On 13/01/15 20:51, Wei Liu wrote:
>
>>> +
>>> + vnuma = d->vnuma;
>>> + printk(" %u vnodes, %u vcpus:\n", vnuma->nr_vnodes,
>>> d->max_vcpus);
>>> + for ( i = 0; i < vnuma->nr_vnodes; i++ )
>>> + {
>>> + err = snprintf(keyhandler_scratch, 12, "%3u",
>>> + vnuma->vnode_to_pnode[i]);
>>> + if ( err < 0 || vnuma->vnode_to_pnode[i] == NUMA_NO_NODE )
>>> + strlcpy(keyhandler_scratch, "???", 3);
>>> +
>>> + printk(" vnode %3u - pnode %s\n", i,
>>> keyhandler_scratch);
>>> + for ( j = 0; j < vnuma->nr_vmemranges; j++ )
>>> + {
>>> + if ( vnuma->vmemrange[j].nid == i )
>>> + {
>>> + printk(" %016"PRIx64" - %016"PRIx64"\n",
>>> + vnuma->vmemrange[j].start,
>>> + vnuma->vmemrange[j].end);
>>> + }
>>> + }
>>> +
>>> + printk(" vcpus: ");
>>> + for ( j = 0, n = 0; j < d->max_vcpus; j++ )
>>> + {
>>> + if ( !(j & 0x3f) )
>>> + process_pending_softirqs();
>>> +
>>> + if ( vnuma->vcpu_to_vnode[j] == i )
>>> + {
>>> + if ( (n + 1) % 8 == 0 )
>>> + printk("%3d\n", j);
>>> + else if ( !(n % 8) && n != 0 )
>>> + printk("%17d ", j);
>>> + else
>>> + printk("%3d ", j);
>>> + n++;
>>> + }
>> Do you have a sample of what this looks like for a VM with more than 8
>> vcpus ?
>>
> (XEN) Memory location of each domain:
> (XEN) Domain 0 (total: 131072):
> (XEN) Node 0: 131072
> (XEN) Domain 6 (total: 768064):
> (XEN) Node 0: 768064
> (XEN) 2 vnodes, 20 vcpus:
> (XEN) vnode 0 - pnode 0
> (XEN) 0000000000000000 - 000000005dc00000
> (XEN) vcpus: 0 1 2 3 4 5 6 7
> (XEN) 8 9
> (XEN) vnode 1 - pnode 0
> (XEN) 000000005dc00000 - 00000000bb000000
> (XEN) 0000000100000000 - 0000000100800000
> (XEN) vcpus: 10 11 12 13 14 15 16 17
> (XEN) 18 19
This is very verbose, and (IMO) moderately difficult to parse even
knowing that the information is.
May I suggest the following sylistic changes:
2 vnodes, 20 vcpus:
1: pnode 0, vcpus {1-9}
0000000000000000 - 000000005dc00000
2: pnode 1, vcpus {10-20}
000000005dc00000 - 00000000bb000000
0000000100000000 - 0000000100800000
You have already stated 2 vnodes, so "vnode $X" is redundant as the list
index. The vcpus are exceedingly likely to be consecutively allocated,
and cpumask_scnprintf() is a very concise way of representing them (and
will reduce your code quite a bit).
I might even be tempted to suggest adding something like "guest physical
layout" to the "2 vnodes, 20 vcpus" line, to make explicitly clear that
the memory ranges are gfns and not mfns/something else. After all, this
debugkey is mixing host and guest memory attributes.
~Andrew
_______________________________________________
Xen-devel mailing list
Xen-devel@lists.xen.org
http://lists.xen.org/xen-devel
