drivers/net/ethernet/ti/icssg/icssg_common.c | 5 +++++ 1 file changed, 5 insertions(+)
Page recycling was removed from the XDP_DROP path in emac_run_xdp() to
avoid conflicts with AF_XDP zero-copy mode, which uses xsk_buff_free()
instead.
However, this causes a memory leak when running XDP programs that drop
packets in non-zero-copy mode (standard page pool mode). The pages are
never returned to the page pool, leading to OOM conditions.
Fix this by handling cleanup in the caller, emac_rx_packet().
When emac_run_xdp() returns ICSSG_XDP_CONSUMED for XDP_DROP, the
caller now recycles the page back to the page pool. The zero-copy
path, emac_rx_packet_zc() already handles cleanup correctly with
xsk_buff_free().
Fixes: 7a64bb388df3 ("net: ti: icssg-prueth: Add AF_XDP zero copy for RX")
Signed-off-by: Meghana Malladi <m-malladi@ti.com>
---
drivers/net/ethernet/ti/icssg/icssg_common.c | 5 +++++
1 file changed, 5 insertions(+)
diff --git a/drivers/net/ethernet/ti/icssg/icssg_common.c b/drivers/net/ethernet/ti/icssg/icssg_common.c
index 0cf9dfe0fa36..0a3cf2f848a5 100644
--- a/drivers/net/ethernet/ti/icssg/icssg_common.c
+++ b/drivers/net/ethernet/ti/icssg/icssg_common.c
@@ -1075,6 +1075,11 @@ static int emac_rx_packet(struct prueth_emac *emac, u32 flow_id, u32 *xdp_state)
xdp_prepare_buff(&xdp, pa, PRUETH_HEADROOM, pkt_len, false);
*xdp_state = emac_run_xdp(emac, &xdp, &pkt_len);
+ if (*xdp_state == ICSSG_XDP_CONSUMED) {
+ page_pool_recycle_direct(pool, page);
+ goto requeue;
+ }
+
if (*xdp_state != ICSSG_XDP_PASS)
goto requeue;
headroom = xdp.data - xdp.data_hard_start;
base-commit: 908c344d5cfac4160f49715da9efacdf5b6a28bd
--
2.43.0
On Wed, Mar 11, 2026 at 03:24:41PM +0530, Meghana Malladi wrote:
> Page recycling was removed from the XDP_DROP path in emac_run_xdp() to
> avoid conflicts with AF_XDP zero-copy mode, which uses xsk_buff_free()
> instead.
>
> However, this causes a memory leak when running XDP programs that drop
> packets in non-zero-copy mode (standard page pool mode). The pages are
> never returned to the page pool, leading to OOM conditions.
>
> Fix this by handling cleanup in the caller, emac_rx_packet().
> When emac_run_xdp() returns ICSSG_XDP_CONSUMED for XDP_DROP, the
> caller now recycles the page back to the page pool. The zero-copy
> path, emac_rx_packet_zc() already handles cleanup correctly with
> xsk_buff_free().
>
> Fixes: 7a64bb388df3 ("net: ti: icssg-prueth: Add AF_XDP zero copy for RX")
> Signed-off-by: Meghana Malladi <m-malladi@ti.com>
Reviewed-by: Simon Horman <horms@kernel.org>
© 2016 - 2026 Red Hat, Inc.