- After finding the bug in the reference model and wasting countless hours going back and forth with FPGA timing optimization and bug tweaks, I realized that the design was fundementally broken. The decision to use two clock domains (high speed) and low speed was correct from the beginning. The FPGA is dreadfully slow, (you definitely don't want to do much logic at 300MHz...), but the handoff between tclk and tclk_div4 was too complicated. The puzzle of having to respond to wait quickly, covering the corner cases, and meeting timing was just too ugly.
- The "new" design goes back to the method of using the high speed logic only for doing a "dumb" parallel to serial converter and preparing all the necessary signals in the low speed domain.
- This feel A LOT cleaner and the it already passes basic tests with the chip reference and the loopback after less than 3 hours of redesign work!
- The TX meets timing but there is still some work to do with wait pushback testing.
- The logic was a mess, causing me to go around in circles for days. In the end, by adding a missing sync circuit (duh!) between the fast and slow clock to align the edges and removing a redundant pipeline stage ("double") the nasty logic just fell away. Looks good now.
-Write bursts mostly works and design looks clean.
-one bug left to fix on streams of writes...
- Old design was not workable with bursting and long waits. The wait signal needs to be very carfully handled since it's asynchronous to the clock.
-The TX needs to be stopped quickly so the sync needs to be done at the high speed clock, not at div4 clock
-Since there are synchronizers here, there should be only one point of sync. This is not completely the case still, but I think??? it should be safe by constructiona at this point.
-bursting working at this point for writes!!!!!
- Using the BUFIO makes another clock domain....FPGAs apparently hate clock domain crossings, avoid them at all cost.
- Now moving back to having on high speed clock domain for logic and DDR blocks, take care of IO alignment in software for TX and RX
- Also, fixed the io_wait path with logic...not sure what I was thinking there. Logic was trivial. The way it was,the io path was going straight into the FIFO as a wait.
- reset was broken!
- need to account for wait
- merging read/write wait for simplicity, otherwise you would need to reset the packets to figure out if it's a read or write transaction...and I don't want to reset every packet throughout the pipe.
- holding rx in reset state until tx is done
- removing reset from all pipeline registers
- removing reset from oddr/iddr
- the idea is to keep things quiet not to block in lots of places. The only real block needed is in the FIFO to keep "noise" from propagating past the link. The link should be kept in a safe reset state until the rx fram is stable and the clock is running so that the pipe can be cleaned out.
* Seems like a useless feature. Why autogenerate the transactions at the transmit side. This should always be done at the receive side to minimize bits moving across the link. Can't really see a use for it anymore so I am removing it.
* If you want to hack the design to reduce latency, you can always grab the raw etx_core and drive signals directly through write port.
* May consider adding a fourth port to etx to allow bypassing the link interfac?
* Add an ifdef to bypass the fifos?
