driver icon indicating copy to clipboard operation
driver copied to clipboard
verified publisher icon linux4wilc

kernel NULL pointer at PC is at debug_thread+0x144/0x2e8 [wilc_sdio]

Open mgkiller7 opened this issue 6 years ago • 13 comments

we use wilc1000 in cpu NUC976（arm926t core）with linux 3.10.108. wilc module work well except for scheduling bug at https://github.com/linux4wilc/driver/issues/83 （Not yet solved）.

There are two ways to to reproduce this "kernel NULL pointer Oops".

**1、Long time ping with packet size = 8000 **

when wilc1000 sucess to connect to a AP, we continuous ping the AP's gateway ip 192.168.50.1 . when sucessfull reach 173206 times(almost 49 hours), the ping cmd fail to ping any more. At this time the wilc driver log out "cfg_get Timed Out", and end with "kernel NULL pointer".

Below is the log: 8008 bytes from 192.168.50.1: seq=42118 ttl=64 time=130.000 ms 8008 bytes from 192.168.50.1: seq=42119 ttl=64 time=1020.000 ms 8008 bytes from 192.168.50.1: seq=42120 ttl=64 time=10.000 ms 8008 bytes from 192.168.50.1: seq=42121 ttl=64 time=1030.000 ms 8008 bytes from 192.168.50.1: seq=42122 ttl=64 time=20.000 ms 8008 bytes from 192.168.50.1: seq=42123 ttl=64 time=90.000 ms [176580.360000] wilc_sdio mmc0:0001:1 wlan0: INFO [debug_thread]*** Debug Thread Running *** 8008 bytes from 192.168.50.1: seq=42124 ttl=64 time=110.000 ms 8008 bytes from 192.168.50.1: seq=42125 ttl=64 time=120.000 ms 8008 bytes from 192.168.50.1: seq=42126 ttl=64 time=1020.000 ms 8008 bytes from 192.168.50.1: seq=42127 ttl=64 time=10.000 ms 8008 bytes from 192.168.50.1: seq=42128 ttl=64 time=170.000 ms 8008 bytes from 192.168.50.1: seq=42129 ttl=64 time=70.000 ms [176586.360000] wilc_sdio mmc0:0001:1 wlan0: INFO [debug_thread]*** Debug Thread Running *** 8008 bytes from 192.168.50.1: seq=42130 ttl=64 time=90.000 ms 8008 bytes from 192.168.50.1: seq=42131 ttl=64 time=200.000 ms 8008 bytes from 192.168.50.1: seq=42132 ttl=64 time=120.000 ms 8008 bytes from 192.168.50.1: seq=42133 ttl=64 time=1020.000 ms 8008 bytes from 192.168.50.1: seq=42134 ttl=64 time=10.000 ms 8008 bytes from 192.168.50.1: seq=42135 ttl=64 time=160.000 ms [176592.360000] wilc_sdio mmc0:0001:1 wlan0: INFO [debug_thread]*** Debug Thread Running *** 8008 bytes from 192.168.50.1: seq=42136 ttl=64 time=70.000 ms 8008 bytes from 192.168.50.1: seq=42137 ttl=64 time=90.000 ms [176598.360000] wilc_sdio mmc0:0001:1 wlan0: INFO [debug_thread]*** Debug Thread Running *** [176600.370000] cfg_get Timed Out [176600.370000] wilc_sdio mmc0:0001:1 wlan0: ERR [wilc_send_config_pkt:1786] Get Timed out [176600.390000] wilc_sdio mmc0:0001:1 wlan0: ERR [wilc_get_statistics:1084] Failed to send scan parameters [176604.360000] wilc_sdio mmc0:0001:1 wlan0: INFO [debug_thread]*** Debug Thread Running *** [176605.370000] cfg_get Timed Out [176605.370000] wilc_sdio mmc0:0001:1 wlan0: ERR [wilc_send_config_pkt:1786] Get Timed out [176605.390000] wilc_sdio mmc0:0001:1 wlan0: ERR [wilc_get_statistics:1084] Failed to send scan parameters [176610.360000] wilc_sdio mmc0:0001:1 wlan0: INFO [debug_thread]*** Debug Thread Running *** [176610.380000] cfg_get Timed Out [176610.380000] wilc_sdio mmc0:0001:1 wlan0: ERR [wilc_send_config_pkt:1786] Get Timed out [176610.400000] wilc_sdio mmc0:0001:1 wlan0: ERR [wilc_get_statistics:1084] Failed to send scan parameters [176615.410000] cfg_get Timed Out [176615.410000] wilc_sdio mmc0:0001:1 wlan0: ERR [wilc_send_config_pkt:1786] Get Timed out [176615.420000] wilc_sdio mmc0:0001:1 wlan0: ERR [wilc_get_statistics:1084] Failed to send scan parameters [176616.360000] wilc_sdio mmc0:0001:1 wlan0: INFO [debug_thread]*** Debug Thread Running *** [176620.400000] cfg_get Timed Out [176620.400000] wilc_sdio mmc0:0001:1 wlan0: ERR [wilc_send_config_pkt:1786] Get Timed out [176620.410000] wilc_sdio mmc0:0001:1 wlan0: ERR [wilc_get_statistics:1084] Failed to send scan parameters [176622.360000] wilc_sdio mmc0:0001:1 wlan0: INFO [debug_thread]*** Debug Thread Running *** [176622.360000] wilc_sdio mmc0:0001:1 wlan0: INFO [debug_thread]<Recover> [176622.380000] wilc_sdio mmc0:0001:1 wlan0: INFO [wilc_mac_close]Mac close [176622.380000] wilc_sdio mmc0:0001:1 wlan0: INFO [wilc_mac_close]Deinitializing wilc [176622.400000] wilc_sdio mmc0:0001:1 wlan0: INFO [wilc_wlan_deinitialize]Deinitializing wilc ... [176622.400000] wilc_sdio mmc0:0001:1 wlan0: INFO [wilc_wlan_deinitialize]Disabling IRQ [176622.440000] wilc_sdio mmc0:0001:1: wilc_sdio_disable_interrupt [176622.450000] wilc_sdio mmc0:0001:1 wlan0: INFO [wilc_wlan_deinitialize]Deinitializing Threads [176622.470000] wilc_sdio mmc0:0001:1 wlan0: INFO [wlan_deinitialize_threads]Deinitializing Threads [176622.470000] wilc_sdio mmc0:0001:1 wlan0: INFO [wlan_deinitialize_threads]Deinitializing Threads [176622.490000] wilc_sdio mmc0:0001:1 wlan0: INFO [wilc_wlan_deinitialize]Deinitializing IRQ [176622.500000] wilc_sdio mmc0:0001:1 wlan0: INFO [wilc_wlan_deinitialize]Deinitializing WILC Wlan [176622.510000] wilc_sdio mmc0:0001:1 wlan0: INFO [wilc_wlan_deinitialize]wilc deinitialization Done [176622.520000] Unable to handle kernel NULL pointer dereference at virtual address 00000080 [176622.530000] pgd = c0004000 [176622.530000] [00000080] *pgd=00000000 [176622.530000] Internal error: Oops: 17 [#1] PREEMPT ARM [176622.530000] Modules linked in: wilc_sdio(C) watchdogd(O) usb_f_rndis u_ether libcomposite [176622.530000] CPU: 0 PID: 1647 Comm: WILC_DEBUG Tainted: G WC O 3.10.108 #5 [176622.530000] task: c38d82c0 ti: c1fd8000 task.ti: c1fd8000 [176622.530000] PC is at debug_thread+0x144/0x2e8 [wilc_sdio] [176622.530000] LR is at wake_up_klogd+0x54/0x7c [176622.530000] pc : [] lr : [] psr: 40000013 [176622.530000] sp : c1fd9f20 ip : c1fd9cf8 fp : c1fd9f64 [176622.530000] r10: 00000002 r9 : bf0349f4 r8 : bf0345ec [176622.530000] r7 : bf034394 r6 : bf02e39e r5 : c1f3da00 r4 : c1ddd2c0 [176622.530000] r3 : 00000000 r2 : 00000001 r1 : 00000000 r0 : 00000000 [176622.530000] Flags: nZcv IRQs on FIQs on Mode SVC_32 ISA ARM Segment kernel [176622.530000] Control: 0005317f Table: 01f30000 DAC: 00000017 [176622.530000] Process WILC_DEBUG (pid: 1647, stack limit = 0xc1fd81b8) [176622.530000] Stack: (0xc1fd9f20 to 0xc1fda000) [176622.530000] 9f20: c1fd9f44 c1fd9f30 c03655d8 c1f39d44 c1ddd350 c38fa3c0 bf0225ac c1f39d44 [176622.530000] 9f40: 00000000 c1ddd2c0 bf0225ac 00000000 00000000 00000000 c1fd9fac c1fd9f68 [176622.530000] 9f60: c0038fa4 bf0225bc c1fd9f94 00000000 c0040920 c1ddd2c0 00000000 c1fd9f7c [176622.530000] 9f80: c1fd9f7c 00000000 c1fd9f88 c1fd9f88 c1f39d44 c0038ef0 00000000 00000000 [176622.530000] 9fa0: 00000000 c1fd9fb0 c000e1f8 c0038f00 00000000 00000000 00000000 00000000 [176622.530000] 9fc0: 00000000 00000000 00000000 00000000 00000000 00000000 00000000 00000000 [176622.530000] 9fe0: 00000000 00000000 00000000 00000000 00000013 00000000 00000000 00000000 [176622.530000] Backtrace: [176622.530000] [] (debug_thread+0x0/0x2e8 [wilc_sdio]) from [] (kthread+0xb4/0xc0) [176622.530000] [] (kthread+0x0/0xc0) from [] (ret_from_fork+0x14/0x3c) [176622.530000] r7:00000000 r6:00000000 r5:c0038ef0 r4:c1f39d44 [176622.530000] Code: e3a0a002 e5930080 ebfffb22 e5943028 (e5930080) [176622.550000] ---[ end trace 30590ca708369d2c ]---

^C --- 192.168.50.1 ping statistics --- 253244 packets transmitted, 173206 packets received, 31% packet loss round-trip min/avg/max = 10.000/238.313/1380.000 ms

2、do a ESD Test to wilc module Another ways to quickly reproduce kernel NULL pointer Oops is to do a ESD Test to wilc module, which will cause wilc_sdio_cmd52 time-out, and also lead to kernel NULL pointer Oops:

[ 405.740000] wilc_sdio mmc0:0001:1 wlan0: INFO [debug_thread]*** Debug Thread Running *** [ 411.740000] wilc_sdio mmc0:0001:1 wlan0: INFO [debug_thread]*** Debug Thread Running *** [ 417.760000] wilc_sdio mmc0:0001:1 wlan0: INFO [debug_thread]*** Debug Thread Running *** [ 423.760000] wilc_sdio mmc0:0001:1 wlan0: INFO [debug_thread]*** Debug Thread Running *** [ 429.760000] wilc_sdio mmc0:0001:1 wlan0: INFO [debug_thread]*** Debug Thread Running *** [ 435.760000] wilc_sdio mmc0:0001:1 wlan0: INFO [debug_thread]*** Debug Thread Running *** [ 441.760000] wilc_sdio mmc0:0001:1 wlan0: INFO [debug_thread]*** Debug Thread Running *** [ 447.760000] wilc_sdio mmc0:0001:1 wlan0: INFO [debug_thread]*** Debug Thread Running *** [ 453.760000] wilc_sdio mmc0:0001:1 wlan0: INFO [debug_thread]*** Debug Thread Running *** [ 455.700000] wilc_sdio mmc0:0001:1: wilc_sdio_cmd52..failed, err(-110) [ 455.700000] wilc_sdio mmc0:0001:1: Failed cmd 52, read reg (000000fa) ... [ 455.710000] wilc_sdio mmc0:0001:1: wilc_sdio_cmd52..failed, err(-110) [ 455.720000] wilc_sdio mmc0:0001:1: Failed cmd52, set 0x10c data... [ 455.730000] wilc_sdio mmc0:0001:1 wlan0: ERR [wilc_wlan_handle_txq:1053] fail read reg vmm_tbl_entry.. [ 455.740000] wilc_sdio mmc0:0001:1: wilc_sdio_cmd52..failed, err(-110) [ 455.750000] wilc_sdio mmc0:0001:1: Failed cmd 52, read reg (000000fc) ... [ 458.700000] cfg_get Timed Out [ 458.700000] wilc_sdio mmc0:0001:1 wlan0: ERR [wilc_send_config_pkt:1786] Get Timed out [ 458.710000] wilc_sdio mmc0:0001:1 wlan0: ERR [wilc_get_statistics:1084] Failed to send scan parameters [ 459.760000] wilc_sdio mmc0:0001:1 wlan0: INFO [debug_thread]*** Debug Thread Running *** [ 460.710000] wilc_sdio mmc0:0001:1: wilc_sdio_cmd52..failed, err(-110) [ 460.710000] wilc_sdio mmc0:0001:1: Failed cmd 52, read reg (000000fa) ... [ 460.720000] wilc_sdio mmc0:0001:1: wilc_sdio_cmd52..failed, err(-110) [ 460.730000] wilc_sdio mmc0:0001:1: Failed cmd52, set 0x10c data... [ 460.740000] wilc_sdio mmc0:0001:1 wlan0: ERR [wilc_wlan_handle_txq:1053] fail read reg vmm_tbl_entry.. [ 460.750000] wilc_sdio mmc0:0001:1: wilc_sdio_cmd52..failed, err(-110) [ 460.760000] wilc_sdio mmc0:0001:1: Failed cmd 52, read reg (000000fc) ... [ 463.720000] cfg_get Timed Out [ 463.720000] wilc_sdio mmc0:0001:1 wlan0: ERR [wilc_send_config_pkt:1786] Get Timed out [ 463.730000] wilc_sdio mmc0:0001:1 wlan0: ERR [wilc_get_statistics:1084] Failed to send scan parameters [ 465.720000] wilc_sdio mmc0:0001:1: wilc_sdio_cmd52..failed, err(-110) [ 465.720000] wilc_sdio mmc0:0001:1: Failed cmd 52, read reg (000000fa) ... [ 465.730000] wilc_sdio mmc0:0001:1: wilc_sdio_cmd52..failed, err(-110) [ 465.740000] wilc_sdio mmc0:0001:1: Failed cmd52, set 0x10c data... [ 465.750000] wilc_sdio mmc0:0001:1 wlan0: ERR [wilc_wlan_handle_txq:1053] fail read reg vmm_tbl_entry.. [ 465.760000] wilc_sdio mmc0:0001:1 wlan0: INFO [debug_thread]*** Debug Thread Running *** [ 465.770000] wilc_sdio mmc0:0001:1: wilc_sdio_cmd52..failed, err(-110) [ 465.780000] wilc_sdio mmc0:0001:1: Failed cmd 52, read reg (000000fc) ... [ 468.720000] cfg_get Timed Out [ 468.720000] wilc_sdio mmc0:0001:1 wlan0: ERR [wilc_send_config_pkt:1786] Get Timed out [ 468.730000] wilc_sdio mmc0:0001:1 wlan0: ERR [wilc_get_statistics:1084] Failed to send scan parameters [ 470.730000] wilc_sdio mmc0:0001:1: wilc_sdio_cmd52..failed, err(-110) [ 470.730000] wilc_sdio mmc0:0001:1: Failed cmd 52, read reg (000000fa) ... [ 470.740000] wilc_sdio mmc0:0001:1: wilc_sdio_cmd52..failed, err(-110) [ 470.750000] wilc_sdio mmc0:0001:1: Failed cmd52, set 0x10c data... [ 470.760000] wilc_sdio mmc0:0001:1 wlan0: ERR [wilc_wlan_handle_txq:1053] fail read reg vmm_tbl_entry.. [ 470.770000] wilc_sdio mmc0:0001:1: wilc_sdio_cmd52..failed, err(-110) [ 470.780000] wilc_sdio mmc0:0001:1: Failed cmd 52, read reg (000000fc) ... [ 471.780000] wilc_sdio mmc0:0001:1 wlan0: INFO [debug_thread]*** Debug Thread Running *** [ 473.730000] cfg_get Timed Out [ 473.730000] wilc_sdio mmc0:0001:1 wlan0: ERR [wilc_send_config_pkt:1786] Get Timed out [ 473.740000] wilc_sdio mmc0:0001:1 wlan0: ERR [wilc_get_statistics:1084] Failed to send scan parameters [ 475.740000] wilc_sdio mmc0:0001:1: wilc_sdio_cmd52..failed, err(-110) [ 475.740000] wilc_sdio mmc0:0001:1: Failed cmd 52, read reg (000000fa) ... [ 475.750000] wilc_sdio mmc0:0001:1: wilc_sdio_cmd52..failed, err(-110) [ 475.760000] wilc_sdio mmc0:0001:1: Failed cmd52, set 0x10c data... [ 475.770000] wilc_sdio mmc0:0001:1 wlan0: ERR [wilc_wlan_handle_txq:1053] fail read reg vmm_tbl_entry.. [ 475.780000] wilc_sdio mmc0:0001:1: wilc_sdio_cmd52..failed, err(-110) [ 475.790000] wilc_sdio mmc0:0001:1: Failed cmd 52, read reg (000000fc) ... [ 477.780000] wilc_sdio mmc0:0001:1 wlan0: INFO [debug_thread]*** Debug Thread Running *** [ 478.740000] cfg_get Timed Out [ 478.740000] wilc_sdio mmc0:0001:1 wlan0: ERR [wilc_send_config_pkt:1786] Get Timed out [ 478.750000] wilc_sdio mmc0:0001:1 wlan0: ERR [wilc_get_statistics:1084] Failed to send scan parameters [ 480.750000] wilc_sdio mmc0:0001:1: wilc_sdio_cmd52..failed, err(-110) [ 480.750000] wilc_sdio mmc0:0001:1: Failed cmd 52, read reg (000000fa) ... [ 480.760000] wilc_sdio mmc0:0001:1: wilc_sdio_cmd52..failed, err(-110) [ 480.770000] wilc_sdio mmc0:0001:1: Failed cmd52, set 0x10c data... [ 480.780000] wilc_sdio mmc0:0001:1 wlan0: ERR [wilc_wlan_handle_txq:1053] fail read reg vmm_tbl_entry.. [ 480.790000] wilc_sdio mmc0:0001:1: wilc_sdio_cmd52..failed, err(-110) [ 480.800000] wilc_sdio mmc0:0001:1: Failed cmd 52, read reg (000000fc) ... [ 483.760000] cfg_get Timed Out [ 483.760000] wilc_sdio mmc0:0001:1 wlan0: ERR [wilc_send_config_pkt:1786] Get Timed out [ 483.770000] wilc_sdio mmc0:0001:1 wlan0: ERR [wilc_get_statistics:1084] Failed to send scan parameters [ 483.780000] wilc_sdio mmc0:0001:1 wlan0: INFO [debug_thread]*** Debug Thread Running *** [ 483.790000] wilc_sdio mmc0:0001:1 wlan0: INFO [debug_thread]<Recover> [ 483.790000] wilc_sdio mmc0:0001:1 wlan0: INFO [wilc_mac_close]Mac close [ 483.800000] wilc_sdio mmc0:0001:1 wlan0: INFO [wilc_mac_close]Deinitializing wilc [ 483.810000] wilc_sdio mmc0:0001:1 wlan0: INFO [wilc_wlan_deinitialize]Deinitializing wilc ... [ 483.820000] wilc_sdio mmc0:0001:1 wlan0: INFO [wilc_wlan_deinitialize]Disabling IRQ [ 483.830000] wilc_sdio mmc0:0001:1: wilc_sdio_disable_interrupt [ 483.840000] wilc_sdio mmc0:0001:1: can't release sdio_irq, err(-110) [ 483.850000] wilc_sdio mmc0:0001:1 wlan0: INFO [wilc_wlan_deinitialize]Deinitializing Threads [ 483.850000] wilc_sdio mmc0:0001:1 wlan0: INFO [wlan_deinitialize_threads]Deinitializing Threads [ 483.870000] wilc_sdio mmc0:0001:1 wlan0: INFO [wlan_deinitialize_threads]Deinitializing Threads [ 483.870000] wilc_sdio mmc0:0001:1 wlan0: INFO [wilc_wlan_deinitialize]Deinitializing IRQ [ 483.880000] wilc_sdio mmc0:0001:1: wilc_sdio_cmd52..failed, err(-110) [ 483.890000] wilc_sdio mmc0:0001:1: Failed cmd 52, read reg (000000fa) ... [ 483.900000] wilc_sdio mmc0:0001:1: wilc_sdio_cmd52..failed, err(-110) [ 483.910000] wilc_sdio mmc0:0001:1: Failed cmd52, set 0x10c data... [ 483.930000] wilc_sdio mmc0:0001:1 wlan0: ERR [wilc_wlan_stop:1570] Error while reading reg [ 483.950000] wilc_sdio mmc0:0001:1: wilc_sdio_cmd52..failed, err(-110) [ 483.950000] wilc_sdio mmc0:0001:1: Failed cmd 52, read reg (000000fc) ... [ 483.970000] wilc_sdio mmc0:0001:1 wlan0: ERR [wilc_wlan_deinitialize:852] failed in wlan_stop [ 483.970000] wilc_sdio mmc0:0001:1 wlan0: INFO [wilc_wlan_deinitialize]Deinitializing WILC Wlan [ 483.990000] wilc_sdio mmc0:0001:1 wlan0: INFO [wilc_wlan_deinitialize]wilc deinitialization Done [ 483.990000] Unable to handle kernel NULL pointer dereference at virtual address 00000080 [ 484.010000] pgd = c0004000 [ 484.010000] [00000080] *pgd=00000000 [ 484.010000] Internal error: Oops: 17 [#1] PREEMPT ARM [ 484.010000] Modules linked in: wilc_sdio(C) qmi_wwan option usb_wwan watchdogd(O) usb_f_rndis u_ether libcomposite [ 484.010000] CPU: 0 PID: 1205 Comm: WILC_DEBUG Tainted: G WC O 3.10.108 #6 [ 484.010000] task: c3876dc0 ti: c1d8c000 task.ti: c1d8c000 [ 484.010000] PC is at debug_thread+0x144/0x2e8 [wilc_sdio] [ 484.010000] LR is at wake_up_klogd+0x54/0x7c [ 484.010000] pc : [] lr : [] psr: 40000013 [ 484.010000] sp : c1d8df20 ip : c1d8dcf8 fp : c1d8df64 [ 484.010000] r10: 00000002 r9 : bf0439f4 r8 : bf0435ec [ 484.010000] r7 : bf043394 r6 : bf03d39e r5 : c1c9ba00 r4 : c1da92c0 [ 484.010000] r3 : 00000000 r2 : 00000001 r1 : 00000000 r0 : 00000000 [ 484.010000] Flags: nZcv IRQs on FIQs on Mode SVC_32 ISA ARM Segment kernel [ 484.010000] Control: 0005317f Table: 01f08000 DAC: 00000017 [ 484.010000] Process WILC_DEBUG (pid: 1205, stack limit = 0xc1d8c1b8) [ 484.010000] Stack: (0xc1d8df20 to 0xc1d8e000) [ 484.010000] df20: c1d8df44 c1d8df30 c0365520 c1e3dd44 c1da9350 c38c23c0 bf0315ac c1e3dd44 [ 484.010000] df40: 00000000 c1da92c0 bf0315ac 00000000 00000000 00000000 c1d8dfac c1d8df68 [ 484.010000] df60: c0038fa4 bf0315bc c1d8df94 00000000 c0040920 c1da92c0 00000000 c1d8df7c [ 484.010000] df80: c1d8df7c 00000000 c1d8df88 c1d8df88 c1e3dd44 c0038ef0 00000000 00000000 [ 484.010000] dfa0: 00000000 c1d8dfb0 c000e1f8 c0038f00 00000000 00000000 00000000 00000000 [ 484.010000] dfc0: 00000000 00000000 00000000 00000000 00000000 00000000 00000000 00000000 [ 484.010000] dfe0: 00000000 00000000 00000000 00000000 00000013 00000000 00000000 00000000 [ 484.010000] Backtrace: [ 484.010000] [] (debug_thread+0x0/0x2e8 [wilc_sdio]) from [] (kthread+0xb4/0xc0) [ 484.010000] [] (kthread+0x0/0xc0) from [] (ret_from_fork+0x14/0x3c) [ 484.010000] r7:00000000 r6:00000000 r5:c0038ef0 r4:c1e3dd44 [ 484.010000] Code: e3a0a002 e5930080 ebfffb22 e5943028 (e5930080) [ 484.040000] ---[ end trace 2a6aa3540f6dd498 ]---

When we dig into wilc driver and disassembly this kernel NULL pointer Oops，we got： debug_thread+0x144 = 0x222c+0x144=0x2370，locate to disassembly dump file:

0000222c <debug_thread>: int debug_running; int recovery_on; int wait_for_recovery; static int (void *arg) { 222c: e1a0c00d mov ip, sp 2230: e92ddff0 push {r4, r5, r6, r7, r8, r9, sl, fp, ip, lr, pc} ... for (i = 0; i < WILC_NUM_CONCURRENT_IFC; i++) wilc_mac_close(wl->vif[i]->ndev); 235c: e5943024 ldr r3, [r4, #36] ; 0x24 2360: e3a0a002 mov sl, #2 2364: e5930080 ldr r0, [r3, #128] ; 0x80 2368: ebfffb22 bl ff8 <wilc_mac_close> 236c: e5943028 ldr r3, [r4, #40] ; 0x28 2370: e5930080 ldr r0, [r3, #128] ; 0x80 // PC is at this line 2374: ebfffb1f bl ff8 <wilc_mac_close> 2378: e3a0300a mov r3, #10 for (i = WILC_NUM_CONCURRENT_IFC; i > 0; i--) { while (wilc_mac_open(wl->vif[i-1]->ndev) && --timeout) 237c: e24a2001 sub r2, sl, #1 2380: e28a9007 add r9, sl, #7

So we guess kernel NULL pointer Oops come from "wilc_mac_close(wl->vif[i]->ndev)", and we modify debug_thread() at wilc_netdev.c by adding NULL pointer check before access ndev of wl->vif[i]，below is my modify： ---- a/linux-3.10.108/drivers/staging/wilc/wilc_netdev.c +++ b/linux-3.10.108/drivers/staging/wilc/wilc_netdev.c @@ -69,9 +69,13 @@ static int debug_thread(void *arg) timeout = 10; recovery_on = 1; wait_for_recovery = 1; -- for (i = 0; i < WILC_NUM_CONCURRENT_IFC; i++) -- wilc_mac_close(wl->vif[i]->ndev); ++ for (i = 0; i < WILC_NUM_CONCURRENT_IFC; i++) { ++ printk("close: wl->vif[%d] == %p\n", i, wl->vif[i]); ++ if(wl->vif[i]) ++ wilc_mac_close(wl->vif[i]->ndev); ++ } for (i = WILC_NUM_CONCURRENT_IFC; i > 0; i--) { ++ if(wl->vif[i-1] == NULL) {printk("open: wl->vif[%d] == %p\n", i-1, wl->vif[i-1]); continue;} while (wilc_mac_open(wl->vif[i-1]->ndev) && --timeout) msleep(100);

After this modify we don't have this kernel NULL pointer Oops any more, and the wilc driver go forward with other errors, we think the wilc driver can't recover itself from wilc_sdio_cmd52-failed, at this point console shell alway be hang if we type “ifconfig wlan0” or type “ifconfig wlan0 down” or type “rmmod wilc-sdio”, the only way we can do is to reboot the whole board.

after we re-power-on or reset the whole board, wilc work well again.

Our questions: 1、This kernel NULL pointer Oops is wilc driver bug？ 2、After we modify debug_thread(void *arg) in wilc_netdev.c to avoid kernel NULL pointer Oops，but wilc_sdio_cmd52-failed remain，how can we fix this issue? 3、How to let wilc driver return to work，except reboot the whole board?

mgkiller7 avatar Jan 06 '20 15:01 mgkiller7

Finally i found out the reason why hang while type "rmmod wilc-sdio" after wilc_sdio_cmd52-failed . it is realate to a infinish loop"while (wait_for_recovery)" in wilc_send_config_pkt(), see my modify below:

int wilc_send_config_pkt(struct wilc_vif *vif, u8 mode, struct wid *wids, u32 count) { int i; int ret = 0; u32 drv = wilc_get_vif_idx(vif); ++ int cnt = 0;

    if (wait_for_recovery) {
            PRINT_INFO(vif->ndev, CORECONFIG_DBG,
                       "Host interface is suspended\n");
            while (wait_for_recovery) {
                    msleep(300);

++ cnt++; ++ if(cnt > 10) { ++ printk("[MCB] wilc_send_config_pkt wait_for_recovery timeout(3s)!!!!\n"); ++ ret = -ETIMEDOUT; ++ return ret; ++ } } ... }

But successfully rmmod wilc-sdio, the console shell frozen and can't input any cmd, but can log out some extern HW irq relate printk. So i think the kernel is soft lockup after rmmod wilc-sdio.

Anyone have a advise？

mgkiller7 avatar Jan 10 '20 08:01 mgkiller7

@mgkiller7,

I think you are not using the latest driver code and in the recent release, the changes were done to handle a crash issue in debug recovery thread. It should help with the scenario you have mentioned in the 1st point(Long time ping issue). Also, the "scheduling while atomic" warning should not be observed in the latest release. The latest release is shared at wilc_linux_15_3. I suggest verifying the behavior with the latest driver and firmware release.

A similar crash is observed in 2 ways but they are different issues.

  1. Long time ping with packet size: In this procedure, It's a temporary failure in SDIO CMD52 and recovery can result in success if the kernel Null pointer crash is avoided. Because during the reinitialized, the chance of SDIO CMD's success is more as its temporary failure. Please verify this scenario with the latest release or with the Kernel NULL pointer Oops fixes.

  2. ESD Test to wilc module: There is a continuous SDIO CMD52 failure scenario and even if the recovery code doesn't crash the driver will fail to reinitialize because of continuous CMD52 failure. For recovery code success, the CMD52 should go thorough during re-initialize otherwise no use.

Regarding the shell hang issue, I suspect its happening because debug_thread and timers are not cleaned completely when there is a failure of recovery code (because of continuous CMD52 failure-point 2).

Changes suggested in wilc_send_config_pkt() for 'while' loop termination is not required if we clear the 'wait_for_recovery' in debug_thread() function. Please try below changes and check if it helps to overcome the shell hang issue.

--- a/drivers/staging/wilc1000/wilc_netdev.c
+++ b/drivers/staging/wilc1000/wilc_netdev.c
@@ -106,6 +106,7 @@ static int debug_thread(void *arg)
			hif_drv->hif_state = HOST_IF_IDLE;
		}
		recovery_on = 0;
+		wait_for_recovery = 0;
	}
	return 0;
}
@@ -1234,6 +1235,12 @@ static int wilc_mac_close(struct net_device *ndev)
		wl->open_ifcs--;
	} else {
		PRINT_ER(ndev, "MAC close called with no opened interfaces\n");
+		if (vif->ndev) {
+			netif_stop_queue(vif->ndev);
+
+			if (!recovery_on)
+				wilc_deinit_host_int(vif->ndev);
+		}
		return 0;
	}

@@ -1318,6 +1325,14 @@ void wilc_netdev_cleanup(struct wilc *wilc)
		}
	}

+	if (debug_running) {
+		if (&wilc->debug_thread_started)
+			complete(&wilc->debug_thread_started);
+		if (wilc->debug_thread) {
+			kthread_stop(wilc->debug_thread);
+			wilc->debug_thread = NULL;
+		}
+	}
	wilc_wfi_deinit_mon_interface(wilc, false);

	flush_workqueue(wilc->hif_workqueue);

Please note these changes may help to just remove the module successfully but to make wilc module work its must to have the success of SDIO CMD.

ajaykathat avatar Jan 10 '20 16:01 ajaykathat

Thank you very much, ajaykathat Sorry for my delay. I will have a try your advise.

mgkiller7 avatar Jan 15 '20 01:01 mgkiller7

Hi， ajaykathat After add the your modify mentioned above to drivers/staging/wilc1000/wilc_netdev.c，in most cases it help to success to rmmod wilc-sdio indeed. But when i do pressure test last night（my test script loop do: insmod wilc-sdio.ko; try to connect a non-existent wifi ap; after 2mintes do ifconfig wlan0 down; rmmod wilc-sdio; sleep 10s and go on next loop）， there is a small probability of rmmod wilc-sdio failure： [12351.270000] wilc_sdio mmc0:0001:1 wlan0: INFO [wilc_netdev_cleanup]Unregistering netdev c393a000 （hang here, at 12351 Seconds(3.4Hours) after start-up）

The normal rmmod log is： [12178.330000] wilc_sdio mmc0:0001:1 wlan0: INFO [wilc_netdev_cleanup]Unregistering netdev c393a000 [12178.410000] De-Initializing Locks [12178.450000] Freeing wiphy [12178.490000] Module_exit Done. [12178.490000] at_pwr_dev: deinit [12178.500000] at_pwr_dev: unregistered [12178.510000] mmc0: card 0001 removed

This morning i observed this rmmod hang，i login my board by telnet，and run “ps” cmd， i got： 。。。 8085 root 501:00 [K_TXQ_TASK] 。。。 (i think K_TXQ_TASK may in a dead loop last for 501Mintes(8.35Hours) after rmmod wilc-sdio )

This kernel thread is “int wilc_txq_task(void *vp)” created in wilc//wilc_netdev.c wilc->txq_thread = kthread_run(wilc_txq_task, (void *)wilc, "K_TXQ_TASK");

hi, ajaykathat I found that some logs are missing when doing ifconfig wlan0 down before rmmod wilc-sdio failure. the normal ifconfig wlan0 down is : [12175.270000] wilc_sdio mmc0:0001:1 wlan0: INFO [wilc_mac_close]Mac close [12175.290000] wilc_sdio mmc0:0001:1 wlan0: INFO [wilc_mac_close]Deinitializing wilc [12175.290000] wilc_sdio mmc0:0001:1 wlan0: INFO [wilc_wlan_deinitialize]Deinitializing wilc ... [12175.310000] wilc_sdio mmc0:0001:1 wlan0: INFO [wilc_wlan_deinitialize]Disabling IRQ [12175.310000] wilc_sdio mmc0:0001:1: wilc_sdio_disable_interrupt [12175.340000] wilc_sdio mmc0:0001:1 wlan0: INFO [wilc_wlan_deinitialize]Deinitializing Threads [12175.340000] wilc_sdio mmc0:0001:1 wlan0: INFO [wlan_deinitialize_threads]Deinitializing Threads [12175.370000] wilc_sdio mmc0:0001:1 wlan0: INFO [wlan_deinitialize_threads]Deinit debug Thread [12175.390000] Exit debug thread [12175.390000] wilc_sdio mmc0:0001:1 wlan0: INFO [wlan_deinitialize_threads]Deinitializing Threads [12175.430000] wilc_sdio mmc0:0001:1 wlan0: INFO [wilc_wlan_deinitialize]Deinitializing IRQ [12175.450000] wilc_sdio mmc0:0001:1 wlan0: INFO [wilc_wlan_deinitialize]Deinitializing WILC Wlan [12175.470000] wilc_sdio mmc0:0001:1 wlan0: INFO [wilc_wlan_deinitialize]wilc deinitialization Done [12175.480000] wilc_sdio mmc0:0001:1 wlan0: INFO [wilc_mgmt_frame_register]Return since mac is closed [12175.490000] wilc_sdio mmc0:0001:1 wlan0: INFO [wilc_mgmt_frame_register]Return since mac is closed [12175.500000] wilc_sdio mmc0:0001:1 wlan0: INFO [wilc_mgmt_frame_register]Return since mac is closed [12175.510000] wilc_sdio mmc0:0001:1 wlan0: INFO [wilc_mgmt_frame_register]Return since mac is closed [12175.520000] wilc_sdio mmc0:0001:1 wlan0: INFO [wilc_mgmt_frame_register]Return since mac is closed [12175.530000] wilc_sdio mmc0:0001:1 wlan0: INFO [wilc_mgmt_frame_register]Return since mac is closed [12175.540000] wilc_sdio mmc0:0001:1 wlan0: INFO [wilc_mgmt_frame_register]Return since mac is closed

The bold log above is the missing log of doing the last ifconfig wlan0 down.

So i think static there must be dead loop in wilc_sdio_disable_interrupt() in wilc/wilc_sdio.c void wilc_sdio_disable_interrupt(struct wilc *dev) { struct sdio_func *func = container_of(dev->dev, struct sdio_func, dev); int ret;

    dev_info(&func->dev, "%s\n", __func__);

    if (sdio_intr_lock  == WILC_SDIO_HOST_IRQ_TAKEN)
            **wait_event_interruptible(sdio_intr_waitqueue,
                               sdio_intr_lock == WILC_SDIO_HOST_NO_TAKEN);**
    sdio_intr_lock  = WILC_SDIO_HOST_DIS_TAKEN;

    sdio_claim_host(func);
    ret = sdio_release_irq(func);
    if (ret < 0)
            dev_err(&func->dev, "can't release sdio_irq, err(%d)\n", ret);
    sdio_release_host(func);
    sdio_intr_lock  = WILC_SDIO_HOST_NO_TAKEN;

}

================================================================ And the "scheduling while atomic" warning is also observed with this the latest release driver&firmware of wilc_linux_15_3 in linux-3.10.108.

。。。 [ 136.480000] wilc_sdio mmc0:0001:1 wlan0: INFO [wilc_wlan_initialize]WILC Firmware Ver = WILC_WIFI_FW_REL_15_3 Build: 11606 。。。

wlan0: Trying to associa[ 144.940000] BUG: scheduling while atomic: wpa_supplicant/1064/0x00000202 te with 04:9[ 144.950000] Modules linked in:2:26:c7:3f:a wilc_sdio(C) watchdogd(O) usb_f_rndis u_ether libcomposite 8 (SSID='ASU[ 144.960000] CPU: 0 PID: 1064 Comm: wpa_supplicant Tainted: G C O 3.10.108 #8 S_A8_2G' fre[ 144.970000] Backtrace: [ 144.970000] [] (dump_backtrace+0x0/0x104) from [] (show_stack+0x18/0x1c) [ 144.980000] r7:c0aef6d8 r6:c1ebe000 r5:c0aef6d8 r4:00000000 [ 144.990000] [] (show_stack+0x0/0x1c) from [] (dump_stack+0x20/0x28) [ 145.000000] [] (dump_stack+0x0/0x28) from [] (__schedule_bug+0x48/0x60) [ 145.010000] [] (__schedule_bug+0x0/0x60) from [] (__schedule+0x60/0x3d0) [ 145.010000] r5:c0aef6d8 r4:c3877340 [ 145.020000] [] (__schedule+0x0/0x3d0) from [] (schedule+0x88/0x8c) [ 145.030000] [] (schedule+0x0/0x8c) from [] (schedule_timeout+0x14c/0x170) [ 145.030000] [] (schedule_timeout+0x0/0x170) from [] (wait_for_common+0x100/0x1a4) [ 145.040000] [] (wait_for_common+0x0/0x1a4) from [] (wait_for_completion_timeout+0x14/0x18) [ 145.050000] r9:c1ebfdb4 r8:00000008 r7:c1d6d3cc r6:c39323c0 r5:00000008 r4:c1d6d2c0 [ 145.060000] [] (wait_for_completion_timeout+0x0/0x18) from [] (cfg_set+0x100/0x144 [wilc_sdio]) [ 145.070000] [] (cfg_set+0x0/0x144 [wilc_sdio]) from [] (wilc_send_config_pkt+0x210/0x280 [wilc_sdio]) [ 145.080000] r9:bf034794 r8:00000001 r7:00000001 r6:c39323c0 r5:00000000 r4:c1ebfdb8 [ 145.090000] [] (wilc_send_config_pkt+0x0/0x280 [wilc_sdio]) from [] (wilc_frame_register+0x94/0x114 [wilc_sdio]) [ 145.100000] [] (wilc_frame_register+0x0/0x114 [wilc_sdio]) from [] (wilc_mgmt_frame_register+0x90/0xa0 [wilc_sdio]) [ 145.110000] r6:c39333a8 r5:c1d6d140 r4:c1fac8e0 [ 145.110000] [] (wilc_mgmt_frame_register+0x0/0xa0 [wilc_sdio]) from [] (cfg80211_mlme_unregister_socket+0xc4/0xf0) [ 145.120000] [] (cfg80211_mlme_unregister_socket+0x0/0xf0) from [] (nl80211_netlink_notify+0xf4/0x10c) [ 145.130000] [] (nl80211_netlink_notify+0x0/0x10c) from [] (notifier_call_chain+0x4c/0x70) [ 145.140000] [] (notifier_call_chain+0x0/0x70) from [] (__atomic_notifier_call_chain+0x40/0x54) [ 145.150000] r9:c35b6de0 r8:00000008 r7:ffffffff r6:c1ebfecc r5:00000001 r4:c0b309dc [ 145.160000] [] (__atomic_notifier_call_chain+0x0/0x54) from [] (atomic_notifier_call_chain+0x20/0x28) [ 145.170000] r7:c35b6de0 r6:c35b6dc0 r5:00000000 r4:c1e80400 [ 145.180000] [] (atomic_notifier_call_chain+0x0/0x28) from [] (netlink_release+0x148/0x284) [ 145.190000] [] (netlink_release+0x0/0x284) from [] (sock_release+0x28/0xa8) [ 145.200000] r6:c3825470 r5:00000000 r4:c35b6dc0 [ 145.200000] [] (sock_release+0x0/0xa8) from [] (sock_close+0x14/0x1c) [ 145.210000] r5:c35ba660 r4:c1f14500 [ 145.210000] [] (sock_close+0x0/0x1c) from [] (__fput+0xfc/0x224) [ 145.220000] [] (__fput+0x0/0x224) from [] (____fput+0x10/0x14) [ 145.230000] [] (____fput+0x0/0x14) from [] (task_work_run+0x90/0xac) [ 145.240000] [] (task_work_run+0x0/0xac) from [] (do_work_pending+0xbc/0xc0) [ 145.250000] r7:00000006 r6:c1ebffb0 r5:c000e308 r4:c1ebe000 [ 145.250000] [] (do_work_pending+0x0/0xc0) from [] (work_pending+0xc/0x20) [ 145.260000] r7:00000006 r6:00000002 r5:000cde70 r4:000cee28 。。。

mgkiller7 avatar Jan 15 '20 08:01 mgkiller7

Hi @mgkiller7, Thanks for the verification. Regarding the failure of 'rmmod' under stress testing issue, I don't think it's related to the suggested fixes. Please share the script to have an idea about the complete test procedure. For the overnight testing did you also enable the failure of CMD52 or CMD52 was getting successful.?

Regarding "scheduling while atomic" warning I don't see with Linux kernel 4.9.36. Let's handle that issue as part of Issue#50

ajaykathat avatar Jan 16 '20 07:01 ajaykathat

Hi, ajaykathat Thanks for your reply. I can‘t observe any CMD52 error in the whole testing. My stress testing script is: ~ # cat /nvm/wilc_scan.sh while true do # install wilc driver insmod /lib/modules/3.10.108/wilc-sdio.ko sleep 5

            # try to connect a non-existent wifi ap
            echo "ctrl_interface=/var/run/wpa_supplicant" > /tmp/wilc_wpa_supplicant.conf
            echo "update_config=1" >> /tmp/wilc_wpa_supplicant.conf
            wpa_supplicant -iwlan0 -Dnl80211 -c /tmp/wilc_wpa_supplicant.conf &
            sleep 5
            wpa_cli scan
            sleep 3
            wpa_cli scan_results
            wpa_cli -p/var/run/wpa_supplicant ap_scan 1
            wpa_cli -p/var/run/wpa_supplicant add_network
            #wpa_cli -p/var/run/wpa_supplicant set_network 0 ssid '"ASUS_A8_2G"'
            wpa_cli -p/var/run/wpa_supplicant set_network 0 ssid '"None-Exist-AP"'
            wpa_cli -p/var/run/wpa_supplicant set_network 0 key_mgmt WPA-PSK
            wpa_cli -p/var/run/wpa_supplicant set_network 0 psk '"12345678"'
            wpa_cli -p/var/run/wpa_supplicant select_network 0
            sleep 30
            #udhcpc -i wlan0 -q

            # down wlan0
            ifconfig wlan0 down

            # uninstall wilc driver
            rmmod wilc-sdio
            
            # sleep 
            sleep 2

done

=================================================================

I found that some logs are missing when doing ifconfig wlan0 down before rmmod wilc-sdio failure. the normal ifconfig wlan0 down is : [12175.270000] wilc_sdio mmc0:0001:1 wlan0: INFO [wilc_mac_close]Mac close [12175.290000] wilc_sdio mmc0:0001:1 wlan0: INFO [wilc_mac_close]Deinitializing wilc [12175.290000] wilc_sdio mmc0:0001:1 wlan0: INFO [wilc_wlan_deinitialize]Deinitializing wilc ... [12175.310000] wilc_sdio mmc0:0001:1 wlan0: INFO [wilc_wlan_deinitialize]Disabling IRQ [12175.310000] wilc_sdio mmc0:0001:1: wilc_sdio_disable_interrupt [12175.340000] wilc_sdio mmc0:0001:1 wlan0: INFO [wilc_wlan_deinitialize]Deinitializing Threads [12175.340000] wilc_sdio mmc0:0001:1 wlan0: INFO [wlan_deinitialize_threads]Deinitializing Threads [12175.370000] wilc_sdio mmc0:0001:1 wlan0: INFO [wlan_deinitialize_threads]Deinit debug Thread [12175.390000] Exit debug thread [12175.390000] wilc_sdio mmc0:0001:1 wlan0: INFO [wlan_deinitialize_threads]Deinitializing Threads [12175.430000] wilc_sdio mmc0:0001:1 wlan0: INFO [wilc_wlan_deinitialize]Deinitializing IRQ [12175.450000] wilc_sdio mmc0:0001:1 wlan0: INFO [wilc_wlan_deinitialize]Deinitializing WILC Wlan [12175.470000] wilc_sdio mmc0:0001:1 wlan0: INFO [wilc_wlan_deinitialize]wilc deinitialization Done [12175.480000] wilc_sdio mmc0:0001:1 wlan0: INFO [wilc_mgmt_frame_register]Return since mac is closed [12175.490000] wilc_sdio mmc0:0001:1 wlan0: INFO [wilc_mgmt_frame_register]Return since mac is closed [12175.500000] wilc_sdio mmc0:0001:1 wlan0: INFO [wilc_mgmt_frame_register]Return since mac is closed [12175.510000] wilc_sdio mmc0:0001:1 wlan0: INFO [wilc_mgmt_frame_register]Return since mac is closed [12175.520000] wilc_sdio mmc0:0001:1 wlan0: INFO [wilc_mgmt_frame_register]Return since mac is closed [12175.530000] wilc_sdio mmc0:0001:1 wlan0: INFO [wilc_mgmt_frame_register]Return since mac is closed [12175.540000] wilc_sdio mmc0:0001:1 wlan0: INFO [wilc_mgmt_frame_register]Return since mac is closed

The bold log above is the missing log of doing the last ifconfig wlan0 down. I can see the debug thread didn't exit yet that time, this means ifconfig wlan0 down is not success complete. So i think this uncomplete ifconfig wlan0 down cause rmmod wilc-sdio hang(mybe some locks didn't release， and ifconfig wlan0 down request it).

i think static there must be dead loop in wilc_sdio_disable_interrupt() in wilc/wilc_sdio.c void wilc_sdio_disable_interrupt(struct wilc *dev) { struct sdio_func *func = container_of(dev->dev, struct sdio_func, dev); int ret;

dev_info(&func->dev, "%s\n", __func__);

if (sdio_intr_lock  == WILC_SDIO_HOST_IRQ_TAKEN)
        **wait_event_interruptible(sdio_intr_waitqueue,
                           sdio_intr_lock == WILC_SDIO_HOST_NO_TAKEN);**
sdio_intr_lock  = WILC_SDIO_HOST_DIS_TAKEN;

sdio_claim_host(func);
ret = sdio_release_irq(func);
if (ret < 0)
        dev_err(&func->dev, "can't release sdio_irq, err(%d)\n", ret);
sdio_release_host(func);
sdio_intr_lock  = WILC_SDIO_HOST_NO_TAKEN;

}

mgkiller7 avatar Jan 16 '20 11:01 mgkiller7

@mgkiller7 during the last two years i noticed over and over again that repeated modprobe & rmmod od doing so while on high cpu load, causes the driver (or the whole system) to hang.

To give you an idea of how we handle this see:


my $driverstatus = `lsmod | grep wilc | grep -v -e "cfg80211" | awk '{print \$3}'`;
chomp($driverstatus);
TRACE("driverstatus: $driverstatus");

if ( $driverstatus eq "-1" ) {
    print("WILC driver hangs");

I've writte a bit about it in https://github.com/linux4wilc/firmware/issues/7

Your issue is rather a side effect for us because reloading the driver is what we do in order to be able to operate the WILC3000 as an AP at all.

Here as well we can force the whole thing to crash just by doing a

ping -s 65000

instead of

ping

Mateusz-Gwara avatar Jan 16 '20 14:01 Mateusz-Gwara

Hi，Mateusz-Gwara Thank you for your sharing. Your skill is good to determine whether the rmmod wilc-sdio in hang，after this hang issue happn then i can reboot my whole board to recovery. But it is not a fundamental solution to this issue.

mgkiller7 avatar Jan 17 '20 05:01 mgkiller7

But it is not a fundamental solution to this issue. ^^this is definitely true

Unfortunately the salesforce support practically stopped working with us, delaying tickets until they expire.

If you find a way to avoid a hanging module which forces the system to be rebooted, I would be very thankful for any information.

Mateusz-Gwara avatar Jan 17 '20 11:01 Mateusz-Gwara

I didn't see hang issue with "ping -s 65500" command. I am using SAMA5D4 Xplained + wilc3000 SDIO and using the latest firmware and driver. I have also tested with WILC3000 SPI and didn't observe the immediate hang issue.

@mgkiller7 If you suspect the hang happening in wilc_sdio_disable_interrupt(), can you print the value of "sdio_intr_lock" variable to confirm? Also, keep one debug print while coming out of wilc_sdio_disable_interrupt().

  1. what is the occurrence rate for this issue. Are you seeing this issue once every ~4 hrs?
  2. Do you also see this issue if you do not start the wpa_supplicant( i.e scan and connect step)?

ajaykathat avatar Jan 17 '20 11:01 ajaykathat

@ajaykathat The hang takes several days sometimes to occur. Streaming a youtube video or webcam feed usually causes problems way faster.

Mateusz-Gwara avatar Jan 17 '20 12:01 Mateusz-Gwara

Hi, Mateusz-Gwara, ajaykathat Since update to the latest wilc driver&firmware, this issue never seen again. Best regards, mgkiller7

mgkiller7 avatar Apr 02 '20 14:04 mgkiller7

@mgkiller7 Did you use the latest dev or master branch driver & fw???

Regards

Mateusz Gwara

Mateusz-Gwara avatar Apr 03 '20 11:04 Mateusz-Gwara