Experimental Physics and
Industrial Control System

Benjamin Franksen <[email protected]> · Tue, 30 Jan 2018 16:52:04 +0100

On 01/30/2018 04:28 PM, Ralph Lange wrote:
> On Tue, Jan 30, 2018 at 4:15 PM, Benjamin Franksen <
> [email protected]> wrote:
>> On 01/30/2018 03:31 PM, Simon Reiter wrote:
>>> after updating from EPICS 3.14.12.6 to 3.16.1 I receive the following
>>> message.
>>> A PV is set by the sequencer (2.2.4). It is actually done, but every
>>> time, the sequencer set the PV, I get this message.
>>>
>>>     sevr=minor ca_array_put(typeToCA(type), count, var->chid, value):
>>> Channel write request failed
>>>
>>> int pv_variable;
>>> assign pv_variable to "{IOC}:timeout:S";
>>> pv_variable = value;
>>> pvPut(pv_variable);
>>>
>>> The connected PV is binary output.
>>
>> From looking at the CA sources it seems that this is ECA_PUTFAIL error
>> code, which is used as the default (catch-all) error code in
>> src/ca/client/oldChannelNotify.cpp e.g. line 525:
>>
>>     catch ( ... )
>>     {
>>         caStatus = ECA_PUTFAIL;
>>     }
>>
>> So we know that /something/ fails and it is not any of the expected
>> exceptions (cacChannel::badString, ... std::bad_alloc). You could insert
>> debug statements in the catch-all clause to print more information about
>> the exception.
>>
>> Does a caput from the command line work also result in an error message?
> 
> Uh-oh,
> There is a (mostly harmless) bug in 3.16.1 that returns failure codes to CA
> put operations even if they succeed.
> The underlying problem has been fixed upstream, and will be released with
> 3.16.2.
> 
> Meanwhile, please refer to the Known Problems page (
> https://epics.anl.gov/base/R3-16/1-docs/KnownProblems.html) that contains a
> patch.

Thanks Ralph, that clears it up then.

Cheers
Ben

-- 
"Make it so they have to reboot after every typo." ― Scott Adams

Experimental Physics and Industrial Control System

Experimental Physics and
Industrial Control System