def __init__(self, data_width, max_line_length):
super().__init__("agg_aligner")
# Capture to the object
self.data_width = data_width
self.max_line_length = max_line_length
self.counter_width = clog2(self.max_line_length)
# Clock and Reset
self._clk = self.clock("clk")
self._rst_n = self.reset("rst_n")
# Inputs
self._in_dat = self.input("in_dat", self.data_width)
self._in_valid = self.input("in_valid", 1)
self._line_length = self.input("line_length", self.counter_width)
self._line_length.add_attribute(
ConfigRegAttr("Line Length/Image Width for alignment"))
# Outputs
self._out_dat = self.output("out_dat", self.data_width)
self._out_valid = self.output("out_valid", 1)
self._out_align = self.output("align", 1)
# Local Signals
self._cnt = self.var("cnt", self.counter_width)
# Generate
self.add_code(self.update_cnt)
self.add_code(self.set_align)
self.wire(self._out_dat, self._in_dat)
self.wire(self._out_valid, self._in_valid)
def visit(self, node):
if isinstance(node, _kratos.Generator):
ports_ = node.get_port_names()
for port_name in ports_:
curr_port = node.get_port(port_name)
attrs = curr_port.find_attribute(
lambda a: isinstance(a, ConfigRegAttr))
annotation_attr = curr_port.find_attribute(
lambda a: isinstance(a, FormalAttr))
if port_name is "mode":
print("Found mode...")
print(attrs)
if len(attrs) != 1:
continue
cr_attr = attrs[0]
doc = cr_attr.get_documentation()
# need to wire it to the instances parent and up
gen = node
parent_gen = gen.parent_generator()
child_port = curr_port
child_gen = gen
top_lvl_cfg = parent_gen is None
while parent_gen is not None:
# create a port based on the target's definition
new_name = child_gen.instance_name + "_" + child_port.name
p = parent_gen.port(child_port, new_name, False)
parent_gen.wire(child_port, p)
# move up the hierarchy
child_port = p
child_gen = parent_gen
parent_gen = parent_gen.parent_generator()
# Only add the attribute if this is a newly created port, not a top-level cfg reg
if top_lvl_cfg is False:
child_port_cra = ConfigRegAttr()
child_port_cra.set_documentation(doc)
child_port.add_attribute(child_port_cra)
if annotation_attr:
ann_att = annotation_attr[0]
annot_type = ann_att.get_formal_ann()
child_port.add_attribute(
FormalAttr(f"{child_port}", annot_type))
def __init__(self, name, data_width, init_val):
super().__init__(f"config_reg_{name}_{data_width}")
self.add_attribute(ConfigRegAttr())
self.init_val = init_val
self.data_width = data_width
self._out = self.output("o_data_out", data_width)
self._clk = self.clock("i_clk")
self._rst_n = self.reset("i_rst_n")
self.add_code(self.set_out)
def __init__(self, data_width, interconnect_output_ports):
super().__init__("Chain", debug=True)
# generator parameters
self.data_width = data_width
self.interconnect_output_ports = interconnect_output_ports
# chain enable configuration register
self._chain_en = self.input("chain_en", 1)
self._chain_en.add_attribute(
ConfigRegAttr("Signal indicating whether to enable chaining"))
self._chain_en.add_attribute(
FormalAttr(self._chain_en.name, FormalSignalConstraint.SET0))
# inputs
self._curr_tile_data_out = self.input(
"curr_tile_data_out",
self.data_width,
size=self.interconnect_output_ports,
packed=True,
explicit_array=True)
self._chain_data_in = self.input("chain_data_in",
self.data_width,
size=self.interconnect_output_ports,
packed=True,
explicit_array=True)
self._accessor_output = self.input("accessor_output",
self.interconnect_output_ports)
self._data_out_tile = self.output("data_out_tile",
self.data_width,
size=self.interconnect_output_ports,
packed=True,
explicit_array=True)
self.add_code(self.set_data_out)
def __init__(self, mem_params, word_width):
super().__init__("lake_mem", debug=True)
################################################################
# PARAMETERS
################################################################
# print("MEM PARAMS ", mem_params)
# basic parameters
self.word_width = word_width
# general memory parameters
self.mem_name = mem_params["name"]
self.capacity = mem_params["capacity"]
self.rw_same_cycle = mem_params["rw_same_cycle"]
self.use_macro = mem_params["use_macro"]
self.macro_name = mem_params["macro_name"]
# number of port types
self.num_read_write_ports = mem_params["num_read_write_ports"]
self.num_read_only_ports = mem_params["num_read_ports"]
self.num_write_only_ports = mem_params["num_write_ports"]
self.num_read_ports = self.num_read_only_ports + self.num_read_write_ports
self.num_write_ports = self.num_write_only_ports + self.num_read_write_ports
# info for port types
self.write_info = mem_params["write_info"]
self.read_info = mem_params["read_info"]
self.read_write_info = mem_params["read_write_info"]
# TODO change - for now, we assume you cannot have read/write and read or write ports
# should be the max of write vs read_write and need to handle more general case
if self.num_read_write_ports == 0:
self.write_width = mem_params["write_port_width"]
self.read_width = mem_params["read_port_width"]
else:
self.write_width = mem_params["read_write_port_width"]
self.read_width = mem_params["read_write_port_width"]
assert self.capacity % self.write_width == 0, \
"Memory capacity is not a multiple of the port width for writes"
assert self.capacity % self.read_width == 0, \
"Memory capacity is not a multiple of the port width for reads"
# innermost dimension for size of memory is the size of whichever port
# type has a wider width between reads and writes
self.mem_size = max(self.read_width, self.write_width)
# this assert has to be true if previous two asserts are true
assert self.capacity % self.mem_size == 0
# this is the last dimension for size of memory - equal to the number
# of the port type with wider width addresses can fit in the memory
self.mem_last_dim = int(self.capacity / self.mem_size)
self.mem_size_bits = max(1, clog2(self.mem_size))
self.mem_last_dim_bits = max(1, clog2(self.mem_last_dim))
# chaining parameters and config regs
self.chaining = mem_params["chaining"]
self.num_chain = mem_params["num_chain"]
self.num_chain_bits = clog2(self.num_chain)
if self.chaining:
self.chain_index = self.var("chain_index",
width=self.num_chain_bits)
self.chain_index.add_attribute(
ConfigRegAttr("Chain index for chaining"))
self.chain_index.add_attribute(
FormalAttr(self.chain_index.name, FormalSignalConstraint.SET0))
# minimum required widths for address signals
if self.mem_size == self.write_width and self.mem_size == self.read_width:
self.write_addr_width = self.mem_last_dim_bits + self.num_chain_bits
self.read_addr_width = self.mem_last_dim_bits + self.num_chain_bits
elif self.mem_size == self.write_width:
self.write_addr_width = self.mem_last_dim_bits + self.num_chain_bits
self.read_addr_width = self.mem_size_bits + self.mem_last_dim_bits + self.num_chain_bits
elif self.mem_size == self.read_width:
self.write_addr_width = self.mem_size_bits + self.mem_last_dim_bits + self.num_chain_bits
self.read_addr_width = self.mem_last_dim_bits + self.num_chain_bits
else:
print("Error occurred! Memory size does not make sense.")
################################################################
# I/O INTERFACE (WITHOUT ADDRESSING) + MEMORY
################################################################
self.clk = self.clock("clk")
# active low asynchornous reset
self.rst_n = self.reset("rst_n", 1)
self.data_in = self.input("data_in",
width=self.word_width,
size=(self.num_write_ports,
self.write_width),
explicit_array=True,
packed=True)
self.chain_en = self.input("chain_en", 1)
# write enable (high: write, low: read when rw_same_cycle = False, else
# only indicates write)
self.write = self.input("write", width=1, size=self.num_write_ports)
self.data_out = self.output("data_out",
width=self.word_width,
size=(self.num_read_ports,
self.read_width),
explicit_array=True,
packed=True)
self.write_chain = self.var("write_chain",
width=1,
size=self.num_write_ports)
if self.use_macro:
self.read_write_addr = self.input("read_write_addr",
width=self.addr_width,
size=self.num_read_write_ports,
explicit_array=True)
sram = SRAM(not self.use_macro, self.macro_name, word_width,
mem_params["read_write_port_width"],
mem_params["capacity"],
mem_params["num_read_write_ports"],
mem_params["num_read_write_ports"],
clog2(mem_params["capacity"]), 0, 1)
self.add_child(
"SRAM_" + mem_params["name"],
sram,
clk=self.clk,
clk_en=1,
mem_data_in_bank=self.data_in,
mem_data_out_bank=self.data_out,
mem_addr_in_bank=self.read_write_addr,
# TODO adjust
mem_cen_in_bank=1,
mem_wen_in_bank=self.write_chain,
wtsel=0,
rtsel=1)
else:
# memory variable (not I/O)
self.memory = self.var("memory",
width=self.word_width,
size=(self.mem_last_dim, self.mem_size),
explicit_array=True,
packed=True)
################################################################
# ADDRESSING I/O AND SIGNALS
################################################################
# I/O is different depending on whether we have read and write ports or
# read/write ports
# we keep address width at 16 to avoid unpacked
# safe_wire errors for addr in hw_top_lake - can change by changing
# default_config_width for those addr gens while accounting for muxing
# bits, but the extra bits are unused anyway
if self.rw_same_cycle:
self.read = self.input("read",
width=1,
size=self.num_read_ports)
else:
self.read = self.var("read", width=1, size=self.num_read_ports)
for i in range(self.num_read_ports):
self.wire(self.read[i], 1)
# TODO change later - same read/write or read and write assumption as above
if self.num_write_only_ports != 0 and self.num_read_only_ports != 0:
# writes
self.write_addr = self.input(
"write_addr",
width=16, # self.write_addr_width,
size=self.num_write_ports,
explicit_array=True)
assert self.write_info[0]["latency"] > 0, \
"Latency for write ports must be greater than 1 clock cycle."
# reads
self.read_addr = self.input(
"read_addr",
width=16, # self.read_addr_width,
size=self.num_read_ports,
explicit_array=True)
# TODO for now assuming all read ports have same latency
# TODO also should add support for other latencies
# rw_same_cycle is not valid here because read/write share the same port
elif self.num_read_write_ports != 0:
self.read_write_addr = self.input(
"read_write_addr",
width=
16, # max(self.read_addr_width, self.write_addr_width),
size=self.num_read_write_ports,
explicit_array=True)
# writes
self.write_addr = self.var(
"write_addr",
width=16, # self.write_addr_width,
size=self.num_read_write_ports,
explicit_array=True)
for p in range(self.num_read_write_ports):
safe_wire(gen=self,
w_to=self.write_addr[p],
w_from=self.read_write_addr[p])
# reads
self.read_addr = self.var("read_addr",
width=self.read_addr_width,
size=self.num_read_write_ports,
explicit_array=True)
for p in range(self.num_read_write_ports):
safe_wire(gen=self,
w_to=self.read_addr[p],
w_from=self.read_write_addr[p])
# TODO in self.read_write_info we should allow for different read
# and write latencies?
self.read_info = self.read_write_info
# TODO just doing chaining for SRAM
if self.chaining and self.num_read_write_ports > 0:
self.wire(
self.write_chain,
# chaining not enabled
(
~self.chain_en |
# chaining enabled
(self.chain_en &
(self.chain_index == self.read_write_addr[
self.write_addr_width + self.num_chain_bits,
self.write_addr_width]))) & self.write)
# chaining not supported
else:
self.wire(self.write_chain, self.write)
if self.use_macro:
self.wire(sram.ports.mem_wen_in_bank, self.write_chain)
self.add_write_data_block()
self.add_read_data_block()
def __init__(self,
data_width=16,
banks=1,
memory_width=64,
rw_same_cycle=False,
read_delay=1,
addr_width=9):
super().__init__("strg_fifo")
# Generation parameters
self.banks = banks
self.data_width = data_width
self.memory_width = memory_width
self.rw_same_cycle = rw_same_cycle
self.read_delay = read_delay
self.addr_width = addr_width
self.fw_int = int(self.memory_width / self.data_width)
# assert banks > 1 or rw_same_cycle is True or self.fw_int > 1, \
# "Can't sustain throughput with this setup. Need potential bandwidth for " + \
# "1 write and 1 read in a cycle - try using more banks or a macro that supports 1R1W"
# Clock and Reset
self._clk = self.clock("clk")
self._rst_n = self.reset("rst_n")
# Inputs + Outputs
self._push = self.input("push", 1)
self._data_in = self.input("data_in", self.data_width)
self._pop = self.input("pop", 1)
self._data_out = self.output("data_out", self.data_width)
self._valid_out = self.output("valid_out", 1)
self._empty = self.output("empty", 1)
self._full = self.output("full", 1)
# get relevant signals from the storage banks
self._data_from_strg = self.input("data_from_strg", self.data_width,
size=(self.banks,
self.fw_int),
explicit_array=True,
packed=True)
self._wen_addr = self.var("wen_addr", self.addr_width,
size=self.banks,
explicit_array=True,
packed=True)
self._ren_addr = self.var("ren_addr", self.addr_width,
size=self.banks,
explicit_array=True,
packed=True)
self._front_combined = self.var("front_combined", self.data_width,
size=self.fw_int,
explicit_array=True,
packed=True)
self._data_to_strg = self.output("data_to_strg", self.data_width,
size=(self.banks,
self.fw_int),
explicit_array=True,
packed=True)
self._wen_to_strg = self.output("wen_to_strg", self.banks)
self._ren_to_strg = self.output("ren_to_strg", self.banks)
self._num_words_mem = self.var("num_words_mem", self.data_width)
if self.banks == 1:
self._curr_bank_wr = self.var("curr_bank_wr", 1)
self.wire(self._curr_bank_wr, kts.const(0, 1))
self._curr_bank_rd = self.var("curr_bank_rd", 1)
self.wire(self._curr_bank_rd, kts.const(0, 1))
else:
self._curr_bank_wr = self.var("curr_bank_wr", kts.clog2(self.banks))
self._curr_bank_rd = self.var("curr_bank_rd", kts.clog2(self.banks))
self._write_queue = self.var("write_queue", self.data_width,
size=(self.banks,
self.fw_int),
explicit_array=True,
packed=True)
# Lets us know if the bank has a write queued up
self._queued_write = self.var("queued_write", self.banks)
self._front_data_out = self.var("front_data_out", self.data_width)
self._front_pop = self.var("front_pop", 1)
self._front_empty = self.var("front_empty", 1)
self._front_full = self.var("front_full", 1)
self._front_valid = self.var("front_valid", 1)
self._front_par_read = self.var("front_par_read", 1)
self._front_par_out = self.var("front_par_out", self.data_width,
size=(self.fw_int,
1),
explicit_array=True,
packed=True)
self._front_rd_ptr = self.var("front_rd_ptr", max(1, clog2(self.fw_int)))
self._front_push = self.var("front_push", 1)
self.wire(self._front_push, self._push & (~self._full | self._pop))
self._front_rf = RegFIFO(data_width=self.data_width,
width_mult=1,
depth=self.fw_int,
parallel=True,
break_out_rd_ptr=True)
# This one breaks out the read pointer so we can properly
# reorder the data to storage
self.add_child("front_rf", self._front_rf,
clk=self._clk,
clk_en=kts.const(1, 1),
rst_n=self._rst_n,
push=self._front_push,
pop=self._front_pop,
empty=self._front_empty,
full=self._front_full,
valid=self._front_valid,
parallel_read=self._front_par_read,
parallel_load=kts.const(0, 1), # We don't need to parallel load the front
parallel_in=0, # Same reason as above
parallel_out=self._front_par_out,
num_load=0,
rd_ptr_out=self._front_rd_ptr)
self.wire(self._front_rf.ports.data_in[0], self._data_in)
self.wire(self._front_data_out, self._front_rf.ports.data_out[0])
self._back_data_in = self.var("back_data_in", self.data_width)
self._back_data_out = self.var("back_data_out", self.data_width)
self._back_push = self.var("back_push", 1)
self._back_empty = self.var("back_empty", 1)
self._back_full = self.var("back_full", 1)
self._back_valid = self.var("back_valid", 1)
self._back_pl = self.var("back_pl", 1)
self._back_par_in = self.var("back_par_in", self.data_width,
size=(self.fw_int,
1),
explicit_array=True,
packed=True)
self._back_num_load = self.var("back_num_load", clog2(self.fw_int) + 1)
self._back_occ = self.var("back_occ", clog2(self.fw_int) + 1)
self._front_occ = self.var("front_occ", clog2(self.fw_int) + 1)
self._back_rf = RegFIFO(data_width=self.data_width,
width_mult=1,
depth=self.fw_int,
parallel=True,
break_out_rd_ptr=False)
self._fw_is_1 = self.var("fw_is_1", 1)
self.wire(self._fw_is_1, kts.const(self.fw_int == 1, 1))
self._back_pop = self.var("back_pop", 1)
if self.fw_int == 1:
self.wire(self._back_pop, self._pop & (~self._empty | self._push) & ~self._back_pl)
else:
self.wire(self._back_pop, self._pop & (~self._empty | self._push))
self.add_child("back_rf", self._back_rf,
clk=self._clk,
clk_en=kts.const(1, 1),
rst_n=self._rst_n,
push=self._back_push,
pop=self._back_pop,
empty=self._back_empty,
full=self._back_full,
valid=self._back_valid,
parallel_read=kts.const(0, 1),
# Only do back load when data is going there
parallel_load=self._back_pl & self._back_num_load.r_or(),
parallel_in=self._back_par_in,
num_load=self._back_num_load)
self.wire(self._back_rf.ports.data_in[0], self._back_data_in)
self.wire(self._back_data_out, self._back_rf.ports.data_out[0])
# send the writes through when a read isn't happening
for i in range(self.banks):
self.add_code(self.send_writes, idx=i)
self.add_code(self.send_reads, idx=i)
# Set the parallel load to back bank - if no delay it's immediate
# if not, it's delayed :)
if self.read_delay == 1:
self._ren_delay = self.var("ren_delay", 1)
self.add_code(self.set_parallel_ld_delay_1)
self.wire(self._back_pl, self._ren_delay)
else:
self.wire(self._back_pl, self._ren_to_strg.r_or())
# Combine front end data - just the items + incoming
# this data is actually based on the rd_ptr from the front fifo
for i in range(self.fw_int):
self.wire(self._front_combined[i], self._front_par_out[self._front_rd_ptr + i])
# This is always true
# self.wire(self._front_combined[self.fw_int - 1], self._data_in)
# prioritize queued writes, otherwise send combined data
for i in range(self.banks):
self.wire(self._data_to_strg[i],
kts.ternary(self._queued_write[i], self._write_queue[i], self._front_combined))
# Wire the thin output from front to thin input to back
self.wire(self._back_data_in, self._front_data_out)
self.wire(self._back_push, self._front_valid)
self.add_code(self.set_front_pop)
# Queue writes
for i in range(self.banks):
self.add_code(self.set_write_queue, idx=i)
# Track number of words in memory
# if self.read_delay == 1:
# self.add_code(self.set_num_words_mem_delay)
# else:
self.add_code(self.set_num_words_mem)
# Track occupancy of the two small fifos
self.add_code(self.set_front_occ)
self.add_code(self.set_back_occ)
if self.banks > 1:
self.add_code(self.set_curr_bank_wr)
self.add_code(self.set_curr_bank_rd)
if self.read_delay == 1:
self._prev_bank_rd = self.var("prev_bank_rd", max(1, kts.clog2(self.banks)))
self.add_code(self.set_prev_bank_rd)
# Parallel load data to back - based on num_load
index_into = self._curr_bank_rd
if self.read_delay == 1:
index_into = self._prev_bank_rd
for i in range(self.fw_int - 1):
# Shift data over if you bypassed from the memory output
self.wire(self._back_par_in[i],
kts.ternary(self._back_num_load == self.fw_int,
self._data_from_strg[index_into][i],
self._data_from_strg[index_into][i + 1]))
self.wire(self._back_par_in[self.fw_int - 1],
kts.ternary(self._back_num_load == self.fw_int,
self._data_from_strg[index_into][self.fw_int - 1],
kts.const(0, self.data_width)))
# Set the parallel read to the front fifo - analogous with trying to write to the memory
self.add_code(self.set_front_par_read)
# Set the number being parallely loaded into the register
self.add_code(self.set_back_num_load)
# Data out and valid out are (in the general case) just the data and valid from the back fifo
# In the case where we have a fresh memory read, it would be from that
bank_idx_read = self._curr_bank_rd
if self.read_delay == 1:
bank_idx_read = self._prev_bank_rd
self.wire(self._data_out,
kts.ternary(self._back_pl, self._data_from_strg[bank_idx_read][0], self._back_data_out))
self.wire(self._valid_out, kts.ternary(self._back_pl, self._pop, self._back_valid))
# Set addresses to storage
for i in range(self.banks):
self.add_code(self.set_wen_addr, idx=i)
self.add_code(self.set_ren_addr, idx=i)
# Now deal with a shared address vs separate addresses
if self.rw_same_cycle:
# Separate
self._wen_addr_out = self.output("wen_addr_out", self.addr_width,
size=self.banks,
explicit_array=True,
packed=True)
self._ren_addr_out = self.output("ren_addr_out", self.addr_width,
size=self.banks,
explicit_array=True,
packed=True)
self.wire(self._wen_addr_out, self._wen_addr)
self.wire(self._ren_addr_out, self._ren_addr)
else:
self._addr_out = self.output("addr_out", self.addr_width,
size=self.banks,
explicit_array=True,
packed=True)
# If sharing the addresses, send read addr with priority
for i in range(self.banks):
self.wire(self._addr_out[i],
kts.ternary(self._wen_to_strg[i], self._wen_addr[i], self._ren_addr[i]))
# Do final empty/full
self._num_items = self.var("num_items", self.data_width)
self.add_code(self.set_num_items)
self._fifo_depth = self.input("fifo_depth", self.data_width)
self._fifo_depth.add_attribute(ConfigRegAttr("Fifo depth..."))
self.wire(self._empty, self._num_items == 0)
self.wire(self._full, self._num_items == (self._fifo_depth))
def __init__(self,
word_width,
input_ports,
output_ports,
memories,
edges):
super().__init__("LakeTop", debug=True)
# parameters
self.word_width = word_width
self.input_ports = input_ports
self.output_ports = output_ports
self.default_config_width = 16
self.cycle_count_width = 16
self.stencil_valid = False
# objects
self.memories = memories
self.edges = edges
# tile enable and clock
self.tile_en = self.input("tile_en", 1)
self.tile_en.add_attribute(ConfigRegAttr("Tile logic enable manifested as clock gate"))
self.tile_en.add_attribute(FormalAttr(self.tile_en.name, FormalSignalConstraint.SET1))
self.clk_mem = self.clock("clk")
self.clk_mem.add_attribute(FormalAttr(self.clk_mem.name, FormalSignalConstraint.CLK))
# chaining
chain_supported = False
for mem in self.memories.keys():
if self.memories[mem]["chaining"]:
chain_supported = True
break
if chain_supported:
self.chain_en = self.input("chain_en", 1)
self.chain_en.add_attribute(ConfigRegAttr("Chaining enable"))
self.chain_en.add_attribute(FormalAttr(self.chain_en.name, FormalSignalConstraint.SET0))
else:
self.chain_en = self.var("chain_en", 1)
self.wire(self.chain_en, 0)
# gate clock with tile_en
gclk = self.var("gclk", 1)
self.gclk = kts.util.clock(gclk)
self.wire(gclk, self.clk_mem & self.tile_en)
self.clk_en = self.clock_en("clk_en", 1)
# active low asynchornous reset
self.rst_n = self.reset("rst_n", 1)
self.rst_n.add_attribute(FormalAttr(self.rst_n.name, FormalSignalConstraint.RSTN))
# data in and out of top level Lake memory object
self.data_in = self.input("data_in",
width=self.word_width,
size=self.input_ports,
explicit_array=True,
packed=True)
self.data_in.add_attribute(FormalAttr(self.data_in.name, FormalSignalConstraint.SEQUENCE))
self.data_out = self.output("data_out",
width=self.word_width,
size=self.output_ports,
explicit_array=True,
packed=True)
self.data_out.add_attribute(FormalAttr(self.data_out.name, FormalSignalConstraint.SEQUENCE))
# global cycle count for accessor comparison
self._cycle_count = self.var("cycle_count", 16)
@always_ff((posedge, self.gclk), (negedge, "rst_n"))
def increment_cycle_count(self):
if ~self.rst_n:
self._cycle_count = 0
else:
self._cycle_count = self._cycle_count + 1
self.add_always(increment_cycle_count)
# info about memories
num_mem = len(memories)
subscript_mems = list(self.memories.keys())
# list of the data out from each memory
self.mem_data_outs = [self.var(f"mem_data_out_{subscript_mems[i]}",
width=self.word_width,
size=self.memories[subscript_mems[i]]
["read_port_width" if "read_port_width" in self.memories[subscript_mems[i]]
else "read_write_port_width"],
explicit_array=True, packed=True) for i in range(num_mem)]
# keep track of write, read_addr, and write_addr vars for read/write memories
# to later check whether there is a write and what to use for the shared port
self.mem_read_write_addrs = {}
# create memory instance for each memory
self.mem_insts = {}
i = 0
for mem in self.memories.keys():
m = mem_inst(self.memories[mem], self.word_width)
self.mem_insts[mem] = m
self.add_child(mem,
m,
clk=self.gclk,
rst_n=self.rst_n,
# put data out in memory data out list
data_out=self.mem_data_outs[i],
chain_en=self.chain_en)
i += 1
# get input and output memories
is_input, is_output = [], []
for mem_name in self.memories.keys():
mem = self.memories[mem_name]
if mem["is_input"]:
is_input.append(mem_name)
if mem["is_output"]:
is_output.append(mem_name)
# TODO direct connection to write doesn't work (?), so have to do this...
self.low = self.var("low", 1)
self.wire(self.low, 0)
# TODO adding multiple ports to 1 memory after talking about mux with compiler team
# set up input memories
for i in range(len(is_input)):
in_mem = is_input[i]
# input addressor / accessor parameters
input_dim = self.memories[in_mem]["input_edge_params"]["dim"]
input_range = self.memories[in_mem]["input_edge_params"]["max_range"]
input_stride = self.memories[in_mem]["input_edge_params"]["max_stride"]
# input port associated with memory
input_port_index = self.memories[in_mem]["input_port"]
self.valid = self.var(
f"input_port{input_port_index}_2{in_mem}_accessor_valid", 1)
self.wire(self.mem_insts[in_mem].ports.write, self.valid)
# hook up data from the specified input port to the memory
safe_wire(self, self.mem_insts[in_mem].ports.data_in[0],
self.data_in[input_port_index])
if self.memories[in_mem]["num_read_write_ports"] > 0:
self.mem_read_write_addrs[in_mem] = {"write": self.valid}
# create IteratorDomain, AddressGenerator, and ScheduleGenerator
# for writes to this input memory
forloop = ForLoop(iterator_support=input_dim,
config_width=max(1, clog2(input_range))) # self.default_config_width)
loop_itr = forloop.get_iter()
loop_wth = forloop.get_cfg_width()
self.add_child(f"input_port{input_port_index}_2{in_mem}_forloop",
forloop,
clk=self.gclk,
rst_n=self.rst_n,
step=self.valid)
newAG = AddrGen(iterator_support=input_dim,
config_width=max(1, clog2(input_stride))) # self.default_config_width)
self.add_child(f"input_port{input_port_index}_2{in_mem}_write_addr_gen",
newAG,
clk=self.gclk,
rst_n=self.rst_n,
step=self.valid,
mux_sel=forloop.ports.mux_sel_out,
restart=forloop.ports.restart)
if self.memories[in_mem]["num_read_write_ports"] == 0:
safe_wire(self, self.mem_insts[in_mem].ports.write_addr[0], newAG.ports.addr_out)
else:
self.mem_read_write_addrs[in_mem]["write_addr"] = newAG.ports.addr_out
newSG = SchedGen(iterator_support=input_dim,
config_width=self.cycle_count_width)
self.add_child(f"input_port{input_port_index}_2{in_mem}_write_sched_gen",
newSG,
clk=self.gclk,
rst_n=self.rst_n,
mux_sel=forloop.ports.mux_sel_out,
finished=forloop.ports.restart,
cycle_count=self._cycle_count,
valid_output=self.valid)
# set up output memories
for i in range(len(is_output)):
out_mem = is_output[i]
# output addressor / accessor parameters
output_dim = self.memories[out_mem]["output_edge_params"]["dim"]
output_range = self.memories[out_mem]["output_edge_params"]["max_range"]
output_stride = self.memories[out_mem]["output_edge_params"]["max_stride"]
# output port associated with memory
output_port_index = self.memories[out_mem]["output_port"]
# hook up data from the memory to the specified output port
self.wire(self.data_out[output_port_index],
self.mem_insts[out_mem].ports.data_out[0][0])
# self.mem_data_outs[subscript_mems.index(out_mem)][0])
self.valid = self.var(f"{out_mem}2output_port{output_port_index}_accessor_valid", 1)
if self.memories[out_mem]["rw_same_cycle"]:
self.wire(self.mem_insts[out_mem].ports.read, self.valid)
# create IteratorDomain, AddressGenerator, and ScheduleGenerator
# for reads from this output memory
forloop = ForLoop(iterator_support=output_dim,
config_width=max(1, clog2(output_range))) # self.default_config_width)
loop_itr = forloop.get_iter()
loop_wth = forloop.get_cfg_width()
self.add_child(f"{out_mem}2output_port{output_port_index}_forloop",
forloop,
clk=self.gclk,
rst_n=self.rst_n,
step=self.valid)
newAG = AddrGen(iterator_support=output_dim,
config_width=max(1, clog2(output_stride))) # self.default_config_width)
self.add_child(f"{out_mem}2output_port{output_port_index}_read_addr_gen",
newAG,
clk=self.gclk,
rst_n=self.rst_n,
step=self.valid,
mux_sel=forloop.ports.mux_sel_out,
restart=forloop.ports.restart)
if self.memories[out_mem]["num_read_write_ports"] == 0:
safe_wire(self, self.mem_insts[out_mem].ports.read_addr[0], newAG.ports.addr_out)
else:
self.mem_read_write_addrs[in_mem]["read_addr"] = newAG.ports.addr_out
newSG = SchedGen(iterator_support=output_dim,
config_width=self.cycle_count_width) # self.default_config_width)
self.add_child(f"{out_mem}2output_port{output_port_index}_read_sched_gen",
newSG,
clk=self.gclk,
rst_n=self.rst_n,
mux_sel=forloop.ports.mux_sel_out,
finished=forloop.ports.restart,
cycle_count=self._cycle_count,
valid_output=self.valid)
# create shared IteratorDomains and accessors as well as
# read/write addressors for memories connected by each edge
for edge in self.edges:
# see how many signals need to be selected between for
# from and to signals for edge
num_mux_from = len(edge["from_signal"])
num_mux_to = len(edge["to_signal"])
# get unique edge_name identifier for hardware modules
edge_name = get_edge_name(edge)
# create forloop and accessor valid output signal
self.valid = self.var(edge_name + "_accessor_valid", 1)
forloop = ForLoop(iterator_support=edge["dim"])
self.forloop = forloop
loop_itr = forloop.get_iter()
loop_wth = forloop.get_cfg_width()
self.add_child(edge_name + "_forloop",
forloop,
clk=self.gclk,
rst_n=self.rst_n,
step=self.valid)
# create input addressor
readAG = AddrGen(iterator_support=edge["dim"],
config_width=self.default_config_width)
self.add_child(f"{edge_name}_read_addr_gen",
readAG,
clk=self.gclk,
rst_n=self.rst_n,
step=self.valid,
mux_sel=forloop.ports.mux_sel_out,
restart=forloop.ports.restart)
# assign read address to all from memories
if self.memories[edge["from_signal"][0]]["num_read_write_ports"] == 0:
# can assign same read addrs to all the memories
for i in range(len(edge["from_signal"])):
safe_wire(self, self.mem_insts[edge["from_signal"][i]].ports.read_addr[0], readAG.ports.addr_out)
else:
for i in range(len(edge["from_signal"])):
self.mem_read_write_addrs[edge["from_signal"][i]]["read_addr"] = readAG.ports.addr_out
# if needing to mux, choose which from memory we get data
# from for to memory data in
if num_mux_from > 1:
num_mux_bits = clog2(num_mux_from)
self.mux_sel = self.var(f"{edge_name}_mux_sel",
width=num_mux_bits)
read_addr_width = max(1, clog2(self.memories[edge["from_signal"][0]]["capacity"]))
# decide which memory to get data from for to memory's data in
safe_wire(self, self.mux_sel,
readAG.ports.addr_out[read_addr_width + num_mux_from - 1, read_addr_width])
comb_mux_from = self.combinational()
# for i in range(num_mux_from):
# TODO want to use a switch statement here, but get add_fn_ln issue
if_mux_sel = IfStmt(self.mux_sel == 0)
for j in range(len(edge["to_signal"])):
# print("TO ", edge["to_signal"][j])
# print("FROM ", edge["from_signal"][i])
if_mux_sel.then_(self.mem_insts[edge["to_signal"][j]].ports.data_in.assign(self.mem_insts[edge["from_signal"][0]].ports.data_out))
if_mux_sel.else_(self.mem_insts[edge["to_signal"][j]].ports.data_in.assign(self.mem_insts[edge["from_signal"][1]].ports.data_out))
comb_mux_from.add_stmt(if_mux_sel)
# no muxing from, data_out from the one and only memory
# goes to all to memories (valid determines whether it is
# actually written)
else:
for j in range(len(edge["to_signal"])):
# print("TO ", edge["to_signal"][j])
# print("FROM ", edge["from_signal"][0])
safe_wire(self,
self.mem_insts[edge["to_signal"][j]].ports.data_in,
# only one memory to read from
self.mem_insts[edge["from_signal"][0]].ports.data_out)
# create output addressor
writeAG = AddrGen(iterator_support=edge["dim"],
config_width=self.default_config_width)
# step, mux_sel, restart may need delayed signals (assigned later)
self.add_child(f"{edge_name}_write_addr_gen",
writeAG,
clk=self.gclk,
rst_n=self.rst_n)
# set write addr for to memories
if self.memories[edge["to_signal"][0]]["num_read_write_ports"] == 0:
for i in range(len(edge["to_signal"])):
safe_wire(self, self.mem_insts[edge["to_signal"][i]].ports.write_addr[0], writeAG.ports.addr_out)
else:
for i in range(len(edge["to_signal"])):
self.mem_read_write_addrs[edge["to_signal"][i]] = {"write": self.valid, "write_addr": writeAG.ports.addr_out}
# calculate necessary delay between from_signal to to_signal
# TODO this may need to be more sophisticated and based on II as well
# TODO just need to add for loops for all the ports
if self.memories[edge["from_signal"][0]]["num_read_write_ports"] == 0:
self.delay = self.memories[edge["from_signal"][0]]["read_info"][0]["latency"]
else:
self.delay = self.memories[edge["from_signal"][0]]["read_write_info"][0]["latency"]
if self.delay > 0:
# signals that need to be delayed due to edge latency
self.delayed_writes = self.var(f"{edge_name}_delayed_writes",
width=self.delay)
self.delayed_mux_sels = self.var(f"{edge_name}_delayed_mux_sels",
width=self.forloop.ports.mux_sel_out.width,
size=self.delay,
explicit_array=True,
packed=True)
self.delayed_restarts = self.var(f"{edge_name}_delayed_restarts",
width=self.delay)
# delay in valid between read from memory and write to next memory
@always_ff((posedge, self.gclk), (negedge, "rst_n"))
def get_delayed_write(self):
if ~self.rst_n:
self.delayed_writes = 0
self.delayed_mux_sels = 0
self.delayed_restarts = 0
else:
for i in range(self.delay - 1):
self.delayed_writes[i + 1] = self.delayed_writes[i]
self.delayed_mux_sels[i + 1] = self.delayed_mux_sels[i]
self.delayed_restarts[i + 1] = self.delayed_restarts[i]
self.delayed_writes[0] = self.valid
self.delayed_mux_sels[0] = self.forloop.ports.mux_sel_out
self.delayed_restarts[0] = self.forloop.ports.restart
self.add_always(get_delayed_write)
# if we have a mux for the destination memories,
# choose which mux to write to
if num_mux_to > 1:
num_mux_bits = clog2(num_mux_to)
self.mux_sel_to = self.var(f"{edge_name}_mux_sel_to",
width=num_mux_bits)
write_addr_width = max(1, clog2(self.memories[edge["to_signal"][0]]["capacity"]))
# decide which destination memory gets written to
safe_wire(self, self.mux_sel_to,
writeAG.ports.addr_out[write_addr_width + num_mux_to - 1, write_addr_width])
# wire the write (or if needed, delayed write) signal to the selected destination memory
# and set write enable low for all other destination memories
comb_mux_to = self.combinational()
for i in range(num_mux_to):
if_mux_sel_to = IfStmt(self.mux_sel_to == i)
if self.delay == 0:
if_mux_sel_to.then_(self.mem_insts[edge["to_signal"][i]].ports.write.assign(self.valid))
else:
if_mux_sel_to.then_(self.mem_insts[edge["to_signal"][i]].ports.write.assign(self.delayed_writes[self.delay - 1]))
if_mux_sel_to.else_(self.mem_insts[edge["to_signal"][i]].ports.write.assign(self.low))
comb_mux_to.add_stmt(if_mux_sel_to)
# no muxing to, just write to the one destination memory
else:
if self.delay == 0:
self.wire(self.mem_insts[edge["to_signal"][0]].ports.write, self.valid)
else:
self.wire(self.mem_insts[edge["to_signal"][0]].ports.write, self.delayed_writes[self.delay - 1])
# assign delayed signals for write addressor if needed
if self.delay == 0:
self.wire(writeAG.ports.step, self.valid)
self.wire(writeAG.ports.mux_sel, self.forloop.ports.mux_sel_out)
self.wire(writeAG.ports.restart, self.forloop.ports.restart)
else:
self.wire(writeAG.ports.step, self.delayed_writes[self.delay - 1])
self.wire(writeAG.ports.mux_sel, self.delayed_mux_sels[self.delay - 1])
self.wire(writeAG.ports.restart, self.delayed_restarts[self.delay - 1])
# create accessor for edge
newSG = SchedGen(iterator_support=edge["dim"],
config_width=self.cycle_count_width) # self.default_config_width)
self.add_child(edge_name + "_sched_gen",
newSG,
clk=self.gclk,
rst_n=self.rst_n,
mux_sel=forloop.ports.mux_sel_out,
finished=forloop.ports.restart,
cycle_count=self._cycle_count,
valid_output=self.valid)
# for read write memories, choose either read or write address based on whether
# we are writing to the memory (whether write enable is high)
read_write_addr_comb = self.combinational()
for mem_name in self.memories:
if mem_name in self.mem_read_write_addrs:
mem_info = self.mem_read_write_addrs[mem_name]
if_write = IfStmt(mem_info["write"] == 1)
addr_width = self.mem_insts[mem_name].ports.read_write_addr[0].width
if_write.then_(self.mem_insts[mem_name].ports.read_write_addr[0].assign(mem_info["write_addr"][addr_width - 1, 0]))
if_write.else_(self.mem_insts[mem_name].ports.read_write_addr[0].assign(mem_info["read_addr"][addr_width - 1, 0]))
read_write_addr_comb.add_stmt(if_write)
# clock enable and flush passes
kts.passes.auto_insert_clock_enable(self.internal_generator)
clk_en_port = self.internal_generator.get_port("clk_en")
clk_en_port.add_attribute(FormalAttr(clk_en_port.name, FormalSignalConstraint.SET1))
self.add_attribute("sync-reset=flush")
kts.passes.auto_insert_sync_reset(self.internal_generator)
flush_port = self.internal_generator.get_port("flush")
# bring config registers up to top level
lift_config_reg(self.internal_generator)
def __init__(self, fetch_width, data_width, int_out_ports):
assert not (fetch_width &
(fetch_width - 1)), "Memory width needs to be a power of 2"
super().__init__("sync_groups")
# Absorb inputs
self.fetch_width = fetch_width
self.data_width = data_width
self.fw_int = int(self.fetch_width / self.data_width)
self.int_out_ports = int_out_ports
self.groups = self.int_out_ports
# Clock and Reset
self._clk = self.clock("clk")
self._rst_n = self.reset("rst_n")
# Inputs
self._ack_in = self.input("ack_in", self.int_out_ports)
self._data_in = self.input("data_in",
self.data_width,
size=(self.int_out_ports, self.fw_int),
explicit_array=True,
packed=True)
self._mem_valid_data = self.input("mem_valid_data", self.int_out_ports)
self._mem_valid_data_out = self.output("mem_valid_data_out",
self.int_out_ports)
self._valid_in = self.input("valid_in", self.int_out_ports)
# Indicates which port belongs to which synchronization group
self._sync_group = self.input("sync_group",
self.int_out_ports,
size=self.int_out_ports,
explicit_array=True,
packed=True)
sync_config = ConfigRegAttr(
"This array of one hot vectors" +
" is used to denote which ports are synchronized to eachother." +
" If multiple ports should output data relative to eachother" +
" one should put them in the same group.")
self._sync_group.add_attribute(sync_config)
# Outputs
self._data_out = self.output("data_out",
self.data_width,
size=(self.int_out_ports, self.fw_int),
explicit_array=True,
packed=True)
self._valid_out = self.output("valid_out", self.int_out_ports)
# Locals
self._sync_agg = self.var("sync_agg",
self.int_out_ports,
size=self.int_out_ports,
explicit_array=True,
packed=True)
self._sync_valid = self.var("sync_valid", self.int_out_ports)
self._data_reg = self.var("data_reg",
self.data_width,
size=(self.int_out_ports, self.fw_int),
explicit_array=True,
packed=True)
self._valid_reg = self.var("valid_reg", self.int_out_ports)
# This signal allows us to orchestrate the synchronization groups
# at the output address controller
self._ren_in = self.input("ren_in", self.int_out_ports)
self._ren_int = self.var("ren_int", self.int_out_ports)
self._rd_sync_gate = self.output("rd_sync_gate", self.int_out_ports)
self._local_gate_bus = self.var("local_gate_bus",
self.int_out_ports,
size=self.int_out_ports,
explicit_array=True,
packed=True)
self._local_gate_bus_n = self.var("local_gate_bus_n",
self.int_out_ports,
size=self.int_out_ports,
explicit_array=True,
packed=True)
self.wire(self._local_gate_bus, ~self._local_gate_bus_n)
self._local_gate_bus_tpose = self.var("local_gate_bus_tpose",
self.int_out_ports,
size=self.int_out_ports,
explicit_array=True,
packed=True)
self._local_gate_reduced = self.var("local_gate_reduced",
self.int_out_ports)
self._local_gate_mask = self.var("local_gate_mask",
self.int_out_ports,
size=self.int_out_ports,
explicit_array=True,
packed=True)
self._group_finished = self.var("group_finished", self.int_out_ports)
self._grp_fin_large = self.var("grp_fin_large",
self.int_out_ports,
size=self.int_out_ports,
explicit_array=True,
packed=True)
self._done = self.var("done", self.int_out_ports)
self._done_alt = self.var("done_alt", self.int_out_ports)
# Output data is ungated
self.wire(self._data_out, self._data_reg)
self.wire(self._rd_sync_gate, self._local_gate_reduced)
# Valid requires gating based on sync_valid
self.wire(self._ren_int, self._ren_in & self._local_gate_reduced)
# Add Code
self.add_code(self.set_sync_agg, unroll_for=True)
self.add_code(self.set_sync_valid)
for i in range(self.int_out_ports):
self.add_code(self.set_sync_stage, idx=i)
self.add_code(self.set_out_valid)
self.add_code(self.set_reduce_gate)
for i in range(self.groups):
self.add_code(self.set_rd_gates, idx=i)
self.add_code(self.set_tpose)
self.add_code(self.set_finished)
self.add_code(self.next_gate_mask, unroll_for=True)
self.add_code(self.set_grp_fin, unroll_for=True)
def __init__(self, agg_height, data_width, mem_width, max_agg_schedule):
super().__init__("aggregation_buffer")
self.agg_height = agg_height
self.data_width = data_width
self.mem_width = mem_width
self.max_agg_schedule = max_agg_schedule # This is the maximum length of the schedule
self.fw_int = int(self.mem_width / self.data_width)
# Clock and Reset
self._clk = self.clock("clk")
self._rst_n = self.reset("rst_n")
# Inputs
# Bring in a single element into an AggregationBuffer w/ valid signaling
self._data_in = self.input("data_in", self.data_width)
self._valid_in = self.input("valid_in", 1)
self._align = self.input("align", 1)
# Outputs
self._data_out = self.output("data_out", self.mem_width)
self._valid_out = self.output("valid_out", 1)
self._data_out_chop = []
for i in range(self.fw_int):
self._data_out_chop.append(
self.output(f"data_out_chop_{i}", self.data_width))
self.add_stmt(self._data_out_chop[i].assign(
self._data_out[(self.data_width * (i + 1)) - 1,
self.data_width * i]))
# CONFIG:
# We receive a periodic (doesn't need to be, but has a maximum schedule,
# so...possibly the schedule is a for loop?
# Tells us where to write successive elements...
self._in_schedule = self.input("in_sched",
max(1, clog2(self.agg_height)),
size=self.max_agg_schedule,
explicit_array=True,
packed=True)
doc = "Input schedule for aggregation buffer. Enumerate which" + \
f" of {self.agg_height} buffers to write to."
self._in_schedule.add_attribute(ConfigRegAttr(doc))
self._in_period = self.input("in_period", clog2(self.max_agg_schedule))
doc = "Input period for aggregation buffer. 1 is a reasonable" + \
" setting for most applications"
self._in_period.add_attribute(ConfigRegAttr(doc))
# ...and which order to output the blocks
self._out_schedule = self.input("out_sched",
max(1, clog2(agg_height)),
size=self.max_agg_schedule,
explicit_array=True,
packed=True)
doc = "Output schedule for aggregation buffer. Enumerate which" + \
f" of {self.agg_height} buffers to write to SRAM from."
self._out_schedule.add_attribute(ConfigRegAttr(doc))
self._out_period = self.input("out_period",
clog2(self.max_agg_schedule))
self._out_period.add_attribute(
ConfigRegAttr("Output period for aggregation buffer"))
self._in_sched_ptr = self.var("in_sched_ptr",
clog2(self.max_agg_schedule))
self._out_sched_ptr = self.var("out_sched_ptr",
clog2(self.max_agg_schedule))
# Local Signals
self._aggs_out = self.var("aggs_out",
self.mem_width,
size=self.agg_height,
packed=True,
explicit_array=True)
self._aggs_sep = []
for i in range(self.agg_height):
self._aggs_sep.append(
self.var(f"aggs_sep_{i}",
self.data_width,
size=self.fw_int,
packed=True))
self._valid_demux = self.var("valid_demux", self.agg_height)
self._align_demux = self.var("align_demux", self.agg_height)
self._next_full = self.var("next_full", self.agg_height)
self._valid_out_mux = self.var("valid_out_mux", self.agg_height)
for i in range(self.agg_height):
# Add in the children aggregators...
self.add_child(f"agg_{i}",
Aggregator(self.data_width,
mem_word_width=self.fw_int),
clk=self._clk,
rst_n=self._rst_n,
in_pixels=self._data_in,
valid_in=self._valid_demux[i],
agg_out=self._aggs_sep[i],
valid_out=self._valid_out_mux[i],
next_full=self._next_full[i],
align=self._align_demux[i])
portlist = []
if self.fw_int == 1:
self.wire(self._aggs_out[i], self._aggs_sep[i])
else:
for j in range(self.fw_int):
portlist.append(self._aggs_sep[i][self.fw_int - 1 - j])
self.wire(self._aggs_out[i], kts.concat(*portlist))
# Sequential code blocks
self.add_code(self.update_in_sched_ptr)
self.add_code(self.update_out_sched_ptr)
# Combinational code blocks
self.add_code(self.valid_demux_comb)
self.add_code(self.align_demux_comb)
self.add_code(self.valid_out_comb)
self.add_code(self.output_data_comb)
def __init__(self, iterator_support=6, config_width=16, use_enable=True):
super().__init__(f"sched_gen_{iterator_support}_{config_width}")
self.iterator_support = iterator_support
self.config_width = config_width
self.use_enable = use_enable
# PORT DEFS: begin
# INPUTS
self._clk = self.clock("clk")
self._rst_n = self.reset("rst_n")
# OUTPUTS
self._valid_output = self.output("valid_output", 1)
# VARS
self._valid_out = self.var("valid_out", 1)
self._cycle_count = self.input("cycle_count", self.config_width)
self._mux_sel = self.input("mux_sel",
max(clog2(self.iterator_support), 1))
self._addr_out = self.var("addr_out", self.config_width)
# Receive signal on last iteration of looping structure and
# gate the output...
self._finished = self.input("finished", 1)
self._valid_gate_inv = self.var("valid_gate_inv", 1)
self._valid_gate = self.var("valid_gate", 1)
self.wire(self._valid_gate, ~self._valid_gate_inv)
# Since dim = 0 is not sufficient, we need a way to prevent
# the controllers from firing on the starting offset
if self.use_enable:
self._enable = self.input("enable", 1)
self._enable.add_attribute(
ConfigRegAttr("Disable the controller so it never fires..."))
self._enable.add_attribute(
FormalAttr(f"{self._enable.name}",
FormalSignalConstraint.SOLVE))
# Otherwise we set it as a 1 and leave it up to synthesis...
else:
self._enable = self.var("enable", 1)
self.wire(self._enable, kratos.const(1, 1))
@always_ff((posedge, "clk"), (negedge, "rst_n"))
def valid_gate_inv_ff():
if ~self._rst_n:
self._valid_gate_inv = 0
# If we are finishing the looping structure, turn this off to implement one-shot
elif self._finished:
self._valid_gate_inv = 1
self.add_code(valid_gate_inv_ff)
# Compare based on minimum of addr + global cycle...
self.c_a_cmp = min(self._cycle_count.width, self._addr_out.width)
# PORT DEFS: end
self.add_child(f"sched_addr_gen",
AddrGen(iterator_support=self.iterator_support,
config_width=self.config_width),
clk=self._clk,
rst_n=self._rst_n,
step=self._valid_out,
mux_sel=self._mux_sel,
addr_out=self._addr_out,
restart=const(0, 1))
self.add_code(self.set_valid_out)
self.add_code(self.set_valid_output)
def __init__(self, iterator_support=6, config_width=16):
super().__init__(f"for_loop_{iterator_support}_{config_width}",
debug=True)
self.iterator_support = iterator_support
self.config_width = config_width
# Create params for instancing this module...
self.iterator_support_par = self.param("ITERATOR_SUPPORT",
clog2(iterator_support) + 1,
value=self.iterator_support)
self.config_width_par = self.param("CONFIG_WIDTH",
clog2(config_width) + 1,
value=self.config_width)
# PORT DEFS: begin
# INPUTS
self._clk = self.clock("clk")
self._rst_n = self.reset("rst_n")
self._ranges = self.input("ranges",
self.config_width,
size=self.iterator_support,
packed=True,
explicit_array=True)
self._ranges.add_attribute(
ConfigRegAttr("Ranges of address generator"))
self._ranges.add_attribute(
FormalAttr(f"{self._ranges.name}", FormalSignalConstraint.SOLVE))
self._dimensionality = self.input("dimensionality",
1 + clog2(self.iterator_support))
self._dimensionality.add_attribute(
ConfigRegAttr("Dimensionality of address generator"))
self._dimensionality.add_attribute(
FormalAttr(f"{self._dimensionality.name}",
FormalSignalConstraint.SOLVE))
self._step = self.input("step", 1)
# OUTPUTS
# PORT DEFS: end
# LOCAL VARIABLES: begin
self._dim_counter = self.var("dim_counter",
self.config_width,
size=self.iterator_support,
packed=True,
explicit_array=True)
self._strt_addr = self.var("strt_addr", self.config_width)
self._counter_update = self.var("counter_update", 1)
self._calc_addr = self.var("calc_addr", self.config_width)
self._max_value = self.var("max_value", self.iterator_support)
self._mux_sel = self.var("mux_sel", max(clog2(self.iterator_support),
1))
self._mux_sel_out = self.output("mux_sel_out",
max(clog2(self.iterator_support), 1))
self.wire(self._mux_sel_out, self._mux_sel)
# LOCAL VARIABLES: end
# GENERATION LOGIC: begin
self._done = self.var("done", 1)
self._clear = self.var("clear", self.iterator_support)
self._inc = self.var("inc", self.iterator_support)
self._inced_cnt = self.var("inced_cnt", self._dim_counter.width)
self.wire(self._inced_cnt, self._dim_counter[self._mux_sel] + 1)
# Next_max_value
self._maxed_value = self.var("maxed_value", 1)
self.wire(
self._maxed_value,
(self._dim_counter[self._mux_sel] == self._ranges[self._mux_sel])
& self._inc[self._mux_sel])
self.add_code(self.set_mux_sel)
for i in range(self.iterator_support):
self.add_code(self.set_clear, idx=i)
self.add_code(self.set_inc, idx=i)
self.add_code(self.dim_counter_update, idx=i)
self.add_code(self.max_value_update, idx=i)
# GENERATION LOGIC: end
self._restart = self.output("restart", 1)
self.wire(self._restart, self._step & (~self._done))
def __init__(
self,
# number of bits in a word
word_width,
# number of words that can be sotred at an address in SRAM
# note fetch_width must be powers of 2
fetch_width,
# total number of transpose buffers
num_tb,
# height of this particular transpose buffer
max_tb_height,
# maximum value for range parameters in nested for loop
# (and as a result, maximum length of indices input vector)
# specifying inner for loop values for output column
# addressing
max_range,
max_range_inner,
max_stride,
tb_iterator_support):
super().__init__("transpose_buffer", debug=True)
#########################
# GENERATION PARAMETERS #
#########################
self.word_width = word_width
self.fetch_width = fetch_width
self.num_tb = num_tb
self.max_tb_height = max_tb_height
self.max_range = max_range
self.max_range_inner = max_range_inner
self.max_stride = max_stride
self.tb_iterator_support = tb_iterator_support
##################################
# BITS FOR GENERATION PARAMETERS #
##################################
self.fetch_width_bits = max(1, clog2(self.fetch_width))
self.num_tb_bits = max(1, clog2(self.num_tb))
self.max_range_bits = max(1, clog2(self.max_range))
self.max_range_inner_bits = max(1, clog2(self.max_range_inner))
self.max_stride_bits = max(1, clog2(self.max_stride))
self.tb_col_index_bits = 2 * max(self.fetch_width_bits,
self.num_tb_bits) + 1
self.max_tb_height_bits2 = max(1, clog2(2 * self.max_tb_height))
self.max_tb_height_bits = max(1, clog2(self.max_tb_height))
self.tb_iterator_support_bits = max(
1,
clog2(self.tb_iterator_support) + 1)
self.max_range_stride_bits2 = max(2 * self.max_range_bits,
2 * self.max_stride_bits)
##########
# INPUTS #
##########
self.clk = self.clock("clk")
# active low asynchronous reset
self.rst_n = self.reset("rst_n", 1)
# data input from SRAM
if self.fetch_width == 1:
self.input_data = self.input("input_data", self.word_width)
else:
self.input_data = self.input("input_data",
width=self.word_width,
size=self.fetch_width,
packed=True)
# valid indicating whether data input from SRAM is valid and
# should be stored in transpose buffer
self.valid_data = self.input("valid_data", 1)
self.ack_in = self.input("ack_in", 1)
self.ren = self.input("ren", 1)
self.mem_valid_data = self.input("mem_valid_data", 1)
###########################
# CONFIGURATION REGISTERS #
###########################
# the range of the outer for loop in nested for loop for output
# column address generation
self.range_outer = self.input("range_outer", self.max_range_bits)
self.range_outer.add_attribute(
ConfigRegAttr("Outer range for output for loop pattern"))
# the range of the inner for loop in nested for loop for output
# column address generation
self.range_inner = self.input("range_inner", self.max_range_inner_bits)
self.range_inner.add_attribute(
ConfigRegAttr("Inner range for output for for loop pattern"))
# stride for the given application
self.stride = self.input("stride", self.max_stride_bits)
self.stride.add_attribute(ConfigRegAttr("Application stride"))
self.tb_height = self.input("tb_height", self.max_tb_height_bits)
self.tb_height.add_attribute(ConfigRegAttr("Transpose Buffer height"))
self.dimensionality = self.input("dimensionality",
self.tb_iterator_support_bits)
self.dimensionality.add_attribute(
ConfigRegAttr("Transpose Buffer dimensionality"))
# specifies inner for loop values for output column
# addressing
self.indices = self.input(
"indices",
width=clog2(2 * self.num_tb * self.fetch_width),
# the length of indices is equal to range_inner,
# so the maximum possible size for self.indices
# is the maximum value of range_inner, which is
# self.max_range_inner
size=self.max_range_inner,
explicit_array=True,
packed=True)
self.indices.add_attribute(
ConfigRegAttr("Output indices for for loop pattern"))
# offset to start output address if we're starting in the middle of a wider
# fetch width word for example
self.starting_addr = self.input("starting_addr", self.fetch_width_bits)
self.starting_addr.add_attribute(ConfigRegAttr("TB starting address"))
###########
# OUTPUTS #
###########
self.col_pixels = self.output("col_pixels",
width=self.word_width,
size=self.max_tb_height,
packed=True,
explicit_array=True)
self.output_valid = self.output("output_valid", 1)
self.rdy_to_arbiter = self.output("rdy_to_arbiter", 1)
###################
# LOCAL VARIABLES #
###################
# transpose buffer
if self.fetch_width == 1:
self.tb = self.var("tb",
width=self.word_width,
size=2 * self.max_tb_height,
packed=True)
else:
self.tb = self.var("tb",
width=self.word_width,
size=[2 * self.max_tb_height, self.fetch_width],
packed=True)
self.tb_valid = self.var("tb_valid", 2 * self.max_tb_height)
self.index_outer = self.var("index_outer", self.max_range_bits)
self.index_inner = self.var("index_inner", self.max_range_inner_bits)
self.input_buf_index = self.var("input_buf_index", 1)
self.out_buf_index = self.var("out_buf_index", 1)
self.switch_out_buf = self.var("switch_out_buf", 1)
self.switch_next_line = self.var("switch_next_line", 1)
self.row_index = self.var("row_index", self.max_tb_height_bits)
self.input_index = self.var("input_index", self.max_tb_height_bits2)
self.output_index_abs = self.var("output_index_abs",
self.max_range_stride_bits2)
if self.fetch_width != 1:
self.output_index_long = self.var("output_index_long",
self.max_range_stride_bits2)
self.output_index = self.var("output_index", self.fetch_width_bits)
self.indices_index_inner = self.var(
"indices_index_inner", clog2(2 * self.num_tb * self.fetch_width))
self.curr_out_start = self.var("curr_out_start",
self.max_range_stride_bits2)
self.start_data = self.var("start_data", 1)
self.old_start_data = self.var("old_start_data", 1)
self.pause_tb = self.var("pause_tb", 1)
self.pause_output = self.var("pause_output", 1)
self.on_next_line = self.var("on_next_line", 1)
self.mask_valid = self.var("mask_valid", 1)
self.pause_tbinv = self.var("pause_tbinv", 1)
self.rdy_to_arbiterinv = self.var("rdy_to_arbiterinv", 1)
self.out_buf_indexinv = self.var("out_buf_indexinv", 1)
##########################
# SEQUENTIAL CODE BLOCKS #
##########################
self.add_code(self.set_index_outer)
self.add_code(self.set_index_inner)
self.add_code(self.set_pause_tb)
self.add_code(self.set_row_index)
self.add_code(self.set_input_buf_index)
self.add_code(self.input_to_tb)
self.add_code(self.output_from_tb)
self.add_code(self.set_output_valid)
self.add_code(self.set_out_buf_index)
self.add_code(self.set_rdy_to_arbiter)
self.add_code(self.set_start_data)
self.add_code(self.set_curr_out_start)
if self.fetch_width != 1:
self.add_code(self.set_output_index)
self.add_code(self.set_old_start_data)
self.add_code(self.set_on_next_line)
#############################
# COMBINATIONAL CODE BLOCKS #
#############################
self.add_code(self.set_pause_output)
self.add_code(self.set_input_index)
self.add_code(self.set_tb_out_indices)
self.add_code(self.set_switch_out_buf)
self.add_code(self.set_switch_next_line)
if self.fetch_width != 1:
self.add_code(self.set_output_index_long)
self.add_code(self.set_mask_valid)
self.add_code(self.set_invs)
def __init__(self,
data_width=16, # CGRA Params
mem_width=64,
mem_depth=512,
banks=1,
input_iterator_support=6, # Addr Controllers
output_iterator_support=6,
input_config_width=16,
output_config_width=16,
interconnect_input_ports=2, # Connection to int
interconnect_output_ports=2,
mem_input_ports=1,
mem_output_ports=1,
read_delay=1, # Cycle delay in read (SRAM vs Register File)
rw_same_cycle=False, # Does the memory allow r+w in same cycle?
agg_height=4,
max_agg_schedule=32,
input_max_port_sched=32,
output_max_port_sched=32,
align_input=1,
max_line_length=128,
max_tb_height=1,
tb_range_max=128,
tb_range_inner_max=5,
tb_sched_max=64,
max_tb_stride=15,
num_tb=1,
tb_iterator_support=2,
multiwrite=1,
num_tiles=1,
max_prefetch=8,
app_ctrl_depth_width=16,
remove_tb=False,
stcl_valid_iter=4):
super().__init__("strg_ub")
self.data_width = data_width
self.mem_width = mem_width
self.mem_depth = mem_depth
self.banks = banks
self.input_iterator_support = input_iterator_support
self.output_iterator_support = output_iterator_support
self.input_config_width = input_config_width
self.output_config_width = output_config_width
self.interconnect_input_ports = interconnect_input_ports
self.interconnect_output_ports = interconnect_output_ports
self.mem_input_ports = mem_input_ports
self.mem_output_ports = mem_output_ports
self.agg_height = agg_height
self.max_agg_schedule = max_agg_schedule
self.input_max_port_sched = input_max_port_sched
self.output_max_port_sched = output_max_port_sched
self.input_port_sched_width = clog2(self.interconnect_input_ports)
self.align_input = align_input
self.max_line_length = max_line_length
assert self.mem_width >= self.data_width, "Data width needs to be smaller than mem"
self.fw_int = int(self.mem_width / self.data_width)
self.num_tb = num_tb
self.max_tb_height = max_tb_height
self.tb_range_max = tb_range_max
self.tb_range_inner_max = tb_range_inner_max
self.max_tb_stride = max_tb_stride
self.tb_sched_max = tb_sched_max
self.tb_iterator_support = tb_iterator_support
self.multiwrite = multiwrite
self.max_prefetch = max_prefetch
self.num_tiles = num_tiles
self.app_ctrl_depth_width = app_ctrl_depth_width
self.remove_tb = remove_tb
self.read_delay = read_delay
self.rw_same_cycle = rw_same_cycle
self.stcl_valid_iter = stcl_valid_iter
# phases = [] TODO
self.address_width = clog2(self.num_tiles * self.mem_depth)
# CLK and RST
self._clk = self.clock("clk")
self._rst_n = self.reset("rst_n")
# INPUTS
self._data_in = self.input("data_in",
self.data_width,
size=self.interconnect_input_ports,
packed=True,
explicit_array=True)
self._wen_in = self.input("wen_in", self.interconnect_input_ports)
self._ren_input = self.input("ren_in", self.interconnect_output_ports)
# Post rate matched
self._ren_in = self.var("ren_in_muxed", self.interconnect_output_ports)
# Processed versions of wen and ren from the app ctrl
self._wen = self.var("wen", self.interconnect_input_ports)
self._ren = self.var("ren", self.interconnect_output_ports)
# Add rate matched
# If one input port, let any output port use the wen_in as the ren_in
# If more, do the same thing but also provide port selection
if self.interconnect_input_ports == 1:
self._rate_matched = self.input("rate_matched", self.interconnect_output_ports)
self._rate_matched.add_attribute(ConfigRegAttr("Rate matched - 1 or 0"))
for i in range(self.interconnect_output_ports):
self.wire(self._ren_in[i],
kts.ternary(self._rate_matched[i],
self._wen_in,
self._ren_input[i]))
else:
self._rate_matched = self.input("rate_matched", 1 + kts.clog2(self.interconnect_input_ports),
size=self.interconnect_output_ports,
explicit_array=True,
packed=True)
self._rate_matched.add_attribute(ConfigRegAttr("Rate matched [input port | on/off]"))
for i in range(self.interconnect_output_ports):
self.wire(self._ren_in[i],
kts.ternary(self._rate_matched[i][0],
self._wen_in[self._rate_matched[i][kts.clog2(self.interconnect_input_ports), 1]],
self._ren_input[i]))
self._arb_wen_en = self.var("arb_wen_en", self.interconnect_input_ports)
self._arb_ren_en = self.var("arb_ren_en", self.interconnect_output_ports)
self._data_from_strg = self.input("data_from_strg",
self.data_width,
size=(self.banks,
self.mem_output_ports,
self.fw_int),
packed=True,
explicit_array=True)
self._mem_valid_data = self.input("mem_valid_data",
self.mem_output_ports,
size=self.banks,
explicit_array=True,
packed=True)
self._out_mem_valid_data = self.var("out_mem_valid_data",
self.mem_output_ports,
size=self.banks,
explicit_array=True,
packed=True)
# We need to signal valids out of the agg buff, only if one exists...
if self.agg_height > 0:
self._to_iac_valid = self.var("ab_to_mem_valid",
self.interconnect_input_ports)
self._data_out = self.output("data_out",
self.data_width,
size=self.interconnect_output_ports,
packed=True,
explicit_array=True)
self._valid_out = self.output("valid_out",
self.interconnect_output_ports)
self._valid_out_alt = self.var("valid_out_alt",
self.interconnect_output_ports)
self._data_to_strg = self.output("data_to_strg",
self.data_width,
size=(self.banks,
self.mem_input_ports,
self.fw_int),
packed=True,
explicit_array=True)
# If we can perform a read and a write on the same cycle,
# this will necessitate a separate read and write address...
if self.rw_same_cycle:
self._wr_addr_out = self.output("wr_addr_out",
self.address_width,
size=(self.banks,
self.mem_input_ports),
explicit_array=True,
packed=True)
self._rd_addr_out = self.output("rd_addr_out",
self.address_width,
size=(self.banks,
self.mem_output_ports),
explicit_array=True,
packed=True)
else:
self._addr_out = self.output("addr_out",
self.address_width,
size=(self.banks,
self.mem_input_ports),
packed=True,
explicit_array=True)
self._cen_to_strg = self.output("cen_to_strg", self.mem_output_ports,
size=self.banks,
explicit_array=True,
packed=True)
self._wen_to_strg = self.output("wen_to_strg", self.mem_input_ports,
size=self.banks,
explicit_array=True,
packed=True)
if self.num_tb > 0:
self._tb_valid_out = self.var("tb_valid_out", self.interconnect_output_ports)
self._port_wens = self.var("port_wens", self.interconnect_input_ports)
####################
##### APP CTRL #####
####################
self._ack_transpose = self.var("ack_transpose",
self.banks,
size=self.interconnect_output_ports,
explicit_array=True,
packed=True)
self._ack_reduced = self.var("ack_reduced",
self.interconnect_output_ports)
self.app_ctrl = AppCtrl(interconnect_input_ports=self.interconnect_input_ports,
interconnect_output_ports=self.interconnect_output_ports,
depth_width=self.app_ctrl_depth_width,
sprt_stcl_valid=True,
stcl_iter_support=self.stcl_valid_iter)
# Some refactoring here for pond to get rid of app controllers...
# This is honestly pretty messy and should clean up nicely when we have the spec...
self._ren_out_reduced = self.var("ren_out_reduced",
self.interconnect_output_ports)
if self.num_tb == 0 or self.remove_tb:
self.wire(self._wen, self._wen_in)
self.wire(self._ren, self._ren_in)
self.wire(self._valid_out, self._valid_out_alt)
self.wire(self._arb_wen_en, self._wen)
self.wire(self._arb_ren_en, self._ren)
else:
self.add_child("app_ctrl", self.app_ctrl,
clk=self._clk,
rst_n=self._rst_n,
wen_in=self._wen_in,
ren_in=self._ren_in,
# ren_update=self._tb_valid_out,
valid_out_data=self._valid_out,
# valid_out_stencil=,
wen_out=self._wen,
ren_out=self._ren)
self.wire(self.app_ctrl.ports.tb_valid, self._tb_valid_out)
self.wire(self.app_ctrl.ports.ren_update, self._tb_valid_out)
self.app_ctrl_coarse = AppCtrl(interconnect_input_ports=self.interconnect_input_ports,
interconnect_output_ports=self.interconnect_output_ports,
depth_width=self.app_ctrl_depth_width)
self.add_child("app_ctrl_coarse", self.app_ctrl_coarse,
clk=self._clk,
rst_n=self._rst_n,
wen_in=self._to_iac_valid, # self._port_wens & self._to_iac_valid, # Gets valid and the ack
ren_in=self._ren_out_reduced,
tb_valid=kts.const(0, 1),
ren_update=self._ack_reduced,
wen_out=self._arb_wen_en,
ren_out=self._arb_ren_en)
###########################
##### INPUT AGG SCHED #####
###########################
###########################################
##### AGGREGATION ALIGNERS (OPTIONAL) #####
###########################################
# These variables are holders and can be swapped out if needed
self._data_consume = self._data_in
self._valid_consume = self._wen
# Zero out if not aligning
if(self.agg_height > 0):
self._align_to_agg = self.var("align_input",
self.interconnect_input_ports)
# Add the aggregation buffer aligners
if(self.align_input):
self._data_consume = self.var("data_consume",
self.data_width,
size=self.interconnect_input_ports,
packed=True,
explicit_array=True)
self._valid_consume = self.var("valid_consume",
self.interconnect_input_ports)
# Make new aggregation aligners for each port
for i in range(self.interconnect_input_ports):
new_child = AggAligner(self.data_width, self.max_line_length)
self.add_child(f"agg_align_{i}", new_child,
clk=self._clk,
rst_n=self._rst_n,
in_dat=self._data_in[i],
in_valid=self._wen[i],
align=self._align_to_agg[i],
out_valid=self._valid_consume[i],
out_dat=self._data_consume[i])
else:
if self.agg_height > 0:
self.wire(self._align_to_agg, const(0, self._align_to_agg.width))
################################################
##### END: AGGREGATION ALIGNERS (OPTIONAL) #####
################################################
if self.agg_height == 0:
self._to_iac_dat = self._data_consume
self._to_iac_valid = self._valid_consume
##################################
##### AGG BUFFERS (OPTIONAL) #####
##################################
# Only instantiate agg_buffer if needed
if(self.agg_height > 0):
self._to_iac_dat = self.var("ab_to_mem_dat",
self.mem_width,
size=self.interconnect_input_ports,
packed=True,
explicit_array=True)
# self._to_iac_valid = self.var("ab_to_mem_valid",
# self.interconnect_input_ports)
self._agg_buffers = []
# Add input aggregations buffers
for i in range(self.interconnect_input_ports):
# add children aggregator buffers...
agg_buffer_new = AggregationBuffer(self.agg_height,
self.data_width,
self.mem_width,
self.max_agg_schedule)
self._agg_buffers.append(agg_buffer_new)
self.add_child(f"agg_in_{i}", agg_buffer_new,
clk=self._clk,
rst_n=self._rst_n,
data_in=self._data_consume[i],
valid_in=self._valid_consume[i],
align=self._align_to_agg[i],
data_out=self._to_iac_dat[i],
valid_out=self._to_iac_valid[i])
#######################################
##### END: AGG BUFFERS (OPTIONAL) #####
#######################################
self._ready_tba = self.var("ready_tba", self.interconnect_output_ports)
####################################
##### INPUT ADDRESS CONTROLLER #####
####################################
self._wen_to_arb = self.var("wen_to_arb", self.mem_input_ports,
size=self.banks,
explicit_array=True,
packed=True)
self._addr_to_arb = self.var("addr_to_arb",
self.address_width,
size=(self.banks,
self.mem_input_ports),
explicit_array=True,
packed=True)
self._data_to_arb = self.var("data_to_arb",
self.data_width,
size=(self.banks,
self.mem_input_ports,
self.fw_int),
explicit_array=True,
packed=True)
# Connect these inputs ports to an address generator
iac = InputAddrCtrl(interconnect_input_ports=self.interconnect_input_ports,
mem_depth=self.mem_depth,
num_tiles=self.num_tiles,
banks=self.banks,
iterator_support=self.input_iterator_support,
address_width=self.address_width,
data_width=self.data_width,
fetch_width=self.mem_width,
multiwrite=self.multiwrite,
strg_wr_ports=self.mem_input_ports,
config_width=self.input_config_width)
self.add_child(f"input_addr_ctrl", iac,
clk=self._clk,
rst_n=self._rst_n,
valid_in=self._to_iac_valid,
# wen_en=kts.concat(*([kts.const(1, 1)] * self.interconnect_input_ports)),
wen_en=self._arb_wen_en,
data_in=self._to_iac_dat,
wen_to_sram=self._wen_to_arb,
addr_out=self._addr_to_arb,
port_out=self._port_wens,
data_out=self._data_to_arb)
#########################################
##### END: INPUT ADDRESS CONTROLLER #####
#########################################
self._arb_acks = self.var("arb_acks",
self.interconnect_output_ports,
size=self.banks,
explicit_array=True,
packed=True)
self._prefetch_step = self.var("prefetch_step", self.interconnect_output_ports)
self._oac_step = self.var("oac_step", self.interconnect_output_ports)
self._oac_valid = self.var("oac_valid", self.interconnect_output_ports)
self._ren_out = self.var("ren_out",
self.interconnect_output_ports,
size=self.banks,
explicit_array=True,
packed=True)
self._ren_out_tpose = self.var("ren_out_tpose",
self.banks,
size=self.interconnect_output_ports,
explicit_array=True,
packed=True)
self._oac_addr_out = self.var("oac_addr_out",
self.address_width,
size=self.interconnect_output_ports,
explicit_array=True,
packed=True)
#####################################
##### OUTPUT ADDRESS CONTROLLER #####
#####################################
oac = OutputAddrCtrl(interconnect_output_ports=self.interconnect_output_ports,
mem_depth=self.mem_depth,
num_tiles=self.num_tiles,
banks=self.banks,
iterator_support=self.output_iterator_support,
address_width=self.address_width,
config_width=self.output_config_width)
if self.remove_tb:
self.wire(self._oac_valid, self._ren)
self.wire(self._oac_step, self._ren)
else:
self.wire(self._oac_valid, self._prefetch_step)
self.wire(self._oac_step, self._ack_reduced)
self.chain_idx_bits = max(1, clog2(num_tiles))
self._enable_chain_output = self.input("enable_chain_output", 1)
self._chain_idx_output = self.input("chain_idx_output", self.chain_idx_bits)
self.add_child(f"output_addr_ctrl", oac,
clk=self._clk,
rst_n=self._rst_n,
valid_in=self._oac_valid,
ren=self._ren_out,
addr_out=self._oac_addr_out,
step_in=self._oac_step)
for i in range(self.interconnect_output_ports):
for j in range(self.banks):
self.wire(self._ren_out_tpose[i][j], self._ren_out[j][i])
##############################
##### READ/WRITE ARBITER #####
##############################
# Hook up the read write arbiters for each bank
self._arb_dat_out = self.var("arb_dat_out",
self.data_width,
size=(self.banks,
self.mem_output_ports,
self.fw_int),
explicit_array=True,
packed=True)
self._arb_port_out = self.var("arb_port_out",
self.interconnect_output_ports,
size=(self.banks,
self.mem_output_ports),
explicit_array=True,
packed=True)
self._arb_valid_out = self.var("arb_valid_out", self.mem_output_ports,
size=self.banks,
explicit_array=True,
packed=True)
self._rd_sync_gate = self.var("rd_sync_gate",
self.interconnect_output_ports)
self.arbiters = []
for i in range(self.banks):
rw_arb = RWArbiter(fetch_width=self.mem_width,
data_width=self.data_width,
memory_depth=self.mem_depth,
num_tiles=self.num_tiles,
int_in_ports=self.interconnect_input_ports,
int_out_ports=self.interconnect_output_ports,
strg_wr_ports=self.mem_input_ports,
strg_rd_ports=self.mem_output_ports,
read_delay=self.read_delay,
rw_same_cycle=self.rw_same_cycle,
separate_addresses=self.rw_same_cycle)
self.arbiters.append(rw_arb)
self.add_child(f"rw_arb_{i}", rw_arb,
clk=self._clk,
rst_n=self._rst_n,
wen_in=self._wen_to_arb[i],
w_data=self._data_to_arb[i],
w_addr=self._addr_to_arb[i],
data_from_mem=self._data_from_strg[i],
mem_valid_data=self._mem_valid_data[i],
out_mem_valid_data=self._out_mem_valid_data[i],
ren_en=self._arb_ren_en,
rd_addr=self._oac_addr_out,
out_data=self._arb_dat_out[i],
out_port=self._arb_port_out[i],
out_valid=self._arb_valid_out[i],
cen_mem=self._cen_to_strg[i],
wen_mem=self._wen_to_strg[i],
data_to_mem=self._data_to_strg[i],
out_ack=self._arb_acks[i])
# Bind the separate addrs
if self.rw_same_cycle:
self.wire(rw_arb.ports.wr_addr_to_mem, self._wr_addr_out[i])
self.wire(rw_arb.ports.rd_addr_to_mem, self._rd_addr_out[i])
else:
self.wire(rw_arb.ports.addr_to_mem, self._addr_out[i])
if self.remove_tb:
self.wire(rw_arb.ports.ren_in, self._ren_out[i])
else:
self.wire(rw_arb.ports.ren_in, self._ren_out[i] & self._rd_sync_gate)
self.num_tb_bits = max(1, clog2(self.num_tb))
self._data_to_sync = self.var("data_to_sync",
self.data_width,
size=(self.interconnect_output_ports,
self.fw_int),
explicit_array=True,
packed=True)
self._valid_to_sync = self.var("valid_to_sync", self.interconnect_output_ports)
self._data_to_tba = self.var("data_to_tba",
self.data_width,
size=(self.interconnect_output_ports,
self.fw_int),
explicit_array=True,
packed=True)
self._valid_to_tba = self.var("valid_to_tba", self.interconnect_output_ports)
self._data_to_pref = self.var("data_to_pref",
self.data_width,
size=(self.interconnect_output_ports,
self.fw_int),
explicit_array=True,
packed=True)
self._valid_to_pref = self.var("valid_to_pref", self.interconnect_output_ports)
#######################
##### DEMUX READS #####
#######################
dmux_rd = DemuxReads(fetch_width=self.mem_width,
data_width=self.data_width,
banks=self.banks,
int_out_ports=self.interconnect_output_ports,
strg_rd_ports=self.mem_output_ports)
self._arb_dat_out_f = self.var("arb_dat_out_f",
self.data_width,
size=(self.banks * self.mem_output_ports,
self.fw_int),
explicit_array=True,
packed=True)
self._arb_port_out_f = self.var("arb_port_out_f",
self.interconnect_output_ports,
size=(self.banks * self.mem_output_ports),
explicit_array=True,
packed=True)
self._arb_valid_out_f = self.var("arb_valid_out_f", self.mem_output_ports * self.banks)
self._arb_mem_valid_data_f = self.var("arb_mem_valid_data_f", self.mem_output_ports * self.banks)
self._arb_mem_valid_data_out = self.var("arb_mem_valid_data_out",
self.interconnect_output_ports)
self._mem_valid_data_sync = self.var("mem_valid_data_sync",
self.interconnect_output_ports)
self._mem_valid_data_pref = self.var("mem_valid_data_pref",
self.interconnect_output_ports)
tmp_cnt = 0
for i in range(self.banks):
for j in range(self.mem_output_ports):
self.wire(self._arb_dat_out_f[tmp_cnt], self._arb_dat_out[i][j])
self.wire(self._arb_port_out_f[tmp_cnt], self._arb_port_out[i][j])
self.wire(self._arb_valid_out_f[tmp_cnt], self._arb_valid_out[i][j])
self.wire(self._arb_mem_valid_data_f[tmp_cnt], self._out_mem_valid_data[i][j])
tmp_cnt = tmp_cnt + 1
# If this is end of the road...
if self.remove_tb:
assert self.fw_int == 1, "Make it easier on me now..."
self.add_child("demux_rds", dmux_rd,
clk=self._clk,
rst_n=self._rst_n,
data_in=self._arb_dat_out_f,
mem_valid_data=self._arb_mem_valid_data_f,
mem_valid_data_out=self._arb_mem_valid_data_out,
valid_in=self._arb_valid_out_f,
port_in=self._arb_port_out_f,
valid_out=self._valid_out_alt)
for i in range(self.interconnect_output_ports):
self.wire(self._data_out[i], dmux_rd.ports.data_out[i])
else:
self.add_child("demux_rds", dmux_rd,
clk=self._clk,
rst_n=self._rst_n,
data_in=self._arb_dat_out_f,
mem_valid_data=self._arb_mem_valid_data_f,
mem_valid_data_out=self._arb_mem_valid_data_out,
valid_in=self._arb_valid_out_f,
port_in=self._arb_port_out_f,
data_out=self._data_to_sync,
valid_out=self._valid_to_sync)
#######################
##### SYNC GROUPS #####
#######################
sync_group = SyncGroups(fetch_width=self.mem_width,
data_width=self.data_width,
int_out_ports=self.interconnect_output_ports)
for i in range(self.interconnect_output_ports):
self.wire(self._ren_out_reduced[i], self._ren_out_tpose[i].r_or())
self.add_child("sync_grp", sync_group,
clk=self._clk,
rst_n=self._rst_n,
data_in=self._data_to_sync,
mem_valid_data=self._arb_mem_valid_data_out,
mem_valid_data_out=self._mem_valid_data_sync,
valid_in=self._valid_to_sync,
data_out=self._data_to_pref,
valid_out=self._valid_to_pref,
ren_in=self._ren_out_reduced,
rd_sync_gate=self._rd_sync_gate,
ack_in=self._ack_reduced)
# This is the end of the line if we aren't using tb
######################
##### PREFETCHER #####
######################
prefetchers = []
for i in range(self.interconnect_output_ports):
pref = Prefetcher(fetch_width=self.mem_width,
data_width=self.data_width,
max_prefetch=self.max_prefetch)
prefetchers.append(pref)
if self.num_tb == 0:
assert self.fw_int == 1, \
"If no transpose buffer, data width needs match memory width"
self.add_child(f"pre_fetch_{i}", pref,
clk=self._clk,
rst_n=self._rst_n,
data_in=self._data_to_pref[i],
mem_valid_data=self._mem_valid_data_sync[i],
mem_valid_data_out=self._mem_valid_data_pref[i],
valid_read=self._valid_to_pref[i],
tba_rdy_in=self._ren[i],
# data_out=self._data_out[i],
valid_out=self._valid_out_alt[i],
prefetch_step=self._prefetch_step[i])
self.wire(self._data_out[i], pref.ports.data_out[0])
else:
self.add_child(f"pre_fetch_{i}", pref,
clk=self._clk,
rst_n=self._rst_n,
data_in=self._data_to_pref[i],
mem_valid_data=self._mem_valid_data_sync[i],
mem_valid_data_out=self._mem_valid_data_pref[i],
valid_read=self._valid_to_pref[i],
tba_rdy_in=self._ready_tba[i],
data_out=self._data_to_tba[i],
valid_out=self._valid_to_tba[i],
prefetch_step=self._prefetch_step[i])
#############################
##### TRANSPOSE BUFFERS #####
#############################
if self.num_tb > 0:
self._tb_data_out = self.var("tb_data_out", self.data_width,
size=self.interconnect_output_ports,
explicit_array=True,
packed=True)
self._tb_valid_out = self.var("tb_valid_out", self.interconnect_output_ports)
for i in range(self.interconnect_output_ports):
tba = TransposeBufferAggregation(word_width=self.data_width,
fetch_width=self.fw_int,
num_tb=self.num_tb,
max_tb_height=self.max_tb_height,
max_range=self.tb_range_max,
max_range_inner=self.tb_range_inner_max,
max_stride=self.max_tb_stride,
tb_iterator_support=self.tb_iterator_support)
self.add_child(f"tba_{i}", tba,
clk=self._clk,
rst_n=self._rst_n,
SRAM_to_tb_data=self._data_to_tba[i],
valid_data=self._valid_to_tba[i],
tb_index_for_data=0,
ack_in=self._valid_to_tba[i],
mem_valid_data=self._mem_valid_data_pref[i],
tb_to_interconnect_data=self._tb_data_out[i],
tb_to_interconnect_valid=self._tb_valid_out[i],
tb_arbiter_rdy=self._ready_tba[i],
tba_ren=self._ren[i])
for i in range(self.interconnect_output_ports):
self.wire(self._data_out[i], self._tb_data_out[i])
# self.wire(self._valid_out[i], self._tb_valid_out[i])
else:
self.wire(self._valid_out, self._valid_out_alt)
####################
##### ADD CODE #####
####################
self.add_code(self.transpose_acks)
self.add_code(self.reduce_acks)
def __init__(self, iterator_support=6, config_width=16):
super().__init__(f"addr_gen_{iterator_support}_{config_width}",
debug=True)
# Store local...
self.iterator_support = iterator_support
self.config_width = config_width
# PORT DEFS: begin
# INPUTS
self._clk = self.clock("clk")
self._rst_n = self.reset("rst_n")
self._restart = self.input("restart", 1)
self._strides = self.input("strides",
self.config_width,
size=self.iterator_support,
packed=True,
explicit_array=True)
self._strides.add_attribute(
ConfigRegAttr("Strides of address generator"))
self._strides.add_attribute(
FormalAttr(f"{self._strides.name}", FormalSignalConstraint.SOLVE))
self._starting_addr = self.input("starting_addr", self.config_width)
self._starting_addr.add_attribute(
ConfigRegAttr("Starting address of address generator"))
self._starting_addr.add_attribute(
FormalAttr(f"{self._starting_addr.name}",
FormalSignalConstraint.SOLVE))
self._step = self.input("step", 1)
self._mux_sel = self.input("mux_sel",
max(clog2(self.iterator_support), 1))
# OUTPUTS
# TODO why is this config width instead of address width?
self._addr_out = self.output("addr_out", self.config_width)
# PORT DEFS: end
# LOCAL VARIABLES: begin
self._strt_addr = self.var("strt_addr", self.config_width)
self._calc_addr = self.var("calc_addr", self.config_width)
self._max_value = self.var("max_value", self.iterator_support)
# LOCAL VARIABLES: end
# GENERATION LOGIC: begin
self.wire(self._strt_addr, self._starting_addr)
self.wire(self._addr_out, self._calc_addr)
self._current_addr = self.var("current_addr", self.config_width)
# Calculate address by taking previous calculation and adding the muxed stride
self.wire(self._calc_addr, self._strt_addr + self._current_addr)
self.add_code(self.calculate_address)
def __init__(self,
interconnect_input_ports,
interconnect_output_ports,
depth_width=16,
sprt_stcl_valid=False,
stcl_cnt_width=16,
stcl_iter_support=4):
super().__init__("app_ctrl", debug=True)
self.int_in_ports = interconnect_input_ports
self.int_out_ports = interconnect_output_ports
self.depth_width = depth_width
self.sprt_stcl_valid = sprt_stcl_valid
self.stcl_cnt_width = stcl_cnt_width
self.stcl_iter_support = stcl_iter_support
# Clock and Reset
self._clk = self.clock("clk")
self._rst_n = self.reset("rst_n")
# IO
self._wen_in = self.input("wen_in", self.int_in_ports)
self._ren_in = self.input("ren_in", self.int_out_ports)
self._ren_update = self.input("ren_update", self.int_out_ports)
self._tb_valid = self.input("tb_valid", self.int_out_ports)
self._valid_out_data = self.output("valid_out_data", self.int_out_ports)
self._valid_out_stencil = self.output("valid_out_stencil", self.int_out_ports)
# Send tb valid to valid out for now...
if self.sprt_stcl_valid:
# Add the config registers to watch
self._ranges = self.input("ranges", self.stcl_cnt_width,
size=self.stcl_iter_support,
packed=True,
explicit_array=True)
self._ranges.add_attribute(ConfigRegAttr("Ranges of stencil valid generator"))
self._threshold = self.input("threshold", self.stcl_cnt_width,
size=self.stcl_iter_support,
packed=True,
explicit_array=True)
self._threshold.add_attribute(ConfigRegAttr("Threshold of stencil valid generator"))
self._dim_counter = self.var("dim_counter", self.stcl_cnt_width,
size=self.stcl_iter_support,
packed=True,
explicit_array=True)
self._update = self.var("update", self.stcl_iter_support)
self.wire(self._update[0], const(1, 1))
for i in range(self.stcl_iter_support - 1):
self.wire(self._update[i + 1],
(self._dim_counter[i] == (self._ranges[i] - 1)) & self._update[i])
for i in range(self.stcl_iter_support):
self.add_code(self.dim_counter_update, idx=i)
# Now we need to just compute stencil valid
threshold_comps = [self._dim_counter[_i] >= self._threshold[_i] for _i in range(self.stcl_iter_support)]
self.wire(self._valid_out_stencil[0], kts.concat(*threshold_comps).r_and())
for i in range(self.int_out_ports - 1):
# self.wire(self._valid_out_stencil[i + 1], 0)
# for multiple ports
self.wire(self._valid_out_stencil[i + 1], kts.concat(*threshold_comps).r_and())
else:
self.wire(self._valid_out_stencil, self._tb_valid)
# Now gate the valid with stencil valid
self.wire(self._valid_out_data, self._tb_valid & self._valid_out_stencil)
self._wr_delay_state_n = self.var("wr_delay_state_n", self.int_out_ports)
self._wen_out = self.output("wen_out", self.int_in_ports)
self._ren_out = self.output("ren_out", self.int_out_ports)
self._write_depth_wo = self.input("write_depth_wo", self.depth_width,
size=self.int_in_ports,
explicit_array=True,
packed=True)
self._write_depth_wo.add_attribute(ConfigRegAttr("Depth of writes"))
self._write_depth_ss = self.input("write_depth_ss", self.depth_width,
size=self.int_in_ports,
explicit_array=True,
packed=True)
self._write_depth_ss.add_attribute(ConfigRegAttr("Depth of writes"))
self._write_depth = self.var("write_depth", self.depth_width,
size=self.int_in_ports,
explicit_array=True,
packed=True)
for i in range(self.int_in_ports):
self.wire(self._write_depth[i],
kts.ternary(self._wr_delay_state_n[i],
self._write_depth_ss[i],
self._write_depth_wo[i]))
self._read_depth = self.input("read_depth", self.depth_width,
size=self.int_out_ports,
explicit_array=True,
packed=True)
self._read_depth.add_attribute(ConfigRegAttr("Depth of reads"))
self._write_count = self.var("write_count", self.depth_width,
size=self.int_in_ports,
explicit_array=True,
packed=True)
self._read_count = self.var("read_count", self.depth_width,
size=self.int_out_ports,
explicit_array=True,
packed=True)
self._write_done = self.var("write_done", self.int_in_ports)
self._write_done_ff = self.var("write_done_ff", self.int_in_ports)
self._read_done = self.var("read_done", self.int_out_ports)
self._read_done_ff = self.var("read_done_ff", self.int_out_ports)
self.in_port_bits = max(1, kts.clog2(self.int_in_ports))
self._input_port = self.input("input_port", self.in_port_bits,
size=self.int_out_ports,
explicit_array=True,
packed=True)
self._input_port.add_attribute(ConfigRegAttr("Relative input port for an output port"))
self.out_port_bits = max(1, kts.clog2(self.int_out_ports))
self._output_port = self.input("output_port", self.out_port_bits,
size=self.int_in_ports,
explicit_array=True,
packed=True)
self._output_port.add_attribute(ConfigRegAttr("Relative output port for an input port"))
self._prefill = self.input("prefill", self.int_out_ports)
self._prefill.add_attribute(ConfigRegAttr("Is the input stream prewritten?"))
for i in range(self.int_out_ports):
self.add_code(self.set_read_done, idx=i)
if self.int_in_ports == 1:
self.add_code(self.set_read_done_ff_one_wr, idx=i)
else:
self.add_code(self.set_read_done_ff, idx=i)
# self._write_done_comb = self.var("write_done_comb", self.int_in_ports)
for i in range(self.int_in_ports):
self.add_code(self.set_write_done, idx=i)
self.add_code(self.set_write_done_ff, idx=i)
for i in range(self.int_in_ports):
self.add_code(self.set_write_cnt, idx=i)
for i in range(self.int_out_ports):
if self.int_in_ports == 1:
self.add_code(self.set_read_cnt_one_wr, idx=i)
else:
self.add_code(self.set_read_cnt, idx=i)
for i in range(self.int_out_ports):
if self.int_in_ports == 1:
self.add_code(self.set_wr_delay_state_one_wr, idx=i)
else:
self.add_code(self.set_wr_delay_state, idx=i)
self._read_on = self.var("read_on", self.int_out_ports)
for i in range(self.int_out_ports):
self.wire(self._read_on[i], self._read_depth[i].r_or())
# If we have prefill enabled, we are skipping the initial delay step...
self.wire(self._ren_out,
(self._wr_delay_state_n | self._prefill) & ~self._read_done_ff & self._ren_in &
self._read_on)
self.wire(self._wen_out, ~self._write_done_ff & self._wen_in)
def add_config_reg(generator, name, description, bitwidth, **kwargs):
cfg_reg = generator.input(name, bitwidth, **kwargs)
cfg_reg.add_attribute(ConfigRegAttr(description))
return cfg_reg
def __init__(self,
fetch_width,
data_width,
max_prefetch):
super().__init__("prefetcher")
# Capture to the object
self.fetch_width = fetch_width
self.data_width = data_width
self.fw_int = int(self.fetch_width / self.data_width)
self.max_prefetch = max_prefetch
# Clock and Reset
self._clk = self.clock("clk")
self._rst_n = self.reset("rst_n")
# Inputs
self._data_in = self.input("data_in",
self.data_width,
size=self.fw_int,
explicit_array=True,
packed=True)
self._valid_read = self.input("valid_read", 1)
self._tba_rdy_in = self.input("tba_rdy_in", 1)
self._input_latency = self.input("input_latency", clog2(self.max_prefetch) + 1)
doc = "This register is set to denote the input latency loop for reads. " + \
"This is sent to an internal fifo and an almost full signal is " + \
"used to pull more reads that the transpose buffers need."
self._input_latency.add_attribute(ConfigRegAttr(doc))
self._max_lat = const(self.max_prefetch - 1,
clog2(self.max_prefetch) + 1)
# Outputs
self._data_out = self.output("data_out",
self.data_width,
size=self.fw_int,
explicit_array=True,
packed=True)
self._valid_out = self.output("valid_out", 1)
self._prefetch_step = self.output("prefetch_step", 1)
# Local Signals
self._cnt = self.var("cnt", clog2(self.max_prefetch) + 1)
self._fifo_empty = self.var("fifo_empty", 1)
self._fifo_full = self.var("fifo_full", 1)
reg_fifo = RegFIFO(data_width=self.data_width,
width_mult=self.fw_int,
depth=self.max_prefetch)
self.add_child("fifo", reg_fifo,
clk=self._clk,
rst_n=self._rst_n,
clk_en=1,
data_in=self._data_in,
data_out=self._data_out,
push=self._valid_read,
pop=self._tba_rdy_in,
empty=self._fifo_empty,
full=self._fifo_full,
valid=self._valid_out)
# Generate
self.add_code(self.update_cnt)
self.add_code(self.set_prefetch_step)
def __init__(
self,
data_width=16, # CGRA Params
mem_depth=32,
default_iterator_support=3,
interconnect_input_ports=2, # Connection to int
interconnect_output_ports=2,
mem_input_ports=1,
mem_output_ports=1,
config_data_width=32,
config_addr_width=8,
cycle_count_width=16,
add_clk_enable=True,
add_flush=True):
super().__init__("pond", debug=True)
self.interconnect_input_ports = interconnect_input_ports
self.interconnect_output_ports = interconnect_output_ports
self.mem_input_ports = mem_input_ports
self.mem_output_ports = mem_output_ports
self.mem_depth = mem_depth
self.data_width = data_width
self.config_data_width = config_data_width
self.config_addr_width = config_addr_width
self.add_clk_enable = add_clk_enable
self.add_flush = add_flush
self.cycle_count_width = cycle_count_width
self.default_iterator_support = default_iterator_support
self.default_config_width = kts.clog2(self.mem_depth)
# inputs
self._clk = self.clock("clk")
self._clk.add_attribute(
FormalAttr(f"{self._clk.name}", FormalSignalConstraint.CLK))
self._rst_n = self.reset("rst_n")
self._rst_n.add_attribute(
FormalAttr(f"{self._rst_n.name}", FormalSignalConstraint.RSTN))
self._clk_en = self.clock_en("clk_en", 1)
# Enable/Disable tile
self._tile_en = self.input("tile_en", 1)
self._tile_en.add_attribute(
ConfigRegAttr("Tile logic enable manifested as clock gate"))
gclk = self.var("gclk", 1)
self._gclk = kts.util.clock(gclk)
self.wire(gclk, kts.util.clock(self._clk & self._tile_en))
self._cycle_count = add_counter(self, "cycle_count",
self.cycle_count_width)
# Create write enable + addr, same for read.
# self._write = self.input("write", self.interconnect_input_ports)
self._write = self.var("write", self.mem_input_ports)
# self._write.add_attribute(ControlSignalAttr(is_control=True))
self._write_addr = self.var("write_addr",
kts.clog2(self.mem_depth),
size=self.interconnect_input_ports,
explicit_array=True,
packed=True)
# Add "_pond" suffix to avoid error during garnet RTL generation
self._data_in = self.input("data_in_pond",
self.data_width,
size=self.interconnect_input_ports,
explicit_array=True,
packed=True)
self._data_in.add_attribute(
FormalAttr(f"{self._data_in.name}",
FormalSignalConstraint.SEQUENCE))
self._data_in.add_attribute(ControlSignalAttr(is_control=False))
self._read = self.var("read", self.mem_output_ports)
self._t_write = self.var("t_write", self.interconnect_input_ports)
self._t_read = self.var("t_read", self.interconnect_output_ports)
# self._read.add_attribute(ControlSignalAttr(is_control=True))
self._read_addr = self.var("read_addr",
kts.clog2(self.mem_depth),
size=self.interconnect_output_ports,
explicit_array=True,
packed=True)
self._s_read_addr = self.var("s_read_addr",
kts.clog2(self.mem_depth),
size=self.interconnect_output_ports,
explicit_array=True,
packed=True)
self._data_out = self.output("data_out_pond",
self.data_width,
size=self.interconnect_output_ports,
explicit_array=True,
packed=True)
self._data_out.add_attribute(
FormalAttr(f"{self._data_out.name}",
FormalSignalConstraint.SEQUENCE))
self._data_out.add_attribute(ControlSignalAttr(is_control=False))
self._valid_out = self.output("valid_out_pond",
self.interconnect_output_ports)
self._valid_out.add_attribute(
FormalAttr(f"{self._valid_out.name}",
FormalSignalConstraint.SEQUENCE))
self._valid_out.add_attribute(ControlSignalAttr(is_control=False))
self._mem_data_out = self.var("mem_data_out",
self.data_width,
size=self.mem_output_ports,
explicit_array=True,
packed=True)
self._s_mem_data_in = self.var("s_mem_data_in",
self.data_width,
size=self.interconnect_input_ports,
explicit_array=True,
packed=True)
self._mem_data_in = self.var("mem_data_in",
self.data_width,
size=self.mem_input_ports,
explicit_array=True,
packed=True)
self._s_mem_write_addr = self.var("s_mem_write_addr",
kts.clog2(self.mem_depth),
size=self.interconnect_input_ports,
explicit_array=True,
packed=True)
self._s_mem_read_addr = self.var("s_mem_read_addr",
kts.clog2(self.mem_depth),
size=self.interconnect_output_ports,
explicit_array=True,
packed=True)
self._mem_write_addr = self.var("mem_write_addr",
kts.clog2(self.mem_depth),
size=self.mem_input_ports,
explicit_array=True,
packed=True)
self._mem_read_addr = self.var("mem_read_addr",
kts.clog2(self.mem_depth),
size=self.mem_output_ports,
explicit_array=True,
packed=True)
if self.interconnect_output_ports == 1:
self.wire(self._data_out[0], self._mem_data_out[0])
else:
for i in range(self.interconnect_output_ports):
self.wire(self._data_out[i], self._mem_data_out[0])
# Valid out is simply passing the read signal through...
self.wire(self._valid_out, self._t_read)
# Create write addressors
for wr_port in range(self.interconnect_input_ports):
RF_WRITE_ITER = ForLoop(
iterator_support=self.default_iterator_support,
config_width=self.cycle_count_width)
RF_WRITE_ADDR = AddrGen(
iterator_support=self.default_iterator_support,
config_width=self.default_config_width)
RF_WRITE_SCHED = SchedGen(
iterator_support=self.default_iterator_support,
config_width=self.cycle_count_width,
use_enable=True)
self.add_child(f"rf_write_iter_{wr_port}",
RF_WRITE_ITER,
clk=self._gclk,
rst_n=self._rst_n,
step=self._t_write[wr_port])
# Whatever comes through here should hopefully just pipe through seamlessly
# addressor modules
self.add_child(f"rf_write_addr_{wr_port}",
RF_WRITE_ADDR,
clk=self._gclk,
rst_n=self._rst_n,
step=self._t_write[wr_port],
mux_sel=RF_WRITE_ITER.ports.mux_sel_out,
restart=RF_WRITE_ITER.ports.restart)
safe_wire(self, self._write_addr[wr_port],
RF_WRITE_ADDR.ports.addr_out)
self.add_child(f"rf_write_sched_{wr_port}",
RF_WRITE_SCHED,
clk=self._gclk,
rst_n=self._rst_n,
mux_sel=RF_WRITE_ITER.ports.mux_sel_out,
finished=RF_WRITE_ITER.ports.restart,
cycle_count=self._cycle_count,
valid_output=self._t_write[wr_port])
# Create read addressors
for rd_port in range(self.interconnect_output_ports):
RF_READ_ITER = ForLoop(
iterator_support=self.default_iterator_support,
config_width=self.cycle_count_width)
RF_READ_ADDR = AddrGen(
iterator_support=self.default_iterator_support,
config_width=self.default_config_width)
RF_READ_SCHED = SchedGen(
iterator_support=self.default_iterator_support,
config_width=self.cycle_count_width,
use_enable=True)
self.add_child(f"rf_read_iter_{rd_port}",
RF_READ_ITER,
clk=self._gclk,
rst_n=self._rst_n,
step=self._t_read[rd_port])
self.add_child(f"rf_read_addr_{rd_port}",
RF_READ_ADDR,
clk=self._gclk,
rst_n=self._rst_n,
step=self._t_read[rd_port],
mux_sel=RF_READ_ITER.ports.mux_sel_out,
restart=RF_READ_ITER.ports.restart)
if self.interconnect_output_ports > 1:
safe_wire(self, self._read_addr[rd_port],
RF_READ_ADDR.ports.addr_out)
else:
safe_wire(self, self._read_addr[rd_port],
RF_READ_ADDR.ports.addr_out)
self.add_child(f"rf_read_sched_{rd_port}",
RF_READ_SCHED,
clk=self._gclk,
rst_n=self._rst_n,
mux_sel=RF_READ_ITER.ports.mux_sel_out,
finished=RF_READ_ITER.ports.restart,
cycle_count=self._cycle_count,
valid_output=self._t_read[rd_port])
self.wire(self._write, self._t_write.r_or())
self.wire(self._mem_write_addr[0],
decode(self, self._t_write, self._s_mem_write_addr))
self.wire(self._mem_data_in[0],
decode(self, self._t_write, self._s_mem_data_in))
self.wire(self._read, self._t_read.r_or())
self.wire(self._mem_read_addr[0],
decode(self, self._t_read, self._s_mem_read_addr))
# ===================================
# Instantiate config hooks...
# ===================================
self.fw_int = 1
self.data_words_per_set = 2**self.config_addr_width
self.sets = int(
(self.fw_int * self.mem_depth) / self.data_words_per_set)
self.sets_per_macro = max(
1, int(self.mem_depth / self.data_words_per_set))
self.total_sets = max(1, 1 * self.sets_per_macro)
self._config_data_in = self.input("config_data_in",
self.config_data_width)
self._config_data_in.add_attribute(ControlSignalAttr(is_control=False))
self._config_data_in_shrt = self.var("config_data_in_shrt",
self.data_width)
self.wire(self._config_data_in_shrt,
self._config_data_in[self.data_width - 1, 0])
self._config_addr_in = self.input("config_addr_in",
self.config_addr_width)
self._config_addr_in.add_attribute(ControlSignalAttr(is_control=False))
self._config_data_out_shrt = self.var("config_data_out_shrt",
self.data_width,
size=self.total_sets,
explicit_array=True,
packed=True)
self._config_data_out = self.output("config_data_out",
self.config_data_width,
size=self.total_sets,
explicit_array=True,
packed=True)
self._config_data_out.add_attribute(
ControlSignalAttr(is_control=False))
for i in range(self.total_sets):
self.wire(
self._config_data_out[i],
self._config_data_out_shrt[i].extend(self.config_data_width))
self._config_read = self.input("config_read", 1)
self._config_read.add_attribute(ControlSignalAttr(is_control=False))
self._config_write = self.input("config_write", 1)
self._config_write.add_attribute(ControlSignalAttr(is_control=False))
self._config_en = self.input("config_en", self.total_sets)
self._config_en.add_attribute(ControlSignalAttr(is_control=False))
self._mem_data_cfg = self.var("mem_data_cfg",
self.data_width,
explicit_array=True,
packed=True)
self._mem_addr_cfg = self.var("mem_addr_cfg",
kts.clog2(self.mem_depth))
# Add config...
stg_cfg_seq = StorageConfigSeq(
data_width=self.data_width,
config_addr_width=self.config_addr_width,
addr_width=kts.clog2(self.mem_depth),
fetch_width=self.data_width,
total_sets=self.total_sets,
sets_per_macro=self.sets_per_macro)
# The clock to config sequencer needs to be the normal clock or
# if the tile is off, we bring the clock back in based on config_en
cfg_seq_clk = self.var("cfg_seq_clk", 1)
self._cfg_seq_clk = kts.util.clock(cfg_seq_clk)
self.wire(cfg_seq_clk, kts.util.clock(self._gclk))
self.add_child(f"config_seq",
stg_cfg_seq,
clk=self._cfg_seq_clk,
rst_n=self._rst_n,
clk_en=self._clk_en | self._config_en.r_or(),
config_data_in=self._config_data_in_shrt,
config_addr_in=self._config_addr_in,
config_wr=self._config_write,
config_rd=self._config_read,
config_en=self._config_en,
wr_data=self._mem_data_cfg,
rd_data_out=self._config_data_out_shrt,
addr_out=self._mem_addr_cfg)
if self.interconnect_output_ports == 1:
self.wire(stg_cfg_seq.ports.rd_data_stg, self._mem_data_out)
else:
self.wire(stg_cfg_seq.ports.rd_data_stg[0], self._mem_data_out[0])
self.RF_GEN = RegisterFile(data_width=self.data_width,
write_ports=self.mem_input_ports,
read_ports=self.mem_output_ports,
width_mult=1,
depth=self.mem_depth,
read_delay=0)
# Now we can instantiate and wire up the register file
self.add_child(f"rf",
self.RF_GEN,
clk=self._gclk,
rst_n=self._rst_n,
data_out=self._mem_data_out)
# Opt in for config_write
self._write_rf = self.var("write_rf", self.mem_input_ports)
self.wire(
self._write_rf[0],
kts.ternary(self._config_en.r_or(), self._config_write,
self._write[0]))
for i in range(self.mem_input_ports - 1):
self.wire(
self._write_rf[i + 1],
kts.ternary(self._config_en.r_or(), kts.const(0, 1),
self._write[i + 1]))
self.wire(self.RF_GEN.ports.wen, self._write_rf)
# Opt in for config_data_in
for i in range(self.interconnect_input_ports):
self.wire(
self._s_mem_data_in[i],
kts.ternary(self._config_en.r_or(), self._mem_data_cfg,
self._data_in[i]))
self.wire(self.RF_GEN.ports.data_in, self._mem_data_in)
# Opt in for config_addr
for i in range(self.interconnect_input_ports):
self.wire(
self._s_mem_write_addr[i],
kts.ternary(self._config_en.r_or(), self._mem_addr_cfg,
self._write_addr[i]))
self.wire(self.RF_GEN.ports.wr_addr, self._mem_write_addr[0])
for i in range(self.interconnect_output_ports):
self.wire(
self._s_mem_read_addr[i],
kts.ternary(self._config_en.r_or(), self._mem_addr_cfg,
self._read_addr[i]))
self.wire(self.RF_GEN.ports.rd_addr, self._mem_read_addr[0])
if self.add_clk_enable:
# self.clock_en("clk_en")
kts.passes.auto_insert_clock_enable(self.internal_generator)
clk_en_port = self.internal_generator.get_port("clk_en")
clk_en_port.add_attribute(ControlSignalAttr(False))
if self.add_flush:
self.add_attribute("sync-reset=flush")
kts.passes.auto_insert_sync_reset(self.internal_generator)
flush_port = self.internal_generator.get_port("flush")
flush_port.add_attribute(ControlSignalAttr(True))
# Finally, lift the config regs...
lift_config_reg(self.internal_generator)
def __init__(self,
interconnect_input_ports=2,
mem_depth=32,
num_tiles=1,
banks=1,
iterator_support=6,
address_width=5,
data_width=16,
fetch_width=16,
multiwrite=1,
strg_wr_ports=2,
config_width=16):
super().__init__("input_addr_ctrl", debug=True)
assert multiwrite >= 1, "Multiwrite must be at least 1..."
self.interconnect_input_ports = interconnect_input_ports
self.mem_depth = mem_depth
self.num_tiles = num_tiles
self.banks = banks
self.iterator_support = iterator_support
self.address_width = address_width
self.port_sched_width = max(1, clog2(self.interconnect_input_ports))
self.data_width = data_width
self.fetch_width = fetch_width
self.fw_int = int(self.fetch_width / self.data_width)
self.multiwrite = multiwrite
self.strg_wr_ports = strg_wr_ports
self.config_width = config_width
self.mem_addr_width = clog2(self.num_tiles * self.mem_depth)
if self.banks > 1:
self.bank_addr_width = clog2(self.banks)
else:
self.bank_addr_width = 0
self.address_width = self.mem_addr_width + self.bank_addr_width
# Clock and Reset
self._clk = self.clock("clk")
self._rst_n = self.reset("rst_n")
# Inputs
# phases = [] TODO
# Take in the valid and data and attach an address + direct to a port
self._valid_in = self.input("valid_in", self.interconnect_input_ports)
self._wen_en = self.input("wen_en", self.interconnect_input_ports)
self._wen_en_saved = self.var("wen_en_saved",
self.interconnect_input_ports)
self._data_in = self.input("data_in",
self.data_width,
size=(self.interconnect_input_ports,
self.fw_int),
explicit_array=True,
packed=True)
self._data_in_saved = self.var("data_in_saved",
self.data_width,
size=(self.interconnect_input_ports,
self.fw_int),
explicit_array=True,
packed=True)
# Outputs
self._wen = self.output("wen_to_sram",
self.strg_wr_ports,
size=self.banks,
explicit_array=True,
packed=True)
wen_full_size = (self.interconnect_input_ports, self.multiwrite)
self._wen_full = self.var("wen_full",
self.banks,
size=wen_full_size,
explicit_array=True,
packed=True)
self._wen_reduced = self.var("wen_reduced",
self.banks,
size=self.interconnect_input_ports,
explicit_array=True,
packed=True)
self._wen_reduced_saved = self.var("wen_reduced_saved",
self.banks,
size=self.interconnect_input_ports,
explicit_array=True,
packed=True)
self._addresses = self.output("addr_out",
self.mem_addr_width,
size=(self.banks, self.strg_wr_ports),
explicit_array=True,
packed=True)
self._data_out = self.output("data_out",
self.data_width,
size=(self.banks, self.strg_wr_ports,
self.fw_int),
explicit_array=True,
packed=True)
self._port_out_exp = self.var("port_out_exp",
self.interconnect_input_ports,
size=self.banks,
explicit_array=True,
packed=True)
self._port_out = self.output("port_out", self.interconnect_input_ports)
self._counter = self.var("counter", self.port_sched_width)
# Wire to port out
for i in range(self.interconnect_input_ports):
new_tmp = []
for j in range(self.banks):
new_tmp.append(self._port_out_exp[j][i])
self.wire(self._port_out[i], kts.concat(*new_tmp).r_or())
self._done = self.var("done",
self.strg_wr_ports,
size=self.banks,
explicit_array=True,
packed=True)
# LOCAL VARS
self._local_addrs = self.var("local_addrs",
self.address_width,
size=(self.interconnect_input_ports,
self.multiwrite),
packed=True,
explicit_array=True)
self._local_addrs_saved = self.var("local_addrs_saved",
self.address_width,
size=(self.interconnect_input_ports,
self.multiwrite),
packed=True,
explicit_array=True)
for i in range(self.interconnect_input_ports):
for j in range(self.banks):
concat_ports = []
for k in range(self.multiwrite):
concat_ports.append(self._wen_full[i][k][j])
self.wire(self._wen_reduced[i][j],
kts.concat(*concat_ports).r_or())
if self.banks == 1 and self.interconnect_input_ports == 1:
self.wire(self._wen_full[0][0][0], self._valid_in)
elif self.banks == 1 and self.interconnect_input_ports > 1:
self.add_code(self.set_wen_single)
else:
self.add_code(self.set_wen_mult)
# MAIN
# Iterate through all banks to priority decode the wen
self.add_code(self.decode_out_lowest)
# Also set the write ports on the storage
if self.strg_wr_ports > 1:
self._idx_cnt = self.var("idx_cnt",
8,
size=(self.banks, self.strg_wr_ports - 1),
explicit_array=True,
packed=True)
for i in range(self.strg_wr_ports - 1):
self.add_code(self.decode_out_alt, idx=i + 1)
# Now we should instantiate the child address generators
# (1 per input port) to send to the sram banks
for i in range(self.interconnect_input_ports):
self.add_child(f"address_gen_{i}",
AddrGen(iterator_support=self.iterator_support,
config_width=self.config_width),
clk=self._clk,
rst_n=self._rst_n,
clk_en=const(1, 1),
flush=const(0, 1),
step=self._valid_in[i])
# Need to check that the address falls into the bank for implicit banking
# Then, obey the input schedule to send the proper Aggregator to the output
# The wen to sram should be that the valid for the selected port is high
# Do the same thing for the output address
assert self.multiwrite <= self.banks and self.multiwrite > 0,\
"Multiwrite should be between 1 and banks"
if self.multiwrite > 1:
size = (self.interconnect_input_ports, self.multiwrite - 1)
self._offsets_cfg = self.input("offsets_cfg",
self.address_width,
size=size,
packed=True,
explicit_array=True)
doc = "These offsets provide the ability to write to multiple banks explicitly"
self._offsets_cfg.add_attribute(ConfigRegAttr(doc))
self.add_code(self.set_multiwrite_addrs)
# to handle multiple input ports going to fewer SRAM write ports
self.add_code(self.set_int_ports_counter)
self.add_code(self.save_mult_int_signals)