[libvirt] [PATCH v3 08/12] conf: Allocate/release 'uid' and 'fid' in PCI address

Yi Min Zhao zyimin at linux.ibm.com
Fri Aug 24 07:00:35 UTC 2018



在 2018/8/23 下午7:01, Andrea Bolognani 写道:
> On Thu, 2018-08-23 at 18:01 +0800, Yi Min Zhao wrote:
>> 在 2018/8/22 下午5:56, Andrea Bolognani 写道:
>>> The patches I said I'd write are now on the list:
>>>
>>>     https://www.redhat.com/archives/libvir-list/2018-August/msg01105.html
>>>
>>> No reviews yet, we'll see whether they get ACKed or NACKed :)
>> I saw them. Could I give my ACKed?
> Feel free to do that, but before pushing I'd like to have Laine's
> ACK since he's the one who introduced the functions in the first
> place.
>
>> I have a question about your previous comment about error report.
>> You thought we should report more specific information.
>>
>>> +    virZPCIDeviceAddressPtr zpci = dev->addr.pci.zpci;
>>> +
>>> +    if (zpci && !dev->pciAddressExtFlags) {
>>> +        virReportError(VIR_ERR_INTERNAL_ERROR, "%s", _("zPCI is not supported."));
>>> +        return -1;
>>> +    }
>> It's called by all device types which possibly use PCI address.
>> I'm not sure how to report device's name in error string.
> I think reporting the address is enough. The idea is to give the
> user some information they can use to figure out which device is
> misconfigured rather than just telling them there is an issue and
> leaving it at that.
>
> I also think we might be at a point where enough big changes have
> been made that the best way forward is probably to post a v3 that
> includes those and then use that as the basis for more fine-tuning
> for minor stuff such as error messages and the like. Does that
> sound reasonable?
>
Yes, it makes sense.

I'm preparing V4 and could post it today.




More information about the libvir-list mailing list