Re: [Qemu-devel] qemu-kvm: remove "boot=on|off" drive parameter compatibility

Re: [Qemu-devel] qemu-kvm: remove "boot=on|off" drive parameter compatibility

[Anthony Liguori]

Jan Kiszka <jan.kiszka@siemens.com> writes:

> On 2012-10-01 11:31, Marcelo Tosatti wrote:
>
> It's not just about default configs. We need to validate if the
> migration formats are truly compatible (qemu-kvm -> QEMU, the other way
> around definitely not). For the command line switches, we could provide
> a wrapper script that translates them into upstream format or simply
> ignores them. That should be harmless to carry upstream.

qemu-kvm has:

 -no-kvm
 -no-kvm-irqchip
 -no-kvm-pit
 -no-kvm-pit-reinjection
 -tdf <- does nothing

There are replacements for all of the above.  If we need to add them to
qemu.git, it's not big deal to add them.

 -drive ...,boot= <- this is ignored

cpu_set command for CPU hotplug which is known broken in qemu-kvm.

testdev which is nice but only used for development

Default nic is rtl8139 vs. e1000.

Some logic to move change the default VGA ram size to 16mb for pc-1.2
(QEMU uses 16mb by default now too).

I think at this point, none of this matters but I added the various
distro maintainers to the thread.

I think it's time for the distros to drop qemu-kvm and just ship
qemu.git.  Is there anything else that needs to happen to make that
switch?

Regards,

Anthony Liguori

>
> But I would really stop worrying about the qemu-kvm code base.
>
> Jan
>
> -- 
> Siemens AG, Corporate Technology, CT RTC ITP SDP-DE
> Corporate Competence Center Embedded Linux
> --
> To unsubscribe from this list: send the line "unsubscribe kvm" in
> the body of a message to majordomo@vger.kernel.org
> More majordomo info at  http://vger.kernel.org/majordomo-info.html

Re: [Qemu-devel] qemu-kvm: remove "boot=on|off" drive parameter compatibility

[Jan Kiszka]

On 2012-10-01 15:19, Anthony Liguori wrote:
> Jan Kiszka <jan.kiszka@siemens.com> writes:
> 
>> On 2012-10-01 11:31, Marcelo Tosatti wrote:
>>
>> It's not just about default configs. We need to validate if the
>> migration formats are truly compatible (qemu-kvm -> QEMU, the other way
>> around definitely not). For the command line switches, we could provide
>> a wrapper script that translates them into upstream format or simply
>> ignores them. That should be harmless to carry upstream.
> 
> qemu-kvm has:
> 
>  -no-kvm
>  -no-kvm-irqchip
>  -no-kvm-pit
>  -no-kvm-pit-reinjection
>  -tdf <- does nothing
> 
> There are replacements for all of the above.  If we need to add them to
> qemu.git, it's not big deal to add them.

But I don't think we should add them to the source code. This can
perfectly be handled my a (disposable) script layer on top of
qemu-system-x86_64 - the namespace (qemu-kvm in most cases) is also free.

> 
>  -drive ...,boot= <- this is ignored
> 
> cpu_set command for CPU hotplug which is known broken in qemu-kvm.

Right, so nothing is lost when migrating to QEMU.

> 
> testdev which is nice but only used for development
> 
> Default nic is rtl8139 vs. e1000.
> 
> Some logic to move change the default VGA ram size to 16mb for pc-1.2
> (QEMU uses 16mb by default now too).

Also nicely manageable in a wrapper.

> 
> I think at this point, none of this matters but I added the various
> distro maintainers to the thread.
> 
> I think it's time for the distros to drop qemu-kvm and just ship
> qemu.git.

+1

Jan

>  Is there anything else that needs to happen to make that
> switch?
> 
> Regards,
> 
> Anthony Liguori

-- 
Siemens AG, Corporate Technology, CT RTC ITP SDP-DE
Corporate Competence Center Embedded Linux

Re: [Qemu-devel] qemu-kvm: remove "boot=on|off" drive parameter compatibility

[Gleb Natapov]

On Mon, Oct 01, 2012 at 03:26:05PM +0200, Jan Kiszka wrote:
> On 2012-10-01 15:19, Anthony Liguori wrote:
> > Jan Kiszka <jan.kiszka@siemens.com> writes:
> > 
> >> On 2012-10-01 11:31, Marcelo Tosatti wrote:
> >>
> >> It's not just about default configs. We need to validate if the
> >> migration formats are truly compatible (qemu-kvm -> QEMU, the other way
> >> around definitely not). For the command line switches, we could provide
> >> a wrapper script that translates them into upstream format or simply
> >> ignores them. That should be harmless to carry upstream.
> > 
> > qemu-kvm has:
> > 
> >  -no-kvm
> >  -no-kvm-irqchip
> >  -no-kvm-pit
> >  -no-kvm-pit-reinjection
> >  -tdf <- does nothing
> > 
> > There are replacements for all of the above.  If we need to add them to
> > qemu.git, it's not big deal to add them.
> 
> But I don't think we should add them to the source code. This can
> perfectly be handled my a (disposable) script layer on top of
> qemu-system-x86_64 - the namespace (qemu-kvm in most cases) is also free.
> 
> > 
> >  -drive ...,boot= <- this is ignored
> > 
> > cpu_set command for CPU hotplug which is known broken in qemu-kvm.
> 
> Right, so nothing is lost when migrating to QEMU.
> 
> > 
> > testdev which is nice but only used for development
> > 
Jan, do you have a plan for testdev device? It would be a pity to have
qemu-kvm just for that.

> > Default nic is rtl8139 vs. e1000.
> > 
> > Some logic to move change the default VGA ram size to 16mb for pc-1.2
> > (QEMU uses 16mb by default now too).
> 
> Also nicely manageable in a wrapper.
> 
> > 
> > I think at this point, none of this matters but I added the various
> > distro maintainers to the thread.
> > 
> > I think it's time for the distros to drop qemu-kvm and just ship
> > qemu.git.
> 
> +1
> 
> Jan
> 
> >  Is there anything else that needs to happen to make that
> > switch?
> > 
> > Regards,
> > 
> > Anthony Liguori
> 
> -- 
> Siemens AG, Corporate Technology, CT RTC ITP SDP-DE
> Corporate Competence Center Embedded Linux

--
			Gleb.

Re: [Qemu-devel] qemu-kvm: remove "boot=on|off" drive parameter compatibility

[Jan Kiszka]

[-- Attachment #1: Type: text/plain, Size: 1511 bytes --]

On 2012-10-03 11:55, Gleb Natapov wrote:
> On Mon, Oct 01, 2012 at 03:26:05PM +0200, Jan Kiszka wrote:
>> On 2012-10-01 15:19, Anthony Liguori wrote:
>>> Jan Kiszka <jan.kiszka@siemens.com> writes:
>>>
>>>> On 2012-10-01 11:31, Marcelo Tosatti wrote:
>>>>
>>>> It's not just about default configs. We need to validate if the
>>>> migration formats are truly compatible (qemu-kvm -> QEMU, the other way
>>>> around definitely not). For the command line switches, we could provide
>>>> a wrapper script that translates them into upstream format or simply
>>>> ignores them. That should be harmless to carry upstream.
>>>
>>> qemu-kvm has:
>>>
>>>  -no-kvm
>>>  -no-kvm-irqchip
>>>  -no-kvm-pit
>>>  -no-kvm-pit-reinjection
>>>  -tdf <- does nothing
>>>
>>> There are replacements for all of the above.  If we need to add them to
>>> qemu.git, it's not big deal to add them.
>>
>> But I don't think we should add them to the source code. This can
>> perfectly be handled my a (disposable) script layer on top of
>> qemu-system-x86_64 - the namespace (qemu-kvm in most cases) is also free.
>>
>>>
>>>  -drive ...,boot= <- this is ignored
>>>
>>> cpu_set command for CPU hotplug which is known broken in qemu-kvm.
>>
>> Right, so nothing is lost when migrating to QEMU.
>>
>>>
>>> testdev which is nice but only used for development
>>>
> Jan, do you have a plan for testdev device? It would be a pity to have
> qemu-kvm just for that.

Nope, not on my schedule.

Jan



[-- Attachment #2: OpenPGP digital signature --]
[-- Type: application/pgp-signature, Size: 259 bytes --]

Re: [Qemu-devel] qemu-kvm: remove "boot=on|off" drive parameter compatibility

[Gleb Natapov]

On Wed, Oct 03, 2012 at 12:06:57PM +0200, Jan Kiszka wrote:
> On 2012-10-03 11:55, Gleb Natapov wrote:
> > On Mon, Oct 01, 2012 at 03:26:05PM +0200, Jan Kiszka wrote:
> >> On 2012-10-01 15:19, Anthony Liguori wrote:
> >>> Jan Kiszka <jan.kiszka@siemens.com> writes:
> >>>
> >>>> On 2012-10-01 11:31, Marcelo Tosatti wrote:
> >>>>
> >>>> It's not just about default configs. We need to validate if the
> >>>> migration formats are truly compatible (qemu-kvm -> QEMU, the other way
> >>>> around definitely not). For the command line switches, we could provide
> >>>> a wrapper script that translates them into upstream format or simply
> >>>> ignores them. That should be harmless to carry upstream.
> >>>
> >>> qemu-kvm has:
> >>>
> >>>  -no-kvm
> >>>  -no-kvm-irqchip
> >>>  -no-kvm-pit
> >>>  -no-kvm-pit-reinjection
> >>>  -tdf <- does nothing
> >>>
> >>> There are replacements for all of the above.  If we need to add them to
> >>> qemu.git, it's not big deal to add them.
> >>
> >> But I don't think we should add them to the source code. This can
> >> perfectly be handled my a (disposable) script layer on top of
> >> qemu-system-x86_64 - the namespace (qemu-kvm in most cases) is also free.
> >>
> >>>
> >>>  -drive ...,boot= <- this is ignored
> >>>
> >>> cpu_set command for CPU hotplug which is known broken in qemu-kvm.
> >>
> >> Right, so nothing is lost when migrating to QEMU.
> >>
> >>>
> >>> testdev which is nice but only used for development
> >>>
> > Jan, do you have a plan for testdev device? It would be a pity to have
> > qemu-kvm just for that.
> 
> Nope, not on my schedule.
> 
Understood :)

--
			Gleb.

Re: [Qemu-devel] qemu-kvm: remove "boot=on|off" drive parameter compatibility

[Lucas Meneghel Rodrigues]

On 10/03/2012 06:55 AM, Gleb Natapov wrote:
> On Mon, Oct 01, 2012 at 03:26:05PM +0200, Jan Kiszka wrote:
>> On 2012-10-01 15:19, Anthony Liguori wrote:
>>> Jan Kiszka <jan.kiszka@siemens.com> writes:
>>>
>>>> On 2012-10-01 11:31, Marcelo Tosatti wrote:
>>>>
>>>> It's not just about default configs. We need to validate if the
>>>> migration formats are truly compatible (qemu-kvm -> QEMU, the other way
>>>> around definitely not). For the command line switches, we could provide
>>>> a wrapper script that translates them into upstream format or simply
>>>> ignores them. That should be harmless to carry upstream.
>>>
>>> qemu-kvm has:
>>>
>>>   -no-kvm
>>>   -no-kvm-irqchip
>>>   -no-kvm-pit
>>>   -no-kvm-pit-reinjection
>>>   -tdf <- does nothing
>>>
>>> There are replacements for all of the above.  If we need to add them to
>>> qemu.git, it's not big deal to add them.
>>
>> But I don't think we should add them to the source code. This can
>> perfectly be handled my a (disposable) script layer on top of
>> qemu-system-x86_64 - the namespace (qemu-kvm in most cases) is also free.
>>
>>>
>>>   -drive ...,boot= <- this is ignored
>>>
>>> cpu_set command for CPU hotplug which is known broken in qemu-kvm.
>>
>> Right, so nothing is lost when migrating to QEMU.
>>
>>>
>>> testdev which is nice but only used for development
>>>
> Jan, do you have a plan for testdev device? It would be a pity to have
> qemu-kvm just for that.

Yep, I did send patches with the testdev device present on qemu-kvm.git 
to qemu.git a while ago, but there were many comments on the review, I 
ended up not implementing everything that was asked and the patches were 
archived.

If nobody wants to step up to port it, I'll re-read the original thread 
and will spin up new patches (and try to go through the end with it). 
Executing the KVM unittests is something that we can't afford to lose, 
so I'd say it's important on this last mile effort to get rid of qemu-kvm.

Re: [Qemu-devel] qemu-kvm: remove "boot=on|off" drive parameter compatibility

[Paolo Bonzini]

Il 03/10/2012 12:57, Lucas Meneghel Rodrigues ha scritto:
> Yep, I did send patches with the testdev device present on qemu-kvm.git
> to qemu.git a while ago, but there were many comments on the review, I
> ended up not implementing everything that was asked and the patches were
> archived.
> 
> If nobody wants to step up to port it, I'll re-read the original thread
> and will spin up new patches (and try to go through the end with it).
> Executing the KVM unittests is something that we can't afford to lose,
> so I'd say it's important on this last mile effort to get rid of qemu-kvm.

Absolutely, IIRC the problem was that testdev did a little bit of
everything... let's see what's the functionality of testdev:

- write (port 0xf1), can be replaced in autotest with:
-device isa-debugcon,iobase=0xf1,chardev=...

- exit code (port 0xf4), see this series:
http://lists.gnu.org/archive/html/qemu-devel/2012-07/msg00818.html

- ram size (port 0xd1).  If we can also patch kvm-unittests, the memory
is available in the CMOS or in fwcfg.  Here is the SeaBIOS code:

        u32 rs = ((inb_cmos(0x34) << 16) | (inb_cmos(0x35) << 24));
        if (rs)
            rs += 16 * 1024 * 1024;
        else
            rs = (((inb_cmos(0x30) << 10) | (inb_cmos(0x31) << 18))
                  + 1 * 1024 * 1024);

The rest (ports 0xe0..0xe7, 0x2000..0x2017, MMIO) can be left in testdev.

Paolo

Re: [Qemu-devel] qemu-kvm: remove "boot=on|off" drive parameter compatibility

[Gleb Natapov]

On Wed, Oct 03, 2012 at 03:19:56PM +0200, Paolo Bonzini wrote:
> Il 03/10/2012 12:57, Lucas Meneghel Rodrigues ha scritto:
> > Yep, I did send patches with the testdev device present on qemu-kvm.git
> > to qemu.git a while ago, but there were many comments on the review, I
> > ended up not implementing everything that was asked and the patches were
> > archived.
> > 
> > If nobody wants to step up to port it, I'll re-read the original thread
> > and will spin up new patches (and try to go through the end with it).
> > Executing the KVM unittests is something that we can't afford to lose,
> > so I'd say it's important on this last mile effort to get rid of qemu-kvm.
> 
> Absolutely, IIRC the problem was that testdev did a little bit of
> everything... let's see what's the functionality of testdev:
> 
> - write (port 0xf1), can be replaced in autotest with:
> -device isa-debugcon,iobase=0xf1,chardev=...
> 
kvm-unit-tests no longer uses 0xf1 for output. It uses serial.

> - exit code (port 0xf4), see this series:
> http://lists.gnu.org/archive/html/qemu-devel/2012-07/msg00818.html
> 
> - ram size (port 0xd1).  If we can also patch kvm-unittests, the memory
> is available in the CMOS or in fwcfg.  Here is the SeaBIOS code:
> 
>         u32 rs = ((inb_cmos(0x34) << 16) | (inb_cmos(0x35) << 24));
>         if (rs)
>             rs += 16 * 1024 * 1024;
>         else
>             rs = (((inb_cmos(0x30) << 10) | (inb_cmos(0x31) << 18))
>                   + 1 * 1024 * 1024);
> 
> The rest (ports 0xe0..0xe7, 0x2000..0x2017, MMIO) can be left in testdev.
> 
> Paolo

--
			Gleb.

Re: [Qemu-devel] qemu-kvm: remove "boot=on|off" drive parameter compatibility

[Jan Kiszka]

On 2012-10-03 15:19, Paolo Bonzini wrote:
> Il 03/10/2012 12:57, Lucas Meneghel Rodrigues ha scritto:
>> Yep, I did send patches with the testdev device present on qemu-kvm.git
>> to qemu.git a while ago, but there were many comments on the review, I
>> ended up not implementing everything that was asked and the patches were
>> archived.
>>
>> If nobody wants to step up to port it, I'll re-read the original thread
>> and will spin up new patches (and try to go through the end with it).
>> Executing the KVM unittests is something that we can't afford to lose,
>> so I'd say it's important on this last mile effort to get rid of qemu-kvm.
> 
> Absolutely, IIRC the problem was that testdev did a little bit of
> everything... let's see what's the functionality of testdev:
> 
> - write (port 0xf1), can be replaced in autotest with:
> -device isa-debugcon,iobase=0xf1,chardev=...
> 
> - exit code (port 0xf4), see this series:
> http://lists.gnu.org/archive/html/qemu-devel/2012-07/msg00818.html
> 
> - ram size (port 0xd1).  If we can also patch kvm-unittests, the memory
> is available in the CMOS or in fwcfg.  Here is the SeaBIOS code:
> 
>         u32 rs = ((inb_cmos(0x34) << 16) | (inb_cmos(0x35) << 24));
>         if (rs)
>             rs += 16 * 1024 * 1024;
>         else
>             rs = (((inb_cmos(0x30) << 10) | (inb_cmos(0x31) << 18))
>                   + 1 * 1024 * 1024);
> 
> The rest (ports 0xe0..0xe7, 0x2000..0x2017, MMIO) can be left in testdev.

IIRC, one of the biggest problem with testdev was its hack to inject
interrupts.

Jan

-- 
Siemens AG, Corporate Technology, CT RTC ITP SDP-DE
Corporate Competence Center Embedded Linux

Re: [Qemu-devel] qemu-kvm: remove "boot=on|off" drive parameter compatibility

[Lucas Meneghel Rodrigues]

On 10/04/2012 07:48 AM, Jan Kiszka wrote:
> On 2012-10-03 15:19, Paolo Bonzini wrote:
>> Il 03/10/2012 12:57, Lucas Meneghel Rodrigues ha scritto:
>>> Yep, I did send patches with the testdev device present on qemu-kvm.git
>>> to qemu.git a while ago, but there were many comments on the review, I
>>> ended up not implementing everything that was asked and the patches were
>>> archived.
>>>
>>> If nobody wants to step up to port it, I'll re-read the original thread
>>> and will spin up new patches (and try to go through the end with it).
>>> Executing the KVM unittests is something that we can't afford to lose,
>>> so I'd say it's important on this last mile effort to get rid of qemu-kvm.
>>
>> Absolutely, IIRC the problem was that testdev did a little bit of
>> everything... let's see what's the functionality of testdev:
>>
>> - write (port 0xf1), can be replaced in autotest with:
>> -device isa-debugcon,iobase=0xf1,chardev=...
>>
>> - exit code (port 0xf4), see this series:
>> http://lists.gnu.org/archive/html/qemu-devel/2012-07/msg00818.html
>>
>> - ram size (port 0xd1).  If we can also patch kvm-unittests, the memory
>> is available in the CMOS or in fwcfg.  Here is the SeaBIOS code:
>>
>>          u32 rs = ((inb_cmos(0x34) << 16) | (inb_cmos(0x35) << 24));
>>          if (rs)
>>              rs += 16 * 1024 * 1024;
>>          else
>>              rs = (((inb_cmos(0x30) << 10) | (inb_cmos(0x31) << 18))
>>                    + 1 * 1024 * 1024);
>>
>> The rest (ports 0xe0..0xe7, 0x2000..0x2017, MMIO) can be left in testdev.
>
> IIRC, one of the biggest problem with testdev was its hack to inject
> interrupts.

Jan, I assume this commit helps to fix this, right?

commit b334ec567f1de9a60349991e7b75083d569ddb0a
Author: Jan Kiszka <jan.kiszka@siemens.com>
Date:   Fri Mar 2 10:30:47 2012 +0100

     qemu-kvm: Use upstream kvm-i8259

     Drop the qemu-kvm version in favor of the equivalent upstream
     implementation. This allows to move the i8259 back into the hwlib.

     Note that this also drops the testdev hack and restores proper
     isa_get_irq. If testdev scripts exist that inject > IRQ15, they need
     fixing. Testing for these interrupts on the PIIX3 makes no practical
     sense anyway as those lines are unused.

     Signed-off-by: Jan Kiszka <jan.kiszka@siemens.com>
     Signed-off-by: Avi Kivity <avi@redhat.com>

Re: [Qemu-devel] qemu-kvm: remove "boot=on|off" drive parameter compatibility

[Jan Kiszka]

On 2012-10-04 14:10, Lucas Meneghel Rodrigues wrote:
> On 10/04/2012 07:48 AM, Jan Kiszka wrote:
>> On 2012-10-03 15:19, Paolo Bonzini wrote:
>>> Il 03/10/2012 12:57, Lucas Meneghel Rodrigues ha scritto:
>>>> Yep, I did send patches with the testdev device present on qemu-kvm.git
>>>> to qemu.git a while ago, but there were many comments on the review, I
>>>> ended up not implementing everything that was asked and the patches were
>>>> archived.
>>>>
>>>> If nobody wants to step up to port it, I'll re-read the original thread
>>>> and will spin up new patches (and try to go through the end with it).
>>>> Executing the KVM unittests is something that we can't afford to lose,
>>>> so I'd say it's important on this last mile effort to get rid of qemu-kvm.
>>>
>>> Absolutely, IIRC the problem was that testdev did a little bit of
>>> everything... let's see what's the functionality of testdev:
>>>
>>> - write (port 0xf1), can be replaced in autotest with:
>>> -device isa-debugcon,iobase=0xf1,chardev=...
>>>
>>> - exit code (port 0xf4), see this series:
>>> http://lists.gnu.org/archive/html/qemu-devel/2012-07/msg00818.html
>>>
>>> - ram size (port 0xd1).  If we can also patch kvm-unittests, the memory
>>> is available in the CMOS or in fwcfg.  Here is the SeaBIOS code:
>>>
>>>          u32 rs = ((inb_cmos(0x34) << 16) | (inb_cmos(0x35) << 24));
>>>          if (rs)
>>>              rs += 16 * 1024 * 1024;
>>>          else
>>>              rs = (((inb_cmos(0x30) << 10) | (inb_cmos(0x31) << 18))
>>>                    + 1 * 1024 * 1024);
>>>
>>> The rest (ports 0xe0..0xe7, 0x2000..0x2017, MMIO) can be left in testdev.
>>
>> IIRC, one of the biggest problem with testdev was its hack to inject
>> interrupts.
> 
> Jan, I assume this commit helps to fix this, right?
> 
> commit b334ec567f1de9a60349991e7b75083d569ddb0a
> Author: Jan Kiszka <jan.kiszka@siemens.com>
> Date:   Fri Mar 2 10:30:47 2012 +0100
> 
>      qemu-kvm: Use upstream kvm-i8259
> 
>      Drop the qemu-kvm version in favor of the equivalent upstream
>      implementation. This allows to move the i8259 back into the hwlib.
> 
>      Note that this also drops the testdev hack and restores proper
>      isa_get_irq. If testdev scripts exist that inject > IRQ15, they need
>      fixing. Testing for these interrupts on the PIIX3 makes no practical
>      sense anyway as those lines are unused.
> 
>      Signed-off-by: Jan Kiszka <jan.kiszka@siemens.com>
>      Signed-off-by: Avi Kivity <avi@redhat.com>

Yes, this improved it a lot as we no longer depend on additional
changes. I'm not sure if there was resistance beyond that.

When cleaning up the code: register_ioport_read must be replaced with
the memory API.

Jan

-- 
Siemens AG, Corporate Technology, CT RTC ITP SDP-DE
Corporate Competence Center Embedded Linux

Re: [Qemu-devel] qemu-kvm: remove "boot=on|off" drive parameter compatibility

[Lucas Meneghel Rodrigues]

On 10/04/2012 09:27 AM, Jan Kiszka wrote:
> On 2012-10-04 14:10, Lucas Meneghel Rodrigues wrote:
>> On 10/04/2012 07:48 AM, Jan Kiszka wrote:
>>> On 2012-10-03 15:19, Paolo Bonzini wrote:
>>>> Il 03/10/2012 12:57, Lucas Meneghel Rodrigues ha scritto:
>>>>> Yep, I did send patches with the testdev device present on qemu-kvm.git
>>>>> to qemu.git a while ago, but there were many comments on the review, I
>>>>> ended up not implementing everything that was asked and the patches were
>>>>> archived.
>>>>>
>>>>> If nobody wants to step up to port it, I'll re-read the original thread
>>>>> and will spin up new patches (and try to go through the end with it).
>>>>> Executing the KVM unittests is something that we can't afford to lose,
>>>>> so I'd say it's important on this last mile effort to get rid of qemu-kvm.
>>>>
>>>> Absolutely, IIRC the problem was that testdev did a little bit of
>>>> everything... let's see what's the functionality of testdev:
>>>>
>>>> - write (port 0xf1), can be replaced in autotest with:
>>>> -device isa-debugcon,iobase=0xf1,chardev=...
>>>>
>>>> - exit code (port 0xf4), see this series:
>>>> http://lists.gnu.org/archive/html/qemu-devel/2012-07/msg00818.html
>>>>
>>>> - ram size (port 0xd1).  If we can also patch kvm-unittests, the memory
>>>> is available in the CMOS or in fwcfg.  Here is the SeaBIOS code:
>>>>
>>>>           u32 rs = ((inb_cmos(0x34) << 16) | (inb_cmos(0x35) << 24));
>>>>           if (rs)
>>>>               rs += 16 * 1024 * 1024;
>>>>           else
>>>>               rs = (((inb_cmos(0x30) << 10) | (inb_cmos(0x31) << 18))
>>>>                     + 1 * 1024 * 1024);
>>>>
>>>> The rest (ports 0xe0..0xe7, 0x2000..0x2017, MMIO) can be left in testdev.
>>>
>>> IIRC, one of the biggest problem with testdev was its hack to inject
>>> interrupts.
>>
>> Jan, I assume this commit helps to fix this, right?
>>
>> commit b334ec567f1de9a60349991e7b75083d569ddb0a
>> Author: Jan Kiszka <jan.kiszka@siemens.com>
>> Date:   Fri Mar 2 10:30:47 2012 +0100
>>
>>       qemu-kvm: Use upstream kvm-i8259
>>
>>       Drop the qemu-kvm version in favor of the equivalent upstream
>>       implementation. This allows to move the i8259 back into the hwlib.
>>
>>       Note that this also drops the testdev hack and restores proper
>>       isa_get_irq. If testdev scripts exist that inject > IRQ15, they need
>>       fixing. Testing for these interrupts on the PIIX3 makes no practical
>>       sense anyway as those lines are unused.
>>
>>       Signed-off-by: Jan Kiszka <jan.kiszka@siemens.com>
>>       Signed-off-by: Avi Kivity <avi@redhat.com>
>
> Yes, this improved it a lot as we no longer depend on additional
> changes. I'm not sure if there was resistance beyond that.
>
> When cleaning up the code: register_ioport_read must be replaced with
> the memory API.

I did look at the MemoryRegionOps/memory_region_init_io and still did 
not figure out how to port things. I'll send a v2 addressing all the 
comments made so far but this one, just to see if people are OK with the 
direction of the full patch, then if you could give me some pointers of 
how to do this conversion, it'd be great.

Re: [Qemu-devel] qemu-kvm: remove "boot=on|off" drive parameter compatibility

[Jan Kiszka]

On 2012-10-04 19:21, Lucas Meneghel Rodrigues wrote:
> On 10/04/2012 09:27 AM, Jan Kiszka wrote:
>> On 2012-10-04 14:10, Lucas Meneghel Rodrigues wrote:
>>> On 10/04/2012 07:48 AM, Jan Kiszka wrote:
>>>> On 2012-10-03 15:19, Paolo Bonzini wrote:
>>>>> Il 03/10/2012 12:57, Lucas Meneghel Rodrigues ha scritto:
>>>>>> Yep, I did send patches with the testdev device present on qemu-kvm.git
>>>>>> to qemu.git a while ago, but there were many comments on the review, I
>>>>>> ended up not implementing everything that was asked and the patches were
>>>>>> archived.
>>>>>>
>>>>>> If nobody wants to step up to port it, I'll re-read the original thread
>>>>>> and will spin up new patches (and try to go through the end with it).
>>>>>> Executing the KVM unittests is something that we can't afford to lose,
>>>>>> so I'd say it's important on this last mile effort to get rid of qemu-kvm.
>>>>>
>>>>> Absolutely, IIRC the problem was that testdev did a little bit of
>>>>> everything... let's see what's the functionality of testdev:
>>>>>
>>>>> - write (port 0xf1), can be replaced in autotest with:
>>>>> -device isa-debugcon,iobase=0xf1,chardev=...
>>>>>
>>>>> - exit code (port 0xf4), see this series:
>>>>> http://lists.gnu.org/archive/html/qemu-devel/2012-07/msg00818.html
>>>>>
>>>>> - ram size (port 0xd1).  If we can also patch kvm-unittests, the memory
>>>>> is available in the CMOS or in fwcfg.  Here is the SeaBIOS code:
>>>>>
>>>>>           u32 rs = ((inb_cmos(0x34) << 16) | (inb_cmos(0x35) << 24));
>>>>>           if (rs)
>>>>>               rs += 16 * 1024 * 1024;
>>>>>           else
>>>>>               rs = (((inb_cmos(0x30) << 10) | (inb_cmos(0x31) << 18))
>>>>>                     + 1 * 1024 * 1024);
>>>>>
>>>>> The rest (ports 0xe0..0xe7, 0x2000..0x2017, MMIO) can be left in testdev.
>>>>
>>>> IIRC, one of the biggest problem with testdev was its hack to inject
>>>> interrupts.
>>>
>>> Jan, I assume this commit helps to fix this, right?
>>>
>>> commit b334ec567f1de9a60349991e7b75083d569ddb0a
>>> Author: Jan Kiszka <jan.kiszka@siemens.com>
>>> Date:   Fri Mar 2 10:30:47 2012 +0100
>>>
>>>       qemu-kvm: Use upstream kvm-i8259
>>>
>>>       Drop the qemu-kvm version in favor of the equivalent upstream
>>>       implementation. This allows to move the i8259 back into the hwlib.
>>>
>>>       Note that this also drops the testdev hack and restores proper
>>>       isa_get_irq. If testdev scripts exist that inject > IRQ15, they need
>>>       fixing. Testing for these interrupts on the PIIX3 makes no practical
>>>       sense anyway as those lines are unused.
>>>
>>>       Signed-off-by: Jan Kiszka <jan.kiszka@siemens.com>
>>>       Signed-off-by: Avi Kivity <avi@redhat.com>
>>
>> Yes, this improved it a lot as we no longer depend on additional
>> changes. I'm not sure if there was resistance beyond that.
>>
>> When cleaning up the code: register_ioport_read must be replaced with
>> the memory API.
> 
> I did look at the MemoryRegionOps/memory_region_init_io and still did 
> not figure out how to port things. I'll send a v2 addressing all the 
> comments made so far but this one, just to see if people are OK with the 
> direction of the full patch, then if you could give me some pointers of 
> how to do this conversion, it'd be great.
> 

See e.g. http://thread.gmane.org/gmane.comp.emulators.qemu/171217

Jan

-- 
Siemens AG, Corporate Technology, CT RTC ITP SDP-DE
Corporate Competence Center Embedded Linux

Re: [Qemu-devel] qemu-kvm: remove "boot=on|off" drive parameter compatibility

[Marcelo Tosatti]

On Mon, Oct 01, 2012 at 08:19:29AM -0500, Anthony Liguori wrote:
> Jan Kiszka <jan.kiszka@siemens.com> writes:
> 
> > On 2012-10-01 11:31, Marcelo Tosatti wrote:
> >
> > It's not just about default configs. We need to validate if the
> > migration formats are truly compatible (qemu-kvm -> QEMU, the other way
> > around definitely not). For the command line switches, we could provide
> > a wrapper script that translates them into upstream format or simply
> > ignores them. That should be harmless to carry upstream.
> 
> qemu-kvm has:
> 
>  -no-kvm
>  -no-kvm-irqchip
>  -no-kvm-pit
>  -no-kvm-pit-reinjection
>  -tdf <- does nothing
> 
> There are replacements for all of the above.  If we need to add them to
> qemu.git, it's not big deal to add them.

At the moment the only purpose of this command line options is for
compability with scripts. My view is that scripts are easily fixed,
so we can just drop them. No need to carry this to QEMU.

>  -drive ...,boot= <- this is ignored
> 
> cpu_set command for CPU hotplug which is known broken in qemu-kvm.
> 
> testdev which is nice but only used for development
>
> Default nic is rtl8139 vs. e1000.

Config file (as suggested earlier on this thread).

> Some logic to move change the default VGA ram size to 16mb for pc-1.2
> (QEMU uses 16mb by default now too).

Config file.

> I think at this point, none of this matters but I added the various
> distro maintainers to the thread.
> 
> I think it's time for the distros to drop qemu-kvm and just ship
> qemu.git.  Is there anything else that needs to happen to make that
> switch?
> 
> Regards,
> 
> Anthony Liguori
> 
> >
> > But I would really stop worrying about the qemu-kvm code base.
> >
> > Jan
> >
> > -- 
> > Siemens AG, Corporate Technology, CT RTC ITP SDP-DE
> > Corporate Competence Center Embedded Linux
> > --
> > To unsubscribe from this list: send the line "unsubscribe kvm" in
> > the body of a message to majordomo@vger.kernel.org
> > More majordomo info at  http://vger.kernel.org/majordomo-info.html

Re: [Qemu-devel] qemu-kvm: remove "boot=on|off" drive parameter compatibility

[Jan Kiszka]

On 2012-10-01 15:31, Marcelo Tosatti wrote:
> On Mon, Oct 01, 2012 at 08:19:29AM -0500, Anthony Liguori wrote:
>> Jan Kiszka <jan.kiszka@siemens.com> writes:
>>
>>> On 2012-10-01 11:31, Marcelo Tosatti wrote:
>>>
>>> It's not just about default configs. We need to validate if the
>>> migration formats are truly compatible (qemu-kvm -> QEMU, the other way
>>> around definitely not). For the command line switches, we could provide
>>> a wrapper script that translates them into upstream format or simply
>>> ignores them. That should be harmless to carry upstream.
>>
>> qemu-kvm has:
>>
>>  -no-kvm
>>  -no-kvm-irqchip
>>  -no-kvm-pit
>>  -no-kvm-pit-reinjection
>>  -tdf <- does nothing
>>
>> There are replacements for all of the above.  If we need to add them to
>> qemu.git, it's not big deal to add them.
> 
> At the moment the only purpose of this command line options is for
> compability with scripts. My view is that scripts are easily fixed,
> so we can just drop them. No need to carry this to QEMU.
> 
>>  -drive ...,boot= <- this is ignored
>>
>> cpu_set command for CPU hotplug which is known broken in qemu-kvm.
>>
>> testdev which is nice but only used for development
>>
>> Default nic is rtl8139 vs. e1000.
> 
> Config file (as suggested earlier on this thread).

If you need to append -config bla, you can also specify the desired NIC
explicitly - I see no value in the former. If we decide to mangle a
qemu-kvm command line before calling QEMU binaries, we can adjust this
variation there. Otherwise it's the same as with all those -kvm*:
Scripts/management tools will need adjustment.

Jan

-- 
Siemens AG, Corporate Technology, CT RTC ITP SDP-DE
Corporate Competence Center Embedded Linux

Re: [Qemu-devel] qemu-kvm: remove "boot=on|off" drive parameter compatibility

[Michael Tokarev]

01.10.2012 17:36, Jan Kiszka wrote:
> On 2012-10-01 15:31, Marcelo Tosatti wrote:

>>> Default nic is rtl8139 vs. e1000.
>>
>> Config file (as suggested earlier on this thread).
> 
> If you need to append -config bla, you can also specify the desired NIC
> explicitly - I see no value in the former. If we decide to mangle a
> qemu-kvm command line before calling QEMU binaries, we can adjust this
> variation there. Otherwise it's the same as with all those -kvm*:
> Scripts/management tools will need adjustment.

I don't think there's a need to "manage" this rtl8139 at
all at this level. Let's declare that "qemu-kvm 1.3+" will
switch from rtl8139 to e1000 by default as a more suitable
in modern world, -- the same way as qemu did the switch
earlier, and be done with that.  I think.

Note that this is JUST for "-net nic" users, which should
be the minority (proper usage is -device rtl8139 or -device
e1000, with explicit model).

Thanks,

/mjt

Re: [Qemu-devel] qemu-kvm: remove "boot=on|off" drive parameter compatibility

[Jan Kiszka]

On 2012-10-01 15:44, Michael Tokarev wrote:
> 01.10.2012 17:36, Jan Kiszka wrote:
>> On 2012-10-01 15:31, Marcelo Tosatti wrote:
> 
>>>> Default nic is rtl8139 vs. e1000.
>>>
>>> Config file (as suggested earlier on this thread).
>>
>> If you need to append -config bla, you can also specify the desired NIC
>> explicitly - I see no value in the former. If we decide to mangle a
>> qemu-kvm command line before calling QEMU binaries, we can adjust this
>> variation there. Otherwise it's the same as with all those -kvm*:
>> Scripts/management tools will need adjustment.
> 
> I don't think there's a need to "manage" this rtl8139 at
> all at this level. Let's declare that "qemu-kvm 1.3+" will
> switch from rtl8139 to e1000 by default as a more suitable
> in modern world, -- the same way as qemu did the switch
> earlier, and be done with that.  I think.
> 
> Note that this is JUST for "-net nic" users, which should
> be the minority (proper usage is -device rtl8139 or -device
> e1000, with explicit model).

Well, the alternative to config files or wrapper scripts is just a
detailed checklist "how to migrate from qemu-kvm to QEMU".

Jan

-- 
Siemens AG, Corporate Technology, CT RTC ITP SDP-DE
Corporate Competence Center Embedded Linux

Re: [Qemu-devel] qemu-kvm: remove "boot=on|off" drive parameter compatibility

[Paolo Bonzini]

Il 01/10/2012 15:19, Anthony Liguori ha scritto:
> I think it's time for the distros to drop qemu-kvm and just ship
> qemu.git.  Is there anything else that needs to happen to make that
> switch?

Perhaps change the default to -machine accel=kvm:tcg?

Paolo

Re: [Qemu-devel] qemu-kvm: remove "boot=on|off" drive parameter compatibility

[Jan Kiszka]

On 2012-10-01 15:38, Paolo Bonzini wrote:
> Il 01/10/2012 15:19, Anthony Liguori ha scritto:
>> I think it's time for the distros to drop qemu-kvm and just ship
>> qemu.git.  Is there anything else that needs to happen to make that
>> switch?
> 
> Perhaps change the default to -machine accel=kvm:tcg?

That's the old discussion again: This way we generate that "silent
failure" (unless you monitor the console output), where users will
complain that "QEMU is so slow" because something is blocking KVM. Maybe
the risk for the latter is lower these days as modules are auto-loaded
now, don't know.

Jan

-- 
Siemens AG, Corporate Technology, CT RTC ITP SDP-DE
Corporate Competence Center Embedded Linux

Re: [Qemu-devel] qemu-kvm: remove "boot=on|off" drive parameter compatibility

[Anthony Liguori]

Paolo Bonzini <pbonzini@redhat.com> writes:

> Il 01/10/2012 15:19, Anthony Liguori ha scritto:
>> I think it's time for the distros to drop qemu-kvm and just ship
>> qemu.git.  Is there anything else that needs to happen to make that
>> switch?
>
> Perhaps change the default to -machine accel=kvm:tcg?
>
> Paolo

I would be in favor of:

#if defined(CONFIG_KVM)
accel=kvm
#else
accel=tcg
#endif

If KVM is available for your target, you almost certainly want to use
it.

I'd be very happy with that change for 1.3.

Regards,

Anthony Liguori

[Qemu-devel] [PATCH] kvm: Set default accelerator to "kvm" if the host supports it

[Jan Kiszka]

If we built a target for a host that supports KVM in principle, set the
default accelerator to KVM as well. This also means the start of QEMU
will fail to start if KVM support turns out to be unavailable at
runtime.

Signed-off-by: Jan Kiszka <jan.kiszka@siemens.com>
---
 kvm-all.c  |    1 +
 kvm-stub.c |    1 +
 kvm.h      |    1 +
 vl.c       |    4 ++--
 4 files changed, 5 insertions(+), 2 deletions(-)

diff --git a/kvm-all.c b/kvm-all.c
index 92a7137..4d5f86c 100644
--- a/kvm-all.c
+++ b/kvm-all.c
@@ -103,6 +103,7 @@ struct KVMState
 #endif
 };
 
+bool kvm_configured = true;
 KVMState *kvm_state;
 bool kvm_kernel_irqchip;
 bool kvm_async_interrupts_allowed;
diff --git a/kvm-stub.c b/kvm-stub.c
index 3c52eb5..86a6451 100644
--- a/kvm-stub.c
+++ b/kvm-stub.c
@@ -17,6 +17,7 @@
 #include "gdbstub.h"
 #include "kvm.h"
 
+bool kvm_configured;
 KVMState *kvm_state;
 bool kvm_kernel_irqchip;
 bool kvm_async_interrupts_allowed;
diff --git a/kvm.h b/kvm.h
index dea2998..9936e5f 100644
--- a/kvm.h
+++ b/kvm.h
@@ -22,6 +22,7 @@
 #include <linux/kvm.h>
 #endif
 
+extern bool kvm_configured;
 extern int kvm_allowed;
 extern bool kvm_kernel_irqchip;
 extern bool kvm_async_interrupts_allowed;
diff --git a/vl.c b/vl.c
index 8d305ca..f557bd1 100644
--- a/vl.c
+++ b/vl.c
@@ -2215,8 +2215,8 @@ static int configure_accelerator(void)
     }
 
     if (p == NULL) {
-        /* Use the default "accelerator", tcg */
-        p = "tcg";
+        /* The default accelerator depends on the availability of KVM. */
+        p = kvm_configured ? "kvm" : "tcg";
     }
 
     while (!accel_initialised && *p != '\0') {
-- 
1.7.3.4

Re: [Qemu-devel] [PATCH] kvm: Set default accelerator to "kvm" if the host supports it

[Anthony Liguori]

Jan Kiszka <jan.kiszka@siemens.com> writes:

> If we built a target for a host that supports KVM in principle, set the
> default accelerator to KVM as well. This also means the start of QEMU
> will fail to start if KVM support turns out to be unavailable at
> runtime.
>
> Signed-off-by: Jan Kiszka <jan.kiszka@siemens.com>
> ---
>  kvm-all.c  |    1 +
>  kvm-stub.c |    1 +
>  kvm.h      |    1 +
>  vl.c       |    4 ++--
>  4 files changed, 5 insertions(+), 2 deletions(-)
>
> diff --git a/kvm-all.c b/kvm-all.c
> index 92a7137..4d5f86c 100644
> --- a/kvm-all.c
> +++ b/kvm-all.c
> @@ -103,6 +103,7 @@ struct KVMState
>  #endif
>  };
>  
> +bool kvm_configured = true;
>  KVMState *kvm_state;
>  bool kvm_kernel_irqchip;
>  bool kvm_async_interrupts_allowed;
> diff --git a/kvm-stub.c b/kvm-stub.c
> index 3c52eb5..86a6451 100644
> --- a/kvm-stub.c
> +++ b/kvm-stub.c
> @@ -17,6 +17,7 @@
>  #include "gdbstub.h"
>  #include "kvm.h"
>  
> +bool kvm_configured;
>  KVMState *kvm_state;
>  bool kvm_kernel_irqchip;
>  bool kvm_async_interrupts_allowed;
> diff --git a/kvm.h b/kvm.h
> index dea2998..9936e5f 100644
> --- a/kvm.h
> +++ b/kvm.h
> @@ -22,6 +22,7 @@
>  #include <linux/kvm.h>
>  #endif
>  
> +extern bool kvm_configured;
>  extern int kvm_allowed;
>  extern bool kvm_kernel_irqchip;
>  extern bool kvm_async_interrupts_allowed;
> diff --git a/vl.c b/vl.c
> index 8d305ca..f557bd1 100644
> --- a/vl.c
> +++ b/vl.c
> @@ -2215,8 +2215,8 @@ static int configure_accelerator(void)
>      }
>  
>      if (p == NULL) {
> -        /* Use the default "accelerator", tcg */
> -        p = "tcg";
> +        /* The default accelerator depends on the availability of KVM. */
> +        p = kvm_configured ? "kvm" : "tcg";
>      }

How about making this an arch_init() function call and then using a #if
defined(KVM_CONFIG) in arch_init.c?

I hate to introduce another global variable if we can avoid it...

Otherwise:

Acked-by: Anthony Liguori <aliguori@us.ibm.com>

Blue/Aurelien, any objections?

Regards,

Anthony Liguori

>  
>      while (!accel_initialised && *p != '\0') {
> -- 
> 1.7.3.4

Re: [Qemu-devel] [PATCH] kvm: Set default accelerator to "kvm" if the host supports it

[Aurelien Jarno]

On Mon, Oct 01, 2012 at 11:20:41AM -0500, Anthony Liguori wrote:
> Jan Kiszka <jan.kiszka@siemens.com> writes:
> 
> > If we built a target for a host that supports KVM in principle, set the
> > default accelerator to KVM as well. This also means the start of QEMU
> > will fail to start if KVM support turns out to be unavailable at
> > runtime.
> >
> > Signed-off-by: Jan Kiszka <jan.kiszka@siemens.com>
> > ---
> >  kvm-all.c  |    1 +
> >  kvm-stub.c |    1 +
> >  kvm.h      |    1 +
> >  vl.c       |    4 ++--
> >  4 files changed, 5 insertions(+), 2 deletions(-)
> >
> > diff --git a/kvm-all.c b/kvm-all.c
> > index 92a7137..4d5f86c 100644
> > --- a/kvm-all.c
> > +++ b/kvm-all.c
> > @@ -103,6 +103,7 @@ struct KVMState
> >  #endif
> >  };
> >  
> > +bool kvm_configured = true;
> >  KVMState *kvm_state;
> >  bool kvm_kernel_irqchip;
> >  bool kvm_async_interrupts_allowed;
> > diff --git a/kvm-stub.c b/kvm-stub.c
> > index 3c52eb5..86a6451 100644
> > --- a/kvm-stub.c
> > +++ b/kvm-stub.c
> > @@ -17,6 +17,7 @@
> >  #include "gdbstub.h"
> >  #include "kvm.h"
> >  
> > +bool kvm_configured;
> >  KVMState *kvm_state;
> >  bool kvm_kernel_irqchip;
> >  bool kvm_async_interrupts_allowed;
> > diff --git a/kvm.h b/kvm.h
> > index dea2998..9936e5f 100644
> > --- a/kvm.h
> > +++ b/kvm.h
> > @@ -22,6 +22,7 @@
> >  #include <linux/kvm.h>
> >  #endif
> >  
> > +extern bool kvm_configured;
> >  extern int kvm_allowed;
> >  extern bool kvm_kernel_irqchip;
> >  extern bool kvm_async_interrupts_allowed;
> > diff --git a/vl.c b/vl.c
> > index 8d305ca..f557bd1 100644
> > --- a/vl.c
> > +++ b/vl.c
> > @@ -2215,8 +2215,8 @@ static int configure_accelerator(void)
> >      }
> >  
> >      if (p == NULL) {
> > -        /* Use the default "accelerator", tcg */
> > -        p = "tcg";
> > +        /* The default accelerator depends on the availability of KVM. */
> > +        p = kvm_configured ? "kvm" : "tcg";
> >      }
> 
> How about making this an arch_init() function call and then using a #if
> defined(KVM_CONFIG) in arch_init.c?
> 
> I hate to introduce another global variable if we can avoid it...
> 
> Otherwise:
> 
> Acked-by: Anthony Liguori <aliguori@us.ibm.com>
> 
> Blue/Aurelien, any objections?
> 

I am not sure this way of doing is really distribution friendly, where
the QEMU package is built for a large variety of hosts, some of them
maybe without KVM support.

I am all for enabling KVM by default, but I think it should fall back to
TCG if it is not enabled (probably with a warning so that the user is
aware of that). On the other hand, if the user explicitly enables KVM
support with -enable-kvm or with accel=kvm, it should fail to start. In 
other words, a bit like the configure script options, by default we 
use auto-detection, but if an option is explicitly enabled, it fails if
it can't be enabled.

-- 
Aurelien Jarno                          GPG: 1024D/F1BCDB73
aurelien@aurel32.net                 http://www.aurel32.net

Re: [Qemu-devel] [PATCH] kvm: Set default accelerator to "kvm" if the host supports it

[Jan Kiszka]

[-- Attachment #1: Type: text/plain, Size: 2670 bytes --]

On 2012-10-01 18:20, Anthony Liguori wrote:
> Jan Kiszka <jan.kiszka@siemens.com> writes:
> 
>> If we built a target for a host that supports KVM in principle, set the
>> default accelerator to KVM as well. This also means the start of QEMU
>> will fail to start if KVM support turns out to be unavailable at
>> runtime.
>>
>> Signed-off-by: Jan Kiszka <jan.kiszka@siemens.com>
>> ---
>>  kvm-all.c  |    1 +
>>  kvm-stub.c |    1 +
>>  kvm.h      |    1 +
>>  vl.c       |    4 ++--
>>  4 files changed, 5 insertions(+), 2 deletions(-)
>>
>> diff --git a/kvm-all.c b/kvm-all.c
>> index 92a7137..4d5f86c 100644
>> --- a/kvm-all.c
>> +++ b/kvm-all.c
>> @@ -103,6 +103,7 @@ struct KVMState
>>  #endif
>>  };
>>  
>> +bool kvm_configured = true;
>>  KVMState *kvm_state;
>>  bool kvm_kernel_irqchip;
>>  bool kvm_async_interrupts_allowed;
>> diff --git a/kvm-stub.c b/kvm-stub.c
>> index 3c52eb5..86a6451 100644
>> --- a/kvm-stub.c
>> +++ b/kvm-stub.c
>> @@ -17,6 +17,7 @@
>>  #include "gdbstub.h"
>>  #include "kvm.h"
>>  
>> +bool kvm_configured;
>>  KVMState *kvm_state;
>>  bool kvm_kernel_irqchip;
>>  bool kvm_async_interrupts_allowed;
>> diff --git a/kvm.h b/kvm.h
>> index dea2998..9936e5f 100644
>> --- a/kvm.h
>> +++ b/kvm.h
>> @@ -22,6 +22,7 @@
>>  #include <linux/kvm.h>
>>  #endif
>>  
>> +extern bool kvm_configured;
>>  extern int kvm_allowed;
>>  extern bool kvm_kernel_irqchip;
>>  extern bool kvm_async_interrupts_allowed;
>> diff --git a/vl.c b/vl.c
>> index 8d305ca..f557bd1 100644
>> --- a/vl.c
>> +++ b/vl.c
>> @@ -2215,8 +2215,8 @@ static int configure_accelerator(void)
>>      }
>>  
>>      if (p == NULL) {
>> -        /* Use the default "accelerator", tcg */
>> -        p = "tcg";
>> +        /* The default accelerator depends on the availability of KVM. */
>> +        p = kvm_configured ? "kvm" : "tcg";
>>      }
> 
> How about making this an arch_init() function call and then using a #if
> defined(KVM_CONFIG) in arch_init.c?
> 
> I hate to introduce another global variable if we can avoid it...

Hacked too quickly. In fact, kvm_configured is simply kvm_available().
However, resistance appear to be too high here.

Jan

> 
> Otherwise:
> 
> Acked-by: Anthony Liguori <aliguori@us.ibm.com>
> 
> Blue/Aurelien, any objections?
> 
> Regards,
> 
> Anthony Liguori
> 
>>  
>>      while (!accel_initialised && *p != '\0') {
>> -- 
>> 1.7.3.4
> --
> To unsubscribe from this list: send the line "unsubscribe kvm" in
> the body of a message to majordomo@vger.kernel.org
> More majordomo info at  http://vger.kernel.org/majordomo-info.html
> 



[-- Attachment #2: OpenPGP digital signature --]
[-- Type: application/pgp-signature, Size: 259 bytes --]

Re: [Qemu-devel] [PATCH] kvm: Set default accelerator to "kvm" if the host supports it

[Blue Swirl]

On Mon, Oct 1, 2012 at 4:20 PM, Anthony Liguori <anthony@codemonkey.ws> wrote:
> Jan Kiszka <jan.kiszka@siemens.com> writes:
>
>> If we built a target for a host that supports KVM in principle, set the
>> default accelerator to KVM as well. This also means the start of QEMU
>> will fail to start if KVM support turns out to be unavailable at
>> runtime.
>>
>> Signed-off-by: Jan Kiszka <jan.kiszka@siemens.com>
>> ---
>>  kvm-all.c  |    1 +
>>  kvm-stub.c |    1 +
>>  kvm.h      |    1 +
>>  vl.c       |    4 ++--
>>  4 files changed, 5 insertions(+), 2 deletions(-)
>>
>> diff --git a/kvm-all.c b/kvm-all.c
>> index 92a7137..4d5f86c 100644
>> --- a/kvm-all.c
>> +++ b/kvm-all.c
>> @@ -103,6 +103,7 @@ struct KVMState
>>  #endif
>>  };
>>
>> +bool kvm_configured = true;
>>  KVMState *kvm_state;
>>  bool kvm_kernel_irqchip;
>>  bool kvm_async_interrupts_allowed;
>> diff --git a/kvm-stub.c b/kvm-stub.c
>> index 3c52eb5..86a6451 100644
>> --- a/kvm-stub.c
>> +++ b/kvm-stub.c
>> @@ -17,6 +17,7 @@
>>  #include "gdbstub.h"
>>  #include "kvm.h"
>>
>> +bool kvm_configured;
>>  KVMState *kvm_state;
>>  bool kvm_kernel_irqchip;
>>  bool kvm_async_interrupts_allowed;
>> diff --git a/kvm.h b/kvm.h
>> index dea2998..9936e5f 100644
>> --- a/kvm.h
>> +++ b/kvm.h
>> @@ -22,6 +22,7 @@
>>  #include <linux/kvm.h>
>>  #endif
>>
>> +extern bool kvm_configured;
>>  extern int kvm_allowed;
>>  extern bool kvm_kernel_irqchip;
>>  extern bool kvm_async_interrupts_allowed;
>> diff --git a/vl.c b/vl.c
>> index 8d305ca..f557bd1 100644
>> --- a/vl.c
>> +++ b/vl.c
>> @@ -2215,8 +2215,8 @@ static int configure_accelerator(void)
>>      }
>>
>>      if (p == NULL) {
>> -        /* Use the default "accelerator", tcg */
>> -        p = "tcg";
>> +        /* The default accelerator depends on the availability of KVM. */
>> +        p = kvm_configured ? "kvm" : "tcg";
>>      }
>
> How about making this an arch_init() function call and then using a #if
> defined(KVM_CONFIG) in arch_init.c?
>
> I hate to introduce another global variable if we can avoid it...
>
> Otherwise:
>
> Acked-by: Anthony Liguori <aliguori@us.ibm.com>
>
> Blue/Aurelien, any objections?

No, maybe a message could be printed that says that the default has
changed, for a few releases.

>
> Regards,
>
> Anthony Liguori
>
>>
>>      while (!accel_initialised && *p != '\0') {
>> --
>> 1.7.3.4

Re: [Qemu-devel] [PATCH] kvm: Set default accelerator to "kvm" if the host supports it

[Peter Maydell]

On 3 October 2012 21:01, Blue Swirl <blauwirbel@gmail.com> wrote:
> On Mon, Oct 1, 2012 at 4:20 PM, Anthony Liguori <anthony@codemonkey.ws> wrote:
>> Jan Kiszka <jan.kiszka@siemens.com> writes:
>>> +        /* The default accelerator depends on the availability of KVM. */
>>> +        p = kvm_configured ? "kvm" : "tcg";
>>>      }

>> Blue/Aurelien, any objections?
>
> No, maybe a message could be printed that says that the default has
> changed, for a few releases.

I've lost track of the conversation, are we currently proposing
the accelerator default to be "kvm" (as per the original patch
you quote here) or "kvm:tcg" ?

I'm not entirely sure which I prefer from an ARM perspective
For some time to come and for a lot of targets (ie any target
CPU except A15), having a default of "kvm" is going to cause
existing working commandlines to stop working. [I expect that
ARM-host qemu binaries will be built with CONFIG_KVM once ARM
KVM support lands, but the same binary will be run on hosts
without virtualization extensions.] On the other hand, perhaps
there just aren't really very many people who run QEMU on
ARM hosts, and so we can ignore them :-)

-- PMM

Re: [Qemu-devel] [PATCH] kvm: Set default accelerator to "kvm" if the host supports it

[Alexander Graf]

On 03.10.2012, at 22:26, Peter Maydell wrote:

> On 3 October 2012 21:01, Blue Swirl <blauwirbel@gmail.com> wrote:
>> On Mon, Oct 1, 2012 at 4:20 PM, Anthony Liguori <anthony@codemonkey.ws> wrote:
>>> Jan Kiszka <jan.kiszka@siemens.com> writes:
>>>> +        /* The default accelerator depends on the availability of KVM. */
>>>> +        p = kvm_configured ? "kvm" : "tcg";
>>>>     }
> 
>>> Blue/Aurelien, any objections?
>> 
>> No, maybe a message could be printed that says that the default has
>> changed, for a few releases.
> 
> I've lost track of the conversation, are we currently proposing
> the accelerator default to be "kvm" (as per the original patch
> you quote here) or "kvm:tcg" ?
> 
> I'm not entirely sure which I prefer from an ARM perspective
> For some time to come and for a lot of targets (ie any target
> CPU except A15), having a default of "kvm" is going to cause
> existing working commandlines to stop working. [I expect that
> ARM-host qemu binaries will be built with CONFIG_KVM once ARM
> KVM support lands, but the same binary will be run on hosts
> without virtualization extensions.] On the other hand, perhaps
> there just aren't really very many people who run QEMU on
> ARM hosts, and so we can ignore them :-)

We get similar problems on PPC. Take the following example:

  $ qemu-system-ppc -M mpc8544ds -kernel uImage -nographic

That command line would work just fine if you run it on a G5 today. However, if we switch to accel=kvm:tcg, it will stop working, because kvm on the G5 can not virtualize an e500 CPU.

I don't know any good way around that one either the way the code is layered today. We only know the type of CPU we want to create after we made the decision whether we do KVM or not.


Alex

Re: [Qemu-devel] [PATCH] kvm: Set default accelerator to "kvm" if the host supports it

[Anthony Liguori]

Alexander Graf <agraf@suse.de> writes:

> On 03.10.2012, at 22:26, Peter Maydell wrote:
>
>> On 3 October 2012 21:01, Blue Swirl <blauwirbel@gmail.com> wrote:
>>> On Mon, Oct 1, 2012 at 4:20 PM, Anthony Liguori <anthony@codemonkey.ws> wrote:
>>>> Jan Kiszka <jan.kiszka@siemens.com> writes:
>>>>> +        /* The default accelerator depends on the availability of KVM. */
>>>>> +        p = kvm_configured ? "kvm" : "tcg";
>>>>>     }
>> 
>>>> Blue/Aurelien, any objections?
>>> 
>>> No, maybe a message could be printed that says that the default has
>>> changed, for a few releases.
>> 
>> I've lost track of the conversation, are we currently proposing
>> the accelerator default to be "kvm" (as per the original patch
>> you quote here) or "kvm:tcg" ?
>> 
>> I'm not entirely sure which I prefer from an ARM perspective
>> For some time to come and for a lot of targets (ie any target
>> CPU except A15), having a default of "kvm" is going to cause
>> existing working commandlines to stop working. [I expect that
>> ARM-host qemu binaries will be built with CONFIG_KVM once ARM
>> KVM support lands, but the same binary will be run on hosts
>> without virtualization extensions.] On the other hand, perhaps
>> there just aren't really very many people who run QEMU on
>> ARM hosts, and so we can ignore them :-)
>
> We get similar problems on PPC. Take the following example:
>
>   $ qemu-system-ppc -M mpc8544ds -kernel uImage -nographic

But do you really expect people to do this?  I have to believe that
people running on PPC hardware and running qemu-system-ppc most likely
want to do KVM...

Kernel development seems unlikely...  The folks that I know doing PPC
kernel development with QEMU still qemu-system-ppc on x86.

Regards,

Anthony Liguori

>
> That command line would work just fine if you run it on a G5 today. However, if we switch to accel=kvm:tcg, it will stop working, because kvm on the G5 can not virtualize an e500 CPU.
>
> I don't know any good way around that one either the way the code is layered today. We only know the type of CPU we want to create after we made the decision whether we do KVM or not.
>
>
> Alex

Re: [Qemu-devel] [PATCH] kvm: Set default accelerator to "kvm" if the host supports it

[Alexander Graf]

On 05.10.2012, at 04:17, Anthony Liguori wrote:

> Alexander Graf <agraf@suse.de> writes:
> 
>> On 03.10.2012, at 22:26, Peter Maydell wrote:
>> 
>>> On 3 October 2012 21:01, Blue Swirl <blauwirbel@gmail.com> wrote:
>>>> On Mon, Oct 1, 2012 at 4:20 PM, Anthony Liguori <anthony@codemonkey.ws> wrote:
>>>>> Jan Kiszka <jan.kiszka@siemens.com> writes:
>>>>>> +        /* The default accelerator depends on the availability of KVM. */
>>>>>> +        p = kvm_configured ? "kvm" : "tcg";
>>>>>>    }
>>> 
>>>>> Blue/Aurelien, any objections?
>>>> 
>>>> No, maybe a message could be printed that says that the default has
>>>> changed, for a few releases.
>>> 
>>> I've lost track of the conversation, are we currently proposing
>>> the accelerator default to be "kvm" (as per the original patch
>>> you quote here) or "kvm:tcg" ?
>>> 
>>> I'm not entirely sure which I prefer from an ARM perspective
>>> For some time to come and for a lot of targets (ie any target
>>> CPU except A15), having a default of "kvm" is going to cause
>>> existing working commandlines to stop working. [I expect that
>>> ARM-host qemu binaries will be built with CONFIG_KVM once ARM
>>> KVM support lands, but the same binary will be run on hosts
>>> without virtualization extensions.] On the other hand, perhaps
>>> there just aren't really very many people who run QEMU on
>>> ARM hosts, and so we can ignore them :-)
>> 
>> We get similar problems on PPC. Take the following example:
>> 
>>  $ qemu-system-ppc -M mpc8544ds -kernel uImage -nographic
> 
> But do you really expect people to do this?  I have to believe that
> people running on PPC hardware and running qemu-system-ppc most likely
> want to do KVM...

Sure. But we wouldn't be able to even tell them what went wrong, as we don't have a negotiation mechanism right now that could tell user space "hey, the CPU you selected is unknown to me".

However, if during cpu init we could add such a check and then fall back to tcg mode if accel=kvm:tcg with a warning, that'd be nice user experience.

We could do the same for ARM. If you do -M beagle on an A15 KVM enabled machine, you would still be able to do so, but KVM tells you it can't emulate an A8 right now. And if in the future KVM learns how to expose an A8 on A15, we could just not bail out and things would magically work.

Apart from that, I like the idea of kvm:tcg with a warning as the default for qemu. We should still have a qemu-kvm binary in the distro that does accel=kvm so people don't accidentally fall back to tcg mode.


Alex

Re: [Qemu-devel] [PATCH] kvm: Set default accelerator to "kvm" if the host supports it

[Peter Maydell]

On 5 October 2012 03:24, Alexander Graf <agraf@suse.de> wrote:
> On 05.10.2012, at 04:17, Anthony Liguori wrote:
>> Alexander Graf <agraf@suse.de> writes:
>>> We get similar problems on PPC. Take the following example:
>>>
>>>  $ qemu-system-ppc -M mpc8544ds -kernel uImage -nographic
>>
>> But do you really expect people to do this?  I have to believe that
>> people running on PPC hardware and running qemu-system-ppc most likely
>> want to do KVM...
>
> Sure. But we wouldn't be able to even tell them what went wrong,
> as we don't have a negotiation mechanism right now that could tell
> user space "hey, the CPU you selected is unknown to me".

On ARM we will at least be able to say what happened, because
our kvm_arch_init_vcpu() will fail (when we try the ioctl to
tell the kernel "be this kind of CPU") and we can print a
message there. However QEMU as a whole doesn't have any way of
falling back to TCG at that point so it just bails out.

-- PMM

Re: [Qemu-devel] [PATCH] kvm: Set default accelerator to "kvm" if the host supports it

[Andreas Färber]

Am 05.10.2012 04:24, schrieb Alexander Graf:
> 
> On 05.10.2012, at 04:17, Anthony Liguori wrote:
> 
>> Alexander Graf <agraf@suse.de> writes:
>>
>>> On 03.10.2012, at 22:26, Peter Maydell wrote:
>>>
>>>> On 3 October 2012 21:01, Blue Swirl <blauwirbel@gmail.com> wrote:
>>>>> On Mon, Oct 1, 2012 at 4:20 PM, Anthony Liguori <anthony@codemonkey.ws> wrote:
>>>>>> Jan Kiszka <jan.kiszka@siemens.com> writes:
>>>>>>> +        /* The default accelerator depends on the availability of KVM. */
>>>>>>> +        p = kvm_configured ? "kvm" : "tcg";
>>>>>>>    }
>>>>
>>>>>> Blue/Aurelien, any objections?
>>>>>
>>>>> No, maybe a message could be printed that says that the default has
>>>>> changed, for a few releases.
>>>>
>>>> I've lost track of the conversation, are we currently proposing
>>>> the accelerator default to be "kvm" (as per the original patch
>>>> you quote here) or "kvm:tcg" ?
>>>>
>>>> I'm not entirely sure which I prefer from an ARM perspective
>>>> For some time to come and for a lot of targets (ie any target
>>>> CPU except A15), having a default of "kvm" is going to cause
>>>> existing working commandlines to stop working. [I expect that
>>>> ARM-host qemu binaries will be built with CONFIG_KVM once ARM
>>>> KVM support lands, but the same binary will be run on hosts
>>>> without virtualization extensions.] On the other hand, perhaps
>>>> there just aren't really very many people who run QEMU on
>>>> ARM hosts, and so we can ignore them :-)
>>>
>>> We get similar problems on PPC. Take the following example:
>>>
>>>  $ qemu-system-ppc -M mpc8544ds -kernel uImage -nographic
>>
>> But do you really expect people to do this?  I have to believe that
>> people running on PPC hardware and running qemu-system-ppc most likely
>> want to do KVM...
> 
> Sure. But we wouldn't be able to even tell them what went wrong, as we don't have a negotiation mechanism right now that could tell user space "hey, the CPU you selected is unknown to me".

Would it help to split out the cpu_model -> CPUClass lookup from
cpu_ppc_init() to invoke a hook or inquire a field indicating KVM support?

Andreas

> 
> However, if during cpu init we could add such a check and then fall back to tcg mode if accel=kvm:tcg with a warning, that'd be nice user experience.
> 
> We could do the same for ARM. If you do -M beagle on an A15 KVM enabled machine, you would still be able to do so, but KVM tells you it can't emulate an A8 right now. And if in the future KVM learns how to expose an A8 on A15, we could just not bail out and things would magically work.
> 
> Apart from that, I like the idea of kvm:tcg with a warning as the default for qemu. We should still have a qemu-kvm binary in the distro that does accel=kvm so people don't accidentally fall back to tcg mode.
> 
> 
> Alex
> 


-- 
SUSE LINUX Products GmbH, Maxfeldstr. 5, 90409 Nürnberg, Germany
GF: Jeff Hawn, Jennifer Guild, Felix Imendörffer; HRB 16746 AG Nürnberg

Re: [Qemu-devel] [PATCH] kvm: Set default accelerator to "kvm" if the host supports it

[Alexander Graf]

On 08.10.2012, at 16:03, Andreas Färber wrote:

> Am 05.10.2012 04:24, schrieb Alexander Graf:
>> 
>> On 05.10.2012, at 04:17, Anthony Liguori wrote:
>> 
>>> Alexander Graf <agraf@suse.de> writes:
>>> 
>>>> On 03.10.2012, at 22:26, Peter Maydell wrote:
>>>> 
>>>>> On 3 October 2012 21:01, Blue Swirl <blauwirbel@gmail.com> wrote:
>>>>>> On Mon, Oct 1, 2012 at 4:20 PM, Anthony Liguori <anthony@codemonkey.ws> wrote:
>>>>>>> Jan Kiszka <jan.kiszka@siemens.com> writes:
>>>>>>>> +        /* The default accelerator depends on the availability of KVM. */
>>>>>>>> +        p = kvm_configured ? "kvm" : "tcg";
>>>>>>>>   }
>>>>> 
>>>>>>> Blue/Aurelien, any objections?
>>>>>> 
>>>>>> No, maybe a message could be printed that says that the default has
>>>>>> changed, for a few releases.
>>>>> 
>>>>> I've lost track of the conversation, are we currently proposing
>>>>> the accelerator default to be "kvm" (as per the original patch
>>>>> you quote here) or "kvm:tcg" ?
>>>>> 
>>>>> I'm not entirely sure which I prefer from an ARM perspective
>>>>> For some time to come and for a lot of targets (ie any target
>>>>> CPU except A15), having a default of "kvm" is going to cause
>>>>> existing working commandlines to stop working. [I expect that
>>>>> ARM-host qemu binaries will be built with CONFIG_KVM once ARM
>>>>> KVM support lands, but the same binary will be run on hosts
>>>>> without virtualization extensions.] On the other hand, perhaps
>>>>> there just aren't really very many people who run QEMU on
>>>>> ARM hosts, and so we can ignore them :-)
>>>> 
>>>> We get similar problems on PPC. Take the following example:
>>>> 
>>>> $ qemu-system-ppc -M mpc8544ds -kernel uImage -nographic
>>> 
>>> But do you really expect people to do this?  I have to believe that
>>> people running on PPC hardware and running qemu-system-ppc most likely
>>> want to do KVM...
>> 
>> Sure. But we wouldn't be able to even tell them what went wrong, as we don't have a negotiation mechanism right now that could tell user space "hey, the CPU you selected is unknown to me".
> 
> Would it help to split out the cpu_model -> CPUClass lookup from
> cpu_ppc_init() to invoke a hook or inquire a field indicating KVM support?

Well, we need to basically determine whether KVM is enabled only after cpu creation of the machine file.


Alex

Re: [Qemu-devel] [PATCH] kvm: Set default accelerator to "kvm" if the host supports it

[Andreas Färber]

Hello Jan,

Am 01.10.2012 16:34, schrieb Jan Kiszka:
> If we built a target for a host that supports KVM in principle, set the
> default accelerator to KVM as well. This also means the start of QEMU
> will fail to start if KVM support turns out to be unavailable at
> runtime.

>From a distro point of view this of course means that we will build
against KVM and that the new KVM default will start to fail for users on
very old hardware. Can't we do a runtime check to select the default?

Would be nice to at least amend the commit message with how they are
expected to remedy that via command line. -machine accel=tcg?

Regards,
Andreas

> 
> Signed-off-by: Jan Kiszka <jan.kiszka@siemens.com>
> ---
>  kvm-all.c  |    1 +
>  kvm-stub.c |    1 +
>  kvm.h      |    1 +
>  vl.c       |    4 ++--
>  4 files changed, 5 insertions(+), 2 deletions(-)
> 
> diff --git a/kvm-all.c b/kvm-all.c
> index 92a7137..4d5f86c 100644
> --- a/kvm-all.c
> +++ b/kvm-all.c
> @@ -103,6 +103,7 @@ struct KVMState
>  #endif
>  };
>  
> +bool kvm_configured = true;
>  KVMState *kvm_state;
>  bool kvm_kernel_irqchip;
>  bool kvm_async_interrupts_allowed;
> diff --git a/kvm-stub.c b/kvm-stub.c
> index 3c52eb5..86a6451 100644
> --- a/kvm-stub.c
> +++ b/kvm-stub.c
> @@ -17,6 +17,7 @@
>  #include "gdbstub.h"
>  #include "kvm.h"
>  
> +bool kvm_configured;
>  KVMState *kvm_state;
>  bool kvm_kernel_irqchip;
>  bool kvm_async_interrupts_allowed;
> diff --git a/kvm.h b/kvm.h
> index dea2998..9936e5f 100644
> --- a/kvm.h
> +++ b/kvm.h
> @@ -22,6 +22,7 @@
>  #include <linux/kvm.h>
>  #endif
>  
> +extern bool kvm_configured;
>  extern int kvm_allowed;
>  extern bool kvm_kernel_irqchip;
>  extern bool kvm_async_interrupts_allowed;
> diff --git a/vl.c b/vl.c
> index 8d305ca..f557bd1 100644
> --- a/vl.c
> +++ b/vl.c
> @@ -2215,8 +2215,8 @@ static int configure_accelerator(void)
>      }
>  
>      if (p == NULL) {
> -        /* Use the default "accelerator", tcg */
> -        p = "tcg";
> +        /* The default accelerator depends on the availability of KVM. */
> +        p = kvm_configured ? "kvm" : "tcg";
>      }
>  
>      while (!accel_initialised && *p != '\0') {
> 


-- 
SUSE LINUX Products GmbH, Maxfeldstr. 5, 90409 Nürnberg, Germany
GF: Jeff Hawn, Jennifer Guild, Felix Imendörffer; HRB 16746 AG Nürnberg

Re: [Qemu-devel] [PATCH] kvm: Set default accelerator to "kvm" if the host supports it

[Daniel P. Berrange]

On Mon, Oct 01, 2012 at 06:43:00PM +0200, Andreas Färber wrote:
> Hello Jan,
> 
> Am 01.10.2012 16:34, schrieb Jan Kiszka:
> > If we built a target for a host that supports KVM in principle, set the
> > default accelerator to KVM as well. This also means the start of QEMU
> > will fail to start if KVM support turns out to be unavailable at
> > runtime.
> 
> From a distro point of view this of course means that we will build
> against KVM and that the new KVM default will start to fail for users on
> very old hardware. Can't we do a runtime check to select the default?

NB, this is *not* only about old hardware. There are plenty of users who
use QEMU inside VMs. One very common usage I know of is image building
tools which are run inside Amazon VMs, using libguestfs & QEMU.

IMHO, default to KVM, fallback to TCG is the most friendly default
behaviour.

Daniel
-- 
|: http://berrange.com      -o-    http://www.flickr.com/photos/dberrange/ :|
|: http://libvirt.org              -o-             http://virt-manager.org :|
|: http://autobuild.org       -o-         http://search.cpan.org/~danberr/ :|
|: http://entangle-photo.org       -o-       http://live.gnome.org/gtk-vnc :|

Re: [Qemu-devel] [PATCH] kvm: Set default accelerator to "kvm" if the host supports it

[Anthony Liguori]

"Daniel P. Berrange" <berrange@redhat.com> writes:

> On Mon, Oct 01, 2012 at 06:43:00PM +0200, Andreas Färber wrote:
>> Hello Jan,
>> 
>> Am 01.10.2012 16:34, schrieb Jan Kiszka:
>> > If we built a target for a host that supports KVM in principle, set the
>> > default accelerator to KVM as well. This also means the start of QEMU
>> > will fail to start if KVM support turns out to be unavailable at
>> > runtime.
>> 
>> From a distro point of view this of course means that we will build
>> against KVM and that the new KVM default will start to fail for users on
>> very old hardware. Can't we do a runtime check to select the default?
>
> NB, this is *not* only about old hardware. There are plenty of users who
> use QEMU inside VMs. One very common usage I know of is image building
> tools which are run inside Amazon VMs, using libguestfs & QEMU.

But libguest can set it's accelerator option to whatever it wants.

If your running QEMU under a VM, it's pretty reasonable to have to use a
special option IMHO.

> IMHO, default to KVM, fallback to TCG is the most friendly default
> behaviour.

Except if a user expects good network performance and can't
understand why they're getting 100kb/s.

Regards,

Anthony Liguori

>
> Daniel
> -- 
> |: http://berrange.com      -o-    http://www.flickr.com/photos/dberrange/ :|
> |: http://libvirt.org              -o-             http://virt-manager.org :|
> |: http://autobuild.org       -o-         http://search.cpan.org/~danberr/ :|
> |: http://entangle-photo.org       -o-       http://live.gnome.org/gtk-vnc :|

Re: [Qemu-devel] [PATCH] kvm: Set default accelerator to "kvm" if the host supports it

[Paolo Bonzini]

> But libguest can set it's accelerator option to whatever it wants.
> 
> If your running QEMU under a VM, it's pretty reasonable to have to
> use a special option IMHO.

It's also reasonable to have consecutive releases change defaults in
a more "friendly" way (i.e. from tcg to kvm:tcg), especially since
we'll get users that formerly used qemu-kvm and never had to specify
neither -machine accel nor --enable-kvm.

> > IMHO, default to KVM, fallback to TCG is the most friendly default
> > behaviour.
> 
> Except if a user expects good network performance and can't
> understand why they're getting 100kb/s.

libguestfs doesn't need network at all (though I wonder if they could
use lxc instead...).

Paolo

Re: [Qemu-devel] [PATCH] kvm: Set default accelerator to "kvm" if the host supports it

[Anthony Liguori]

Paolo Bonzini <pbonzini@redhat.com> writes:

>> But libguest can set it's accelerator option to whatever it wants.
>> 
>> If your running QEMU under a VM, it's pretty reasonable to have to
>> use a special option IMHO.
>
> It's also reasonable to have consecutive releases change defaults in
> a more "friendly" way (i.e. from tcg to kvm:tcg), especially since
> we'll get users that formerly used qemu-kvm and never had to specify
> neither -machine accel nor --enable-kvm.

I agree with you except for the 'kvm:tcg' part.

>
>> > IMHO, default to KVM, fallback to TCG is the most friendly default
>> > behaviour.
>> 
>> Except if a user expects good network performance and can't
>> understand why they're getting 100kb/s.
>
> libguestfs doesn't need network at all (though I wonder if they could
> use lxc instead...).

FWIW, libguestfs already uses -machine accel=kvm:tcg so we can
completely the libguestfs use-case for this discussion.

Regards,

Anthony Liguori

>
> Paolo

Re: [Qemu-devel] [PATCH] kvm: Set default accelerator to "kvm" if the host supports it

[Markus Armbruster]

"Daniel P. Berrange" <berrange@redhat.com> writes:

> On Mon, Oct 01, 2012 at 06:43:00PM +0200, Andreas Färber wrote:
>> Hello Jan,
>> 
>> Am 01.10.2012 16:34, schrieb Jan Kiszka:
>> > If we built a target for a host that supports KVM in principle, set the
>> > default accelerator to KVM as well. This also means the start of QEMU
>> > will fail to start if KVM support turns out to be unavailable at
>> > runtime.
>> 
>> From a distro point of view this of course means that we will build
>> against KVM and that the new KVM default will start to fail for users on
>> very old hardware. Can't we do a runtime check to select the default?
>
> NB, this is *not* only about old hardware. There are plenty of users who
> use QEMU inside VMs. One very common usage I know of is image building
> tools which are run inside Amazon VMs, using libguestfs & QEMU.
>
> IMHO, default to KVM, fallback to TCG is the most friendly default
> behaviour.

Friendly perhaps, generating an infinite series of questions "why is my
guest slow as molasses?" certainly.

And for each instance of the question, there's an unknown number of
users who give QEMU a quick try, screw up KVM unknowingly, observe the
glacial speed, and conclude it's crap.

Re: [Qemu-devel] [PATCH] kvm: Set default accelerator to "kvm" if the host supports it

[Aurelien Jarno]

On Tue, Oct 02, 2012 at 09:46:12AM +0200, Markus Armbruster wrote:
> "Daniel P. Berrange" <berrange@redhat.com> writes:
> 
> > On Mon, Oct 01, 2012 at 06:43:00PM +0200, Andreas Färber wrote:
> >> Hello Jan,
> >> 
> >> Am 01.10.2012 16:34, schrieb Jan Kiszka:
> >> > If we built a target for a host that supports KVM in principle, set the
> >> > default accelerator to KVM as well. This also means the start of QEMU
> >> > will fail to start if KVM support turns out to be unavailable at
> >> > runtime.
> >> 
> >> From a distro point of view this of course means that we will build
> >> against KVM and that the new KVM default will start to fail for users on
> >> very old hardware. Can't we do a runtime check to select the default?
> >
> > NB, this is *not* only about old hardware. There are plenty of users who
> > use QEMU inside VMs. One very common usage I know of is image building
> > tools which are run inside Amazon VMs, using libguestfs & QEMU.
> >
> > IMHO, default to KVM, fallback to TCG is the most friendly default
> > behaviour.
> 
> Friendly perhaps, generating an infinite series of questions "why is my
> guest slow as molasses?" certainly.
> 
> And for each instance of the question, there's an unknown number of
> users who give QEMU a quick try, screw up KVM unknowingly, observe the
> glacial speed, and conclude it's crap.
> 

That's why it should not fallback silently to TCG, but warn the user
about that.

On the other hand, on a machine without KVM support (which might just be
because of local policy admin policy which doesn't provide access the 
/dev/kvm, nested virtualization, etc.), if QEMU fails to start (while
previous versions were working), the user can conclude that QEMU is
crap.

-- 
Aurelien Jarno                          GPG: 1024D/F1BCDB73
aurelien@aurel32.net                 http://www.aurel32.net

Re: [Qemu-devel] [PATCH] kvm: Set default accelerator to "kvm" if the host supports it

[Michael Tokarev]

On 02.10.2012 11:46, Markus Armbruster wrote:
> "Daniel P. Berrange" <berrange@redhat.com> writes:

>> IMHO, default to KVM, fallback to TCG is the most friendly default
>> behaviour.
> 
> Friendly perhaps, generating an infinite series of questions "why is my
> guest slow as molasses?" certainly.

With a warning about "switching to slow emulation mode because .."
printed at startup that becomes a non-issue, because there's no
reason to ask more questions about why it is slow - it already
said why.  Yes some may try to ask what to do, which is different.

Every howto nowadays mentions kvm modules and /dev/kvm device
permissions.

> And for each instance of the question, there's an unknown number of
> users who give QEMU a quick try, screw up KVM unknowingly, observe the
> glacial speed, and conclude it's crap.

This is, again, I think, unfair.  With the warning message it becomes
more or less obvious.

If you're talking about users who run it with -daemonize argument -
this is a) stupid to do when TRYING it out, so it's not a big deal
to lose another stupid user, and b) qemu should init everything
first and throw all warnings and fatal errors before daemonizing,
if this is not the case it should be fixed in the code.

And if you're talking about management software (libvirt and others),
it controls all the required privileges already and explicitly
requests acceleration and other stuff.

So the best thing to do is what Daniel, Aurelien, Paolo and others
are suggested: accel=kvm:tcg with a warning.

Thanks,

/mjt

Re: [Qemu-devel] [PATCH] kvm: Set default accelerator to "kvm" if the host supports it

[Jan Kiszka]

[-- Attachment #1: Type: text/plain, Size: 2164 bytes --]

On 2012-10-03 08:58, Michael Tokarev wrote:
> On 02.10.2012 11:46, Markus Armbruster wrote:
>> "Daniel P. Berrange" <berrange@redhat.com> writes:
> 
>>> IMHO, default to KVM, fallback to TCG is the most friendly default
>>> behaviour.
>>
>> Friendly perhaps, generating an infinite series of questions "why is my
>> guest slow as molasses?" certainly.
> 
> With a warning about "switching to slow emulation mode because .."
> printed at startup that becomes a non-issue, because there's no
> reason to ask more questions about why it is slow - it already
> said why.  Yes some may try to ask what to do, which is different.
> 
> Every howto nowadays mentions kvm modules and /dev/kvm device
> permissions.
> 
>> And for each instance of the question, there's an unknown number of
>> users who give QEMU a quick try, screw up KVM unknowingly, observe the
>> glacial speed, and conclude it's crap.
> 
> This is, again, I think, unfair.  With the warning message it becomes
> more or less obvious.
> 
> If you're talking about users who run it with -daemonize argument -
> this is a) stupid to do when TRYING it out, so it's not a big deal
> to lose another stupid user, and b) qemu should init everything
> first and throw all warnings and fatal errors before daemonizing,
> if this is not the case it should be fixed in the code.
> 
> And if you're talking about management software (libvirt and others),
> it controls all the required privileges already and explicitly
> requests acceleration and other stuff.
> 
> So the best thing to do is what Daniel, Aurelien, Paolo and others
> are suggested: accel=kvm:tcg with a warning.

Well, we had a lot of problems with such a fallback in the past, but I
think we had no proper warnings back then.

I'm not fully believing in users will always realize the console
message. I would therefore suggest to change the window title of QEMU as
well if we fail to initialize some accelerator. Something like "QEMU
without KVM" (could be "QEMU without $FAILED_ACCEL" in the end, i.e. not
just for KVM). If that makes sense for everyone, I'll hack the required
patches.

Jan



[-- Attachment #2: OpenPGP digital signature --]
[-- Type: application/pgp-signature, Size: 259 bytes --]

Re: [Qemu-devel] qemu-kvm: remove "boot=on|off" drive parameter compatibility

[Alexander Graf]

On 01.10.2012, at 15:19, Anthony Liguori wrote:

> Jan Kiszka <jan.kiszka@siemens.com> writes:
> 
>> On 2012-10-01 11:31, Marcelo Tosatti wrote:
>> 
>> It's not just about default configs. We need to validate if the
>> migration formats are truly compatible (qemu-kvm -> QEMU, the other way
>> around definitely not). For the command line switches, we could provide
>> a wrapper script that translates them into upstream format or simply
>> ignores them. That should be harmless to carry upstream.
> 
> qemu-kvm has:
> 
> -no-kvm

Should be easy to have around as backwards compat hack. All it needs to do is set accel=tcg.

> -no-kvm-irqchip
> -no-kvm-pit
> -no-kvm-pit-reinjection

Those are quite important, as we need cmdline backwards compatibility.

> -tdf <- does nothing
> 
> There are replacements for all of the above.  If we need to add them to
> qemu.git, it's not big deal to add them.
> 
> -drive ...,boot= <- this is ignored

It's ignored, but useful for certain things. I don't know how many of our users use boot= today, but it's certainly still in the code, and supported. I honestly wouldn't mind to carry a SUSE specific patch that implements boot= for now until we can deem it deprecated enough that we can drop it.

> cpu_set command for CPU hotplug which is known broken in qemu-kvm.
> 
> testdev which is nice but only used for development
> 
> Default nic is rtl8139 vs. e1000.

Couldn't we have a machine option that tells us that -M pc-1.2 is really -M pc-kvm-1.2? That way we could implement the device difference, right?


Alex

> 
> Some logic to move change the default VGA ram size to 16mb for pc-1.2
> (QEMU uses 16mb by default now too).
> 
> I think at this point, none of this matters but I added the various
> distro maintainers to the thread.
> 
> I think it's time for the distros to drop qemu-kvm and just ship
> qemu.git.  Is there anything else that needs to happen to make that
> switch?
> 
> Regards,
> 
> Anthony Liguori
> 
>> 
>> But I would really stop worrying about the qemu-kvm code base.
>> 
>> Jan
>> 
>> -- 
>> Siemens AG, Corporate Technology, CT RTC ITP SDP-DE
>> Corporate Competence Center Embedded Linux
>> --
>> To unsubscribe from this list: send the line "unsubscribe kvm" in
>> the body of a message to majordomo@vger.kernel.org
>> More majordomo info at  http://vger.kernel.org/majordomo-info.html
> 

Re: [Qemu-devel] qemu-kvm: remove "boot=on|off" drive parameter compatibility

[Paolo Bonzini]

Il 01/10/2012 16:07, Alexander Graf ha scritto:
>>> -drive ...,boot= <- this is ignored
> 
> It's ignored, but useful for certain things. I don't know how many of
> our users use boot= today, but it's certainly still in the code, and
> supported. I honestly wouldn't mind to carry a SUSE specific patch
> that implements boot= for now until we can deem it deprecated enough
> that we can drop it.

Extboot is not going to be backported to QEMU, but SeaBIOS can now boot
from all emulated device models except MegaSAS.  So we may at most have
just the option, for example making it an alias for bootindex=1 on the
corresponding device.

Paolo

Re: [Qemu-devel] qemu-kvm: remove "boot=on|off" drive parameter compatibility

[Bruce Rogers]

 >>> On 10/1/2012 at 07:19 AM, Anthony Liguori <anthony@codemonkey.ws> wrote: 
> Jan Kiszka <jan.kiszka@siemens.com> writes:
> 
>> On 2012-10-01 11:31, Marcelo Tosatti wrote:
>>
>> It's not just about default configs. We need to validate if the
>> migration formats are truly compatible (qemu-kvm -> QEMU, the other way
>> around definitely not). For the command line switches, we could provide
>> a wrapper script that translates them into upstream format or simply
>> ignores them. That should be harmless to carry upstream.
> 
> qemu-kvm has:
> 
>  -no-kvm
>  -no-kvm-irqchip
>  -no-kvm-pit
>  -no-kvm-pit-reinjection
>  -tdf <- does nothing
> 
> There are replacements for all of the above.  If we need to add them to
> qemu.git, it's not big deal to add them.
> 
>  -drive ...,boot= <- this is ignored
> 
> cpu_set command for CPU hotplug which is known broken in qemu-kvm.
> 
> testdev which is nice but only used for development
> 
> Default nic is rtl8139 vs. e1000.
> 
> Some logic to move change the default VGA ram size to 16mb for pc-1.2
> (QEMU uses 16mb by default now too).
> 
> I think at this point, none of this matters but I added the various
> distro maintainers to the thread.
> 
> I think it's time for the distros to drop qemu-kvm and just ship
> qemu.git.  Is there anything else that needs to happen to make that
> switch?

We are seriously considering moving to qemu.git for our SP3 release of
SUSE SLES 11. There are just a handful of patches that provide the backwards
compatibility we need to maintain (default to kvm, default nic model,
vga ram size), so assuming there is a 100% commitment to fully supporting
kvm in qemu going forward (which I don't doubt) I think this is a good time
for us to make that switch.

Bruce

Re: [Qemu-devel] qemu-kvm: remove "boot=on|off" drive parameter compatibility

[Daniel P. Berrange]

On Mon, Oct 01, 2012 at 08:19:29AM -0500, Anthony Liguori wrote:
> Jan Kiszka <jan.kiszka@siemens.com> writes:
> I think at this point, none of this matters but I added the various
> distro maintainers to the thread.
> 
> I think it's time for the distros to drop qemu-kvm and just ship
> qemu.git.  Is there anything else that needs to happen to make that
> switch?

If that is upstream's recommendation, then I see no issue with switching
Fedora 19 / RHEL-7 to use qemu.git instead of qemu-kvm.git.

Regards,
Daniel
-- 
|: http://berrange.com      -o-    http://www.flickr.com/photos/dberrange/ :|
|: http://libvirt.org              -o-             http://virt-manager.org :|
|: http://autobuild.org       -o-         http://search.cpan.org/~danberr/ :|
|: http://entangle-photo.org       -o-       http://live.gnome.org/gtk-vnc :|