KVM Development List

Zachary Amsden | 1 Aug 2010 01:55

Picon

Re: 2.6.35-rc1 regression with pvclock and smp guests

On 07/31/2010 06:36 AM, Arjan Koers wrote:
> On 2010-07-31 13:53, Arjan Koers wrote:
>    
>> The kernel boots successfully when CONFIG_PRINTK_TIME is not set.
>>
>>      
> The problem occurs when this message is printed:
>
> [    0.016000] kvm-clock: cpu 1, msr 0:1511c01, secondary cpu clock
>
> When I disable that printk, the kernel boots with
> CONFIG_PRINTK_TIME=y
>
> --- a/arch/x86/kernel/kvmclock.c
> +++ b/arch/x86/kernel/kvmclock.c
>  <at>  <at>  -131,8 +131,8  <at>  <at>  static int kvm_register_clock(char *txt)
>   	int low, high;
>   	low = (int)__pa(&per_cpu(hv_clock, cpu)) | 1;
>   	high = ((u64)__pa(&per_cpu(hv_clock, cpu))>>  32);
> -	printk(KERN_INFO "kvm-clock: cpu %d, msr %x:%x, %s\n",
> -	       cpu, high, low, txt);
> +	/*printk(KERN_INFO "kvm-clock: cpu %d, msr %x:%x, %s\n",
> +	       cpu, high, low, txt);*/
>
>   	return native_write_msr_safe(msr_kvm_system_time, low, high);
>   }
>
> So the problem appears to be that the clock of the second CPU
> is used too soon (or that clock setup should finish earlier).
>

(Continue reading)

Avi Kivity | 1 Aug 2010 09:51

Picon

Re: enabling X86_FEATURE_ARCH_PERFMON in guest

  On 07/30/2010 08:02 PM, David S. Ahern wrote:
> How do I get X86_FEATURE_ARCH_PERFMON enabled for a guest?
>
> I've tried "-cpu host,+perfmon" and "-cpu host,+arch_perfmon", but both
> get rejected with an error: CPU feature perfmon not found
>

qemu doesn't know about the perfmon feature bits.  Even if it did, kvm 
doesn't, so it wouldn't work.

Jes is working on this.

--

-- 
error compiling committee.c: too many arguments to function

--
To unsubscribe from this list: send the line "unsubscribe kvm" in
the body of a message to majordomo <at> vger.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html

Gleb Natapov | 1 Aug 2010 10:28

Picon

Re: [PATCH 1/2] KVM: x86 emulator: don't update vcpu state if instruction is restarted.

On Sat, Jul 31, 2010 at 08:25:13PM +0300, Avi Kivity wrote:
>  On 07/29/2010 03:11 PM, Gleb Natapov wrote:
> >No need to update vcpu state since instruction is in the middle of the
> >emulation.
> >
> >Signed-off-by: Gleb Natapov<gleb <at> redhat.com>
> >---
> >  arch/x86/kvm/x86.c |   31 +++++++++++++------------------
> >  1 files changed, 13 insertions(+), 18 deletions(-)
> >
> >diff --git a/arch/x86/kvm/x86.c b/arch/x86/kvm/x86.c
> >index 76fbc32..7e5f075 100644
> >--- a/arch/x86/kvm/x86.c
> >+++ b/arch/x86/kvm/x86.c
> > <at>  <at>  -4057,32 +4057,27  <at>  <at>  restart:
> >  		return handle_emulation_failure(vcpu);
> >  	}
> >
> >-	toggle_interruptibility(vcpu, vcpu->arch.emulate_ctxt.interruptibility);
> >-	kvm_x86_ops->set_rflags(vcpu, vcpu->arch.emulate_ctxt.eflags);
> >-	memcpy(vcpu->arch.regs, c->regs, sizeof c->regs);
> >-	kvm_rip_write(vcpu, vcpu->arch.emulate_ctxt.eip);
> >+	r = EMULATE_DONE;
> >
> >-	if (vcpu->arch.emulate_ctxt.exception>= 0) {
> >+	if (vcpu->arch.emulate_ctxt.exception>= 0)
> >  		inject_emulated_exception(vcpu);
> >-		return EMULATE_DONE;
> >-	}
> >-

(Continue reading)

Avi Kivity | 1 Aug 2010 10:54

Picon

Re: [PATCH 1/2] KVM: x86 emulator: don't update vcpu state if instruction is restarted.

  On 08/01/2010 11:28 AM, Gleb Natapov wrote:
> On Sat, Jul 31, 2010 at 08:25:13PM +0300, Avi Kivity wrote:
>>   On 07/29/2010 03:11 PM, Gleb Natapov wrote:
>>> No need to update vcpu state since instruction is in the middle of the
>>> emulation.
>>>
>>> Signed-off-by: Gleb Natapov<gleb <at> redhat.com>
>>> ---
>>>   arch/x86/kvm/x86.c |   31 +++++++++++++------------------
>>>   1 files changed, 13 insertions(+), 18 deletions(-)
>>>
>>> diff --git a/arch/x86/kvm/x86.c b/arch/x86/kvm/x86.c
>>> index 76fbc32..7e5f075 100644
>>> --- a/arch/x86/kvm/x86.c
>>> +++ b/arch/x86/kvm/x86.c
>>>  <at>  <at>  -4057,32 +4057,27  <at>  <at>  restart:
>>>   		return handle_emulation_failure(vcpu);
>>>   	}
>>>
>>> -	toggle_interruptibility(vcpu, vcpu->arch.emulate_ctxt.interruptibility);
>>> -	kvm_x86_ops->set_rflags(vcpu, vcpu->arch.emulate_ctxt.eflags);
>>> -	memcpy(vcpu->arch.regs, c->regs, sizeof c->regs);
>>> -	kvm_rip_write(vcpu, vcpu->arch.emulate_ctxt.eip);
>>> +	r = EMULATE_DONE;
>>>
>>> -	if (vcpu->arch.emulate_ctxt.exception>= 0) {
>>> +	if (vcpu->arch.emulate_ctxt.exception>= 0)
>>>   		inject_emulated_exception(vcpu);
>>> -		return EMULATE_DONE;
>>> -	}

(Continue reading)

Gleb Natapov | 1 Aug 2010 11:01

Picon

Re: [PATCH 1/2] KVM: x86 emulator: don't update vcpu state if instruction is restarted.

On Sun, Aug 01, 2010 at 11:54:38AM +0300, Avi Kivity wrote:
>  On 08/01/2010 11:28 AM, Gleb Natapov wrote:
> >On Sat, Jul 31, 2010 at 08:25:13PM +0300, Avi Kivity wrote:
> >>  On 07/29/2010 03:11 PM, Gleb Natapov wrote:
> >>>No need to update vcpu state since instruction is in the middle of the
> >>>emulation.
> >>>
> >>>Signed-off-by: Gleb Natapov<gleb <at> redhat.com>
> >>>---
> >>>  arch/x86/kvm/x86.c |   31 +++++++++++++------------------
> >>>  1 files changed, 13 insertions(+), 18 deletions(-)
> >>>
> >>>diff --git a/arch/x86/kvm/x86.c b/arch/x86/kvm/x86.c
> >>>index 76fbc32..7e5f075 100644
> >>>--- a/arch/x86/kvm/x86.c
> >>>+++ b/arch/x86/kvm/x86.c
> >>> <at>  <at>  -4057,32 +4057,27  <at>  <at>  restart:
> >>>  		return handle_emulation_failure(vcpu);
> >>>  	}
> >>>
> >>>-	toggle_interruptibility(vcpu, vcpu->arch.emulate_ctxt.interruptibility);
> >>>-	kvm_x86_ops->set_rflags(vcpu, vcpu->arch.emulate_ctxt.eflags);
> >>>-	memcpy(vcpu->arch.regs, c->regs, sizeof c->regs);
> >>>-	kvm_rip_write(vcpu, vcpu->arch.emulate_ctxt.eip);
> >>>+	r = EMULATE_DONE;
> >>>
> >>>-	if (vcpu->arch.emulate_ctxt.exception>= 0) {
> >>>+	if (vcpu->arch.emulate_ctxt.exception>= 0)
> >>>  		inject_emulated_exception(vcpu);
> >>>-		return EMULATE_DONE;

(Continue reading)

Avi Kivity | 1 Aug 2010 11:14

Picon

Re: [PATCH 1/2] KVM: x86 emulator: don't update vcpu state if instruction is restarted.

  On 08/01/2010 12:01 PM, Gleb Natapov wrote:
>>>
>>> It uses rip _during_ pio. And pio emulation changes rip
>>> only at the end of emulation.
>> But non-emulated pio does a skip_emulated_instruction() immediately
>> (or so the code in kvm-tpr-opt.c assumes:
>>
> Indeed, this is bug in non-emulated pio.

But userspace depends on this bug.

> But the patch does not change
> rip behaviour for emulated pio. vcpu->arch.emulate_ctxt.eip is updated
> only at the end of emulation.

That will lead to failures if the emulator is used for the kvm-tpr-opt 
pio (which may happen with big real mode).

--

-- 
error compiling committee.c: too many arguments to function

--
To unsubscribe from this list: send the line "unsubscribe kvm" in
the body of a message to majordomo <at> vger.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html

Gleb Natapov | 1 Aug 2010 11:24

Picon

Re: [PATCH 1/2] KVM: x86 emulator: don't update vcpu state if instruction is restarted.

On Sun, Aug 01, 2010 at 12:14:40PM +0300, Avi Kivity wrote:
>  On 08/01/2010 12:01 PM, Gleb Natapov wrote:
> >>>
> >>>It uses rip _during_ pio. And pio emulation changes rip
> >>>only at the end of emulation.
> >>But non-emulated pio does a skip_emulated_instruction() immediately
> >>(or so the code in kvm-tpr-opt.c assumes:
> >>
> >Indeed, this is bug in non-emulated pio.
> 
> But userspace depends on this bug.
We can fix that, or make it smarter. Look for io instruction at
rip/rip-2 and use rip accordingly for instance.

> 
> >But the patch does not change
> >rip behaviour for emulated pio. vcpu->arch.emulate_ctxt.eip is updated
> >only at the end of emulation.
> 
> That will lead to failures if the emulator is used for the
> kvm-tpr-opt pio (which may happen with big real mode).
> 
IIRC it was always this way in emulator. I'd rather fix userspace than
break emulator.

--
			Gleb.
--
To unsubscribe from this list: send the line "unsubscribe kvm" in
the body of a message to majordomo <at> vger.kernel.org

(Continue reading)

Avi Kivity | 1 Aug 2010 12:00

Picon

Re: [PATCH 1/2] KVM: x86 emulator: don't update vcpu state if instruction is restarted.

  On 08/01/2010 12:24 PM, Gleb Natapov wrote:
> On Sun, Aug 01, 2010 at 12:14:40PM +0300, Avi Kivity wrote:
>>   On 08/01/2010 12:01 PM, Gleb Natapov wrote:
>>>>> It uses rip _during_ pio. And pio emulation changes rip
>>>>> only at the end of emulation.
>>>> But non-emulated pio does a skip_emulated_instruction() immediately
>>>> (or so the code in kvm-tpr-opt.c assumes:
>>>>
>>> Indeed, this is bug in non-emulated pio.
>> But userspace depends on this bug.
> We can fix that, or make it smarter. Look for io instruction at
> rip/rip-2 and use rip accordingly for instance.

That requires everyone to update, or suffer major breakage.

>>> But the patch does not change
>>> rip behaviour for emulated pio. vcpu->arch.emulate_ctxt.eip is updated
>>> only at the end of emulation.
>> That will lead to failures if the emulator is used for the
>> kvm-tpr-opt pio (which may happen with big real mode).
>>
> IIRC it was always this way in emulator. I'd rather fix userspace than
> break emulator.

It wasn't a problem because the emulator wasn't (and still isn't) used 
for this.  But it has the potential to break badly once we make 
emulate_invalid_guest_state=1 the default.

--

-- 
error compiling committee.c: too many arguments to function

(Continue reading)

Gleb Natapov | 1 Aug 2010 12:53

Picon

Re: [PATCH 1/2] KVM: x86 emulator: don't update vcpu state if instruction is restarted.

On Sun, Aug 01, 2010 at 01:00:11PM +0300, Avi Kivity wrote:
>  On 08/01/2010 12:24 PM, Gleb Natapov wrote:
> >On Sun, Aug 01, 2010 at 12:14:40PM +0300, Avi Kivity wrote:
> >>  On 08/01/2010 12:01 PM, Gleb Natapov wrote:
> >>>>>It uses rip _during_ pio. And pio emulation changes rip
> >>>>>only at the end of emulation.
> >>>>But non-emulated pio does a skip_emulated_instruction() immediately
> >>>>(or so the code in kvm-tpr-opt.c assumes:
> >>>>
> >>>Indeed, this is bug in non-emulated pio.
> >>But userspace depends on this bug.
> >We can fix that, or make it smarter. Look for io instruction at
> >rip/rip-2 and use rip accordingly for instance.
> 
> That requires everyone to update, or suffer major breakage.
> 
They will suffer major breakage when they update to a kvm that calls to
kvm-tpr-opt.c from emulator anyway.

> >>>But the patch does not change
> >>>rip behaviour for emulated pio. vcpu->arch.emulate_ctxt.eip is updated
> >>>only at the end of emulation.
> >>That will lead to failures if the emulator is used for the
> >>kvm-tpr-opt pio (which may happen with big real mode).
> >>
> >IIRC it was always this way in emulator. I'd rather fix userspace than
> >break emulator.
> 
> It wasn't a problem because the emulator wasn't (and still isn't)
> used for this.  But it has the potential to break badly once we make

(Continue reading)

Avi Kivity | 1 Aug 2010 14:17

Picon

Re: [PATCH 1/2] KVM: x86 emulator: don't update vcpu state if instruction is restarted.

  On 08/01/2010 01:53 PM, Gleb Natapov wrote:
>
>> That requires everyone to update, or suffer major breakage.
>>
> They will suffer major breakage when they update to a kvm that calls to
> kvm-tpr-opt.c from emulator anyway.

Why?

>>> IIRC it was always this way in emulator. I'd rather fix userspace than
>>> break emulator.
>> It wasn't a problem because the emulator wasn't (and still isn't)
>> used for this.  But it has the potential to break badly once we make
>> emulate_invalid_guest_state=1 the default.
>>
> So what can we do about it?
>

Keep the existing behaviour.

--

-- 
error compiling committee.c: too many arguments to function

--
To unsubscribe from this list: send the line "unsubscribe kvm" in
the body of a message to majordomo <at> vger.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html

Group

gmane.comp.emulators.kvm.devel.

Project Web Page

KVM Development List

Search Archive

Page

1 | 2 | 3 | 4 | 5 | 6 | 7 | 8 | 9 | 10 | 11 | 154

Articles in period: 1539

August 2010

Language

Change language

Options

Current view: Threads only / Showing only 20 lines / Not hiding cited text.
Change to All messages, whole messages, or hide cited text.

Post a message
NNTP Newsgroup
Classic Gmane web interface
XML

XML

RSS Feed
List Information

About Gmane

Recent entries

[Bug 58771] New: VM performance degradation after KVM QEMU migration o (24 May 2013)
Attn:webmail customers Online Services.!! (24 May 2013)
USB Dongle problems using passthrough. (24 May 2013)
[PATCH] uio/uio_pci_generic: fix error return code in probe() (23 May 2013)
[PATCH] qemu-kvm: fix unmatched RAM alloction/free (18 Jan 2009)
[PATCH 1/1] KVM: add kvm_para_available to asm-generic/kvm_para.h (22 May 2013)
[nVMX ] A couple more tests (22 May 2013)
[PATCH] kvm: add detail error message when fail to add ioeventfd (22 May 2013)
[PATCH] kvm tools: add status notification hook to virtio-mmio (21 May 2013)
I/O port permission bit inheritance between threads (20 May 2013)
KVM call agenda for 2013-05-21 (20 May 2013)
[RFC PATCH v2 0/2] Hyper-V timers (19 May 2013)
[PATCH 00/18] KVM/MIPS32: Support for the new Virtualization ASE (VZ-A (19 May 2013)
[PATCH v3 01/13] nEPT: Support LOAD_IA32_EFER entry/exit controls for (19 May 2013)
[PATCH v3 0/4] KVM/MIPS32: Fixes for Linux 3.10 (18 May 2013)
Genuine Cruzer Blade USB flash drive 8GB Flash Disk (17 May 2013)
Problems while booting a linux system on fast models based CortexA15 (18 May 2013)
[PATCH v2 0/3] KVM/MIPS32: Fixes for Linux 3.10 (17 May 2013)
[PATCH] KVM/MIPS32: Export min_low_pfn. (17 May 2013)
[PATCH 0/8] s390/kvm fixes (17 May 2013)
[PATCH v6 0/7] KVM: MMU: fast zap all shadow pages (16 May 2013)

Archives

742	May 2013
1164	April 2013
928	March 2013
764	February 2013
1075	January 2013
634	December 2012
682	November 2012
990	October 2012
1079	September 2012
1155	August 2012
927	July 2012
897	June 2012
834	May 2012
666	April 2012
1020	March 2012
996	February 2012
1030	January 2012
1165	December 2011
772	November 2011
758	October 2011
721	September 2011
1238	August 2011
1430	July 2011
1287	June 2011
1686	May 2011
1150	April 2011
955	March 2011
1206	February 2011
1289	January 2011
1165	December 2010
1631	November 2010
1200	October 2010
1058	September 2010
1539	August 2010
1119	July 2010
1367	June 2010
1342	May 2010
1317	April 2010
1579	March 2010
1352	February 2010
1059	January 2010
1004	December 2009
712	November 2009
1425	October 2009
1191	September 2009
1692	August 2009
1853	July 2009
2130	June 2009
2358	May 2009
1650	April 2009
1133	March 2009
1057	February 2009
969	January 2009
1587	December 2008
1397	November 2008
1310	October 2008
1545	September 2008
861	August 2008
1213	July 2008
903	June 2008
1109	May 2008
1605	April 2008
1382	March 2008
1105	February 2008
1428	January 2008
1018	December 2007
1199	November 2007
1442	October 2007
1083	September 2007
1091	August 2007
991	July 2007
785	June 2007
820	May 2007
705	April 2007
396	March 2007
436	February 2007
481	January 2007
317	December 2006
323	November 2006
7	October 2006
1	February 2006
1	August 2005
2	March 2004

Design

Zawodny | Right Menu | Left Menu

Your Own Design

Paste the URL of your CSS below. Download CSS template