Instantiate fake tensor modules outside of diplomacy

Adding them to the Diplomacy graph will cause to widen source widths
which can have area implications.

This gets rid of the need to do addResource() calls to the manually
generated Verilog files.  Their module parameters should be kept the
same as what's used in the parent Verilog module, however.

src/main/scala/radiance/tile/RadianceTile.scala

@@ -192,6 +192,8 @@ class RadianceTile private (
   //     "We recommend setting nSrcIds to at least 16."
   // )
 
+  val tensorTagWidth = 4 // hardcoded
+
   // Replicates some of the logic of how Vortex determines the tag width of
   // memory requests so that Chisel and Verilog are in agreement on bitwidths.
   // See VX_gpu_pkg.sv
@@ -380,12 +382,6 @@ class RadianceTile private (
     tlMasterXbar.node :=* AddressOrNode(base) :=* dcacheNode
   }
 
-  // Instantiate a fake TensorCoreDecoupled module to force unique-ification of
-  // module names in the Chisel-generated Verilog.  This should be disabled for
-  // synthesis runs
-  val tensor = LazyModule(new radiance.core.TensorCoreDecoupledTL)
-  tlMasterXbar.node :=* tensor.node
-
   /* below are copied from rocket */
 
   val tile_master_blocker =
@@ -743,13 +739,13 @@ class RadianceTileModuleImp(outer: RadianceTile)
     def connectTensor = {
       val tcb0 = new {
         val addr = core.io.tc_a_bits_address(31, 0)
-        val tag = core.io.tc_a_bits_tag(3, 0)
+        val tag = core.io.tc_a_bits_tag(outer.tensorTagWidth - 1, 0)
         val aValid = core.io.tc_a_valid(0)
         val dReady = core.io.tc_d_ready(0)
       }
       val tcb1 = new {
         val addr = core.io.tc_a_bits_address(63, 32)
-        val tag = core.io.tc_a_bits_tag(7, 4)
+        val tag = core.io.tc_a_bits_tag(4 + outer.tensorTagWidth - 1, 4)
         val aValid = core.io.tc_a_valid(1)
         val dReady = core.io.tc_d_ready(1)
       }
@@ -759,8 +755,8 @@ class RadianceTileModuleImp(outer: RadianceTile)
         val adapter = Module(
           new VortexTLAdapter(
             outer.smemSourceWidth,
-            new VortexBundleA(tagWidth = 4, dataWidth = 32 * 8),
+            new VortexBundleA(tagWidth = outer.tensorTagWidth, dataWidth = 32 * 8),
-            new VortexBundleD(tagWidth = 4, dataWidth = 32 * 8),
+            new VortexBundleD(tagWidth = outer.tensorTagWidth, dataWidth = 32 * 8),
             client
           )
         )
@@ -850,8 +846,31 @@ class RadianceTileModuleImp(outer: RadianceTile)
   // TODO: generalize for useVxCache
   if (!outer.radianceParams.useVxCache) {}
 
-  // connect io.start and io.finish of the fake TensorCoreDecoupled module
+  // Instantiate a fake tensor core module to force unique-ification of module
-  outer.tensor.module.io.start := false.B
+  // names in the Chisel-generated Verilog.  These should be left out for
+  // synthesis runs, although these will likely be optimized-out if the inputs
+  // are tied to low.
+
+  val tensorNumSourceIds = (1 << outer.tensorTagWidth)
+  val tensor = Module(new radiance.core.TensorCoreDecoupled(
+                      8, 8, half = true, tensorNumSourceIds))
+  tensor.io.initiate.valid := false.B
+  tensor.io.initiate.bits := DontCare
+  tensor.io.respA.valid := false.B
+  tensor.io.respA.bits := DontCare
+  tensor.io.respB.valid := false.B
+  tensor.io.respB.bits := DontCare
+  tensor.io.respC := DontCare
+  tensor.io.reqA.ready := false.B
+  tensor.io.reqB.ready := false.B
+  tensor.io.writeback.ready := false.B
+
+  val dpu = Module(new radiance.core.TensorDotProductUnit(8, half = true))
+  dpu.io.in.valid := false.B
+  dpu.io.in.bits.a := DontCare
+  dpu.io.in.bits.b := DontCare
+  dpu.io.in.bits.c := DontCare
+  dpu.io.stall := false.B
 
   // // RoCC
   // if (outer.roccs.size > 0) {

src/main/scala/radiance/tile/VortexCore.scala

@@ -342,11 +342,11 @@ class Vortex(tile: RadianceTile)(implicit p: Parameters)
   // tensor core
   // this module is referenced from inside the Verilog RTL of the core
   // pipeline.
-  if (tile.radianceParams.core.tensorCoreFP16) {
+  // if (tile.radianceParams.core.tensorCoreFP16) {
-    addResource("/vsrc/TensorDotProductUnit.sv")
+  //   addResource("/vsrc/TensorDotProductUnit.sv")
-  } else {
+  // } else {
-    addResource("/vsrc/TensorDotProductUnitFP32.sv")
+  //   addResource("/vsrc/TensorDotProductUnitFP32.sv")
-  }
+  // }
 
   // fpnew
   // compile order matters; package definitions (ex. fpnew_pkg) should be