riscv/rocket-chip
1
0
Fork 0
Code Releases Activity

axi4: simplify Fragmenter by using user bits

Browse Source
This commit is contained in:
Wesley W. Terpstra 2017-04-20 15:20:20 -07:00
parent 7a1d107c9e
commit e100a943ea
1 changed files with 34 additions and 119 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

153
src/main/scala/uncore/axi4/Fragmenter.scala
View File

@ -10,8 +10,7 @@ import diplomacy._
import scala.math.{min,max}
import uncore.tilelink2.{leftOR, rightOR, UIntToOH1, OH1ToOH}
// lite: masters all use only one ID => reads will not be interleaved
class AXI4Fragmenter(lite: Boolean = false, maxInFlight: => Int = 32, combinational: Boolean = true)(implicit p: Parameters) extends LazyModule
class AXI4Fragmenter()(implicit p: Parameters) extends LazyModule
{
val maxBeats = 1 << AXI4Parameters.lenBits
def expandTransfer(x: TransferSizes, beatBytes: Int, alignment: BigInt) =
@ -19,11 +18,11 @@ class AXI4Fragmenter(lite: Boolean = false, maxInFlight: => Int = 32, combinatio
def mapSlave(s: AXI4SlaveParameters, beatBytes: Int) = s.copy(
supportsWrite = expandTransfer(s.supportsWrite, beatBytes, s.minAlignment),
supportsRead = expandTransfer(s.supportsRead, beatBytes, s.minAlignment),
interleavedId = if (lite) Some(0) else s.interleavedId) // see AXI4FragmenterSideband for !lite case
interleavedId = None) // this breaks interleaving guarantees
def mapMaster(m: AXI4MasterParameters) = m.copy(aligned = true)
val node = AXI4AdapterNode(
masterFn = { mp => mp.copy(masters = mp.masters.map(m => mapMaster(m))) },
masterFn = { mp => mp.copy(masters = mp.masters.map(m => mapMaster(m)), userBits = mp.userBits + 1) },
slaveFn = { sp => sp.copy(slaves = sp.slaves .map(s => mapSlave(s, sp.beatBytes))) })
lazy val module = new LazyModuleImp(this) {
@ -40,9 +39,6 @@ class AXI4Fragmenter(lite: Boolean = false, maxInFlight: => Int = 32, combinatio
val master = edgeIn.master
val masters = master.masters
// If the user claimed this was a lite interface, then there must be only one Id
require (!lite || master.endId == 1)
// We don't support fragmenting to sub-beat accesses
slaves.foreach { s =>
require (!s.supportsRead || s.supportsRead.contains(beatBytes))
@ -139,154 +135,73 @@ class AXI4Fragmenter(lite: Boolean = false, maxInFlight: => Int = 32, combinatio
val readSizes1 = slaves.map(s => s.supportsRead .max/beatBytes-1)
val writeSizes1 = slaves.map(s => s.supportsWrite.max/beatBytes-1)
// Indirection variables for inputs and outputs; makes transformation application easier
// Irrevocable queues in front because we want to accept the request before responses come back
val (in_ar, ar_last, _) = fragment(Queue.irrevocable(in.ar, 1, flow=true), readSizes1)
val (in_aw, aw_last, w_beats) = fragment(Queue.irrevocable(in.aw, 1, flow=true), writeSizes1)
val in_w = in.w
val in_r = in.r
val in_b = in.b
val out_ar = Wire(out.ar)
val out_aw = out.aw
val out_w = out.w
val out_r = Wire(out.r)
val out_b = Wire(out.b)
val depth = if (combinational) 1 else 2
// In case a slave ties arready := rready, we need a queue to break the combinational loop
// between the two branches (in_ar => {out_ar => out_r, sideband} => in_r).
if (in.ar.bits.getWidth < in.r.bits.getWidth) {
out.ar <> Queue(out_ar, depth, flow=combinational)
out_r <> out.r
} else {
out.ar <> out_ar
out_r <> Queue(out.r, depth, flow=combinational)
}
// In case a slave ties awready := bready or wready := bready, we need this queue
out_b <> Queue(out.b, depth, flow=combinational)
// Sideband to track which transfers were the last fragment
def sideband() = if (lite) {
Module(new Queue(Bool(), maxInFlight, flow=combinational)).io
} else {
Module(new AXI4FragmenterSideband(maxInFlight, flow=combinational)).io
}
val sideband_ar_r = sideband()
val sideband_aw_b = sideband()
// AR flow control
out_ar.valid := in_ar.valid && sideband_ar_r.enq.ready
in_ar.ready := sideband_ar_r.enq.ready && out_ar.ready
sideband_ar_r.enq.valid := in_ar.valid && out_ar.ready
out_ar.bits := in_ar.bits
sideband_ar_r.enq.bits := ar_last
// AR flow control; super easy
out.ar <> in_ar
out.ar.bits.user.get := Cat(in_ar.bits.user.toList ++ Seq(ar_last))
// When does W channel start counting a new transfer
val wbeats_latched = RegInit(Bool(false))
val wbeats_ready = Wire(Bool())
val wbeats_valid = Wire(Bool())
when (wbeats_valid && wbeats_ready) { wbeats_latched := Bool(true) }
when (out_aw.fire()) { wbeats_latched := Bool(false) }
when (out.aw.fire()) { wbeats_latched := Bool(false) }
// AW flow control
out_aw.valid := in_aw.valid && sideband_aw_b.enq.ready && (wbeats_ready || wbeats_latched)
in_aw.ready := sideband_aw_b.enq.ready && out_aw.ready && (wbeats_ready || wbeats_latched)
sideband_aw_b.enq.valid := in_aw.valid && out_aw.ready && (wbeats_ready || wbeats_latched)
out.aw.valid := in_aw.valid && (wbeats_ready || wbeats_latched)
in_aw.ready := out.aw.ready && (wbeats_ready || wbeats_latched)
wbeats_valid := in_aw.valid && !wbeats_latched
out_aw.bits := in_aw.bits
sideband_aw_b.enq.bits := aw_last
out.aw.bits := in_aw.bits
out.aw.bits.user.get := Cat(in_aw.bits.user.toList ++ Seq(aw_last))
// We need to inject 'last' into the W channel fragments, count!
val w_counter = RegInit(UInt(0, width = AXI4Parameters.lenBits+1))
val w_idle = w_counter === UInt(0)
val w_todo = Mux(w_idle, Mux(wbeats_valid, w_beats, UInt(0)), w_counter)
val w_last = w_todo === UInt(1)
w_counter := w_todo - out_w.fire()
assert (!out_w.fire() || w_todo =/= UInt(0)) // underflow impossible
w_counter := w_todo - out.w.fire()
assert (!out.w.fire() || w_todo =/= UInt(0)) // underflow impossible
// W flow control
wbeats_ready := w_idle
out_w.valid := in_w.valid && (!wbeats_ready || wbeats_valid)
in_w.ready := out_w.ready && (!wbeats_ready || wbeats_valid)
out_w.bits := in_w.bits
out_w.bits.last := w_last
out.w.valid := in.w.valid && (!wbeats_ready || wbeats_valid)
in.w.ready := out.w.ready && (!wbeats_ready || wbeats_valid)
out.w.bits := in.w.bits
out.w.bits.last := w_last
// We should also recreate the last last
assert (!out_w.valid || !in_w.bits.last || w_last)
assert (!out.w.valid || !in.w.bits.last || w_last)
// R flow control
val r_last = out_r.bits.last
in_r.valid := out_r.valid && (!r_last || sideband_ar_r.deq.valid)
out_r.ready := in_r.ready && (!r_last || sideband_ar_r.deq.valid)
sideband_ar_r.deq.ready := r_last && out_r.valid && in_r.ready
in_r.bits := out_r.bits
in_r.bits.last := r_last && sideband_ar_r.deq.bits
val r_last = out.r.bits.user.get(0)
in.r <> out.r
in.r.bits.last := out.r.bits.last && r_last
in.r.bits.user.foreach { _ := out.r.bits.user.get >> 1 }
// B flow control
val b_last = sideband_aw_b.deq.bits
in_b.valid := out_b.valid && sideband_aw_b.deq.valid && b_last
out_b.ready := sideband_aw_b.deq.valid && (!b_last || in_b.ready)
sideband_aw_b.deq.ready := out_b.valid && (!b_last || in_b.ready)
in_b.bits := out_b.bits
val b_last = out.b.bits.user.get(0)
in.b <> out.b
in.b.valid := out.b.valid && b_last
out.b.ready := in.b.ready || !b_last
in.b.bits.user.foreach { _ := out.b.bits.user.get >> 1 }
// Merge errors from dropped B responses
val r_resp = RegInit(UInt(0, width = AXI4Parameters.respBits))
val resp = out_b.bits.resp | r_resp
when (out_b.fire()) { r_resp := Mux(b_last, UInt(0), resp) }
in_b.bits.resp := resp
val error = RegInit(Vec.fill(edgeIn.master.endId) { UInt(0, width = AXI4Parameters.respBits)})
in.b.bits.resp := out.b.bits.resp | error(out.b.bits.id)
(error zip UIntToOH(out.b.bits.id, edgeIn.master.endId).toBools) foreach { case (reg, sel) =>
when (sel && out.b.fire()) { reg := Mux(b_last, UInt(0), reg | out.b.bits.resp) }
}
}
/* We want to put barriers between the fragments of a fragmented transfer and all other transfers.
* This lets us use very little state to reassemble the fragments (else we need one FIFO per ID).
* Furthermore, because all the fragments share the same AXI ID, they come back contiguously.
* This guarantees that no other R responses might get mixed between fragments, ensuring that the
* interleavedId for the slaves remains unaffected by the fragmentation transformation.
* Of course, if you need to fragment, this means there is a potentially hefty serialization cost.
* However, this design allows full concurrency in the common no-fragmentation-needed scenario.
*/
class AXI4FragmenterSideband(maxInFlight: Int, flow: Boolean = false) extends Module
{
val io = new QueueIO(Bool(), maxInFlight)
io.count := UInt(0)
val PASS = UInt(2, width = 2) // allow 'last=1' bits to enque, on 'last=0' if count>0 block else accept+FIND
val FIND = UInt(0, width = 2) // allow 'last=0' bits to enque, accept 'last=1' and switch to WAIT
val WAIT = UInt(1, width = 2) // block all access till count=0
val state = RegInit(PASS)
val count = RegInit(UInt(0, width = log2Up(maxInFlight)))
val full = count === UInt(maxInFlight-1)
val empty = count === UInt(0)
val last = count === UInt(1)
io.deq.bits := state(1) || (last && state(0)) // PASS || (last && WAIT)
io.deq.valid := !empty
io.enq.ready := !full && (empty || (state === FIND) || (state === PASS && io.enq.bits))
// WAIT => count > 0
assert (state =/= WAIT || count =/= UInt(0))
if (flow) {
when (io.enq.valid) {
io.deq.valid := Bool(true)
when (empty) { io.deq.bits := io.enq.bits }
}
}
count := count + io.enq.fire() - io.deq.fire()
switch (state) {
is(PASS) { when (io.enq.valid && !io.enq.bits && empty) { state := FIND } }
is(FIND) { when (io.enq.valid && io.enq.bits && !full) { state := Mux(empty, PASS, WAIT) } }
is(WAIT) { when (last && io.deq.ready) { state := PASS } }
}
}
}
object AXI4Fragmenter
{
// applied to the AXI4 source node; y.node := AXI4Fragmenter()(x.node)
def apply(lite: Boolean = false, maxInFlight: => Int = 32, combinational: Boolean = true)(x: AXI4OutwardNode)(implicit p: Parameters, sourceInfo: SourceInfo): AXI4OutwardNode = {
val fragmenter = LazyModule(new AXI4Fragmenter(lite, maxInFlight, combinational))
def apply()(x: AXI4OutwardNode)(implicit p: Parameters, sourceInfo: SourceInfo): AXI4OutwardNode = {
val fragmenter = LazyModule(new AXI4Fragmenter)
fragmenter.node := x
fragmenter.node
}