all lists on lists.proxmox.com
 help / color / mirror / Atom feed
From: David Riley <d.riley@proxmox.com>
To: Stefan Hanreich <s.hanreich@proxmox.com>, pve-devel@lists.proxmox.com
Subject: Re: [PATCH pve-access-control v5 5/7] fix #7520: sdn: add VNet ACL migration
Date: Thu, 16 Jul 2026 09:25:31 +0200	[thread overview]
Message-ID: <89967b7d-5756-46cc-a833-2c5c814e416a@proxmox.com> (raw)
In-Reply-To: <029ecefd-ae81-49ac-89c4-783dae029941@proxmox.com>

Thanks for taking a look.

Comments inline.

On 7/15/26 5:00 PM, Stefan Hanreich wrote:
> comments inline
>
> On 7/7/26 2:05 PM, David Riley wrote:
>> Add routine to enable the relocation of ACL entries when their
>> corresponding SDN resources are moved. This prevents stale entries
>> when VNets are moved between zones.
>>
>> This is a preparatory change that does not have an immediate effect on
>> its own, but lays the foundation for safely relocating ACL nodes and
>> rejecting conflicting destination paths.
>>
>> Link: https://bugzilla.proxmox.com/show_bug.cgi?id=7520
>> Signed-off-by: David Riley <d.riley@proxmox.com>
>> ---
>>   src/PVE/AccessControl.pm | 84 ++++++++++++++++++++++++++++++++++++++++
>>   1 file changed, 84 insertions(+)
>>
>> diff --git a/src/PVE/AccessControl.pm b/src/PVE/AccessControl.pm
>> index 1047226..4faeb8f 100644
>> --- a/src/PVE/AccessControl.pm
>> +++ b/src/PVE/AccessControl.pm
>> @@ -2045,6 +2045,90 @@ sub remove_sdn_resource_access {
>>       lock_user_config($delete_resource_fn, "SDN ACL cleanup failed");
>>   }
>>   
>> +sub migrate_sdn_resource_access {
>> +    my ($migrations) = @_;
>> +    # [ { src_path => 'zones/<zoney>/<vnet>', dest_path => 'zones/<zonex>/<vnet>' } ]
> this claims the value is a string
>
>> +
>> +    return if !$migrations;
>> +
>> +    my $migrate_fn = sub {
>> +        my $usercfg = cfs_read_file("user.cfg");
>> +        my $modified = 0;
>> +
>> +        my @prepared_moves = ();
>> +        for my $migration (@$migrations) {
>> +
>> +            my $invalid_key = 0;
>> +
>> +            for my $key (qw(src_path dest_path)) {
>> +                if (!defined($migration->{$key})) {
>> +                    warn "Invalid '$key': must be an array reference.\n";
> but this says it must be an array reference? also, if I misunderstand
> and this should indeed be an array reference then we should explicitly
> check for that?

Nice catch, this was indeed a leftover from an earlier iteration where I used arrays.
I switched to strings in this version, so I'll update the warning to correctly check
for and report missing path keys (e.g.: 'Missing src_path/dest_path') in v6.

>
>> +                    $invalid_key = 1;
>> +                    last;
>> +                }
>> +            }
>> +
>> +            next if $invalid_key;
>> +
>> +            my $src_path = "sdn/$migration->{src_path}";
>> +            my $dest_path = "sdn/$migration->{dest_path}";
>> +
>> +            my ($src_parent, $src_last_key, $acl_node) =
>> +                find_acl_tree_node_parent($usercfg->{acl_root}, $src_path);
>> +
>> +            if (defined($src_parent) && defined($src_last_key) && defined($acl_node)) {
>> +                # probe dest for conflicts
>> +                my (undef, undef, $existing_dest_node) =
>> +                    find_acl_tree_node_parent($usercfg->{acl_root}, $dest_path);
>> +
>> +                if (acl_tree_has_permissions($existing_dest_node)) {
>> +                    my $conflict_path = "/$dest_path";
>> +                    my $source_path = "/$src_path";
>> +                    die
>> +                        "Destination '$conflict_path' already has permissions configured (Source:"
>> +                        . " '$source_path'). Please remove the target ACLs manually before"
>> +                        . " retrying.\n";
>> +                }
>> +
>> +                # stage move
>> +                push(
>> +                    @prepared_moves,
>> +                    {
>> +                        src_parent => $src_parent,
>> +                        src_last_key => $src_last_key,
>> +                        dest_path => $dest_path,
>> +                        node => $acl_node,
>> +                    },
>> +                );
>> +            }
>> +        }
>> +
>> +        for my $move (@prepared_moves) {
>> +            delete $move->{src_parent}->{children}->{ $move->{src_last_key} };
>> +
>> +            my $current = $usercfg->{acl_root};
>> +            my @dest_segments = split('/', $move->{dest_path});
>> +            my $dest_last_key = pop @dest_segments;
>> +
>> +            for my $segment (@dest_segments) {
>> +                $current->{children}->{$segment} = {}
>> +                    if !defined($current->{children}->{$segment});
>> +
>> +                $current = $current->{children}->{$segment};
>> +            }
>> +            $current->{children}->{$dest_last_key} = $move->{node};
>> +
>> +            $modified = 1;
> nit: wouldn't an early return if prepared moves is empty save us from
> tracking this here?


Good point, that simplifies the logic nicely. I'll add an early return if '@prepared_moves'
is empty and drop the '$modified' flag entirely for v6.

>
>> +        }
>> +
>> +        if ($modified) {
>> +            cfs_write_file("user.cfg", $usercfg);
>> +        }
>> +    };
>> +
>> +    lock_user_config($migrate_fn, "ACL migration failed");
>> +}
>> +
>>   my $USER_CONTROLLED_TFA_TYPES = {
>>       u2f => 1,
>>       oath => 1,
>
>
>
>




  reply	other threads:[~2026-07-16  7:25 UTC|newest]

Thread overview: 18+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2026-07-07 12:03 [PATCH access-control/network v5 0/7] fix #7520: sdn: prune orphaned ACLs and handle VNet migrations David Riley
2026-07-07 12:03 ` [PATCH pve-access-control v5 1/7] permission: add acl tree node parent lookup without node creation David Riley
2026-07-15 14:59   ` Stefan Hanreich
2026-07-16  7:31     ` David Riley
2026-07-07 12:03 ` [PATCH pve-access-control v5 2/7] fix #7520: sdn: add pruning SDN resources ACLs David Riley
2026-07-07 12:04 ` [PATCH pve-access-control v5 3/7] test: add sdn acl pruning tests David Riley
2026-07-16  7:32   ` Daniel Kral
2026-07-16  8:05     ` David Riley
2026-07-07 12:04 ` [PATCH pve-access-control v5 4/7] permission: add helper to check for active permissions David Riley
2026-07-07 12:04 ` [PATCH pve-access-control v5 5/7] fix #7520: sdn: add VNet ACL migration David Riley
2026-07-15 14:59   ` Stefan Hanreich
2026-07-16  7:25     ` David Riley [this message]
2026-07-07 12:04 ` [PATCH pve-access-control v5 6/7] test: add sdn acl migration tests David Riley
2026-07-07 12:04 ` [PATCH pve-network v5 7/7] fix #7520: config: prune orphaned and relocate moved VNet ACLs David Riley
2026-07-15 15:00   ` Stefan Hanreich
2026-07-16  7:11     ` David Riley
2026-07-15 15:11 ` [PATCH access-control/network v5 0/7] fix #7520: sdn: prune orphaned ACLs and handle VNet migrations Stefan Hanreich
2026-07-16 10:47 ` superseded: " David Riley

Reply instructions:

You may reply publicly to this message via plain-text email
using any one of the following methods:

* Save the following mbox file, import it into your mail client,
  and reply-to-all from there: mbox

  Avoid top-posting and favor interleaved quoting:
  https://en.wikipedia.org/wiki/Posting_style#Interleaved_style

* Reply using the --to, --cc, and --in-reply-to
  switches of git-send-email(1):

  git send-email \
    --in-reply-to=89967b7d-5756-46cc-a833-2c5c814e416a@proxmox.com \
    --to=d.riley@proxmox.com \
    --cc=pve-devel@lists.proxmox.com \
    --cc=s.hanreich@proxmox.com \
    /path/to/YOUR_REPLY

  https://kernel.org/pub/software/scm/git/docs/git-send-email.html

* If your mail client supports setting the In-Reply-To header
  via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line before the message body.
This is an external index of several public inboxes,
see mirroring instructions on how to clone and mirror
all data and code used by this external index.
Service provided by Proxmox Server Solutions GmbH | Privacy | Legal