Re: Qualcomm Atheros QCA6174 monitoring mode receiving no data

[Patrick Daume <patrick.daume@dedrone.com>]

On Mi, 2019-06-05 at 17:38 +0200, Patrick Daume wrote:
> I test it with modprobe option rawmode=1 and it seems to work.
> 
> solution:
> 
> linux kernel 5.1.0-rc7-wt-ath+
> firmware: WLAN.RM.4.4.1.c3-00030
> 
> sudo modprobe ath10k_core rawmode=1
> sudo modprobe ath10k_pci
> 
> 
> But if i call tcpdump for a while i got some error traces in dmesg.
> How
> can i fix this?
> 
> [  281.101682]  pinctrl_intel [last unloaded: cfg80211]
> [  281.101694] CPU: 1 PID: 0 Comm: swapper/1 Tainted:
> G        W         5.1.0-rc7-wt-ath+ #3
> [  281.101697] Hardware name:  /, BIOS 5.11 03/24/2016
> [  281.101703] RIP: 0010:skb_trim+0x2d/0x30
> [  281.101708] Code: 00 00 39 77 70 77 02 f3 c3 8b 47 74 85 c0 75 1a
> 48 8b 87 c8 00 00 00 48 2b 87 c0 00 00 00 89 77 70 01 c6 89 b7 b8 00
> 00 00 c3 <0f> 0b c3 0f 1f 44 00 00 8b 87 e8 00 00 00 8b 96 d0 00 00
> 00 01 d0
> [  281.101712] RSP: 0018:ffff9d47dda83ce0 EFLAGS: 00010206
> [  281.101717] RAX: 00000000000004cc RBX: 0000000000008000 RCX:
> ffff9d47dda83e60
> [  281.101720] RDX: 0000000000000000 RSI: 0000000000000ae0 RDI:
> ffff9d47d3eebf00
> [  281.101723] RBP: 0000000000000004 R08: 0000000000000000 R09:
> 0000000000000000
> [  281.101726] R10: fffff2c05158d840 R11: ffff9d43b447e4b8 R12:
> 0000000000000000
> [  281.101729] R13: 0000000000004000 R14: ffff9d47d0421620 R15:
> ffff9d47d3eebf00
> [  281.101734] FS:  0000000000000000(0000) GS:ffff9d47dda80000(0000)
> knlGS:0000000000000000
> [  281.101737] CS:  0010 DS: 0000 ES: 0000 CR0: 0000000080050033
> [  281.101741] CR2: 000056110f0580a4 CR3: 00000003a120e003 CR4:
> 00000000003606e0
> [  281.101744] DR0: 0000000000000000 DR1: 0000000000000000 DR2:
> 0000000000000000
> [  281.101747] DR3: 0000000000000000 DR6: 00000000fffe0ff0 DR7:
> 0000000000000400
> [  281.101749] Call Trace:
> [  281.101755]  <IRQ>
> [  281.101798]  ath10k_htt_rx_h_mpdu+0x6a5/0xab0 [ath10k_core]
> [  281.101813]  ? swiotlb_map+0x6b/0x1c0
> [  281.101846]  ath10k_htt_txrx_compl_task+0xc4f/0x1490 [ath10k_core]
> [  281.101864]  ath10k_pci_napi_poll+0x51/0x130 [ath10k_pci]
> [  281.101876]  net_rx_action+0x27e/0x3d0
> [  281.101888]  __do_softirq+0xfd/0x2d8
> [  281.101899]  irq_exit+0xae/0xb0
> [  281.101906]  do_IRQ+0x52/0xe0
> [  281.101913]  common_interrupt+0xf/0xf
> [  281.101917]  </IRQ>
> [  281.101928] RIP: 0010:cpuidle_enter_state+0xc4/0x460
> [  281.101932] Code: e8 71 f7 90 ff 80 7c 24 0f 00 74 17 9c 58 0f 1f
> 44 00 00 f6 c4 02 0f 85 65 03 00 00 31 ff e8 b3 f6 96 ff fb 66 0f 1f
> 44 00 00 <45> 85 e4 0f 88 35 03 00 00 4c 2b 7c 24 10 48 ba cf f7 53
> e3 a5 9b
> [  281.101936] RSP: 0018:ffffacef41943e88 EFLAGS: 00000246 ORIG_RAX:
> ffffffffffffffdd
> [  281.101941] RAX: ffff9d47ddaa2c80 RBX: ffffffffad2d7860 RCX:
> 000000000000001f
> [  281.101944] RDX: 0000000000000000 RSI: 000000002d958403 RDI:
> 0000000000000000
> [  281.101947] RBP: ffff9d47ddaac100 R08: fffffffca8f126a7 R09:
> ffff9d47ddaa1ca4
> [  281.101950] R10: 0000000000000018 R11: 0000000000002615 R12:
> 0000000000000008
> [  281.101953] R13: 0000000000000008 R14: ffffffffad2d7860 R15:
> 0000004172f2f453
> [  281.101965]  ? cpuidle_enter_state+0x9f/0x460
> [  281.101972]  do_idle+0x202/0x280
> [  281.101980]  cpu_startup_entry+0x19/0x20
> [  281.101990]  start_secondary+0x17a/0x1d0
> [  281.101999]  secondary_startup_64+0xa4/0xb0
> [  281.102006] ---[ end trace 23dc6ddb093396c5 ]---
> [  306.126454] device wlp2s0 entered promiscuous mode
> [  317.700835] device wlp2s0 left promiscuous mode








htt.h:2226

/*
 * Should be: sizeof(struct htt_host_rx_desc) + max rx MSDU size,
>> * sizeof(struct htt_host_rx_desc) 360 bytes + max MPDU/~MDSU size
802.11AC 16.384 bytes (2^14) + (2^9)
 * rounded up to a cache line size.
 */
>> #define HTT_RX_BUF_SIZE 16896
#define HTT_RX_MSDU_SIZE (HTT_RX_BUF_SIZE - (int)sizeof(struct
htt_rx_desc))





htt_rx.c:1390, ath10k_htt_rx_h_undecap_raw

if (WARN_ON_ONCE(!(is_first && is_last) && !msdu_limit_err))
	return;

/* We see zero length msdus and greater max lenght msdus,
 * not sure why. At least don't try to trim it further.
 */
>> if (unlikely(msdu->len < FCS_LEN || msdu->len > HTT_RX_MSDU_SIZE))
>>	return;

skb_trim(msdu, msdu->len - FCS_LEN);


_______________________________________________
ath10k mailing list
ath10k@lists.infradead.org
http://lists.infradead.org/mailman/listinfo/ath10k