Уже второй раз после сбоя по питанию перестает работать один и тот же SG-17R. При этом остается доступным SSH-подключение, а вот демоны, стартующие после dropbear или падают в core dump или просто зависают. Перепрошивка помогает восстановить работоспособность.
В первом случае расследование проводить не было возможности - нужно было срочно восстановить работу.
Во втором случае убиение процесса svd_if вручную по SSH позволило продолжить загрузку без проблем. Правда, работоспособность VF-каналов после этого не проверялась.
В обоих случаях были предупреждения о сбоях CRC на jffs2. Очень похоже что именно она и портится.
На всякий случай логи загрузки со второго случая.
1Jan 1 07:00:36 (none) syslog.info syslogd started: BusyBox v1.1.2
2Jan 1 07:00:36 (none) user.notice kernel: klogd started: BusyBox v1.1.2 (2013.08.16-11:19+0000)
3Jan 1 07:00:36 (none) user.notice kernel: Linux version 2.6.16 (df@dfsig) (gcc version 4.1.0) #1 Fri Aug 16 18:34:49 NOVT 2013
4Jan 1 07:00:36 (none) user.warn kernel: CPU revision is: 0001800b
5Jan 1 07:00:36 (none) user.warn kernel: ADM5120 board setup
6Jan 1 07:00:36 (none) user.warn kernel: Altera firmware load: fail or device not present
7Jan 1 07:00:36 (none) user.warn kernel: Determined physical RAM map:
8Jan 1 07:00:36 (none) user.warn kernel: memory: 01d0f000 @ 002f1000 (usable)
9Jan 1 07:00:36 (none) user.debug kernel: On node 0 totalpages: 8192
10Jan 1 07:00:36 (none) user.debug kernel: DMA zone: 8192 pages, LIFO batch:1
11Jan 1 07:00:36 (none) user.debug kernel: DMA32 zone: 0 pages, LIFO batch:0
12Jan 1 07:00:36 (none) user.debug kernel: Normal zone: 0 pages, LIFO batch:0
13Jan 1 07:00:36 (none) user.debug kernel: HighMem zone: 0 pages, LIFO batch:0
14Jan 1 07:00:36 (none) user.warn kernel: Built 1 zonelists
15Jan 1 07:00:36 (none) user.notice kernel: Kernel command line: root=/dev/mtdblock2 rootfstype=jffs2 console=ttyS0
16Jan 1 07:00:36 (none) user.warn kernel: Primary instruction cache 8kB, physically tagged, 2-way, linesize 16 bytes.
17Jan 1 07:00:36 (none) user.warn kernel: Primary data cache 8kB, 2-way, linesize 16 bytes.
18Jan 1 07:00:36 (none) user.warn kernel: Synthesized TLB refill handler (20 instructions).
19Jan 1 07:00:36 (none) user.warn kernel: Synthesized TLB load handler fastpath (32 instructions).
20Jan 1 07:00:36 (none) user.warn kernel: Synthesized TLB store handler fastpath (32 instructions).
21Jan 1 07:00:36 (none) user.warn kernel: Synthesized TLB modify handler fastpath (31 instructions).
22Jan 1 07:00:36 (none) user.warn kernel: PID hash table entries: 256 (order: 8, 4096 bytes)
23Jan 1 07:00:36 (none) user.warn kernel: CPU clock: 175MHz
24Jan 1 07:00:36 (none) user.warn kernel: Dentry cache hash table entries: 8192 (order: 3, 32768 bytes)
25Jan 1 07:00:36 (none) user.warn kernel: Inode-cache hash table entries: 4096 (order: 2, 16384 bytes)
26Jan 1 07:00:36 (none) user.info kernel: Memory: 29424k/29756k available (2253k kernel code, 316k reserved, 382k data, 272k init, 0k highmem)
27Jan 1 07:00:36 (none) user.debug kernel: Calibrating delay loop... 169.47 BogoMIPS (lpj=84736)
28Jan 1 07:00:36 (none) user.warn kernel: Mount-cache hash table entries: 512
29Jan 1 07:00:36 (none) user.warn kernel: Checking for 'wait' instruction... available.
30Jan 1 07:00:36 (none) user.info kernel: NET: Registered protocol family 16
31Jan 1 07:00:36 (none) user.warn kernel: System has PCI BIOS
32Jan 1 07:00:36 (none) user.warn kernel: ^IPCI slot#5 IRQ enabled
33Jan 1 07:00:36 (none) user.warn kernel: TC classifier action (bugs to netdev@vger.kernel.org cc hadi@cyberus.ca)
34Jan 1 07:00:36 (none) user.info kernel: devfs: 2004-01-31 Richard Gooch (rgooch@atnf.csiro.au)
35Jan 1 07:00:36 (none) user.info kernel: devfs: boot_options: 0x1
36Jan 1 07:00:36 (none) user.info kernel: JFFS2 version 2.2. (NAND) (SUMMARY) (C) 2001-2003 Red Hat, Inc.
37Jan 1 07:00:36 (none) user.info kernel: Initializing Cryptographic API
38Jan 1 07:00:36 (none) user.info kernel: io scheduler noop registered
39Jan 1 07:00:36 (none) user.info kernel: io scheduler cfq registered (default)
40Jan 1 07:00:36 (none) user.info kernel: ADM5120 LED & GPIO driver
41Jan 1 07:00:36 (none) user.info kernel: adm5120_wdt: Watchdog Timer version 0.1
42Jan 1 07:00:36 (none) user.notice kernel: ADM5120 UART driver. Port nr=2
43Jan 1 07:00:36 (none) user.info kernel: ttyS0 at I/O 0x12600000 (irq = 1) is a ADM5120
44Jan 1 07:00:36 (none) user.info kernel: ttyS1 at I/O 0x12800000 (irq = 2) is a ADM5120
45Jan 1 07:00:36 (none) user.info kernel: PPP generic driver version 2.4.2
46Jan 1 07:00:36 (none) user.info kernel: PPP Deflate Compression module registered
47Jan 1 07:00:36 (none) user.info kernel: PPP BSD Compression module registered
48Jan 1 07:00:36 (none) user.info kernel: PPP MPPE Compression module registered
49Jan 1 07:00:36 (none) user.info kernel: NET: Registered protocol family 24
50Jan 1 07:00:36 (none) user.info kernel: HDLC support module revision 1.18
51Jan 1 07:00:36 (none) user.info kernel: Cronyx Ltd, Synchronous PPP and CISCO HDLC (c) 1994
52Jan 1 07:00:36 (none) user.info kernel: Linux port (c) 1998 Building Number Three Ltd & Jan "Yenya" Kasprzak.
53Jan 1 07:00:36 (none) user.info kernel: NAND device: Manufacturer ID: 0xec, Chip ID: 0x75 (Samsung NAND 32MiB 3,3V 8-bit)
54Jan 1 07:00:36 (none) user.warn kernel: Bad block table not found for chip 0
55Jan 1 07:00:36 (none) user.warn kernel: Bad block table not found for chip 0
56Jan 1 07:00:36 (none) user.info kernel: Scanning device for bad blocks
57Jan 1 07:00:36 (none) user.warn kernel: Bad eraseblock 0 at 0x00000000
58Jan 1 07:00:36 (none) user.warn kernel: Bad eraseblock 1965 at 0x01eb4000
59Jan 1 07:00:36 (none) user.debug kernel: Bad block table written to 0x01ffc000, version 0x01
60Jan 1 07:00:36 (none) user.debug kernel: Bad block table written to 0x01ff8000, version 0x01
61Jan 1 07:00:36 (none) user.notice kernel: Creating 3 MTD partitions on "NAND 32MiB 3,3V 8-bit":
62Jan 1 07:00:36 (none) user.notice kernel: 0x00000000-0x00020000 : "Boot Partition"
63Jan 1 07:00:36 (none) user.notice kernel: 0x00020000-0x00220000 : "Kernel partition"
64Jan 1 07:00:36 (none) user.notice kernel: 0x00220000-0x02000000 : "Root partition"
65Jan 1 07:00:36 (none) user.info kernel: NET: Registered protocol family 2
66Jan 1 07:00:36 (none) user.warn kernel: IP route cache hash table entries: 512 (order: -1, 2048 bytes)
67Jan 1 07:00:36 (none) user.warn kernel: TCP established hash table entries: 2048 (order: 1, 8192 bytes)
68Jan 1 07:00:36 (none) user.warn kernel: TCP bind hash table entries: 2048 (order: 1, 8192 bytes)
69Jan 1 07:00:36 (none) user.info kernel: TCP: Hash tables configured (established 2048 bind 2048)
70Jan 1 07:00:36 (none) user.info kernel: TCP reno registered
71Jan 1 07:00:36 (none) user.warn kernel: ip_tables: (C) 2000-2006 Netfilter Core Team
72Jan 1 07:00:36 (none) user.info kernel: TCP bic registered
73Jan 1 07:00:36 (none) user.info kernel: TCP cubic registered
74Jan 1 07:00:36 (none) user.info kernel: TCP westwood registered
75Jan 1 07:00:36 (none) user.info kernel: TCP highspeed registered
76Jan 1 07:00:36 (none) user.info kernel: TCP htcp registered
77Jan 1 07:00:36 (none) user.info kernel: Initializing IPsec netlink socket
78Jan 1 07:00:36 (none) user.info kernel: NET: Registered protocol family 1
79Jan 1 07:00:36 (none) user.info kernel: NET: Registered protocol family 17
80Jan 1 07:00:36 (none) user.info kernel: NET: Registered protocol family 15
81Jan 1 07:00:36 (none) user.warn kernel: Load linux Bridge driver (VLAN capable)
82Jan 1 07:00:36 (none) user.notice kernel: Bridge firewalling registered
83Jan 1 07:00:36 (none) user.info kernel: 802.1Q VLAN Support v1.8 Ben Greear <greearb@candelatech.com>
84Jan 1 07:00:36 (none) user.info kernel: All bugs added by David S. Miller <davem@redhat.com>
85Jan 1 07:00:36 (none) user.warn kernel: Empty flash at 0x01dd8004 ends at 0x01dd81e8
86Jan 1 07:00:36 (none) user.warn kernel: Empty flash at 0x01ddc004 ends at 0x01ddc1e8
87Jan 1 07:00:36 (none) user.warn kernel: VFS: Mounted root (jffs2 filesystem) readonly.
88Jan 1 07:00:36 (none) user.info kernel: Mounted devfs on /dev
89Jan 1 07:00:36 (none) user.info kernel: Freeing unused kernel memory: 272k freed
90Jan 1 07:00:36 (none) user.warn kernel: Algorithmics/MIPS FPU Emulator v1.5
91Jan 1 07:00:36 (none) user.warn kernel: .
92Jan 1 07:00:36 (none) user.warn kernel: ADM5120 Switch Module Init V1.3
93Jan 1 07:00:36 (none) user.warn kernel: ADM5120 MAC: 00:11:22:33:44:55
94Jan 1 07:00:42 (none) user.info kernel: sgatab

0x55:0x9c) board found in slot [2]
95Jan 1 07:00:42 (none) user.warn kernel: PCI: Enabling device 0000:00:02.0 (0000 -> 0002)
96Jan 1 07:00:42 (none) user.info kernel: sgatab: id=9c at bus - 0x00 func - 0x0
97Jan 1 07:00:42 (none) user.info kernel: sgatab: irq 6, subsystem id 0x2, memory addr 0x11400000
98Jan 1 07:00:49 (none) user.notice kernel: Load DSLAM switch control driver
99Jan 1 07:00:50 (none) user.notice kernel: Unload DSLAM switch control driver
100Jan 1 07:00:57 (none) authpriv.info dropbear[771]: Running in background
101Jan 1 07:01:06 (none) user.notice kernel: JFFS2 notice: (60) read_dnode: header CRC failed on node at 0x1ca43c0: read 0xffffffff, calculated 0xb7c33014
102Jan 1 07:02:32 (none) authpriv.info dropbear[869]: Child connection from 10.99.192.200:31949
103Jan 1 07:02:37 (none) authpriv.notice dropbear[869]: password auth succeeded for 'root' from 10.99.192.200:31949