From mboxrd@z Thu Jan 1 00:00:00 1970
Return-Path: <f.ebner@proxmox.com>
Received: from firstgate.proxmox.com (firstgate.proxmox.com [212.224.123.68])
(using TLSv1.3 with cipher TLS_AES_256_GCM_SHA384 (256/256 bits)
key-exchange X25519 server-signature RSA-PSS (2048 bits))
(No client certificate requested)
by lists.proxmox.com (Postfix) with ESMTPS id 72F7E90737
for <pve-devel@lists.proxmox.com>; Fri, 3 Feb 2023 14:46:39 +0100 (CET)
Received: from firstgate.proxmox.com (localhost [127.0.0.1])
by firstgate.proxmox.com (Proxmox) with ESMTP id 5CDC654A2
for <pve-devel@lists.proxmox.com>; Fri, 3 Feb 2023 14:46:39 +0100 (CET)
Received: from proxmox-new.maurer-it.com (proxmox-new.maurer-it.com
[94.136.29.106])
(using TLSv1.3 with cipher TLS_AES_256_GCM_SHA384 (256/256 bits)
key-exchange X25519 server-signature RSA-PSS (2048 bits))
(No client certificate requested)
by firstgate.proxmox.com (Proxmox) with ESMTPS
for <pve-devel@lists.proxmox.com>; Fri, 3 Feb 2023 14:46:38 +0100 (CET)
Received: from proxmox-new.maurer-it.com (localhost.localdomain [127.0.0.1])
by proxmox-new.maurer-it.com (Proxmox) with ESMTP id 0BCA4452D4;
Fri, 3 Feb 2023 14:46:38 +0100 (CET)
Message-ID: <f9758e76-2a9c-cef9-412e-1b85a5d70c8b@proxmox.com>
Date: Fri, 3 Feb 2023 14:46:36 +0100
MIME-Version: 1.0
User-Agent: Mozilla/5.0 (X11; Linux x86_64; rv:102.0) Gecko/20100101
Thunderbird/102.6.0
From: Fiona Ebner <f.ebner@proxmox.com>
To: pve-devel@lists.proxmox.com, "aderumier@odiso.com" <aderumier@odiso.com>
References: <20230202110344.840195-1-aderumier@odiso.com>
<20230202110344.840195-12-aderumier@odiso.com>
Content-Language: en-US
In-Reply-To: <20230202110344.840195-12-aderumier@odiso.com>
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: 7bit
X-SPAM-LEVEL: Spam detection results: 0
AWL 0.042 Adjusted score from AWL reputation of From: address
BAYES_00 -1.9 Bayes spam probability is 0 to 1%
KAM_DMARC_STATUS 0.01 Test Rule for DKIM or SPF Failure with Strict Alignment
NICE_REPLY_A -0.09 Looks like a legit reply (A)
SPF_HELO_NONE 0.001 SPF: HELO does not publish an SPF Record
SPF_PASS -0.001 SPF: sender matches SPF record
Subject: Re: [pve-devel] [PATCH v3 qemu-server 11/13] memory: add virtio-mem
support
X-BeenThere: pve-devel@lists.proxmox.com
X-Mailman-Version: 2.1.29
Precedence: list
List-Id: Proxmox VE development discussion <pve-devel.lists.proxmox.com>
List-Unsubscribe: <https://lists.proxmox.com/cgi-bin/mailman/options/pve-devel>,
<mailto:pve-devel-request@lists.proxmox.com?subject=unsubscribe>
List-Archive: <http://lists.proxmox.com/pipermail/pve-devel/>
List-Post: <mailto:pve-devel@lists.proxmox.com>
List-Help: <mailto:pve-devel-request@lists.proxmox.com?subject=help>
List-Subscribe: <https://lists.proxmox.com/cgi-bin/mailman/listinfo/pve-devel>,
<mailto:pve-devel-request@lists.proxmox.com?subject=subscribe>
X-List-Received-Date: Fri, 03 Feb 2023 13:46:39 -0000
Am 02.02.23 um 12:03 schrieb Alexandre Derumier:
> @@ -157,6 +168,109 @@ sub get_current_memory {
> return $memory->{current};
> }
>
> +sub get_virtiomem_block_size {
> + my ($conf) = @_;
> +
> + my $sockets = $conf->{sockets} || 1;
> + my $MAX_MEM = get_max_mem($conf);
> + my $static_memory = get_static_mem($conf, $sockets);
> + my $memory = get_current_memory($conf->{memory});
> +
> + #virtiomem can map 32000 block size.
> + #try to use lowest blocksize, lower = more chance to unplug memory.
> + my $blocksize = ($MAX_MEM - $static_memory) / 32000;
> + #2MB is the minimum to be aligned with THP
> + $blocksize = 2 if $blocksize < 2;
> + $blocksize = 2**(ceil(log($blocksize)/log(2)));
> + #Linux guest kernel only support 4MiB block currently (kernel <= 6.2)
> + $blocksize = 4 if $blocksize < 4;
> +
> + return $blocksize;
> +}
> +
> +my sub get_virtiomem_total_current_size {
> + my ($mems) = @_;
> + my $size = 0;
> + for my $mem (values %$mems) {
> + $size += $mem->{current};
> + }
> + return $size;
> +}
> +
> +my sub balance_virtiomem {
The function should die if not all memory can be plugged.
> + my ($vmid, $virtiomems, $blocksize, $target_virtiomem_total) = @_;
> +
> + my $nb_virtiomem = keys %$virtiomems;
Style nit: explicit scalar() would be nice.
> +
> + print"try to balance memory on $nb_virtiomem virtiomems\n";
> + die "error. no more available blocks in virtiomem to balance the remaining memory" if $target_virtiomem_total < 0;
> +
> + #if we can't share exactly the same amount, we add the remainder on last node
> + my $virtiomem_target_aligned = int( $target_virtiomem_total / $nb_virtiomem / $blocksize) * $blocksize;
> + my $virtiomem_target_remaining = $target_virtiomem_total - ($virtiomem_target_aligned * ($nb_virtiomem-1));
> +
> + my $i = 0;
> + foreach my $id (sort keys %$virtiomems) {
> + my $virtiomem = $virtiomems->{$id};
> + $i++;
> + my $virtiomem_target = $i == $nb_virtiomem ? $virtiomem_target_remaining : $virtiomem_target_aligned;
> + $virtiomem->{completed} = 0;
> + $virtiomem->{retry} = 0;
> + $virtiomem->{target} = $virtiomem_target;
> +
> + print "virtiomem$id: set-requested-size : $virtiomem_target\n";
> + mon_cmd($vmid, 'qom-set', path => "/machine/peripheral/virtiomem$id", property => "requested-size", value => $virtiomem_target * 1024 * 1024);
Style nit: some lines are over 100 chars and some quite a bit
> + }
> +
> + while (1) {
> +
> + sleep 1;
> + my $total_finished = 0;
> +
> + foreach my $id (keys %$virtiomems) {
> +
> + my $virtiomem = $virtiomems->{$id};
> +
> + if ($virtiomem->{error} || $virtiomem->{completed}) {
> + $total_finished++;
> + next;
> + }
> +
> + my $size = mon_cmd($vmid, 'qom-get', path => "/machine/peripheral/virtiomem$id", property => "size");
> + $virtiomem->{current} = $size / 1024 / 1024;
> + print"virtiomem$id: virtiomem->last: $virtiomem->{last} virtiomem->current: $virtiomem->{current} virtio_mem_target:$virtiomem->{target}\n";
> +
> + if($virtiomem->{current} == $virtiomem->{target}) {
> + print"virtiomem$id: completed\n";
> + $virtiomem->{completed} = 1;
> + next;
> + }
> +
> + if($virtiomem->{current} != $virtiomem->{last}) {
> + #if value has changed, but not yet completed
> + print "virtiomem$id: changed but don't not reach target yet\n";
> + $virtiomem->{retry} = 0;
> + $virtiomem->{last} = $virtiomem->{current};
> + next;
> + }
> +
> + if($virtiomem->{retry} >= 5) {
> + print "virtiomem$id: too many retry. set error\n";
> + $virtiomem->{error} = 1;
> + #as change is async, we don't want that value change after the api call
> + eval {
> + mon_cmd($vmid, 'qom-set', path => "/machine/peripheral/virtiomem$id", property => "requested-size", value => $virtiomem->{current} * 1024 *1024);
> + };
> + }
> + print"virtiomem$id: increase retry: $virtiomem->{retry}\n";
> + $virtiomem->{retry}++;
> + }
> +
> + my $nb_virtiomem = keys %$virtiomems;
Redeclares variable. The number can't change from before, or am I
missing something?
> + return if $total_finished == $nb_virtiomem;
Style nit: could also use
while ($total_finished != $nb_virtiomem)
at the beginning of the loop.
> + }
> +}
> +
> sub get_numa_node_list {
> my ($conf) = @_;
> my @numa_map;
> @@ -237,7 +351,39 @@ sub qemu_memory_hotplug {
> my $MAX_MEM = get_max_mem($conf);
> die "you cannot add more memory than max mem $MAX_MEM MB!\n" if $value > $MAX_MEM;
>
> - if ($value > $memory) {
> + my $confmem = parse_memory($conf->{memory});
You can just re-use the existing $oldmem?
> +
> + if ($confmem->{virtio}) {
> + my $blocksize = get_virtiomem_block_size($conf);
> +
> + my $virtiomems = {};
> +
> + for (my $i = 0; $i < $sockets; $i++) {
> + my $size = mon_cmd($vmid, 'qom-get', path => "/machine/peripheral/virtiomem$i", property => "size");
> + $size = $size / 1024 /1024;
> + $virtiomems->{$i} = {
> + current => $size,
> + last => $size,
> + error => 0,
> + completed => 0,
> + retry => 0
> + };
> + }
> +
> + my $target_virtiomem_total = $value - $static_memory;
> + my $err;
> + eval {
> + balance_virtiomem($vmid, $virtiomems, $blocksize, $target_virtiomem_total);
> + };
> + $err = $@ if $@;
> +
> + my $current_memory = $static_memory + get_virtiomem_total_current_size($virtiomems);
> + $newmem->{current} = $current_memory;
> + $conf->{memory} = print_memory($newmem);
> + PVE::QemuConfig->write_config($vmid, $conf);
> + die $err if $err;
> +
> + } elsif ($value > $memory) {
>
> my $numa_hostmap;
>
> @@ -441,17 +590,42 @@ sub config {
> }
>
> if ($hotplug) {
> - foreach_dimm($conf, $vmid, $memory, $static_memory, sub {
> - my ($conf, $vmid, $name, $dimm_size, $numanode, $current_size, $memory) = @_;
>
> - my $mem_object = print_mem_object($conf, "mem-$name", $dimm_size);
> + my $confmem = parse_memory($conf->{memory});
> +
> + if ($confmem->{'virtio'}) {
>
> - push @$cmd, "-object" , $mem_object;
> - push @$cmd, "-device", "pc-dimm,id=$name,memdev=mem-$name,node=$numanode";
> + my $MAX_MEM = get_max_mem($conf);
> + my $node_maxmem = ($MAX_MEM - $static_memory) / $sockets;
> + my $node_mem = ($memory - $static_memory) / $sockets;
Nit: If the number of $sockets is not a power of 2, I think this breaks.
But I guess we already don't support it. Running current version without
your patches (for a VM with memory hotplug):
root@pve701 ~ # qm set 131 --sockets 3
update VM 131: -sockets 3
root@pve701 ~ # qm set 131 -memory 8192
update VM 131: -memory 8192
root@pve701 ~ # qm start 131
kvm: total memory for NUMA nodes (0x3fffffff) should equal RAM size
(0x40000000)
start failed: QEMU exited with code 1
I guess we can just fix it up together with the existing rounding issue
if/when somebody complains about it ;)
> + my $blocksize = get_virtiomem_block_size($conf);
>
> - die "memory size ($memory) must be aligned to $dimm_size for hotplugging\n"
> - if $current_size > $memory;
> - });
> + die "memory need to be a multiple of $blocksize MiB with maxmemory $MAX_MEM MiB when virtiomem is enabled\n"
> + if $memory % $blocksize != 0;
> +
> + for (my $i = 0; $i < $sockets; $i++) {
> +
> + my $id = "virtiomem$i";
> + my $mem_object = print_mem_object($conf, "mem-$id", $node_maxmem);
> + push @$cmd, "-object" , "$mem_object,reserve=off";
> +
> + my $mem_device = "virtio-mem-pci,block-size=${blocksize}M,requested-size=${node_mem}M,id=$id,memdev=mem-$id,node=$i";
> + $mem_device .= ",prealloc=on" if $conf->{hugepages};
> + $mem_devices->{$id} = $mem_device;
> + }
> + } else {
> + foreach_dimm($conf, $vmid, $memory, $static_memory, sub {
> + my ($conf, $vmid, $name, $dimm_size, $numanode, $current_size, $memory) = @_;
> +
> + my $mem_object = print_mem_object($conf, "mem-$name", $dimm_size);
> +
> + push @$cmd, "-object" , $mem_object;
> + push @$cmd, "-device", "pc-dimm,id=$name,memdev=mem-$name,node=$numanode";
> +
> + die "memory size ($memory) must be aligned to $dimm_size for hotplugging\n"
> + if $current_size > $memory;
> + });
> + }
> }
> }
>