Update dev with master

build.sc

@@ -185,6 +185,35 @@ trait Emulator extends Cross.Module2[String, String] {
     }
   }
 
+  object litexgenerate extends Module {
+    def compile = T {
+      os.proc("firtool",
+        generator.chirrtl().path,
+        s"--annotation-file=${generator.chiselAnno().path}",
+        "--disable-annotation-unknown",
+        "-dedup",
+        "-O=debug",
+        "--split-verilog",
+        "--preserve-values=named",
+        "--output-annotation-file=mfc.anno.json",
+        "--lowering-options=disallowLocalVariables",
+        s"-o=${T.dest}"
+      ).call(T.dest)
+      PathRef(T.dest)
+    }
+
+    def rtls = T {
+      os.read(compile().path / "filelist.f").split("\n").map(str =>
+        try {
+          os.Path(str)
+        } catch {
+          case e: IllegalArgumentException if e.getMessage.contains("is not an absolute path") =>
+            compile().path / str.stripPrefix("./")
+        }
+      ).filter(p => p.ext == "v" || p.ext == "sv").map(PathRef(_)).toSeq
+    }
+  }
+
   object mfccompiler extends Module {
     def compile = T {
       os.proc("firtool",
@@ -233,7 +262,7 @@ trait Emulator extends Cross.Module2[String, String] {
         "debug_rob.cc",
         "emulator.cc",
         "remote_bitbang.cc",
-        ).map(c => PathRef(csrcDir().path / c))
+      ).map(c => PathRef(csrcDir().path / c))
     }
 
     def CMakeListsString = T {
@@ -347,6 +376,39 @@ object emulator extends Cross[Emulator](
   ("freechips.rocketchip.system.TestHarness", "freechips.rocketchip.system.DefaultBConfig"),
   ("freechips.rocketchip.system.TestHarness", "freechips.rocketchip.system.DefaultRV32BConfig"),
 
+  // Litex
+  ("freechips.rocketchip.system.TestHarness", "freechips.rocketchip.system.LitexConfigSmall1x1"),
+  ("freechips.rocketchip.system.TestHarness", "freechips.rocketchip.system.LitexConfigSmall1x2"),
+  ("freechips.rocketchip.system.TestHarness", "freechips.rocketchip.system.LitexConfigSmall1x4"),
+  ("freechips.rocketchip.system.TestHarness", "freechips.rocketchip.system.LitexConfigSmall1x8"),
+  ("freechips.rocketchip.system.TestHarness", "freechips.rocketchip.system.LitexConfigSmall2x1"),
+  ("freechips.rocketchip.system.TestHarness", "freechips.rocketchip.system.LitexConfigSmall2x2"),
+  ("freechips.rocketchip.system.TestHarness", "freechips.rocketchip.system.LitexConfigSmall2x4"),
+  ("freechips.rocketchip.system.TestHarness", "freechips.rocketchip.system.LitexConfigSmall2x8"),
+  ("freechips.rocketchip.system.TestHarness", "freechips.rocketchip.system.LitexConfigSmall4x1"),
+  ("freechips.rocketchip.system.TestHarness", "freechips.rocketchip.system.LitexConfigSmall4x2"),
+  ("freechips.rocketchip.system.TestHarness", "freechips.rocketchip.system.LitexConfigSmall4x4"),
+  ("freechips.rocketchip.system.TestHarness", "freechips.rocketchip.system.LitexConfigSmall4x8"),
+  ("freechips.rocketchip.system.TestHarness", "freechips.rocketchip.system.LitexConfigSmall8x1"),
+  ("freechips.rocketchip.system.TestHarness", "freechips.rocketchip.system.LitexConfigSmall8x2"),
+  ("freechips.rocketchip.system.TestHarness", "freechips.rocketchip.system.LitexConfigSmall8x4"),
+  ("freechips.rocketchip.system.TestHarness", "freechips.rocketchip.system.LitexConfigSmall8x8"),
+  ("freechips.rocketchip.system.TestHarness", "freechips.rocketchip.system.LitexConfigBig1x1"),
+  ("freechips.rocketchip.system.TestHarness", "freechips.rocketchip.system.LitexConfigBig1x2"),
+  ("freechips.rocketchip.system.TestHarness", "freechips.rocketchip.system.LitexConfigBig1x4"),
+  ("freechips.rocketchip.system.TestHarness", "freechips.rocketchip.system.LitexConfigBig1x8"),
+  ("freechips.rocketchip.system.TestHarness", "freechips.rocketchip.system.LitexConfigBig2x1"),
+  ("freechips.rocketchip.system.TestHarness", "freechips.rocketchip.system.LitexConfigBig2x2"),
+  ("freechips.rocketchip.system.TestHarness", "freechips.rocketchip.system.LitexConfigBig2x4"),
+  ("freechips.rocketchip.system.TestHarness", "freechips.rocketchip.system.LitexConfigBig2x8"),
+  ("freechips.rocketchip.system.TestHarness", "freechips.rocketchip.system.LitexConfigBig4x1"),
+  ("freechips.rocketchip.system.TestHarness", "freechips.rocketchip.system.LitexConfigBig4x2"),
+  ("freechips.rocketchip.system.TestHarness", "freechips.rocketchip.system.LitexConfigBig4x4"),
+  ("freechips.rocketchip.system.TestHarness", "freechips.rocketchip.system.LitexConfigBig4x8"),
+  ("freechips.rocketchip.system.TestHarness", "freechips.rocketchip.system.LitexConfigBig8x1"),
+  ("freechips.rocketchip.system.TestHarness", "freechips.rocketchip.system.LitexConfigBig8x2"),
+  ("freechips.rocketchip.system.TestHarness", "freechips.rocketchip.system.LitexConfigBig8x4"),
+  ("freechips.rocketchip.system.TestHarness", "freechips.rocketchip.system.LitexConfigBig8x8"),
 )
 
 object `runnable-riscv-test` extends mill.Cross[RiscvTest](

src/main/resources/vsrc/TraceSinkMonitor.v

@@ -0,0 +1,36 @@
+module TraceSinkMonitor
+#(
+    parameter FILE_NAME = "trace_sink_monitor.txt"
+)
+(
+    input clk,
+    input reset,
+    input in_fire,
+    input[7:0] in_byte
+);
+
+`ifndef SYNTHESIS
+
+integer file;
+
+initial begin
+    file = $fopen(FILE_NAME, "w");
+    if (file == 0) begin
+        $display("Failed to open %s", FILE_NAME);
+        $finish;
+    end
+end
+
+always @(posedge clk) begin
+    if (in_fire & ~reset) begin
+        $fwrite(file, "%c", in_byte);
+    end
+end 
+
+final begin
+    $fclose(file);
+end
+
+`endif
+
+endmodule

src/main/scala/devices/debug/Debug.scala

@@ -660,7 +660,7 @@ class TLDebugModuleOuter(device: Device)(implicit p: Parameters) extends LazyMod
       val hartResetReg = RegNext(next=hartResetNxt, init=0.U.asTypeOf(hartResetNxt))
 
       for (component <- 0 until nComponents) {
-        hartResetNxt(component) := DMCONTROLReg.hartreset & hartSelected(component)
+        hartResetNxt(component) := DMCONTROLNxt.hartreset & hartSelected(component)
         io.hartResetReq.get(component) := hartResetReg(component)
       }
     }

src/main/scala/rocket/ALU.scala

@@ -81,6 +81,8 @@ abstract class AbstractALU(implicit p: Parameters) extends CoreModule()(p) {
 }
 
 class ALU(implicit p: Parameters) extends AbstractALU()(p) {
+  override def desiredName = "RocketALU"
+
   // ADD, SUB
   val in2_inv = Mux(isSub(io.fn), ~io.in2, io.in2)
   val in1_xor_in2 = io.in1 ^ in2_inv

src/main/scala/rocket/CSR.scala

@@ -149,7 +149,7 @@ class PTBR(implicit p: Parameters) extends CoreBundle()(p) {
     case 32 => (1, 9)
     case 64 => (4, 16)
   }
-  require(modeBits + maxASIdBits + maxPAddrBits - pgIdxBits == xLen)
+  require(!usingVM || modeBits + maxASIdBits + maxPAddrBits - pgIdxBits == xLen)
 
   val mode = UInt(modeBits.W)
   val asid = UInt(maxASIdBits.W)
@@ -272,8 +272,8 @@ class CSRFileIO(hasBeu: Boolean)(implicit p: Parameters) extends CoreBundle
   val csr_stall = Output(Bool()) // stall retire for wfi
   val rw_stall = Output(Bool()) // stall rw, rw will have no effect while rw_stall
   val eret = Output(Bool())
+  val trap_return = Output(Bool())
   val singleStep = Output(Bool())
-
   val status = Output(new MStatus())
   val hstatus = Output(new HStatus())
   val gstatus = Output(new MStatus())
@@ -998,6 +998,7 @@ class CSRFile(
   io.hgatp := reg_hgatp
   io.vsatp := reg_vsatp
   io.eret := insn_call || insn_break || insn_ret
+  io.trap_return := insn_ret
   io.singleStep := reg_dcsr.step && !reg_debug
   io.status := reg_mstatus
   io.status.sd := io.status.fs.andR || io.status.xs.andR || io.status.vs.andR

src/main/scala/rocket/Instructions.scala

@@ -582,8 +582,8 @@ object Instructions {
   def VMADC_VXM          = BitPat("b0100010??????????100?????1010111")
   def VMADD_VV           = BitPat("b101001???????????010?????1010111")
   def VMADD_VX           = BitPat("b101001???????????110?????1010111")
-  def VMAND_MM           = BitPat("b011001???????????010?????1010111")
-  def VMANDN_MM          = BitPat("b011000???????????010?????1010111")
+  def VMAND_MM           = BitPat("b0110011??????????010?????1010111")
+  def VMANDN_MM          = BitPat("b0110001??????????010?????1010111")
   def VMAX_VV            = BitPat("b000111???????????000?????1010111")
   def VMAX_VX            = BitPat("b000111???????????100?????1010111")
   def VMAXU_VV           = BitPat("b000110???????????000?????1010111")
@@ -605,10 +605,10 @@ object Instructions {
   def VMIN_VX            = BitPat("b000101???????????100?????1010111")
   def VMINU_VV           = BitPat("b000100???????????000?????1010111")
   def VMINU_VX           = BitPat("b000100???????????100?????1010111")
-  def VMNAND_MM          = BitPat("b011101???????????010?????1010111")
-  def VMNOR_MM           = BitPat("b011110???????????010?????1010111")
-  def VMOR_MM            = BitPat("b011010???????????010?????1010111")
-  def VMORN_MM           = BitPat("b011100???????????010?????1010111")
+  def VMNAND_MM          = BitPat("b0111011??????????010?????1010111")
+  def VMNOR_MM           = BitPat("b0111101??????????010?????1010111")
+  def VMOR_MM            = BitPat("b0110101??????????010?????1010111")
+  def VMORN_MM           = BitPat("b0111001??????????010?????1010111")
   def VMSBC_VV           = BitPat("b0100111??????????000?????1010111")
   def VMSBC_VVM          = BitPat("b0100110??????????000?????1010111")
   def VMSBC_VX           = BitPat("b0100111??????????100?????1010111")
@@ -653,8 +653,8 @@ object Instructions {
   def VMV_V_V            = BitPat("b010111100000?????000?????1010111")
   def VMV_V_X            = BitPat("b010111100000?????100?????1010111")
   def VMV_X_S            = BitPat("b0100001?????00000010?????1010111")
-  def VMXNOR_MM          = BitPat("b011111???????????010?????1010111")
-  def VMXOR_MM           = BitPat("b011011???????????010?????1010111")
+  def VMXNOR_MM          = BitPat("b0111111??????????010?????1010111")
+  def VMXOR_MM           = BitPat("b0110111??????????010?????1010111")
   def VNCLIP_WI          = BitPat("b101111???????????011?????1010111")
   def VNCLIP_WV          = BitPat("b101111???????????000?????1010111")
   def VNCLIP_WX          = BitPat("b101111???????????100?????1010111")

src/main/scala/rocket/PTW.scala

@@ -303,7 +303,7 @@ class PTW(n: Int)(implicit edge: TLEdgeOut, p: Parameters) extends CoreModule()(
   val (pte, invalid_paddr, invalid_gpa) = {
     val tmp = mem_resp_data.asTypeOf(new PTE())
     val res = WireDefault(tmp)
-    res.ppn := Mux(do_both_stages && !stage2, tmp.ppn(vpnBits.min(tmp.ppn.getWidth)-1, 0), tmp.ppn(ppnBits-1, 0))
+    res.ppn := Mux(do_both_stages && !stage2, tmp.ppn(vpnBits.min(tmp.ppn.getWidth)-1, 0), tmp.ppn(ppnBits.min(tmp.ppn.getWidth)-1, 0))
     when (tmp.r || tmp.w || tmp.x) {
       // for superpage mappings, make sure PPN LSBs are zero
       for (i <- 0 until pgLevels-1)

src/main/scala/rocket/RocketCore.scala

@@ -11,6 +11,7 @@ import freechips.rocketchip.tile._
 import freechips.rocketchip.util._
 import freechips.rocketchip.util.property
 import scala.collection.mutable.ArrayBuffer
+import freechips.rocketchip.trace._
 
 case class RocketCoreParams(
   xLen: Int = 64,
@@ -56,7 +57,8 @@ case class RocketCoreParams(
   debugROB: Option[DebugROBParams] = None, // if size < 1, SW ROB, else HW ROB
   haveCease: Boolean = true, // non-standard CEASE instruction
   haveSimTimeout: Boolean = true, // add plusarg for simulation timeout
-  vector: Option[RocketCoreVectorParams] = None
+  vector: Option[RocketCoreVectorParams] = None,
+  enableTraceCoreIngress: Boolean = false
 ) extends CoreParams {
   val lgPauseCycles = 5
   val haveFSDirty = false
@@ -131,6 +133,8 @@ class CoreInterrupts(val hasBeu: Boolean)(implicit p: Parameters) extends TileIn
 trait HasRocketCoreIO extends HasRocketCoreParameters {
   implicit val p: Parameters
   def nTotalRoCCCSRs: Int
+  def traceIngressParams = TraceCoreParams(nGroups = 1, iretireWidth = coreParams.retireWidth, 
+                                            xlen = coreParams.xLen, iaddrWidth = coreParams.xLen) 
   val io = IO(new CoreBundle()(p) {
     val hartid = Input(UInt(hartIdLen.W))
     val reset_vector = Input(UInt(resetVectorLen.W))
@@ -146,6 +150,7 @@ trait HasRocketCoreIO extends HasRocketCoreParameters {
     val wfi = Output(Bool())
     val traceStall = Input(Bool())
     val vector = if (usingVector) Some(Flipped(new VectorCoreIO)) else None
+    val trace_core_ingress = if (rocketParams.enableTraceCoreIngress) Some(Output(new TraceCoreInterface(traceIngressParams))) else None
   })
 }
 
@@ -301,6 +306,7 @@ class Rocket(tile: RocketTile)(implicit p: Parameters) extends CoreModule()(p)
   val wb_reg_raw_inst = Reg(UInt())
   val wb_reg_wdata = Reg(Bits())
   val wb_reg_rs2 = Reg(Bits())
+  val wb_reg_br_taken = Reg(Bool())
   val take_pc_wb = Wire(Bool())
   val wb_reg_wphit           = Reg(Vec(nBreakpoints, Bool()))
 
@@ -723,6 +729,7 @@ class Rocket(tile: RocketTile)(implicit p: Parameters) extends CoreModule()(p)
     wb_reg_hfence_v := mem_ctrl.mem_cmd === M_HFENCEV
     wb_reg_hfence_g := mem_ctrl.mem_cmd === M_HFENCEG
     wb_reg_pc := mem_reg_pc
+    wb_reg_br_taken := mem_br_taken
     wb_reg_wphit := mem_reg_wphit | bpu.io.bpwatch.map { bpw => (bpw.rvalid(0) && mem_reg_load) || (bpw.wvalid(0) && mem_reg_store) }
     wb_reg_set_vconfig := mem_reg_set_vconfig
   }
@@ -825,6 +832,27 @@ class Rocket(tile: RocketTile)(implicit p: Parameters) extends CoreModule()(p)
                  wb_reg_wdata))))
   when (rf_wen) { rf.write(rf_waddr, rf_wdata) }
 
+  if (rocketParams.enableTraceCoreIngress) {
+    val trace_ingress = Module(new TraceCoreIngress(traceIngressParams))
+    trace_ingress.io.in.valid := wb_valid || wb_xcpt
+    trace_ingress.io.in.taken := wb_reg_br_taken
+    trace_ingress.io.in.is_branch := wb_ctrl.branch
+    trace_ingress.io.in.is_jal := wb_ctrl.jal
+    trace_ingress.io.in.is_jalr := wb_ctrl.jalr
+    trace_ingress.io.in.insn := wb_reg_inst
+    trace_ingress.io.in.pc := wb_reg_pc
+    trace_ingress.io.in.is_compressed := !wb_reg_raw_inst(1, 0).andR // 2'b11 is uncompressed, everything else is compressed
+    trace_ingress.io.in.interrupt := csr.io.trace(0).interrupt && csr.io.trace(0).exception
+    trace_ingress.io.in.exception := !csr.io.trace(0).interrupt && csr.io.trace(0).exception
+    trace_ingress.io.in.trap_return := csr.io.trap_return
+
+    io.trace_core_ingress.get.group(0) <> trace_ingress.io.out
+    io.trace_core_ingress.get.priv := csr.io.trace(0).priv 
+    io.trace_core_ingress.get.tval := csr.io.tval
+    io.trace_core_ingress.get.cause := csr.io.cause
+    io.trace_core_ingress.get.time := csr.io.time
+  }
+
   // hook up control/status regfile
   csr.io.ungated_clock := clock
   csr.io.decode(0).inst := id_inst(0)
@@ -1176,6 +1204,7 @@ class Rocket(tile: RocketTile)(implicit p: Parameters) extends CoreModule()(p)
       !div.io.req.ready || // mul/div in flight
       usingFPU.B && !io.fpu.fcsr_rdy || // long-latency FPU in flight
       io.dmem.replay_next || // long-latency load replaying
+      id_rocc_busy || // RoCC command in flight
       (!long_latency_stall && (ibuf.io.inst(0).valid || io.imem.resp.valid)) // instruction pending
 
     assert(!(ex_pc_valid || mem_pc_valid || wb_pc_valid) || clock_en)

src/main/scala/rocket/TLB.scala

@@ -403,7 +403,7 @@ class TLB(instruction: Boolean, lgMaxSize: Int, cfg: TLBConfig)(implicit edge: T
   val vsatp_mode_mismatch  = priv_v && (vstage1_en =/= v_entries_use_stage1) && !io.req.bits.passthrough
 
   // share a single physical memory attribute checker (unshare if critical path)
-  val refill_ppn = io.ptw.resp.bits.pte.ppn(ppnBits-1, 0)
+  val refill_ppn = if (usingVM) io.ptw.resp.bits.pte.ppn(ppnBits-1, 0) else 0.U
   /** refill signal */
   val do_refill = usingVM.B && io.ptw.resp.valid
   /** sfence invalidate refill */

src/main/scala/subsystem/Cluster.scala

@@ -13,7 +13,7 @@ import freechips.rocketchip.interrupts.{IntIdentityNode, IntSyncIdentityNode, Nu
 import freechips.rocketchip.prci.{ClockCrossingType, NoCrossing, ClockSinkParameters, ClockGroupIdentityNode, BundleBridgeBlockDuringReset}
 import freechips.rocketchip.tile.{RocketTile, NMI, TraceBundle}
 import freechips.rocketchip.tilelink.TLWidthWidget
-import freechips.rocketchip.util.TraceCoreInterface
+import freechips.rocketchip.trace.TraceCoreInterface
 
 import scala.collection.immutable.SortedMap
 

src/main/scala/subsystem/HasHierarchicalElements.scala

@@ -18,7 +18,7 @@ import freechips.rocketchip.tile.{TileParams, TilePRCIDomain, BaseTile, NMI, Tra
 import freechips.rocketchip.tilelink.{TLNode, TLBuffer, TLCacheCork, TLTempNode, TLFragmenter}
 import freechips.rocketchip.prci.{ClockCrossingType, ClockGroup, ResetCrossingType, ClockGroupNode, ClockDomain}
 import freechips.rocketchip.rocket.TracedInstruction
-import freechips.rocketchip.util.TraceCoreInterface
+import freechips.rocketchip.trace.TraceCoreInterface
 
 import scala.collection.immutable.SortedMap
 

src/main/scala/subsystem/HasTiles.scala

@@ -15,7 +15,7 @@ import freechips.rocketchip.tile.{MaxHartIdBits, BaseTile, InstantiableTileParam
 import freechips.rocketchip.tilelink.TLWidthWidget
 import freechips.rocketchip.prci.{ClockGroup, BundleBridgeBlockDuringReset, NoCrossing, SynchronousCrossing, CreditedCrossing, RationalCrossing, AsynchronousCrossing}
 import freechips.rocketchip.rocket.TracedInstruction
-import freechips.rocketchip.util.TraceCoreInterface
+import freechips.rocketchip.trace.TraceCoreInterface
 
 import scala.collection.immutable.SortedMap
 

src/main/scala/subsystem/HierarchicalElementPRCIDomain.scala

@@ -12,7 +12,7 @@ import freechips.rocketchip.interrupts.{IntInwardNode, IntOutwardNode}
 import freechips.rocketchip.prci.{ClockCrossingType, ResetCrossingType, ResetDomain, ClockSinkNode, ClockSinkParameters, ClockIdentityNode, FixedClockBroadcast, ClockDomain}
 import freechips.rocketchip.tile.{RocketTile, TraceBundle}
 import freechips.rocketchip.tilelink.{TLInwardNode, TLOutwardNode}
-import freechips.rocketchip.util.TraceCoreInterface
+import freechips.rocketchip.trace.TraceCoreInterface
 
 import freechips.rocketchip.tilelink.TLClockDomainCrossing
 import freechips.rocketchip.tilelink.TLResetDomainCrossing

src/main/scala/subsystem/Litex.scala

@@ -0,0 +1,53 @@
+// See LICENSE.SiFive for license details.
+// See LICENSE.Berkeley for license details.
+
+package freechips.rocketchip.subsystem
+
+import chisel3.util._
+
+import org.chipsalliance.cde.config._
+import org.chipsalliance.diplomacy.lazymodule._
+
+import freechips.rocketchip.devices.debug.{DebugModuleKey, DefaultDebugModuleParams, ExportDebug, JTAG, APB}
+import freechips.rocketchip.devices.tilelink.{
+  BuiltInErrorDeviceParams, BootROMLocated, BootROMParams, CLINTKey, DevNullDevice, CLINTParams, PLICKey, PLICParams, DevNullParams
+}
+import freechips.rocketchip.prci.{SynchronousCrossing, AsynchronousCrossing, RationalCrossing, ClockCrossingType}
+import freechips.rocketchip.diplomacy.{
+  AddressSet, MonitorsEnabled,
+}
+import freechips.rocketchip.resources.{
+  DTSModel, DTSCompat, DTSTimebase, BigIntHexContext
+}
+import freechips.rocketchip.tile.{
+  MaxHartIdBits, RocketTileParams, BuildRoCC, AccumulatorExample, OpcodeSet, TranslatorExample, CharacterCountExample, BlackBoxExample
+}
+import freechips.rocketchip.util.ClockGateModelFile
+import scala.reflect.ClassTag
+
+class WithLitexMemPort extends Config((site, here, up) => {
+  case ExtMem => Some(MemoryPortParams(MasterPortParams(
+                      base = x"8000_0000",
+                      size = x"8000_0000",
+                      beatBytes = site(MemoryBusKey).beatBytes,
+                      idBits = 4), 1))
+})
+
+class WithLitexMMIOPort extends Config((site, here, up) => {
+  case ExtBus => Some(MasterPortParams(
+                      base = x"1000_0000",
+                      size = x"7000_0000",
+                      beatBytes = site(SystemBusKey).beatBytes,
+                      idBits = 4))
+})
+
+class WithLitexSlavePort extends Config((site, here, up) => {
+  case ExtIn  => Some(SlavePortParams(
+                      beatBytes = site(SystemBusKey).beatBytes,
+                      idBits = 8,
+                      sourceBits = 4))
+})
+
+class WithNBitMemoryBus(dataBits: Int) extends Config((site, here, up) => {
+  case MemoryBusKey => up(MemoryBusKey, site).copy(beatBytes = dataBits/8)
+})