On 08/26/2013 05:11 PM, Pavel Hrdina wrote:
On 26.8.2013 07:11, Alex Jia wrote:
> On 08/23/2013 06:22 PM, Guannan Ren wrote:
>> virsh cpu-stats guest --start 0 --count 3
>> It outputs right but the return value is 1 rather than 0
>> echo $?
>> 1
>
> It's ok with libvirt-0.10.2-23.el6.x86_64 on RHEL6.y, but failed with
> libvirt-1.1.1-2.el7.x86_64
> on RHEL7.y, it's a little wired, the commit 73b89ed8 is introduced since
> 2012-07-25, and
> Michal only switches it to c99 initialization of vshCmdDef after that,
> except this, nobody changes
> the cmdCPUStats(), maybe, we have different patches about cmdCPUStats()
> between RHEL6.y and RHEL7.y?
This bug has been introduced by commit a54f25a9.
Yes, thanks, I gave a error keyword 'cmdCPUStats' on the following cmdline.
# git blame tools/virsh-domain.c | grep cmdCPUStats
>
>> Found by running libvirt-autotest
>> ./run -t libvirt --tests virsh_cpu_stats
>> ---
>> tools/virsh-domain.c | 10 +++++++---
>> 1 file changed, 7 insertions(+), 3 deletions(-)
>>
>> diff --git a/tools/virsh-domain.c b/tools/virsh-domain.c
>> index b29f934..bcf495c 100644
>> --- a/tools/virsh-domain.c
>> +++ b/tools/virsh-domain.c
>> @@ -6350,7 +6350,9 @@ cmdCPUStats(vshControl *ctl, const vshCmd *cmd)
>>
>> if (!nparams) {
>> vshPrint(ctl, "%s", _("No per-CPU stats
available"));
>> - goto do_show_total;
>> + if (show_total)
>> + goto do_show_total;
>> + goto cleanup;
>> }
>>
>> if (VIR_ALLOC_N(params, nparams * MIN(show_count, 128))< 0)
>> @@ -6389,10 +6391,12 @@ cmdCPUStats(vshControl *ctl, const vshCmd *cmd)
>> }
>> VIR_FREE(params);
>>
>> -do_show_total:
>> - if (!show_total)
>> + if (!show_total) {
>> + ret = true;
>> goto cleanup;
>> + }
>>
>> +do_show_total:
>> /* get supported num of parameter for total statistics */
>> if ((nparams = virDomainGetCPUStats(dom, NULL, 0, -1, 1,
>> flags))< 0)
>> goto failed_stats;
>
> --
> libvir-list mailing list
> libvir-list(a)redhat.com
>
https://www.redhat.com/mailman/listinfo/libvir-list