Johan Ström | 19 Sep 08:57 2014
Picon

Issues with FTDI based USB-Serial dongle & DS2480B

Hi,

I've noticed some problems using a FTDI based USB serial dongle together
with a DS2480B based adapter (also known as DS9097U).
On startup the device was not recognized at all, complaining about wrong
responses. Anyone else seen these issues?

I've debugged the problem and found the cause.
On startup, this is what happens:

  DEBUG: ow_ds9097U.c:(287) Attempt 0 of 3 to initialize the DS9097U
TRAFFIC OUT <write> bus=0 (/dev/cua-labdesk)
Byte buffer anonymous, length=1
--000: C1
   <.>
  DEBUG: ow_ds9097U.c:(381) Send the initial reset to the bus master.
TRAFFIC OUT <write> bus=0 (/dev/cua-labdesk)
Byte buffer anonymous, length=1
--000: 71
   <q>
TRAFFIC OUT <write> bus=0 (/dev/cua-labdesk)
Byte buffer anonymous, length=1
--000: 0F
   <.>
  DEBUG: ow_tcp_read.c:(64) attempt 1 bytes Time: 5.000000 seconds
TRAFFIC IN  <NETREAD> bus=0 (/dev/cua-labdesk)
Byte buffer anonymous, length=1
--000: 70
   <p>
  DEBUG: ow_tcp_read.c:(114) read: 1 - 0 = 1
(Continue reading)

Loren Amelang | 15 Sep 01:30 2014
Picon

Re: unruly temperatures - Vol 100, Issue 6

Hakan,

I had an episode (on a BBB with I2C connection) where for a couple of weeks
the temperature resolution would randomly drop from 12-bit to 1/2 degree C.
Looking at the owserver presentation of all of the various bit resolutions,
sometimes the lower resolutions were not proper representations of the
higher-bit values. I couldn't tell if some values were delayed, or just
plain offset, but depending on which resolution happened to get written into
my filesystem where my logger reads them, the temperature could appear to
jump by 1/2 degree. I verified that the values in the filesystem files were
being logged correctly. It would happen for a few minutes or a few hours,
and then fix itself. And then without me doing anything at all, it just
stopped happening!

Looking at your graph, it appears you usually get full resolution gradual
changes on either side of your 1/2 degree steps. My graphs were always more
like the later part of your blue line - chattering by 1/2 degree where there
should have been a smooth transition.

I have no clues why this started or stopped...

| Loren Amelang | loren <at> pacific.net |

On Saturday, September 13, 2014 at 10:29 PM,
owfs-developers-request <at> lists.sourceforge.net wrote:

> 1. unruly temperatures (H?kan Elmqvist)
>
> ----------------------------------------------------------------------
>
(Continue reading)

Håkan Elmqvist | 12 Sep 17:18 2014

unruly temperatures

Suddenly my temperature readings have started to jump up and down about 0.5 degree C at infrequent but random points of time. See:
http://smeden.org/jumps.jpg
(the green trace is the temperature of my watermaker, the blue sea- and the red outdoor-temperature)
I have had this system since 2005 but not seen this behaviour before. Today I am using owfs 2.9p5. The sensors are a mix of old and new ones.
Has anyone seen this behaviour? Is there any explanation?
Thanks in advance
H

--
Håkan Elmqvist
Sunnerdahlsv 7
167 62 Bromma
Telefon: +46 (8) 80 18 81, +46 (704) 56 74 81, +46 (176) 84 077
epost: hakelm <at> smeden.org
Kryptera mera!
------------------------------------------------------------------------------
Want excitement?
Manually upgrade your production database.
When you want reliability, choose Perforce
Perforce version control. Predictably reliable.
http://pubads.g.doubleclick.net/gampad/clk?id=157508191&iu=/4140/ostg.clktrk
_______________________________________________
Owfs-developers mailing list
Owfs-developers <at> lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/owfs-developers
Stefano Miccoli | 12 Sep 00:40 2014

allow independent installation of perl5/php/python in ownet

Hi:

you all know that there are 2 sets of perl/python/php bindings:

*) one based on swig in module/swig/{perl5,php,python} that accesses the 1wire bus directly
*) one based on the owserver protocol in module/ownet/{perl5,php,python}

The problem with the current configure script is that disabling swig or one of the swig bindings
(./configure --enable-swig=no  or ./configure --enable-owperl=no --enable-php=no --enable-python=no) also prevents the installation of the corresponding ownet binding. This is really a nuisance, since it is often useful to install just the bindings from ownet, without being forced to compile/install the whole swig thing. 

The enclosed patch tries to solve this by decoupling the --with-python --with-perl5 --with-php options from the corresponding --enable-ow{perl5,php,python} options.

So, considering for example python, if this patch is applied:

--with-python=no disables both module/ownet/python and module/swig/python
--enable-swig=no disables the whole module/swig
--enable-ownet=no disables the whole module/ownet

--enable-owpython=no disables only module/swig/python

This latter options is somehow redundant with my patch: in fact  --with-python --enable-swig and --enable-ownet are clearly orthogonal to each other, while --enable-owpython lies in the space spanned by --with-python and --enable-swig. However I was unsure if it would be acceptable to drop this well known option.

Stefano

------------------------------------------------------------------------------
Want excitement?
Manually upgrade your production database.
When you want reliability, choose Perforce
Perforce version control. Predictably reliable.
http://pubads.g.doubleclick.net/gampad/clk?id=157508191&iu=/4140/ostg.clktrk
_______________________________________________
Owfs-developers mailing list
Owfs-developers <at> lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/owfs-developers
Howell, Larry (Contractor | 10 Sep 17:43 2014

code not able to access owfs

Our project is developing an embedded system running Linux 2.6.35 on an i.MX53.  Fuse is built in to the kernel.  The system utilizes plugin devices that store configuration data on internal DS2505s.  We’re using owfs-2.8p15 and writing application code based on the owcapi.  The system uses a DS2482-800 bus master and the initialization parameters are ‘i2c=/dev/i2c-2:0 –m mnt/1wire --allow_other’.  The filesystem mounts with expected permissions, and non-root users can access the filesystem with no problem from the bash shell.  However, the application code fails all attempts at filesystem access with a ‘permission denied’ error.  Searching the owfs man pages hasn’t provided any clues about this problem.  Any pointers or suggestions would be greatly appreciated.

 

Thanks for your help!

 

Larry Howell

 

------------------------------------------------------------------------------
Want excitement?
Manually upgrade your production database.
When you want reliability, choose Perforce
Perforce version control. Predictably reliable.
http://pubads.g.doubleclick.net/gampad/clk?id=157508191&iu=/4140/ostg.clktrk
_______________________________________________
Owfs-developers mailing list
Owfs-developers <at> lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/owfs-developers
Mick Sulley | 6 Sep 00:53 2014

DS2413 Power

Just testing a Sheepwalk dual channel IO board, based on DS2413 and it 
does not have a 'power' register.  Is there any way to tell if it is 
running parasitic or has a 5v supply?

Thanks
Mick

------------------------------------------------------------------------------
Slashdot TV.  
Video for Nerds.  Stuff that matters.
http://tv.slashdot.org/
Colin Law | 4 Sep 22:25 2014
Picon

Owserver hang on TPLink WR703N with OpenWRT

Hi

I have put OpenWRT (AA) on a TPLink WR703N and installed owserver from
the openwrt repository (owserver version 2.8p13-1) with remarkably few
problems.  Using a LinkUSB all seems well initially.  Unfortunately
after a few minutes operation owserver hangs.  Running with log level
9 the log contains:
....
 CALL: ow_parsename.c:FS_ParsedName_anywhere(95)
path=[/28.601DE1020000/temperature12]
  DEBUG: ow_cache.c:Cache_Get_Device(927) Looking for device 28 60 1D
E1 02 00 00 C4
  DEBUG: ow_cache.c:Cache_Get_Common(1083) Search in cache sn 28 60 1D
E1 02 00 00 C4 pointer=0x77c8f900 index=0 size=4
  DEBUG: ow_cache.c:Cache_Get_Common(1099) Value found in cache.
Remaining life: 4 seconds.
  DEBUG: ow_presence.c:CheckPresence(76) Found device on bus 0
  DEBUG: ow_read.c:adjust_file_size(329) file_length=12 offset=0 size=1
  DEBUG: ow_cache.c:Cache_Get_Common(1083) Search in cache sn 28 60 1D
E1 02 00 00 C4 pointer=0x77c79130 index=0 size=8
  DEBUG: ow_cache.c:Cache_Get_Common(1115) Value found in cache, but
expired by 99 seconds.
  DEBUG: ow_cache.c:Cache_Get_Simul_Time(986) Looking for conversion
time 28 60 1D E1 02 00 00 C4
  DEBUG: ow_cache.c:Cache_Get_Common(1083) Search in cache sn 00 00 00
00 00 00 00 00 pointer=0x77c8f8f8 index=0 size=0
  DEBUG: ow_cache.c:Cache_Get_Common(1119) Value not found in cache
  DEBUG: ow_cache.c:Cache_Get_Internal(956) 28 60 1D E1 02 00 00 C4 size=4
  DEBUG: ow_cache.c:Cache_Get_Common(1083) Search in cache sn 28 60 1D
E1 02 00 00 C4 pointer=0x77c78290 index=-2 size=4
  DEBUG: ow_cache.c:Cache_Get_Common(1119) Value not found in cache
  DEBUG: ow_select.c:BUS_select(66) Selecting a path (and device)
path=/28.601DE1020000/temperature12 SN=28 60 1D E1 02 00 00 C4 last
path=00 00 00 00 00 00 00 00
  DEBUG: ow_select.c:BUS_select(79) Continuing root branch
  DEBUG: loop.c:Ping_or_Send(112) Taking too long, send a keep-alive pulse
  DEBUG: to_client.c:ToClient(56) payload=-1 size=0, ret=0, sg=0x0 offset=0
  DEBUG: to_client.c:ToClient(63) Send delay message
  DEBUG: loop.c:Ping_or_Send(112) Taking too long, send a keep-alive pulse
  DEBUG: to_client.c:ToClient(56) payload=-1 size=0, ret=0, sg=0x0 offset=0
  DEBUG: to_client.c:ToClient(63) Send delay message
  DEBUG: loop.c:Ping_or_Send(112) Taking too long, send a keep-alive pulse
  DEBUG: to_client.c:ToClient(56) payload=-1 size=0, ret=0, sg=0x0 offset=0
  DEBUG: to_client.c:ToClient(63) Send delay message
... ad infinitum  (or at least ad nauseam).

It fails on two different 1-wire networks and two LinkUSB adaptors,
one of which has been running with owserver on a Sheeva Plug for
several years.

Any suggestions gratefully accepted.

Colin

------------------------------------------------------------------------------
Slashdot TV.  
Video for Nerds.  Stuff that matters.
http://tv.slashdot.org/
Stefano Miccoli | 28 Aug 13:00 2014

bug in owlib/src/c/ow_net_server.c

Hi,

commit 2e36ac0ecefb45ff4b741a9afe4d33a40025e346
Author: Paul Alfille <paul.alfille <at> gmail.com>
Date:   Tue Aug 26 15:23:36 2014 -0400

    Coverity changes
    
    Not sure if these are really issues, but making changes for Coverity
    scanning error.


The above commit introduces a Coverity induced bug!

in file module/owlib/src/c/ow_net_server.c memory cleanup belongs to the child thread (routine ProcessAcceptSocket) and not to the calling routine (ProcessListenSocket).

In fact pthread_create returns before processing in ProcessAcceptSocket is ended, so struct asd must not be freed in the calling routine.

A patch that undoes the modification is enclosed.

Stefano

------------------------------------------------------------------------------
Slashdot TV.  
Video for Nerds.  Stuff that matters.
http://tv.slashdot.org/
_______________________________________________
Owfs-developers mailing list
Owfs-developers <at> lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/owfs-developers
Stefano Miccoli | 24 Aug 14:38 2014

patch: owserver crash due to unintialized Mutex.aliasfind_mutex

Hi,

my owserver (current development git version) crashes on OSX when one tries to read
/settings/alias/list. (This does not happens on Linux).

I found out that this is due to an uninitialized member in the global struct Mutex, see
module/owlib/src/include/ow_mutexes.h On the hunt for similar errors I found other struct Mutex
members that are not initialized and/or not used.

The attached patch fixes this problem.

Actions are

1) init Mutex.aliaslist_mutex
2) delete reference to unused Mutex.simul_mutex, Mutex.aliasfind_mutex
3) correct a fprintf call from  fprintf(stderr, buf) to fprintf(stderr, "%s", buf). This latter
correction may seem unnecessary, but the CLANG compiler flags fprintf(stderr, buf) as a security
problem, and I think that although the injection of malicious code into 'buf' is very unlikely, it is
always better to stay on the safe side.

Regards, 

Stefano


------------------------------------------------------------------------------
Slashdot TV.  
Video for Nerds.  Stuff that matters.
http://tv.slashdot.org/
_______________________________________________
Owfs-developers mailing list
Owfs-developers <at> lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/owfs-developers
Jan Kandziora | 23 Aug 22:34 2014
Picon
Picon

Problems with multiple owservers when network components fail

Hi all,

I have problems with the network timeout in a system with multiple
owservers.

The x86 master and three Raspberry Pi are each one equipped
with a self-made (sources identical) Linux distribution running
owfs-2.9p0. They all have DS9490 host adaptors and all run owservers to
access it. On the master, there is a Tcl application which accesses the
owservers through owtcl. The system runs smoothly as long all components
are there and functional.

But hardware may fail and that isn't handled gracefully enough by owfs
by now. When I pull a network cable to one of the raspis and do

% ow get alarm

on the master or even scan a single bus

% ow get bus.1/alarm

which happens to be "away", that command takes at least some seconds,
but sometimes up to a minute to return. I tried tweaking the
settings/timeout/network parameter, but it doesn't seem to do anything.

I'd wish that owfs would return *an error* and do it *much sooner*
(<<one second) so I can react to that failure on the application level
before watchdogs on the other hardware on the still working buses trigger.

Question: How to tweak the owserver handling so I get a shorter timeout
and an error message from owfs about the network failure?

Kind regards

	Jan

------------------------------------------------------------------------------
Slashdot TV.  
Video for Nerds.  Stuff that matters.
http://tv.slashdot.org/
Colin Law | 23 Aug 14:45 2014
Picon

OWFS on TP-Link TL-WR703N

Hi

I have seen suggestions that a cheap wifi 1-wire hub can be made using
a TL-WR703N and a USB/1-wire adaptor. using openWRT and owserver.

I can't find any reports by anyone that has actually done this
however.  Has anybody here had a go or got a reference to someone that
has?

Colin

------------------------------------------------------------------------------
Slashdot TV.  
Video for Nerds.  Stuff that matters.
http://tv.slashdot.org/

Gmane