Re: [pve-devel] [PATCH qemu-server 3/4] check_local_resources: add more info per mapped device

[Dominik Csapak <d.csapak@proxmox.com>]

On 3/22/24 14:37, Fiona Ebner wrote:
> Am 18.03.24 um 12:18 schrieb Dominik Csapak:
>> such as the mapping name and if it's marked for live-migration (pci only)
>>
>> Signed-off-by: Dominik Csapak <d.csapak@proxmox.com>
>> ---
>>   PVE/API2/Qemu.pm   |  2 +-
>>   PVE/QemuMigrate.pm |  5 +++--
>>   PVE/QemuServer.pm  | 10 ++++++----
>>   3 files changed, 10 insertions(+), 7 deletions(-)
>>
>> diff --git a/PVE/API2/Qemu.pm b/PVE/API2/Qemu.pm
>> index 497987ff..4ecaeb91 100644
>> --- a/PVE/API2/Qemu.pm
>> +++ b/PVE/API2/Qemu.pm
>> @@ -4516,7 +4516,7 @@ __PACKAGE__->register_method({
>>   	$res->{local_disks} = [ values %$local_disks ];;
>>   
>>   	$res->{local_resources} = $local_resources;
>> -	$res->{'mapped-resources'} = $mapped_resources;
>> +	$res->{'mapped-resources'} =  [ map { "$_->{key}" } $mapped_resources->@* ];
> 
> Or should it become a hash? Then you could use 'keys' instead of map and
> a 'key' property.

yes that makes sense. Since we change the internal structure, we have
to adapt all call sites anyway...


> 
>>   
>>   	return $res;
>>   
>> diff --git a/PVE/QemuMigrate.pm b/PVE/QemuMigrate.pm
>> index 8d9b35ae..6fe8157d 100644
>> --- a/PVE/QemuMigrate.pm
>> +++ b/PVE/QemuMigrate.pm
>> @@ -232,7 +232,7 @@ sub prepare {
>>       my ($loc_res, $mapped_res, $missing_mappings_by_node) = PVE::QemuServer::check_local_resources($conf, 1);
>>       my $blocking_resources = [];
>>       for my $res ($loc_res->@*) {
>> -	if (!grep($res, $mapped_res->@*)) {
> 
> Seems like this is currently broken. I.e. it should be $res eq $_

yep, seems i have to improve my tests running with that
(maybe i can whip up some automated ones to send along)

> 
>> +	if (!grep { $_->{key} eq $res } $mapped_res->@*) {
>>   	    push $blocking_resources->@*, $res;
>>   	}
>>       }
>> @@ -246,8 +246,9 @@ sub prepare {
>>   
>>       if (scalar($mapped_res->@*)) {
>>   	my $missing_mappings = $missing_mappings_by_node->{$self->{node}};
>> +	my $mapped_text = join(", ", map { $_->{key} } $mapped_res->@*);
>>   	if ($running) {
>> -	    die "can't migrate running VM which uses mapped devices: " . join(", ", $mapped_res->@*) . "\n";
>> +	    die "can't migrate running VM which uses mapped devices: " . $mapped_text . "\n";
> 
> Style nit: no need for the concatenation anymore, can just use the
> variable inside the string.
> 
>>   	} elsif (scalar($missing_mappings->@*)) {
>>   	    die "can't migrate to '$self->{node}': missing mapped devices " . join(", ", $missing_mappings->@*) . "\n";
>>   	} else {
>> diff --git a/PVE/QemuServer.pm b/PVE/QemuServer.pm
>> index 6e2c8052..ef3aee20 100644
>> --- a/PVE/QemuServer.pm
>> +++ b/PVE/QemuServer.pm
>> @@ -2600,14 +2600,16 @@ sub check_local_resources {
>>   	    next if $entry->{host} && $entry->{host} =~ m/^spice$/i;
>>   	    if ($entry->{mapping}) {
>>   		$add_missing_mapping->('usb', $k, $entry->{mapping});
>> -		push @$mapped_res, $k;
>> +		push @$mapped_res, { key => $k, device => $entry->{mapping} };
> 
> Should we name the second key 'name'/'mapping'/'id' instead of 'device'?

yeah name is probably better here...

> 
>>   	    }
>>   	}
>>   	if ($k =~ m/^hostpci/) {
>>   	    my $entry = parse_property_string('pve-qm-hostpci', $conf->{$k});
>> -	    if ($entry->{mapping}) {
>> -		$add_missing_mapping->('pci', $k, $entry->{mapping});
>> -		push @$mapped_res, $k;
>> +	    if (my $name = $entry->{mapping}) {
>> +		$add_missing_mapping->('pci', $k, $name);
>> +		my $mapped_device = { key => $k, device => $name };
>> +		$mapped_device->{'live-migration'} = 1 if $pci_map->{ids}->{$name}->{'live-migration-capable'};
> 
> Style nit: line too long
> 
>> +		push @$mapped_res, $mapped_device;
>>   	    }
>>   	}
>>   	# sockets are safe: they will recreated be on the target side post-migrate