FEC DMA bug

ToxicAvenger — Mon, 16 Dec 2013 13:13:19 GMT

Hello.

Im working with FEC on i.MX53 in bridge configuration (wlan + eth in bridge).

System is Android 10.2 from Adeneo, im using brctl util to setup bridge.

When bridge is starting, device hangs and im receiving error (kernel panic):

kernel BUG at arch/arm/mm/dma-mapping.c:409!

Unable to handle kernel NULL pointer dereference at virtual address 00000000

pgd = 80004000

[00000000] *pgd=00000000

Internal error: Oops: 817 [#1] PREEMPT

...

[<80035e64>] (__bug+0x18/0x24) from [<80037894>] (___dma_single_cpu_to_dev+0x34/0x60)

[<80037894>] (___dma_single_cpu_to_dev+0x34/0x60) from [<80252474>] (fec_enet_start_xmit+0x13c/0x1dc)

[<80252474>] (fec_enet_start_xmit+0x13c/0x1dc) from [<80364598>] (dev_hard_start_xmit+0x2ac/0x3b8)

[<80364598>] (dev_hard_start_xmit+0x2ac/0x3b8) from [<80377ca0>] (sch_direct_xmit+0x8c/0x22c)

[<80377ca0>] (sch_direct_xmit+0x8c/0x22c) from [<80364b4c>] (dev_queue_xmit+0x370/0x5ac)

[<80364b4c>] (dev_queue_xmit+0x370/0x5ac) from [<8042da7c>] (br_dev_queue_push_xmit+0x88/0x90)

[<8042da7c>] (br_dev_queue_push_xmit+0x88/0x90) from [<80432c68>] (br_nf_post_routing+0x208/0x22c)

[<80432c68>] (br_nf_post_routing+0x208/0x22c) from [<8037f8b8>] (nf_iterate+0x54/0x98)

[<8037f8b8>] (nf_iterate+0x54/0x98) from [<8037fae8>] (nf_hook_slow+0x74/0x118)

[<8037fae8>] (nf_hook_slow+0x74/0x118) from [<8042dac8>] (br_forward_finish+0x44/0x6c)

[<8042dac8>] (br_forward_finish+0x44/0x6c) from [<80432dac>] (br_nf_forward_finish+0x120/0x138)

[<80432dac>] (br_nf_forward_finish+0x120/0x138) from [<80433160>] (br_nf_forward_ip+0x260/0x2a0)

[<80433160>] (br_nf_forward_ip+0x260/0x2a0) from [<8037f8b8>] (nf_iterate+0x54/0x98)

[<8037f8b8>] (nf_iterate+0x54/0x98) from [<8037fae8>] (nf_hook_slow+0x74/0x118)

[<8037fae8>] (nf_hook_slow+0x74/0x118) from [<8042db80>] (__br_forward+0x90/0xb8)

[<8042db80>] (__br_forward+0x90/0xb8) from [<8042d7cc>] (deliver_clone+0x44/0x4c)

[<8042d7cc>] (deliver_clone+0x44/0x4c) from [<8042e7b0>] (br_handle_frame_finish+0x184/0x220)

[<8042e7b0>] (br_handle_frame_finish+0x184/0x220) from [<80433338>] (br_nf_pre_routing_finish_ipv6+0x140/0x154)

[<80433338>] (br_nf_pre_routing_finish_ipv6+0x140/0x154) from [<80433b40>] (br_nf_pre_routing+0x40c/0x770)

[<80433b40>] (br_nf_pre_routing+0x40c/0x770) from [<8037f8b8>] (nf_iterate+0x54/0x98)

[<8037f8b8>] (nf_iterate+0x54/0x98) from [<8037fae8>] (nf_hook_slow+0x74/0x118)

[<8037fae8>] (nf_hook_slow+0x74/0x118) from [<8042ea40>] (br_handle_frame+0x1f4/0x244)

[<8042ea40>] (br_handle_frame+0x1f4/0x244) from [<80363414>] (__netif_receive_skb+0x1e0/0x324)

[<80363414>] (__netif_receive_skb+0x1e0/0x324) from [<803635e0>] (process_backlog+0x88/0x134)

[<803635e0>] (process_backlog+0x88/0x134) from [<803638c4>] (net_rx_action+0x5c/0x144)

[<803638c4>] (net_rx_action+0x5c/0x144) from [<8005cccc>] (__do_softirq+0x70/0xf8)

[<8005cccc>] (__do_softirq+0x70/0xf8) from [<8005ce3c>] (do_softirq+0x40/0x5c)

[<8005ce3c>] (do_softirq+0x40/0x5c) from [<8005ced0>] (run_ksoftirqd+0x78/0x15c)

[<8005ced0>] (run_ksoftirqd+0x78/0x15c) from [<8006d838>] (kthread+0x78/0x80)

[<8006d838>] (kthread+0x78/0x80) from [<800339e0>] (kernel_thread_exit+0x0/0x8)

Code: e92d4010 e59f000c eb113dc9 e3a03000 (e5833000)

---[ end trace 10069177ba2a70fb ]---

Is there any patch for that situation?

Best regards,

Toxic

i.MX ProcessorsのトピックFEC DMA bug