You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
I understand that for Mem, a readWrite port can always host a read or a write port, so in principle the RAM with 1 readSync and 1 readWrite port should be able to be blackboxed into a Ram_2wrs. This is not possible currently:
class spinal.core.blackboxAll$ is not able to blackbox toplevel/bindProtoToCoreCtrl_txPktBuffer : Mem[151*512 bits]
write ports : 0
readAsync ports : 0
readSync ports : 1
readWrite ports : 1
-> Unblackboxable memory topology
The text was updated successfully, but these errors were encountered:
I think in practice RAM like this would be possible to infer by the EDA tool, so there's not a strict need to blackbox this case. I'm currently using the blackbox flow to make sure I don't accidentally create RAM that could not be inferred by Vivado; do we have some sort of checks for this purpose?
But shoudnl't as the idea is to be very accurate about what it is (Ram_1r_1wrs which doesn't exists)
I think in practice RAM like this would be possible to infer by the EDA tool,
Ahh on the big vendors yes, but on smaller things / darker place, things can break very easily
For vivado, i only use blackboxing when using byte enable or multiple write ports
I understand that for Mem, a readWrite port can always host a read or a write port, so in principle the RAM with 1 readSync and 1 readWrite port should be able to be blackboxed into a
Ram_2wrs
. This is not possible currently:The text was updated successfully, but these errors were encountered: