Re: [mpls-linux-general] Bug about number of POPs

[James R. L. <jl...@mi...>]

I think there is a limitation of 8 instructions per insegment  Maybe this
should be dynamic or a compile time option.  The specific case of N POPs =
should
probably be handled by a single instruction.  Thsi would probably work
better for you (I think the LSR MIB has a NPOP on the insegment right?)

Thanks for filing a bug report!

On Fri, Aug 22, 2003 at 11:39:45AM +0200, Stefan Winter wrote:
> Hello,
>=20
> while implementing an snmp agent or mpls-linux, I have come across an a=
nnoying=20
> limitation in mpls-linux. If you generate a new InSegment with more tha=
n 7=20
> POPs (mplsadm2 -A -I gen:11111:0 -i pop:pop:pop:.....:peek), mplsadm2 g=
oes=20
> wild. If the number of POPs is in [8:9], the -i directive is simply ign=
ored=20
> and the new segment just reads "POP PEEK". If the number of POPs is >=3D=
10,=20
> mplsadm2 is aborted with SIGSEGV. I use mpls-linux 1.172.
> Even worse, I embedded some of the mplsadm2 code in my app, and then it=
 did=20
> not SEGV, but overwrote memory in its vicinity which led to really weir=
d=20
> errors.
> I have seen this behaviour on 2 independent installations (on the secon=
d one,=20
> even 8 and 9 POPs led to SIGSEGV).
> Can someone verify this bug? I=B4d open a tracker item on SourceForge t=
hen.
>=20
> Greetings,
>=20
> Stefan Winter
>=20
>=20
>=20
> -------------------------------------------------------
> This SF.net email is sponsored by: VM Ware
> With VMware you can run multiple operating systems on a single machine.
> WITHOUT REBOOTING! Mix Linux / Windows / Novell virtual machines
> at the same time. Free trial click here:http://www.vmware.com/wl/offer/=
358/0
> _______________________________________________
> mpls-linux-general mailing list
> mpl...@li...
> https://lists.sourceforge.net/lists/listinfo/mpls-linux-general

--=20
James R. Leu