvm/i8254.c |7 +
Christian Borntraeger wrote:
> Hello Avi,
>
> in commit fb56dbb31c4738a3918db81fd24da732ce3b4ae6 you changed
> include/linux/Kbuild:
> snip
> KVM: Export include/linux/kvm.h only if $ARCH actually supports KVM
> Currently, make headers_check barfs due to , which
> includes, not existing.
Am Freitag, 7. März 2008 schrieb Avi Kivity:
> As I'm about to disappear for a week, consider a patch to remove the
> config dependency and add asm-*/kvm.h pre-acked for mainline. Maybe the
> presence of those empty asm-*/kvm.h files will encourage further kvm
> ports to *.
Something like the
Avi Kivity wrote:
> Glauber Costa wrote:
>> This patch writes 0 (actually, what really matters is that the
>> LSB is cleared) to the system time msr before rebooting/shutting down
>> the machine.
>>
>> Without it, we can have a random memory location being written
>> when the guest comes back
>>
>>
Avi Kivity wrote:
> Glauber Costa wrote:
>> This patch writes 0 (actually, what really matters is that the
>> LSB is cleared) to the system time msr before rebooting/shutting down
>> the machine.
>>
>> Without it, we can have a random memory location being written
>> when the guest comes back
>>
* [EMAIL PROTECTED] <[EMAIL PROTECTED]> [2008-03-06 18:03]:
> On Thu, Mar 6, 2008 at 6:19 PM, Ryan Harper <[EMAIL PROTECTED]> wrote:
>
> >
> > (dethklok) kvm-63 % ./kvm_stat
> > Please mount debugfs ('mount -t debugfs debugfs /sys/kernel/debug')
> > and ensure the kvm modules are loaded
> > (dethk
On Tue, Mar 04, 2008 at 02:35:21PM -0800, Christoph Lameter wrote:
> It is the atomic dead end that we want to avoid. And your patch is exactly
> that. Both the invalidate_page and the RCU locks us into this.
I preferred to answer with code to avoid any possible misunderstanding
(I through alread
This combines the non-sleep-capable RCU locking of #v9 with a seqlock
so the mmu notifier fast path will require zero cacheline
writes/bouncing while still providing mmu_notifier_unregister and
allowing to schedule inside the mmu notifier methods. If we drop
mmu_notifier_unregister we can as well d
This is a rediff of Christoph's plain i_mmap_lock2rwsem patch on top
of #v9 1/4 + 2/4 + 3/4 (hence this is called 4/4). This is mostly to
show that after 3/4, any patch that plugs on the EMM patchset will
plug nicely on top of my MMU notifer patchset too.
The patch trigger bug checks here in modpr
This patch a llows machine_crash_shutdown to
be replaced, just like any of the other functions
in machine_ops
Signed-off-by: Glauber Costa <[EMAIL PROTECTED]>
---
arch/x86/kernel/crash.c |3 ++-
arch/x86/kernel/reboot.c |7 ++-
include/asm-x86/reboot.h |1 +
3 files changed, 9 in
Avi,
I tracked down the kexec paths that requires overloading of machine_ops to two.
So here's a simpler version, that'll probably be a best fit.
thanks
-
This SF.net email is sponsored by: Microsoft
Defy all challenges. M
This patch writes 0 (actually, what really matters is that the
LSB is cleared) to the system time msr before shutting down
the machine for kexec.
Without it, we can have a random memory location being written
when the guest comes back
It overrides the functions shutdown, used in the path of kerne
Glauber Costa wrote:
>>
>> Why not go all the way and to _restart the same way?
>>
> Because it got a parameter, and doing it in the same macro would make
> my beautiful macros ugly.
> Using another one, to pass the argument, didn't seem justifiable to
> me, since there were just one of its kind.
Glauber Costa wrote:
> as for kexec, it uses precisely the shutdown function, doesn't it?
>
> Or is it crash_shutdown?
> Humm, /me looks, and I think it's the later, right?
>
Only on crash-triggered kexecs. It can also happen via sys_reboot().
Which, it appears, goes through machine_shutdown().
Christian Borntraeger wrote:
> Am Freitag, 7. März 2008 schrieb Avi Kivity:
>
>> As I'm about to disappear for a week, consider a patch to remove the
>> config dependency and add asm-*/kvm.h pre-acked for mainline. Maybe the
>> presence of those empty asm-*/kvm.h files will encourage further
Avi Kivity wrote:
> Glauber Costa wrote:
>> as for kexec, it uses precisely the shutdown function, doesn't it?
>>
>> Or is it crash_shutdown?
>> Humm, /me looks, and I think it's the later, right?
>>
>
> Only on crash-triggered kexecs. It can also happen via sys_reboot().
> Which, it appears, g
Yang, Sheng wrote:
> Hi
>
> Here is the latest in kernel PIT patch. Not much change from last edition.
>
> One known issue is on 2.6.9 pae guest(e.g. RHEL4), you need "clock=pit"
> kernel
> parameter to get the correct time. That's because the kernel is too active
> to "fix the lost interrupt"
Hi all,
As far as I see I can configure 4 cpu´s in a guest (using the -smp option)
but this seems to be only virtual, means the process on the host only uses
one physical cpu.
Background: I want to run just one guest, using the full cpu power of a host
(quadcore). Now I only get 25 % cpu
KVM fully uses all 4 CPUs.
1. You need to make sure KVM is activated (not Qemu).
2. You need to have the guest workload to be multi-threaded.
--
-Alexey Eremenko "Technologov"
-
This SF.net email is sponsored by: Microsoft
Spray your juice deeper into her.-
This SF.net email is sponsored by: Microsoft
Defy all challenges. Microsoft(R) Visual Studio 2008.
http://clk.atdmt.com/MRT/go/vse012070mrt/direct/01/__
On Fri, Mar 07, 2008 at 05:52:42PM +0100, Peter Zijlstra wrote:
> hlist_del_rcu(&mn->hlist)
>
> > + rcu_read_unlock();
>
> kfree(mn);
>
> > young |= mn->ops->clear_flush_young(mn, mm, address);
>
> *BANG*
My objective was to allow mmu_notifier_register/unr
On Fri, Mar 07, 2008 at 07:01:35PM +0100, Peter Zijlstra wrote:
> The reason Christoph can do without RCU is because he doesn't allow
> unregister, and as soon as you drop that you'll end up with something
Not sure to follow, what do you mean "he doesn't allow"? We'll also
have to rip unregister r
On Wed, Mar 05, 2008 at 04:22:11PM -0800, Christoph Lameter wrote:
> + if (e->callback) {
> + x = e->callback(e, mm, op, start, end);
> + if (x)
> + return x;
[..]
> +
> + if (emm_notify(mm, emm_referenced, address,
Glauber Costa wrote:
> This patch a llows machine_crash_shutdown to
> be replaced, just like any of the other functions
> in machine_ops
>
>
er, against what tree is this? doesn't apply to kvm.git.
--
Any sufficiently difficult bug is indistinguishable from a feature.
--
2008/3/5, Zdenek Kabelac <[EMAIL PROTECTED]>:
> 2008/3/5, Avi Kivity <[EMAIL PROTECTED]>:
>
> > Andi Kleen wrote:
> > > Avi Kivity <[EMAIL PROTECTED]> writes:
> > >
> > >> Most likely movs emulation is broken for long counts. Please post a
> > >> disassembly of copy_user_generic_string to
Avi Kivity wrote:
> Glauber Costa wrote:
>> This patch a llows machine_crash_shutdown to
>> be replaced, just like any of the other functions
>> in machine_ops
>>
>>
> er, against what tree is this? doesn't apply to kvm.git.
>
It'd be kvm.git with the machine_ops non-static functions patch.
Ho
On Fri, Mar 07, 2008 at 07:45:52PM +0100, Andrea Arcangeli wrote:
> On Fri, Mar 07, 2008 at 07:01:35PM +0100, Peter Zijlstra wrote:
> > The reason Christoph can do without RCU is because he doesn't allow
> > unregister, and as soon as you drop that you'll end up with something
>
> Not sure to foll
On Fri, 7 Mar 2008, Andrea Arcangeli wrote:
> This below simple patch invalidates the "invalidate_page" part, the
> next patch will invalidate the RCU part, and btw in a way that doesn't
> forbid unregistering the mmu notifiers at runtime (like your brand new
> EMM does).
Sounds good.
> The reas
On Fri, 7 Mar 2008, Andrea Arcangeli wrote:
> This combines the non-sleep-capable RCU locking of #v9 with a seqlock
> so the mmu notifier fast path will require zero cacheline
> writes/bouncing while still providing mmu_notifier_unregister and
> allowing to schedule inside the mmu notifier methods
On Fri, 7 Mar 2008, Andrea Arcangeli wrote:
> I didn't look into this but it shows how it would be risky to make
> this change in .25. It's a bit strange that the bugcheck triggers
Yes this was never intended for .25. I think we need to split this into a
copule of patches. One needs to get rid o
On Fri, 7 Mar 2008, Andrea Arcangeli wrote:
> In the meantime I've also been thinking that we could need the
> write_seqlock in mmu_notifier_register, to know when to restart the
> loop if somebody does a mmu_notifier_register;
> synchronize_rcu(). Otherwise there's no way to be sure the mmu
> not
On Fri, 7 Mar 2008, Andrea Arcangeli wrote:
> PS. this problem I pointed out of _end possibly called before _begin
> is the same for #v9 and EMM V1 as far as I can tell.
Hmmm.. We could just push that on the driver saying that is has to
tolerate it. Otherwise how can we solve this?
---
On Fri, 7 Mar 2008, Andrea Arcangeli wrote:
> This is a replacement for the previously posted 3/4, one of the pieces
> to allow the mmu notifier methods to sleep.
Looks good. That is what we talked about last week. What guarantees now
that we see the cacheline referenced after the cacheline that
On Fri, 2008-03-07 at 13:26 +0100, Christian Borntraeger wrote:
> +unifdef-$(CONFIG_HAVE_KVM) += kvm.h
> unifdef-y += llc.h
> unifdef-y += loop.h
> snip--
>
> This patch does not work. Kbuild (scripts/Makefile.headersinst) does
> not check the config file, so kvm.h is never installed.
>
On Thursday 06 March 2008 22:45:44 Glauber Costa wrote:
> Use LSB of the address passed through the msr to enable/disable
> the clock. Setting it to 1 enables it, setting it to 0 disables it.
>
> As the guest data structures are aligned anyway, this
> won't be a problem, as this bit is free.
>
> Gu
中层经理系统化管理 成就企业目标(证书班2天)
深圳 3月29-30日、4月26-27日、5月31-6月1日 深圳金百合大酒店
上海 4月12-13日、5月17-18日、6月14-15日 上海新梅华东大酒店
━━━
培训对象：
企业副总、各部门经理、主管、各级中层管理人员、新提拔的、从专业人才转型到管理的、
进一步想提高管理绩效的、晋升到高层管理以及其它预备管理人员
培训证书：
培训合格者颁发由香港光华管理学院签署的《职业经理人培训证书》
主办单
51 matches
Mail list logo
