unexpected behavior of incisiv with disable in for loop

[shubham_dce]

Hi,

I am very new to this forum, so please pardon any decorum that i may have violated.

Here is the reference code : 

always @(posedge clk or negedge resetn)
begin
  if(~resetn)
  begin
    for(int j=0; j<64; j++)
      enable_data[j] <= 1'b1;
  end
  else
  begin
    if(data_accepted) 
      begin: ENABLE_DATA_BLOCK
         -> event_1;
         for (int k=0; k<64; k++)
           begin
              -> event_2;
              if (enable_data[k] && queue_pos_available[k])
                 begin
                     -> event_3;
                     enable_data [k] <=0;
                     -> event_4;
                     disable ENABLE_DATA_BLOCK;
                 end
           end // for loop
       end // if(data_accepted)
   end // else block
end // always...

While running the above code in incisiv (version incisiv/15.20.016) i am facing the following issue:
with all the condition being true that is data_accepted=1, enable_data[k]=1 and queue_pos_available[k]=1
enable_data[k] is not taking the assignment 0. event_3 and event_4 both are triggering. 

I ran the code by removing disable to break and the code ran correctly. So, i think there is a problem by using disable. I can't go with this fix as break is not synthesisable. My lint-run is not passing with the change of break. 

Thinking that there is a problem with the use of disable, i tried breaking the loop by setting the loop variable to its maximum value as:
.
.
.
if (enable_data[k] && queue_pos_available[k])
                 begin
                     -> event_3;
                     enable_data [k] <=0;
                     -> event_4;
                     k=64;  // i am using blocking statement as i wanted to break the loop immediately.
                 end
.
.

I want to add a point that the code is running fine with VCS and modelsim simulators. 

If any-one can direct me of what can be the issue here, it will be of real help. 
Thanks for your time.
 

Regards,

Shubham verma

[uwes]

there are a few points to add here

 

1. since this a specific request for a particular tool please raise a request with your vendor 

2. fork/disable etc are subject process control and are very close to grey area's of the vlog spec and/or races - so you might see differences in behviour

3. a much better solution than to disable the block are one of the two following:

a) recode the for loop to a while loop 

>for (int k=0; k<64; k++)

int k=0; int flag=1;

while(k<64 && flag) do k++; ....  // set flag to break done

or 

B) wrap the for loop in a task and then terminate the task with a "return" upon the condition instead of a break

 

/uwe

[shubham_dce]

Hi uwe,

Thanks for your reply.

Please find my replies: 

On 1/16/2017 at 2:14 PM, uwes said:

. since this a specific request for a particular tool please raise a request with your vendor 

 

- i am trying to take the issue to incisiv team. My intention to ask here was to familiarize with any similarity anyone has faced

On 1/16/2017 at 2:14 PM, uwes said:

a much better solution than to disable the block are one of the two followi

i tried to use while loop but the issue seems to persist. I am trying different solutions to see if i am able to get ahead.

I will update if i found a way.

 

Thanks,

Shubham Verma