1 """ Pipeline and BufferedHandshake implementation, conforming to the same API.
2 For multi-input and multi-output variants, see multipipe.
4 Associated development bugs:
5 * http://bugs.libre-riscv.org/show_bug.cgi?id=64
6 * http://bugs.libre-riscv.org/show_bug.cgi?id=57
11 a strategically very important function that is identical in function
12 to nmigen's Signal.eq function, except it may take objects, or a list
13 of objects, or a tuple of objects, and where objects may also be
19 stage requires compliance with a strict API that may be
20 implemented in several means, including as a static class.
21 the methods of a stage instance must be as follows:
23 * ispec() - Input data format specification
24 returns an object or a list or tuple of objects, or
25 a Record, each object having an "eq" function which
26 takes responsibility for copying by assignment all
28 * ospec() - Output data format specification
29 requirements as for ospec
30 * process(m, i) - Processes an ispec-formatted object
31 returns a combinatorial block of a result that
32 may be assigned to the output, by way of the "eq"
34 * setup(m, i) - Optional function for setting up submodules
35 may be used for more complex stages, to link
36 the input (i) to submodules. must take responsibility
37 for adding those submodules to the module (m).
38 the submodules must be combinatorial blocks and
39 must have their inputs and output linked combinatorially.
41 Both StageCls (for use with non-static classes) and Stage (for use
42 by static classes) are abstract classes from which, for convenience
43 and as a courtesy to other developers, anything conforming to the
44 Stage API may *choose* to derive.
49 A useful combinatorial wrapper around stages that chains them together
50 and then presents a Stage-API-conformant interface. By presenting
51 the same API as the stages it wraps, it can clearly be used recursively.
56 A convenience class that takes an input shape, output shape, a
57 "processing" function and an optional "setup" function. Honestly
58 though, there's not much more effort to just... create a class
59 that returns a couple of Records (see ExampleAddRecordStage in
65 A convenience class that takes a single function as a parameter,
66 that is chain-called to create the exact same input and output spec.
67 It has a process() function that simply returns its input.
69 Instances of this class are completely redundant if handed to
70 StageChain, however when passed to UnbufferedPipeline they
71 can be used to introduce a single clock delay.
76 The base class for pipelines. Contains previous and next ready/valid/data.
77 Also has an extremely useful "connect" function that can be used to
78 connect a chain of pipelines and present the exact same prev/next
84 A simple stalling clock-synchronised pipeline that has no buffering
85 (unlike BufferedHandshake). Data flows on *every* clock cycle when
86 the conditions are right (this is nominally when the input is valid
87 and the output is ready).
89 A stall anywhere along the line will result in a stall back-propagating
90 down the entire chain. The BufferedHandshake by contrast will buffer
91 incoming data, allowing previous stages one clock cycle's grace before
94 An advantage of the UnbufferedPipeline over the Buffered one is
95 that the amount of logic needed (number of gates) is greatly
96 reduced (no second set of buffers basically)
98 The disadvantage of the UnbufferedPipeline is that the valid/ready
99 logic, if chained together, is *combinatorial*, resulting in
100 progressively larger gate delay.
102 PassThroughHandshake:
105 A Control class that introduces a single clock delay, passing its
106 data through unaltered. Unlike RegisterPipeline (which relies
107 on UnbufferedPipeline and PassThroughStage) it handles ready/valid
113 A convenience class that, because UnbufferedPipeline introduces a single
114 clock delay, when its stage is a PassThroughStage, it results in a Pipeline
115 stage that, duh, delays its (unmodified) input by one clock cycle.
120 nmigen implementation of buffered pipeline stage, based on zipcpu:
121 https://zipcpu.com/blog/2017/08/14/strategies-for-pipelining.html
123 this module requires quite a bit of thought to understand how it works
124 (and why it is needed in the first place). reading the above is
125 *strongly* recommended.
127 unlike john dawson's IEEE754 FPU STB/ACK signalling, which requires
128 the STB / ACK signals to raise and lower (on separate clocks) before
129 data may proceeed (thus only allowing one piece of data to proceed
130 on *ALTERNATE* cycles), the signalling here is a true pipeline
131 where data will flow on *every* clock when the conditions are right.
133 input acceptance conditions are when:
134 * incoming previous-stage strobe (p.i_valid) is HIGH
135 * outgoing previous-stage ready (p.o_ready) is LOW
137 output transmission conditions are when:
138 * outgoing next-stage strobe (n.o_valid) is HIGH
139 * outgoing next-stage ready (n.i_ready) is LOW
141 the tricky bit is when the input has valid data and the output is not
142 ready to accept it. if it wasn't for the clock synchronisation, it
143 would be possible to tell the input "hey don't send that data, we're
144 not ready". unfortunately, it's not possible to "change the past":
145 the previous stage *has no choice* but to pass on its data.
147 therefore, the incoming data *must* be accepted - and stored: that
148 is the responsibility / contract that this stage *must* accept.
149 on the same clock, it's possible to tell the input that it must
150 not send any more data. this is the "stall" condition.
152 we now effectively have *two* possible pieces of data to "choose" from:
153 the buffered data, and the incoming data. the decision as to which
154 to process and output is based on whether we are in "stall" or not.
155 i.e. when the next stage is no longer ready, the output comes from
156 the buffer if a stall had previously occurred, otherwise it comes
157 direct from processing the input.
159 this allows us to respect a synchronous "travelling STB" with what
160 dan calls a "buffered handshake".
162 it's quite a complex state machine!
167 Synchronised pipeline, Based on:
168 https://github.com/ZipCPU/dbgbus/blob/master/hexbus/rtl/hbdeword.v
171 from nmigen
import Signal
, Cat
, Const
, Mux
, Module
, Value
172 from nmigen
.cli
import verilog
, rtlil
173 from nmigen
.lib
.fifo
import SyncFIFO
, SyncFIFOBuffered
174 from nmigen
.hdl
.ast
import ArrayProxy
175 from nmigen
.hdl
.rec
import Record
, Layout
177 from abc
import ABCMeta
, abstractmethod
178 from collections
.abc
import Sequence
179 from queue
import Queue
182 class RecordObject(Record
):
183 def __init__(self
, layout
=None, name
=None):
184 Record
.__init
__(self
, layout
=layout
or [], name
=None)
186 def __setattr__(self
, k
, v
):
187 if k
in dir(Record
) or "fields" not in self
.__dict
__:
188 return object.__setattr
__(self
, k
, v
)
190 if isinstance(v
, Record
):
191 newlayout
= {k
: (k
, v
.layout
)}
193 newlayout
= {k
: (k
, v
.shape())}
194 self
.layout
.fields
.update(newlayout
)
197 for x
in self
.fields
.values():
202 """ contains signals that come *from* the previous stage (both in and out)
203 * i_valid: previous stage indicating all incoming data is valid.
204 may be a multi-bit signal, where all bits are required
205 to be asserted to indicate "valid".
206 * o_ready: output to next stage indicating readiness to accept data
207 * i_data : an input - added by the user of this class
210 def __init__(self
, i_width
=1, stage_ctl
=False):
211 self
.stage_ctl
= stage_ctl
212 self
.i_valid
= Signal(i_width
, name
="p_i_valid") # prev >>in self
213 self
._o
_ready
= Signal(name
="p_o_ready") # prev <<out self
214 self
.i_data
= None # XXX MUST BE ADDED BY USER
216 self
.s_o_ready
= Signal(name
="p_s_o_rdy") # prev <<out self
217 self
.trigger
= Signal(reset_less
=True)
221 """ public-facing API: indicates (externally) that stage is ready
224 return self
.s_o_ready
# set dynamically by stage
225 return self
._o
_ready
# return this when not under dynamic control
227 def _connect_in(self
, prev
, direct
=False, fn
=None):
228 """ internal helper function to connect stage to an input source.
229 do not use to connect stage-to-stage!
231 i_valid
= prev
.i_valid
if direct
else prev
.i_valid_test
232 i_data
= fn(prev
.i_data
) if fn
is not None else prev
.i_data
233 return [self
.i_valid
.eq(i_valid
),
234 prev
.o_ready
.eq(self
.o_ready
),
235 eq(self
.i_data
, i_data
),
239 def i_valid_test(self
):
240 vlen
= len(self
.i_valid
)
242 # multi-bit case: valid only when i_valid is all 1s
243 all1s
= Const(-1, (len(self
.i_valid
), False))
244 i_valid
= (self
.i_valid
== all1s
)
246 # single-bit i_valid case
247 i_valid
= self
.i_valid
249 # when stage indicates not ready, incoming data
250 # must "appear" to be not ready too
252 i_valid
= i_valid
& self
.s_o_ready
256 def elaborate(self
, platform
):
258 m
.d
.comb
+= self
.trigger
.eq(self
.i_valid_test
& self
.o_ready
)
262 return [self
.i_data
.eq(i
.i_data
),
263 self
.o_ready
.eq(i
.o_ready
),
264 self
.i_valid
.eq(i
.i_valid
)]
269 if hasattr(self
.i_data
, "ports"):
270 yield from self
.i_data
.ports()
271 elif isinstance(self
.i_data
, Sequence
):
272 yield from self
.i_data
281 """ contains the signals that go *to* the next stage (both in and out)
282 * o_valid: output indicating to next stage that data is valid
283 * i_ready: input from next stage indicating that it can accept data
284 * o_data : an output - added by the user of this class
286 def __init__(self
, stage_ctl
=False):
287 self
.stage_ctl
= stage_ctl
288 self
.o_valid
= Signal(name
="n_o_valid") # self out>> next
289 self
.i_ready
= Signal(name
="n_i_ready") # self <<in next
290 self
.o_data
= None # XXX MUST BE ADDED BY USER
292 self
.d_valid
= Signal(reset
=1) # INTERNAL (data valid)
293 self
.trigger
= Signal(reset_less
=True)
296 def i_ready_test(self
):
298 return self
.i_ready
& self
.d_valid
301 def connect_to_next(self
, nxt
):
302 """ helper function to connect to the next stage data/valid/ready.
303 data/valid is passed *TO* nxt, and ready comes *IN* from nxt.
304 use this when connecting stage-to-stage
306 return [nxt
.i_valid
.eq(self
.o_valid
),
307 self
.i_ready
.eq(nxt
.o_ready
),
308 eq(nxt
.i_data
, self
.o_data
),
311 def _connect_out(self
, nxt
, direct
=False, fn
=None):
312 """ internal helper function to connect stage to an output source.
313 do not use to connect stage-to-stage!
315 i_ready
= nxt
.i_ready
if direct
else nxt
.i_ready_test
316 o_data
= fn(nxt
.o_data
) if fn
is not None else nxt
.o_data
317 return [nxt
.o_valid
.eq(self
.o_valid
),
318 self
.i_ready
.eq(i_ready
),
319 eq(o_data
, self
.o_data
),
322 def elaborate(self
, platform
):
324 m
.d
.comb
+= self
.trigger
.eq(self
.i_ready_test
& self
.o_valid
)
330 if hasattr(self
.o_data
, "ports"):
331 yield from self
.o_data
.ports()
332 elif isinstance(self
.o_data
, Sequence
):
333 yield from self
.o_data
342 """ a helper class for iterating twin-argument compound data structures.
344 Record is a special (unusual, recursive) case, where the input may be
345 specified as a dictionary (which may contain further dictionaries,
346 recursively), where the field names of the dictionary must match
347 the Record's field spec. Alternatively, an object with the same
348 member names as the Record may be assigned: it does not have to
351 ArrayProxy is also special-cased, it's a bit messy: whilst ArrayProxy
352 has an eq function, the object being assigned to it (e.g. a python
353 object) might not. despite the *input* having an eq function,
354 that doesn't help us, because it's the *ArrayProxy* that's being
355 assigned to. so.... we cheat. use the ports() function of the
356 python object, enumerate them, find out the list of Signals that way,
359 def iterator2(self
, o
, i
):
360 if isinstance(o
, dict):
361 yield from self
.dict_iter2(o
, i
)
363 if not isinstance(o
, Sequence
):
365 for (ao
, ai
) in zip(o
, i
):
366 #print ("visit", fn, ao, ai)
367 if isinstance(ao
, Record
):
368 yield from self
.record_iter2(ao
, ai
)
369 elif isinstance(ao
, ArrayProxy
) and not isinstance(ai
, Value
):
370 yield from self
.arrayproxy_iter2(ao
, ai
)
374 def dict_iter2(self
, o
, i
):
375 for (k
, v
) in o
.items():
376 print ("d-iter", v
, i
[k
])
380 def _not_quite_working_with_all_unit_tests_record_iter2(self
, ao
, ai
):
381 print ("record_iter2", ao
, ai
, type(ao
), type(ai
))
382 if isinstance(ai
, Value
):
383 if isinstance(ao
, Sequence
):
385 for o
, i
in zip(ao
, ai
):
388 for idx
, (field_name
, field_shape
, _
) in enumerate(ao
.layout
):
389 if isinstance(field_shape
, Layout
):
393 if hasattr(val
, field_name
): # check for attribute
394 val
= getattr(val
, field_name
)
396 val
= val
[field_name
] # dictionary-style specification
397 yield from self
.iterator2(ao
.fields
[field_name
], val
)
399 def record_iter2(self
, ao
, ai
):
400 for idx
, (field_name
, field_shape
, _
) in enumerate(ao
.layout
):
401 if isinstance(field_shape
, Layout
):
405 if hasattr(val
, field_name
): # check for attribute
406 val
= getattr(val
, field_name
)
408 val
= val
[field_name
] # dictionary-style specification
409 yield from self
.iterator2(ao
.fields
[field_name
], val
)
411 def arrayproxy_iter2(self
, ao
, ai
):
413 op
= getattr(ao
, p
.name
)
414 print ("arrayproxy - p", p
, p
.name
)
415 yield from self
.iterator2(op
, p
)
419 """ a helper class for iterating single-argument compound data structures.
422 def iterate(self
, i
):
423 """ iterate a compound structure recursively using yield
425 if not isinstance(i
, Sequence
):
428 print ("iterate", ai
)
429 if isinstance(ai
, Record
):
430 print ("record", list(ai
.layout
))
431 yield from self
.record_iter(ai
)
432 elif isinstance(ai
, ArrayProxy
) and not isinstance(ai
, Value
):
433 yield from self
.array_iter(ai
)
437 def record_iter(self
, ai
):
438 for idx
, (field_name
, field_shape
, _
) in enumerate(ai
.layout
):
439 if isinstance(field_shape
, Layout
):
443 if hasattr(val
, field_name
): # check for attribute
444 val
= getattr(val
, field_name
)
446 val
= val
[field_name
] # dictionary-style specification
447 print ("recidx", idx
, field_name
, field_shape
, val
)
448 yield from self
.iterate(val
)
450 def array_iter(self
, ai
):
452 yield from self
.iterate(p
)
456 """ makes signals equal: a helper routine which identifies if it is being
457 passed a list (or tuple) of objects, or signals, or Records, and calls
458 the objects' eq function.
461 for (ao
, ai
) in Visitor2().iterator2(o
, i
):
463 if not isinstance(rres
, Sequence
):
470 """ flattens a compound structure recursively using Cat
472 from nmigen
.tools
import flatten
473 #res = list(flatten(i)) # works (as of nmigen commit f22106e5) HOWEVER...
474 res
= list(Visitor().iterate(i
)) # needed because input may be a sequence
478 class StageCls(metaclass
=ABCMeta
):
479 """ Class-based "Stage" API. requires instantiation (after derivation)
481 see "Stage API" above.. Note: python does *not* require derivation
482 from this class. All that is required is that the pipelines *have*
483 the functions listed in this class. Derivation from this class
484 is therefore merely a "courtesy" to maintainers.
487 def ispec(self
): pass # REQUIRED
489 def ospec(self
): pass # REQUIRED
491 #def setup(self, m, i): pass # OPTIONAL
493 def process(self
, i
): pass # REQUIRED
496 class Stage(metaclass
=ABCMeta
):
497 """ Static "Stage" API. does not require instantiation (after derivation)
499 see "Stage API" above. Note: python does *not* require derivation
500 from this class. All that is required is that the pipelines *have*
501 the functions listed in this class. Derivation from this class
502 is therefore merely a "courtesy" to maintainers.
514 #def setup(m, i): pass
521 class RecordBasedStage(Stage
):
522 """ convenience class which provides a Records-based layout.
523 honestly it's a lot easier just to create a direct Records-based
524 class (see ExampleAddRecordStage)
526 def __init__(self
, in_shape
, out_shape
, processfn
, setupfn
=None):
527 self
.in_shape
= in_shape
528 self
.out_shape
= out_shape
529 self
.__process
= processfn
530 self
.__setup
= setupfn
531 def ispec(self
): return Record(self
.in_shape
)
532 def ospec(self
): return Record(self
.out_shape
)
533 def process(seif
, i
): return self
.__process
(i
)
534 def setup(seif
, m
, i
): return self
.__setup
(m
, i
)
537 class StageChain(StageCls
):
538 """ pass in a list of stages, and they will automatically be
539 chained together via their input and output specs into a
542 the end result basically conforms to the exact same Stage API.
544 * input to this class will be the input of the first stage
545 * output of first stage goes into input of second
546 * output of second goes into input into third (etc. etc.)
547 * the output of this class will be the output of the last stage
549 def __init__(self
, chain
, specallocate
=False):
551 self
.specallocate
= specallocate
554 return self
.chain
[0].ispec()
557 return self
.chain
[-1].ospec()
559 def _specallocate_setup(self
, m
, i
):
560 for (idx
, c
) in enumerate(self
.chain
):
561 if hasattr(c
, "setup"):
562 c
.setup(m
, i
) # stage may have some module stuff
563 o
= self
.chain
[idx
].ospec() # last assignment survives
564 m
.d
.comb
+= eq(o
, c
.process(i
)) # process input into "o"
565 if idx
== len(self
.chain
)-1:
567 i
= self
.chain
[idx
+1].ispec() # new input on next loop
568 m
.d
.comb
+= eq(i
, o
) # assign to next input
569 return o
# last loop is the output
571 def _noallocate_setup(self
, m
, i
):
572 for (idx
, c
) in enumerate(self
.chain
):
573 if hasattr(c
, "setup"):
574 c
.setup(m
, i
) # stage may have some module stuff
575 i
= o
= c
.process(i
) # store input into "o"
576 return o
# last loop is the output
578 def setup(self
, m
, i
):
579 if self
.specallocate
:
580 self
.o
= self
._specallocate
_setup
(m
, i
)
582 self
.o
= self
._noallocate
_setup
(m
, i
)
584 def process(self
, i
):
585 return self
.o
# conform to Stage API: return last-loop output
589 """ Common functions for Pipeline API
591 def __init__(self
, stage
=None, in_multi
=None, stage_ctl
=False):
592 """ Base class containing ready/valid/data to previous and next stages
594 * p: contains ready/valid to the previous stage
595 * n: contains ready/valid to the next stage
597 Except when calling Controlbase.connect(), user must also:
598 * add i_data member to PrevControl (p) and
599 * add o_data member to NextControl (n)
603 # set up input and output IO ACK (prev/next ready/valid)
604 self
.p
= PrevControl(in_multi
, stage_ctl
)
605 self
.n
= NextControl(stage_ctl
)
607 # set up the input and output data
608 if stage
is not None:
609 self
.p
.i_data
= stage
.ispec() # input type
610 self
.n
.o_data
= stage
.ospec()
612 def connect_to_next(self
, nxt
):
613 """ helper function to connect to the next stage data/valid/ready.
615 return self
.n
.connect_to_next(nxt
.p
)
617 def _connect_in(self
, prev
):
618 """ internal helper function to connect stage to an input source.
619 do not use to connect stage-to-stage!
621 return self
.p
._connect
_in
(prev
.p
)
623 def _connect_out(self
, nxt
):
624 """ internal helper function to connect stage to an output source.
625 do not use to connect stage-to-stage!
627 return self
.n
._connect
_out
(nxt
.n
)
629 def connect(self
, pipechain
):
630 """ connects a chain (list) of Pipeline instances together and
631 links them to this ControlBase instance:
633 in <----> self <---> out
636 [pipe1, pipe2, pipe3, pipe4]
639 out---in out--in out---in
641 Also takes care of allocating i_data/o_data, by looking up
642 the data spec for each end of the pipechain. i.e It is NOT
643 necessary to allocate self.p.i_data or self.n.o_data manually:
644 this is handled AUTOMATICALLY, here.
646 Basically this function is the direct equivalent of StageChain,
647 except that unlike StageChain, the Pipeline logic is followed.
649 Just as StageChain presents an object that conforms to the
650 Stage API from a list of objects that also conform to the
651 Stage API, an object that calls this Pipeline connect function
652 has the exact same pipeline API as the list of pipline objects
655 Thus it becomes possible to build up larger chains recursively.
656 More complex chains (multi-input, multi-output) will have to be
659 eqs
= [] # collated list of assignment statements
661 # connect inter-chain
662 for i
in range(len(pipechain
)-1):
664 pipe2
= pipechain
[i
+1]
665 eqs
+= pipe1
.connect_to_next(pipe2
)
667 # connect front of chain to ourselves
669 self
.p
.i_data
= front
.stage
.ispec()
670 eqs
+= front
._connect
_in
(self
)
672 # connect end of chain to ourselves
674 self
.n
.o_data
= end
.stage
.ospec()
675 eqs
+= end
._connect
_out
(self
)
679 def _postprocess(self
, i
): # XXX DISABLED
680 return i
# RETURNS INPUT
681 if hasattr(self
.stage
, "postprocess"):
682 return self
.stage
.postprocess(i
)
685 def set_input(self
, i
):
686 """ helper function to set the input data
688 return eq(self
.p
.i_data
, i
)
691 return self
.p
.ports() + self
.n
.ports()
693 def _elaborate(self
, platform
):
694 """ handles case where stage has dynamic ready/valid functions
698 if self
.stage
is not None and hasattr(self
.stage
, "setup"):
699 self
.stage
.setup(m
, self
.p
.i_data
)
701 if not self
.p
.stage_ctl
:
704 # intercept the previous (outgoing) "ready", combine with stage ready
705 m
.d
.comb
+= self
.p
.s_o_ready
.eq(self
.p
._o
_ready
& self
.stage
.d_ready
)
707 # intercept the next (incoming) "ready" and combine it with data valid
708 sdv
= self
.stage
.d_valid(self
.n
.i_ready
)
709 m
.d
.comb
+= self
.n
.d_valid
.eq(self
.n
.i_ready
& sdv
)
714 class BufferedHandshake(ControlBase
):
715 """ buffered pipeline stage. data and strobe signals travel in sync.
716 if ever the input is ready and the output is not, processed data
717 is shunted in a temporary register.
719 Argument: stage. see Stage API above
721 stage-1 p.i_valid >>in stage n.o_valid out>> stage+1
722 stage-1 p.o_ready <<out stage n.i_ready <<in stage+1
723 stage-1 p.i_data >>in stage n.o_data out>> stage+1
729 input data p.i_data is read (only), is processed and goes into an
730 intermediate result store [process()]. this is updated combinatorially.
732 in a non-stall condition, the intermediate result will go into the
733 output (update_output). however if ever there is a stall, it goes
734 into r_data instead [update_buffer()].
736 when the non-stall condition is released, r_data is the first
737 to be transferred to the output [flush_buffer()], and the stall
740 on the next cycle (as long as stall is not raised again) the
741 input may begin to be processed and transferred directly to output.
744 def elaborate(self
, platform
):
745 self
.m
= ControlBase
._elaborate
(self
, platform
)
747 result
= self
.stage
.ospec()
748 r_data
= self
.stage
.ospec()
750 # establish some combinatorial temporaries
751 o_n_validn
= Signal(reset_less
=True)
752 n_i_ready
= Signal(reset_less
=True, name
="n_i_rdy_data")
753 nir_por
= Signal(reset_less
=True)
754 nir_por_n
= Signal(reset_less
=True)
755 p_i_valid
= Signal(reset_less
=True)
756 nir_novn
= Signal(reset_less
=True)
757 nirn_novn
= Signal(reset_less
=True)
758 por_pivn
= Signal(reset_less
=True)
759 npnn
= Signal(reset_less
=True)
760 self
.m
.d
.comb
+= [p_i_valid
.eq(self
.p
.i_valid_test
),
761 o_n_validn
.eq(~self
.n
.o_valid
),
762 n_i_ready
.eq(self
.n
.i_ready_test
),
763 nir_por
.eq(n_i_ready
& self
.p
._o
_ready
),
764 nir_por_n
.eq(n_i_ready
& ~self
.p
._o
_ready
),
765 nir_novn
.eq(n_i_ready | o_n_validn
),
766 nirn_novn
.eq(~n_i_ready
& o_n_validn
),
767 npnn
.eq(nir_por | nirn_novn
),
768 por_pivn
.eq(self
.p
._o
_ready
& ~p_i_valid
)
771 # store result of processing in combinatorial temporary
772 self
.m
.d
.comb
+= eq(result
, self
.stage
.process(self
.p
.i_data
))
774 # if not in stall condition, update the temporary register
775 with self
.m
.If(self
.p
.o_ready
): # not stalled
776 self
.m
.d
.sync
+= eq(r_data
, result
) # update buffer
778 # data pass-through conditions
779 with self
.m
.If(npnn
):
780 o_data
= self
._postprocess
(result
)
781 self
.m
.d
.sync
+= [self
.n
.o_valid
.eq(p_i_valid
), # valid if p_valid
782 eq(self
.n
.o_data
, o_data
), # update output
784 # buffer flush conditions (NOTE: can override data passthru conditions)
785 with self
.m
.If(nir_por_n
): # not stalled
786 # Flush the [already processed] buffer to the output port.
787 o_data
= self
._postprocess
(r_data
)
788 self
.m
.d
.sync
+= [self
.n
.o_valid
.eq(1), # reg empty
789 eq(self
.n
.o_data
, o_data
), # flush buffer
791 # output ready conditions
792 self
.m
.d
.sync
+= self
.p
._o
_ready
.eq(nir_novn | por_pivn
)
797 class SimpleHandshake(ControlBase
):
798 """ simple handshake control. data and strobe signals travel in sync.
799 implements the protocol used by Wishbone and AXI4.
801 Argument: stage. see Stage API above
803 stage-1 p.i_valid >>in stage n.o_valid out>> stage+1
804 stage-1 p.o_ready <<out stage n.i_ready <<in stage+1
805 stage-1 p.i_data >>in stage n.o_data out>> stage+1
810 Inputs Temporary Output Data
811 ------- ---------- ----- ----
812 P P N N PiV& ~NiR& N P
819 0 0 1 0 0 0 0 1 process(i_data)
820 0 0 1 1 0 0 0 1 process(i_data)
824 0 1 1 0 0 0 0 1 process(i_data)
825 0 1 1 1 0 0 0 1 process(i_data)
829 1 0 1 0 0 0 0 1 process(i_data)
830 1 0 1 1 0 0 0 1 process(i_data)
832 1 1 0 0 1 0 1 0 process(i_data)
833 1 1 0 1 1 1 1 0 process(i_data)
834 1 1 1 0 1 0 1 1 process(i_data)
835 1 1 1 1 1 0 1 1 process(i_data)
839 def elaborate(self
, platform
):
840 self
.m
= m
= ControlBase
._elaborate
(self
, platform
)
843 result
= self
.stage
.ospec()
845 # establish some combinatorial temporaries
846 n_i_ready
= Signal(reset_less
=True, name
="n_i_rdy_data")
847 p_i_valid_p_o_ready
= Signal(reset_less
=True)
848 p_i_valid
= Signal(reset_less
=True)
849 m
.d
.comb
+= [p_i_valid
.eq(self
.p
.i_valid_test
),
850 n_i_ready
.eq(self
.n
.i_ready_test
),
851 p_i_valid_p_o_ready
.eq(p_i_valid
& self
.p
.o_ready
),
854 # store result of processing in combinatorial temporary
855 m
.d
.comb
+= eq(result
, self
.stage
.process(self
.p
.i_data
))
857 # previous valid and ready
858 with m
.If(p_i_valid_p_o_ready
):
859 o_data
= self
._postprocess
(result
)
860 m
.d
.sync
+= [r_busy
.eq(1), # output valid
861 eq(self
.n
.o_data
, o_data
), # update output
863 # previous invalid or not ready, however next is accepting
864 with m
.Elif(n_i_ready
):
865 o_data
= self
._postprocess
(result
)
866 m
.d
.sync
+= [eq(self
.n
.o_data
, o_data
)]
867 # TODO: could still send data here (if there was any)
868 #m.d.sync += self.n.o_valid.eq(0) # ...so set output invalid
869 m
.d
.sync
+= r_busy
.eq(0) # ...so set output invalid
871 m
.d
.comb
+= self
.n
.o_valid
.eq(r_busy
)
872 # if next is ready, so is previous
873 m
.d
.comb
+= self
.p
._o
_ready
.eq(n_i_ready
)
878 class UnbufferedPipeline(ControlBase
):
879 """ A simple pipeline stage with single-clock synchronisation
880 and two-way valid/ready synchronised signalling.
882 Note that a stall in one stage will result in the entire pipeline
885 Also that unlike BufferedHandshake, the valid/ready signalling does NOT
886 travel synchronously with the data: the valid/ready signalling
887 combines in a *combinatorial* fashion. Therefore, a long pipeline
888 chain will lengthen propagation delays.
890 Argument: stage. see Stage API, above
892 stage-1 p.i_valid >>in stage n.o_valid out>> stage+1
893 stage-1 p.o_ready <<out stage n.i_ready <<in stage+1
894 stage-1 p.i_data >>in stage n.o_data out>> stage+1
902 p.i_data : StageInput, shaped according to ispec
904 p.o_data : StageOutput, shaped according to ospec
906 r_data : input_shape according to ispec
907 A temporary (buffered) copy of a prior (valid) input.
908 This is HELD if the output is not ready. It is updated
910 result: output_shape according to ospec
911 The output of the combinatorial logic. it is updated
912 COMBINATORIALLY (no clock dependence).
916 Inputs Temp Output Data
938 1 1 0 0 0 1 1 process(i_data)
939 1 1 0 1 1 1 0 process(i_data)
940 1 1 1 0 0 1 1 process(i_data)
941 1 1 1 1 0 1 1 process(i_data)
944 Note: PoR is *NOT* involved in the above decision-making.
947 def elaborate(self
, platform
):
948 self
.m
= m
= ControlBase
._elaborate
(self
, platform
)
950 data_valid
= Signal() # is data valid or not
951 r_data
= self
.stage
.ospec() # output type
954 p_i_valid
= Signal(reset_less
=True)
955 pv
= Signal(reset_less
=True)
956 buf_full
= Signal(reset_less
=True)
957 m
.d
.comb
+= p_i_valid
.eq(self
.p
.i_valid_test
)
958 m
.d
.comb
+= pv
.eq(self
.p
.i_valid
& self
.p
.o_ready
)
959 m
.d
.comb
+= buf_full
.eq(~self
.n
.i_ready_test
& data_valid
)
961 m
.d
.comb
+= self
.n
.o_valid
.eq(data_valid
)
962 m
.d
.comb
+= self
.p
._o
_ready
.eq(~data_valid | self
.n
.i_ready_test
)
963 m
.d
.sync
+= data_valid
.eq(p_i_valid | buf_full
)
966 m
.d
.sync
+= eq(r_data
, self
.stage
.process(self
.p
.i_data
))
967 o_data
= self
._postprocess
(r_data
)
968 m
.d
.comb
+= eq(self
.n
.o_data
, o_data
)
972 class UnbufferedPipeline2(ControlBase
):
973 """ A simple pipeline stage with single-clock synchronisation
974 and two-way valid/ready synchronised signalling.
976 Note that a stall in one stage will result in the entire pipeline
979 Also that unlike BufferedHandshake, the valid/ready signalling does NOT
980 travel synchronously with the data: the valid/ready signalling
981 combines in a *combinatorial* fashion. Therefore, a long pipeline
982 chain will lengthen propagation delays.
984 Argument: stage. see Stage API, above
986 stage-1 p.i_valid >>in stage n.o_valid out>> stage+1
987 stage-1 p.o_ready <<out stage n.i_ready <<in stage+1
988 stage-1 p.i_data >>in stage n.o_data out>> stage+1
993 p.i_data : StageInput, shaped according to ispec
995 p.o_data : StageOutput, shaped according to ospec
997 buf : output_shape according to ospec
998 A temporary (buffered) copy of a valid output
999 This is HELD if the output is not ready. It is updated
1002 Inputs Temp Output Data
1004 P P N N ~NiR& N P (buf_full)
1009 0 0 0 0 0 0 1 process(i_data)
1010 0 0 0 1 1 1 0 reg (odata, unchanged)
1011 0 0 1 0 0 0 1 process(i_data)
1012 0 0 1 1 0 0 1 process(i_data)
1014 0 1 0 0 0 0 1 process(i_data)
1015 0 1 0 1 1 1 0 reg (odata, unchanged)
1016 0 1 1 0 0 0 1 process(i_data)
1017 0 1 1 1 0 0 1 process(i_data)
1019 1 0 0 0 0 1 1 process(i_data)
1020 1 0 0 1 1 1 0 reg (odata, unchanged)
1021 1 0 1 0 0 1 1 process(i_data)
1022 1 0 1 1 0 1 1 process(i_data)
1024 1 1 0 0 0 1 1 process(i_data)
1025 1 1 0 1 1 1 0 reg (odata, unchanged)
1026 1 1 1 0 0 1 1 process(i_data)
1027 1 1 1 1 0 1 1 process(i_data)
1030 Note: PoR is *NOT* involved in the above decision-making.
1033 def elaborate(self
, platform
):
1034 self
.m
= m
= ControlBase
._elaborate
(self
, platform
)
1036 buf_full
= Signal() # is data valid or not
1037 buf
= self
.stage
.ospec() # output type
1040 p_i_valid
= Signal(reset_less
=True)
1041 m
.d
.comb
+= p_i_valid
.eq(self
.p
.i_valid_test
)
1043 m
.d
.comb
+= self
.n
.o_valid
.eq(buf_full | p_i_valid
)
1044 m
.d
.comb
+= self
.p
._o
_ready
.eq(~buf_full
)
1045 m
.d
.sync
+= buf_full
.eq(~self
.n
.i_ready_test
& self
.n
.o_valid
)
1047 o_data
= Mux(buf_full
, buf
, self
.stage
.process(self
.p
.i_data
))
1048 o_data
= self
._postprocess
(o_data
)
1049 m
.d
.comb
+= eq(self
.n
.o_data
, o_data
)
1050 m
.d
.sync
+= eq(buf
, self
.n
.o_data
)
1055 class PassThroughStage(StageCls
):
1056 """ a pass-through stage which has its input data spec equal to its output,
1057 and "passes through" its data from input to output.
1059 def __init__(self
, iospecfn
):
1060 self
.iospecfn
= iospecfn
1061 def ispec(self
): return self
.iospecfn()
1062 def ospec(self
): return self
.iospecfn()
1063 def process(self
, i
): return i
1066 class PassThroughHandshake(ControlBase
):
1067 """ A control block that delays by one clock cycle.
1069 Inputs Temporary Output Data
1070 ------- ------------------ ----- ----
1071 P P N N PiV& PiV| NiR| pvr N P (pvr)
1072 i o i o PoR ~PoR ~NoV o o
1076 0 0 0 0 0 1 1 0 1 1 odata (unchanged)
1077 0 0 0 1 0 1 0 0 1 0 odata (unchanged)
1078 0 0 1 0 0 1 1 0 1 1 odata (unchanged)
1079 0 0 1 1 0 1 1 0 1 1 odata (unchanged)
1081 0 1 0 0 0 0 1 0 0 1 odata (unchanged)
1082 0 1 0 1 0 0 0 0 0 0 odata (unchanged)
1083 0 1 1 0 0 0 1 0 0 1 odata (unchanged)
1084 0 1 1 1 0 0 1 0 0 1 odata (unchanged)
1086 1 0 0 0 0 1 1 1 1 1 process(in)
1087 1 0 0 1 0 1 0 0 1 0 odata (unchanged)
1088 1 0 1 0 0 1 1 1 1 1 process(in)
1089 1 0 1 1 0 1 1 1 1 1 process(in)
1091 1 1 0 0 1 1 1 1 1 1 process(in)
1092 1 1 0 1 1 1 0 0 1 0 odata (unchanged)
1093 1 1 1 0 1 1 1 1 1 1 process(in)
1094 1 1 1 1 1 1 1 1 1 1 process(in)
1099 def elaborate(self
, platform
):
1100 self
.m
= m
= ControlBase
._elaborate
(self
, platform
)
1102 r_data
= self
.stage
.ospec() # output type
1105 p_i_valid
= Signal(reset_less
=True)
1106 pvr
= Signal(reset_less
=True)
1107 m
.d
.comb
+= p_i_valid
.eq(self
.p
.i_valid_test
)
1108 m
.d
.comb
+= pvr
.eq(p_i_valid
& self
.p
.o_ready
)
1110 m
.d
.comb
+= self
.p
.o_ready
.eq(~self
.n
.o_valid | self
.n
.i_ready_test
)
1111 m
.d
.sync
+= self
.n
.o_valid
.eq(p_i_valid | ~self
.p
.o_ready
)
1113 odata
= Mux(pvr
, self
.stage
.process(self
.p
.i_data
), r_data
)
1114 m
.d
.sync
+= eq(r_data
, odata
)
1115 r_data
= self
._postprocess
(r_data
)
1116 m
.d
.comb
+= eq(self
.n
.o_data
, r_data
)
1121 class RegisterPipeline(UnbufferedPipeline
):
1122 """ A pipeline stage that delays by one clock cycle, creating a
1123 sync'd latch out of o_data and o_valid as an indirect byproduct
1124 of using PassThroughStage
1126 def __init__(self
, iospecfn
):
1127 UnbufferedPipeline
.__init
__(self
, PassThroughStage(iospecfn
))
1130 class FIFOControl(ControlBase
):
1131 """ FIFO Control. Uses SyncFIFO to store data, coincidentally
1132 happens to have same valid/ready signalling as Stage API.
1134 i_data -> fifo.din -> FIFO -> fifo.dout -> o_data
1137 def __init__(self
, depth
, stage
, in_multi
=None, stage_ctl
=False,
1138 fwft
=True, buffered
=False, pipe
=False):
1141 * depth: number of entries in the FIFO
1142 * stage: data processing block
1143 * fwft : first word fall-thru mode (non-fwft introduces delay)
1144 * buffered: use buffered FIFO (introduces extra cycle delay)
1146 NOTE 1: FPGAs may have trouble with the defaults for SyncFIFO
1147 (fwft=True, buffered=False)
1149 NOTE 2: i_data *must* have a shape function. it can therefore
1150 be a Signal, or a Record, or a RecordObject.
1152 data is processed (and located) as follows:
1154 self.p self.stage temp fn temp fn temp fp self.n
1155 i_data->process()->result->cat->din.FIFO.dout->cat(o_data)
1157 yes, really: cat produces a Cat() which can be assigned to.
1158 this is how the FIFO gets de-catted without needing a de-cat
1162 assert not (fwft
and buffered
), "buffered cannot do fwft"
1166 self
.buffered
= buffered
1169 ControlBase
.__init
__(self
, stage
, in_multi
, stage_ctl
)
1171 def elaborate(self
, platform
):
1172 self
.m
= m
= ControlBase
._elaborate
(self
, platform
)
1174 # make a FIFO with a signal of equal width to the o_data.
1175 (fwidth
, _
) = self
.n
.o_data
.shape()
1177 fifo
= SyncFIFOBuffered(fwidth
, self
.fdepth
)
1179 fifo
= Queue(fwidth
, self
.fdepth
, fwft
=self
.fwft
, pipe
=self
.pipe
)
1180 m
.submodules
.fifo
= fifo
1182 # store result of processing in combinatorial temporary
1183 result
= self
.stage
.ospec()
1184 m
.d
.comb
+= eq(result
, self
.stage
.process(self
.p
.i_data
))
1186 # connect previous rdy/valid/data - do cat on i_data
1187 # NOTE: cannot do the PrevControl-looking trick because
1188 # of need to process the data. shaaaame....
1189 m
.d
.comb
+= [fifo
.we
.eq(self
.p
.i_valid_test
),
1190 self
.p
.o_ready
.eq(fifo
.writable
),
1191 eq(fifo
.din
, cat(result
)),
1194 # connect next rdy/valid/data - do cat on o_data
1195 connections
= [self
.n
.o_valid
.eq(fifo
.readable
),
1196 fifo
.re
.eq(self
.n
.i_ready_test
),
1198 if self
.fwft
or self
.buffered
:
1199 m
.d
.comb
+= connections
1201 m
.d
.sync
+= connections
# unbuffered fwft mode needs sync
1202 o_data
= cat(self
.n
.o_data
).eq(fifo
.dout
)
1203 o_data
= self
._postprocess
(o_data
)
1210 class UnbufferedPipeline(FIFOControl
):
1211 def __init__(self
, stage
, in_multi
=None, stage_ctl
=False):
1212 FIFOControl
.__init
__(self
, 1, stage
, in_multi
, stage_ctl
,
1213 fwft
=True, pipe
=False)
1215 # aka "BreakReadyStage" XXX had to set fwft=True to get it to work
1216 class PassThroughHandshake(FIFOControl
):
1217 def __init__(self
, stage
, in_multi
=None, stage_ctl
=False):
1218 FIFOControl
.__init
__(self
, 1, stage
, in_multi
, stage_ctl
,
1219 fwft
=True, pipe
=True)
1221 # this is *probably* BufferedHandshake, although test #997 now succeeds.
1222 class BufferedHandshake(FIFOControl
):
1223 def __init__(self
, stage
, in_multi
=None, stage_ctl
=False):
1224 FIFOControl
.__init
__(self
, 2, stage
, in_multi
, stage_ctl
,
1225 fwft
=True, pipe
=False)
1229 # this is *probably* SimpleHandshake (note: memory cell size=0)
1230 class SimpleHandshake(FIFOControl):
1231 def __init__(self, stage, in_multi=None, stage_ctl=False):
1232 FIFOControl.__init__(self, 0, stage, in_multi, stage_ctl,
1233 fwft=True, pipe=False)