Category Archives: Frequency dependent bugs
CDC methodology – “Be afraid”
This is (yet) another CDC (clock domain crossing) bug and a simpler version of synchronising an asynchronous event non-atomically More correctly this type of bug is called “CDC re-convergence“. CDC appears to be favourite interview question, but only in a
CDC methodology – “Be afraid”
This is (yet) another CDC (clock domain crossing) bug and a simpler version of synchronising an asynchronous event non-atomically More correctly this type of bug is called “CDC re-convergence“. CDC appears to be favourite interview question, but only in a
There is loads of margin in the ASIC process …. isn’t there?
This was a device that was complete, validated, in production and so out of our lives. But a customer was seeing very occasional failures where the watchdog would fire meaning the processor had died, hung, got lost, gone off into
There is loads of margin in the ASIC process …. isn’t there?
This was a device that was complete, validated, in production and so out of our lives. But a customer was seeing very occasional failures where the watchdog would fire meaning the processor had died, hung, got lost, gone off into
Synchronising an asynchronous event non-atomically
The dangers of passing information between asynchronous clock domains are well known and well documented (see Ran Ginosar’s “14 Ways to Fool Your Synchronizer“). Much of this information had been heeded in this circuit. But the synchonisation of a single
Synchronising an asynchronous event non-atomically
The dangers of passing information between asynchronous clock domains are well known and well documented (see Ran Ginosar’s “14 Ways to Fool Your Synchronizer“). Much of this information had been heeded in this circuit. But the synchonisation of a single
When silicon fails.. 