Implement SPI driver on ATmega2560 #4045

danielamkaer · 2015-10-05T07:54:18Z

No description provided.

jnohlgard · 2015-10-05T08:28:03Z

@danielamkaer Please add a prefix to your commit message, like "cpu/atmega2560: Add SPI driver implementation", and also squash your whitespace fixes into the main commit.

I don't have the hardware available right now so I can't test. @N8Fear or @LudwigOrtmann, do you have any SPI devices that you can wire up to the Arduino?

thomaseichinger · 2015-10-05T08:35:33Z

boards/arduino-mega2560/include/periph_conf.h

 #define SPI_1_EN            0

+// SPI 0 Registers


Please use C comment style. // -> /* */

thomaseichinger · 2015-10-05T08:42:42Z

cpu/atmega2560/periph/spi.c

+#if SPI_0_EN
+        case SPI_0:
+            SPI_0_DDR |= (1<<SPI_0_MOSI)|(1<<SPI_0_SCK);
+            SPI_0_DDR &= ~((1<<SPI_0_MISO));


These two lines could be moved to spi_conf_pins?

The configuration of the pins depends on whether it is master/slave, no parameter for spi_conf_pins to configure that

Ah, didn't see SPI_0_SCK changed lines.

thomaseichinger · 2015-10-05T08:45:59Z

cpu/atmega2560/periph/spi.c

+    switch (dev) {
+#if SPI_0_EN
+    case SPI_0:
+        SPI_0_DR = out;


I'm not a ATmega expert, but wouldn't it be better to check if a previously set data register was already written out?

As far as I can read from the data sheet, there isn't a way to check if the outgoing data register is empty. However the function will not return until SPIF is set, which indicates that the operation is finished.

OlegHahm · 2015-10-05T08:52:13Z

Welcome to RIOT, @danielamkaer!

thomaseichinger · 2015-10-05T08:52:21Z

@danielamkaer Great job! Please see some minor remarks above.

LudwigKnuepfer · 2015-10-07T15:35:41Z

I have, but I probably won't get around to testing this PR in the immediate future.

PeterKietzmann · 2015-10-22T10:59:17Z

Now that I gave up my Atmega2560 board, testing this PR might also be a job for @ReneHerthel :-)

LudwigKnuepfer · 2015-10-22T20:11:55Z

With

diff --git a/boards/arduino-mega2560/Makefile.features b/boards/arduino-mega2560/Makefile.features
index 6d25415..ad1a466 100644
--- a/boards/arduino-mega2560/Makefile.features
+++ b/boards/arduino-mega2560/Makefile.features
@@ -1,4 +1,5 @@
 FEATURES_PROVIDED += periph_uart
 FEATURES_PROVIDED += periph_gpio
 FEATURES_PROVIDED += periph_timer
+FEATURES_PROVIDED += periph_spi
 FEATURES_MCU_GROUP = avr8

I get

Building application "periph_spi" for "arduino-mega2560" with MCU "atmega2560".

"make" -C /home/lo/RIOT/boards/arduino-mega2560
"make" -C /home/lo/RIOT/core
"make" -C /home/lo/RIOT/cpu/atmega2560
"make" -C /home/lo/RIOT/cpu/atmega2560/periph
"make" -C /home/lo/RIOT/cpu/atmega_common
"make" -C /home/lo/RIOT/drivers
"make" -C /home/lo/RIOT/sys
"make" -C /home/lo/RIOT/sys/auto_init
"make" -C /home/lo/RIOT/sys/shell
"make" -C /home/lo/RIOT/sys/shell/commands
"make" -C /home/lo/RIOT/sys/tsrb
"make" -C /home/lo/RIOT/sys/uart_stdio
/home/lo/RIOT/tests/periph_spi/bin/arduino-mega2560/periph_spi.a(main.o): In function `slave_on_cs':
/home/lo/RIOT/tests/periph_spi/main.c:135: undefined reference to `spi_transmission_begin'
collect2: error: ld returned 1 exit status
/home/lo/RIOT/tests/periph_spi/../../Makefile.include:213: recipe for target 'all' failed
make: *** [all] Error 1

Please ping me when this is resolved.

danielamkaer · 2015-10-26T07:55:35Z

@LudwigKnuepfer try now.

LudwigKnuepfer · 2015-10-26T16:25:52Z

@danielamkaer: On a general note regarding our development procedures - please don't instantly squash your commits, it makes it easier for reviewers to see what changed. The "needs squashing" label is not there to alert you, but to prevent merging of unsquashed PRs.

danielamkaer · 2015-10-27T08:00:36Z

Okay @LudwigKnuepfer - I'll remember for the future. This PR + 2 others are my first ever contributions to open source projects.

LudwigKnuepfer · 2015-10-27T17:41:17Z

@danielamkaer No sweat, you'll (hopefully) get more feedback like this during your first steps. And remember: although some work has gone into our processes and policies, they are not perfect and can benefit from your input. So don't refrain from speaking out if you see room for improvement.

LudwigKnuepfer · 2015-10-27T20:27:21Z

Can't test, don't know how to map GPIO pins to board layout anymore due to changes in RIOT =(
@haukepetersen I suspect you had your hands in this. Please tutor me ;) (Or even better: write some documentation...)

danielamkaer · 2015-10-28T09:33:40Z

@LudwigKnuepfer - Lookup using https://www.arduino.cc/en/Hacking/PinMapping2560

Translate pin 4 (PG5) on Arduino into GPIO_PIN(PORT_G, 5)

danielamkaer · 2015-11-19T07:51:39Z

Any progress?

haukepetersen · 2015-11-19T12:47:05Z

@LudwigKnuepfer I guess we can shed some light on the pin mapping: we don't have a sub-set statically mapped anymore, but you can access any MCU pin freely using the GPIO_PIN(port, pin) macro. As @danielamkaer got it right, if you e.g. want to set pin PG.05 you just call gpio_set(GPIO_PIN(PORT_G, 5) or gpio_set(GPIO(5, 5)). The port enum and the GPIO_PIN macros are defined in cpu/atmega2560/cpu_periph.h.

Now mapping the acutal MCU pins to Arduino pins is another issue, just refer to the Arduino doc (e.g. the link posted above) for details. But I have an idea how to make this easier - will prepare a PR...

haukepetersen · 2015-11-19T13:09:19Z

how about this: #4313. Now you can just use gpio_xx(ARDUINO_PIN_X)

LudwigKnuepfer · 2015-11-22T11:43:27Z

Sorry, I did not have time for testing.. I would not feel hurt if anyone else tests this ;)

zhuoshuguo · 2016-01-23T10:10:52Z

@danielamkaer Ping. A kind reminder for rebase.

danielamkaer · 2016-01-30T17:00:04Z

@zhuoshuguo I need some help I think. I'm trying to do
$ git pull upstream master --rebase
but I run into a lot of conflicts, which are not caused by my changes.

LudwigKnuepfer · 2016-02-20T10:14:00Z

@danielamkaer
I just did a local git rebase upstream/master with your branch and that went fine.
Maybe you want to git reset --hard origin/create_spi_driver_for_atmega2560 in order to reset your branch (throwing away all new changes in case there are any, so be careful) before.

OlegHahm · 2016-02-26T16:59:54Z

@thomaseichinger, @haukepetersen, @PeterKietzmann, anyone for a review?

PeterKietzmann · 2016-02-29T12:47:38Z

Yes, in general but not so much time currently. At the latest during the next Hack'n'ACK

kYc0o · 2016-02-29T14:20:14Z

Maybe @aabadie would like to try this? (:

aabadie · 2016-02-29T17:43:11Z

Thanks @kYc0o. Why not, if I have time.

haukepetersen · 2016-03-09T14:09:51Z

Hi, just tested this PR, for me it does not work. Connecting a logic analyzer to pins 50-52 does not show me any pin activity when sending data on the bus...

haukepetersen · 2016-03-09T17:20:18Z

While starting to debug the driver, I did so many changes, that I decided to make a new PR -> #5020.

@danielamkaer: Sorry again for starting to actually test this so late! It's the good old thing that you only start with something until you really need it...

haukepetersen · 2016-03-10T13:04:43Z

closed in favor of #5020

jnohlgard assigned N8Fear Oct 5, 2015

thomaseichinger reviewed Oct 5, 2015
View reviewed changes

danielamkaer force-pushed the create_spi_driver_for_atmega2560 branch from ffc05f3 to 3664b33 Compare October 5, 2015 08:41

thomaseichinger reviewed Oct 5, 2015
View reviewed changes

danielamkaer force-pushed the create_spi_driver_for_atmega2560 branch from 3664b33 to 5678835 Compare October 5, 2015 08:44

thomaseichinger reviewed Oct 5, 2015
View reviewed changes

danielamkaer force-pushed the create_spi_driver_for_atmega2560 branch 2 times, most recently from be465fe to 297bf0c Compare October 5, 2015 09:10

danielamkaer force-pushed the create_spi_driver_for_atmega2560 branch from 36a575d to 4e77d79 Compare October 12, 2015 06:43

danielamkaer force-pushed the create_spi_driver_for_atmega2560 branch 2 times, most recently from 33013c6 to 9cf1db7 Compare October 26, 2015 07:54

danielamkaer force-pushed the create_spi_driver_for_atmega2560 branch from 9cf1db7 to 8ed35ff Compare October 26, 2015 07:57

danielamkaer force-pushed the create_spi_driver_for_atmega2560 branch 2 times, most recently from 6e78880 to f24c0b4 Compare October 28, 2015 08:28

OlegHahm modified the milestone: Release 2016.03 Dec 8, 2015

zhuoshuguo unassigned N8Fear Jan 23, 2016

miri64 assigned thomaseichinger Jan 26, 2016

cpu/atmega2560: SPI driver

ddde457

danielamkaer force-pushed the create_spi_driver_for_atmega2560 branch from f24c0b4 to ddde457 Compare February 25, 2016 08:43

PeterKietzmann added the Community: Hack'n'ACK candidate This PR is a candidate for review and discussion during one of RIOT's monthly Hack'n'ACK parties label Feb 29, 2016

haukepetersen mentioned this pull request Mar 9, 2016

cpu/atmega2560: added SPI driver #5020

Merged

haukepetersen closed this Mar 10, 2016

Provide feedback

Saved searches

Use saved searches to filter your results more quickly

Implement SPI driver on ATmega2560 #4045

Implement SPI driver on ATmega2560 #4045

danielamkaer commented Oct 5, 2015

jnohlgard commented Oct 5, 2015

thomaseichinger Oct 5, 2015

thomaseichinger Oct 5, 2015

danielamkaer Oct 5, 2015

thomaseichinger Oct 5, 2015

thomaseichinger Oct 5, 2015

danielamkaer Oct 5, 2015

OlegHahm commented Oct 5, 2015

thomaseichinger commented Oct 5, 2015

LudwigKnuepfer commented Oct 7, 2015

PeterKietzmann commented Oct 22, 2015

LudwigKnuepfer commented Oct 22, 2015

danielamkaer commented Oct 26, 2015

LudwigKnuepfer commented Oct 26, 2015

danielamkaer commented Oct 27, 2015

LudwigKnuepfer commented Oct 27, 2015

LudwigKnuepfer commented Oct 27, 2015

danielamkaer commented Oct 28, 2015

danielamkaer commented Nov 19, 2015

haukepetersen commented Nov 19, 2015

haukepetersen commented Nov 19, 2015

LudwigKnuepfer commented Nov 22, 2015

zhuoshuguo commented Jan 23, 2016

danielamkaer commented Jan 30, 2016

LudwigKnuepfer commented Feb 20, 2016

OlegHahm commented Feb 26, 2016

PeterKietzmann commented Feb 29, 2016

kYc0o commented Feb 29, 2016

aabadie commented Feb 29, 2016

haukepetersen commented Mar 9, 2016

haukepetersen commented Mar 9, 2016

haukepetersen commented Mar 10, 2016

Implement SPI driver on ATmega2560 #4045

Implement SPI driver on ATmega2560 #4045

Conversation

danielamkaer commented Oct 5, 2015

jnohlgard commented Oct 5, 2015

thomaseichinger Oct 5, 2015

Choose a reason for hiding this comment

thomaseichinger Oct 5, 2015

Choose a reason for hiding this comment

danielamkaer Oct 5, 2015

Choose a reason for hiding this comment

thomaseichinger Oct 5, 2015

Choose a reason for hiding this comment

thomaseichinger Oct 5, 2015

Choose a reason for hiding this comment

danielamkaer Oct 5, 2015

Choose a reason for hiding this comment

OlegHahm commented Oct 5, 2015

thomaseichinger commented Oct 5, 2015

LudwigKnuepfer commented Oct 7, 2015

PeterKietzmann commented Oct 22, 2015

LudwigKnuepfer commented Oct 22, 2015

danielamkaer commented Oct 26, 2015

LudwigKnuepfer commented Oct 26, 2015

danielamkaer commented Oct 27, 2015

LudwigKnuepfer commented Oct 27, 2015

LudwigKnuepfer commented Oct 27, 2015

danielamkaer commented Oct 28, 2015

danielamkaer commented Nov 19, 2015

haukepetersen commented Nov 19, 2015

haukepetersen commented Nov 19, 2015

LudwigKnuepfer commented Nov 22, 2015

zhuoshuguo commented Jan 23, 2016

danielamkaer commented Jan 30, 2016

LudwigKnuepfer commented Feb 20, 2016

OlegHahm commented Feb 26, 2016

PeterKietzmann commented Feb 29, 2016

kYc0o commented Feb 29, 2016

aabadie commented Feb 29, 2016

haukepetersen commented Mar 9, 2016

haukepetersen commented Mar 9, 2016

haukepetersen commented Mar 10, 2016