nextpnr/machxo2/facade_import.py
2023-10-02 14:49:17 +02:00

696 lines
26 KiB
Python

#!/usr/bin/env python3
import argparse
import json
import sys
from os import path
tiletype_names = dict()
gfx_wire_ids = dict()
gfx_wire_names = list()
parser = argparse.ArgumentParser(description="import MachXO2 routing and bels from Project Trellis")
parser.add_argument("device", type=str, help="target device")
parser.add_argument("-p", "--constids", type=str, help="path to constids.inc")
parser.add_argument("-g", "--gfxh", type=str, help="path to gfx.h")
parser.add_argument("-L", "--libdir", type=str, action="append", help="extra Python library path")
args = parser.parse_args()
sys.path += args.libdir
import pytrellis
import database
import pip_classes
import timing_dbs
with open(args.gfxh) as f:
state = 0
for line in f:
if state == 0 and line.startswith("enum GfxTileWireId"):
state = 1
elif state == 1 and line.startswith("};"):
state = 0
elif state == 1 and (line.startswith("{") or line.strip() == ""):
pass
elif state == 1:
idx = len(gfx_wire_ids)
name = line.strip().rstrip(",")
gfx_wire_ids[name] = idx
gfx_wire_names.append(name)
def wire_type(name):
return "WIRE_TYPE_NONE"
# Get the index for a tiletype
def get_tiletype_index(name):
if name in tiletype_names:
return tiletype_names[name]
idx = len(tiletype_names)
tiletype_names[name] = idx
return idx
def package_shortname(long_name, family):
if long_name.startswith("CABGA"):
if (family == "MachXO"):
return "B" + long_name[5:]
else:
return "BG" + long_name[5:]
elif long_name.startswith("CSBGA"):
if (family == "MachXO"):
return "M" + long_name[5:]
else:
return "MG" + long_name[5:]
elif long_name.startswith("CSFBGA"):
return "MG" + long_name[6:]
elif long_name.startswith("UCBGA"):
return "UMG" + long_name[5:]
elif long_name.startswith("FPBGA"):
return "FG" + long_name[5:]
elif long_name.startswith("FTBGA"):
if (family == "MachXO"):
return "FT" + long_name[5:]
else:
return "FTG" + long_name[5:]
elif long_name.startswith("WLCSP"):
if (family == "MachXO3D"):
return "UTG" + long_name[5:]
else:
return "UWG" + long_name[5:]
elif long_name.startswith("TQFP"):
if (family == "MachXO"):
return "T" + long_name[4:]
else:
return "TG" + long_name[4:]
elif long_name.startswith("QFN"):
if (family == "MachXO3D"):
return "SG" + long_name[3:]
else:
if long_name[3]=="8":
return "QN" + long_name[3:]
else:
return "SG" + long_name[3:]
else:
print("unknown package name " + long_name)
sys.exit(-1)
constids = dict()
class BinaryBlobAssembler:
def l(self, name, ltype = None, export = False):
if ltype is None:
print("label %s" % (name,))
else:
print("label %s %s" % (name, ltype))
def r(self, name, comment):
if comment is None:
print("ref %s" % (name,))
else:
print("ref %s %s" % (name, comment))
def r_slice(self, name, length, comment):
if comment is None:
print("ref %s" % (name,))
else:
print("ref %s %s" % (name, comment))
print ("u32 %d" % (length, ))
def s(self, s, comment):
assert "|" not in s
print("str |%s| %s" % (s, comment))
def u8(self, v, comment):
assert -128 <= int(v) <= 127
if comment is None:
print("u8 %d" % (v,))
else:
print("u8 %d %s" % (v, comment))
def u16(self, v, comment):
# is actually used as signed 16 bit
assert -32768 <= int(v) <= 32767
if comment is None:
print("u16 %d" % (v,))
else:
print("u16 %d %s" % (v, comment))
def u32(self, v, comment):
if comment is None:
print("u32 %d" % (v,))
else:
print("u32 %d %s" % (v, comment))
def pre(self, s):
print("pre %s" % s)
def post(self, s):
print("post %s" % s)
def push(self, name):
print("push %s" % name)
def pop(self):
print("pop")
def get_bel_index(rg, loc, name):
tile = rg.tiles[loc]
idx = 0
for bel in tile.bels:
if rg.to_str(bel.name) == name:
return idx
idx += 1
# FIXME: I/O pins can be missing in various rows. Is there a nice way to
# assert on each device size?
return None
packages = {}
pindata = []
variants = {}
def process_devices_db(family, device):
devicefile = path.join(database.get_db_root(), "devices.json")
with open(devicefile, 'r') as f:
devicedb = json.load(f)
for varname, vardata in sorted(devicedb["families"][family]["devices"][device]["variants"].items()):
variants[varname] = vardata
def process_pio_db(rg, device):
piofile = path.join(database.get_db_root(), dev_family[device], dev_names[device], "iodb.json")
with open(piofile, 'r') as f:
piodb = json.load(f)
for pkgname, pkgdata in sorted(piodb["packages"].items()):
pins = []
for name, pinloc in sorted(pkgdata.items()):
x = pinloc["col"]
y = pinloc["row"]
loc = pytrellis.Location(x, y)
pio = "PIO" + pinloc["pio"]
bel_idx = get_bel_index(rg, loc, pio)
if bel_idx is not None:
pins.append((name, loc, bel_idx))
packages[pkgname] = pins
for metaitem in piodb["pio_metadata"]:
x = metaitem["col"]
y = metaitem["row"]
loc = pytrellis.Location(x, y)
pio = "PIO" + metaitem["pio"]
bank = metaitem["bank"]
if "function" in metaitem:
pinfunc = metaitem["function"]
else:
pinfunc = None
dqs = -1
if "dqs" in metaitem:
pass
# tdqs = metaitem["dqs"]
# if tdqs[0] == "L":
# dqs = 0
# elif tdqs[0] == "R":
# dqs = 2048
# suffix_size = 0
# while tdqs[-(suffix_size+1)].isdigit():
# suffix_size += 1
# dqs |= int(tdqs[-suffix_size:])
bel_idx = get_bel_index(rg, loc, pio)
if bel_idx is not None:
pindata.append((loc, bel_idx, bank, pinfunc, dqs))
speed_grade_names = {
"MachXO2": ["1", "2", "3", "4", "5", "6"],
"MachXO3": ["5", "6"],
"MachXO3D": ["2", "3", "5", "6"]
}
speed_grade_cells = {}
speed_grade_pips = {}
pip_class_to_idx = {"default": 0, "zero": 1}
timing_port_xform = {
"RAD0": "A0",
"RAD1": "B0",
"RAD2": "C0",
"RAD3": "D0",
}
delay_db = {}
# Convert from Lattice-style grouped SLICE to new nextpnr split style SLICE
def postprocess_timing_data(cells):
def delay_diff(x, y):
return (x[0] - y[0], x[1] - y[1])
split_cells = {}
comb_delays = {}
comb_delays[("A", "F")] = delay_db["SLICE"][("A0", "F0")]
comb_delays[("B", "F")] = delay_db["SLICE"][("B0", "F0")]
comb_delays[("C", "F")] = delay_db["SLICE"][("C0", "F0")]
comb_delays[("D", "F")] = delay_db["SLICE"][("D0", "F0")]
comb_delays[("A", "OFX")] = delay_db["SLICE"][("A0", "OFX0")]
comb_delays[("B", "OFX")] = delay_db["SLICE"][("B0", "OFX0")]
comb_delays[("C", "OFX")] = delay_db["SLICE"][("C0", "OFX0")]
comb_delays[("D", "OFX")] = delay_db["SLICE"][("D0", "OFX0")]
comb_delays[("M", "OFX")] = delay_db["SLICE"][("M0", "OFX0")] # worst case
comb_delays[("F1", "OFX")] = delay_diff(delay_db["SLICE"][("A1", "OFX0")],
delay_db["SLICE"][("A1", "F1")])
comb_delays[("FXA", "OFX")] = delay_db["SLICE"][("FXA", "OFX1")]
comb_delays[("FXB", "OFX")] = delay_db["SLICE"][("FXB", "OFX1")]
split_cells["TRELLIS_COMB"] = comb_delays
carry0_delays = {}
carry0_delays[("A", "F")] = delay_db["SLICE"][("A0", "F0")]
carry0_delays[("B", "F")] = delay_db["SLICE"][("B0", "F0")]
carry0_delays[("C", "F")] = delay_db["SLICE"][("C0", "F0")]
carry0_delays[("D", "F")] = delay_db["SLICE"][("D0", "F0")]
carry0_delays[("A", "FCO")] = delay_db["SLICE"][("A0", "FCO")]
carry0_delays[("B", "FCO")] = delay_db["SLICE"][("B0", "FCO")]
carry0_delays[("C", "FCO")] = delay_db["SLICE"][("C0", "FCO")]
carry0_delays[("D", "FCO")] = delay_db["SLICE"][("D0", "FCO")]
carry0_delays[("FCI", "F")] = delay_db["SLICE"][("FCI", "F0")]
carry0_delays[("FCI", "FCO")] = delay_db["SLICE"][("FCI", "FCO")]
split_cells["TRELLIS_COMB_CARRY0"] = carry0_delays
carry1_delays = {}
carry1_delays[("A", "F")] = delay_db["SLICE"][("A1", "F1")]
carry1_delays[("B", "F")] = delay_db["SLICE"][("B1", "F1")]
carry1_delays[("C", "F")] = delay_db["SLICE"][("C1", "F1")]
carry1_delays[("D", "F")] = delay_db["SLICE"][("D1", "F1")]
carry1_delays[("A", "FCO")] = delay_db["SLICE"][("A1", "FCO")]
carry1_delays[("B", "FCO")] = delay_db["SLICE"][("B1", "FCO")]
carry1_delays[("C", "FCO")] = delay_db["SLICE"][("C1", "FCO")]
carry1_delays[("D", "FCO")] = delay_db["SLICE"][("D1", "FCO")]
carry1_delays[("FCI", "F")] = delay_diff(delay_db["SLICE"][("FCI", "F1")], delay_db["SLICE"][("FCI", "FCO")])
carry1_delays[("FCI", "FCO")] = (0, 0)
split_cells["TRELLIS_COMB_CARRY1"] = carry1_delays
for celltype, celldelays in sorted(split_cells.items()):
delays = []
setupholds = []
for (from_pin, to_pin), (min_delay, max_delay) in sorted(celldelays.items()):
delays.append((constids[from_pin], constids[to_pin], min_delay, max_delay))
cells.append((constids[celltype], delays, setupholds))
def process_timing_data(family):
for grade in speed_grade_names[family]:
with open(timing_dbs.cells_db_path(family, grade)) as f:
cell_data = json.load(f)
cells = []
for cell, cdata in sorted(cell_data.items()):
celltype = constids[cell.replace(":", "_").replace("=", "_").replace(",", "_")]
delays = []
setupholds = []
delay_db[cell] = {}
for entry in cdata:
if entry["type"] == "Width":
continue
elif entry["type"] == "IOPath":
from_pin = entry["from_pin"][1] if type(entry["from_pin"]) is list else entry["from_pin"]
if from_pin in timing_port_xform:
from_pin = timing_port_xform[from_pin]
to_pin = entry["to_pin"]
if to_pin in timing_port_xform:
to_pin = timing_port_xform[to_pin]
min_delay = min(entry["rising"][0], entry["falling"][0])
max_delay = min(entry["rising"][2], entry["falling"][2])
delay_db[cell][(from_pin, to_pin)] = (min_delay, max_delay)
delays.append((constids[from_pin], constids[to_pin], min_delay, max_delay))
elif entry["type"] == "SetupHold":
if type(entry["pin"]) is list:
continue
pin = constids[entry["pin"]]
clock = constids[entry["clock"][1]]
min_setup = entry["setup"][0]
max_setup = entry["setup"][2]
min_hold = entry["hold"][0]
max_hold = entry["hold"][2]
setupholds.append((pin, clock, min_setup, max_setup, min_hold, max_hold))
else:
assert False, entry["type"]
cells.append((celltype, delays, setupholds))
postprocess_timing_data(cells)
pip_class_delays = []
for i in range(len(pip_class_to_idx)):
pip_class_delays.append((50, 50, 0, 0))
pip_class_delays[pip_class_to_idx["zero"]] = (0, 0, 0, 0)
with open(timing_dbs.interconnect_db_path(family, grade)) as f:
interconn_data = json.load(f)
for pipclass, pipdata in sorted(interconn_data.items()):
min_delay = pipdata["delay"][0] * 1.1
max_delay = pipdata["delay"][2] * 1.1
min_fanout = pipdata["fanout"][0]
max_fanout = pipdata["fanout"][2]
if grade == speed_grade_names[family][0]:
pip_class_to_idx[pipclass] = len(pip_class_delays)
pip_class_delays.append((min_delay, max_delay, min_fanout, max_fanout))
else:
if pipclass in pip_class_to_idx:
pip_class_delays[pip_class_to_idx[pipclass]] = (min_delay, max_delay, min_fanout, max_fanout)
speed_grade_cells[grade] = cells
speed_grade_pips[grade] = pip_class_delays
def get_pip_class(wire_from, wire_to):
if "FCO" in wire_from or "FCI" in wire_to:
return pip_class_to_idx["zero"]
if "F5" in wire_from or "FX" in wire_from or "FXA" in wire_to or "FXB" in wire_to:
return pip_class_to_idx["zero"]
if "JCLK" in wire_from:
wire_from = wire_from.replace("JCLK","CLK")
class_name = pip_classes.get_pip_class(wire_from, wire_to)
if class_name is None or class_name not in pip_class_to_idx:
class_name = "default"
return pip_class_to_idx[class_name]
def write_database(family, dev_name, chip, rg, endianness):
def write_loc(loc, sym_name):
bba.u16(loc.x, "%s.x" % sym_name)
bba.u16(loc.y, "%s.y" % sym_name)
# Use Lattice naming conventions, so convert to 1-based col indexing.
def get_wire_name(loc, idx):
tile = rg.tiles[loc]
return "R{}C{}_{}".format(loc.y, loc.x + 1, rg.to_str(tile.wires[idx].name))
# Before doing anything, ensure sorted routing graph iteration matches
# y, x
loc_iter = list(sorted(rg.tiles, key=lambda l : (l.y, l.x)))
i = 1 # Drop (-2, -2) location.
for y in range(0, max_row+1):
for x in range(0, max_col+1):
l = loc_iter[i]
assert((y, x) == (l.y, l.x))
i = i + 1
bba = BinaryBlobAssembler()
bba.pre('#include "nextpnr.h"')
bba.pre('#include "embed.h"')
bba.pre('NEXTPNR_NAMESPACE_BEGIN')
bba.post('EmbeddedFile chipdb_file_%s("machxo2/chipdb-%s.bin", chipdb_blob_%s);' % (dev_name, dev_name, dev_name))
bba.post('NEXTPNR_NAMESPACE_END')
bba.push("chipdb_blob_%s" % args.device)
bba.r("chip_info", "chip_info")
# Nominally should be in order, but support situations where python
# decides to iterate over rg.tiles out-of-order.
for l in loc_iter:
t = rg.tiles[l]
# Do not include special globals location for now.
if (l.x, l.y) == (-2, -2):
continue
if len(t.arcs) > 0:
bba.l("loc%d_%d_pips" % (l.y, l.x), "PipInfoPOD")
for arc in t.arcs:
write_loc(arc.srcWire.rel, "src")
write_loc(arc.sinkWire.rel, "dst")
bba.u16(arc.srcWire.id, "src_idx {}".format(get_wire_name(arc.srcWire.rel, arc.srcWire.id)))
bba.u16(arc.sinkWire.id, "dst_idx {}".format(get_wire_name(arc.sinkWire.rel, arc.sinkWire.id)))
src_name = get_wire_name(arc.srcWire.rel, arc.srcWire.id)
snk_name = get_wire_name(arc.sinkWire.rel, arc.sinkWire.id)
bba.u16(get_pip_class(src_name, snk_name), "timing_class")
bba.u8(get_tiletype_index(rg.to_str(arc.tiletype)), "tile_type")
cls = arc.cls
bba.u8(cls, "pip_type")
bba.u16(arc.lutperm_flags, "lutperm_flags")
bba.u16(0, "padding")
if len(t.wires) > 0:
for wire_idx in range(len(t.wires)):
wire = t.wires[wire_idx]
if len(wire.arcsDownhill) > 0:
bba.l("loc%d_%d_wire%d_downpips" % (l.y, l.x, wire_idx), "PipLocatorPOD")
for dp in wire.arcsDownhill:
write_loc(dp.rel, "rel_loc")
bba.u32(dp.id, "index")
if len(wire.arcsUphill) > 0:
bba.l("loc%d_%d_wire%d_uppips" % (l.y, l.x, wire_idx), "PipLocatorPOD")
for up in wire.arcsUphill:
write_loc(up.rel, "rel_loc")
bba.u32(up.id, "index")
if len(wire.belPins) > 0:
bba.l("loc%d_%d_wire%d_belpins" % (l.y, l.x, wire_idx), "BelPortPOD")
for bp in wire.belPins:
write_loc(bp.bel.rel, "rel_bel_loc")
bba.u32(bp.bel.id, "bel_index")
bba.u32(constids[rg.to_str(bp.pin)], "port")
bba.l("loc%d_%d_wires" % (l.y, l.x), "WireInfoPOD")
for wire_idx in range(len(t.wires)):
wire = t.wires[wire_idx]
bba.s(rg.to_str(wire.name), "name")
bba.u16(constids[wire_type(rg.to_str(wire.name))], "type")
if ("TILE_WIRE_" + rg.to_str(wire.name)) in gfx_wire_ids:
bba.u16(gfx_wire_ids["TILE_WIRE_" + rg.to_str(wire.name)], "tile_wire")
else:
bba.u16(0, "tile_wire")
bba.r_slice("loc%d_%d_wire%d_uppips" % (l.y, l.x, wire_idx) if len(wire.arcsUphill) > 0 else None, len(wire.arcsUphill), "pips_uphill")
bba.r_slice("loc%d_%d_wire%d_downpips" % (l.y, l.x, wire_idx) if len(wire.arcsDownhill) > 0 else None, len(wire.arcsDownhill), "pips_downhill")
bba.r_slice("loc%d_%d_wire%d_belpins" % (l.y, l.x, wire_idx) if len(wire.belPins) > 0 else None, len(wire.belPins), "bel_pins")
if len(t.bels) > 0:
for bel_idx in range(len(t.bels)):
bel = t.bels[bel_idx]
bba.l("loc%d_%d_bel%d_wires" % (l.y, l.x, bel_idx), "BelWirePOD")
for pin in bel.wires:
write_loc(pin.wire.rel, "rel_wire_loc")
bba.u32(pin.wire.id, "wire_index")
bba.u32(constids[rg.to_str(pin.pin)], "port")
bba.u32(int(pin.dir), "type")
bba.l("loc%d_%d_bels" % (l.y, l.x), "BelInfoPOD")
for bel_idx in range(len(t.bels)):
bel = t.bels[bel_idx]
bba.s(rg.to_str(bel.name), "name")
bba.u32(constids[rg.to_str(bel.type)], "type")
bba.u32(bel.z, "z")
bba.r_slice("loc%d_%d_bel%d_wires" % (l.y, l.x, bel_idx), len(bel.wires), "bel_wires")
bba.l("tiles", "TileTypePOD")
for l in loc_iter:
t = rg.tiles[l]
if (l.y, l.x) == (-2, -2):
continue
bba.r_slice("loc%d_%d_bels" % (l.y, l.x) if len(t.bels) > 0 else None, len(t.bels), "bel_data")
bba.r_slice("loc%d_%d_wires" % (l.y, l.x) if len(t.wires) > 0 else None, len(t.wires), "wire_data")
bba.r_slice("loc%d_%d_pips" % (l.y, l.x) if len(t.arcs) > 0 else None, len(t.arcs), "pips_data")
for y in range(0, max_row+1):
for x in range(0, max_col+1):
bba.l("tile_info_%d_%d" % (x, y), "TileNamePOD")
for tile in chip.get_tiles_by_position(y, x):
bba.s(tile.info.name, "name")
bba.u16(get_tiletype_index(tile.info.type), "type_idx")
bba.u16(0, "padding")
bba.l("tiles_info", "TileInfoPOD")
for y in range(0, max_row+1):
for x in range(0, max_col+1):
bba.r_slice("tile_info_%d_%d" % (x, y), len(chip.get_tiles_by_position(y, x)), "tile_names")
for package, pkgdata in sorted(packages.items()):
bba.l("package_data_%s" % package, "PackagePinPOD")
for pin in pkgdata:
name, loc, bel_idx = pin
bba.s(name, "name")
write_loc(loc, "abs_loc")
bba.u32(bel_idx, "bel_index")
bba.l("package_data", "PackageInfoPOD")
for package, pkgdata in sorted(packages.items()):
bba.s(package, "name")
bba.r_slice("package_data_%s" % package, len(pkgdata), "pin_data")
bba.l("pio_info", "PIOInfoPOD")
for pin in pindata:
loc, bel_idx, bank, func, dqs = pin
write_loc(loc, "abs_loc")
bba.u32(bel_idx, "bel_index")
if func is not None and func != "WRITEN":
bba.s(func, "function_name")
else:
bba.r(None, "function_name")
# TODO: io_grouping? And DQS.
bba.u16(bank, "bank")
bba.u16(dqs, "dqsgroup")
bba.l("tiletype_names", "RelPtr<char>")
for tt, idx in sorted(tiletype_names.items(), key=lambda x: x[1]):
bba.s(tt, "name")
for grade in speed_grade_names[family]:
for cell in speed_grade_cells[grade]:
celltype, delays, setupholds = cell
if len(delays) > 0:
bba.l("cell_%d_delays_%s" % (celltype, grade))
for delay in delays:
from_pin, to_pin, min_delay, max_delay = delay
bba.u32(from_pin, "from_pin")
bba.u32(to_pin, "to_pin")
bba.u32(min_delay, "min_delay")
bba.u32(max_delay, "max_delay")
if len(setupholds) > 0:
bba.l("cell_%d_setupholds_%s" % (celltype, grade))
for sh in setupholds:
pin, clock, min_setup, max_setup, min_hold, max_hold = sh
bba.u32(pin, "sig_port")
bba.u32(clock, "clock_port")
bba.u32(min_setup, "min_setup")
bba.u32(max_setup, "max_setup")
bba.u32(min_hold, "min_hold")
bba.u32(max_hold, "max_hold")
bba.l("cell_timing_data_%s" % grade)
for cell in speed_grade_cells[grade]:
celltype, delays, setupholds = cell
bba.u32(celltype, "cell_type")
bba.r_slice("cell_%d_delays_%s" % (celltype, grade) if len(delays) > 0 else None, len(delays), "delays")
bba.r_slice("cell_%d_setupholds_%s" % (celltype, grade) if len(delays) > 0 else None, len(setupholds), "setupholds")
bba.l("pip_timing_data_%s" % grade)
for pipclass in speed_grade_pips[grade]:
min_delay, max_delay, min_fanout, max_fanout = pipclass
bba.u32(min_delay, "min_delay")
bba.u32(max_delay, "max_delay")
bba.u32(min_fanout, "min_fanout")
bba.u32(max_fanout, "max_fanout")
bba.l("speed_grade_data")
for grade in speed_grade_names[family]:
bba.r_slice("cell_timing_data_%s" % grade, len(speed_grade_cells[grade]), "cell_timings")
bba.r_slice("pip_timing_data_%s" % grade, len(speed_grade_pips[grade]), "pip_classes")
for name, var_data in sorted(variants.items()):
bba.l("supported_packages_%s" % name, "PackageSupportedPOD")
for package in var_data["packages"]:
bba.s(package, "name")
bba.s(package_shortname(package, chip.info.family), "short_name")
bba.l("supported_speed_grades_%s" % name, "SpeedSupportedPOD")
for speed in var_data["speeds"]:
bba.u16(speed, "speed")
bba.u16(speed_grade_names[family].index(str(speed)), "index")
bba.l("supported_suffixes_%s" % name, "SuffixeSupportedPOD")
for suffix in var_data["suffixes"]:
bba.s(suffix, "suffix")
bba.l("variant_data", "VariantInfoPOD")
for name, var_data in sorted(variants.items()):
bba.s(name, "variant_name")
bba.r_slice("supported_packages_%s" % name, len(var_data["packages"]), "supported_packages")
bba.r_slice("supported_speed_grades_%s" % name, len(var_data["speeds"]), "supported_speed_grades")
bba.r_slice("supported_suffixes_%s" % name, len(var_data["suffixes"]), "supported_suffixes")
bba.l("spine_info", "SpineInfoPOD")
spines = chip.global_data_machxo2.spines
for spine in spines:
bba.u32(spine.row, "row")
bba.l("chip_info")
bba.s(chip.info.family, "family")
bba.s(chip.info.name, "device_name")
bba.u32(max_col + 1, "width")
bba.u32(max_row + 1, "height")
bba.u32((max_col + 1) * (max_row + 1), "num_tiles")
bba.u32(const_id_count, "const_id_count")
bba.r_slice("tiles", (max_col + 1) * (max_row + 1), "tiles")
bba.r_slice("tiletype_names", len(tiletype_names), "tiletype_names")
bba.r_slice("package_data", len(packages), "package_info")
bba.r_slice("pio_info", len(pindata), "pio_info")
bba.r_slice("tiles_info", (max_col + 1) * (max_row + 1), "tile_info")
bba.r_slice("variant_data", len(variants), "variant_info")
bba.r_slice("spine_info", len(spines), "spine_info")
bba.r_slice("speed_grade_data", len(speed_grade_names[family]), "speed_grades")
bba.pop()
return bba
dev_family = {
"256X": "MachXO",
"640X": "MachXO",
"1200X":"MachXO",
"2280X":"MachXO",
"256": "MachXO2",
"640": "MachXO2",
"1200": "MachXO2",
"2000": "MachXO2",
"4000": "MachXO2",
"7000": "MachXO2",
"1300": "MachXO3",
"2100": "MachXO3",
"4300": "MachXO3",
"6900": "MachXO3",
"9400": "MachXO3",
"4300D": "MachXO3D",
"9400D": "MachXO3D"
}
dev_names = {
"256X": "LCMXO256",
"640X": "LCMXO640",
"1200X":"LCMXO1200",
"2280X":"LCMXO2280",
"256": "LCMXO2-256",
"640": "LCMXO2-640",
"1200": "LCMXO2-1200",
"2000": "LCMXO2-2000",
"4000": "LCMXO2-4000",
"7000": "LCMXO2-7000",
"1300": "LCMXO3-1300",
"2100": "LCMXO3-2100",
"4300": "LCMXO3-4300",
"6900": "LCMXO3-6900",
"9400": "LCMXO3-9400",
"4300D": "LCMXO3D-4300",
"9400D": "LCMXO3D-9400"
}
def main():
global max_row, max_col, const_id_count
pytrellis.load_database(database.get_db_root())
args = parser.parse_args()
const_id_count = 1 # count ID_NONE
with open(args.constids) as f:
for line in f:
line = line.replace("(", " ")
line = line.replace(")", " ")
line = line.split()
if len(line) == 0:
continue
assert len(line) == 2
assert line[0] == "X"
idx = len(constids) + 1
constids[line[1]] = idx
const_id_count += 1
constids["SLICE"] = constids["TRELLIS_SLICE"]
constids["PIO"] = constids["TRELLIS_IO"]
chip = pytrellis.Chip(dev_names[args.device])
rg = pytrellis.make_optimized_chipdb(chip, include_lutperm_pips=True, split_slice_mode=True)
max_row = chip.get_max_row()
max_col = chip.get_max_col()
process_timing_data(dev_family[args.device])
process_pio_db(rg, args.device)
process_devices_db(chip.info.family, chip.info.name)
bba = write_database(dev_family[args.device],args.device, chip, rg, "le")
if __name__ == "__main__":
main()