[pve-devel] [PATCH v3 qemu-server 1/2] add virtio-mem support

[Alexandre Derumier <aderumier@odiso.com>]

This patch add virtio-mem support, through a new maxmemory option.

a 4GB static memory is needed for DMA+boot memory, as this memory
is almost always un-unpluggeable.

1 virtio-mem pci device is setup for each numa node on pci.4 bridge

virtio-mem use a fixed blocksize with 32k max blocksize,
so blocksize is computed from the maxmemory/32000 with a minimum of
2MB to map THP.
(lower blocksize = more chance to unplug memory).

Signed-off-by: Alexandre Derumier <aderumier@odiso.com>
---
 PVE/QemuServer.pm        |   9 ++-
 PVE/QemuServer/Memory.pm | 141 ++++++++++++++++++++++++++++++---------
 PVE/QemuServer/PCI.pm    |   8 +++
 3 files changed, 126 insertions(+), 32 deletions(-)

diff --git a/PVE/QemuServer.pm b/PVE/QemuServer.pm
index 9a20647..d0f6d1f 100644
--- a/PVE/QemuServer.pm
+++ b/PVE/QemuServer.pm
@@ -338,6 +338,13 @@ my $confdesc = {
 	maximum => 262144,
 	default => 'cgroup v1: 1024, cgroup v2: 100',
     },
+    'memory_max' => {
+	optional => 1,
+	type => 'integer',
+	description => "Max hotpluggable virtio-mem memory",
+	minimum => 4096,
+	default => undef,
+    },
     memory => {
 	optional => 1,
 	type => 'integer',
@@ -3858,7 +3865,7 @@ sub config_to_command {
 	push @$cmd, get_cpu_options($conf, $arch, $kvm, $kvm_off, $machine_version, $winversion, $gpu_passthrough);
     }
 
-    PVE::QemuServer::Memory::config($conf, $vmid, $sockets, $cores, $defaults, $hotplug_features, $cmd);
+    PVE::QemuServer::Memory::config($conf, $vmid, $sockets, $cores, $defaults, $hotplug_features, $cmd, $devices, $bridges, $arch, $machine_type);
 
     push @$cmd, '-S' if $conf->{freeze};
 
diff --git a/PVE/QemuServer/Memory.pm b/PVE/QemuServer/Memory.pm
index 013917e..ed7eff6 100644
--- a/PVE/QemuServer/Memory.pm
+++ b/PVE/QemuServer/Memory.pm
@@ -8,9 +8,45 @@ use PVE::Exception qw(raise raise_param_exc);
 
 use PVE::QemuServer;
 use PVE::QemuServer::Monitor qw(mon_cmd);
+use PVE::QemuServer::PCI qw(print_pci_addr);
 
 my $MAX_NUMA = 8;
-my $STATICMEM = 1024;
+
+my sub get_static_mem {
+    my ($conf, $defaults) = @_;
+
+    my $sockets = 1;
+    $sockets = $conf->{smp} if $conf->{smp}; # old style - no longer iused
+    $sockets = $conf->{sockets} if $conf->{sockets};
+    my $hotplug_features = PVE::QemuServer::parse_hotplug_features(defined($conf->{hotplug}) ? $conf->{hotplug} : '1');
+
+    my $static_memory = 0;
+
+    if ($hotplug_features->{memory} || $conf->{'memory_max'}) {
+       $static_memory = 1024;
+       $static_memory = $static_memory * $sockets if ($conf->{hugepages} && $conf->{hugepages} == 1024);
+       $static_memory = 4096 if $conf->{'memory_max'};
+    } else {
+       $static_memory = $conf->{memory} || $defaults->{memory};
+    }
+
+    return $static_memory;
+}
+
+my sub get_virtiomem_block_size {
+    my ($conf, $static_memory) = @_;
+
+    my $maxmemory = $conf->{'memory_max'};
+    return undef if !$maxmemory;
+
+    #virtiomem can map 32000 block size. try to use lowerst blocksize, lower = more chance to unplug memory.
+    my $blocksize = ($maxmemory - $static_memory) / 32000;
+    #round next power of 2
+    $blocksize = 2**(int(log($blocksize)/log(2))+1);
+    #2MB is the minimum to be aligned with THP
+    $blocksize = 2 if $blocksize < 2;
+    return $blocksize;
+}
 
 my $_host_bits;
 my sub get_host_phys_address_bits {
@@ -58,7 +94,14 @@ my sub get_max_mem {
     # remove 20 bits to get MB and half that as QEMU needs some overhead
     my $bits_to_max_mem = int(1 << ($bits - 21));
 
-    return $bits_to_max_mem > 4*1024*1024 ? 4*1024*1024 : $bits_to_max_mem;
+    my $max_mem = $bits_to_max_mem > 4*1024*1024 ? 4*1024*1024 : $bits_to_max_mem;
+
+    if($conf->{'memory_max'}) {
+	die "memory_max can't be bigger than $max_mem MB" if $conf->{'memory_max'} > $max_mem;
+	return $conf->{'memory_max'};
+    }
+
+    return $max_mem;
 }
 
 sub get_numa_node_list {
@@ -152,6 +195,8 @@ sub foreach_reverse_dimm {
     }
 }
 
+
+
 sub qemu_memory_hotplug {
     my ($vmid, $conf, $defaults, $opt, $value) = @_;
 
@@ -164,14 +209,46 @@ sub qemu_memory_hotplug {
     $value = $defaults->{memory} if !$value;
     return $value if $value == $memory;
 
-    my $static_memory = $STATICMEM;
-    $static_memory = $static_memory * $sockets if ($conf->{hugepages} && $conf->{hugepages} == 1024);
+    my $static_memory = get_static_mem($conf, $defaults);
+    my $max_mem = get_max_mem($conf);
 
     die "memory can't be lower than $static_memory MB" if $value < $static_memory;
-    my $MAX_MEM = get_max_mem($conf);
-    die "you cannot add more memory than max mem $MAX_MEM MB!\n" if $memory > $MAX_MEM;
+    die "you cannot add more memory than max mem $max_mem MB!\n" if $value > $max_mem;
+
+   if ($conf->{'memory_max'}) {
+
+	die "memory size need to be multiple of 32MB when memory_max is defined" if $value % 32 != 0;
+
+	my $requested_size = ($value - $static_memory) / $sockets * 1024 * 1024;
 
-    if ($value > $memory) {
+	my $totalsize = $static_memory;
+	my $err = undef;
+
+	for (my $i = 0; $i < $sockets; $i++)  {
+
+	    my $id = "virtiomem$i";
+	    my $retry = 0;
+	    mon_cmd($vmid, 'qom-set', path => "/machine/peripheral/$id", property => "requested-size", value => int($requested_size));
+
+	    my $size = 0;
+	    while (1) {
+		sleep 1;
+		$size = mon_cmd($vmid, 'qom-get', path => "/machine/peripheral/$id", property => "size");
+		$err = 1 if $retry > 5;
+		last if $size eq $requested_size || $retry > 5;
+		$retry++;
+	    }
+	    $totalsize += ($size / 1024 / 1024 );
+	}
+	#update conf after each succesful module unplug
+	if($err) {
+	    $conf->{memory} = $totalsize;
+	    PVE::QemuConfig->write_config($vmid, $conf);
+	    raise_param_exc({ 'memory' => "error modify virtio memory" }) if $err;
+	}
+	return $totalsize;
+
+    } elsif($value > $memory) {
 
 	my $numa_hostmap;
 
@@ -266,33 +343,27 @@ sub qemu_dimm_list {
 }
 
 sub config {
-    my ($conf, $vmid, $sockets, $cores, $defaults, $hotplug_features, $cmd) = @_;
+    my ($conf, $vmid, $sockets, $cores, $defaults, $hotplug_features, $cmd, $devices, $bridges, $arch, $machine_type) = @_;
 
     my $memory = $conf->{memory} || $defaults->{memory};
-    my $static_memory = 0;
+    my $static_memory = get_static_mem($conf, $defaults);
 
-    if ($hotplug_features->{memory}) {
+    if ($hotplug_features->{memory} || $conf->{'memory_max'}) {
 	die "NUMA needs to be enabled for memory hotplug\n" if !$conf->{numa};
-	my $MAX_MEM = get_max_mem($conf);
-	die "Total memory is bigger than ${MAX_MEM}MB\n" if $memory > $MAX_MEM;
+	my $max_mem = get_max_mem($conf);
+	die "Total memory is bigger than ${max_mem}MB\n" if $memory > $max_mem;
 
 	for (my $i = 0; $i < $MAX_NUMA; $i++) {
 	    die "cannot enable memory hotplugging with custom NUMA topology\n"
 		if $conf->{"numa$i"};
 	}
 
-	my $sockets = 1;
-	$sockets = $conf->{sockets} if $conf->{sockets};
-
-	$static_memory = $STATICMEM;
-	$static_memory = $static_memory * $sockets if ($conf->{hugepages} && $conf->{hugepages} == 1024);
-
-	die "minimum memory must be ${static_memory}MB\n" if($memory < $static_memory);
-	push @$cmd, '-m', "size=${static_memory},slots=255,maxmem=${MAX_MEM}M";
+	die "memory size need to be multiple of 32MB when memory_max is defined" if $conf->{'memory_max'} && ($memory % 32 != 0);
+	my $cmdstr = "size=${static_memory},maxmem=${max_mem}M";
+	$cmdstr .= ",slots=255" if !$conf->{'memory_max'};
+	push @$cmd, '-m', $cmdstr;
 
     } else {
-
-	$static_memory = $memory;
 	push @$cmd, '-m', $static_memory;
     }
 
@@ -359,7 +430,21 @@ sub config {
 	}
     }
 
-    if ($hotplug_features->{memory}) {
+    if ($conf->{'memory_max'}) {
+	my $node_maxmem = ($conf->{'memory_max'} - $static_memory) / $sockets;
+	my $node_mem = ($memory - $static_memory) / $sockets;
+	my $blocksize = get_virtiomem_block_size($conf, $static_memory);
+
+	for (my $i = 0; $i < $sockets; $i++)  {
+
+	    my $id = "virtiomem$i";
+	    my $pciaddr = print_pci_addr($id, $bridges, $arch, $machine_type);
+	    my $mem_object = print_mem_object($conf, "mem-$id", $node_maxmem);
+
+	    push @$cmd, "-object" , $mem_object;
+	    push @$devices, "-device", "virtio-mem-pci,block-size=${blocksize}M,requested-size=${node_mem}M,id=$id,memdev=mem-$id,node=$i$pciaddr";
+	}
+    } elsif ($hotplug_features->{memory}) {
 	foreach_dimm($conf, $vmid, $memory, $sockets, sub {
 	    my ($conf, $vmid, $name, $dimm_size, $numanode, $current_size, $memory) = @_;
 
@@ -485,20 +570,14 @@ sub hugepages_topology {
 
     my $defaults = PVE::QemuServer::load_defaults();
     my $memory = $conf->{memory} || $defaults->{memory};
-    my $static_memory = 0;
+    my $static_memory = get_static_mem($conf, $defaults);
+
     my $sockets = 1;
     $sockets = $conf->{smp} if $conf->{smp}; # old style - no longer iused
     $sockets = $conf->{sockets} if $conf->{sockets};
     my $numa_custom_topology = undef;
     my $hotplug_features = PVE::QemuServer::parse_hotplug_features(defined($conf->{hotplug}) ? $conf->{hotplug} : '1');
 
-    if ($hotplug_features->{memory}) {
-	$static_memory = $STATICMEM;
-	$static_memory = $static_memory * $sockets if ($conf->{hugepages} && $conf->{hugepages} == 1024);
-    } else {
-	$static_memory = $memory;
-    }
-
     #custom numa topology
     for (my $i = 0; $i < $MAX_NUMA; $i++) {
 	next if !$conf->{"numa$i"};
diff --git a/PVE/QemuServer/PCI.pm b/PVE/QemuServer/PCI.pm
index a18b974..0187c74 100644
--- a/PVE/QemuServer/PCI.pm
+++ b/PVE/QemuServer/PCI.pm
@@ -249,6 +249,14 @@ sub get_pci_addr_map {
 	'scsihw2' => { bus => 4, addr => 1 },
 	'scsihw3' => { bus => 4, addr => 2 },
 	'scsihw4' => { bus => 4, addr => 3 },
+	'virtiomem0' => { bus => 4, addr => 4 },
+	'virtiomem1' => { bus => 4, addr => 5 },
+	'virtiomem2' => { bus => 4, addr => 6 },
+	'virtiomem3' => { bus => 4, addr => 7 },
+	'virtiomem4' => { bus => 4, addr => 8 },
+	'virtiomem5' => { bus => 4, addr => 9 },
+	'virtiomem6' => { bus => 4, addr => 10 },
+	'virtiomem7' => { bus => 4, addr => 11 },
     } if !defined($pci_addr_map);
     return $pci_addr_map;
 }
-- 
2.30.2