mkiss in ubuntu kernels?

mkiss in ubuntu kernels?

[Collin Hockey]

Hi everyone, and thanks in advance for any help.

I've been working for the past several days on getting a Kantronics
9612 paired with an iCom IC-910 for packet communication over KISS
working using the linux AX.25 stack. I have successfully set up
kissattach and related utilities to be able to send data back and
forth at 1200 baud.

I also need to communicate at 9600 baud. For the Kantronics TNC, this
requires talking to the radio over Port 2. My current understanding is
that packet routing is done using parts of the KISS protocol and that
this is achieved by using mkiss to split the physical serial port into
two pseudo ttys, which I can then use kissattach on separately to talk
to each port.

All of the documentation I have been able to find though assumes the
old style psuedo-tty setup, where master and slave pairs are numbered,
and plenty are pre-loaded by the kernel for use by user programs.
However, at least in my kernel (2.6.28), the client program is
supposed to request the single pseudo-tty master file, and the kernel
creates a new slave for each connection. There tends to be a pile of
these in my pseudo-tty slave directory, so I never know which one to
connect to. We do need to automate and script most of these
connections, so I do need to reliably be able to tell which slave tty
I need to connect to.

Hopefully I've explained the problem well enough to be understood, and
I would appreciate any help or suggestions you might have.

Thanks again,
Collin Hockey

Re: mkiss in ubuntu kernels?

[Alan Crosswell]

You only kissattach once;  if the TNC supports mkiss then the port is  
encoded in the kiss header for each packet.
/a


On Sep 10, 2009, at 6:44 PM, Collin Hockey <hockeyc@umich.edu> wrote:

> Hi everyone, and thanks in advance for any help.
>
> I've been working for the past several days on getting a Kantronics
> 9612 paired with an iCom IC-910 for packet communication over KISS
> working using the linux AX.25 stack. I have successfully set up
> kissattach and related utilities to be able to send data back and
> forth at 1200 baud.
>
> I also need to communicate at 9600 baud. For the Kantronics TNC, this
> requires talking to the radio over Port 2. My current understanding is
> that packet routing is done using parts of the KISS protocol and that
> this is achieved by using mkiss to split the physical serial port into
> two pseudo ttys, which I can then use kissattach on separately to talk
> to each port.
>
> All of the documentation I have been able to find though assumes the
> old style psuedo-tty setup, where master and slave pairs are numbered,
> and plenty are pre-loaded by the kernel for use by user programs.
> However, at least in my kernel (2.6.28), the client program is
> supposed to request the single pseudo-tty master file, and the kernel
> creates a new slave for each connection. There tends to be a pile of
> these in my pseudo-tty slave directory, so I never know which one to
> connect to. We do need to automate and script most of these
> connections, so I do need to reliably be able to tell which slave tty
> I need to connect to.
>
> Hopefully I've explained the problem well enough to be understood, and
> I would appreciate any help or suggestions you might have.
>
> Thanks again,
> Collin Hockey
> --
> To unsubscribe from this list: send the line "unsubscribe linux- 
> hams" in
> the body of a message to majordomo@vger.kernel.org
> More majordomo info at  http://vger.kernel.org/majordomo-info.html
>
>

Re: mkiss in ubuntu kernels?

[Collin Hockey]

I was looking at the information in the mkiss man page
(http://linux.die.net/man/8/mkiss), it seems to claim to be the one
attached to the physical serial port, routing KISS frames through two
or more pseudo ttys whose slave ends are attached with kissattach to
separate ax.25 network interfaces.

Am I really confused?

On Thu, Sep 10, 2009 at 10:39 PM, Alan Crosswell <alan@columbia.edu> wrote:
> You only kissattach once;  if the TNC supports mkiss then the port is
> encoded in the kiss header for each packet.
> /a
>
>
> On Sep 10, 2009, at 6:44 PM, Collin Hockey <hockeyc@umich.edu> wrote:
>
>> Hi everyone, and thanks in advance for any help.
>>
>> I've been working for the past several days on getting a Kantronics
>> 9612 paired with an iCom IC-910 for packet communication over KISS
>> working using the linux AX.25 stack. I have successfully set up
>> kissattach and related utilities to be able to send data back and
>> forth at 1200 baud.
>>
>> I also need to communicate at 9600 baud. For the Kantronics TNC, this
>> requires talking to the radio over Port 2. My current understanding is
>> that packet routing is done using parts of the KISS protocol and that
>> this is achieved by using mkiss to split the physical serial port into
>> two pseudo ttys, which I can then use kissattach on separately to talk
>> to each port.
>>
>> All of the documentation I have been able to find though assumes the
>> old style psuedo-tty setup, where master and slave pairs are numbered,
>> and plenty are pre-loaded by the kernel for use by user programs.
>> However, at least in my kernel (2.6.28), the client program is
>> supposed to request the single pseudo-tty master file, and the kernel
>> creates a new slave for each connection. There tends to be a pile of
>> these in my pseudo-tty slave directory, so I never know which one to
>> connect to. We do need to automate and script most of these
>> connections, so I do need to reliably be able to tell which slave tty
>> I need to connect to.
>>
>> Hopefully I've explained the problem well enough to be understood, and
>> I would appreciate any help or suggestions you might have.
>>
>> Thanks again,
>> Collin Hockey
>> --
>> To unsubscribe from this list: send the line "unsubscribe linux-hams" in
>> the body of a message to majordomo@vger.kernel.org
>> More majordomo info at  http://vger.kernel.org/majordomo-info.html
>>
>>
>
--
To unsubscribe from this list: send the line "unsubscribe linux-hams" in
the body of a message to majordomo@vger.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html

Re: mkiss in ubuntu kernels?

[Alan Crosswell]

No, probably me that's got it wrong.

On Sep 10, 2009, at 10:48 PM, Collin Hockey <hockeyc@umich.edu> wrote:

> I was looking at the information in the mkiss man page
> (http://linux.die.net/man/8/mkiss), it seems to claim to be the one
> attached to the physical serial port, routing KISS frames through two
> or more pseudo ttys whose slave ends are attached with kissattach to
> separate ax.25 network interfaces.
>
> Am I really confused?
>
> On Thu, Sep 10, 2009 at 10:39 PM, Alan Crosswell <alan@columbia.edu>  
> wrote:
>> You only kissattach once;  if the TNC supports mkiss then the port is
>> encoded in the kiss header for each packet.
>> /a
>>
>>
>> On Sep 10, 2009, at 6:44 PM, Collin Hockey <hockeyc@umich.edu> wrote:
>>
>>> Hi everyone, and thanks in advance for any help.
>>>
>>> I've been working for the past several days on getting a Kantronics
>>> 9612 paired with an iCom IC-910 for packet communication over KISS
>>> working using the linux AX.25 stack. I have successfully set up
>>> kissattach and related utilities to be able to send data back and
>>> forth at 1200 baud.
>>>
>>> I also need to communicate at 9600 baud. For the Kantronics TNC,  
>>> this
>>> requires talking to the radio over Port 2. My current  
>>> understanding is
>>> that packet routing is done using parts of the KISS protocol and  
>>> that
>>> this is achieved by using mkiss to split the physical serial port  
>>> into
>>> two pseudo ttys, which I can then use kissattach on separately to  
>>> talk
>>> to each port.
>>>
>>> All of the documentation I have been able to find though assumes the
>>> old style psuedo-tty setup, where master and slave pairs are  
>>> numbered,
>>> and plenty are pre-loaded by the kernel for use by user programs.
>>> However, at least in my kernel (2.6.28), the client program is
>>> supposed to request the single pseudo-tty master file, and the  
>>> kernel
>>> creates a new slave for each connection. There tends to be a pile of
>>> these in my pseudo-tty slave directory, so I never know which one to
>>> connect to. We do need to automate and script most of these
>>> connections, so I do need to reliably be able to tell which slave  
>>> tty
>>> I need to connect to.
>>>
>>> Hopefully I've explained the problem well enough to be understood,  
>>> and
>>> I would appreciate any help or suggestions you might have.
>>>
>>> Thanks again,
>>> Collin Hockey
>>> --
>>> To unsubscribe from this list: send the line "unsubscribe linux- 
>>> hams" in
>>> the body of a message to majordomo@vger.kernel.org
>>> More majordomo info at  http://vger.kernel.org/majordomo-info.html
>>>
>>>
>>
>

Re: mkiss in ubuntu kernels?

[Curt, WE7U]

On Thu, 10 Sep 2009, Collin Hockey wrote:

> I was looking at the information in the mkiss man page
> (http://linux.die.net/man/8/mkiss), it seems to claim to be the one
> attached to the physical serial port, routing KISS frames through two
> or more pseudo ttys whose slave ends are attached with kissattach to
> separate ax.25 network interfaces.

Did you look up the Linux AX.25 HOWTO document?  That may give you
additional practical info for setting up MKISS.

-- 
Curt, WE7U.                         <http://www.eskimo.com/~archer>
    APRS:  Where it's at!                    <http://www.xastir.org>
   Lotto:  A tax on people who are bad at math. - unknown
Windows:  Microsoft's tax on computer illiterates. - WE7U.
The world DOES revolve around me:  I picked the coordinate system!"

Re: mkiss in ubuntu kernels?

[Collin Hockey]

Hi Curt,

Yes, I did go through that tutorial. It got me going to the point
where I am now :)

I think I understand how mkiss works, but my problem is with the newer
linux kernel pseudo tty setup. Mkiss doesn't report which slave tty it
has connected each port to because it was designed at a time when
there were a bunch of matched pairs (ptm0 and pts0 for instance, m
being for master and s being for slave), whereas now, a program opens
the single master pseudo tty and it gets a new slave tty from the OS.
I'm trying to find a reliable way to figure out which tty mkiss has
connected each port to.

Perhaps I'm barking completely up the wrong tree, but that's my
understanding of it anyway.

Thanks a lot for your answers, they're definitely helping me get
closer to a solution.

Collin Hockey

On Thu, Sep 10, 2009 at 11:51 PM, Curt, WE7U <archer@eskimo.com> wrote:
> On Thu, 10 Sep 2009, Collin Hockey wrote:
>
>> I was looking at the information in the mkiss man page
>> (http://linux.die.net/man/8/mkiss), it seems to claim to be the one
>> attached to the physical serial port, routing KISS frames through two
>> or more pseudo ttys whose slave ends are attached with kissattach to
>> separate ax.25 network interfaces.
>
> Did you look up the Linux AX.25 HOWTO document?  That may give you
> additional practical info for setting up MKISS.
>
> --
> Curt, WE7U.                         <http://www.eskimo.com/~archer>
>   APRS:  Where it's at!                    <http://www.xastir.org>
>  Lotto:  A tax on people who are bad at math. - unknown
> Windows:  Microsoft's tax on computer illiterates. - WE7U.
> The world DOES revolve around me:  I picked the coordinate system!"
>
--
To unsubscribe from this list: send the line "unsubscribe linux-hams" in
the body of a message to majordomo@vger.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html

Re: mkiss in ubuntu kernels?

[Matti Aarnio]

On Fri, Sep 11, 2009 at 01:10:01AM -0400, Collin Hockey wrote:
> Hi Curt,
> 
> Yes, I did go through that tutorial. It got me going to the point
> where I am now :)
> 
> I think I understand how mkiss works, but my problem is with the newer
> linux kernel pseudo tty setup. Mkiss doesn't report which slave tty it
> has connected each port to because it was designed at a time when
> there were a bunch of matched pairs (ptm0 and pts0 for instance, m
> being for master and s being for slave), whereas now, a program opens
> the single master pseudo tty and it gets a new slave tty from the OS.
> I'm trying to find a reliable way to figure out which tty mkiss has
> connected each port to.
> 
> Perhaps I'm barking completely up the wrong tree, but that's my
> understanding of it anyway.
> 
> Thanks a lot for your answers, they're definitely helping me get
> closer to a solution.

You are not at wrong tree, that program is from era before present
pseudo-tty system, when the stupid way it does things did work.

mkiss users are very rare indeed...


I have rewritten mkiss a bit to use modern pty:s, and it is in this
package:

  http://ham.zmailer.org/oh2mqk/libax25/ax25-tools-0.0.9-3.tar.gz

but to use it, you will need also my  libax25  from same directory.
After you take it into use, the information on HOWTO is plain wrong.
See  "man mkiss".   The mkiss will also need associated kissattach.

I did also rework the  kissattach   so that you do not need to set
IP address on AX.25 interface.

My libax25 changes have made it to  cvs@cvs.linux-ax25.org:/home/ax25-cvs
but tools changes are still in the pipeline..  Apparently I have to push
at them again.

> Collin Hockey

   73 de Matti, OH2MQK


> On Thu, Sep 10, 2009 at 11:51 PM, Curt, WE7U <archer@eskimo.com> wrote:
> > On Thu, 10 Sep 2009, Collin Hockey wrote:
> >
> >> I was looking at the information in the mkiss man page
> >> (http://linux.die.net/man/8/mkiss), it seems to claim to be the one
> >> attached to the physical serial port, routing KISS frames through two
> >> or more pseudo ttys whose slave ends are attached with kissattach to
> >> separate ax.25 network interfaces.
> >
> > Did you look up the Linux AX.25 HOWTO document?  That may give you
> > additional practical info for setting up MKISS.
> >
> > --
> > Curt, WE7U.                         <http://www.eskimo.com/~archer>
> >   APRS:  Where it's at!                    <http://www.xastir.org>
> >  Lotto:  A tax on people who are bad at math. - unknown
> > Windows:  Microsoft's tax on computer illiterates. - WE7U.
> > The world DOES revolve around me:  I picked the coordinate system!"
> >
> --
> To unsubscribe from this list: send the line "unsubscribe linux-hams" in
> the body of a message to majordomo@vger.kernel.org
> More majordomo info at  http://vger.kernel.org/majordomo-info.html
--
To unsubscribe from this list: send the line "unsubscribe linux-hams" in
the body of a message to majordomo@vger.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html

Re: mkiss in ubuntu kernels?

[Collin Hockey]

Matti,

These improvements to mkiss and kissattach are great. After a few
minutes fussing with the install and getting my other apps reading the
right axports file, everything is working wonderfully. Thanks a ton
everyone, hopefully someday soon I'll be able to answer some questions
as well :-)

Collin

On Fri, Sep 11, 2009 at 3:38 AM, Matti Aarnio <matti.aarnio@zmailer.org> wrote:
> On Fri, Sep 11, 2009 at 01:10:01AM -0400, Collin Hockey wrote:
>> Hi Curt,
>>
>> Yes, I did go through that tutorial. It got me going to the point
>> where I am now :)
>>
>> I think I understand how mkiss works, but my problem is with the newer
>> linux kernel pseudo tty setup. Mkiss doesn't report which slave tty it
>> has connected each port to because it was designed at a time when
>> there were a bunch of matched pairs (ptm0 and pts0 for instance, m
>> being for master and s being for slave), whereas now, a program opens
>> the single master pseudo tty and it gets a new slave tty from the OS.
>> I'm trying to find a reliable way to figure out which tty mkiss has
>> connected each port to.
>>
>> Perhaps I'm barking completely up the wrong tree, but that's my
>> understanding of it anyway.
>>
>> Thanks a lot for your answers, they're definitely helping me get
>> closer to a solution.
>
> You are not at wrong tree, that program is from era before present
> pseudo-tty system, when the stupid way it does things did work.
>
> mkiss users are very rare indeed...
>
>
> I have rewritten mkiss a bit to use modern pty:s, and it is in this
> package:
>
>  http://ham.zmailer.org/oh2mqk/libax25/ax25-tools-0.0.9-3.tar.gz
>
> but to use it, you will need also my  libax25  from same directory.
> After you take it into use, the information on HOWTO is plain wrong.
> See  "man mkiss".   The mkiss will also need associated kissattach.
>
> I did also rework the  kissattach   so that you do not need to set
> IP address on AX.25 interface.
>
> My libax25 changes have made it to  cvs@cvs.linux-ax25.org:/home/ax25-cvs
> but tools changes are still in the pipeline..  Apparently I have to push
> at them again.
>
>> Collin Hockey
>
>   73 de Matti, OH2MQK
>
>
>> On Thu, Sep 10, 2009 at 11:51 PM, Curt, WE7U <archer@eskimo.com> wrote:
>> > On Thu, 10 Sep 2009, Collin Hockey wrote:
>> >
>> >> I was looking at the information in the mkiss man page
>> >> (http://linux.die.net/man/8/mkiss), it seems to claim to be the one
>> >> attached to the physical serial port, routing KISS frames through two
>> >> or more pseudo ttys whose slave ends are attached with kissattach to
>> >> separate ax.25 network interfaces.
>> >
>> > Did you look up the Linux AX.25 HOWTO document?  That may give you
>> > additional practical info for setting up MKISS.
>> >
>> > --
>> > Curt, WE7U.                         <http://www.eskimo.com/~archer>
>> >   APRS:  Where it's at!                    <http://www.xastir.org>
>> >  Lotto:  A tax on people who are bad at math. - unknown
>> > Windows:  Microsoft's tax on computer illiterates. - WE7U.
>> > The world DOES revolve around me:  I picked the coordinate system!"
>> >
>> --
>> To unsubscribe from this list: send the line "unsubscribe linux-hams" in
>> the body of a message to majordomo@vger.kernel.org
>> More majordomo info at  http://vger.kernel.org/majordomo-info.html
> --
> To unsubscribe from this list: send the line "unsubscribe linux-hams" in
> the body of a message to majordomo@vger.kernel.org
> More majordomo info at  http://vger.kernel.org/majordomo-info.html
>
--
To unsubscribe from this list: send the line "unsubscribe linux-hams" in
the body of a message to majordomo@vger.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html

Re: mkiss in ubuntu kernels?

[David A. Ranch]

I hope those patches make it upstream and into the main kernel sources.
 Until that day comes, I had similar problems for bringing up a JNOS
setup.  I hacked it in using net2kiss and some kludges in my init scripts:


   #start net2kiss
   PTSRPORT=`/usr/sbin/net2kiss -i sm1 /dev/ptmx &`
   PTSPORT=`echo $PTSRPORT | awk -F / '{print $4}'`
   echo -en "[net2kiss on /dev/pts/$PTSPORT ] "


Hope this helps someone

--David, KI6ZHD


> These improvements to mkiss and kissattach are great. After a few
> minutes fussing with the install and getting my other apps reading the
> right axports file, everything is working wonderfully. Thanks a ton
> everyone, hopefully someday soon I'll be able to answer some questions
> as well :-)
> 
> Collin
> 
> On Fri, Sep 11, 2009 at 3:38 AM, Matti Aarnio <matti.aarnio@zmailer.org> wrote:
>> On Fri, Sep 11, 2009 at 01:10:01AM -0400, Collin Hockey wrote:
>>> Hi Curt,
>>>
>>> Yes, I did go through that tutorial. It got me going to the point
>>> where I am now :)
>>>
>>> I think I understand how mkiss works, but my problem is with the newer
>>> linux kernel pseudo tty setup. Mkiss doesn't report which slave tty it
>>> has connected each port to because it was designed at a time when
>>> there were a bunch of matched pairs (ptm0 and pts0 for instance, m
>>> being for master and s being for slave), whereas now, a program opens
>>> the single master pseudo tty and it gets a new slave tty from the OS.
>>> I'm trying to find a reliable way to figure out which tty mkiss has
>>> connected each port to.
>>>
>>> Perhaps I'm barking completely up the wrong tree, but that's my
>>> understanding of it anyway.
>>>
>>> Thanks a lot for your answers, they're definitely helping me get
>>> closer to a solution.
>> You are not at wrong tree, that program is from era before present
>> pseudo-tty system, when the stupid way it does things did work.
>>
>> mkiss users are very rare indeed...
>>
>>
>> I have rewritten mkiss a bit to use modern pty:s, and it is in this
>> package:
>>
>>  http://ham.zmailer.org/oh2mqk/libax25/ax25-tools-0.0.9-3.tar.gz
>>
>> but to use it, you will need also my  libax25  from same directory.
>> After you take it into use, the information on HOWTO is plain wrong.
>> See  "man mkiss".   The mkiss will also need associated kissattach.
>>
>> I did also rework the  kissattach   so that you do not need to set
>> IP address on AX.25 interface.
>>
>> My libax25 changes have made it to  cvs@cvs.linux-ax25.org:/home/ax25-cvs
>> but tools changes are still in the pipeline..  Apparently I have to push
>> at them again.
>>
>>> Collin Hockey
>>   73 de Matti, OH2MQK
>>
>>
>>> On Thu, Sep 10, 2009 at 11:51 PM, Curt, WE7U <archer@eskimo.com> wrote:
>>>> On Thu, 10 Sep 2009, Collin Hockey wrote:
>>>>
>>>>> I was looking at the information in the mkiss man page
>>>>> (http://linux.die.net/man/8/mkiss), it seems to claim to be the one
>>>>> attached to the physical serial port, routing KISS frames through two
>>>>> or more pseudo ttys whose slave ends are attached with kissattach to
>>>>> separate ax.25 network interfaces.
>>>> Did you look up the Linux AX.25 HOWTO document?  That may give you
>>>> additional practical info for setting up MKISS.
>>>>
>>>> --
>>>> Curt, WE7U.                         <http://www.eskimo.com/~archer>
>>>>   APRS:  Where it's at!                    <http://www.xastir.org>
>>>>  Lotto:  A tax on people who are bad at math. - unknown
>>>> Windows:  Microsoft's tax on computer illiterates. - WE7U.
>>>> The world DOES revolve around me:  I picked the coordinate system!"
>>>>
>>> --
>>> To unsubscribe from this list: send the line "unsubscribe linux-hams" in
>>> the body of a message to majordomo@vger.kernel.org
>>> More majordomo info at  http://vger.kernel.org/majordomo-info.html
>> --
>> To unsubscribe from this list: send the line "unsubscribe linux-hams" in
>> the body of a message to majordomo@vger.kernel.org
>> More majordomo info at  http://vger.kernel.org/majordomo-info.html
>>
> --
> To unsubscribe from this list: send the line "unsubscribe linux-hams" in
> the body of a message to majordomo@vger.kernel.org
> More majordomo info at  http://vger.kernel.org/majordomo-info.html