On Sat, 30 Sep 2023, Bjoern A. Zeeb wrote:

Hi,

I just pushed

URL: https://cgit.FreeBSD.org/src/commit/?id=16e688b2a403a3dbc05de3b9f8a3132838613790
        LinuxKPI: 802.11: fix problem removing sta

into main.  Anyone on main, can you please test that the annoying
firmware crash is gone (or rather things at least improved).

I've been running a service netif restart wlan0 / service wpa_supplicant
restart wlan0 in a loop over the morning and not hit the fw crash myself
anymore.

There is at least one more bug lingering (likely currently mitigated by
other code; I have a local assert which got triggered once in the above
testing loop)) which I'll try to sort out the next 48 hours but any testing
now will be highly appreciated also for the upcoming 14.0-R.

If I can get any positive feedback I'll feel more at ease merging it all
to stable/14 and stable/13.

Feedback best here or on your open PR or the current "tracking" PR [1].

[1] https://bugs.freebsd.org/bugzilla/show_bug.cgi?id=274003





In case you wonder about the other problem;  seems to be related to
removing the node from net80211 (node_reclaim -> node_free).

I'll let you know when that is fixed.

Fatal trap 9: general protection fault while in kernel mode cpuid = 0; apic id = 00 instruction pointer = 0x20:0xffffffff80cf17a1 stack pointer = 0x28:0xfffffe009076e9b0 frame pointer = 0x28:0xfffffe009076e9c0 code segment = base 0x0, limit 0xfffff, type 0x1b = DPL 0, pres 1, long 1, def32 0, gran 1 processor eflags = interrupt enabled, resume, IOPL = 0 current process = 6700 (wpa_supplicant) rdi: fffffe00ac7c9000 rsi: fffff80001059050 rdx: 0000000000000002 rcx: fffffe008ca59c80 r8: fffff801433af948 r9: 0000000021268000 rax: deadc0dedeadc0de rbx: fffffe00ac7c9000 rbp: fffffe009076e9c0 r10: 0000000000000000 r11: 0000000000010000 r12: fffffe00ac7c9000 r13: fffffe00ac5c3638 r14: deadc0dedeadc0de r15: fffffe00ac5bd000 trap number = 9 panic: general protection fault cpuid = 0 time = 1696062479 KDB: stack backtrace: db_trace_self_wrapper() at db_trace_self_wrapper+0x2b/frame 0xfffffe009076e6f0 vpanic() at vpanic+0x132/frame 0xfffffe009076e820 panic() at panic+0x43/frame 0xfffffe009076e880 trap_fatal() at trap_fatal+0x40c/frame 0xfffffe009076e8e0 calltrap() at calltrap+0x8/frame 0xfffffe009076e8e0 --- trap 0x9, rip = 0xffffffff80cf17a1, rsp = 0xfffffe009076e9b0, rbp = 0xfffffe009076e9c0 --- node_free() at node_free+0x11/frame 0xfffffe009076e9c0 ieee80211_sta_join1() at ieee80211_sta_join1+0xbf/frame 0xfffffe009076ea10
ieee80211_sta_join() at ieee80211_sta_join+0x427/frame 0xfffffe009076ea60
ieee80211_ioctl_setmlme() at ieee80211_ioctl_setmlme+0x104/frame 
0xfffffe009076eb00
ieee80211_ioctl_set80211() at ieee80211_ioctl_set80211+0x5ce/frame 
0xfffffe009076eb70
ieee80211_ioctl() at ieee80211_ioctl+0x327/frame 0xfffffe009076ebe0
ifioctl() at ifioctl+0x940/frame 0xfffffe009076ecd0 kern_ioctl() at kern_ioctl+0x286/frame 0xfffffe009076ed30 sys_ioctl() at sys_ioctl+0x152/frame 0xfffffe009076ee00 amd64_syscall() at amd64_syscall+0x14f/frame 0xfffffe009076ef30 fast_syscall_common() at fast_syscall_common+0xf8/frame 0xfffffe009076ef30
--- syscall (54, FreeBSD ELF64, ioctl), rip = 0xf5284d980fa, rsp = 
0xf5282168fd8, rbp = 0xf5282169030 ---
KDB: enter: panic [ thread pid 6700 tid 100110 ]
Stopped at      kdb_enter+0x32: movq    $0,0xe2a6a3(%rip)


--
Bjoern A. Zeeb                                                     r15:7

Reply via email to