[LARTC] Compiling sch_tbf.c in user space

[LARTC] Compiling sch_tbf.c in user space

[vinay mathew]

Hi,
I am new to kernel programming.
My task is to route packets from different hosts
through one single token bucket filter as we are
trying to model a wireless scenario with many hosts
communicating and sharing the same bandwidth.So the
idea we came up with was to route packets from
different hosts to user space using Netfilter and then
send them through a common tbf implemented in user
space in order to model a simple bandwidth shaping
scenario.
Routing packets to user space has been easy using the
Netfitler hooks and libipq.
My problem now is that i need to implement a tbf in
user space.I was thinking if its possible to modify
the sch_tbf.c code to make it work in user space so
that i can compile it as part of my user space program
that handles the packets sent to user space thro
netfilter.Ive tried modifying this code a bit but it
gives me a huge number of errors when i try to compile
it as a simple C program in user space.
If this is not possible does any one have a simple
basic implementation of the token bucket filter that i
can use in my program to test this scenario or any
other ideas on how to implement this tbf in user
space.
Thanx a million in advance.......and im sorry i didnt
know exactly where to post the query so thats why i
posted to both lists.
thanx and kind regards
vinay


	
		
__________________________________ 
Do you Yahoo!? 
Yahoo! Mail - You care about security. So do we. 
http://promotions.yahoo.com/new_mail
_______________________________________________
LARTC mailing list / LARTC@mailman.ds9a.nl
http://mailman.ds9a.nl/mailman/listinfo/lartc HOWTO: http://lartc.org/

Compiling sch_tbf.c in user space

[vinay mathew]

Hi,
I am new to kernel programming.
My task is to route packets from different hosts
through one single token bucket filter as we are
trying to model a wireless scenario with many hosts
communicating and sharing the same bandwidth.So the
idea we came up with was to route packets from
different hosts to user space using Netfilter and then
send them through a common tbf implemented in user
space in order to model a simple bandwidth shaping
scenario.
Routing packets to user space has been easy using the
Netfitler hooks and libipq.
My problem now is that i need to implement a tbf in
user space.I was thinking if its possible to modify
the sch_tbf.c code to make it work in user space so
that i can compile it as part of my user space program
that handles the packets sent to user space thro
netfilter.Ive tried modifying this code a bit but it
gives me a huge number of errors when i try to compile
it as a simple C program in user space.
If this is not possible does any one have a simple
basic implementation of the token bucket filter that i
can use in my program to test this scenario or any
other ideas on how to implement this tbf in user
space.
Thanx a million in advance.......and im sorry i didnt
know exactly where to post the query so thats why i
posted to both lists.
thanx and kind regards
vinay


	
		
__________________________________ 
Do you Yahoo!? 
Yahoo! Mail - You care about security. So do we. 
http://promotions.yahoo.com/new_mail

Re: [LARTC] Compiling sch_tbf.c in user space

[Ed Wildgoose]

>If this is not possible does any one have a simple
>basic implementation of the token bucket filter that i
>can use in my program to test this scenario or any
>other ideas on how to implement this tbf in user
>space.
>  
>

It's not as neat, but IMQ might help here?

Ed W
_______________________________________________
LARTC mailing list / LARTC@mailman.ds9a.nl
http://mailman.ds9a.nl/mailman/listinfo/lartc HOWTO: http://lartc.org/

Re: Compiling sch_tbf.c in user space

[Ed Wildgoose]

>If this is not possible does any one have a simple
>basic implementation of the token bucket filter that i
>can use in my program to test this scenario or any
>other ideas on how to implement this tbf in user
>space.
>  
>

It's not as neat, but IMQ might help here?

Ed W
_______________________________________________
LARTC mailing list / LARTC@mailman.ds9a.nl
http://mailman.ds9a.nl/mailman/listinfo/lartc HOWTO: http://lartc.org/

Re: [LARTC] Compiling sch_tbf.c in user space

[Catalin(ux aka Dino) BOIE]

> Hi,
Hi!

> I am new to kernel programming.
> My task is to route packets from different hosts
> through one single token bucket filter as we are
> trying to model a wireless scenario with many hosts
> communicating and sharing the same bandwidth.So the

But why do you want to go to userspace for this simulation. Why not in 
kernel. If you want to simulate wireless, take a lokk at netem scheduler.

> idea we came up with was to route packets from
> different hosts to user space using Netfilter and then
> send them through a common tbf implemented in user
> space in order to model a simple bandwidth shaping
> scenario.
> Routing packets to user space has been easy using the
> Netfitler hooks and libipq.
> My problem now is that i need to implement a tbf in
> user space.I was thinking if its possible to modify
> the sch_tbf.c code to make it work in user space so
> that i can compile it as part of my user space program
> that handles the packets sent to user space thro
> netfilter.Ive tried modifying this code a bit but it
> gives me a huge number of errors when i try to compile
> it as a simple C program in user space.
> If this is not possible does any one have a simple
> basic implementation of the token bucket filter that i
> can use in my program to test this scenario or any
> other ideas on how to implement this tbf in user
> space.
> Thanx a million in advance.......and im sorry i didnt
> know exactly where to post the query so thats why i
> posted to both lists.
> thanx and kind regards
> vinay
>
>
>
>
> __________________________________
> Do you Yahoo!?
> Yahoo! Mail - You care about security. So do we.
> http://promotions.yahoo.com/new_mail
> _______________________________________________
> LARTC mailing list / LARTC@mailman.ds9a.nl
> http://mailman.ds9a.nl/mailman/listinfo/lartc HOWTO: http://lartc.org/
>

---
Catalin(ux aka Dino) BOIE
catab at deuroconsult.ro
http://kernel.umbrella.ro/
_______________________________________________
LARTC mailing list / LARTC@mailman.ds9a.nl
http://mailman.ds9a.nl/mailman/listinfo/lartc HOWTO: http://lartc.org/

Re: Compiling sch_tbf.c in user space

[Catalin(ux aka Dino) BOIE]

> Hi,
Hi!

> I am new to kernel programming.
> My task is to route packets from different hosts
> through one single token bucket filter as we are
> trying to model a wireless scenario with many hosts
> communicating and sharing the same bandwidth.So the

But why do you want to go to userspace for this simulation. Why not in 
kernel. If you want to simulate wireless, take a lokk at netem scheduler.

> idea we came up with was to route packets from
> different hosts to user space using Netfilter and then
> send them through a common tbf implemented in user
> space in order to model a simple bandwidth shaping
> scenario.
> Routing packets to user space has been easy using the
> Netfitler hooks and libipq.
> My problem now is that i need to implement a tbf in
> user space.I was thinking if its possible to modify
> the sch_tbf.c code to make it work in user space so
> that i can compile it as part of my user space program
> that handles the packets sent to user space thro
> netfilter.Ive tried modifying this code a bit but it
> gives me a huge number of errors when i try to compile
> it as a simple C program in user space.
> If this is not possible does any one have a simple
> basic implementation of the token bucket filter that i
> can use in my program to test this scenario or any
> other ideas on how to implement this tbf in user
> space.
> Thanx a million in advance.......and im sorry i didnt
> know exactly where to post the query so thats why i
> posted to both lists.
> thanx and kind regards
> vinay
>
>
>
>
> __________________________________
> Do you Yahoo!?
> Yahoo! Mail - You care about security. So do we.
> http://promotions.yahoo.com/new_mail
> _______________________________________________
> LARTC mailing list / LARTC@mailman.ds9a.nl
> http://mailman.ds9a.nl/mailman/listinfo/lartc HOWTO: http://lartc.org/
>

---
Catalin(ux aka Dino) BOIE
catab at deuroconsult.ro
http://kernel.umbrella.ro/
_______________________________________________
LARTC mailing list / LARTC@mailman.ds9a.nl
http://mailman.ds9a.nl/mailman/listinfo/lartc HOWTO: http://lartc.org/

Re: [LARTC] Compiling sch_tbf.c in user space

[vinay mathew]

--- "Catalin(ux aka Dino) BOIE" <util@deuroconsult.ro>
wrote:

> > Hi,
> Hi!
> 
> > I am new to kernel programming.
> > My task is to route packets from different hosts
> > through one single token bucket filter as we are
> > trying to model a wireless scenario with many
> hosts
> > communicating and sharing the same bandwidth.So
> the
> 
> But why do you want to go to userspace for this
> simulation. Why not in 
> kernel. If you want to simulate wireless, take a
> lokk at netem scheduler.

I have already looked at netem and am using that for
modelling parameters like delay, jitter etc.
But the situation for bandwidth modelling is not there
in netem.The reason im doin it in user space is that i
need to route all packets coming into the system
through a single tbf.Using the kernel and tc system
the tbf is attached to a particular interface and
hence i do not have a common tbf for all interfaces.I
want a common tbf to shape all the traffic before
sending the packets out on the respective interfaces.I
have just come across the idea of using IMQ devices
for this task.I believ it is possible for the packets
coming in from different interfaces to be passed thro
a common tbf attached to an IMQ before sending them
out on the final interfaces.Im just not sure if the
packets routed to thro to the imq are again routed to
the correct interface finally.Do u have any idea on
this one.I would greatly appreciate any help on this.
thanx in advance
> 
> > idea we came up with was to route packets from
> > different hosts to user space using Netfilter and
> then
> > send them through a common tbf implemented in user
> > space in order to model a simple bandwidth shaping
> > scenario.
> > Routing packets to user space has been easy using
> the
> > Netfitler hooks and libipq.
> > My problem now is that i need to implement a tbf
> in
> > user space.I was thinking if its possible to
> modify
> > the sch_tbf.c code to make it work in user space
> so
> > that i can compile it as part of my user space
> program
> > that handles the packets sent to user space thro
> > netfilter.Ive tried modifying this code a bit but
> it
> > gives me a huge number of errors when i try to
> compile
> > it as a simple C program in user space.
> > If this is not possible does any one have a simple
> > basic implementation of the token bucket filter
> that i
> > can use in my program to test this scenario or any
> > other ideas on how to implement this tbf in user
> > space.
> > Thanx a million in advance.......and im sorry i
> didnt
> > know exactly where to post the query so thats why
> i
> > posted to both lists.
> > thanx and kind regards
> > vinay
> >
> >
> >
> >
> > __________________________________
> > Do you Yahoo!?
> > Yahoo! Mail - You care about security. So do we.
> > http://promotions.yahoo.com/new_mail
> > _______________________________________________
> > LARTC mailing list / LARTC@mailman.ds9a.nl
> > http://mailman.ds9a.nl/mailman/listinfo/lartc
> HOWTO: http://lartc.org/
> >
> 
> ---
> Catalin(ux aka Dino) BOIE
> catab at deuroconsult.ro
> http://kernel.umbrella.ro/
> 


==Vinay Mathew
Technische Universität - Braunschweig,
Zimmer 41215, Rebenring 64, 38106 Braunschweig,
Deutschland.
Ph No:+4917622086705






		
__________________________________ 
Do you Yahoo!? 
All your favorites on one personal page – Try My Yahoo!
http://my.yahoo.com 
_______________________________________________
LARTC mailing list / LARTC@mailman.ds9a.nl
http://mailman.ds9a.nl/mailman/listinfo/lartc HOWTO: http://lartc.org/

Re: [LARTC] Compiling sch_tbf.c in user space

[vinay mathew]

--- "Catalin(ux aka Dino) BOIE" <util@deuroconsult.ro>
wrote:

> > Hi,
> Hi!
> 
> > I am new to kernel programming.
> > My task is to route packets from different hosts
> > through one single token bucket filter as we are
> > trying to model a wireless scenario with many
> hosts
> > communicating and sharing the same bandwidth.So
> the
> 
> But why do you want to go to userspace for this
> simulation. Why not in 
> kernel. If you want to simulate wireless, take a
> lokk at netem scheduler.

I have already looked at netem and am using that for
modelling parameters like delay, jitter etc.
But the situation for bandwidth modelling is not there
in netem.The reason im doin it in user space is that i
need to route all packets coming into the system
through a single tbf.Using the kernel and tc system
the tbf is attached to a particular interface and
hence i do not have a common tbf for all interfaces.I
want a common tbf to shape all the traffic before
sending the packets out on the respective interfaces.I
have just come across the idea of using IMQ devices
for this task.I believ it is possible for the packets
coming in from different interfaces to be passed thro
a common tbf attached to an IMQ before sending them
out on the final interfaces.Im just not sure if the
packets routed to thro to the imq are again routed to
the correct interface finally.Do u have any idea on
this one.I would greatly appreciate any help on this.
thanx in advance
> 
> > idea we came up with was to route packets from
> > different hosts to user space using Netfilter and
> then
> > send them through a common tbf implemented in user
> > space in order to model a simple bandwidth shaping
> > scenario.
> > Routing packets to user space has been easy using
> the
> > Netfitler hooks and libipq.
> > My problem now is that i need to implement a tbf
> in
> > user space.I was thinking if its possible to
> modify
> > the sch_tbf.c code to make it work in user space
> so
> > that i can compile it as part of my user space
> program
> > that handles the packets sent to user space thro
> > netfilter.Ive tried modifying this code a bit but
> it
> > gives me a huge number of errors when i try to
> compile
> > it as a simple C program in user space.
> > If this is not possible does any one have a simple
> > basic implementation of the token bucket filter
> that i
> > can use in my program to test this scenario or any
> > other ideas on how to implement this tbf in user
> > space.
> > Thanx a million in advance.......and im sorry i
> didnt
> > know exactly where to post the query so thats why
> i
> > posted to both lists.
> > thanx and kind regards
> > vinay
> >
> >
> >
> >
> > __________________________________
> > Do you Yahoo!?
> > Yahoo! Mail - You care about security. So do we.
> > http://promotions.yahoo.com/new_mail
> > _______________________________________________
> > LARTC mailing list / LARTC@mailman.ds9a.nl
> > http://mailman.ds9a.nl/mailman/listinfo/lartc
> HOWTO: http://lartc.org/
> >
> 
> ---
> Catalin(ux aka Dino) BOIE
> catab at deuroconsult.ro
> http://kernel.umbrella.ro/
> 


=====
Vinay Mathew
Technische Universität - Braunschweig,
Zimmer 41215, Rebenring 64, 38106 Braunschweig,
Deutschland.
Ph No:+4917622086705






		
__________________________________ 
Do you Yahoo!? 
All your favorites on one personal page – Try My Yahoo!
http://my.yahoo.com 

Re: [LARTC] Compiling sch_tbf.c in user space

[vinay mathew]

Thanx for the idea Ed.Can u clarify another point for
me  if u have an idea about it cos im new to this.I
understand that i can route packets from different
interfaces to a common tbf attached to an IMQ
device.Now i need to know whether the packets after
passing through this IMQ device are then automatically
forwarded to the correct actual hardware interfaces to
be further delivered to their respective destinations.
Example: my router has 4 interface cards
eth0,1,2,3.Packets from 0 are destined to 2 and
packets from 1 to 3.However all packets are first to
be routed through a common shaping mechanism which im
assuming now to be a tbf attached to the IMQ
device.This is the scenario i am currently trying to
model by sending the packets to user space and thro a
tbf implemented in user space which i think is turning
out to be a little complex.Thanx again for the help
kind regards
vinay
--- Ed Wildgoose <lists@wildgooses.com> wrote:

> 
> >If this is not possible does any one have a simple
> >basic implementation of the token bucket filter
> that i
> >can use in my program to test this scenario or any
> >other ideas on how to implement this tbf in user
> >space.
> >  
> >
> 
> It's not as neat, but IMQ might help here?
> 
> Ed W
> 



		
__________________________________ 
Do you Yahoo!? 
Yahoo! Mail - Helps protect you from nasty viruses. 
http://promotions.yahoo.com/new_mail
_______________________________________________
LARTC mailing list / LARTC@mailman.ds9a.nl
http://mailman.ds9a.nl/mailman/listinfo/lartc HOWTO: http://lartc.org/

Re: [LARTC] Compiling sch_tbf.c in user space

[vinay mathew]

Thanx for the idea Ed.Can u clarify another point for
me  if u have an idea about it cos im new to this.I
understand that i can route packets from different
interfaces to a common tbf attached to an IMQ
device.Now i need to know whether the packets after
passing through this IMQ device are then automatically
forwarded to the correct actual hardware interfaces to
be further delivered to their respective destinations.
Example: my router has 4 interface cards
eth0,1,2,3.Packets from 0 are destined to 2 and
packets from 1 to 3.However all packets are first to
be routed through a common shaping mechanism which im
assuming now to be a tbf attached to the IMQ
device.This is the scenario i am currently trying to
model by sending the packets to user space and thro a
tbf implemented in user space which i think is turning
out to be a little complex.Thanx again for the help
kind regards
vinay
--- Ed Wildgoose <lists@wildgooses.com> wrote:

> 
> >If this is not possible does any one have a simple
> >basic implementation of the token bucket filter
> that i
> >can use in my program to test this scenario or any
> >other ideas on how to implement this tbf in user
> >space.
> >  
> >
> 
> It's not as neat, but IMQ might help here?
> 
> Ed W
> 



		
__________________________________ 
Do you Yahoo!? 
Yahoo! Mail - Helps protect you from nasty viruses. 
http://promotions.yahoo.com/new_mail

Re: [LARTC] Compiling sch_tbf.c in user space

[Catalin(ux aka Dino) BOIE]

[-- Attachment #1: Type: TEXT/PLAIN, Size: 3269 bytes --]

> I have already looked at netem and am using that for
> modelling parameters like delay, jitter etc.
> But the situation for bandwidth modelling is not there
> in netem.The reason im doin it in user space is that i
> need to route all packets coming into the system
> through a single tbf.Using the kernel and tc system
> the tbf is attached to a particular interface and
> hence i do not have a common tbf for all interfaces.I
> want a common tbf to shape all the traffic before
> sending the packets out on the respective interfaces.I
> have just come across the idea of using IMQ devices
> for this task.I believ it is possible for the packets
> coming in from different interfaces to be passed thro
> a common tbf attached to an IMQ before sending them
> out on the final interfaces.Im just not sure if the
> packets routed to thro to the imq are again routed to
> the correct interface finally.Do u have any idea on
> this one.I would greatly appreciate any help on this.
> thanx in advance

You can take a look at Jamal's actions, especialy redirect.
I didn't use IMQ, sorry.

http://lwn.net/Articles/121407/

>>
>>> idea we came up with was to route packets from
>>> different hosts to user space using Netfilter and
>> then
>>> send them through a common tbf implemented in user
>>> space in order to model a simple bandwidth shaping
>>> scenario.
>>> Routing packets to user space has been easy using
>> the
>>> Netfitler hooks and libipq.
>>> My problem now is that i need to implement a tbf
>> in
>>> user space.I was thinking if its possible to
>> modify
>>> the sch_tbf.c code to make it work in user space
>> so
>>> that i can compile it as part of my user space
>> program
>>> that handles the packets sent to user space thro
>>> netfilter.Ive tried modifying this code a bit but
>> it
>>> gives me a huge number of errors when i try to
>> compile
>>> it as a simple C program in user space.
>>> If this is not possible does any one have a simple
>>> basic implementation of the token bucket filter
>> that i
>>> can use in my program to test this scenario or any
>>> other ideas on how to implement this tbf in user
>>> space.
>>> Thanx a million in advance.......and im sorry i
>> didnt
>>> know exactly where to post the query so thats why
>> i
>>> posted to both lists.
>>> thanx and kind regards
>>> vinay
>>>
>>>
>>>
>>>
>>> __________________________________
>>> Do you Yahoo!?
>>> Yahoo! Mail - You care about security. So do we.
>>> http://promotions.yahoo.com/new_mail
>>> _______________________________________________
>>> LARTC mailing list / LARTC@mailman.ds9a.nl
>>> http://mailman.ds9a.nl/mailman/listinfo/lartc
>> HOWTO: http://lartc.org/
>>>
>>
>> ---
>> Catalin(ux aka Dino) BOIE
>> catab at deuroconsult.ro
>> http://kernel.umbrella.ro/
>>
>
>
> =====
> Vinay Mathew
> Technische Universität - Braunschweig,
> Zimmer 41215, Rebenring 64, 38106 Braunschweig,
> Deutschland.
> Ph No:+4917622086705
>
>
>
>
>
>
>
> __________________________________
> Do you Yahoo!?
> All your favorites on one personal page – Try My Yahoo!
> http://my.yahoo.com
>

---
Catalin(ux aka Dino) BOIE
catab at deuroconsult.ro
http://kernel.umbrella.ro/

Re: Compiling sch_tbf.c in user space

[Catalin(ux aka Dino) BOIE]

[-- Attachment #1: Type: TEXT/PLAIN, Size: 3269 bytes --]

> I have already looked at netem and am using that for
> modelling parameters like delay, jitter etc.
> But the situation for bandwidth modelling is not there
> in netem.The reason im doin it in user space is that i
> need to route all packets coming into the system
> through a single tbf.Using the kernel and tc system
> the tbf is attached to a particular interface and
> hence i do not have a common tbf for all interfaces.I
> want a common tbf to shape all the traffic before
> sending the packets out on the respective interfaces.I
> have just come across the idea of using IMQ devices
> for this task.I believ it is possible for the packets
> coming in from different interfaces to be passed thro
> a common tbf attached to an IMQ before sending them
> out on the final interfaces.Im just not sure if the
> packets routed to thro to the imq are again routed to
> the correct interface finally.Do u have any idea on
> this one.I would greatly appreciate any help on this.
> thanx in advance

You can take a look at Jamal's actions, especialy redirect.
I didn't use IMQ, sorry.

http://lwn.net/Articles/121407/

>>
>>> idea we came up with was to route packets from
>>> different hosts to user space using Netfilter and
>> then
>>> send them through a common tbf implemented in user
>>> space in order to model a simple bandwidth shaping
>>> scenario.
>>> Routing packets to user space has been easy using
>> the
>>> Netfitler hooks and libipq.
>>> My problem now is that i need to implement a tbf
>> in
>>> user space.I was thinking if its possible to
>> modify
>>> the sch_tbf.c code to make it work in user space
>> so
>>> that i can compile it as part of my user space
>> program
>>> that handles the packets sent to user space thro
>>> netfilter.Ive tried modifying this code a bit but
>> it
>>> gives me a huge number of errors when i try to
>> compile
>>> it as a simple C program in user space.
>>> If this is not possible does any one have a simple
>>> basic implementation of the token bucket filter
>> that i
>>> can use in my program to test this scenario or any
>>> other ideas on how to implement this tbf in user
>>> space.
>>> Thanx a million in advance.......and im sorry i
>> didnt
>>> know exactly where to post the query so thats why
>> i
>>> posted to both lists.
>>> thanx and kind regards
>>> vinay
>>>
>>>
>>>
>>>
>>> __________________________________
>>> Do you Yahoo!?
>>> Yahoo! Mail - You care about security. So do we.
>>> http://promotions.yahoo.com/new_mail
>>> _______________________________________________
>>> LARTC mailing list / LARTC@mailman.ds9a.nl
>>> http://mailman.ds9a.nl/mailman/listinfo/lartc
>> HOWTO: http://lartc.org/
>>>
>>
>> ---
>> Catalin(ux aka Dino) BOIE
>> catab at deuroconsult.ro
>> http://kernel.umbrella.ro/
>>
>
>
> =====
> Vinay Mathew
> Technische Universität - Braunschweig,
> Zimmer 41215, Rebenring 64, 38106 Braunschweig,
> Deutschland.
> Ph No:+4917622086705
>
>
>
>
>
>
>
> __________________________________
> Do you Yahoo!?
> All your favorites on one personal page – Try My Yahoo!
> http://my.yahoo.com
>

---
Catalin(ux aka Dino) BOIE
catab at deuroconsult.ro
http://kernel.umbrella.ro/