Re: [Qemu-devel] [PATCH 0/2] Make simpletrace work on Windows

[hkran <hkran@vnet.linux.ibm.com>]

On 09/20/2011 06:15 PM, Zhi Yong Wu wrote:
> On Tue, Sep 20, 2011 at 5:57 PM, Stefan Hajnoczi
> <stefanha@linux.vnet.ibm.com>  wrote:
>> On Tue, Sep 20, 2011 at 05:05:45PM +0800, hkran wrote:
>>> On 09/09/2011 05:37 PM, Stefan Hajnoczi wrote:
>>>> The 'simple' trace backend uses pthreads and does not work on Windows.  These
>>>> patches switch from pthreads to glib so that the code builds on all platforms
>>>> supported by glib.
>>>>
>>>> Only one thing I'm unhappy about: the simpletrace write-out thread used to
>>>> block all signals.  I have removed that code and don't expect glib to do it for
>>>> me.  I'm not sure if there is a problem if signal handlers are invoked in the
>>>> write-out thread instead of a QEMU thread.  Any thoughts?
>>>>
>>>> Stefan Hajnoczi (2):
>>>>    trace: portable simple trace backend using glib
>>>>    trace: use binary file open mode in simpletrace
>>>>
>>>>   trace/simple.c |   58 ++++++++++++++++++++++++++-----------------------------
>>>>   1 files changed, 27 insertions(+), 31 deletions(-)
>>>>
>>> Stefan,
>>>
>>> I applied the patch and make&install it.
>>>
>>> After a round of running of the qemu with the patch, a trace file is
>>> here, but when I want to open it like this,
>>> ./simpletrace.py trace-events trace-29948    //trace-29948 is my tracefile
>>>   an error occurs:
>>>
>>> Traceback (most recent call last):
>>>    File "./simpletrace.py", line 151, in<module>
>>>      run(Formatter())
>>>    File "./simpletrace.py", line 131, in run
>>>      events = parse_events(open(sys.argv[1], 'r'))
>>> IOError: [Errno 2] No such file or directory: 'trace-events'
>>>
>>> Am I using it in a right way?
>> Looks like your current working directory is scripts/ so simpletrace.py
>> will be unable to find the trace-events file which is in the parent
>> directory.
>>
>> Usually I stay in QEMU's root directory and just run:
>> $ qemu # ...generate the trace
>> $ scripts/simpletrace.py trace-events trace-$PID
> I know how to define my own event and play with it now. Very helpful
> for me to debug my functions. thanks.
>
>>> Additionally, There is something about WIN32 in patch, How can I
>>> compile a qemu running on windows? Could you give a reference?
>> Search for 'mingw' in qemu-doc.texi for instructions.
>>
>> Stefan
>>
>>
>
>
It took much long time to setup a mingw environment on windows to build 
qemu with the patch.
I enable the events bdrv_aio_readv and bdrv_aio_writev...... finally I 
got a trace, I paste a piece of that here:

bdrv_aio_readv 1172.216 bs=0x12b84a0 sector_num=0x538607 nb_sectors=0x40 
opaque=0xe1a1d40
bdrv_aio_readv 1.676 bs=0x12b96e0 sector_num=0x538607 nb_sectors=0x40 
opaque=0xe1a1d40
bdrv_aio_readv 811.555 bs=0x12b84a0 sector_num=0x538647 nb_sectors=0x37 
opaque=0xe1a1d40
bdrv_aio_readv 2.515 bs=0x12b96e0 sector_num=0x538647 nb_sectors=0x37 
opaque=0xe1a1d40
bdrv_aio_writev 3549944.197 bs=0x12b84a0 sector_num=0xf2b3f 
nb_sectors=0x48 opaque=0xe1a1d40
bdrv_aio_writev 4.190 bs=0x12b96e0 sector_num=0xf2b3f nb_sectors=0x48 
opaque=0xe1a1d40
bdrv_aio_writev 35316.500 bs=0x12b84a0 sector_num=0x13f7b7 
nb_sectors=0x60 opaque=0xe1a1d40
bdrv_aio_writev 3.911 bs=0x12b96e0 sector_num=0x13f7b7 nb_sectors=0x60 
opaque=0xe1a1d40
bdrv_aio_writev 27754.924 bs=0x12b84a0 sector_num=0x9efdf 
nb_sectors=0x50 opaque=0xe1a1d40
bdrv_aio_writev 3.911 bs=0x12b96e0 sector_num=0x9efdf nb_sectors=0x50 
opaque=0xe1a1d40
bdrv_aio_writev 30819.839 bs=0x12b84a0 sector_num=0xf36a7 
nb_sectors=0x48 opaque=0xe1a1d40

It looks working fine when windows serving a host.

For another thing, if I use virtio driver in my case like this:

qemu-system-i386.exe -m 786 -drive 
file=/c/setup/iso/xp_shanghai.img,if=virtio -sdl -net nic,model=virtio 
-net user

the guest can startup but soon go to crash without any traces. My question:
Can not virtio drivers be applied in the case that windows serve as a host?
furthermore, are there other limitations when qemu is running on windows?