Bug 831150

Created attachment 591160 [details]
Failed when using serial

Tried w/ debug w/ NET as well 

The job logs complains "System.Exception: Kernel debugging over the network fails due to the non-supported NIC 0.For a list of supported NICs please visit http://go.microsoft.com/fwlink/?LinkID=230754 or run the test from another debug connection. at KernelDebugLogoTestSetup.DebugSetupVerifier.DebugSetupVerify()"

I check the link ,found e1000 in support list ,or it means the Target VM (SUT) which I use virtio netkvm driver does not support kernel debug ?

(In reply to comment #2)
> Tried w/ debug w/ NET as well 
> 
> The job logs complains "System.Exception: Kernel debugging over the network
> fails due to the non-supported NIC 0.For a list of supported NICs please
> visit http://go.microsoft.com/fwlink/?LinkID=230754 or run the test from
> another debug connection. at
> KernelDebugLogoTestSetup.DebugSetupVerifier.DebugSetupVerify()"
> 
> I check the link ,found e1000 in support list ,or it means the Target VM
> (SUT) which I use virtio netkvm driver does not support kernel debug ?

Do you have e1000 installed during SVVP configuration?
If yes, why?

(In reply to comment #5)
> (In reply to comment #2)
> > Tried w/ debug w/ NET as well 
> > 
> > The job logs complains "System.Exception: Kernel debugging over the network
> > fails due to the non-supported NIC 0.For a list of supported NICs please
> > visit http://go.microsoft.com/fwlink/?LinkID=230754 or run the test from
> > another debug connection. at
> > KernelDebugLogoTestSetup.DebugSetupVerifier.DebugSetupVerify()"
> > 
> > I check the link ,found e1000 in support list ,or it means the Target VM
> > (SUT) which I use virtio netkvm driver does not support kernel debug ?
> 
> Do you have e1000 installed during SVVP configuration?
> If yes, why?
1 virtio+ 1 e1000 . I use e1000 network for remote debug .According to Vadim ,this is newly supported in win8.

In theory ,we should support remote debug via serial , usb . 
I tried remote debug via serial ,no idea why it failed .That's the reason I report this bug

My question is - if this is an "if implemented" requirement. Meaning that if we don't have debug capable NIC then we don't have to pass this test. But if there is such NIC in the system (e1000) - so we should be able to pass it. So no need to try and pass this test with NIC.

Also - according to job configuration:
3.in Job Configuration .Type 
Transport =serial
SerialTargetPort = 1
SerialHostPort = 1

So they are actually testing debugging through serial. Are you able to break into Windows 2012 guest with kernel debugger (windbg) manually?

(In reply to comment #7)
> My question is - if this is an "if implemented" requirement. Meaning that if
> we don't have debug capable NIC then we don't have to pass this test. But if
> there is such NIC in the system (e1000) - so we should be able to pass it.
> So no need to try and pass this test with NIC.

I know this ,But failed to remote debug it via serial 

> 
> Also - according to job configuration:
> 3.in Job Configuration .Type 
> Transport =serial
> SerialTargetPort = 1
> SerialHostPort = 1
That's what I have done before Test

> 
> So they are actually testing debugging through serial. Are you able to break
> into Windows 2012 guest with kernel debugger (windbg) manually?
Tried .
1.Start VM1 w/ -serial tcp:0:4444,server,nowait
2.Start the win2012 w/ -serial tcp:0:4444
3.enable debug mode in #msconfig
4.Reboot

Actual Results:
win2012 guests come into a black screen and never can be login the guests .

(In reply to comment #8)
> (In reply to comment #7)
> 
> Actual Results:
> win2012 guests come into a black screen and never can be login the guests .

So I think this should be the main focus to pass the test. First to be able to break into guest with serial connection.

(In reply to comment #9)
> (In reply to comment #8)
> > (In reply to comment #7)
> > 
> > Actual Results:
> > win2012 guests come into a black screen and never can be login the guests .
> 
> So I think this should be the main focus to pass the test. First to be able
> to break into guest with serial connection.

Let me update this job's status to make sure we are talking about the same thing :)
  
1.As you knows ,This job is focus on remote kernel debug .M$ support following TransportType : NET 1394 USB2 USB3 Serial .
2.We can choose one of the Transport type to run this job ,if this kind of transport type passed ,then this job will pass
3.I tried this job w/ serial support first ,CLI referring comment #8, The job failed ,logs referring to comment #1 ,and I try to do this manually ,but windows server 2012 guests stuck during boot when I enable debug mode. That's the reason I report the bug
4.I tried this job w/ Net protocol ,the job could pass .pls ignore comment #2, When I update comment #2 ,I only add 1 virtio netcard .after add one more e1000 netcard and use it as the remote debug transport ,no error occurs.
5.I still keep this bug because I  want someone could help me point out why  via serial this job failed

Mike, 

Thanks a lot for clarification!

Actually being unable to use serial for debugging bothers me a lot.
I was confused and thought that we cannot pass this HCK job at all. I suggest to open another BZ and describe there the problem with debugging guest over the serial and close this one (otherwise it looks like a blocker for SVVP).

Thanks,
Yan.

(In reply to comment #11)
> Mike, 
> 
> Thanks a lot for clarification!
> 
> Actually being unable to use serial for debugging bothers me a lot.
> I was confused and thought that we cannot pass this HCK job at all. I
> suggest to open another BZ and describe there the problem with debugging
> guest over the serial and close this one (otherwise it looks like a blocker
> for SVVP).
> 
> Thanks,
> Yan.

I did lots of research on this job ,and put everything on this bug bug forgot to draw a conclusion . Sorry for confusing you ,
Close this one ,I will reopen new bug for win8 guests can not be remote kernel debug via serial tomorrow 

Mike