Bogus bind 4.9.7-0 ?

[Kai Szymanski <kai netwave de>]

Hello.

On one of our nameserver bind (redhat 4.2) with latest update to 4.9.7-0
installed. The named crashes several time with the following log:

Sep 29 22:22:17 kernel: kfree of non-kmalloced memory: 03568414, next= 020f1000, order=5
Sep 29 22:22:17 kernel: kfree of non-kmalloced memory: 03568414, next= 020f1000, order=5
Sep 29 22:22:17 kernel: general protection: 0000
Sep 29 22:22:17 kernel: CPU:    0
Sep 29 22:22:17 kernel: EIP:    0010:[<020f1414>]
Sep 29 22:22:17 kernel: EFLAGS: 00010206
Sep 29 22:22:17 kernel: eax: 020f1414   ebx: 03568414   ecx: 00000000   edx: 00000000
Sep 29 22:22:17 kernel: esi: 00000124   edi: 00000000   ebp: 78ae44de   esp: 05f95d54
Sep 29 22:22:17 kernel: ds: 0018   es: 0018   fs: 002b   gs: 002b   ss: 0018
Sep 29 22:22:17 kernel: Process named-xfer (pid: 10173, process nr: 34, stackpage=05f95000)
Sep 29 22:22:17 kernel: Stack: 00138e8b 03568414 03568414 036220a8 0013918b 03568414 036220a8 036220a8 
Sep 29 22:22:17 kernel:        03568414 00000000 00148792 036220a8 00000000 03568414 0362382c 33809d64 
Sep 29 22:22:17 kernel:        03623848 03623701 00000000 000000ff 00000019 000043dd 2a80a2c2 00000000 
Sep 29 22:22:17 kernel: Call Trace: [sock_wfree+35/44] [kfree_skb+183/244] [tcp_ack+1382/2244] [tcp_rcv+2051/2404] [ip_rcv+1007/1312] [net_bh+252/284] [do_bottom_half+59/96] 
Sep 29 22:22:17 kernel:        [handle_bottom_half+11/24] [kfree_skb+171/244] [cleanup_rbuf+83/148] [tcp_recvmsg+1002/1036] [inet_recvmsg+114/136] [sys_recv+250/272] [sys_socketcall+525/732] [system_call+85/124] 
Sep 29 22:22:17 kernel: Code: 10 88 56 03 0c 4c 8b 00 00 00 00 00 00 00 00 00 50 66 dc 07 
Sep 29 22:22:17 kernel: Aiee, killing interrupt handler
Sep 29 22:22:20 kernel: send_head NULL in tcp_retransmit_time
Sep 29 22:22:22 kernel: invalid operand: 0000
Sep 29 22:22:22 kernel: CPU:    0
Sep 29 22:22:22 kernel: EIP:    0010:[<00000017>]
Sep 29 22:22:22 kernel: EFLAGS: 00010202
Sep 29 22:22:22 kernel: eax: 00000004   ebx: 03568414   ecx: 036054fc   edx: 00000206
Sep 29 22:22:22 kernel: esi: 00ec2c14   edi: 00000000   ebp: 00000001   esp: 00f18f00
Sep 29 22:22:22 kernel: ds: 0018   es: 0018   fs: 002b   gs: 002b   ss: 0018
Sep 29 22:22:22 kernel: Process sendmail (pid: 386, process nr: 32, stackpage=00f18000)
Sep 29 22:22:22 kernel: Stack: 00f18f04 00146919 03568414 0362374c 00ec2c0c 00ec2ccc 00145277 03568414 
Sep 29 22:22:22 kernel:        00000000 00ec2c0c 0000000a 00000000 001468d6 00ec2c0c 00ec2c0c 00ec3b90 
Sep 29 22:22:22 kernel:        00ec3b90 001507d5 00ec2c0c 00000000 00ec3b00 00000000 00136ca0 00ec3b90 
Sep 29 22:22:22 kernel: Call Trace: [tcp_close+165/536] [tcp_close_pending+87/104] [tcp_close+98/536] [inet_release+97/108] [sock_release+92/156] [sock_close+37/44] [__fput+28/64] 
Sep 29 22:22:22 kernel:        [close_fp+76/92] [sys_close+68/80] [system_call+85/124] 
Sep 29 22:22:22 kernel: Code: f0 4c e1 00 f0 6f ef 00 f0 a5 fe 00 f0 87 e9 00 f0 6f ef 00 

Some informations about the software:

	Red Hat 4.2 with latest patches from ftp.redhat.com
	Kernel 2.0.35
	bind & Utils 4.9.7-0 (round about 700 domains, most of the as
			     secondary)

Is this a know bug of bind ? How can i solve the hangs ?

Thanks for your help.

Best regards,
  Kai Szymanski.