[libvirt] [V2 Patch] virsh: Change log level order

Supriya Kannery supriyak at in.ibm.com
Thu Jun 30 08:32:56 UTC 2011


On 05/27/2011 01:39 AM, Eric Blake wrote:
> On 05/20/2011 03:36 AM, Daniel P. Berrange wrote:
>> On Fri, May 20, 2011 at 03:05:28PM +0530, Supriya Kannery wrote:
>>> Change log level order so that messages at all other levels get logged
>>> for "DEBUG" level. Replace magic numbers with corresponding VSH_ERR_xx.
>>>
>>>   } vshErrorLevel;
>> NACK I don't think this bit is a good idea. In fact I think it
>> should be changed to match the log levels in src/util/logging.h
>> exactly, so we have the same behaviour for virsh and the rest
>> of libvirt.
>>
>>> @@ -2146,7 +2146,8 @@ cmdDominfo(vshControl *ctl, const vshCmd
>>>
>>>       /* Check and display whether the domain is persistent or not */
>>>       persistent = virDomainIsPersistent(dom);
>>> -    vshDebug(ctl, 5, "Domain persistent flag value: %d\n", persistent);
>>> +    vshDebug(ctl, VSH_ERR_DEBUG, "Domain persistent flag value: %d\n",
>>> +             persistent);
>>>       if (persistent<  0)
>> ACK to all these changes though, hardcoding the numbers instead of using
>> the enums was clearly bad.
>
> Unfortunately, I tried to apply this patch in order to pick out just the
> magic number removals, but your mailer mungled it beyond the point that
> git could understand it.  Can you fix the issues we've identified,
> including splitting this patch into two pieces (one with just the magic
> number removal, another with my suggestion of factoring the log level
> comparison into a helper function, and then just tweaking the helper
> function and command line -d parsing to nicely map log levels to the
> same values as for libvirtd)?

Sent patches separately for replacement of magic numbers and aligning
virsh loglevel usage to that of libvirt. Please review.
https://www.redhat.com/archives/libvir-list/2011-June/msg01574.html

> Also, since your mailer seems to be conspiring against us, have you
> tried 'git send-email'?  Or attaching rather than sending inline?
> Practice sending mail to yourself, and running that file through 'git
> am' to see what sort of problems we might have applying your patch.  Or,
> you could post a link to a public repository (github or repo.or.cz are
> both decent choices) with your commits pre-applied, to bypass your
> mailer idiosyncrasies.
>

This time verified the patches by sending to myself and applying using
'git am'.




More information about the libvir-list mailing list