[libvirt] [PATCH] virsh: report error if vcpu number exceed the guest maxvcpu number

John Ferlan jferlan at redhat.com
Wed Jul 1 20:29:03 UTC 2015



On 06/28/2015 10:10 PM, Luyao Huang wrote:
> If usr pass a vcpu which exceed guest maxvcpu number, virsh client
> will only output an header like this:
> 
>  # virsh vcpupin test3 1000
>  VCPU: CPU Affinity
>  ----------------------------------
> 
> After this patch:
> 
>  # virsh vcpupin test3 1000
>  error: vcpu 1000 is out of range of cpu count 2
> 
> Signed-off-by: Luyao Huang <lhuang at redhat.com>
> ---
>  tools/virsh-domain.c | 5 +++++
>  1 file changed, 5 insertions(+)
> 

Seemed odd that this check wasn't there - so I did some digging...

Pavel's commit id '81dd81e' removed a check that seems to be what is
desired in this path (or was there before his change):

    if (vcpu >= info.nrVirtCpu) {
        vshError(ctl, "%s", _("vcpupin: vCPU index out of range."));
        goto cleanup;
        return false;
    }

As part of this commit, you'll note there was a test change in
"tests/vcpupin":

 # An out-of-range vCPU number deserves a diagnostic, too.
 $abs_top_builddir/tools/virsh --connect test:///default vcpupin test
100 0,1 > out 2>&1
 test $? = 1 || fail=1
 cat <<\EOF > exp || fail=1
-error: vcpupin: vCPU index out of range.
+error: invalid argument: requested vcpu is higher than allocated vcpus


Searching on their error message lands one in test_driver.c/
testDomainPinVcpu().  So something specific for a set path, but the path
you're hitting is the get path.


FWIW: If a similar test was run on my system I get:
# virsh vcpupin $dom 1000 0,1
error: invalid argument: vcpu 1000 is out of range of live cpu count 2

#



So, if I understand everything that was done - then while your change is
mostly correct, I think you could perhaps message the error similar to
the vshCPUCountCollect failure (see the attached patch)


Before I make that change for you - hopefully Pavel can take a look as
well to be sure I haven't missed something.

With any luck we this could be addressed before the 1.2.17 release, but
if not since it's been a regression since 1.2.13 and no one's noticed,
then another release probably won't hurt.

John


> diff --git a/tools/virsh-domain.c b/tools/virsh-domain.c
> index 27d62e9..681fc1a 100644
> --- a/tools/virsh-domain.c
> +++ b/tools/virsh-domain.c
> @@ -6497,6 +6497,11 @@ cmdVcpuPin(vshControl *ctl, const vshCmd *cmd)
>              goto cleanup;
>          }
>  
> +        if (got_vcpu && vcpu >= ncpus) {
> +            vshError(ctl, _("vcpu %d is out of range of cpu count %d"), vcpu, ncpus);
> +            goto cleanup;
> +        }
> +
>          cpumaplen = VIR_CPU_MAPLEN(maxcpu);
>          cpumap = vshMalloc(ctl, ncpus * cpumaplen);
>          if ((ncpus = virDomainGetVcpuPinInfo(dom, ncpus, cpumap,
> 

-------------- next part --------------
A non-text attachment was scrubbed...
Name: 0001-virsh-report-error-if-vcpu-number-exceed-the-guest-m.patch
Type: text/x-patch
Size: 1857 bytes
Desc: not available
URL: <http://listman.redhat.com/archives/libvir-list/attachments/20150701/5d7c91b2/attachment-0001.bin>


More information about the libvir-list mailing list