Question on generated clock constraints ?

[pavanks]

Hi,

Do we need to specify uncertainty value on generated clocks inside the design or
Will the tool propagate the constraints and calculate this ?

 

[huckle]

How is the uncertainty defined with the source clock? Is it with -from [get_clock xx] and -to [get_clock xx]?
If yes, then your answer is NO, the generated clock is a different clock and will not be affected. You get where you defined, not on other clocks.

 

[pavanks]

How is the uncertainty defined with the source clock? Is it with -from [get_clock xx] and -to [get_clock xx]?
If yes, then your answer is NO, the generated clock is a different clock and will not be affected. You get where you defined, not on other clocks.

You meant here the master clock right?
There is uncertainty defined on the source clock(master clock). Will the tool calculate the uncertainty on the generated clock
through propagation?

 

[lostinxlation]

Why don't you do an experiment ?
I bet there is no such thing like IEEE standard or so on whether the clock uncertainty propagates and it's tool dependent.
The best way is trying out with the tool you are using.

 

[shobhit]

You need to specify theuncertainity to the generated clock seperately using the set_clock_uncertainity command, the tool will not pick uncertainity value from its master clock. it can only pick the latency and waveforms from the masterclock