1 \documentclass[slidestop
]{beamer
}
2 \usepackage{beamerthemesplit
}
6 \title{Pin Multiplexer
}
8 \author{Luke Kenneth Casson Leighton
}
15 \huge{Pin Multiplexer
}\\
17 \Large{Auto-generating documentation, code \\
18 and resources for a Pinmux
}\\
20 \Large{Saving time and money for SoC / EC designers\\
21 in the RISC-V Ecosystem and beyond
}\\
23 \Large{[proposed for
] Chennai
9th RISC-V Workshop
}\\
30 \frame{\frametitle{Credits and Acknowledgements
}
33 \item TODO
\vspace{10pt
}
38 \frame{\frametitle{Glossary
}
41 \item GPIO: general-purpose reconfigureable I/O (Input/Output).
42 \item Pin: an I/O pad. May be driven (input) or may drive (output).
43 \item FN: term for a single-wire "function", such as UART
\_TX,
44 I2C
\_SDA, SDMMC
\_D0 etc. may be an input, output or both
45 (bi-directional case: two wires are
{\bf always
} allocated, one
46 for input to the function and one for output from the function).
47 \item Bus: a group of bi-directional functions (SDMMC D0 to D3)
48 where the direction is ganged and
{\bf under the Bus's control
}
49 \item Input Priority Muxer: a multiplexer with N selector
50 wires and N associated inputs. The lowest (highest?) indexed
51 "selector" enabled results in its
52 input being routed to the output.
53 \item Output Muxer: a many-to-one "redirector" where any one
54 output "routed" to the input, based on a selector "address".
59 \frame{\frametitle{Why, How and What is a Pinmux?
}
62 \item Why? To save cost, increase yield, and to target multiple
63 markets with the same design, thereby increasing uptake
64 and consequently taking advantage of volume pricing.
\vspace{4pt
}
66 Summary: it's all about making more money!
\vspace{4pt
}
67 \item How? By multiplexing many more functions (
100 to
1,
200) than there
68 are actual available pins (
48 to
500), the required chip package
69 is far less costly and the chip more desirable
\vspace{4pt
}
70 \item What? A many-to-many dynamically-configureable router of
71 I/O functions to I/O pins
\vspace{4pt
}
72 \item \bf{Note: actual muxing is deceptively simple, but like
73 a DRAM cell it's actually about the ancillaries / extras
}
78 \frame{\frametitle{Associated Extras
}
81 \item Design Specification (what markets to target)
82 \item Scenario analysis (
{\bf whether
} the chip will fit "markets")
83 \item Documentation: Summary sheet, Technical Reference Manual.
85 \item Control Interface (AXI4 / Wishbone / TileLink / other)
87 \item Linux kernel drivers, DTB, libopencm3, Arduino libraries etc.
91 \item Shakti M-Class has
160 pins with a
99.5\% full
4-way mux
92 \item Almost
640-way routing,
6 "scenarios" (
7th TBD),
93 100+ page Manual needed,
94 \bf{17,
500 lines of auto-generated code
}
99 \frame{\frametitle{Example:
7 banks,
4-way mux,
160 pins
}
101 \includegraphics[height=
1.7in
]{example_pinmux.jpg
}
104 \item { \bf 17,
500 lines of auto-generated HDL (and climbing)
}
105 \item { \bf 12,
500 lines of auto-generated Summary/Analysis
}
106 \item Technical Reference Manual expected to be
100+ pages
115 ALL of these
\vspace{20pt
}\\
116 can be
\vspace{20pt
}\\
117 auto-generated
\vspace{30pt
}
120 (from the Design Specification, after Scenario Analysis)
125 \frame{\frametitle{Reduce workload, reduce duplication, reduce risk and cost
}
128 \item Auto-generate everything: documentation, code, libraries etc.
130 \item Standardise: similar to PLIC, propose GPIO and Pinmux\\
131 saves engineering effort, design effort and much more
133 \item Standardise format of configuration registers:
134 saves code duplication effort (multiple software environments)
136 \item Add support for multiple code formats: Chisel3 (SiFive IOF),
137 BSV (Bluespec), Verilog, VHDL, MyHDL.
139 \item Multiple auto-generated code-formats permits cross-validation:\\
140 auto-generated test suite in one HDL can validate a muxer
141 generated for a different target HDL.
147 \frame{\frametitle{Design Spec and Scenario Analysis
}
150 \item Analyse the target markets that the chip will sell in\\
151 (multiple markets increases sales volume, reduces chip cost)
153 \item Create a formal (python-based) specification for the pinmux
155 \item Add scenarios then check that they meet the requirements\\
156 { \bf (before spending money on hardware engineers!)
}
158 \item Scenarios represent target markets: ICs to be connected\\
159 (GPS, NAND IC, WIFI etc. May require draft schematics
160 drawn up, or client-supplied schematics analysed).
162 \item Analyse the scenarios: if pins are missing, alter and repeat.\\
164 \item Eventually the pinmux meets all requirements...\\
165 { \bf without spending USD \$
5-
50m to find out it doesn't!
}
171 \frame{\frametitle{Muxer cases to handle (One/Many to One/Many) etc.
}
174 \item One FN outputs to Many Pins: no problem\\
175 (weird configuration by end-user, but no damage to ASIC)
176 \item One Pin to Many FN inputs: no problem\\
177 (weird configuration by end-user, but no damage to ASIC)
178 \item Many Pins to One FN input:
{\bf Priority Mux needed
}\\
179 No priority mux: Pin1 = HI, Pin0 = LO, ASIC is damaged
180 \item Many FN outputs simultaneously to one Pin:
{\bf does not occur
}\\
181 (not desirable and not possible, as part of the pinmux design)
182 \item Some FNs (I2C
\_SDA, SD
\_D0.
.3) are I/O Buses\\
183 Bi-directional control of the Pin must be handed to the
185 \item Nice to have: Bus sets pintype, signal strength etc.\\
186 e.g. selecting SD/MMC doesn't need manual pin-config.\\
187 \bf{caveat: get that wrong and the ASIC can't be sold
}
192 \frame{\frametitle{Pin Configuration, input and output
}
194 In/out:
{\bf Note: these all require multiplexing
}
196 \item Output-Enable (aka Input disable): switches pad to In or Out
197 \item Output (actually an input wire controlling pin's level, HI/LO)
198 \item Input (actually an output wire set based on pin's driven level)
200 Characteristics:
{\bf Note: these do not require multiplexing
}
202 \item Output current level:
10mA /
20mA /
30mA /
40mA
203 \item Input hysteresis: low / middle / high. Stops signal noise
204 \item Pin characteristics: CMOS Push-Push / Open-Drain
205 \item Pull-up enable: built-in
10k (
50k?) resistor
206 \item Pull-down enable: built-in
10k (
50k?) resistor
207 \item Muxing and IRQ Edge-detection not part of the I/O pin
208 \item Other? (impedance? not normally part of commercial pinmux)
213 \frame{\frametitle{Standard GPIO
4-way in/out Mux and I/O pad
}
215 \includegraphics[height=
2.5in
]{../shakti/m_class/mygpiomux.jpg
}\\
216 {\bf 4-in,
4-out, pullup/down, hysteresis, edge-detection (EINT)
}
220 \frame{\frametitle{Separating Pin Configuration, input and output
}
223 \item Standard Mux design
{\bf cannot deal with many-to-one inputs
}\\
224 (SiFive IOF source code from Freedom U310 cannot, either)
226 \item I/O pad configuration conflated with In-Muxer conflated with
227 Out-Muxer conflated with GPIO conflated with EINT.
230 {\bf IMPORTANT to separate all of these out:
233 \item EINTs to be totally separate FNs. managed by RISC-V PLIC\\
234 (If every GPIO was an EINT it would mean
100+ IRQs)
236 \item GPIO In/Out/Direction treated just like any other FN\\
237 (but happen to have AXI4 - or other - memory-mapping)
239 \item Pad configuration separated and given one-to-one Registers\\
240 (SRAMs set by AXI4 to control mux, pullup, current etc.)
244 \frame{\frametitle{Register-to-pad "control" settings
}
246 \includegraphics[height=
2.5in
]{reg_gpio_cap_ctrl.jpg
}\\
247 {\bf pullup/down, hysteresis, current, edge-detection
}
252 \frame{\frametitle{GPIO (only): Simplified I/O pad Diagram (FN only)
}
254 \includegraphics[height=
2.5in
]{reg_gpio_pinblock.jpg
}\\
255 {\bf 3 wires: IN, OUT, OUTEN (also = !INEN)
}
260 \frame{\frametitle{Output Muxer (very simple)
}
262 \includegraphics[height=
1.1in
]{reg_gpio_out_mux.jpg
}\\
263 {\bf Ouput Muxer using
2-bit address selection
}\\
266 \item Very straightforward (deceptively so, like SRAM cells)
267 \item Used in both OUT routing and Direction-control routing\\
268 (same address for each, connected to same FNs)
269 \item More complex pinmux will have
3-bit addressing (
8 FNs)\\
270 (Note: not all outputs will be connected, depends on pinmux)
275 \frame{\frametitle{In/Out muxing, direction control: GPIO just a FN
}
277 \includegraphics[height=
2.5in
]{reg_gpio_fn_ctrl.jpg
}\\
278 {\bf Note: function can control I/O direction (bus)
}
283 \frame{\frametitle{Direction Control: Function not bi-directional (bus)
}
285 \includegraphics[height=
2.5in
]{reg_gpio_fn_ctrl2.jpg
}\\
286 Note: Function
{\bf does not
} control I/O direction
291 \frame{\frametitle{Output (and OUTEN) Wiring.
2 pins,
2 GPIO,
2 Fns
}
293 \includegraphics[height=
2.5in
]{reg_gpio_out_wiring.jpg
}\\
294 {\bf Reg0 for Pin0, Reg1 for Pin1, Output and OUTEN same mux
}
299 \frame{\frametitle{Input Selection and Priority Muxing
}
301 \includegraphics[height=
0.75in
]{reg_gpio_comparator.jpg
}\\
302 {\bf Muxer enables input selection
}\\
304 \includegraphics[height=
1.25in
]{reg_gpio_in_prioritymux.jpg
}\\
305 {\bf However multiple inputs must be prioritised
}
310 \frame{\frametitle{Input Priority-Mux Wiring: very different from Out-Mux
}
312 \includegraphics[height=
2.5in
]{reg_gpio_in_wiring.jpg
}\\
313 {\bf Pin Mux selection vals NOT same as FN selection vals
}
318 \frame{\frametitle{Input Priority-Mux Wiring
}
321 \item In-Muxer selection number (
0,
1,
2,
3) obviously has to match
322 with Out-Muxer order (otherwise a bi-directional FN
323 needs different Mux-register settings for
324 selecting either IN or OUT)
326 \item Priority-mux selection values do not actually matter,
327 and have nothing to do with the actual Muxer settings.
329 \item GPIO FN's input muxer is nothing more than an AND gate\\
330 (you never route more than one pin to one GPIO)
332 \item Any other FN with only
1:
1 In also an AND gate \\
333 (this just always happens to be true for GPIO)
335 \item Not all FNs have input capability: clearly they will not
336 be included in the In-Muxing.
341 \frame{\frametitle{Summary
}
344 \item Actual muxing, like SRAM cells, is deceptively simple
346 \item Actual pinmuxes are enormous: auto-generation essential
348 \item HDLs completely unsuited to auto-generation task\\
349 (TRM, docs):
{ \bf Modern OO language features needed
}
351 \item Scenario Analysis / Verification and auto-generation of
352 different HDLs far easier in a Modern OO language
354 \item Standardisation for RISC-V saves implementors from huge
355 duplication cost (HDL, firmware, docs, maintenance)
357 \item { \bf Ultimately it's about saving money and reducing risk
}
364 {\Huge The end
\vspace{20pt
}\\
365 Thank you
\vspace{20pt
}\\
366 Questions?
\vspace{20pt
}
371 \item http://libre-riscv.org/shakti/m
\_class/pinmux/