Skip to content

xdp: add XDP extension to skb #6241

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Closed
wants to merge 2 commits into from
Closed

xdp: add XDP extension to skb #6241

Show file tree
Hide file tree
Changes from all commits
Commits
Show all changes
2 commits
Select commit Hold shift + click to select a range
c0a0c26
xdp: add XDP extension to skb
Oct 28, 2025
8154dd8
xsk: avoid data corruption on cq descriptor number
Oct 28, 2025
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
3 changes: 3 additions & 0 deletions include/linux/skbuff.h
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -4910,6 +4910,9 @@ enum skb_ext_id {
#endif
#if IS_ENABLED(CONFIG_INET_PSP)
SKB_EXT_PSP,
#endif
#if IS_ENABLED(CONFIG_XDP_SOCKETS)
SKB_EXT_XDP,
#endif
SKB_EXT_NUM, /* must be last */
};
Expand Down
5 changes: 5 additions & 0 deletions include/net/xdp_sock.h
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -120,6 +120,11 @@ struct xsk_tx_metadata_ops {
void (*tmo_request_launch_time)(u64 launch_time, void *priv);
};

struct xdp_skb_ext {
u32 num_descs;
struct list_head addrs_list;
};

#ifdef CONFIG_XDP_SOCKETS

int xsk_generic_rcv(struct xdp_sock *xs, struct xdp_buff *xdp);
Expand Down
4 changes: 4 additions & 0 deletions net/core/skbuff.c
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -81,6 +81,7 @@
#include <net/page_pool/helpers.h>
#include <net/psp/types.h>
#include <net/dropreason.h>
#include <net/xdp_sock.h>

#include <linux/uaccess.h>
#include <trace/events/skb.h>
Expand Down Expand Up @@ -5066,6 +5067,9 @@ static const u8 skb_ext_type_len[] = {
#if IS_ENABLED(CONFIG_INET_PSP)
[SKB_EXT_PSP] = SKB_EXT_CHUNKSIZEOF(struct psp_skb_ext),
#endif
#if IS_ENABLED(CONFIG_XDP_SOCKETS)
[SKB_EXT_XDP] = SKB_EXT_CHUNKSIZEOF(struct xdp_skb_ext),
#endif
};

static __always_inline unsigned int skb_ext_total_length(void)
Expand Down
1 change: 1 addition & 0 deletions net/xdp/Kconfig
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -2,6 +2,7 @@
config XDP_SOCKETS
bool "XDP sockets"
depends on BPF_SYSCALL
select SKB_EXTENSIONS
default n
help
XDP sockets allows a channel between XDP programs and
Expand Down
70 changes: 49 additions & 21 deletions net/xdp/xsk.c
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -41,15 +41,8 @@ struct xsk_addr_node {
struct list_head addr_node;
};

struct xsk_addr_head {
u32 num_descs;
struct list_head addrs_list;
};

static struct kmem_cache *xsk_tx_generic_cache;

#define XSKCB(skb) ((struct xsk_addr_head *)((skb)->cb))

void xsk_set_rx_need_wakeup(struct xsk_buff_pool *pool)
{
if (pool->cached_need_wakeup & XDP_WAKEUP_RX)
Expand Down Expand Up @@ -562,6 +555,7 @@ static void xsk_cq_submit_addr_locked(struct xsk_buff_pool *pool,
struct sk_buff *skb)
{
struct xsk_addr_node *pos, *tmp;
struct xdp_skb_ext *ext;
u32 descs_processed = 0;
unsigned long flags;
u32 idx;
Expand All @@ -573,14 +567,16 @@ static void xsk_cq_submit_addr_locked(struct xsk_buff_pool *pool,
(u64)(uintptr_t)skb_shinfo(skb)->destructor_arg);
descs_processed++;

if (unlikely(XSKCB(skb)->num_descs > 1)) {
list_for_each_entry_safe(pos, tmp, &XSKCB(skb)->addrs_list, addr_node) {
ext = skb_ext_find(skb, SKB_EXT_XDP);
if (unlikely(ext)) {
list_for_each_entry_safe(pos, tmp, &ext->addrs_list, addr_node) {
xskq_prod_write_addr(pool->cq, idx + descs_processed,
pos->addr);
descs_processed++;
list_del(&pos->addr_node);
kmem_cache_free(xsk_tx_generic_cache, pos);
}
skb_ext_del(skb, SKB_EXT_XDP);
}
xskq_prod_submit_n(pool->cq, descs_processed);
spin_unlock_irqrestore(&pool->cq_lock, flags);
Expand All @@ -597,12 +593,19 @@ static void xsk_cq_cancel_locked(struct xsk_buff_pool *pool, u32 n)

static void xsk_inc_num_desc(struct sk_buff *skb)
{
XSKCB(skb)->num_descs++;
struct xdp_skb_ext *ext;

ext = skb_ext_find(skb, SKB_EXT_XDP);
if (ext)
ext->num_descs++;
}

static u32 xsk_get_num_desc(struct sk_buff *skb)
{
return XSKCB(skb)->num_descs;
struct xdp_skb_ext *ext;

ext = skb_ext_find(skb, SKB_EXT_XDP);
return ext ? ext->num_descs : 1;
}

static void xsk_destruct_skb(struct sk_buff *skb)
Expand All @@ -621,12 +624,9 @@ static void xsk_destruct_skb(struct sk_buff *skb)
static void xsk_skb_init_misc(struct sk_buff *skb, struct xdp_sock *xs,
u64 addr)
{
BUILD_BUG_ON(sizeof(struct xsk_addr_head) > sizeof(skb->cb));
INIT_LIST_HEAD(&XSKCB(skb)->addrs_list);
skb->dev = xs->dev;
skb->priority = READ_ONCE(xs->sk.sk_priority);
skb->mark = READ_ONCE(xs->sk.sk_mark);
XSKCB(skb)->num_descs = 0;
skb->destructor = xsk_destruct_skb;
skb_shinfo(skb)->destructor_arg = (void *)(uintptr_t)addr;
}
Expand All @@ -636,12 +636,15 @@ static void xsk_consume_skb(struct sk_buff *skb)
struct xdp_sock *xs = xdp_sk(skb->sk);
u32 num_descs = xsk_get_num_desc(skb);
struct xsk_addr_node *pos, *tmp;
struct xdp_skb_ext *ext;

if (unlikely(num_descs > 1)) {
list_for_each_entry_safe(pos, tmp, &XSKCB(skb)->addrs_list, addr_node) {
ext = skb_ext_find(skb, SKB_EXT_XDP);
if (unlikely(ext)) {
list_for_each_entry_safe(pos, tmp, &ext->addrs_list, addr_node) {
list_del(&pos->addr_node);
kmem_cache_free(xsk_tx_generic_cache, pos);
}
skb_ext_del(skb, SKB_EXT_XDP);
}

skb->destructor = sock_wfree;
Expand Down Expand Up @@ -727,6 +730,18 @@ static struct sk_buff *xsk_build_skb_zerocopy(struct xdp_sock *xs,
return ERR_PTR(err);
}
} else {
struct xdp_skb_ext *ext;

ext = skb_ext_find(skb, SKB_EXT_XDP);
if (!ext) {
ext = skb_ext_add(skb, SKB_EXT_XDP);
if (!ext)
return ERR_PTR(-ENOMEM);
memset(ext, 0, sizeof(*ext));
INIT_LIST_HEAD(&ext->addrs_list);
ext->num_descs = 1;
}

xsk_addr = kmem_cache_zalloc(xsk_tx_generic_cache, GFP_KERNEL);
if (!xsk_addr)
return ERR_PTR(-ENOMEM);
Expand All @@ -736,7 +751,8 @@ static struct sk_buff *xsk_build_skb_zerocopy(struct xdp_sock *xs,
* would be dropped, which implies freeing all list elements
*/
xsk_addr->addr = desc->addr;
list_add_tail(&xsk_addr->addr_node, &XSKCB(skb)->addrs_list);
list_add_tail(&xsk_addr->addr_node, &ext->addrs_list);
xsk_inc_num_desc(skb);
}

len = desc->len;
Expand Down Expand Up @@ -814,6 +830,7 @@ static struct sk_buff *xsk_build_skb(struct xdp_sock *xs,
} else {
int nr_frags = skb_shinfo(skb)->nr_frags;
struct xsk_addr_node *xsk_addr;
struct xdp_skb_ext *ext;
struct page *page;
u8 *vaddr;

Expand All @@ -828,6 +845,19 @@ static struct sk_buff *xsk_build_skb(struct xdp_sock *xs,
goto free_err;
}

ext = skb_ext_find(skb, SKB_EXT_XDP);
if (!ext) {
ext = skb_ext_add(skb, SKB_EXT_XDP);
if (!ext) {
__free_page(page);
err = -ENOMEM;
goto free_err;
}
memset(ext, 0, sizeof(*ext));
INIT_LIST_HEAD(&ext->addrs_list);
ext->num_descs = 1;
}

xsk_addr = kmem_cache_zalloc(xsk_tx_generic_cache, GFP_KERNEL);
if (!xsk_addr) {
__free_page(page);
Expand All @@ -843,12 +873,11 @@ static struct sk_buff *xsk_build_skb(struct xdp_sock *xs,
refcount_add(PAGE_SIZE, &xs->sk.sk_wmem_alloc);

xsk_addr->addr = desc->addr;
list_add_tail(&xsk_addr->addr_node, &XSKCB(skb)->addrs_list);
list_add_tail(&xsk_addr->addr_node, &ext->addrs_list);
xsk_inc_num_desc(skb);
}
}

xsk_inc_num_desc(skb);

return skb;

free_err:
Expand All @@ -857,7 +886,6 @@ static struct sk_buff *xsk_build_skb(struct xdp_sock *xs,

if (err == -EOVERFLOW) {
/* Drop the packet */
xsk_inc_num_desc(xs->skb);
xsk_drop_skb(xs->skb);
xskq_cons_release(xs->tx);
} else {
Expand Down
Loading