[Qemu-devel] bad USB tablet update rate on qemu-1.0

[Qemu-devel] bad USB tablet update rate on qemu-1.0

[Erik Rull]

Hi all,

coming from qemu 0.14 the usbdevice tablet update rate gets really bad in 
qemu-1.0 with the same guest.

The cursor more jumps with some delay than moving smoothly.

What could be the reason? Are there tablet options that might remove the 
jumping?

Best regards,

Erik

Re: [Qemu-devel] bad USB tablet update rate on qemu-1.0

[Anthony Liguori]

On 12/19/2011 03:33 PM, Erik Rull wrote:
> Hi all,
>
> coming from qemu 0.14 the usbdevice tablet update rate gets really bad in
> qemu-1.0 with the same guest.

What's the specific guest?

Regards,

Anthony Liguori

>
> The cursor more jumps with some delay than moving smoothly.
>
> What could be the reason? Are there tablet options that might remove the jumping?
>
> Best regards,
>
> Erik
>

Re: [Qemu-devel] bad USB tablet update rate on qemu-1.0

[Erik Rull]

Anthony Liguori wrote:
> On 12/19/2011 03:33 PM, Erik Rull wrote:
>> Hi all,
>>
>> coming from qemu 0.14 the usbdevice tablet update rate gets really bad in
>> qemu-1.0 with the same guest.
>
> What's the specific guest?
>
> Regards,
>
> Anthony Liguori
>

It's a Windows XP guest. It was fine in 0.14

Thanks.

Best regards,

Erik

Re: [Qemu-devel] bad USB tablet update rate on qemu-1.0

[Erik Rull]

Erik Rull wrote:
> Anthony Liguori wrote:
>> On 12/19/2011 03:33 PM, Erik Rull wrote:
>>> Hi all,
>>>
>>> coming from qemu 0.14 the usbdevice tablet update rate gets really bad in
>>> qemu-1.0 with the same guest.
>>
>> What's the specific guest?
>>
>> Regards,
>>
>> Anthony Liguori
>>
>
> It's a Windows XP guest. It was fine in 0.14
>
> Thanks.
>
> Best regards,
>
> Erik
>

Any progress here? I tested it on another CPU board there it was worse - 
only 1-2 cursor updates per second :-(
I tried to use the wacom-tablet but didn't find a driver that works.

Best regards,

Erik

Re: [Qemu-devel] bad USB tablet update rate on qemu-1.0

[Erik Rull]

Erik Rull wrote:
> Erik Rull wrote:
>> Anthony Liguori wrote:
>>> On 12/19/2011 03:33 PM, Erik Rull wrote:
>>>> Hi all,
>>>>
>>>> coming from qemu 0.14 the usbdevice tablet update rate gets really bad in
>>>> qemu-1.0 with the same guest.
>>>
>>> What's the specific guest?
>>>
>>> Regards,
>>>
>>> Anthony Liguori
>>>
>>
>> It's a Windows XP guest. It was fine in 0.14
>>
>> Thanks.
>>
>> Best regards,
>>
>> Erik
>>
>
> Any progress here? I tested it on another CPU board there it was worse -
> only 1-2 cursor updates per second :-(
> I tried to use the wacom-tablet but didn't find a driver that works.
>
> Best regards,
>
> Erik
>

Hi all,

today I tested some stuff with VNC for remote access of the VM - there the 
update rate of the USB tablet (mouse cursor) was really great. So the 
question is now:
- Same host operating system for 0.14 and 1.0
- Same command line parameters beside the vnc option
Why is the native mouse on the qemu 0.14 system good for both vnc and 
native and on the qemu 1.0 system only good for vnc and bad for the native 
hardware mouse?

Maybe this helps finding the source of evil...
If you have ideas what to change / patches to test, just let me know.

Best regards,

Erik

Re: [Qemu-devel] bad USB tablet update rate on qemu-1.0

[Erik Rull]

Erik Rull wrote:
> Erik Rull wrote:
>> Anthony Liguori wrote:
>>> On 12/19/2011 03:33 PM, Erik Rull wrote:
>>>> Hi all,
>>>>
>>>> coming from qemu 0.14 the usbdevice tablet update rate gets really bad in
>>>> qemu-1.0 with the same guest.
>>>
>>> What's the specific guest?
>>>
>>> Regards,
>>>
>>> Anthony Liguori
>>>
>>
>> It's a Windows XP guest. It was fine in 0.14
>>
>> Thanks.
>>
>> Best regards,
>>
>> Erik
>>
>
> Any progress here? I tested it on another CPU board there it was worse -
> only 1-2 cursor updates per second :-(
> I tried to use the wacom-tablet but didn't find a driver that works.
>
> Best regards,
>
> Erik
>

Hi all,

today I tested some stuff with VNC for remote access of the VM - there the 
update rate of the USB tablet (mouse cursor) was really great. So the 
question is now:
- Same host operating system for 0.14 and 1.0
- Same command line parameters beside the vnc option
Why is the native mouse on the qemu 0.14 system good for both vnc and 
native and on the qemu 1.0 system only good for vnc and bad for the native 
hardware mouse?

Maybe this helps finding the source of evil...
If you have ideas what to change / patches to test, just let me know.

Best regards,

Erik

Re: [Qemu-devel] bad USB tablet update rate on qemu-1.0

[Erik Rull]

Erik Rull wrote:
> Erik Rull wrote:
>> Erik Rull wrote:
>>> Anthony Liguori wrote:
>>>> On 12/19/2011 03:33 PM, Erik Rull wrote:
>>>>> Hi all,
>>>>>
>>>>> coming from qemu 0.14 the usbdevice tablet update rate gets really bad in
>>>>> qemu-1.0 with the same guest.
>>>>
>>>> What's the specific guest?
>>>>
>>>> Regards,
>>>>
>>>> Anthony Liguori
>>>>
>>>
>>> It's a Windows XP guest. It was fine in 0.14
>>>
>>> Thanks.
>>>
>>> Best regards,
>>>
>>> Erik
>>>
>>
>> Any progress here? I tested it on another CPU board there it was worse -
>> only 1-2 cursor updates per second :-(
>> I tried to use the wacom-tablet but didn't find a driver that works.
>>
>> Best regards,
>>
>> Erik
>>
>
> Hi all,
>
> today I tested some stuff with VNC for remote access of the VM - there the
> update rate of the USB tablet (mouse cursor) was really great. So the
> question is now:
> - Same host operating system for 0.14 and 1.0
> - Same command line parameters beside the vnc option
> Why is the native mouse on the qemu 0.14 system good for both vnc and
> native and on the qemu 1.0 system only good for vnc and bad for the native
> hardware mouse?
>
> Maybe this helps finding the source of evil...
> If you have ideas what to change / patches to test, just let me know.
>
> Best regards,
>
> Erik
>

Hi all,

I additionally tested -std vga to exclude that it might be related to the 
cirrus emulation - but same result. So there seems to be a difference 
between the captured cursor for the native X-Windows window and the VNC 
window that occured somewhere between 0.14 and 1.0.

Best regards,

Erik

Re: [Qemu-devel] bad USB tablet update rate on qemu-1.0

[Andreas Färber]

Am 19.01.2012 17:40, schrieb Erik Rull:
> Erik Rull wrote:
>> Erik Rull wrote:
>>> Erik Rull wrote:
>>>> Anthony Liguori wrote:
>>>>> On 12/19/2011 03:33 PM, Erik Rull wrote:
>>>>>> Hi all,
>>>>>>
>>>>>> coming from qemu 0.14 the usbdevice tablet update rate gets really
>>>>>> bad in
>>>>>> qemu-1.0 with the same guest.
>>>>>
>>>>> What's the specific guest?
>>>>>
>>>>> Regards,
>>>>>
>>>>> Anthony Liguori
>>>>>
>>>>
>>>> It's a Windows XP guest. It was fine in 0.14
>>>>
>>>> Thanks.
>>>>
>>>> Best regards,
>>>>
>>>> Erik
>>>>
>>>
>>> Any progress here? I tested it on another CPU board there it was worse -
>>> only 1-2 cursor updates per second :-(
>>> I tried to use the wacom-tablet but didn't find a driver that works.
>>>
>>> Best regards,
>>>
>>> Erik
>>>
>>
>> Hi all,
>>
>> today I tested some stuff with VNC for remote access of the VM - there
>> the
>> update rate of the USB tablet (mouse cursor) was really great. So the
>> question is now:
>> - Same host operating system for 0.14 and 1.0
>> - Same command line parameters beside the vnc option
>> Why is the native mouse on the qemu 0.14 system good for both vnc and
>> native and on the qemu 1.0 system only good for vnc and bad for the
>> native
>> hardware mouse?
>>
>> Maybe this helps finding the source of evil...
>> If you have ideas what to change / patches to test, just let me know.
>>
>> Best regards,
>>
>> Erik
>>
> 
> Hi all,
> 
> I additionally tested -std vga to exclude that it might be related to
> the cirrus emulation - but same result. So there seems to be a
> difference between the captured cursor for the native X-Windows window
> and the VNC window that occured somewhere between 0.14 and 1.0.

Then try `git bisect start v1.0 v0.14.0' to find out when exactly the
perceived behavior changed. :)

Andreas

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

Re: [Qemu-devel] bad USB tablet update rate on qemu-1.0

[Erik Rull]

Andreas Färber wrote:
> Am 19.01.2012 17:40, schrieb Erik Rull:
>> Erik Rull wrote:
>>> Erik Rull wrote:
>>>> Erik Rull wrote:
>>>>> Anthony Liguori wrote:
>>>>>> On 12/19/2011 03:33 PM, Erik Rull wrote:
>>>>>>> Hi all,
>>>>>>>
>>>>>>> coming from qemu 0.14 the usbdevice tablet update rate gets really
>>>>>>> bad in
>>>>>>> qemu-1.0 with the same guest.
>>>>>>
>>>>>> What's the specific guest?
>>>>>>
>>>>>> Regards,
>>>>>>
>>>>>> Anthony Liguori
>>>>>>
>>>>>
>>>>> It's a Windows XP guest. It was fine in 0.14
>>>>>
>>>>> Thanks.
>>>>>
>>>>> Best regards,
>>>>>
>>>>> Erik
>>>>>
>>>>
>>>> Any progress here? I tested it on another CPU board there it was worse -
>>>> only 1-2 cursor updates per second :-(
>>>> I tried to use the wacom-tablet but didn't find a driver that works.
>>>>
>>>> Best regards,
>>>>
>>>> Erik
>>>>
>>>
>>> Hi all,
>>>
>>> today I tested some stuff with VNC for remote access of the VM - there
>>> the
>>> update rate of the USB tablet (mouse cursor) was really great. So the
>>> question is now:
>>> - Same host operating system for 0.14 and 1.0
>>> - Same command line parameters beside the vnc option
>>> Why is the native mouse on the qemu 0.14 system good for both vnc and
>>> native and on the qemu 1.0 system only good for vnc and bad for the
>>> native
>>> hardware mouse?
>>>
>>> Maybe this helps finding the source of evil...
>>> If you have ideas what to change / patches to test, just let me know.
>>>
>>> Best regards,
>>>
>>> Erik
>>>
>>
>> Hi all,
>>
>> I additionally tested -std vga to exclude that it might be related to
>> the cirrus emulation - but same result. So there seems to be a
>> difference between the captured cursor for the native X-Windows window
>> and the VNC window that occured somewhere between 0.14 and 1.0.
>
> Then try `git bisect start v1.0 v0.14.0' to find out when exactly the
> perceived behavior changed. :)
>
> Andreas
>

Hi Andreas,

thanks for the hint, I haven't used git up to now. Very interesting feature.
I just did a clone of the current qemu-kvm (which I use) and started to 
bisect, but got an error, where I don't know how to proceed:
erik@debian:~/qemu-test/qemu-kvm$ git bisect good qemu-kvm-0.14.0
You need to start by "git bisect start"
Do you want me to do it for you [Y/n]?
erik@debian:~/qemu-test/qemu-kvm$ git bisect bad qemu-kvm-1.0
Bisecting: 2043 revisions left to test after this
fatal: Entry 'roms/seabios' not uptodate. Cannot merge.
erik@debian:~/qemu-test/qemu-kvm$

Any ideas?
I did a git pull, but I just got the response that it is already up to date...

Best regards,

Erik

Re: [Qemu-devel] bad USB tablet update rate on qemu-1.0

[Andreas Färber]

Hi Erik,

Am 19.01.2012 20:15, schrieb Erik Rull:
> Andreas Färber wrote:
>> Am 19.01.2012 17:40, schrieb Erik Rull:
>>> [...] there seems to be a
>>> difference between the captured cursor for the native X-Windows window
>>> and the VNC window that occured somewhere between 0.14 and 1.0.
>>
>> Then try `git bisect start v1.0 v0.14.0' to find out when exactly the
>> perceived behavior changed. :)

> I just did a clone of the current qemu-kvm (which I use) and started to
> bisect, but got an error, where I don't know how to proceed:
> erik@debian:~/qemu-test/qemu-kvm$ git bisect good qemu-kvm-0.14.0
> You need to start by "git bisect start"
> Do you want me to do it for you [Y/n]?
> erik@debian:~/qemu-test/qemu-kvm$ git bisect bad qemu-kvm-1.0
> Bisecting: 2043 revisions left to test after this
> fatal: Entry 'roms/seabios' not uptodate. Cannot merge.
> erik@debian:~/qemu-test/qemu-kvm$

Hm, did you maybe previously do a `git submodule init'? You may need to
run `git submodule update' then (which may fail as recently for roms/SLOF).

Otherwise, generally when it does not compile you can try `git bisect
skip' to try a different commit.

Andreas

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

Re: [Qemu-devel] bad USB tablet update rate on qemu-1.0

[Erik Rull]

Hi Andreas,

Andreas Färber wrote:
> Hi Erik,
>
> Am 19.01.2012 20:15, schrieb Erik Rull:
>> Andreas Färber wrote:
>>> Am 19.01.2012 17:40, schrieb Erik Rull:
>>>> [...] there seems to be a
>>>> difference between the captured cursor for the native X-Windows window
>>>> and the VNC window that occured somewhere between 0.14 and 1.0.
>>>
>>> Then try `git bisect start v1.0 v0.14.0' to find out when exactly the
>>> perceived behavior changed. :)
>
>> I just did a clone of the current qemu-kvm (which I use) and started to
>> bisect, but got an error, where I don't know how to proceed:
>> erik@debian:~/qemu-test/qemu-kvm$ git bisect good qemu-kvm-0.14.0
>> You need to start by "git bisect start"
>> Do you want me to do it for you [Y/n]?
>> erik@debian:~/qemu-test/qemu-kvm$ git bisect bad qemu-kvm-1.0
>> Bisecting: 2043 revisions left to test after this
>> fatal: Entry 'roms/seabios' not uptodate. Cannot merge.
>> erik@debian:~/qemu-test/qemu-kvm$
>
> Hm, did you maybe previously do a `git submodule init'? You may need to
> run `git submodule update' then (which may fail as recently for roms/SLOF).
>
> Otherwise, generally when it does not compile you can try `git bisect
> skip' to try a different commit.
>
> Andreas

Updated to the latest git version from github.com and it's fine now. I did 
a test bisectioning by just using bisect good and bad randomized. My test 
system is available again on Monday. I will keep you updated.

Best regards,

Erik

Re: [Qemu-devel] bad USB tablet update rate on qemu-1.0

[erik.rull]

Hi all,

I'm really sorry, but the bisectioning does not work for the versions that
I want to test.

I get a bunch of errors when bisectioning.

e.g. continuosly:
erik@debian:~/qemu-test/qemu-kvm$ make
  CC    x86_64-softmmu/monitor.o
In file included from /home/erik/qemu-test/qemu-kvm/qmp-commands.h:19,
                 from /home/erik/qemu-test/qemu-kvm/monitor.c:3145:
/home/erik/qemu-test/qemu-kvm/qapi-types.h:19:34: error:
qapi/qapi-types-core.h: No such file or directory
In file included from /home/erik/qemu-test/qemu-kvm/qmp-commands.h:19,
                 from /home/erik/qemu-test/qemu-kvm/monitor.c:3145:
/home/erik/qemu-test/qemu-kvm/qapi-types.h:21: error: expected expression
before 'typedef'
make[1]: *** [monitor.o] Error 1

Additionally there seem to be differences between the 1.0 version in git
and the 1.0 version for download - there it compiles... *argh*

Please help.
I tried several git bisect skip but that didn't help.
The commit that was tested is:
Bisecting: a merge base must be tested
[2d2339f995d7176dcb2de10d162aed323a1ffbf3] Merge commit
'f487d6278f75f84378833b8c3a67443346d639dc' into upstream-merge

I thought that only working/compiling stuff gets committed?

Btw. the master does not compile either!

Best regards,

Erik