Am 10.05.2010 um 14:42 schrieb Kenneth Lerman:
> My guess is that the commented out Interp::synch() should be something like:
>
> GET_EXTERNAL_QUEUE_EMPTY(); // possibly check that this returns true
> set_probe_data(&_setup);
> _setup.probe_flag = OFF;
I just tried your suggestion - no change. GET_EXTERNAL_QUEUE_EMPTY() does
return true, but doesnt wait either.
Actually I dont think the code from Interp::read needs duplication here because
upon the next read it falls into set_probe_data() since the probe flag is ON.
So the problem isnt that this code isnt executed, to the contrary it's executed
too early by immediately doing the next read() after the execute(). So we need
code to wait here until the probe operation finished one way or the other, then
let read() do its work.
-Michael
BTW I'm debugging this on emc2-sim running on virtualbox on a Macbook - no
probe needed to test the probing code ;-)
>
> In other words, duplicate the code at the beginning of Interp::read. My guess
> is that we should also have the code (from read) that checks and handles the
> toolchange_flag and the input_flag (also from the beginning of Interp::read).
>
> Ken
>
> Michael Haberler wrote:
>> narrowing down more, I tentatively have the condition where the interpreter
>> should wait for a G38 to finish - I'm looking for a suggestion how to make
>> it actually wait
>>
>> Interp::synch() isnt it - any suggestions?
>>
>> -Michael
>>
>> rs274_pre.cc near line 25
>>
>> logDebug("!!!KL Open file is:%s:", _setup.filename);
>> logDebug("MDImode = %d", MDImode);
>> while(MDImode && _setup.call_level) // we are still in a subroutine
>> {
>> status = read(0); // reads from current file and calls parse
>> logDebug("status = %d\n",status);
>> if (status != INTERP_OK)
>> {
>> return status;
>> }
>> int exec_status = execute();
>> /* the following condition is true if a G38 was issued and
>> * the next interpreter read should wait for it to finish
>> */
>> if ((exec_status == INTERP_EXECUTE_FINISH) &&
>> (_setup.probe_flag == ON)) {
>> printf("execute(%s) returns INTERP_EXECUTE_FINISH ,
>> probe_flag ON\n",
>> _setup.blocktext);
>> /* this doesnt work: Interp::synch();
>> */
>> }
>> }
>> Am 10.05.2010 um 01:55 schrieb Michael Haberler:
>>
>>
>>
>>> I've looked into this a bit more - it is definitely a race condition.
>>>
>>> The interpreter inner loop which reads and executes a o-word block fails to
>>> wait for a G38 to finish and immediately executes the next command which
>>> explains the erratic coordinate results observed thereafter (near
>>> emc/rs274ngc/rs274ngc_pre.cc line 241). It also behaves differently from
>>> normal main loop executing an .ngc file (no calls to
>>> emcTaskPlanSynch(),emcTaskPlanClearWait(),emcTaskPlanRead() etc).
>>>
>>>
>>
>>
>>
>
------------------------------------------------------------------------------
_______________________________________________
Emc-developers mailing list
[email protected]
https://lists.sourceforge.net/lists/listinfo/emc-developers
