Message ID | 20220317185159.1661469-1-horatiu.vultur@microchip.com (mailing list archive) |
---|---|
Headers | show |
Series | net: lan966x: Add support for FDMA | expand |
Hi Horatiu, > Currently when injecting or extracting a frame from CPU, the frame > is given to the HW each word at a time. There is another way to > inject/extract frames from CPU using FDMA(Frame Direct Memory Access). > In this way the entire frame is given to the HW. This improves both > RX and TX bitrate. I wanted to test this. ping and such works fine and I'm also seeing fdma interrupts. But as soon as I try iperf3 I get a skb_panic (due to frame size?). Hope that splash below helps. -michael [ 159.778850] skbuff: skb_over_panic: text:c07960c0 len:106 put:4 head:c8cc6e00 data:c8cc6e9a tail:0xc8cc6f04 end:0xc8cc6f00 dev:eth0 [ 159.788067] ------------[ cut here ]------------ [ 159.792575] kernel BUG at net/core/skbuff.c:113! .. [ 160.620995] Backtrace: [ 160.623426] [<c0a62350>] (skb_panic) from [<c08876f8>] (skb_put+0x54/0x58) [ 160.630284] [<c08876a4>] (skb_put) from [<c07960c0>] (lan966x_fdma_xmit+0x108/0x504) [ 160.638011] r5:00000000 r4:00000000 [ 160.641566] [<c0795fb8>] (lan966x_fdma_xmit) from [<c078cdf8>] (lan966x_port_xmit+0x1b0/0x450) [ 160.650171] r10:c1960160 r9:c1960040 r8:00000001 r7:0000001c r6:c196b000 r5:c891a780 [ 160.657973] r4:00000000 [ 160.660488] [<c078cc48>] (lan966x_port_xmit) from [<c08a3e24>] (dev_hard_start_xmit+0x114/0x248) [ 160.669267] r10:c0f016e8 r9:c0f060cc r8:c0e8d350 r7:c196ba00 r6:00000000 r5:c196b000 [ 160.677069] r4:c891a780 [ 160.679584] [<c08a3d10>] (dev_hard_start_xmit) from [<c0904728>] (sch_direct_xmit+0x11c/0x31c) [ 160.688188] r10:2a01000a r9:00000000 r8:00000000 r7:c196b000 r6:c196ba00 r5:c891a780 [ 160.695991] r4:c1e57400 [ 160.698507] [<c090460c>] (sch_direct_xmit) from [<c08a4520>] (__dev_queue_xmit+0x508/0xaac) [ 160.706850] r9:c8cc6f00 r8:0000004a r7:00000001 r6:00000000 r5:c1e57400 r4:c891a780 [ 160.714565] [<c08a4018>] (__dev_queue_xmit) from [<c08a4ad8>] (dev_queue_xmit+0x14/0x18) [ 160.722648] r10:2a01000a r9:00000010 r8:0000004a r7:c1e7038c r6:00000000 r5:c891a780 [ 160.730451] r4:c1e70300 [ 160.732966] [<c08a4ac4>] (dev_queue_xmit) from [<c092b0c0>] (ip_finish_output2+0x270/0x5dc) [ 160.741300] [<c092ae50>] (ip_finish_output2) from [<c092ceb8>] (__ip_finish_output+0x9c/0x144) [ 160.749903] r9:c8cdef00 r8:c10833c0 r7:000005dc r6:c8924000 r5:c891a780 r4:c891a780 [ 160.757618] [<c092ce1c>] (__ip_finish_output) from [<c092d01c>] (ip_output+0xbc/0xc4) [ 160.765441] r10:2a01000a r9:c8cdef00 r8:c8cc6e00 r7:c196b000 r6:c8924000 r5:c10833c0 [ 160.773244] r4:c891a780 [ 160.775759] [<c092cf60>] (ip_output) from [<c092ab10>] (ip_local_out+0x60/0x6c) [ 160.783058] r7:00000000 r6:c8924000 r5:c10833c0 r4:c891a780 [ 160.788692] [<c092aab0>] (ip_local_out) from [<c092ad74>] (ip_build_and_send_pkt+0x110/0x1ec) [ 160.797206] r7:00000000 r6:c8d40000 r5:000000c4 r4:c891a780 [ 160.802841] [<c092ac64>] (ip_build_and_send_pkt) from [<c0952bb8>] (tcp_v4_send_synack+0xec/0x138) [ 160.811792] r10:c8cdef00 r9:c0b9b614 r8:c10833c0 r7:8b01000a r6:2a01000a r5:c8d40000 [ 160.819595] r4:c8924000 [ 160.822110] [<c0952acc>] (tcp_v4_send_synack) from [<c0941520>] (tcp_conn_request+0x6cc/0x984) [ 160.830711] r7:c891ab40 r6:c0952acc r5:c8d40000 r4:c8924000 [ 160.836345] [<c0940e54>] (tcp_conn_request) from [<c095299c>] (tcp_v4_conn_request+0x38/0x84) [ 160.844863] r10:c892902a r9:c892903e r8:c10833c0 r7:00000000 r6:00000000 r5:c891ab40 [ 160.852666] r4:c8d40000 [ 160.855181] [<c0952964>] (tcp_v4_conn_request) from [<c09d05a0>] (tcp_v6_conn_request+0x12c/0x138) [ 160.864122] [<c09d0474>] (tcp_v6_conn_request) from [<c0948870>] (tcp_rcv_state_process+0x29c/0xf44) [ 160.873240] r5:c891ab40 r4:c8d40000 [ 160.876794] [<c09485d4>] (tcp_rcv_state_process) from [<c0953374>] (tcp_v4_do_rcv+0x110/0x22c) [ 160.885398] r10:c892902a r9:c892903e r8:c10833c0 r7:00000000 r6:c8d40000 r5:c891ab40 [ 160.893201] r4:c8d40000 [ 160.895716] [<c0953264>] (tcp_v4_do_rcv) from [<c0955158>] (tcp_v4_rcv+0xb08/0xba8) [ 160.903362] r7:00000000 r6:c8d40000 r5:00000000 r4:c891ab40 [ 160.908998] [<c0954650>] (tcp_v4_rcv) from [<c09273f4>] (ip_protocol_deliver_rcu+0x34/0x1d8) [ 160.917428] r10:c10833c0 r9:c891ab40 r8:00000000 r7:c10833c0 r6:c0f06f64 r5:c891ab40 [ 160.925230] r4:00000006 [ 160.927746] [<c09273c0>] (ip_protocol_deliver_rcu) from [<c0927640>] (ip_local_deliver+0xa8/0xf0) [ 160.936609] r8:c0f01c0c r7:c0f01c0c r6:00000000 r5:c0f01bc4 r4:c891ab40 [ 160.943283] [<c0927598>] (ip_local_deliver) from [<c0926794>] (ip_sublist_rcv_finish+0x44/0x58) [ 160.951967] r5:c0f01bc4 r4:c0f01bc4 [ 160.955522] [<c0926750>] (ip_sublist_rcv_finish) from [<c0926d58>] (ip_list_rcv_finish.constprop.0+0x10c/0x134) [ 160.965598] r7:c0f01c0c r6:c0f01bc4 r5:c0f01c0c r4:c0f01c0c [ 160.971232] [<c0926c4c>] (ip_list_rcv_finish.constprop.0) from [<c09277cc>] (ip_list_rcv+0xec/0x114) [ 160.980358] r10:c0f01c5c r9:c10833c0 r8:c0f01c0c r7:c0f01c5c r6:c10833c0 r5:c196b000 [ 160.988160] r4:c891ab40 [ 160.990676] [<c09276e0>] (ip_list_rcv) from [<c08a6328>] (__netif_receive_skb_list_core+0x90/0x214) [ 160.999714] r10:c196b000 r9:c196b000 r8:00000000 r7:c0f01c5c r6:c0f07450 r5:c891acc0 [ 161.007517] r4:c09276e0 [ 161.010032] [<c08a6298>] (__netif_receive_skb_list_core) from [<c08a6670>] (netif_receive_skb_list_internal+0x1c4/0x2c4) [ 161.020893] r10:c10833b8 r9:c0f08a00 r8:00000000 r7:c0f01d44 r6:c0f01d44 r5:c0f01d44 [ 161.028696] r4:00000000 [ 161.031211] [<c08a64ac>] (netif_receive_skb_list_internal) from [<c08a67a8>] (netif_receive_skb_list+0x38/0xe4) [ 161.041291] r10:c1963de0 r9:f08527d0 r8:f08527c0 r7:00000003 r6:c1960040 r5:c103d920 [ 161.049094] r4:c0f01d44 [ 161.051609] [<c08a6770>] (netif_receive_skb_list) from [<c0795af4>] (lan966x_fdma_napi_poll+0x3ec/0x490) [ 161.061081] r9:f08527d0 r8:f08527c0 r7:00000003 r6:c1960040 r5:c1962de0 r4:00000040 [ 161.068796] [<c0795708>] (lan966x_fdma_napi_poll) from [<c08a6ee4>] (__napi_poll+0x34/0x1dc) [ 161.077226] r10:ef7e6080 r9:c0f01db4 r8:c0f03d00 r7:c0f01db3 r6:00000040 r5:c1963de0 [ 161.085029] r4:00000001 [ 161.087545] [<c08a6eb0>] (__napi_poll) from [<c08a7290>] (net_rx_action+0xec/0x288) [ 161.095190] r7:2e957000 r6:ffffc936 r5:0000012c r4:c1963de0 [ 161.100825] [<c08a71a4>] (net_rx_action) from [<c010143c>] (__do_softirq+0x13c/0x384) [ 161.108647] r10:c0f08a00 r9:00000008 r8:00000100 r7:c1037d40 r6:00000003 r5:00000004 [ 161.116451] r4:c0f0308c [ 161.118966] [<c0101300>] (__do_softirq) from [<c013bc9c>] (irq_exit+0xac/0xdc) [ 161.126182] r10:c0f04f78 r9:c0f08a00 r8:00000000 r7:c0f01ed4 r6:00000000 r5:c0f01ea0 [ 161.133985] r4:c0f08a00 [ 161.136500] [<c013bbf0>] (irq_exit) from [<c0a62c84>] (generic_handle_arch_irq+0x48/0x4c) [ 161.144663] r5:c0f01ea0 r4:c0e8d358 [ 161.148218] [<c0a62c3c>] (generic_handle_arch_irq) from [<c0100be8>] (__irq_svc+0x88/0xb0)
The 03/18/2022 12:07, Michael Walle wrote: > > Hi Horatiu, Hi Michael, > > > Currently when injecting or extracting a frame from CPU, the frame > > is given to the HW each word at a time. There is another way to > > inject/extract frames from CPU using FDMA(Frame Direct Memory Access). > > In this way the entire frame is given to the HW. This improves both > > RX and TX bitrate. > > I wanted to test this. ping and such works fine and I'm also > seeing fdma interrupts. Thanks for testing this also on your board. > But as soon as I try iperf3 I get a skb_panic > (due to frame size?). Hope that splash below helps. I have not seen this issue. But it looks like it is a problem that there is no more space to add the FCS. Can you tell me how you run iperf3 so I can also try it? Also I have a small diff that might fix the issue: --- --- a/drivers/net/ethernet/microchip/lan966x/lan966x_fdma.c +++ b/drivers/net/ethernet/microchip/lan966x/lan966x_fdma.c @@ -534,6 +534,8 @@ int lan966x_fdma_xmit(struct sk_buff *skb, __be32 *ifh, struct net_device *dev) struct lan966x_tx_dcb *next_dcb, *dcb; struct lan966x_tx *tx = &lan966x->tx; struct lan966x_db *next_db; + int needed_headroom; + int needed_tailroom; dma_addr_t dma_addr; int next_to_use; int err; @@ -554,10 +556,11 @@ int lan966x_fdma_xmit(struct sk_buff *skb, __be32 *ifh, struct net_device *dev) /* skb processing */ skb_tx_timestamp(skb); - if (skb_headroom(skb) < IFH_LEN * sizeof(u32)) { - err = pskb_expand_head(skb, - IFH_LEN * sizeof(u32) - skb_headroom(skb), - 0, GFP_ATOMIC); + needed_headroom = max_t(int, IFH_LEN * sizeof(u32) - skb_headroom(skb), 0); + needed_tailroom = max_t(int, ETH_FCS_LEN - skb_tailroom(skb), 0); + if (needed_headroom || needed_tailroom) { + err = pskb_expand_head(skb, needed_headroom, needed_tailroom, + GFP_ATOMIC); if (unlikely(err)) { dev->stats.tx_dropped++; err = NETDEV_TX_OK; --- > > -michael > > [ 159.778850] skbuff: skb_over_panic: text:c07960c0 len:106 put:4 head:c8cc6e00 data:c8cc6e9a tail:0xc8cc6f04 end:0xc8cc6f00 dev:eth0 > [ 159.788067] ------------[ cut here ]------------ > [ 159.792575] kernel BUG at net/core/skbuff.c:113! > .. > [ 160.620995] Backtrace: > [ 160.623426] [<c0a62350>] (skb_panic) from [<c08876f8>] (skb_put+0x54/0x58) > [ 160.630284] [<c08876a4>] (skb_put) from [<c07960c0>] (lan966x_fdma_xmit+0x108/0x504) > [ 160.638011] r5:00000000 r4:00000000 > [ 160.641566] [<c0795fb8>] (lan966x_fdma_xmit) from [<c078cdf8>] (lan966x_port_xmit+0x1b0/0x450) > [ 160.650171] r10:c1960160 r9:c1960040 r8:00000001 r7:0000001c r6:c196b000 r5:c891a780 > [ 160.657973] r4:00000000 > [ 160.660488] [<c078cc48>] (lan966x_port_xmit) from [<c08a3e24>] (dev_hard_start_xmit+0x114/0x248) > [ 160.669267] r10:c0f016e8 r9:c0f060cc r8:c0e8d350 r7:c196ba00 r6:00000000 r5:c196b000 > [ 160.677069] r4:c891a780 > [ 160.679584] [<c08a3d10>] (dev_hard_start_xmit) from [<c0904728>] (sch_direct_xmit+0x11c/0x31c) > [ 160.688188] r10:2a01000a r9:00000000 r8:00000000 r7:c196b000 r6:c196ba00 r5:c891a780 > [ 160.695991] r4:c1e57400 > [ 160.698507] [<c090460c>] (sch_direct_xmit) from [<c08a4520>] (__dev_queue_xmit+0x508/0xaac) > [ 160.706850] r9:c8cc6f00 r8:0000004a r7:00000001 r6:00000000 r5:c1e57400 r4:c891a780 > [ 160.714565] [<c08a4018>] (__dev_queue_xmit) from [<c08a4ad8>] (dev_queue_xmit+0x14/0x18) > [ 160.722648] r10:2a01000a r9:00000010 r8:0000004a r7:c1e7038c r6:00000000 r5:c891a780 > [ 160.730451] r4:c1e70300 > [ 160.732966] [<c08a4ac4>] (dev_queue_xmit) from [<c092b0c0>] (ip_finish_output2+0x270/0x5dc) > [ 160.741300] [<c092ae50>] (ip_finish_output2) from [<c092ceb8>] (__ip_finish_output+0x9c/0x144) > [ 160.749903] r9:c8cdef00 r8:c10833c0 r7:000005dc r6:c8924000 r5:c891a780 r4:c891a780 > [ 160.757618] [<c092ce1c>] (__ip_finish_output) from [<c092d01c>] (ip_output+0xbc/0xc4) > [ 160.765441] r10:2a01000a r9:c8cdef00 r8:c8cc6e00 r7:c196b000 r6:c8924000 r5:c10833c0 > [ 160.773244] r4:c891a780 > [ 160.775759] [<c092cf60>] (ip_output) from [<c092ab10>] (ip_local_out+0x60/0x6c) > [ 160.783058] r7:00000000 r6:c8924000 r5:c10833c0 r4:c891a780 > [ 160.788692] [<c092aab0>] (ip_local_out) from [<c092ad74>] (ip_build_and_send_pkt+0x110/0x1ec) > [ 160.797206] r7:00000000 r6:c8d40000 r5:000000c4 r4:c891a780 > [ 160.802841] [<c092ac64>] (ip_build_and_send_pkt) from [<c0952bb8>] (tcp_v4_send_synack+0xec/0x138) > [ 160.811792] r10:c8cdef00 r9:c0b9b614 r8:c10833c0 r7:8b01000a r6:2a01000a r5:c8d40000 > [ 160.819595] r4:c8924000 > [ 160.822110] [<c0952acc>] (tcp_v4_send_synack) from [<c0941520>] (tcp_conn_request+0x6cc/0x984) > [ 160.830711] r7:c891ab40 r6:c0952acc r5:c8d40000 r4:c8924000 > [ 160.836345] [<c0940e54>] (tcp_conn_request) from [<c095299c>] (tcp_v4_conn_request+0x38/0x84) > [ 160.844863] r10:c892902a r9:c892903e r8:c10833c0 r7:00000000 r6:00000000 r5:c891ab40 > [ 160.852666] r4:c8d40000 > [ 160.855181] [<c0952964>] (tcp_v4_conn_request) from [<c09d05a0>] (tcp_v6_conn_request+0x12c/0x138) > [ 160.864122] [<c09d0474>] (tcp_v6_conn_request) from [<c0948870>] (tcp_rcv_state_process+0x29c/0xf44) > [ 160.873240] r5:c891ab40 r4:c8d40000 > [ 160.876794] [<c09485d4>] (tcp_rcv_state_process) from [<c0953374>] (tcp_v4_do_rcv+0x110/0x22c) > [ 160.885398] r10:c892902a r9:c892903e r8:c10833c0 r7:00000000 r6:c8d40000 r5:c891ab40 > [ 160.893201] r4:c8d40000 > [ 160.895716] [<c0953264>] (tcp_v4_do_rcv) from [<c0955158>] (tcp_v4_rcv+0xb08/0xba8) > [ 160.903362] r7:00000000 r6:c8d40000 r5:00000000 r4:c891ab40 > [ 160.908998] [<c0954650>] (tcp_v4_rcv) from [<c09273f4>] (ip_protocol_deliver_rcu+0x34/0x1d8) > [ 160.917428] r10:c10833c0 r9:c891ab40 r8:00000000 r7:c10833c0 r6:c0f06f64 r5:c891ab40 > [ 160.925230] r4:00000006 > [ 160.927746] [<c09273c0>] (ip_protocol_deliver_rcu) from [<c0927640>] (ip_local_deliver+0xa8/0xf0) > [ 160.936609] r8:c0f01c0c r7:c0f01c0c r6:00000000 r5:c0f01bc4 r4:c891ab40 > [ 160.943283] [<c0927598>] (ip_local_deliver) from [<c0926794>] (ip_sublist_rcv_finish+0x44/0x58) > [ 160.951967] r5:c0f01bc4 r4:c0f01bc4 > [ 160.955522] [<c0926750>] (ip_sublist_rcv_finish) from [<c0926d58>] (ip_list_rcv_finish.constprop.0+0x10c/0x134) > [ 160.965598] r7:c0f01c0c r6:c0f01bc4 r5:c0f01c0c r4:c0f01c0c > [ 160.971232] [<c0926c4c>] (ip_list_rcv_finish.constprop.0) from [<c09277cc>] (ip_list_rcv+0xec/0x114) > [ 160.980358] r10:c0f01c5c r9:c10833c0 r8:c0f01c0c r7:c0f01c5c r6:c10833c0 r5:c196b000 > [ 160.988160] r4:c891ab40 > [ 160.990676] [<c09276e0>] (ip_list_rcv) from [<c08a6328>] (__netif_receive_skb_list_core+0x90/0x214) > [ 160.999714] r10:c196b000 r9:c196b000 r8:00000000 r7:c0f01c5c r6:c0f07450 r5:c891acc0 > [ 161.007517] r4:c09276e0 > [ 161.010032] [<c08a6298>] (__netif_receive_skb_list_core) from [<c08a6670>] (netif_receive_skb_list_internal+0x1c4/0x2c4) > [ 161.020893] r10:c10833b8 r9:c0f08a00 r8:00000000 r7:c0f01d44 r6:c0f01d44 r5:c0f01d44 > [ 161.028696] r4:00000000 > [ 161.031211] [<c08a64ac>] (netif_receive_skb_list_internal) from [<c08a67a8>] (netif_receive_skb_list+0x38/0xe4) > [ 161.041291] r10:c1963de0 r9:f08527d0 r8:f08527c0 r7:00000003 r6:c1960040 r5:c103d920 > [ 161.049094] r4:c0f01d44 > [ 161.051609] [<c08a6770>] (netif_receive_skb_list) from [<c0795af4>] (lan966x_fdma_napi_poll+0x3ec/0x490) > [ 161.061081] r9:f08527d0 r8:f08527c0 r7:00000003 r6:c1960040 r5:c1962de0 r4:00000040 > [ 161.068796] [<c0795708>] (lan966x_fdma_napi_poll) from [<c08a6ee4>] (__napi_poll+0x34/0x1dc) > [ 161.077226] r10:ef7e6080 r9:c0f01db4 r8:c0f03d00 r7:c0f01db3 r6:00000040 r5:c1963de0 > [ 161.085029] r4:00000001 > [ 161.087545] [<c08a6eb0>] (__napi_poll) from [<c08a7290>] (net_rx_action+0xec/0x288) > [ 161.095190] r7:2e957000 r6:ffffc936 r5:0000012c r4:c1963de0 > [ 161.100825] [<c08a71a4>] (net_rx_action) from [<c010143c>] (__do_softirq+0x13c/0x384) > [ 161.108647] r10:c0f08a00 r9:00000008 r8:00000100 r7:c1037d40 r6:00000003 r5:00000004 > [ 161.116451] r4:c0f0308c > [ 161.118966] [<c0101300>] (__do_softirq) from [<c013bc9c>] (irq_exit+0xac/0xdc) > [ 161.126182] r10:c0f04f78 r9:c0f08a00 r8:00000000 r7:c0f01ed4 r6:00000000 r5:c0f01ea0 > [ 161.133985] r4:c0f08a00 > [ 161.136500] [<c013bbf0>] (irq_exit) from [<c0a62c84>] (generic_handle_arch_irq+0x48/0x4c) > [ 161.144663] r5:c0f01ea0 r4:c0e8d358 > [ 161.148218] [<c0a62c3c>] (generic_handle_arch_irq) from [<c0100be8>] (__irq_svc+0x88/0xb0)
Am 2022-03-18 13:10, schrieb Horatiu Vultur: > The 03/18/2022 12:07, Michael Walle wrote: >> > Currently when injecting or extracting a frame from CPU, the frame >> > is given to the HW each word at a time. There is another way to >> > inject/extract frames from CPU using FDMA(Frame Direct Memory Access). >> > In this way the entire frame is given to the HW. This improves both >> > RX and TX bitrate. >> >> I wanted to test this. ping and such works fine and I'm also >> seeing fdma interrupts. > > Thanks for testing this also on your board. > >> But as soon as I try iperf3 I get a skb_panic >> (due to frame size?). Hope that splash below helps. > > I have not seen this issue. But it looks like it is a problem that > there > is no more space to add the FCS. > Can you tell me how you run iperf3 so I can also try it? oh, I forgot to include the commandline. # on the remote computer $ iperf3 --version iperf 3.6 (cJSON 1.5.2) Linux eddie01 4.19.0-18-686-pae #1 SMP Debian 4.19.208-1 (2021-09-29) i686 Optional features available: CPU affinity setting, IPv6 flow label, SCTP, TCP congestion algorithm setting, sendfile / zerocopy, socket pacing, authentication $ iperf3 -s # on the board $ iperf3 --version iperf 3.10.1 (cJSON 1.7.13) Linux buildroot 5.17.0-rc8-next-20220316-00058-gc6cb0628f2a6-dirty #385 SMP Fri Mar 18 13:34:26 CET 2022 armv7l Optional features available: CPU affinity setting, IPv6 flow label, TCP congestion algorithm setting, sendfile / zerocopy, socket pacing, bind to device, support IPv4 don't fragment $ iperf3 -c eddie01 > Also I have a small diff that might fix the issue: > --- > --- a/drivers/net/ethernet/microchip/lan966x/lan966x_fdma.c > +++ b/drivers/net/ethernet/microchip/lan966x/lan966x_fdma.c > @@ -534,6 +534,8 @@ int lan966x_fdma_xmit(struct sk_buff *skb, __be32 > *ifh, struct net_device *dev) > struct lan966x_tx_dcb *next_dcb, *dcb; > struct lan966x_tx *tx = &lan966x->tx; > struct lan966x_db *next_db; > + int needed_headroom; > + int needed_tailroom; > dma_addr_t dma_addr; > int next_to_use; > int err; > @@ -554,10 +556,11 @@ int lan966x_fdma_xmit(struct sk_buff *skb, > __be32 *ifh, struct net_device *dev) > > /* skb processing */ > skb_tx_timestamp(skb); btw. skb_tx_timestamp() should be as close to the handover of the frame to the hardware as possible, no? > - if (skb_headroom(skb) < IFH_LEN * sizeof(u32)) { > - err = pskb_expand_head(skb, > - IFH_LEN * sizeof(u32) - > skb_headroom(skb), > - 0, GFP_ATOMIC); > + needed_headroom = max_t(int, IFH_LEN * sizeof(u32) - > skb_headroom(skb), 0); > + needed_tailroom = max_t(int, ETH_FCS_LEN - skb_tailroom(skb), > 0); > + if (needed_headroom || needed_tailroom) { > + err = pskb_expand_head(skb, needed_headroom, > needed_tailroom, > + GFP_ATOMIC); > if (unlikely(err)) { > dev->stats.tx_dropped++; > err = NETDEV_TX_OK; Indeed this will fix the issue: # iperf3 -c eddie01 Connecting to host eddie01, port 5201 [ 5] local 10.0.1.143 port 55342 connected to 10.0.1.42 port 5201 [ ID] Interval Transfer Bitrate Retr Cwnd [ 5] 0.00-1.01 sec 43.8 MBytes 364 Mbits/sec 0 245 KBytes [ 5] 1.01-2.02 sec 43.8 MBytes 364 Mbits/sec 0 246 KBytes [ 5] 2.02-3.03 sec 43.8 MBytes 364 Mbits/sec 0 259 KBytes # iperf3 -R -c eddie01 Connecting to host eddie01, port 5201 Reverse mode, remote host eddie01 is sending [ 5] local 10.0.1.143 port 55346 connected to 10.0.1.42 port 5201 [ ID] Interval Transfer Bitrate [ 5] 0.00-1.00 sec 28.6 MBytes 240 Mbits/sec [ 5] 1.00-2.00 sec 28.9 MBytes 242 Mbits/sec [ 5] 2.00-3.00 sec 28.7 MBytes 241 Mbits/sec -michael