Re: [Xen-devel] [PATCH 3/3] xen: Document XEN_SYSCTL_CPUPOOL_OP_RMCPU anomalous EBUSY result

[Juergen Gross <jgross@xxxxxxxx>]

On 15/04/16 16:12, Ian Jackson wrote:
> Dario Faggioli writes ("Re: [PATCH 3/3] xen: Document 
> XEN_SYSCTL_CPUPOOL_OP_RMCPU anomalous EBUSY result"):
>> However, a different return value for the super special case of
>> temporary pinning override could maybe be selected. I'm having a look,
>> and although I don't find one that could be seen as a perfect fit (that
>> would be EBUSY, which is taken!), what about one of these:
>>
>>   EEXIST        17      /* File exists */
>>   ENOTEMPTY     39      /* Directory not empty */
>>   EROFS         30      /* Read-only file system */
>>   ENOSPC        28      /* No space left on device */
> 
> Does the hypervisor currently use EAGAIN for anything ?

Yes. Especially XEN_SYSCTL_CPUPOOL_OP_RMCPU can return it already.
Before you are asking: Jan didn't want it to return EAGAIN in the
temporary pinning case as the hypervisor couldn't guarantee that
the situation will be resolved (opposed to the other usage of
EAGAIN).


Juergen


_______________________________________________
Xen-devel mailing list
Xen-devel@xxxxxxxxxxxxx
http://lists.xen.org/xen-devel