* [PATCH v4 1/3] net/af-xdp: Remove XDP program cleanup logic
@ 2025-07-04 13:05 Daniel Borkmann via
2025-07-04 13:05 ` [PATCH v4 2/3] net/af-xdp: Fix up cleanup path upon failure in queue creation Daniel Borkmann via
` (2 more replies)
0 siblings, 3 replies; 7+ messages in thread
From: Daniel Borkmann via @ 2025-07-04 13:05 UTC (permalink / raw)
To: qemu-devel; +Cc: daniel, Ilya Maximets, Jason Wang, Anton Protopopov
There are two issues with the XDP program removal in af_xdp_cleanup():
1) Starting from libxdp 1.3.0 [0] the XDP program gets automatically
detached when we call xsk_socket__delete() for the last successfully
configured queue. libxdp internally keeps track of that. For QEMU
we require libxdp >= 1.4.0. Given QEMU is not loading the program,
lets also not attempt to remove it and delegate this instead.
2) The removal logic is incorrect anyway because we are setting n_queues
into the last queue that never has xdp_flags on failure, so the logic
is always skipped since the non-zero test for s->xdp_flags in
af_xdp_cleanup() fails.
Fixes: cb039ef3d9e3 ("net: add initial support for AF_XDP network backend")
Suggested-by: Ilya Maximets <i.maximets@ovn.org>
Signed-off-by: Daniel Borkmann <daniel@iogearbox.net>
Cc: Ilya Maximets <i.maximets@ovn.org>
Cc: Jason Wang <jasowang@redhat.com>
Cc: Anton Protopopov <aspsk@isovalent.com>
Link: https://github.com/xdp-project/xdp-tools/commit/38c2914988fd5c1ef65f2381fc8af9f3e8404e2b [0]
---
net/af-xdp.c | 12 ------------
1 file changed, 12 deletions(-)
diff --git a/net/af-xdp.c b/net/af-xdp.c
index 01c5fb914e..c5d3b6a953 100644
--- a/net/af-xdp.c
+++ b/net/af-xdp.c
@@ -49,7 +49,6 @@ typedef struct AFXDPState {
char *buffer;
struct xsk_umem *umem;
- uint32_t n_queues;
uint32_t xdp_flags;
bool inhibit;
} AFXDPState;
@@ -274,14 +273,6 @@ static void af_xdp_cleanup(NetClientState *nc)
s->umem = NULL;
qemu_vfree(s->buffer);
s->buffer = NULL;
-
- /* Remove the program if it's the last open queue. */
- if (!s->inhibit && nc->queue_index == s->n_queues - 1 && s->xdp_flags
- && bpf_xdp_detach(s->ifindex, s->xdp_flags, NULL) != 0) {
- fprintf(stderr,
- "af-xdp: unable to remove XDP program from '%s', ifindex: %d\n",
- s->ifname, s->ifindex);
- }
}
static int af_xdp_umem_create(AFXDPState *s, int sock_fd, Error **errp)
@@ -490,12 +481,9 @@ int net_init_af_xdp(const Netdev *netdev,
pstrcpy(s->ifname, sizeof(s->ifname), opts->ifname);
s->ifindex = ifindex;
- s->n_queues = queues;
if (af_xdp_umem_create(s, sock_fds ? sock_fds[i] : -1, errp)
|| af_xdp_socket_create(s, opts, errp)) {
- /* Make sure the XDP program will be removed. */
- s->n_queues = i;
error_propagate(errp, err);
goto err;
}
--
2.43.0
^ permalink raw reply related [flat|nested] 7+ messages in thread
* [PATCH v4 2/3] net/af-xdp: Fix up cleanup path upon failure in queue creation
2025-07-04 13:05 [PATCH v4 1/3] net/af-xdp: Remove XDP program cleanup logic Daniel Borkmann via
@ 2025-07-04 13:05 ` Daniel Borkmann via
2025-07-09 11:26 ` Ilya Maximets
2025-07-04 13:05 ` [PATCH v4 3/3] net/af-xdp: Support pinned map path for AF_XDP sockets Daniel Borkmann via
2025-07-09 11:24 ` [PATCH v4 1/3] net/af-xdp: Remove XDP program cleanup logic Ilya Maximets
2 siblings, 1 reply; 7+ messages in thread
From: Daniel Borkmann via @ 2025-07-04 13:05 UTC (permalink / raw)
To: qemu-devel; +Cc: daniel, Ilya Maximets, Jason Wang, Anton Protopopov
While testing, it turned out that upon error in the queue creation loop,
we never trigger the af_xdp_cleanup() handler. This is because we pass
errp instead of a local err pointer into the various AF_XDP setup functions
instead of a scheme like:
bool fn(..., Error **errp)
{
Error *err = NULL;
foo(arg, &err);
if (err) {
handle the error...
error_propagate(errp, err);
return false;
}
...
}
With a conversion into the above format, the af_xdp_cleanup() handler is
called as expected. Note the error_propagate() handles a NULL err internally.
Fixes: cb039ef3d9e3 ("net: add initial support for AF_XDP network backend")
Signed-off-by: Daniel Borkmann <daniel@iogearbox.net>
Cc: Ilya Maximets <i.maximets@ovn.org>
Cc: Jason Wang <jasowang@redhat.com>
Cc: Anton Protopopov <aspsk@isovalent.com>
---
net/af-xdp.c | 6 +++---
1 file changed, 3 insertions(+), 3 deletions(-)
diff --git a/net/af-xdp.c b/net/af-xdp.c
index c5d3b6a953..29c5ad16cd 100644
--- a/net/af-xdp.c
+++ b/net/af-xdp.c
@@ -482,9 +482,8 @@ int net_init_af_xdp(const Netdev *netdev,
pstrcpy(s->ifname, sizeof(s->ifname), opts->ifname);
s->ifindex = ifindex;
- if (af_xdp_umem_create(s, sock_fds ? sock_fds[i] : -1, errp)
- || af_xdp_socket_create(s, opts, errp)) {
- error_propagate(errp, err);
+ if (af_xdp_umem_create(s, sock_fds ? sock_fds[i] : -1, &err) ||
+ af_xdp_socket_create(s, opts, &err)) {
goto err;
}
}
@@ -506,6 +505,7 @@ int net_init_af_xdp(const Netdev *netdev,
err:
if (nc0) {
qemu_del_net_client(nc0);
+ error_propagate(errp, err);
}
return -1;
--
2.43.0
^ permalink raw reply related [flat|nested] 7+ messages in thread
* [PATCH v4 3/3] net/af-xdp: Support pinned map path for AF_XDP sockets
2025-07-04 13:05 [PATCH v4 1/3] net/af-xdp: Remove XDP program cleanup logic Daniel Borkmann via
2025-07-04 13:05 ` [PATCH v4 2/3] net/af-xdp: Fix up cleanup path upon failure in queue creation Daniel Borkmann via
@ 2025-07-04 13:05 ` Daniel Borkmann via
2025-07-09 12:29 ` Ilya Maximets
2025-07-09 11:24 ` [PATCH v4 1/3] net/af-xdp: Remove XDP program cleanup logic Ilya Maximets
2 siblings, 1 reply; 7+ messages in thread
From: Daniel Borkmann via @ 2025-07-04 13:05 UTC (permalink / raw)
To: qemu-devel; +Cc: daniel, Ilya Maximets, Jason Wang, Anton Protopopov
Extend 'inhibit=on' setting with the option to specify a pinned XSK map
path along with a starting index (default 0) to push the created XSK
sockets into. Example usage:
# ./build/qemu-system-x86_64 [...] \
-netdev af-xdp,ifname=enp2s0f0np0,id=net0,mode=native,queues=2,start-queue=14,inhibit=on,map-path=/sys/fs/bpf/xsks_map,map-start-index=14 \
-device virtio-net-pci,netdev=net0 [...]
This is useful for the case where an existing XDP program with XSK map
is present on the AF_XDP supported phys device and the XSK map is not
yet populated. For example, the former could have been pre-loaded onto
the netdevice by a control plane, which later launches QEMU to populate
it with XSK sockets.
Normally, the main idea behind 'inhibit=on' is that the QEMU instance
doesn't need to have a lot of privileges to use the pre-loaded program
and the pre-created sockets, but this mentioned use-case here is different
where QEMU still needs privileges to create the sockets.
The 'map-start-index' parameter is optional and defaults to 0. It allows
flexible placement of the XSK sockets, and is up to the user to specify
when the XDP program with XSK map was already preloaded. In the simplest
case the queue-to-map-slot mapping is just 1:1 based on ctx->rx_queue_index
but the user might as well have a different scheme (or smaller map size,
e.g. ctx->rx_queue_index % max_size) to push the inbound traffic to one
of the XSK sockets.
Note that the bpf_xdp_query_id() is now only tested for 'inhibit=off'
since only in the latter case the libxdp takes care of installing the
XDP program which was installed based on the s->xdp_flags pointing to
either driver or skb mode. For 'inhibit=on' we don't make any assumptions
and neither go down the path of probing all possible options in which
way the user installed the XDP program.
Signed-off-by: Daniel Borkmann <daniel@iogearbox.net>
Cc: Ilya Maximets <i.maximets@ovn.org>
Cc: Jason Wang <jasowang@redhat.com>
Cc: Anton Protopopov <aspsk@isovalent.com>
---
net/af-xdp.c | 78 +++++++++++++++++++++++++++++++++++++++++++++----
qapi/net.json | 29 +++++++++++-------
qemu-options.hx | 23 +++++++++++++--
3 files changed, 113 insertions(+), 17 deletions(-)
diff --git a/net/af-xdp.c b/net/af-xdp.c
index 29c5ad16cd..005117c336 100644
--- a/net/af-xdp.c
+++ b/net/af-xdp.c
@@ -51,6 +51,10 @@ typedef struct AFXDPState {
uint32_t xdp_flags;
bool inhibit;
+
+ char *map_path;
+ int map_fd;
+ uint32_t map_start_index;
} AFXDPState;
#define AF_XDP_BATCH_SIZE 64
@@ -260,6 +264,7 @@ static void af_xdp_send(void *opaque)
static void af_xdp_cleanup(NetClientState *nc)
{
AFXDPState *s = DO_UPCAST(AFXDPState, nc, nc);
+ int idx;
qemu_purge_queued_packets(nc);
@@ -273,6 +278,18 @@ static void af_xdp_cleanup(NetClientState *nc)
s->umem = NULL;
qemu_vfree(s->buffer);
s->buffer = NULL;
+
+ if (s->map_fd >= 0) {
+ idx = nc->queue_index + s->map_start_index;
+ if (bpf_map_delete_elem(s->map_fd, &idx)) {
+ fprintf(stderr, "af-xdp: unable to remove AF_XDP socket from "
+ "map %s\n", s->map_path);
+ }
+ close(s->map_fd);
+ s->map_fd = -1;
+ }
+ g_free(s->map_path);
+ s->map_path = NULL;
}
static int af_xdp_umem_create(AFXDPState *s, int sock_fd, Error **errp)
@@ -336,7 +353,6 @@ static int af_xdp_socket_create(AFXDPState *s,
};
int queue_id, error = 0;
- s->inhibit = opts->has_inhibit && opts->inhibit;
if (s->inhibit) {
cfg.libxdp_flags |= XSK_LIBXDP_FLAGS__INHIBIT_PROG_LOAD;
}
@@ -387,6 +403,35 @@ static int af_xdp_socket_create(AFXDPState *s,
return 0;
}
+static int af_xdp_update_xsk_map(AFXDPState *s, Error **errp)
+{
+ int xsk_fd, idx, error = 0;
+
+ if (!s->map_path) {
+ return 0;
+ }
+
+ s->map_fd = bpf_obj_get(s->map_path);
+ if (s->map_fd < 0) {
+ error = errno;
+ } else {
+ xsk_fd = xsk_socket__fd(s->xsk);
+ idx = s->nc.queue_index + s->map_start_index;
+ if (bpf_map_update_elem(s->map_fd, &idx, &xsk_fd, 0)) {
+ error = errno;
+ }
+ }
+
+ if (error) {
+ error_setg_errno(errp, error,
+ "failed to insert AF_XDP socket into map %s",
+ s->map_path);
+ return -1;
+ }
+
+ return 0;
+}
+
/* NetClientInfo methods. */
static NetClientInfo net_af_xdp_info = {
.type = NET_CLIENT_DRIVER_AF_XDP,
@@ -441,6 +486,7 @@ int net_init_af_xdp(const Netdev *netdev,
int64_t i, queues;
Error *err = NULL;
AFXDPState *s;
+ bool inhibit;
ifindex = if_nametoindex(opts->ifname);
if (!ifindex) {
@@ -456,8 +502,21 @@ int net_init_af_xdp(const Netdev *netdev,
return -1;
}
- if ((opts->has_inhibit && opts->inhibit) != !!opts->sock_fds) {
- error_setg(errp, "'inhibit=on' requires 'sock-fds' and vice versa");
+ inhibit = opts->has_inhibit && opts->inhibit;
+ if (inhibit && !opts->sock_fds && !opts->map_path) {
+ error_setg(errp, "'inhibit=on' requires 'sock-fds' or 'map-path'");
+ return -1;
+ }
+ if (!inhibit && (opts->sock_fds || opts->map_path)) {
+ error_setg(errp, "'sock-fds' and 'map-path' require 'inhibit=on'");
+ return -1;
+ }
+ if (opts->sock_fds && opts->map_path) {
+ error_setg(errp, "'sock-fds' and 'map-path' are mutually exclusive");
+ return -1;
+ }
+ if (!opts->map_path && opts->has_map_start_index) {
+ error_setg(errp, "'map-start-index' requires 'map-path'");
return -1;
}
@@ -481,14 +540,23 @@ int net_init_af_xdp(const Netdev *netdev,
pstrcpy(s->ifname, sizeof(s->ifname), opts->ifname);
s->ifindex = ifindex;
+ s->inhibit = inhibit;
+
+ s->map_path = g_strdup(opts->map_path);
+ s->map_fd = -1;
+ s->map_start_index = 0;
+ if (opts->has_map_start_index && opts->map_start_index > 0) {
+ s->map_start_index = opts->map_start_index;
+ }
if (af_xdp_umem_create(s, sock_fds ? sock_fds[i] : -1, &err) ||
- af_xdp_socket_create(s, opts, &err)) {
+ af_xdp_socket_create(s, opts, &err) ||
+ af_xdp_update_xsk_map(s, &err)) {
goto err;
}
}
- if (nc0) {
+ if (nc0 && !inhibit) {
s = DO_UPCAST(AFXDPState, nc, nc0);
if (bpf_xdp_query_id(s->ifindex, s->xdp_flags, &prog_id) || !prog_id) {
error_setg_errno(errp, errno,
diff --git a/qapi/net.json b/qapi/net.json
index 97ea183981..3d80a9cacd 100644
--- a/qapi/net.json
+++ b/qapi/net.json
@@ -454,25 +454,34 @@
# (default: 0).
#
# @inhibit: Don't load a default XDP program, use one already loaded
-# to the interface (default: false). Requires @sock-fds.
+# to the interface (default: false). Requires @sock-fds or @map-path.
#
# @sock-fds: A colon (:) separated list of file descriptors for
# already open but not bound AF_XDP sockets in the queue order.
# One fd per queue. These descriptors should already be added
-# into XDP socket map for corresponding queues. Requires
-# @inhibit.
+# into XDP socket map for corresponding queues. @sock-fds and
+# @map-path are mutually exclusive. Requires @inhibit.
+#
+# @map-path: The path to a pinned xsk map to push file descriptors
+# for bound AF_XDP sockets into. @map-path and @sock-fds are
+# mutually exclusive. Requires @inhibit. (Since 10.1)
+#
+# @map-start-index: Use @map-path to insert xsk sockets starting from
+# this index number (default: 0). Requires @map-path. (Since 10.1)
#
# Since: 8.2
##
{ 'struct': 'NetdevAFXDPOptions',
'data': {
- 'ifname': 'str',
- '*mode': 'AFXDPMode',
- '*force-copy': 'bool',
- '*queues': 'int',
- '*start-queue': 'int',
- '*inhibit': 'bool',
- '*sock-fds': 'str' },
+ 'ifname': 'str',
+ '*mode': 'AFXDPMode',
+ '*force-copy': 'bool',
+ '*queues': 'int',
+ '*start-queue': 'int',
+ '*inhibit': 'bool',
+ '*sock-fds': 'str',
+ '*map-path': 'str',
+ '*map-start-index': 'int' },
'if': 'CONFIG_AF_XDP' }
##
diff --git a/qemu-options.hx b/qemu-options.hx
index 1f862b19a6..0fd4fd8d46 100644
--- a/qemu-options.hx
+++ b/qemu-options.hx
@@ -2909,6 +2909,7 @@ DEF("netdev", HAS_ARG, QEMU_OPTION_netdev,
#ifdef CONFIG_AF_XDP
"-netdev af-xdp,id=str,ifname=name[,mode=native|skb][,force-copy=on|off]\n"
" [,queues=n][,start-queue=m][,inhibit=on|off][,sock-fds=x:y:...:z]\n"
+ " [,map-path=/path/to/socket/map][,map-start-index=i]\n"
" attach to the existing network interface 'name' with AF_XDP socket\n"
" use 'mode=MODE' to specify an XDP program attach mode\n"
" use 'force-copy=on|off' to force XDP copy mode even if device supports zero-copy (default: off)\n"
@@ -2916,6 +2917,8 @@ DEF("netdev", HAS_ARG, QEMU_OPTION_netdev,
" with inhibit=on,\n"
" use 'sock-fds' to provide file descriptors for already open AF_XDP sockets\n"
" added to a socket map in XDP program. One socket per queue.\n"
+ " use 'map-path' to provide the socket map location to populate AF_XDP sockets with\n"
+ " beginning from the 'map-start-index' specified index (default: 0) (Since 10.1)\n"
" use 'queues=n' to specify how many queues of a multiqueue interface should be used\n"
" use 'start-queue=m' to specify the first queue that should be used\n"
#endif
@@ -3610,7 +3613,7 @@ SRST
# launch QEMU instance
|qemu_system| linux.img -nic vde,sock=/tmp/myswitch
-``-netdev af-xdp,id=str,ifname=name[,mode=native|skb][,force-copy=on|off][,queues=n][,start-queue=m][,inhibit=on|off][,sock-fds=x:y:...:z]``
+``-netdev af-xdp,id=str,ifname=name[,mode=native|skb][,force-copy=on|off][,queues=n][,start-queue=m][,inhibit=on|off][,sock-fds=x:y:...:z][,map-path=/path/to/socket/map][,map-start-index=i]``
Configure AF_XDP backend to connect to a network interface 'name'
using AF_XDP socket. A specific program attach mode for a default
XDP program can be forced with 'mode', defaults to best-effort,
@@ -3650,7 +3653,8 @@ SRST
-netdev af-xdp,id=n1,ifname=eth0,queues=1,start-queue=1
XDP program can also be loaded externally. In this case 'inhibit' option
- should be set to 'on' and 'sock-fds' provided with file descriptors for
+ should be set to 'on'. Either 'sock-fds' or 'map-path' can be used with
+ 'inhibit' enabled. 'sock-fds' can be provided with file descriptors for
already open but not bound XDP sockets already added to a socket map for
corresponding queues. One socket per queue.
@@ -3659,6 +3663,21 @@ SRST
|qemu_system| linux.img -device virtio-net-pci,netdev=n1 \\
-netdev af-xdp,id=n1,ifname=eth0,queues=3,inhibit=on,sock-fds=15:16:17
+ For the 'inhibit' option set to 'on' used together with 'map-path' it is
+ expected that the XDP program with the socket map is already loaded on
+ the networking device and the map pinned into BPF file system. The path
+ to the pinned map is then passed to QEMU which then creates the file
+ descriptors and inserts them into the existing socket map.
+
+ .. parsed-literal::
+
+ |qemu_system| linux.img -device virtio-net-pci,netdev=n1 \\
+ -netdev af-xdp,id=n1,ifname=eth0,queues=2,inhibit=on,map-path=/sys/fs/bpf/xsks_map
+
+ Additionally, 'map-start-index' can be used to specify the start offset
+ for insertion into the socket map. The combination of 'map-path' and
+ 'sock-fds' together is not supported.
+
``-netdev vhost-user,chardev=id[,vhostforce=on|off][,queues=n]``
Establish a vhost-user netdev, backed by a chardev id. The chardev
should be a unix domain socket backed one. The vhost-user uses a
--
2.43.0
^ permalink raw reply related [flat|nested] 7+ messages in thread
* Re: [PATCH v4 1/3] net/af-xdp: Remove XDP program cleanup logic
2025-07-04 13:05 [PATCH v4 1/3] net/af-xdp: Remove XDP program cleanup logic Daniel Borkmann via
2025-07-04 13:05 ` [PATCH v4 2/3] net/af-xdp: Fix up cleanup path upon failure in queue creation Daniel Borkmann via
2025-07-04 13:05 ` [PATCH v4 3/3] net/af-xdp: Support pinned map path for AF_XDP sockets Daniel Borkmann via
@ 2025-07-09 11:24 ` Ilya Maximets
2 siblings, 0 replies; 7+ messages in thread
From: Ilya Maximets @ 2025-07-09 11:24 UTC (permalink / raw)
To: Daniel Borkmann, qemu-devel; +Cc: i.maximets, Jason Wang, Anton Protopopov
On 7/4/25 3:05 PM, Daniel Borkmann wrote:
> There are two issues with the XDP program removal in af_xdp_cleanup():
>
> 1) Starting from libxdp 1.3.0 [0] the XDP program gets automatically
> detached when we call xsk_socket__delete() for the last successfully
> configured queue. libxdp internally keeps track of that. For QEMU
> we require libxdp >= 1.4.0. Given QEMU is not loading the program,
> lets also not attempt to remove it and delegate this instead.
>
> 2) The removal logic is incorrect anyway because we are setting n_queues
> into the last queue that never has xdp_flags on failure, so the logic
> is always skipped since the non-zero test for s->xdp_flags in
> af_xdp_cleanup() fails.
>
> Fixes: cb039ef3d9e3 ("net: add initial support for AF_XDP network backend")
> Suggested-by: Ilya Maximets <i.maximets@ovn.org>
> Signed-off-by: Daniel Borkmann <daniel@iogearbox.net>
> Cc: Ilya Maximets <i.maximets@ovn.org>
> Cc: Jason Wang <jasowang@redhat.com>
> Cc: Anton Protopopov <aspsk@isovalent.com>
> Link: https://github.com/xdp-project/xdp-tools/commit/38c2914988fd5c1ef65f2381fc8af9f3e8404e2b [0]
> ---
Thanks!
Reviewed-by: Ilya Maximets <i.maximets@ovn.org>
^ permalink raw reply [flat|nested] 7+ messages in thread
* Re: [PATCH v4 2/3] net/af-xdp: Fix up cleanup path upon failure in queue creation
2025-07-04 13:05 ` [PATCH v4 2/3] net/af-xdp: Fix up cleanup path upon failure in queue creation Daniel Borkmann via
@ 2025-07-09 11:26 ` Ilya Maximets
0 siblings, 0 replies; 7+ messages in thread
From: Ilya Maximets @ 2025-07-09 11:26 UTC (permalink / raw)
To: Daniel Borkmann, qemu-devel; +Cc: i.maximets, Jason Wang, Anton Protopopov
On 7/4/25 3:05 PM, Daniel Borkmann wrote:
> While testing, it turned out that upon error in the queue creation loop,
> we never trigger the af_xdp_cleanup() handler. This is because we pass
> errp instead of a local err pointer into the various AF_XDP setup functions
> instead of a scheme like:
>
> bool fn(..., Error **errp)
> {
> Error *err = NULL;
>
> foo(arg, &err);
> if (err) {
> handle the error...
> error_propagate(errp, err);
> return false;
> }
> ...
> }
>
> With a conversion into the above format, the af_xdp_cleanup() handler is
> called as expected. Note the error_propagate() handles a NULL err internally.
>
> Fixes: cb039ef3d9e3 ("net: add initial support for AF_XDP network backend")
> Signed-off-by: Daniel Borkmann <daniel@iogearbox.net>
> Cc: Ilya Maximets <i.maximets@ovn.org>
> Cc: Jason Wang <jasowang@redhat.com>
> Cc: Anton Protopopov <aspsk@isovalent.com>
> ---
> net/af-xdp.c | 6 +++---
> 1 file changed, 3 insertions(+), 3 deletions(-)
>
> diff --git a/net/af-xdp.c b/net/af-xdp.c
> index c5d3b6a953..29c5ad16cd 100644
> --- a/net/af-xdp.c
> +++ b/net/af-xdp.c
> @@ -482,9 +482,8 @@ int net_init_af_xdp(const Netdev *netdev,
> pstrcpy(s->ifname, sizeof(s->ifname), opts->ifname);
> s->ifindex = ifindex;
>
> - if (af_xdp_umem_create(s, sock_fds ? sock_fds[i] : -1, errp)
> - || af_xdp_socket_create(s, opts, errp)) {
> - error_propagate(errp, err);
> + if (af_xdp_umem_create(s, sock_fds ? sock_fds[i] : -1, &err) ||
> + af_xdp_socket_create(s, opts, &err)) {
> goto err;
> }
> }
We should also convert the "no XDP program loaded", I suppose.
> @@ -506,6 +505,7 @@ int net_init_af_xdp(const Netdev *netdev,
> err:
> if (nc0) {
> qemu_del_net_client(nc0);
> + error_propagate(errp, err);
> }
>
> return -1;
^ permalink raw reply [flat|nested] 7+ messages in thread
* Re: [PATCH v4 3/3] net/af-xdp: Support pinned map path for AF_XDP sockets
2025-07-04 13:05 ` [PATCH v4 3/3] net/af-xdp: Support pinned map path for AF_XDP sockets Daniel Borkmann via
@ 2025-07-09 12:29 ` Ilya Maximets
2025-07-11 9:49 ` Daniel Borkmann via
0 siblings, 1 reply; 7+ messages in thread
From: Ilya Maximets @ 2025-07-09 12:29 UTC (permalink / raw)
To: Daniel Borkmann, qemu-devel; +Cc: i.maximets, Jason Wang, Anton Protopopov
On 7/4/25 3:05 PM, Daniel Borkmann wrote:
> Extend 'inhibit=on' setting with the option to specify a pinned XSK map
> path along with a starting index (default 0) to push the created XSK
> sockets into. Example usage:
>
> # ./build/qemu-system-x86_64 [...] \
> -netdev af-xdp,ifname=enp2s0f0np0,id=net0,mode=native,queues=2,start-queue=14,inhibit=on,map-path=/sys/fs/bpf/xsks_map,map-start-index=14 \
> -device virtio-net-pci,netdev=net0 [...]
>
> This is useful for the case where an existing XDP program with XSK map
> is present on the AF_XDP supported phys device and the XSK map is not
> yet populated. For example, the former could have been pre-loaded onto
> the netdevice by a control plane, which later launches QEMU to populate
> it with XSK sockets.
>
> Normally, the main idea behind 'inhibit=on' is that the QEMU instance
> doesn't need to have a lot of privileges to use the pre-loaded program
> and the pre-created sockets, but this mentioned use-case here is different
> where QEMU still needs privileges to create the sockets.
>
> The 'map-start-index' parameter is optional and defaults to 0. It allows
> flexible placement of the XSK sockets, and is up to the user to specify
> when the XDP program with XSK map was already preloaded. In the simplest
> case the queue-to-map-slot mapping is just 1:1 based on ctx->rx_queue_index
> but the user might as well have a different scheme (or smaller map size,
> e.g. ctx->rx_queue_index % max_size) to push the inbound traffic to one
> of the XSK sockets.
>
> Note that the bpf_xdp_query_id() is now only tested for 'inhibit=off'
> since only in the latter case the libxdp takes care of installing the
> XDP program which was installed based on the s->xdp_flags pointing to
> either driver or skb mode. For 'inhibit=on' we don't make any assumptions
> and neither go down the path of probing all possible options in which
> way the user installed the XDP program.
>
> Signed-off-by: Daniel Borkmann <daniel@iogearbox.net>
> Cc: Ilya Maximets <i.maximets@ovn.org>
> Cc: Jason Wang <jasowang@redhat.com>
> Cc: Anton Protopopov <aspsk@isovalent.com>
> ---
> net/af-xdp.c | 78 +++++++++++++++++++++++++++++++++++++++++++++----
> qapi/net.json | 29 +++++++++++-------
> qemu-options.hx | 23 +++++++++++++--
> 3 files changed, 113 insertions(+), 17 deletions(-)
Thnaks, Daniel! I have just a couple of small nits below that I missed in v3,
the rest looks good and is working fine.
>
> diff --git a/net/af-xdp.c b/net/af-xdp.c
> index 29c5ad16cd..005117c336 100644
> --- a/net/af-xdp.c
> +++ b/net/af-xdp.c
> @@ -51,6 +51,10 @@ typedef struct AFXDPState {
>
> uint32_t xdp_flags;
> bool inhibit;
> +
> + char *map_path;
> + int map_fd;
> + uint32_t map_start_index;
> } AFXDPState;
>
> #define AF_XDP_BATCH_SIZE 64
> @@ -260,6 +264,7 @@ static void af_xdp_send(void *opaque)
> static void af_xdp_cleanup(NetClientState *nc)
> {
> AFXDPState *s = DO_UPCAST(AFXDPState, nc, nc);
> + int idx;
>
> qemu_purge_queued_packets(nc);
>
> @@ -273,6 +278,18 @@ static void af_xdp_cleanup(NetClientState *nc)
> s->umem = NULL;
> qemu_vfree(s->buffer);
> s->buffer = NULL;
> +
> + if (s->map_fd >= 0) {
> + idx = nc->queue_index + s->map_start_index;
> + if (bpf_map_delete_elem(s->map_fd, &idx)) {
> + fprintf(stderr, "af-xdp: unable to remove AF_XDP socket from "
> + "map %s\n", s->map_path);
nit :I'd suggest to keep the "map" on a previous line. We have some space for it.
> + }
> + close(s->map_fd);
> + s->map_fd = -1;
> + }
> + g_free(s->map_path);
> + s->map_path = NULL;
> }
>
> static int af_xdp_umem_create(AFXDPState *s, int sock_fd, Error **errp)
> @@ -336,7 +353,6 @@ static int af_xdp_socket_create(AFXDPState *s,
> };
> int queue_id, error = 0;
>
> - s->inhibit = opts->has_inhibit && opts->inhibit;
> if (s->inhibit) {
> cfg.libxdp_flags |= XSK_LIBXDP_FLAGS__INHIBIT_PROG_LOAD;
> }
> @@ -387,6 +403,35 @@ static int af_xdp_socket_create(AFXDPState *s,
> return 0;
> }
>
> +static int af_xdp_update_xsk_map(AFXDPState *s, Error **errp)
> +{
> + int xsk_fd, idx, error = 0;
> +
> + if (!s->map_path) {
> + return 0;
> + }
> +
> + s->map_fd = bpf_obj_get(s->map_path);
> + if (s->map_fd < 0) {
> + error = errno;
> + } else {
> + xsk_fd = xsk_socket__fd(s->xsk);
> + idx = s->nc.queue_index + s->map_start_index;
> + if (bpf_map_update_elem(s->map_fd, &idx, &xsk_fd, 0)) {
> + error = errno;
> + }
> + }
> +
> + if (error) {
> + error_setg_errno(errp, error,
> + "failed to insert AF_XDP socket into map %s",
> + s->map_path);
> + return -1;
nit: Maybe remove this line and return 'error' below?
> + }
> +
> + return 0;
> +}
> +
> /* NetClientInfo methods. */
> static NetClientInfo net_af_xdp_info = {
> .type = NET_CLIENT_DRIVER_AF_XDP,
> @@ -441,6 +486,7 @@ int net_init_af_xdp(const Netdev *netdev,
> int64_t i, queues;
> Error *err = NULL;
> AFXDPState *s;
> + bool inhibit;
>
> ifindex = if_nametoindex(opts->ifname);
> if (!ifindex) {
> @@ -456,8 +502,21 @@ int net_init_af_xdp(const Netdev *netdev,
> return -1;
> }
>
> - if ((opts->has_inhibit && opts->inhibit) != !!opts->sock_fds) {
> - error_setg(errp, "'inhibit=on' requires 'sock-fds' and vice versa");
> + inhibit = opts->has_inhibit && opts->inhibit;
> + if (inhibit && !opts->sock_fds && !opts->map_path) {
> + error_setg(errp, "'inhibit=on' requires 'sock-fds' or 'map-path'");
> + return -1;
> + }
> + if (!inhibit && (opts->sock_fds || opts->map_path)) {
> + error_setg(errp, "'sock-fds' and 'map-path' require 'inhibit=on'");
> + return -1;
> + }
> + if (opts->sock_fds && opts->map_path) {
> + error_setg(errp, "'sock-fds' and 'map-path' are mutually exclusive");
> + return -1;
> + }
> + if (!opts->map_path && opts->has_map_start_index) {
> + error_setg(errp, "'map-start-index' requires 'map-path'");
> return -1;
> }
>
> @@ -481,14 +540,23 @@ int net_init_af_xdp(const Netdev *netdev,
>
> pstrcpy(s->ifname, sizeof(s->ifname), opts->ifname);
> s->ifindex = ifindex;
> + s->inhibit = inhibit;
> +
> + s->map_path = g_strdup(opts->map_path);
> + s->map_fd = -1;
> + s->map_start_index = 0;
> + if (opts->has_map_start_index && opts->map_start_index > 0) {
We should error out if the user specified a negative value. I'd suggest
to add the check to the list of user-input validation above instead of
silently ignoring the incorrect value. And then we could skip the value
check here.
> + s->map_start_index = opts->map_start_index;
> + }
>
> if (af_xdp_umem_create(s, sock_fds ? sock_fds[i] : -1, &err) ||
> - af_xdp_socket_create(s, opts, &err)) {
> + af_xdp_socket_create(s, opts, &err) ||
> + af_xdp_update_xsk_map(s, &err)) {
> goto err;
> }
> }
>
> - if (nc0) {
> + if (nc0 && !inhibit) {
> s = DO_UPCAST(AFXDPState, nc, nc0);
> if (bpf_xdp_query_id(s->ifindex, s->xdp_flags, &prog_id) || !prog_id) {
> error_setg_errno(errp, errno,
> diff --git a/qapi/net.json b/qapi/net.json
> index 97ea183981..3d80a9cacd 100644
> --- a/qapi/net.json
> +++ b/qapi/net.json
> @@ -454,25 +454,34 @@
> # (default: 0).
> #
> # @inhibit: Don't load a default XDP program, use one already loaded
> -# to the interface (default: false). Requires @sock-fds.
> +# to the interface (default: false). Requires @sock-fds or @map-path.
> #
> # @sock-fds: A colon (:) separated list of file descriptors for
> # already open but not bound AF_XDP sockets in the queue order.
> # One fd per queue. These descriptors should already be added
> -# into XDP socket map for corresponding queues. Requires
> -# @inhibit.
> +# into XDP socket map for corresponding queues. @sock-fds and
> +# @map-path are mutually exclusive. Requires @inhibit.
> +#
> +# @map-path: The path to a pinned xsk map to push file descriptors
> +# for bound AF_XDP sockets into. @map-path and @sock-fds are
> +# mutually exclusive. Requires @inhibit. (Since 10.1)
> +#
> +# @map-start-index: Use @map-path to insert xsk sockets starting from
> +# this index number (default: 0). Requires @map-path. (Since 10.1)
> #
> # Since: 8.2
> ##
> { 'struct': 'NetdevAFXDPOptions',
> 'data': {
> - 'ifname': 'str',
> - '*mode': 'AFXDPMode',
> - '*force-copy': 'bool',
> - '*queues': 'int',
> - '*start-queue': 'int',
> - '*inhibit': 'bool',
> - '*sock-fds': 'str' },
> + 'ifname': 'str',
> + '*mode': 'AFXDPMode',
> + '*force-copy': 'bool',
> + '*queues': 'int',
> + '*start-queue': 'int',
> + '*inhibit': 'bool',
> + '*sock-fds': 'str',
> + '*map-path': 'str',
> + '*map-start-index': 'int' },
> 'if': 'CONFIG_AF_XDP' }
>
> ##
> diff --git a/qemu-options.hx b/qemu-options.hx
> index 1f862b19a6..0fd4fd8d46 100644
> --- a/qemu-options.hx
> +++ b/qemu-options.hx
> @@ -2909,6 +2909,7 @@ DEF("netdev", HAS_ARG, QEMU_OPTION_netdev,
> #ifdef CONFIG_AF_XDP
> "-netdev af-xdp,id=str,ifname=name[,mode=native|skb][,force-copy=on|off]\n"
> " [,queues=n][,start-queue=m][,inhibit=on|off][,sock-fds=x:y:...:z]\n"
> + " [,map-path=/path/to/socket/map][,map-start-index=i]\n"
> " attach to the existing network interface 'name' with AF_XDP socket\n"
> " use 'mode=MODE' to specify an XDP program attach mode\n"
> " use 'force-copy=on|off' to force XDP copy mode even if device supports zero-copy (default: off)\n"
> @@ -2916,6 +2917,8 @@ DEF("netdev", HAS_ARG, QEMU_OPTION_netdev,
> " with inhibit=on,\n"
> " use 'sock-fds' to provide file descriptors for already open AF_XDP sockets\n"
> " added to a socket map in XDP program. One socket per queue.\n"
> + " use 'map-path' to provide the socket map location to populate AF_XDP sockets with\n"
nit: it feels like we need some punctuation sign after the 'with', otherwise
it reads as "populate AF_XDP sockets with beginning from the", which makes no
sense.
> + " beginning from the 'map-start-index' specified index (default: 0) (Since 10.1)\n"
> " use 'queues=n' to specify how many queues of a multiqueue interface should be used\n"
> " use 'start-queue=m' to specify the first queue that should be used\n"
> #endif
> @@ -3610,7 +3613,7 @@ SRST
> # launch QEMU instance
> |qemu_system| linux.img -nic vde,sock=/tmp/myswitch
>
> -``-netdev af-xdp,id=str,ifname=name[,mode=native|skb][,force-copy=on|off][,queues=n][,start-queue=m][,inhibit=on|off][,sock-fds=x:y:...:z]``
> +``-netdev af-xdp,id=str,ifname=name[,mode=native|skb][,force-copy=on|off][,queues=n][,start-queue=m][,inhibit=on|off][,sock-fds=x:y:...:z][,map-path=/path/to/socket/map][,map-start-index=i]``
> Configure AF_XDP backend to connect to a network interface 'name'
> using AF_XDP socket. A specific program attach mode for a default
> XDP program can be forced with 'mode', defaults to best-effort,
> @@ -3650,7 +3653,8 @@ SRST
> -netdev af-xdp,id=n1,ifname=eth0,queues=1,start-queue=1
>
> XDP program can also be loaded externally. In this case 'inhibit' option
> - should be set to 'on' and 'sock-fds' provided with file descriptors for
> + should be set to 'on'. Either 'sock-fds' or 'map-path' can be used with
> + 'inhibit' enabled. 'sock-fds' can be provided with file descriptors for
> already open but not bound XDP sockets already added to a socket map for
> corresponding queues. One socket per queue.
>
> @@ -3659,6 +3663,21 @@ SRST
> |qemu_system| linux.img -device virtio-net-pci,netdev=n1 \\
> -netdev af-xdp,id=n1,ifname=eth0,queues=3,inhibit=on,sock-fds=15:16:17
>
> + For the 'inhibit' option set to 'on' used together with 'map-path' it is
> + expected that the XDP program with the socket map is already loaded on
> + the networking device and the map pinned into BPF file system. The path
> + to the pinned map is then passed to QEMU which then creates the file
> + descriptors and inserts them into the existing socket map.
> +
> + .. parsed-literal::
> +
> + |qemu_system| linux.img -device virtio-net-pci,netdev=n1 \\
> + -netdev af-xdp,id=n1,ifname=eth0,queues=2,inhibit=on,map-path=/sys/fs/bpf/xsks_map
> +
> + Additionally, 'map-start-index' can be used to specify the start offset
> + for insertion into the socket map. The combination of 'map-path' and
> + 'sock-fds' together is not supported.
> +
> ``-netdev vhost-user,chardev=id[,vhostforce=on|off][,queues=n]``
> Establish a vhost-user netdev, backed by a chardev id. The chardev
> should be a unix domain socket backed one. The vhost-user uses a
^ permalink raw reply [flat|nested] 7+ messages in thread
* Re: [PATCH v4 3/3] net/af-xdp: Support pinned map path for AF_XDP sockets
2025-07-09 12:29 ` Ilya Maximets
@ 2025-07-11 9:49 ` Daniel Borkmann via
0 siblings, 0 replies; 7+ messages in thread
From: Daniel Borkmann via @ 2025-07-11 9:49 UTC (permalink / raw)
To: Ilya Maximets, qemu-devel; +Cc: Jason Wang, Anton Protopopov
Hey Ilya,
On 7/9/25 2:29 PM, Ilya Maximets wrote:
[...]
> Thnaks, Daniel! I have just a couple of small nits below that I missed in v3,
> the rest looks good and is working fine.
>
>>
>> diff --git a/net/af-xdp.c b/net/af-xdp.c
>> index 29c5ad16cd..005117c336 100644
>> --- a/net/af-xdp.c
>> +++ b/net/af-xdp.c
>> @@ -51,6 +51,10 @@ typedef struct AFXDPState {
>>
>> uint32_t xdp_flags;
>> bool inhibit;
>> +
>> + char *map_path;
>> + int map_fd;
>> + uint32_t map_start_index;
>> } AFXDPState;
>>
>> #define AF_XDP_BATCH_SIZE 64
>> @@ -260,6 +264,7 @@ static void af_xdp_send(void *opaque)
>> static void af_xdp_cleanup(NetClientState *nc)
>> {
>> AFXDPState *s = DO_UPCAST(AFXDPState, nc, nc);
>> + int idx;
>>
>> qemu_purge_queued_packets(nc);
>>
>> @@ -273,6 +278,18 @@ static void af_xdp_cleanup(NetClientState *nc)
>> s->umem = NULL;
>> qemu_vfree(s->buffer);
>> s->buffer = NULL;
>> +
>> + if (s->map_fd >= 0) {
>> + idx = nc->queue_index + s->map_start_index;
>> + if (bpf_map_delete_elem(s->map_fd, &idx)) {
>> + fprintf(stderr, "af-xdp: unable to remove AF_XDP socket from "
>> + "map %s\n", s->map_path);
>
> nit :I'd suggest to keep the "map" on a previous line. We have some space for it.
fixed.
>> + }
>> + close(s->map_fd);
>> + s->map_fd = -1;
>> + }
>> + g_free(s->map_path);
>> + s->map_path = NULL;
>> }
>>
>> static int af_xdp_umem_create(AFXDPState *s, int sock_fd, Error **errp)
>> @@ -336,7 +353,6 @@ static int af_xdp_socket_create(AFXDPState *s,
>> };
>> int queue_id, error = 0;
>>
>> - s->inhibit = opts->has_inhibit && opts->inhibit;
>> if (s->inhibit) {
>> cfg.libxdp_flags |= XSK_LIBXDP_FLAGS__INHIBIT_PROG_LOAD;
>> }
>> @@ -387,6 +403,35 @@ static int af_xdp_socket_create(AFXDPState *s,
>> return 0;
>> }
>>
>> +static int af_xdp_update_xsk_map(AFXDPState *s, Error **errp)
>> +{
>> + int xsk_fd, idx, error = 0;
>> +
>> + if (!s->map_path) {
>> + return 0;
>> + }
>> +
>> + s->map_fd = bpf_obj_get(s->map_path);
>> + if (s->map_fd < 0) {
>> + error = errno;
>> + } else {
>> + xsk_fd = xsk_socket__fd(s->xsk);
>> + idx = s->nc.queue_index + s->map_start_index;
>> + if (bpf_map_update_elem(s->map_fd, &idx, &xsk_fd, 0)) {
>> + error = errno;
>> + }
>> + }
>> +
>> + if (error) {
>> + error_setg_errno(errp, error,
>> + "failed to insert AF_XDP socket into map %s",
>> + s->map_path);
>> + return -1;
>
> nit: Maybe remove this line and return 'error' below?
I left this as-is given the rest of the code has similar style as here
and always returns -1 on errors. error would be positive here and we'd
have to return -error, but still it seems somewhat out of place imho.
>> + }
>> +
>> + return 0;
>> +}
>> +
>> /* NetClientInfo methods. */
>> static NetClientInfo net_af_xdp_info = {
>> .type = NET_CLIENT_DRIVER_AF_XDP,
>> @@ -441,6 +486,7 @@ int net_init_af_xdp(const Netdev *netdev,
>> int64_t i, queues;
>> Error *err = NULL;
>> AFXDPState *s;
>> + bool inhibit;
>>
>> ifindex = if_nametoindex(opts->ifname);
>> if (!ifindex) {
>> @@ -456,8 +502,21 @@ int net_init_af_xdp(const Netdev *netdev,
>> return -1;
>> }
>>
>> - if ((opts->has_inhibit && opts->inhibit) != !!opts->sock_fds) {
>> - error_setg(errp, "'inhibit=on' requires 'sock-fds' and vice versa");
>> + inhibit = opts->has_inhibit && opts->inhibit;
>> + if (inhibit && !opts->sock_fds && !opts->map_path) {
>> + error_setg(errp, "'inhibit=on' requires 'sock-fds' or 'map-path'");
>> + return -1;
>> + }
>> + if (!inhibit && (opts->sock_fds || opts->map_path)) {
>> + error_setg(errp, "'sock-fds' and 'map-path' require 'inhibit=on'");
>> + return -1;
>> + }
>> + if (opts->sock_fds && opts->map_path) {
>> + error_setg(errp, "'sock-fds' and 'map-path' are mutually exclusive");
>> + return -1;
>> + }
>> + if (!opts->map_path && opts->has_map_start_index) {
>> + error_setg(errp, "'map-start-index' requires 'map-path'");
>> return -1;
>> }
>>
>> @@ -481,14 +540,23 @@ int net_init_af_xdp(const Netdev *netdev,
>>
>> pstrcpy(s->ifname, sizeof(s->ifname), opts->ifname);
>> s->ifindex = ifindex;
>> + s->inhibit = inhibit;
>> +
>> + s->map_path = g_strdup(opts->map_path);
>> + s->map_fd = -1;
>> + s->map_start_index = 0;
>> + if (opts->has_map_start_index && opts->map_start_index > 0) {
>
> We should error out if the user specified a negative value. I'd suggest
> to add the check to the list of user-input validation above instead of
> silently ignoring the incorrect value. And then we could skip the value
> check here.
fixed.
>> + s->map_start_index = opts->map_start_index;
>> + }
>>
>> if (af_xdp_umem_create(s, sock_fds ? sock_fds[i] : -1, &err) ||
>> - af_xdp_socket_create(s, opts, &err)) {
>> + af_xdp_socket_create(s, opts, &err) ||
>> + af_xdp_update_xsk_map(s, &err)) {
>> goto err;
>> }
>> }
>>
>> - if (nc0) {
>> + if (nc0 && !inhibit) {
>> s = DO_UPCAST(AFXDPState, nc, nc0);
>> if (bpf_xdp_query_id(s->ifindex, s->xdp_flags, &prog_id) || !prog_id) {
>> error_setg_errno(errp, errno,
>> diff --git a/qapi/net.json b/qapi/net.json
>> index 97ea183981..3d80a9cacd 100644
>> --- a/qapi/net.json
>> +++ b/qapi/net.json
>> @@ -454,25 +454,34 @@
>> # (default: 0).
>> #
>> # @inhibit: Don't load a default XDP program, use one already loaded
>> -# to the interface (default: false). Requires @sock-fds.
>> +# to the interface (default: false). Requires @sock-fds or @map-path.
>> #
>> # @sock-fds: A colon (:) separated list of file descriptors for
>> # already open but not bound AF_XDP sockets in the queue order.
>> # One fd per queue. These descriptors should already be added
>> -# into XDP socket map for corresponding queues. Requires
>> -# @inhibit.
>> +# into XDP socket map for corresponding queues. @sock-fds and
>> +# @map-path are mutually exclusive. Requires @inhibit.
>> +#
>> +# @map-path: The path to a pinned xsk map to push file descriptors
>> +# for bound AF_XDP sockets into. @map-path and @sock-fds are
>> +# mutually exclusive. Requires @inhibit. (Since 10.1)
>> +#
>> +# @map-start-index: Use @map-path to insert xsk sockets starting from
>> +# this index number (default: 0). Requires @map-path. (Since 10.1)
>> #
>> # Since: 8.2
>> ##
>> { 'struct': 'NetdevAFXDPOptions',
>> 'data': {
>> - 'ifname': 'str',
>> - '*mode': 'AFXDPMode',
>> - '*force-copy': 'bool',
>> - '*queues': 'int',
>> - '*start-queue': 'int',
>> - '*inhibit': 'bool',
>> - '*sock-fds': 'str' },
>> + 'ifname': 'str',
>> + '*mode': 'AFXDPMode',
>> + '*force-copy': 'bool',
>> + '*queues': 'int',
>> + '*start-queue': 'int',
>> + '*inhibit': 'bool',
>> + '*sock-fds': 'str',
>> + '*map-path': 'str',
>> + '*map-start-index': 'int' },
>> 'if': 'CONFIG_AF_XDP' }
>>
>> ##
>> diff --git a/qemu-options.hx b/qemu-options.hx
>> index 1f862b19a6..0fd4fd8d46 100644
>> --- a/qemu-options.hx
>> +++ b/qemu-options.hx
>> @@ -2909,6 +2909,7 @@ DEF("netdev", HAS_ARG, QEMU_OPTION_netdev,
>> #ifdef CONFIG_AF_XDP
>> "-netdev af-xdp,id=str,ifname=name[,mode=native|skb][,force-copy=on|off]\n"
>> " [,queues=n][,start-queue=m][,inhibit=on|off][,sock-fds=x:y:...:z]\n"
>> + " [,map-path=/path/to/socket/map][,map-start-index=i]\n"
>> " attach to the existing network interface 'name' with AF_XDP socket\n"
>> " use 'mode=MODE' to specify an XDP program attach mode\n"
>> " use 'force-copy=on|off' to force XDP copy mode even if device supports zero-copy (default: off)\n"
>> @@ -2916,6 +2917,8 @@ DEF("netdev", HAS_ARG, QEMU_OPTION_netdev,
>> " with inhibit=on,\n"
>> " use 'sock-fds' to provide file descriptors for already open AF_XDP sockets\n"
>> " added to a socket map in XDP program. One socket per queue.\n"
>> + " use 'map-path' to provide the socket map location to populate AF_XDP sockets with\n"
>
> nit: it feels like we need some punctuation sign after the 'with', otherwise
> it reads as "populate AF_XDP sockets with beginning from the", which makes no
> sense.
ok, reworded slightly.
Thanks,
Daniel
^ permalink raw reply [flat|nested] 7+ messages in thread
end of thread, other threads:[~2025-07-11 9:49 UTC | newest]
Thread overview: 7+ messages (download: mbox.gz follow: Atom feed
-- links below jump to the message on this page --
2025-07-04 13:05 [PATCH v4 1/3] net/af-xdp: Remove XDP program cleanup logic Daniel Borkmann via
2025-07-04 13:05 ` [PATCH v4 2/3] net/af-xdp: Fix up cleanup path upon failure in queue creation Daniel Borkmann via
2025-07-09 11:26 ` Ilya Maximets
2025-07-04 13:05 ` [PATCH v4 3/3] net/af-xdp: Support pinned map path for AF_XDP sockets Daniel Borkmann via
2025-07-09 12:29 ` Ilya Maximets
2025-07-11 9:49 ` Daniel Borkmann via
2025-07-09 11:24 ` [PATCH v4 1/3] net/af-xdp: Remove XDP program cleanup logic Ilya Maximets
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).