Hi i implemented a sample captive portal on FreeBSD 13.2-RELEASE releng/13.2-n254617-525ecfdad597 GENERIC amd64 and i used IPFW radix tables and dummynet When i log in to captive portal FreeBSD experiences Fatal trap shown below Fatal trap 12: page fault while in kernel mode cpuid = 3; apic id = 03 fault virtual address = 0x2fc fault code = supervisor read data, page not present instruction pointer = 0x20:0xffffffff810ada90 stack pointer = 0x28:0xfffffe00c23302e0 Fatal trap 12: page fault while in kernel mode frame pointer = 0x28:0xfffffe00c23302e0 cpuid = 6; apic id = 06 fault virtual address = 0x2fe fault code = supervisor read data, page not present instruction pointer = 0x20:0xffffffff810ada90 stack pointer = 0x28:0xfffffe00c23d52e0 frame pointer = 0x28:0xfffffe00c23d52e0 code segment = base rx0, limit 0xfffff, type 0x1b code segment = base rx0, limit 0xfffff, type 0x1b = DPL 0, pres 1, long 1, def32 0, gran 1 processor eflags = interrupt enabled, resume, IOPL = 0 current process = 1654 (nginx) trap number = 12 panic: page fault cpuid = 3 time = 1687084663 KDB: stack backtrace: #0 0xffffffff80c53dc5 at kdb_backtrace+0x65 #1 0xffffffff80c06741 at vpanic+0x151 #2 0xffffffff80c065e3 at panic+0x43 #3 0xffffffff810b1fa7 at trap_fatal+0x387 #4 0xffffffff810b1fff at trap_pfault+0x4f #5 0xffffffff81088e78 at calltrap+0x8 #6 0xffffffff80c9a99f at m_pullup+0x1af #7 0xffffffff821b3372 at ipfw_chk+0x1082 #8 0xffffffff821b7f2c at ipfw_check_frame+0x13c #9 0xffffffff80d429c7 at pfil_run_hooks+0x97 #10 0xffffffff80d239e4 at ether_output_frame+0x94 #11 0xffffffff80d23864 at ether_output+0x684 #12 0xffffffff80dbd736 at ip_output+0x1316 #13 0xffffffff80dd43af at tcp_output+0x1dbf #14 0xffffffff80de638d at tcp_usr_send+0x17d #15 0xffffffff80c03f54 at vn_sendfile+0x12a4 #16 0xffffffff80c04e97 at sendfile+0x117 #17 0xffffffff810b289c at amd64_syscall+0x10c Uptime: 20h5m39s # ipfw table all list --- table(test_lan), set(0) --- --- table(lan_ips_2), set(0) --- 192.168.30.1/32 0 --- table(lan_ips_5), set(0) --- 10.0.100.1/32 0 --- table(test_captiveportal), set(0) --- em3 2000 em2 5000 --- table(test_auth_up_2), set(0) --- --- table(test_auth_up_5), set(0) --- --- table(test_auth_down_2), set(0) --- --- table(test_auth_down_5), set(0) --- --- table(test_blocked_mac_2), set(0) --- --- table(test_blocked_mac_5), set(0) --- --- table(test_allowedmacup_2), set(0) --- 78:60:4e:e8:ca:f0/48 1004 5c:d1:d7:ec:a4:24/48 1004 38:1e:5b:b4:8b:e0/48 1004 88:90:47:20:ff:52/48 1004 --- table(test_allowedmacup_5), set(0) --- --- table(test_blocked_mac_all), set(0) --- --- table(test_allowedmacdown_2), set(0) --- 78:60:4e:e8:ca:f0/48 4 5c:d1:d7:ec:a4:24/48 4 38:1e:5b:b4:8b:e0/48 4 88:90:47:20:ff:52/48 4 --- table(test_allowedmacdown_5), set(0) --- # kldstat Id Refs Address Size Name 1 69 0xffffffff80200000 1f3e2d0 kernel 2 4 0xffffffff82140000 47978 ipfw.ko 3 2 0xffffffff82188000 71770 pf.ko 4 1 0xffffffff821fa000 f8a0 carp.ko 5 1 0xffffffff82600000 3c4778 zfs.ko 6 1 0xffffffff82520000 3250 ichsmb.ko 7 1 0xffffffff82524000 2180 smbus.ko 8 1 0xffffffff82527000 12520 dummynet.ko 9 1 0xffffffff8253a000 42a0 ipfw_nat.ko 10 1 0xffffffff8253f000 c852 libalias.ko 11 1 0xffffffff8254c000 2240 pflog.ko 12 1 0xffffffff8254f000 2224 speaker.ko 14 1 0xffffffff8255a000 2548 if_enc.ko 16 1 0xffffffff8258c000 52c0 ng_pppoe.ko 17 8 0xffffffff82592000 aac8 netgraph.ko 18 1 0xffffffff8259d000 39c0 ng_socket.ko 19 1 0xffffffff825a1000 43c4 ng_mppc.ko 20 1 0xffffffff825a6000 20b0 rc4.ko 21 1 0xffffffff825a9000 2398 ng_iface.ko 22 1 0xffffffff825ac000 61e8 ng_ppp.ko 23 1 0xffffffff825b3000 2138 ng_tee.ko 24 1 0xffffffff825b6000 31c8 ng_ether.ko 25 1 0xffffffff825ba000 3468 ipdivert.ko 26 1 0xffffffff825be000 2138 ng_tcpmss.ko em0@pci0:2:0:0: class=0x020000 rev=0x00 hdr=0x00 vendor=0x8086 device=0x10d3 subvendor=0x8086 subdevice=0x0000 vendor = 'Intel Corporation' device = '82574L Gigabit Network Connection' class = network subclass = ethernet
Hi, Can you try to set: sysctl kern.ipc.mb_use_ext_pgs=0
(In reply to Andrey V. Elsukov from comment #2) Yes, Thank you. I applied your sysctl tunable and captive portal worked as expected since 18 hours. may i know is that a final solution? and can it cause an another problem?
(In reply to Alfa from comment #3) This is temporary workaround, it is not yet fixed.