Ted Hong

@meheff -- let me know if you'd like to request additional changes. Otherwise, I can start the process of merging this.

Note that this is not exactly the same as adding an input delay on each receive node. Multiple input channels may be sequentially dependent in the IR, but may be...

An update on this: 1. Currently we have mechanisms for packaging the binary releases, but the actual release is pending going through an internal license review. 2. XLS is currently...

Currently we have the JitChannelQueueWrapper to help with this. Only reason it wasn't incorporated into JitChannelQueueWrapper is the need to have access to the ProcJit object. https://github.com/google/xls/blob/main/xls/flows/ir_wrapper.h#L38

For this check, there will probably need a looser spec for procs used only as a test driver and a stricter one for IR conversion.

Duplicate of https://github.com/google/xls/issues/575

Related https://github.com/google/xls/issues/911 https://github.com/google/xls/issues/899 https://github.com/google/xls/issues/1385 https://github.com/google/xls/issues/1386

Related https://github.com/google/xls/issues/911 https://github.com/google/xls/issues/899 https://github.com/google/xls/issues/1385 https://github.com/google/xls/issues/1386

For this, I'm proposing adding a "cycle count" oriented construct into the DSL. We already have it in the IR, and this issue would be to expose it.

Another usage for this is to be able to parameterize structs with struct parametric. For example -- for SRAMS (https://github.com/google/xls/blob/adb92be6099de5486b7810cdbb4935a67c16cd72/xls/examples/ram.x#L32-L36) Grouping all the parameters in a struct can simply instantiation...