Seeed WioTerminal support #1124

sago35 · 2020-05-20T12:17:54Z

I added Seeed WioTerminal support to TinyGo.
https://wiki.seeedstudio.com/Wio-Terminal-Getting-Started/

ATSAMD5x DataSheet: https://files.seeedstudio.com/wiki/Wio-Terminal/res/ATSAMD51.pdf
PinOut: https://files.seeedstudio.com/wiki/Wio-Terminal/img/WioT-Pinout.jpg
Schematics: https://files.seeedstudio.com/wiki/Wio-Terminal/res/Wio-Terminal-Schematics.pdf

Feature
- ATSAMD51P19 (120MHz / 128 Pins / 512KB / 192KB)
- 4MB External Flash
- Realtek RTL8720DN (Dual Band 2.4GHz / 5GHz WiFi, BLE 5.0)
- USB OTG
- LCD, WIFI, BT, IMU, Microphone, Speaker, microSD Card, Light Sensor, 5-Way Switch, Infrared Emitter (IR 940nm), Crypto-authentication

Current Status:

status
- 4MB External Flash
  - tinygo.org/x/drivers/examples/flash/console/qspi
- WiFi / BT
  - not for this PR
- USB OTG
  - not for this PR
- src/examples/adc
- src/examples/blinky1
- src/examples/button
- src/examples/echo
- src/examples/pininterrupt
- src/examples/pwm
- src/examples/serial
- src/examples/systick
- LCD
  - tinygo.org/x/drivers/examples/ili9341/basic
  - tinygo.org/x/drivers/examples/ili9341/pyportal_boing
  - tinygo.org/x/drivers/examples/ili9341/scroll
- IMU
  - tinygo.org/x/drivers/examples/lis3dh
- Microphone
  - MIC input (PC30 port ADC)
- Speaker
  - tinygo.org/x/drivers/examples/buzzer
- microSD Card
  - not for this PR
    - SPI sdcard driver drivers#155
- Light Sensor
  - PD01 port ADC
- 5-Way Switch and 3 x buttons
  - PD20, PD12, PD09, PD08, PD10
  - PC26, PC27, PC28
- Infrared Emmitter (IR 940nm)
  - PB31
- ATECC608 : I2C CrypitoAuthentication Device
  - ~~not for this PR~~ It wasn't implemented on the board.
TODO
- pinPadMapping[32] to pinPadMpping[64] for SERCOMx
- configure UART1
- interrupts pins
- arm: make FPU configuration consistent arm: make FPU configuration consistent #1105
- rebase after sam: fix register access for interrupts pins in samd51 #1141 has been merged.

aykevl · 2020-05-20T21:46:39Z

src/machine/board_wioterminal.go

+	D0  = PB17 // UART0 RX/PWM available
+	D1  = PB16 // UART0 TX/PWM available
+	D4  = PA14 // PWM available
+	D5  = PA16 // PWM available
+	D6  = PA18 // PWM available
+	D8  = PB03 // built-in neopixel
+	D9  = PA19 // PWM available
+	D10 = PA20 // can be used for PWM or UART1 TX
+	D11 = PA21 // can be used for PWM or UART1 RX
+	D12 = PA22 // PWM available
+	D13 = PA23 // PWM available
+	D21 = PA13 // PWM available
+	D22 = PA12 // PWM available
+	D23 = PB22 // PWM available
+	D24 = PB23 // PWM available
+	D25 = PA17 // PWM available


I could not easily find these in the schematics. Are these pins available on the WioTerminal?

I haven't made it to the correct value yet.

The schematic and pinouts can be found below.
https://files.seeedstudio.com/wiki/Wio-Terminal/res/Wio-Terminal-Schematics.pdf
https://files.seeedstudio.com/wiki/Wio-Terminal/img/WioT-Pinout.jpg

Is the definition of Dxx correct to set D0 to D8 in the Pinout Diagram?
Is it right to fit the Arduino?

Eventually, I'm going to adjust to the following.

https://github.com/Seeed-Studio/ArduinoCore-samd/blob/master/variants/wio_terminal/variant.h
https://github.com/Seeed-Studio/ArduinoCore-samd/blob/master/variants/wio_terminal/variant.cpp

For example.

#define D0 (0ul) const PinDescription g_APinDescription[0] = // 0..8 - Raspberry pi Digital & AD pins {PORTB, 8, PIO_ANALOG, (PIN_ATTR_ANALOG | PIN_ATTR_PWM_E), ADC_Channel2, TC4_CH0, TC4_CH0, EXTERNAL_INT_8}, // 0/RPI_D0/RPI_A0 }

I'm going to set it as follows.

D0 = PB08

aykevl · 2020-05-20T21:48:30Z

src/machine/board_wioterminal.go

+	SPI0_MISO_PIN = PB00 // MISO: SERCOM5/PAD[2]
+)
+
+// SPI on the Feather M4.


It looks like you copied this from the Feather M4?

Are you sure that all the definitions in this file are correct for the WioTerminal?

I'm going to try to get the definition right from now on.
I plan to adjust to the following.

https://github.com/Seeed-Studio/ArduinoCore-samd/blob/master/variants/wio_terminal/variant.h.
https://github.com/Seeed-Studio/ArduinoCore-samd/blob/master/variants/wio_terminal/variant.cpp

We follow the board markings in TinyGo. If there are no markings on the bard, using the same pin names as Arduino or MicroPython use is fine.

aykevl · 2020-05-20T21:56:08Z

src/machine/machine_atsamd51.go

 }

+// Transfer2 writes/reads a single uint16 using the SPI interface.
+func (spi SPI) Transfer2(w uint16) (uint16, error) {


I see you have extended the SPI peripheral with 16-bit and 32-bit methods.

Does this provide an improvement in speed?

Is this really necessary for WioTerminal support?

I'm asking this, because eventually I would like to switch to DMA and interrupt based SPI transfers. That would make it unnecessary to add Transfer2 and Transfer4 methods, as all the traffic is handled by DMA (which is the fastest possible transfer, except maybe for very short transfers).

Does this provide an improvement in speed?
Is this really necessary for WioTerminal support?

Yes.

In the demo of pyportal_boing, ILI9341 is too slow at about 10 fps when using Transfer().
Using Transfer2() makes the ILI9341 32 fps and Transfer4() makes it 36 fps.

The processing time excluding DrawRGBBitmap() was 12.5ms, so the time it takes for one DrawRGBBitmap() is as follows

Transfer : 100 ms

Transfer2 : 31.25 ms

Transfer3 : 27.8 ms

Also, I think that DMA should be enabled at some point.
But without a DMA, it's too slow to use Trasnfer4.
So I'd like to enable Transfer4 once, if possible.

You may be able to use a trick like here to improve the speed:
#562
What it does, is that it makes use of double buffering support in the hardware. This makes it possible to continuously stream out data. Maybe the SAM D5x also supports double buffering in the DATA register.

By using INTEFLAG.DRE, I was able to improve performance without using 32bit-extensions.
ILI9341's pyportal_boing is now 43 fps instead of 36 fps.

ILI9341's pyportal_boing is now 43 fps instead of 36 fps.

I think you possibly meant "36 fps instead of 43 fps" 😸

The double-buffer-like processing of DRE checks seems to be faster than the 4-byte writes ( the 32B-extensions).

board cpu clock driver fps

pyportal samd51j20 120Mhz parallel 51 fps

wioterminal samd51p19 120Mhz spi (first commit) 10 fps

wioterminal samd51p19 120Mhz spi with 32B extensions 36 fps

wioterminal samd51p19 120Mhz spi with DRE check and without 32B extensions 43 fps

That's a nice improvement! And it should benefit anyone using SPI on the atsamd51.
Can you make a separate PR with those changes?

Sorry, now I see you already made the change in tinygo-org/drivers#153.

Maybe the case of just sending (spi.Tx(w, nil)) can be done faster in the same way.
I put that code in ILI9341, but I think it can be done by moving it from spi.go to machine_atsamd51.go (maybe atsamd21 can do it the same way).

sago35 · 2020-05-21T13:22:35Z

What should RESET_MAGIC_VALUE be?

deadprogram · 2020-05-22T08:26:56Z

The address RESET_MAGIC_VALUE is used by the bootloader to trigger the mode to receive new code. Do you happen to know where the source code for the uf2 bootloader used by the WioTerminal?

https://github.com/adafruit/uf2-samdx1/blob/master/inc/uf2.h#L249 is where the original value that we are using comes from.

deadprogram · 2020-05-22T08:27:40Z

@sago35 can you please rebase against the dev branch? That should allow all of the tests to pass now.

sago35 · 2020-05-23T07:30:21Z

The address RESET_MAGIC_VALUE is used by the bootloader to trigger the mode to receive new code. Do you happen to know where the source code for the uf2 bootloader used by the WioTerminal?

I think it's probably this.

https://github.com/Seeed-Studio/ArduinoCore-samd/blob/master/cores/arduino/Reset.cpp#L59

This meant that no changes were needed.
It can be rewritten in tinygo flash.
If I change the value, you won't be able to do the tinygo flash.

deadprogram · 2020-05-23T09:39:23Z

Since it is the same value for the WioTerminal, no changes needed, you can just use the same value.

sago35 · 2020-05-24T09:44:08Z

@aykevl
When I try to configure UART1, the settings in machine_atsamd51.go and wioterminal are different.
I think I'll move this around to board_*.go, is that okay?

Maybe I'll change the file below.

board_feather-m4.go
- tx: PB16, rx: PB17
- https://learn.adafruit.com/adafruit-feather-m4-express-atsamd51/downloads
board_itsybitsy-m4.go
- tx: PA17, rx: PA16
- https://learn.adafruit.com/introducing-adafruit-itsybitsy-m4/downloads
board_metro-m4-airlift.go
- tx: PA22, rx: PA23
- https://learn.adafruit.com/adafruit-metro-m4-express-airlift-wifi/downloads
board_pybadge.go
- tx: PB16, rx: PB17
- https://learn.adafruit.com/adafruit-pybadge/downloads
board_pyportal.go
- ~~tx: PB16, rx: PB17~~ tx:PB12, rx: PB13
- https://learn.adafruit.com/adafruit-pyportal/downloads

sago35 · 2020-05-27T22:49:11Z

@deadprogram
I think that the remaining TODO is how to do UART1.
If no changes are needed, the PR is complete.

And apart from this PR, I'm working on the following

SPI SDcard driver
- SPI sdcard driver drivers#155

I'm hoping to make the following as well, but I think it will take some time.

RTL8720 wifi driver

sago35 · 2020-05-27T23:10:19Z

I tried to write the following, but machine.UART1.handleInterrupt is a private function, so I got an error.
After all, it may be good to move the setting of UART1 to board_xxx.go.

error message: uart\main.go:37:75: machine.UART1.handleInterrupt undefined (type machine.UART has no field or method handleInterrupt)

func main() {
	uart1.Configure(machine.UARTConfig{TX: tx1, RX: rx1})
	machine.UART1 = machine.UART{
		Buffer: machine.NewRingBuffer(),
		Bus:    sam.SERCOM2_USART_INT,
		SERCOM: 2,
	}
	machine.UART1.Interrupt = interrupt.New(sam.IRQ_SERCOM2_2, machine.UART1.handleInterrupt)

deadprogram · 2020-05-28T08:42:44Z

Hi @sago35 the interrupt handler for the UART would normally be in the machine package for that board, not in a user program. Hence why it is not exposed publicly.

Is there a reason to not add the UART1 implementation for the WioTerminal this way?

sago35 · 2020-05-28T10:38:58Z

Hi @deadprogram
(I'm not good at English, so I'm relying on translation tools)

#1124 (comment)
the interrupt handler for the UART would normally be in the machine package for that board, not in a user program. Hence why it is not exposed publicly.
Is there a reason to not add the UART1 implementation for the WioTerminal this way?

This implementation tried to see if it could override the UART1 settings without changing machine_atsamd51.go.
However, as a result, I still think it is necessary to change board_xxx.go because the setting cannot be overwritten.

I also think that UART is good with machine package at this point.
The problem that I am thinking about now is that UART1 of machine_atsamd51.go is set for itsybitsy-m4 only.

My suggestion is to move the setting of UART1 (and UART2 if necessary) to board_itsybitsy-m4.go or board_feather-m4.go etc. instead of machine_atsamd51.go.
Do you think this proposal is a good one?

For example, the setting of UART1 needs to be different for each board as follows.

itsybitsy-m4

	UART1 = UART{
		Buffer: NewRingBuffer(),
		Bus:    sam.SERCOM3_USART_INT,
		SERCOM: 3,
	}

	UART1.Interrupt = interrupt.New(sam.IRQ_SERCOM3_2, UART1.handleInterrupt)

wioterminal

	UART1 = UART{
		Buffer: NewRingBuffer(),
		Bus:    sam.SERCOM2_USART_INT,
		SERCOM: 2,
	}

	UART1.Interrupt = interrupt.New(sam.IRQ_SERCOM2_2, UART1.handleInterrupt)

deadprogram · 2020-05-28T11:00:07Z

I think you have the correct idea. Please look at a similar SAMD21 implementation:

https://github.com/tinygo-org/tinygo/blob/master/src/machine/board_feather-m0.go#L57-L68

sago35 · 2020-05-28T11:14:46Z

OK, make SAMD51's UART in the same way.

sago35 · 2020-05-28T13:17:18Z

UART settings have been corrected.
I kept as much of the original settings as I could, but fixed the obviously different parts.

I have confirmed that feather-m4 and wioterminal work correctly.
The PyPortal is connected to ESP32 and is not sure if it can be configured correctly.
I don't have istybitsy-m4 and metro-m4-airlift and pybadge, so I can't test them.

sago35 · 2020-05-28T13:21:08Z

I think the only remaining task is to set up the interrupts pins.
If #1139 is merged immediately, this PR will address the interrupts pins.
If it takes a while to check #1139, set the SAMD51P19 interrupts pins in #1139.

sago35 · 2020-05-29T22:38:56Z

The following are now supported.

interrupts pins
arm: make FPU configuration consistent

sago35 · 2020-05-29T22:40:19Z

The only thing left to do is to

rebase after sam: fix register access for interrupts pins in samd51 #1141 has been merged

My source code changes are now complete.
I will force-push after #1141 is merged, but I don't think the source code will change.
Please review.
@deadprogram @aykevl

sago35 · 2020-05-31T23:27:17Z

This PR is now complete.
@deadprogram

deadprogram · 2020-06-03T10:46:07Z

OK @sago35 thanks again for working on this. I have a few ideas on how to split this PR into more easily reviewable/mergable pieces:

Make the "Standardize SAMD51 UART Settings" dcbfb3d into a separate PR
Make "Extend SAMD51 pinPadMapping" 7b767d7 a separate PR
Make "Add SAMD51 ADC settings" d47bc08 into a separate PR

Once those are merged, probably the rest of the commits in this current PR can probably then just be squashed into a single PR that is "Add WioTerminal support"

What do you think?

sago35 · 2020-06-03T12:23:21Z

I will try to split the PR.

sago35 · 2020-06-03T13:01:55Z

Hi @deadprogram
I'm going to break it down into three PRs below.
git rebase worked well.
2 will be created after #1152 has been merged.

add support for SAMD51 with 128 pins Add support for SAMD51 with 128 pins #1152
Standardize SAMD51 UART Settings
see WioTerminal support

deadprogram · 2020-06-05T11:30:43Z

Now that #1155 has been merged, this branch needs rebasing to resolve merge conflicts. Thanks!

Seeed-Studio/ArduinoCore-samd@b7a8798 Seeed-Studio/ArduinoCore-samd@6d8c226

sago35 · 2020-06-05T12:44:07Z

Hi @deadprogram
Rebase is done.
The CI test was OK.
Please review.

targets/wioterminal.json

deadprogram · 2020-06-06T09:59:42Z

Thanks for all of the effort and changes to add this board @sago35 great work!

Now squash/merging. Thank you!

sago35 mentioned this pull request May 20, 2020

ILI9341 TFT driver (SPI) tinygo-org/drivers#153

Merged

aykevl reviewed May 20, 2020

View reviewed changes

sago35 force-pushed the wioterminal branch from 658e08b to b352450 Compare May 21, 2020 11:10

sago35 force-pushed the wioterminal branch from 7fbd33b to 205cb70 Compare May 22, 2020 08:48

sago35 force-pushed the wioterminal branch from 205cb70 to 402be87 Compare May 24, 2020 02:37

sago35 requested review from aykevl and deadprogram May 27, 2020 22:49

sago35 mentioned this pull request May 28, 2020

sam: add support for pin change interrupts (samd5x) #1139

Merged

sago35 force-pushed the wioterminal branch from 6ea6137 to 7ecb963 Compare May 28, 2020 11:13

sago35 force-pushed the wioterminal branch from 3c3502f to b5dd618 Compare May 29, 2020 22:34

sago35 force-pushed the wioterminal branch from b5dd618 to 6f7757f Compare May 30, 2020 22:07

sago35 changed the title ~~[WIP] Seeed WioTerminal support~~ Seeed WioTerminal support May 31, 2020

sago35 mentioned this pull request Jun 3, 2020

Add support for SAMD51 with 128 pins #1152

Merged

sago35 mentioned this pull request Jun 3, 2020

Standardize SAMD51 UART settings #1155

Merged

sago35 added 18 commits June 5, 2020 21:08

Copy feather-m4 to wioterminal

dac6f1b

Add support wioterminal

4d18470

Add support for buttons

63269b0

Fix USBCDC vars

e08a28d

Add definition of adc pins

147f6e5

OUTPUT_CTR and I2C0

d181430

Add suport SPI0 and SPI3

34fbbc7

Update port settings

c1f45ce

Update pin names

dbdc4bf

Fix UART settings

a31d475

make FPU configuration consistent

8b96ae4

Add test of pin change interrupts

6f9b0cf

Fix pin configuration error

7870897

Seeed-Studio/ArduinoCore-samd@b7a8798 Seeed-Studio/ArduinoCore-samd@6d8c226

Add comment about RTL8720

f61ab19

Add support for Seeed Wio Terminal board

5811674

Add smoketest

0c7bfff

Add comment about RTL8720

3dabf4c

machine: support wioterminal boards on play.tinygo.org

a2ce365

sago35 force-pushed the wioterminal branch from 68ccfc3 to a2ce365 Compare June 5, 2020 12:16

deadprogram reviewed Jun 5, 2020

View reviewed changes

targets/wioterminal.json Show resolved Hide resolved

deadprogram merged commit c5a8967 into tinygo-org:dev Jun 6, 2020

sago35 mentioned this pull request Jun 17, 2020

Seeed XIAO support #1170

Merged

6 tasks

niaow added this to the v0.14 milestone Jun 27, 2020

board	cpu	clock	driver	fps
pyportal	samd51j20	120Mhz	parallel	51 fps
wioterminal	samd51p19	120Mhz	spi (first commit)	10 fps
wioterminal	samd51p19	120Mhz	spi with 32B extensions	36 fps
wioterminal	samd51p19	120Mhz	spi with DRE check and without 32B extensions	43 fps

Seeed WioTerminal support #1124

Seeed WioTerminal support #1124

Uh oh!

Conversation

sago35 commented May 20, 2020 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Uh oh!

Choose a reason for hiding this comment

Uh oh!

sago35 May 20, 2020 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

aykevl May 20, 2020 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

sago35 May 20, 2020 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

sago35 commented May 21, 2020

Uh oh!

deadprogram commented May 22, 2020

Uh oh!

deadprogram commented May 22, 2020

Uh oh!

sago35 commented May 23, 2020 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Uh oh!

deadprogram commented May 23, 2020

Uh oh!

sago35 commented May 24, 2020 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Uh oh!

sago35 commented May 27, 2020

Uh oh!

sago35 commented May 27, 2020

Uh oh!

deadprogram commented May 28, 2020

Uh oh!

sago35 commented May 28, 2020 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Uh oh!

deadprogram commented May 28, 2020

Uh oh!

sago35 commented May 28, 2020

Uh oh!

sago35 commented May 28, 2020

Uh oh!

sago35 commented May 28, 2020

Uh oh!

sago35 commented May 29, 2020

Uh oh!

sago35 commented May 29, 2020

Uh oh!

sago35 commented May 31, 2020

Uh oh!

deadprogram commented Jun 3, 2020

Uh oh!

sago35 commented May 20, 2020 •

edited

Loading

sago35 May 20, 2020 •

edited

Loading

aykevl May 20, 2020 •

edited

Loading

sago35 May 20, 2020 •

edited

Loading

sago35 commented May 23, 2020 •

edited

Loading

sago35 commented May 24, 2020 •

edited

Loading

sago35 commented May 28, 2020 •

edited

Loading

sago35 commented Jun 3, 2020 •

edited

Loading