[vfio-users] OVMF: Boots only first time after driver installation

[S B]

Thank you, I gziped the files, because they are really huge.
21734 is the first working boot.
21819 is the second boot, which does not work.
​
 trace-21734.gz
<https://drive.google.com/file/d/0BxYQ1T4T4Bx1SG5tR2t2VlBjR1E/view?usp=drive_web>
​​
 trace-21819.gz
<https://drive.google.com/file/d/0BxYQ1T4T4Bx1M3dNdzFmc094Qk0/view?usp=drive_web>
​

2016-01-05 20:07 GMT+01:00 Alex Williamson <alex.l.williamson at gmail.com>:

> On Tue, Jan 5, 2016 at 11:49 AM, S B <sb9sb0 at gmail.com> wrote:
>
>> The file hw/misc/vfio.c does not exist anymore. I only found DEBUG_VFIO
>> in include/hw/vfio/vfio-common.h.
>> I compiled the via makepkg and included the patch below, but the output
>> does not change.
>> I checked the line while compiling and it is changed. I compiled it twice
>> and installed it, but the output does not change.
>> Maybe it did sth wrong but I do not find my mistake. Any hints?
>>
>
> I'm not sure what you hope to accomplish from this, but vfio has switched
> to tracing for debug:
>
> http://git.qemu.org/?p=qemu.git;a=blob;f=docs/tracing.txt
>
> vfio tracing can be enabled by recompiling with the trace backend
> specified in that doc and using "vfio*" in the events file and possibly
> narrowing it down with various function name refinements.  This will result
> in an obscene amount of logging, you're not going to casually spot any
> problems, but if you know something about pci and you're willing to track
> the device state and the source code paths, have at it, the more the
> merrier.  Typically you'll also want to add x-mmap=off to the vfio-pci
> options in QEMU so that we can trace all mmio accesses to the device.
> Thanks,
>
> Alex
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://listman.redhat.com/archives/vfio-users/attachments/20160106/1a75d361/attachment.htm>