The patches I mentioned are already built into main (RTEMS7). And that gives us the problem
that UDP traffic stops after I establish a TCP connection on port 5075 and then close it.
> On 19. Nov 2025, at 16:04, Heinz Junkes (FHI) via Tech-talk <[email protected]> wrote:
>
> Ok, I applied the patch from Chris
>
> remotes/kiwiIOP/chris/close-optional-on-ref-count
>
> Then the correction from SEB
>
> remotes/seb/rtems-6-fix-libio-iop-ref-count
>
> and that is the result at the moment :
>
>
> RTEMS Shell on /dev/console. Use 'help' to list commands.
> erSHLL [/] # r: fcntl: Bad file number
> err: fcntl: Bad file number
> err: fcntl: Bad file number
> err: fcntl: Bad file number
> err: mve0: set_mtu: Bad file number
> err: no valid interfaces found
> warning: no interfaces have a carrier
> err: poll: Bad file number
> err: BSD Program: Could not close file 24 or could not remove it from list of open files
> err: BSD Program: Could not close file 23 or could not remove it from list of open files
> err: BSD Program: Could not close file 22 or could not remove it from list of open files
>
> :-(
>
> Heinz
>
>> On 19. Nov 2025, at 03:26, Michael Davidsaver <[email protected]> wrote:
>>
>> On 11/14/25 8:21 AM, Heinz Junkes (FHI) wrote:
>>> I have now narrowed it down a little further (hopefully)
>>>
>>> When I boot the target and use the debug messages, I get a constant stream of events:
>> ...
>>>
>>> When I now display pvlist on another host, I see the target (141.14.128.12):
>>> (I can repeat this several times)
>>>
>>> (base) hactar:~ junkes$ pvlist
>>> GUID 0x0224EB2FE0ED7782E367B71A version 2: tcp@[ 141.14.128.12:5075 ]
>>>
>> ...
>>>
>>> However, when I run pvlist with GUID, the debug output on the target stops immediately:
>>>
>>> on the client:
>>>
>>> (base) hactar:~ junkes$ pvlist 0x0224EB2FE0ED7782E367B71A
>> When in "guid" mode, pvlist first sends a UDP discovery ping (same as the no argument case), then opens a TCP connection if it receives a reply with a matching GUID. So the network traffic of the preceding should be the same as:
>>> pvlist
>>> pvlist 141.14.128.12:5075
>>
>> If a TCP connection is the trigger, you could try to open a connection with eg. "telnet" or "netcat" and see if the UDP traffic stop coincides on TCP connect or disconnect? Or maybe you have to send some bytes?
>>
>
- Replies:
- Re: pvxs Access/list problem Michael Davidsaver via Tech-talk
- References:
- Re: pvxs Access/list problem Heinz Junkes (FHI) via Tech-talk
- Fwd: pvxs Access/list problem Heinz Junkes (FHI) via Tech-talk
- Re: Fwd: pvxs Access/list problem Michael Davidsaver via Tech-talk
- Re: pvxs Access/list problem Michael Davidsaver via Tech-talk
- Re: pvxs Access/list problem Heinz Junkes (FHI) via Tech-talk
- Navigate by Date:
- Prev:
Re: Streaming from quadEM to HDF5 Mark Rivers via Tech-talk
- Next:
Disable alarm at IOC via PVAccess David Bracey via Tech-talk
- Index:
1994
1995
1996
1997
1998
1999
2000
2001
2002
2003
2004
2005
2006
2007
2008
2009
2010
2011
2012
2013
2014
2015
2016
2017
2018
2019
2020
2021
2022
2023
2024
<2025>
2026
- Navigate by Thread:
- Prev:
Re: pvxs Access/list problem Heinz Junkes (FHI) via Tech-talk
- Next:
Re: pvxs Access/list problem Michael Davidsaver via Tech-talk
- Index:
1994
1995
1996
1997
1998
1999
2000
2001
2002
2003
2004
2005
2006
2007
2008
2009
2010
2011
2012
2013
2014
2015
2016
2017
2018
2019
2020
2021
2022
2023
2024
<2025>
2026
|