Re: [PATCH net-next v6 1/9] vsock: a per-net vsock NS mode state
From: Stefano Garzarella <sgarzare@redhat.com>
Date: 2025-09-26 16:10:12
Also in:
kvm, linux-hyperv, linux-kselftest, lkml, virtualization
On Tue, Sep 16, 2025 at 04:43:45PM -0700, Bobby Eshleman wrote:
quoted hunk ↗ jump to hunk
From: Bobby Eshleman <redacted> Add the per-net vsock NS mode state. This only adds the structure for holding the mode and some of the functions for setting/getting and checking the mode, but does not integrate the functionality yet. Signed-off-by: Bobby Eshleman <redacted> --- Changes in v6: - add orig_net_mode to store mode at creation time which will be used to avoid breakage when namespace changes mode during socket/VM lifespan Changes in v5: - use /proc/sys/net/vsock/ns_mode instead of /proc/net/vsock_ns_mode - change from net->vsock.ns_mode to net->vsock.mode - change vsock_net_set_mode() to vsock_net_write_mode() - vsock_net_write_mode() returns bool for write success to avoid need to use vsock_net_mode_can_set() - remove vsock_net_mode_can_set() --- MAINTAINERS | 1 + include/net/af_vsock.h | 55 +++++++++++++++++++++++++++++++++++++++++++++ include/net/net_namespace.h | 4 ++++ include/net/netns/vsock.h | 20 +++++++++++++++++ 4 files changed, 80 insertions(+)diff --git a/MAINTAINERS b/MAINTAINERS index 47bc35743f22..bc53c67e0926 100644 --- a/MAINTAINERS +++ b/MAINTAINERS@@ -26634,6 +26634,7 @@ L: netdev@vger.kernel.orgS: Maintained F: drivers/vhost/vsock.c F: include/linux/virtio_vsock.h +F: include/net/netns/vsock.h F: include/uapi/linux/virtio_vsock.h F: net/vmw_vsock/virtio_transport.c F: net/vmw_vsock/virtio_transport_common.cdiff --git a/include/net/af_vsock.h b/include/net/af_vsock.h index d40e978126e3..2857e97699de 100644 --- a/include/net/af_vsock.h +++ b/include/net/af_vsock.h@@ -10,6 +10,7 @@#include <linux/kernel.h> #include <linux/workqueue.h> +#include <net/netns/vsock.h> #include <net/sock.h> #include <uapi/linux/vm_sockets.h>@@ -65,6 +66,7 @@ struct vsock_sock {u32 peer_shutdown; bool sent_request; bool ignore_connecting_rst; + enum vsock_net_mode orig_net_mode;
Why `orig_` prefix? Maybe I need to review all the series, but it's a bit confusing for now. I guess it's related to the new behaviour to not change the mode of already created sockets (which I like), but IMHO this variable prefix is confusing.It seems we will have another field with the "actual_" mode, but if it's immutable, I'd avoid that prefix.
quoted hunk ↗ jump to hunk
/* Protected by lock_sock(sk) */ u64 buffer_size;@@ -256,4 +258,57 @@ static inline bool vsock_msgzerocopy_allow(const struct vsock_transport *t){ return t->msgzerocopy_allow && t->msgzerocopy_allow(); } + +static inline enum vsock_net_mode vsock_net_mode(struct net *net) +{ + enum vsock_net_mode ret; + + spin_lock_bh(&net->vsock.lock); + ret = net->vsock.mode; + spin_unlock_bh(&net->vsock.lock); + return ret; +} + +static inline bool vsock_net_write_mode(struct net *net, u8 mode) +{ + bool ret; + + spin_lock_bh(&net->vsock.lock); + + if (net->vsock.written) { + ret = false; + goto skip; + } + + net->vsock.mode = mode; + net->vsock.written = true; + ret = true; + +skip: + spin_unlock_bh(&net->vsock.lock); + return ret; +} + +/* Return true if vsock_sock passes the mode rules for a given net and + * orig_net_mode. Otherwise, return false. + * + * net is the current net namespace of the object being checked. orig_net_mode + * is the mode of net when the object was created.
`orig_net_mode` is also explained in the next paragraph, should we remove from here?
+ *
+ * orig_net_mode is the mode of arg 'net' at the time of creation for the
+ * object being checked. For example, if searching for a vsock_sock then
+ * orig_net_mode is arg net's mode at the time the vsock_sock was created.
+ *
+ * Read more about modes in the comment header of net/vmw_vsock/af_vsock.c.
+ */
+static inline bool vsock_net_check_mode(struct vsock_sock *vsk, struct net *net,
+ enum vsock_net_mode orig_net_mode)
+{
+ struct net *vsk_net = sock_net(sk_vsock(vsk));
+
+ if (net_eq(vsk_net, net))
+ return true;
+
+ return orig_net_mode == VSOCK_NET_MODE_GLOBAL && vsk->orig_net_mode == VSOCK_NET_MODE_GLOBAL;nit: I'd rewrite in this way, just because it seems easy to read to me, but again not strong opinion, this is fine: return orig_net_mode == VSOCK_NET_MODE_GLOBAL && orig_net_mode == vsk->orig_net_mode;
quoted hunk ↗ jump to hunk
+} #endif /* __AF_VSOCK_H__ */diff --git a/include/net/net_namespace.h b/include/net/net_namespace.h index 025a7574b275..005c0da4fb62 100644 --- a/include/net/net_namespace.h +++ b/include/net/net_namespace.h@@ -37,6 +37,7 @@#include <net/netns/smc.h> #include <net/netns/bpf.h> #include <net/netns/mctp.h> +#include <net/netns/vsock.h> #include <net/net_trackers.h> #include <linux/ns_common.h> #include <linux/idr.h>@@ -196,6 +197,9 @@ struct net {/* Move to a better place when the config guard is removed. */ struct mutex rtnl_mutex; #endif +#if IS_ENABLED(CONFIG_VSOCKETS) + struct netns_vsock vsock; +#endif } __randomize_layout; #include <linux/seq_file_net.h>diff --git a/include/net/netns/vsock.h b/include/net/netns/vsock.h new file mode 100644 index 000000000000..d4593c0b8dc4 --- /dev/null +++ b/include/net/netns/vsock.h@@ -0,0 +1,20 @@ +/* SPDX-License-Identifier: GPL-2.0 */ +#ifndef __NET_NET_NAMESPACE_VSOCK_H +#define __NET_NET_NAMESPACE_VSOCK_H + +#include <linux/types.h> + +enum vsock_net_mode { + VSOCK_NET_MODE_GLOBAL, + VSOCK_NET_MODE_LOCAL, +}; + +struct netns_vsock { + struct ctl_table_header *vsock_hdr;
sysctl_header, or sysctl_hdr, or sysctl_ctl ? I'd remove `vsock_` prefix and make more clear its used for sysctl.
+ spinlock_t lock; + + /* protected by lock */ + enum vsock_net_mode mode; + bool written;
I would call this `mode_set` or `mode_locked`, I mean with `mode_` prefix to make sure we don't need to rename it when we will add new fields in the future. Thanks Stefano
+}; +#endif /* __NET_NET_NAMESPACE_VSOCK_H */ -- 2.47.3