search

raspberrypi / linux

Kernel source tree for Raspberry Pi-provided kernel builds. Issues unrelated to the linux kernel should be posted on the community forum at https://forums.raspberrypi.com/
Other
11.02k stars 4.95k forks source link

[RPi 4 B][4.19.y] WiFi AP makes Kernel panic #3114

Closed kamarya closed 5 years ago

kamarya commented 5 years ago

System Raspberry Pi 4 B Linux Kernel Branch: 4.19.y Linux Kernel Version: 4.19.58 SHA1 : d5dc848c982dff2e020f294e384447efe6ea6617 Distribution: Yocto Poky (meta-raspberrypi)

Context of the issue

Running hostapd on the on-board WiFi interface as an access point. The AP works fine in the beginning and the traffic passes through. However after a few minutes RPi freezes and I get kernel panics about hostapd (see logs). Before having the panic about hostapd I get these interesting lines:

[  979.419611] bcm2835-dma fe007000.dma: swiotlb buffer is full (sz: 1536 bytes)
[  979.426928] bcm2835-dma fe007000.dma: DMA: Out of SW-IOMMU space for 1536 bytes
[  979.434559] mmc-bcm2835 fe300000.mmcnr: dma_map_sg returned zero length
[  989.661972] mmc1: Timeout waiting for hardware interrupt.
[  995.168392] ***** psta->aid (54) out of bounds

I also get kernel messages at the boot time complaining about cfg80211 (line 752 in sme.c) This might be related to the fix of @agherzan (see SHA1 above) and it might still need more fixes. I have not tested this with other versions.

Kernel logs

root@raspberrypi4-64:~# [  965.706974] WARNING: CPU: 0 PID: 158 at net/wireless/sme.c:752 __cfg80211_connect_result+0x2fc/0x328 [cfg80211]
[  965.717220] Modules linked in: nfc bnep bluetooth ecdh_generic ipv6 iptable_filter xt_state xt_conntrack iptable_nat ipt_MASQUERADE nf_nat_ipv4 nf_nat nf_conntrack nf_defrag_ipv6 nf_defrag_ipv4 ip_tables x_tables 88XXau(O) brcmfmac brcmutil vc4 sha256_generic bcm2835_v4l2(C) bcm2835_mmal_vchiq(C) v4l2_common cfg80211 videobuf2_vmalloc snd_soc_core videobuf2_memops videobuf2_v4l2 videobuf2_common snd_compress videodev snd_pcm_dmaengine rfkill snd_pcm media snd_timer snd vc_sm_cma(C) vchiq(C) uio_pdrv_genirq uio fixed
[  965.763812] CPU: 0 PID: 158 Comm: kworker/u8:2 Tainted: G        WC O      4.19.58 #1
[  965.771756] Hardware name: Raspberry Pi 4 Model B Rev 1.1 (DT)
[  965.777737] Workqueue: cfg80211 cfg80211_event_work [cfg80211]
[  965.783662] pstate: 40000005 (nZcv daif -PAN -UAO)
[  965.788589] pc : __cfg80211_connect_result+0x2fc/0x328 [cfg80211]
[  965.794836] lr : __cfg80211_connect_result+0x1d4/0x328 [cfg80211]
[  965.801016] sp : ffffff800a633cc0
[  965.804376] x29: ffffff800a633cc0 x28: ffffff8008c46000 
[  965.809774] x27: ffffff800915bcb8 x26: ffffffc07a0f6fb8 
[  965.815172] x25: dead000000000100 x24: dead000000000200 
[  965.820568] x23: ffffff800a633d08 x22: ffffff8008c48688 
[  965.825965] x21: ffffffc07cdae000 x20: ffffffc07a346c00 
[  965.831361] x19: ffffffc079bd8c18 x18: ffffffc077c49694 
[  965.836757] x17: 0000000000000000 x16: 0000000000000000 
[  965.842154] x15: 0000000000000003 x14: 0100081000087f00 
[  965.847550] x13: 0000000000000000 x12: 0000000000000028 
[  965.852945] x11: 0101010101010101 x10: 7f7f7f7f7f7f7f7f 
[  965.858340] x9 : fefefefefefefeff x8 : 0000000000000001 
[  965.863736] x7 : 0000000000000000 x6 : ffffff8008c48000 
[  965.869131] x5 : 0000000000000000 x4 : ffffff800888acb0 
[  965.874526] x3 : ffffff8008c48dd0 x2 : 0000000000000000 
[  965.879922] x1 : 0000000000000000 x0 : 0000000000000000 
[  965.885317] Call trace:
[  965.887862]  __cfg80211_connect_result+0x2fc/0x328 [cfg80211]
[  965.893751]  cfg80211_process_wdev_events+0x120/0x188 [cfg80211]
[  965.899903]  cfg80211_process_rdev_events+0x40/0x98 [cfg80211]
[  965.905876]  cfg80211_event_work+0x24/0x38 [cfg80211]
[  965.911005]  process_one_work+0x1c4/0x420
[  965.915071]  worker_thread+0x50/0x408
[  965.918788]  kthread+0x108/0x138
[  965.922066]  ret_from_fork+0x10/0x1c
[  965.925690] ---[ end trace c74340692d12f570 ]---
[  979.419611] bcm2835-dma fe007000.dma: swiotlb buffer is full (sz: 1536 bytes)
[  979.426928] bcm2835-dma fe007000.dma: DMA: Out of SW-IOMMU space for 1536 bytes
[  979.434559] mmc-bcm2835 fe300000.mmcnr: dma_map_sg returned zero length
[  989.661972] mmc1: Timeout waiting for hardware interrupt.
[  995.168392] ***** psta->aid (54) out of bounds
[ 1088.478020] INFO: task hostapd:336 blocked for more than 120 seconds.
[ 1088.484597]       Tainted: G        WC O      4.19.58 #1
[ 1088.490037] "echo 0 > /proc/sys/kernel/hung_task_timeout_secs" disables this message.
[ 1088.498043] hostapd         D    0   336      1 0x00000000
[ 1088.503661] Call trace:
[ 1088.506206]  __switch_to+0xa8/0xf0
[ 1088.509668]  __schedule+0x254/0x858
[ 1088.513235]  schedule+0x3c/0xa0
[ 1088.516489]  __mmc_claim_host+0xb8/0x200
[ 1088.520500]  sdio_claim_host+0x30/0x50
[ 1088.524392]  brcmf_sdio_bus_txctl+0x120/0x1f0 [brcmfmac]
[ 1088.529838]  brcmf_proto_bcdc_msg.isra.0+0xd0/0xe8 [brcmfmac]
[ 1088.535761]  brcmf_proto_bcdc_query_dcmd+0x94/0x1b8 [brcmfmac]
[ 1088.541733]  brcmf_fil_cmd_data+0x8c/0x180 [brcmfmac]
[ 1088.546945]  brcmf_fil_iovar_data_get+0x78/0x128 [brcmfmac]
[ 1088.552650]  brcmf_cfg80211_get_station+0xb0/0x528 [brcmfmac]
[ 1088.558615]  nl80211_get_station+0xac/0x2d0 [cfg80211]
[ 1088.563877]  genl_family_rcv_msg+0x18c/0x350
[ 1088.568258]  genl_rcv_msg+0xa0/0xd8
[ 1088.571821]  netlink_rcv_skb+0x60/0x120
[ 1088.575761]  genl_rcv+0x3c/0x50
[ 1088.578984]  netlink_unicast+0x188/0x200
[ 1088.583012]  netlink_sendmsg+0x168/0x340
[ 1088.587018]  sock_sendmsg+0x34/0x50
[ 1088.590606]  ___sys_sendmsg+0x274/0x2a8
[ 1088.594520]  __sys_sendmsg+0x7c/0xd0
[ 1088.598196]  __arm64_sys_sendmsg+0x2c/0x38
[ 1088.602377]  el0_svc_common+0x88/0x1a8
[ 1088.606230]  el0_svc_handler+0x38/0x78
[ 1088.610055]  el0_svc+0x8/0xc
[ 1088.613025] INFO: task kworker/u8:3:395 blocked for more than 120 seconds.
[ 1088.620066]       Tainted: G        WC O      4.19.58 #1
[ 1088.625475] "echo 0 > /proc/sys/kernel/hung_task_timeout_secs" disables this message.
[ 1088.633475] kworker/u8:3    D    0   395      2 0x00000028
[ 1088.639143] Workqueue: cfg80211 cfg80211_event_work [cfg80211]
[ 1088.645089] Call trace:
[ 1088.647593]  __switch_to+0xa8/0xf0
[ 1088.651069]  __schedule+0x254/0x858
[ 1088.654633]  schedule+0x3c/0xa0
[ 1088.657824]  schedule_preempt_disabled+0x1c/0x30
[ 1088.662533]  __mutex_lock.isra.0+0x1a4/0x4f8
[ 1088.666891]  __mutex_lock_slowpath+0x24/0x30
[ 1088.671247]  mutex_lock+0x48/0x58
[ 1088.674633]  rtnl_lock+0x1c/0x28
[ 1088.677968]  cfg80211_event_work+0x1c/0x38 [cfg80211]
[ 1088.683136]  process_one_work+0x1c4/0x420
[ 1088.687227]  worker_thread+0x50/0x408
[ 1088.690961]  kthread+0x108/0x138
[ 1088.694259]  ret_from_fork+0x10/0x1c
[ 1088.697896] INFO: task kworker/u8:4:514 blocked for more than 120 seconds.
[ 1088.704894]       Tainted: G        WC O      4.19.58 #1
[ 1088.710301] "echo 0 > /proc/sys/kernel/hung_task_timeout_secs" disables this message.
[ 1088.718266] kworker/u8:4    D    0   514      2 0x00000028
[ 1088.723899] Workqueue: brcmf_wq/mmc1:0001:1 brcmf_sdio_dataworker [brcmfmac]
[ 1088.731080] Call trace:
[ 1088.733567]  __switch_to+0xa8/0xf0
[ 1088.737061]  __schedule+0x254/0x858
[ 1088.740621]  schedule+0x3c/0xa0
[ 1088.743834]  schedule_timeout+0x324/0x448
[ 1088.747927]  wait_for_common+0xc8/0x188
[ 1088.751842]  wait_for_completion+0x28/0x38
[ 1088.756023]  mmc_wait_for_req_done+0x30/0x138
[ 1088.760468]  mmc_wait_for_req+0xb0/0x108
[ 1088.764497]  brcmf_sdiod_sglist_rw+0x260/0x560 [brcmfmac]
[ 1088.770023]  brcmf_sdiod_send_pkt+0x104/0x120 [brcmfmac]
[ 1088.775462]  brcmf_sdio_dataworker+0x17a8/0x2218 [brcmfmac]
[ 1088.781138]  process_one_work+0x1c4/0x420
[ 1088.785230]  worker_thread+0x50/0x408
[ 1088.788966]  kthread+0x108/0x138
[ 1088.792266]  ret_from_fork+0x10/0x1c
[ 1088.795916] INFO: task kworker/0:2:560 blocked for more than 120 seconds.
[ 1088.802824]       Tainted: G        WC O      4.19.58 #1
[ 1088.808233] "echo 0 > /proc/sys/kernel/hung_task_timeout_secs" disables this message.
[ 1088.816195] kworker/0:2     D    0   560      2 0x00000028
[ 1088.821796] Workqueue: events linkwatch_event
[ 1088.826243] Call trace:
[ 1088.828729]  __switch_to+0xa8/0xf0
[ 1088.832203]  __schedule+0x254/0x858
[ 1088.835765]  schedule+0x3c/0xa0
[ 1088.838976]  schedule_preempt_disabled+0x1c/0x30
[ 1088.843687]  __mutex_lock.isra.0+0x1a4/0x4f8
[ 1088.848042]  __mutex_lock_slowpath+0x24/0x30
[ 1088.852398]  mutex_lock+0x48/0x58
[ 1088.855786]  rtnl_lock+0x1c/0x28
[ 1088.859083]  linkwatch_event+0x14/0x40
[ 1088.862908]  process_one_work+0x1c4/0x420
[ 1088.866992]  worker_thread+0x50/0x408
[ 1088.870731]  kthread+0x108/0x138
[ 1088.874036]  ret_from_fork+0x10/0x1c
[ 1209.310017] INFO: task hostapd:336 blocked for more than 120 seconds.
[ 1209.316582]       Tainted: G        WC O      4.19.58 #1
[ 1209.322003] "echo 0 > /proc/sys/kernel/hung_task_timeout_secs" disables this message.
[ 1209.329986] hostapd         D    0   336      1 0x00000000
[ 1209.335604] Call trace:
[ 1209.338118]  __switch_to+0xa8/0xf0
[ 1209.341577]  __schedule+0x254/0x858
[ 1209.345142]  schedule+0x3c/0xa0
[ 1209.348354]  __mmc_claim_host+0xb8/0x200
[ 1209.352361]  sdio_claim_host+0x30/0x50
[ 1209.356223]  brcmf_sdio_bus_txctl+0x120/0x1f0 [brcmfmac]
[ 1209.361667]  brcmf_proto_bcdc_msg.isra.0+0xd0/0xe8 [brcmfmac]
[ 1209.367549]  brcmf_proto_bcdc_query_dcmd+0x94/0x1b8 [brcmfmac]
[ 1209.373520]  brcmf_fil_cmd_data+0x8c/0x180 [brcmfmac]
[ 1209.378694]  brcmf_fil_iovar_data_get+0x78/0x128 [brcmfmac]
[ 1209.384399]  brcmf_cfg80211_get_station+0xb0/0x528 [brcmfmac]
[ 1209.390329]  nl80211_get_station+0xac/0x2d0 [cfg80211]
[ 1209.395570]  genl_family_rcv_msg+0x18c/0x350
[ 1209.399927]  genl_rcv_msg+0xa0/0xd8
[ 1209.403495]  netlink_rcv_skb+0x60/0x120
[ 1209.407412]  genl_rcv+0x3c/0x50
[ 1209.410616]  netlink_unicast+0x188/0x200
[ 1209.414619]  netlink_sendmsg+0x168/0x340
[ 1209.418623]  sock_sendmsg+0x34/0x50
[ 1209.422185]  ___sys_sendmsg+0x274/0x2a8
[ 1209.426098]  __sys_sendmsg+0x7c/0xd0
[ 1209.429729]  __arm64_sys_sendmsg+0x2c/0x38
[ 1209.433910]  el0_svc_common+0x88/0x1a8
[ 1209.437737]  el0_svc_handler+0x38/0x78
[ 1209.441573]  el0_svc+0x8/0xc
[ 1209.444535] INFO: task kworker/u8:3:395 blocked for more than 120 seconds.
[ 1209.451548]       Tainted: G        WC O      4.19.58 #1
[ 1209.456957] "echo 0 > /proc/sys/kernel/hung_task_timeout_secs" disables this message.
[ 1209.464924] kworker/u8:3    D    0   395      2 0x00000028
[ 1209.470591] Workqueue: cfg80211 cfg80211_event_work [cfg80211]
[ 1209.476537] Call trace:
[ 1209.479047]  __switch_to+0xa8/0xf0
[ 1209.482525]  __schedule+0x254/0x858
[ 1209.486087]  schedule+0x3c/0xa0
[ 1209.489278]  schedule_preempt_disabled+0x1c/0x30
[ 1209.493988]  __mutex_lock.isra.0+0x1a4/0x4f8
[ 1209.498345]  __mutex_lock_slowpath+0x24/0x30
[ 1209.502703]  mutex_lock+0x48/0x58
[ 1209.506091]  rtnl_lock+0x1c/0x28
[ 1209.509427]  cfg80211_event_work+0x1c/0x38 [cfg80211]
[ 1209.514578]  process_one_work+0x1c4/0x420
[ 1209.518668]  worker_thread+0x50/0x408
[ 1209.522406]  kthread+0x108/0x138
[ 1209.525686]  ret_from_fork+0x10/0x1c
[ 1209.529337] INFO: task kworker/2:0:402 blocked for more than 120 seconds.
[ 1209.536247]       Tainted: G        WC O      4.19.58 #1
[ 1209.541655] "echo 0 > /proc/sys/kernel/hung_task_timeout_secs" disables this message.
[ 1209.549623] kworker/2:0     D    0   402      2 0x00000028
[ 1209.555290] Workqueue: ipv6_addrconf addrconf_verify_work [ipv6]
[ 1209.561418] Call trace:
[ 1209.563926]  __switch_to+0xa8/0xf0
[ 1209.567413]  __schedule+0x254/0x858
[ 1209.570974]  schedule+0x3c/0xa0
[ 1209.574186]  schedule_preempt_disabled+0x1c/0x30
[ 1209.578896]  __mutex_lock.isra.0+0x1a4/0x4f8
[ 1209.583251]  __mutex_lock_slowpath+0x24/0x30
[ 1209.587608]  mutex_lock+0x48/0x58
[ 1209.590993]  rtnl_lock+0x1c/0x28
[ 1209.594347]  addrconf_verify_work+0x14/0x28 [ipv6]
[ 1209.599231]  process_one_work+0x1c4/0x420
[ 1209.603319]  worker_thread+0x50/0x408
[ 1209.607058]  kthread+0x108/0x138
[ 1209.610358]  ret_from_fork+0x10/0x1c
[ 1209.614014] INFO: task kworker/u8:4:514 blocked for more than 120 seconds.
[ 1209.621023]       Tainted: G        WC O      4.19.58 #1
[ 1209.626431] "echo 0 > /proc/sys/kernel/hung_task_timeout_secs" disables this message.
[ 1209.634395] kworker/u8:4    D    0   514      2 0x00000028
[ 1209.640031] Workqueue: brcmf_wq/mmc1:0001:1 brcmf_sdio_dataworker [brcmfmac]
[ 1209.647214] Call trace:
[ 1209.649701]  __switch_to+0xa8/0xf0
[ 1209.653179]  __schedule+0x254/0x858
[ 1209.656744]  schedule+0x3c/0xa0
[ 1209.659954]  schedule_timeout+0x324/0x448
[ 1209.664046]  wait_for_common+0xc8/0x188
[ 1209.667960]  wait_for_completion+0x28/0x38
[ 1209.672142]  mmc_wait_for_req_done+0x30/0x138
[ 1209.676585]  mmc_wait_for_req+0xb0/0x108
[ 1209.680614]  brcmf_sdiod_sglist_rw+0x260/0x560 [brcmfmac]
[ 1209.686143]  brcmf_sdiod_send_pkt+0x104/0x120 [brcmfmac]
[ 1209.691582]  brcmf_sdio_dataworker+0x17a8/0x2218 [brcmfmac]
[ 1209.697259]  process_one_work+0x1c4/0x420
[ 1209.701349]  worker_thread+0x50/0x408
[ 1209.705088]  kthread+0x108/0x138
[ 1209.708388]  ret_from_fork+0x10/0x1c
[ 1209.712038] INFO: task kworker/0:2:560 blocked for more than 120 seconds.
[ 1209.718945]       Tainted: G        WC O      4.19.58 #1
[ 1209.724356] "echo 0 > /proc/sys/kernel/hung_task_timeout_secs" disables this message.
[ 1209.732335] kworker/0:2     D    0   560      2 0x00000028
[ 1209.737940] Workqueue: events linkwatch_event
[ 1209.742400] Call trace:
[ 1209.744886]  __switch_to+0xa8/0xf0
[ 1209.748358]  __schedule+0x254/0x858
[ 1209.751920]  schedule+0x3c/0xa0
[ 1209.755128]  schedule_preempt_disabled+0x1c/0x30
[ 1209.759838]  __mutex_lock.isra.0+0x1a4/0x4f8
[ 1209.764193]  __mutex_lock_slowpath+0x24/0x30
[ 1209.768549]  mutex_lock+0x48/0x58
[ 1209.771934]  rtnl_lock+0x1c/0x28
[ 1209.775230]  linkwatch_event+0x14/0x40
[ 1209.779054]  process_one_work+0x1c4/0x420
[ 1209.783143]  worker_thread+0x50/0x408
[ 1209.786880]  kthread+0x108/0x138
[ 1209.790178]  ret_from_fork+0x10/0x1c

Warning at the boot pointing to net/wireless/sme.c:752 __cfg80211_connect_result

[   17.137453] WARNING: CPU: 3 PID: 59 at net/wireless/sme.c:752 __cfg80211_connect_result+0x2fc/0x328 [cfg80211]
[   17.147611] Modules linked in: ip_tables x_tables 88XXau(O) brcmfmac brcmutil bcm2835_v4l2(C) sha256_generic vc4 bcm2835_mmal_vchiq(C) v4l2_common videobuf2_vmalloc cfg80211 snd_soc_core videobuf2_memops videobuf2_v4l2 videobuf2_common videodev snd_compress snd_pcm_dmaengine rfkill snd_pcm media snd_timer vc_sm_cma(C) snd vchiq(C) uio_pdrv_genirq uio fixed
[   17.179770] CPU: 3 PID: 59 Comm: kworker/u8:1 Tainted: G         C O      4.19.58 #1
[   17.187626] Hardware name: Raspberry Pi 4 Model B Rev 1.1 (DT)
[   17.193620] Workqueue: cfg80211 cfg80211_event_work [cfg80211]
[   17.199546] pstate: 20000005 (nzCv daif -PAN -UAO)
[   17.204474] pc : __cfg80211_connect_result+0x2fc/0x328 [cfg80211]
[   17.210723] lr : __cfg80211_connect_result+0x1d4/0x328 [cfg80211]
[   17.216903] sp : ffffff800915bcc0
[   17.220263] x29: ffffff800915bcc0 x28: ffffff8008c46000 
[   17.225662] x27: ffffffc07d4d0070 x26: ffffffc07d4d0020 
[   17.231059] x25: dead000000000100 x24: dead000000000200 
[   17.236456] x23: ffffff800915bd08 x22: ffffff8008c48688 
[   17.241854] x21: ffffffc07aaf2000 x20: ffffffc07c664000 
[   17.247250] x19: ffffffc07c662218 x18: 0000000000000000 
[   17.252645] x17: 0000000000000000 x16: 0000000000000002 
[   17.258041] x15: 0000000000000000 x14: 6f9dcd0f5a2c0001 
[   17.263436] x13: 0000000000000040 x12: 0000000000000228 
[   17.268832] x11: 0000000000000000 x10: 0000000000001920 
[   17.274226] x9 : ffffff800915bb10 x8 : ffffffc07cf15380 
[   17.279623] x7 : 0000000000000000 x6 : ffffff8008c48000 
[   17.285019] x5 : 0000000000000000 x4 : ffffff80080d2820 
[   17.290413] x3 : ffffff8008c48dd0 x2 : 0000000000000000 
[   17.295808] x1 : 0000000000000000 x0 : 0000000000000000 
[   17.301204] Call trace:
[   17.303748]  __cfg80211_connect_result+0x2fc/0x328 [cfg80211]
[   17.309637]  cfg80211_process_wdev_events+0x120/0x188 [cfg80211]
[   17.315789]  cfg80211_process_rdev_events+0x40/0x98 [cfg80211]
[   17.321762]  cfg80211_event_work+0x24/0x38 [cfg80211]
[   17.326892]  process_one_work+0x1c4/0x420
[   17.330958]  worker_thread+0x50/0x408
[   17.334676]  kthread+0x108/0x138
[   17.337954]  ret_from_fork+0x10/0x1c
[   17.341579] ---[ end trace 4d1e1b8a1f49bad1 ]---

The same issue with version 4.19.60 (SHA1 20565c94ca877a099a3c9c5fa39a0380f3d16491) and the warning message at the boot.

[   17.224367] WARNING: CPU: 1 PID: 7 at net/wireless/sme.c:752 __cfg80211_connect_result+0x2fc/0x328 [cfg80211]
[   17.234434] Modules linked in: iptable_nat ipt_MASQUERADE nf_nat_ipv4 nf_nat nf_conntrack nf_defrag_ipv6 nf_defrag_ipv4 ip_tables x_tables 88XXau(O) brcmfmac brcmutil bcm2835_v4l2(C) sha256_generic bcm2835_mmal_vchiq(C) vc4 v4l2_common videobuf2_vmalloc cfg80211 videobuf2_memops videobuf2_v4l2 snd_soc_core videobuf2_common videodev snd_compress rfkill snd_pcm_dmaengine snd_pcm media snd_timer vc_sm_cma(C) snd vchiq(C) uio_pdrv_genirq uio fixed
[   17.274460] CPU: 1 PID: 7 Comm: kworker/u8:0 Tainted: G         C O      4.19.60 #1
[   17.282229] Hardware name: Raspberry Pi 4 Model B Rev 1.1 (DT)
[   17.288211] Workqueue: cfg80211 cfg80211_event_work [cfg80211]
[   17.294137] pstate: 40000005 (nZcv daif -PAN -UAO)
[   17.299064] pc : __cfg80211_connect_result+0x2fc/0x328 [cfg80211]
[   17.305311] lr : __cfg80211_connect_result+0x1d4/0x328 [cfg80211]
[   17.311491] sp : ffffff800805bcc0
[   17.314850] x29: ffffff800805bcc0 x28: ffffff8008c46000 
[   17.320249] x27: ffffffc07d4d0070 x26: ffffffc07d4d0020 
[   17.325646] x25: dead000000000100 x24: dead000000000200 
[   17.331042] x23: ffffff800805bd08 x22: ffffff8008c48688 
[   17.336437] x21: ffffffc07a769000 x20: ffffffc07c6b4800 
[   17.341832] x19: ffffffc07c095618 x18: 0000000000000000 
[   17.347227] x17: 0000000000000000 x16: 0000000000000000 
[   17.352622] x15: 0000000000000000 x14: 0100081000087f00 
[   17.358017] x13: 0000000000000000 x12: 0000000000000020 
[   17.363413] x11: 0101010101010101 x10: ffffffff7f7fff7f 
[   17.368809] x9 : fefefebf7965a607 x8 : 0000000000000001 
[   17.374204] x7 : ffffffc07d5a2c00 x6 : ffffff8008c48000 
[   17.379600] x5 : 0000000000000000 x4 : ffffff800888def0 
[   17.384995] x3 : ffffff8008c48dd0 x2 : 0000000000000000 
[   17.390389] x1 : 0000000000000000 x0 : 0000000000000000 
[   17.395784] Call trace:
[   17.398331]  __cfg80211_connect_result+0x2fc/0x328 [cfg80211]
[   17.404222]  cfg80211_process_wdev_events+0x120/0x188 [cfg80211]
[   17.410374]  cfg80211_process_rdev_events+0x40/0x98 [cfg80211]
[   17.416346]  cfg80211_event_work+0x24/0x38 [cfg80211]
[   17.421476]  process_one_work+0x1c4/0x420
[   17.425541]  worker_thread+0x50/0x408
[   17.429260]  kthread+0x108/0x138
[   17.432536]  ret_from_fork+0x10/0x1c
[   17.436160] ---[ end trace 955367a198e4a3fd ]---
ED6E0F17 commented 5 years ago

This is interesting: you seem to have hit issue #1988 in a new way, and I REALLY want to get that one closed.

It might be safe to close this as a duplicate of that one, since I am fairly sure that the wifi issue is not related to your broken 64bit dma

kamarya commented 5 years ago

Any idea about why the kernel freezes ? and about hostapd timeout panic message repeating after the freeze?

I am not sure where these come from.

ED6E0F17 commented 5 years ago

kamarya: The DMA fails 14 seconds after the wifi "cfg80211_connect_result" issue in the above logs, but you suggest that these Warnings happen more often than the "Hung Task." - The "swiotlb buffer is full" message is the one you should focus on, and check in your dmesg that the DMA is being set up properly.

kamarya commented 5 years ago

I have built again the kernel with the latest SHA1 9798c94b34db4a32707f7f501616dbcbd088a696 on Aug 21 2019. Broadcom WiFi in AP mode stops working after a while due to the same issue.

yaroslavros commented 5 years ago

https://github.com/raspberrypi/linux/pull/3164 is not merged yet. Can you manually apply https://github.com/raspberrypi/linux/pull/3164/commits/9dfbe9f500e85d2f751a240f6144785f237d795a and https://github.com/raspberrypi/linux/pull/3164/commits/1395119b339e946fb87f4ff9da2c85c15947734d to give it another try?

kamarya commented 5 years ago

3164 is not merged yet. Can you manually apply 9dfbe9f and 1395119 to give it another try?

I will give it a try as soon as I am back home :)

kamarya commented 5 years ago

@yaroslavros after applying the proposed patch in #3164 on top of 9798c94 the issue has not been reproduced yet. The fix looks promising and I will keep testing it.

agherzan commented 5 years ago

We should be able to close this issue now that the DMA leaks PR was merged.