Add XDP support for the PCIe FDMA path. The PCIe XDP implementation
operates on contiguous ATU-mapped buffers with memcpy-based XDP_TX
transmit, unlike the platform path which uses page_pool.
Only XDP_ACT_BASIC (XDP_PASS, XDP_DROP, XDP_TX, XDP_ABORTED) is
supported; XDP_REDIRECT and NDO_XMIT are not available on the PCIe path,
as they, to my knowledge, require page_pool-backed buffers.
Signed-off-by: Daniel Machon <daniel.machon@microchip.com>
---
.../ethernet/microchip/lan966x/lan966x_fdma_pci.c | 136 ++++++++++++++++++---
.../net/ethernet/microchip/lan966x/lan966x_main.c | 11 +-
.../net/ethernet/microchip/lan966x/lan966x_main.h | 10 ++
.../net/ethernet/microchip/lan966x/lan966x_xdp.c | 6 +
4 files changed, 140 insertions(+), 23 deletions(-)
diff --git a/drivers/net/ethernet/microchip/lan966x/lan966x_fdma_pci.c b/drivers/net/ethernet/microchip/lan966x/lan966x_fdma_pci.c
index 4d69beb41c0c..f2c8c6aa3d4f 100644
--- a/drivers/net/ethernet/microchip/lan966x/lan966x_fdma_pci.c
+++ b/drivers/net/ethernet/microchip/lan966x/lan966x_fdma_pci.c
@@ -1,5 +1,7 @@
// SPDX-License-Identifier: GPL-2.0+
+#include <linux/bpf_trace.h>
+
#include "fdma_api.h"
#include "lan966x_main.h"
@@ -68,10 +70,110 @@ static int lan966x_fdma_pci_tx_alloc(struct lan966x_tx *tx)
return 0;
}
+static int lan966x_fdma_pci_get_next_dcb(struct fdma *fdma)
+{
+ struct fdma_db *db;
+
+ for (int i = 0; i < fdma->n_dcbs; i++) {
+ db = fdma_db_get(fdma, i, 0);
+
+ if (!fdma_db_is_done(db))
+ continue;
+ if (fdma_is_last(fdma, &fdma->dcbs[i]))
+ continue;
+
+ return i;
+ }
+
+ return -1;
+}
+
+static int lan966x_fdma_pci_xmit_xdpf(struct lan966x_port *port,
+ void *ptr, u32 len)
+{
+ struct lan966x *lan966x = port->lan966x;
+ struct lan966x_tx *tx = &lan966x->tx;
+ struct fdma *fdma = &tx->fdma;
+ int next_to_use, ret = 0;
+ void *virt_addr;
+ __be32 *ifh;
+
+ spin_lock(&lan966x->tx_lock);
+
+ next_to_use = lan966x_fdma_pci_get_next_dcb(fdma);
+
+ if (next_to_use < 0) {
+ netif_stop_queue(port->dev);
+ ret = NETDEV_TX_BUSY;
+ goto out;
+ }
+
+ ifh = ptr;
+ memset(ifh, 0, IFH_LEN_BYTES);
+ lan966x_ifh_set_bypass(ifh, 1);
+ lan966x_ifh_set_port(ifh, BIT_ULL(port->chip_port));
+
+ /* Get the virtual addr of the next DB and copy frame to it. */
+ virt_addr = fdma_dataptr_virt_addr_contiguous(fdma, next_to_use, 0);
+ memcpy(virt_addr, ptr, len);
+
+ fdma_dcb_add(fdma,
+ next_to_use,
+ 0,
+ FDMA_DCB_STATUS_SOF |
+ FDMA_DCB_STATUS_EOF |
+ FDMA_DCB_STATUS_BLOCKO(0) |
+ FDMA_DCB_STATUS_BLOCKL(len));
+
+ /* Start the transmission. */
+ lan966x_fdma_tx_start(tx);
+
+out:
+ spin_unlock(&lan966x->tx_lock);
+
+ return ret;
+}
+
+static int lan966x_xdp_pci_run(struct lan966x_port *port, void *data,
+ u32 data_len)
+{
+ struct bpf_prog *xdp_prog = port->xdp_prog;
+ struct lan966x *lan966x = port->lan966x;
+ struct xdp_buff xdp;
+ u32 act;
+
+ xdp_init_buff(&xdp, lan966x->rx.max_mtu, &port->xdp_rxq);
+
+ xdp_prepare_buff(&xdp,
+ data - XDP_PACKET_HEADROOM,
+ IFH_LEN_BYTES + XDP_PACKET_HEADROOM,
+ data_len - IFH_LEN_BYTES,
+ false);
+
+ act = bpf_prog_run_xdp(xdp_prog, &xdp);
+ switch (act) {
+ case XDP_PASS:
+ return FDMA_PASS;
+ case XDP_TX:
+ return lan966x_fdma_pci_xmit_xdpf(port, data, data_len) ?
+ FDMA_DROP : FDMA_TX;
+ default:
+ bpf_warn_invalid_xdp_action(port->dev, xdp_prog, act);
+ fallthrough;
+ case XDP_ABORTED:
+ trace_xdp_exception(port->dev, xdp_prog, act);
+ fallthrough;
+ case XDP_DROP:
+ return FDMA_DROP;
+ }
+}
+
static int lan966x_fdma_pci_rx_check_frame(struct lan966x_rx *rx, u64 *src_port)
{
struct lan966x *lan966x = rx->lan966x;
struct fdma *fdma = &rx->fdma;
+ struct lan966x_port *port;
+ struct fdma_db *db;
void *virt_addr;
virt_addr = fdma_dataptr_virt_addr_contiguous(fdma,
@@ -83,7 +185,15 @@ static int lan966x_fdma_pci_rx_check_frame(struct lan966x_rx *rx, u64 *src_port)
if (WARN_ON(*src_port >= lan966x->num_phys_ports))
return FDMA_ERROR;
- return FDMA_PASS;
+ port = lan966x->ports[*src_port];
+ if (!lan966x_xdp_port_present(port))
+ return FDMA_PASS;
+
+ db = fdma_db_next_get(fdma);
+
+ return lan966x_xdp_pci_run(port,
+ virt_addr,
+ FDMA_DCB_STATUS_BLOCKL(db->status));
}
static struct sk_buff *lan966x_fdma_pci_rx_get_frame(struct lan966x_rx *rx,
@@ -133,24 +243,6 @@ static struct sk_buff *lan966x_fdma_pci_rx_get_frame(struct lan966x_rx *rx,
return skb;
}
-static int lan966x_fdma_pci_get_next_dcb(struct fdma *fdma)
-{
- struct fdma_db *db;
-
- for (int i = 0; i < fdma->n_dcbs; i++) {
- db = fdma_db_get(fdma, i, 0);
-
- if (!fdma_db_is_done(db))
- continue;
- if (fdma_is_last(fdma, &fdma->dcbs[i]))
- continue;
-
- return i;
- }
-
- return -1;
-}
-
static int lan966x_fdma_pci_xmit(struct sk_buff *skb, __be32 *ifh,
struct net_device *dev)
{
@@ -225,6 +317,12 @@ static int lan966x_fdma_pci_napi_poll(struct napi_struct *napi, int weight)
case FDMA_ERROR:
fdma_dcb_advance(fdma);
goto allocate_new;
+ case FDMA_TX:
+ fdma_dcb_advance(fdma);
+ continue;
+ case FDMA_DROP:
+ fdma_dcb_advance(fdma);
+ continue;
}
skb = lan966x_fdma_pci_rx_get_frame(rx, src_port);
fdma_dcb_advance(fdma);
diff --git a/drivers/net/ethernet/microchip/lan966x/lan966x_main.c b/drivers/net/ethernet/microchip/lan966x/lan966x_main.c
index fc14738774ec..b42e044da735 100644
--- a/drivers/net/ethernet/microchip/lan966x/lan966x_main.c
+++ b/drivers/net/ethernet/microchip/lan966x/lan966x_main.c
@@ -877,10 +877,13 @@ static int lan966x_probe_port(struct lan966x *lan966x, u32 p,
port->phylink = phylink;
- if (lan966x->fdma)
- dev->xdp_features = NETDEV_XDP_ACT_BASIC |
- NETDEV_XDP_ACT_REDIRECT |
- NETDEV_XDP_ACT_NDO_XMIT;
+ if (lan966x->fdma) {
+ dev->xdp_features = NETDEV_XDP_ACT_BASIC;
+
+ if (!lan966x_is_pci(lan966x))
+ dev->xdp_features |= NETDEV_XDP_ACT_REDIRECT |
+ NETDEV_XDP_ACT_NDO_XMIT;
+ }
err = register_netdev(dev);
if (err) {
diff --git a/drivers/net/ethernet/microchip/lan966x/lan966x_main.h b/drivers/net/ethernet/microchip/lan966x/lan966x_main.h
index 8fcc51133417..2491fe937e36 100644
--- a/drivers/net/ethernet/microchip/lan966x/lan966x_main.h
+++ b/drivers/net/ethernet/microchip/lan966x/lan966x_main.h
@@ -596,6 +596,16 @@ int lan966x_qsys_sw_status(struct lan966x *lan966x);
#if IS_ENABLED(CONFIG_MCHP_LAN966X_PCI)
extern const struct lan966x_fdma_ops lan966x_fdma_pci_ops;
+
+static inline bool lan966x_is_pci(struct lan966x *lan966x)
+{
+ return lan966x->ops == &lan966x_fdma_pci_ops;
+}
+#else
+static inline bool lan966x_is_pci(struct lan966x *lan966x)
+{
+ return false;
+}
#endif
int lan966x_lag_port_join(struct lan966x_port *port,
diff --git a/drivers/net/ethernet/microchip/lan966x/lan966x_xdp.c b/drivers/net/ethernet/microchip/lan966x/lan966x_xdp.c
index 9ee61db8690b..9b3356ba6ba8 100644
--- a/drivers/net/ethernet/microchip/lan966x/lan966x_xdp.c
+++ b/drivers/net/ethernet/microchip/lan966x/lan966x_xdp.c
@@ -27,6 +27,12 @@ static int lan966x_xdp_setup(struct net_device *dev, struct netdev_bpf *xdp)
if (old_xdp == new_xdp)
goto out;
+ /* PCIe FDMA uses contiguous buffers, so no page_pool reload
+ * is needed.
+ */
+ if (lan966x_is_pci(lan966x))
+ goto out;
+
err = lan966x_fdma_reload_page_pool(lan966x);
if (err) {
xchg(&port->xdp_prog, old_prog);
--
2.34.1
> +static int lan966x_xdp_pci_run(struct lan966x_port *port, void *data,
> + u32 data_len)
> +{
> + struct bpf_prog *xdp_prog = port->xdp_prog;
> + struct lan966x *lan966x = port->lan966x;
> + struct xdp_buff xdp;
> + u32 act;
> +
> + xdp_init_buff(&xdp, lan966x->rx.max_mtu, &port->xdp_rxq);
> +
> + xdp_prepare_buff(&xdp,
> + data - XDP_PACKET_HEADROOM,
> + IFH_LEN_BYTES + XDP_PACKET_HEADROOM,
> + data_len - IFH_LEN_BYTES,
> + false);
> +
> + act = bpf_prog_run_xdp(xdp_prog, &xdp);
> + switch (act) {
> + case XDP_PASS:
> + return FDMA_PASS;
> + case XDP_TX:
> + return lan966x_fdma_pci_xmit_xdpf(port, data, data_len) ?
> + FDMA_DROP : FDMA_TX;
What if the BPF program modifies packet boundaries (e.g.,
headroom/tailroom adjustment)? After bpf_prog_run_xdp(), xdp.data and
xdp.data_end may differ from the original data and data_len, but
lan966x_fdma_pci_xmit_xdpf() is called with the original values.
Wouldn't any adjustments made by the XDP program be silently lost?
> + default:
> + bpf_warn_invalid_xdp_action(port->dev, xdp_prog, act);
> + fallthrough;
> + case XDP_ABORTED:
> + trace_xdp_exception(port->dev, xdp_prog, act);
> + fallthrough;
> + case XDP_DROP:
> + return FDMA_DROP;
> + }
> +}
> +
Hi Mohsin,
> > +static int lan966x_xdp_pci_run(struct lan966x_port *port, void *data,
> > + u32 data_len)
> > +{
> > + struct bpf_prog *xdp_prog = port->xdp_prog;
> > + struct lan966x *lan966x = port->lan966x;
> > + struct xdp_buff xdp;
> > + u32 act;
> > +
> > + xdp_init_buff(&xdp, lan966x->rx.max_mtu, &port->xdp_rxq);
> > +
> > + xdp_prepare_buff(&xdp,
> > + data - XDP_PACKET_HEADROOM,
> > + IFH_LEN_BYTES + XDP_PACKET_HEADROOM,
> > + data_len - IFH_LEN_BYTES,
> > + false);
> > +
> > + act = bpf_prog_run_xdp(xdp_prog, &xdp);
> > + switch (act) {
> > + case XDP_PASS:
> > + return FDMA_PASS;
> > + case XDP_TX:
> > + return lan966x_fdma_pci_xmit_xdpf(port, data, data_len) ?
> > + FDMA_DROP : FDMA_TX;
>
> What if the BPF program modifies packet boundaries (e.g.,
> headroom/tailroom adjustment)? After bpf_prog_run_xdp(), xdp.data and
> xdp.data_end may differ from the original data and data_len, but
> lan966x_fdma_pci_xmit_xdpf() is called with the original values.
> Wouldn't any adjustments made by the XDP program be silently lost?
>
I think you are absolutely right. Thanks for catching this. Actually, we might
have the same issue in the platform XDP path also.
I will fix this in v2.
> > + default:
> > + bpf_warn_invalid_xdp_action(port->dev, xdp_prog, act);
> > + fallthrough;
> > + case XDP_ABORTED:
> > + trace_xdp_exception(port->dev, xdp_prog, act);
> > + fallthrough;
> > + case XDP_DROP:
> > + return FDMA_DROP;
> > + }
> > +}
> > +
/Daniel
© 2016 - 2026 Red Hat, Inc.