10.07.2018, 10:39, "Sriram" <sriram.ec@xxxxxxxxx>:
Hi,
Hello,
I m working on linux-3.12.19 kernel. In our board, we have an IPsec tunnel which carries IPv6 traffic inside
whoops! you are working with a very old and unsupported kernel version. please you are sure that your payload packages are not a larger then 512 octets.
I think that even if the kernel is old and I don't have a bug on the Linux kernel, there may be a problem with your code.
Did you set limits?
clas 8 bits, flow 20 bits, header 8 bits, hop limits 8 (255) bits.
Do different method:
For example: https://github.com/gih900/IPv6--TCP-Frag-Test-Rig
Packet structure looks like below,IPv4(OuterIP) || ESP || IPv6(Inner IP) || IPv6 Frag HDR || UDP || Payload.I have written an application which constructs Inner IPv6 packets and writes to raw socket. Through application I m generating IPv6 packets of length not more than 1340. If length of the IPv6 packet is more than 1340, application does the IPv6 fragmentation and 2 fragments are written to raw socket. If I run this application using a script continously, there is no issue observed for around 30-40 mins. After that we see kernel panic with the stack trace like below.Unable to handle kernel paging request for data at address 0x000004d0Faulting instruction address: 0xc000000000789578Oops: Kernel access of bad area, sig: 11 [#1]SMP NR_CPUS=24 CoreNet GenericModules linked in: shm(O) l1d(O) hetmgr(O) sl_reset(O) cu_led(O) airv1588(O)CPU: 7 PID: 2512 Comm: pdcpUlSockTh Tainted: G W O 3.12.19-rt30 #1task: c0000003fb42dc40 ti: c0000003ef9a0000 task.ti: c0000003ef9a0000NIP: c000000000789578 LR: c00000000078ca48 CTR: c00000000078c8f0REGS: c0000003ef9a2f10 TRAP: 0300 Tainted: G W O (3.12.19-rt30)MSR: 0000000080029000 <CE,EE,ME> CR: 24008422 XER: 20000000SOFTE: 1DEAR: 00000000000004d0, ESR: 0000000000000100GPR00: c00000000078ca48 c0000003ef9a3190 c000000000d6b3a0 c0000003ef9a33a0GPR04: 0000000000000000 c0000003ef9a3740 000000000000000a c0000003ef9f1c80GPR08: c0000003ef9f1c80 00000000000004d0 0000000000000001 000000007abc6dbbGPR12: 0000019f1a2654f8 c00000000fff7480 00000000107ce2ec 000000001075fc10GPR16: 0000000000000011 c000000000d56200 0000000000000000 000000001d8a2000GPR20: 0000000000000000 0000000000000040 c000000000e74b48 c0000003ef9a3200GPR24: c000000000cbc480 c0000003ef9a3740 c0000003ef9a33a0 c0000003ef9a3740GPR28: 000000000000000a c000000074b86000 c0000003ef9f1c80 000000000000000aNIP [c000000000789578] .xfrm_resolve_and_create_bundle+0x78/0xbb4LR [c00000000078ca48] .xfrm_bundle_lookup+0x158/0x6f8Call Trace:[c0000003ef9a3190] [c0000003ef9a32c0] 0xc0000003ef9a32c0 (unreliable)[c0000003ef9a3330] [c00000000078ca48] .xfrm_bundle_lookup+0x158/0x6f8[c0000003ef9a3400] [c0000000006d1a9c] .flow_cache_lookup+0x43c/0x4f4[c0000003ef9a34e0] [c00000000078d348] .xfrm_lookup+0x260/0x6a4[c0000003ef9a3610] [c0000000007a7edc] .ip6_dst_lookup_flow+0x8c/0xc0[c0000003ef9a36b0] [c0000000007ca3d0] .rawv6_sendmsg+0x264/0xcc8[c0000003ef9a3820] [c000000000758c18] .inet_sendmsg+0x98/0x110[c0000003ef9a38c0] [c000000000694124] .sock_sendmsg+0x94/0xf4[c0000003ef9a3a20] [c000000000694738] .___sys_sendmsg+0x340/0x350[c0000003ef9a3c20] [c0000000006990c8] .__sys_sendmsg+0x58/0xb4[c0000003ef9a3d00] [c0000000006dac94] .compat_sys_sendmsg+0x48/0x70[c0000003ef9a3d90] [c0000000006db228] .compat_sys_socketcall+0x2b4/0x310[c0000003ef9a3e30] [c0000000000005a4] syscall_exit+0x0/0x8cInstruction dump:fbc1fff0 fbe1fff8 f821fe61 7c7a1b78 7cdf3378 7c942378 7cb92b78 7cfe3b7860000000 60000000 e93a0000 3ae10070 <e9290000> 7f43d378 7e84a378 7f25cb78
I think many programs written and worked the current kernel ipv6 infrastructure and not have a root cause problems.
stability from at least version 4.10.
Kindly let me know if there is any patch available to fix this issue.If further details are required, I can post.Regards,Sriram
_______________________________________________ Kernelnewbies mailing list Kernelnewbies@xxxxxxxxxxxxxxxxx https://lists.kernelnewbies.org/mailman/listinfo/kernelnewbies
