[c-nsp] CSCek62005

[David Freedman]

Oh, thats strange, this bug was implicated for a crash caused by 
removing an active prefix list that was not used for the label 
allocation feature (just for normal BGP), this was
12.2(31)SB10 on 7206VXR NPE-G1

#sh ver | in ROM
  router uptime is 4 days, 17 minutes
System returned to ROM by bus error at PC 0x61FA34DC, address 0xB0D0B0D 
at 16:53:13 BST Fri May 9 2008

#show region | in 0xA|0xB
  0xA0000000  0xADFFFFFF   234881024  Local  R/W    main:(main_k1)

#sh stack | begin Traceb
-Traceback= 61FA34DC 610A1A64 60D43AA4 6089EDF0 608BA068 609C16F4 609C16E0
$0 : 00000000, AT : 630D0000, v0 : 00000000, v1 : 00000044
a0 : 0B0D0B0D, a1 : 0B0D0B0D, a2 : 651FE6FC, a3 : 00000248
t0 : 00000018, t1 : 3400FF01, t2 : 3400E100, t3 : FFFF00FF
t4 : 60A0CC40, t5 : 6323ED68, t6 : 6323ED64, t7 : 6323ED60
s0 : 662E01D4, s1 : 662C83C0, s2 : 662C2674, s3 : 00000000
s4 : 662C84C0, s5 : 62ED0000, s6 : 00000001, s7 : 62E30000
t8 : 6323EDC0, t9 : 00000000, k0 : 72136AAC, k1 : 609F6640
gp : 630DB964, sp : 72754148, s8 : 63619BC0, ra : 610A1A64
EPC  : 61FA34DC, ErrorEPC : BFC018D4, SREG     : 3400FF03
MDLO : 00000002, MDHI     : 00000000, BadVaddr : 0B0D0B0D
DATA_START : 0x61FC70C0
Cause 00000008 (Code 0x2): TLB (load or instruction fetch) exception

Dave.



Rodney Dunn wrote:
> I've never seen it.
> 
> It appears to be part of the new local label prefix filter feature of
> some sort.
> 
> ie:
> 
> 76k-123-3#config t
> Enter configuration commands, one per line.  End with CNTL/Z.
> 76k-123-3(config)#ip prefix
> 76k-123-3(config)#ip prefix-list p0 permit 1.1.1.1/32
> 76k-123-3(config)#ip prefix-list p0 permit 1.1.1.2/32
> 76k-123-3(config)#end                                
> 76k-123-3#config t         
> 76k-123-3(config)#mpls ldp label 
> 76k-123-3(config-ldp-lbl)#allocate ?
>   global  Specify global Routing/Forwarding instance
> 
> 76k-123-3(config-ldp-lbl)#allocate global ?
>   host-routes  allocate local label for host routes only
>   prefix-list  Specify a prefix list for local label filtering
>   
> 
> 76k-123-3(config-ldp-lbl)#allocate global prefix-list p0 
> 76k-123-3(config-ldp-lbl)#end
> 
> 76k-123-3#
> 76k-123-3#sh ip prefix
> ip prefix-list p0: 2 entries
>    seq 5 permit 1.1.1.1/32
>    seq 10 permit 1.1.1.2/32
> 
> 
> 76k-123-3#config t
> Enter configuration commands, one per line.  End with CNTL/Z.
> 76k-123-3(config)#no mpls ldp label 
> 76k-123-3(config)#end
> 76k-123-3#sh ip prefix
> ip prefix-list p0: 2 entries
>    seq 5 permit 1.1.1.1/32
>    seq 10 permit 1.1.1.2/32
> ip prefix-list p2: 1 entries
>    seq 5 permit 18.18.18.18/32
> ip prefix-list p4: 1 entries
>    seq 5 deny 0.0.0.0/0
> ip prefix-list p9: 1 entries
>    seq 5 deny 0.0.0.0/0 le 1
> 76k-123-3#config t
> Enter configuration commands, one per line.  End with CNTL/Z.
> 76k-123-3(config)#no ip prefix-list p0
> 76k-123-3(config)#end
> 76k-123-3#sh ip prefix-list 
> .....
> 
> crash
> 
> 
> It's fixed in 12.0S branch but hasn't been picked up anywhere later.
> 
> Rodney
> 
> On Tue, May 13, 2008 at 03:32:04PM +0100, David Freedman wrote:
>> Does anybody from cisco here know what exact conditions trigger this and 
>> why there appears to be no "first-fixed-in" candidate for 12.0SY ?
>> 
>> Thanks,
>> 
>> Dave.
>> 
>> _______________________________________________
>> cisco-nsp mailing list  cisco-nsp at puck.nether.net
>> https://puck.nether.net/mailman/listinfo/cisco-nsp
>> archive at http://puck.nether.net/pipermail/cisco-nsp/
> _______________________________________________
> cisco-nsp mailing list  cisco-nsp at puck.nether.net
> https://puck.nether.net/mailman/listinfo/cisco-nsp
> archive at http://puck.nether.net/pipermail/cisco-nsp/
>