libmaple/examples/test-usart-dma.cpp

212 lines
7.5 KiB
C++
Raw Permalink Normal View History

/**
* @file examples/test-usart-dma.cpp
* @author Marti Bolivar <mbolivar@leaflabs.com>
*
* Simple test of DMA used with a USART receiver.
*
* Configures a USART receiver for use with DMA. Received bytes are
* placed into a buffer, with an interrupt firing when the buffer is
* full. At that point, the USART transmitter will print the contents
* of the byte buffer. The buffer is continually filled and refilled
* in this manner.
*
* This example isn't very robust; don't use it in production. In
* particular, since the buffer keeps filling (DMA_CIRC_MODE is set),
* if you keep sending characters after filling the buffer, you'll
* overwrite earlier bytes; this may happen before those earlier bytes
* are done printing. (Typing quickly and seeing how it affects the
* output is a fun way to make sense of how the interrupts and the
* main thread of execution interleave.)
*
* This code is released into the public domain.
*/
Move public headers to include directories; related cleanups. Move libmaple/*.h to (new) libmaple/include/libmaple/. The new accepted way to include a libmaple header foo.h is with: #include <libmaple/foo.h> This is more polite in terms of the include namespace. It also allows us to e.g. implement the Arduino SPI library at all (which has header SPI.h; providing it was previously impossible on case-insensitive filesystems due to libmaple's spi.h). Similarly for Wirish. The old include style (#include "header.h") is now deprecated. libmaple/*.h: - Change include guard #defines from _FOO_H_ to _LIBMAPLE_FOO_H_. - Add license headers where they're missing - Add conditional extern "C" { ... } blocks where they're missing (they aren't always necessary, but we might was well do it against the future, while we're at it.). - Change includes from #include "foo.h" to #include <libmaple/foo.h>. - Move includes after extern "C". - Remove extra trailing newlines Note that this doesn't include the headers under libmaple/usb/ or libmaple/usb/usb_lib. These will get fixed later. libmaple/*.c: - Change includes from #include "foo.h" to #include <libmaple/foo.h>. Makefile: - Add I$(LIBMAPLE_PATH)/include/libmaple to GLOBAL_FLAGS. This allows for users (including Wirish) to migrate their code, but should go away ASAP, since it slows down compilation. Wirish: - Move wirish/**/*.h to (new) wirish/include/wirish/. This ignores the USB headers, which, as usual, are getting handled after everything else. - Similarly generify wirish/boards/ structure. For each supported board "foo", move wirish/boards/foo.h and wirish/boards/foo.cpp to wirish/boards/foo/include/board/board.h and wirish/boards/foo/board.cpp, respectively. Also remove the #ifdef hacks around the .cpp files. - wirish/rules.mk: put wirish/boards/foo/include in the include path (and add wirish/boards/foo/board.cpp to the list of sources to be compiled). This allows saying: #include <board/board.h> instead of the hack currently in place. We can allow the user to override this setting later to make adding custom board definitions easier. - Disable -Werror in libmaple/rules.mk, as the current USB warnings don't let the olimex_stm32_h103 board compile. We can re-enable -Werror once we've moved the board-specific bits out of libmaple proper. libraries, examples: - Update includes accordingly. - Miscellaneous cosmetic fixups. Signed-off-by: Marti Bolivar <mbolivar@leaflabs.com>
2011-11-15 18:45:43 +01:00
#include <libmaple/dma.h>
#include <libmaple/usart.h>
#include <libmaple/gpio.h>
Move public headers to include directories; related cleanups. Move libmaple/*.h to (new) libmaple/include/libmaple/. The new accepted way to include a libmaple header foo.h is with: #include <libmaple/foo.h> This is more polite in terms of the include namespace. It also allows us to e.g. implement the Arduino SPI library at all (which has header SPI.h; providing it was previously impossible on case-insensitive filesystems due to libmaple's spi.h). Similarly for Wirish. The old include style (#include "header.h") is now deprecated. libmaple/*.h: - Change include guard #defines from _FOO_H_ to _LIBMAPLE_FOO_H_. - Add license headers where they're missing - Add conditional extern "C" { ... } blocks where they're missing (they aren't always necessary, but we might was well do it against the future, while we're at it.). - Change includes from #include "foo.h" to #include <libmaple/foo.h>. - Move includes after extern "C". - Remove extra trailing newlines Note that this doesn't include the headers under libmaple/usb/ or libmaple/usb/usb_lib. These will get fixed later. libmaple/*.c: - Change includes from #include "foo.h" to #include <libmaple/foo.h>. Makefile: - Add I$(LIBMAPLE_PATH)/include/libmaple to GLOBAL_FLAGS. This allows for users (including Wirish) to migrate their code, but should go away ASAP, since it slows down compilation. Wirish: - Move wirish/**/*.h to (new) wirish/include/wirish/. This ignores the USB headers, which, as usual, are getting handled after everything else. - Similarly generify wirish/boards/ structure. For each supported board "foo", move wirish/boards/foo.h and wirish/boards/foo.cpp to wirish/boards/foo/include/board/board.h and wirish/boards/foo/board.cpp, respectively. Also remove the #ifdef hacks around the .cpp files. - wirish/rules.mk: put wirish/boards/foo/include in the include path (and add wirish/boards/foo/board.cpp to the list of sources to be compiled). This allows saying: #include <board/board.h> instead of the hack currently in place. We can allow the user to override this setting later to make adding custom board definitions easier. - Disable -Werror in libmaple/rules.mk, as the current USB warnings don't let the olimex_stm32_h103 board compile. We can re-enable -Werror once we've moved the board-specific bits out of libmaple proper. libraries, examples: - Update includes accordingly. - Miscellaneous cosmetic fixups. Signed-off-by: Marti Bolivar <mbolivar@leaflabs.com>
2011-11-15 18:45:43 +01:00
#include <wirish/wirish.h>
/*
* Configuration and state
*/
// Serial port and DMA configuration. You can change these to suit
// your purposes.
HardwareSerial *serial = &Serial2;
#define USART_DMA_DEV DMA1
#if STM32_MCU_SERIES == STM32_SERIES_F1
// On STM32F1 microcontrollers (like what's on Maple and Maple Mini),
// dma tubes are channels.
#define USART_RX_DMA_TUBE DMA_CH6
#elif (STM32_MCU_SERIES == STM32_SERIES_F2 || \
STM32_MCU_SERIES == STM32_SERIES_F4)
// On STM32F2 and STM32F4 microcontrollers (Maple 2 will have an F4),
// dma tubes are streams.
#define USART_RX_DMA_TUBE DMA_S5
#else
#error "unsupported stm32 series"
#endif
// The serial port will make a DMA request each time it receives data.
// This is the dma_request_src we use to tell the DMA tube to handle
// that DMA request.
#define USART_DMA_REQ_SRC DMA_REQ_SRC_USART2_RX
#define BAUD 9600
// This will store the DMA configuration for USART RX.
dma_tube_config tube_config;
// This will store received USART characters.
#define BUF_SIZE 20
char rx_buf[BUF_SIZE];
// The interrupt handler, rx_dma_irq(), sets this to 1.
volatile uint32 irq_fired = 0;
// Used to store DMA interrupt status register (ISR) bits inside
// rx_dma_irq(). This helps explain what's going on inside loop(); see
// comments below.
volatile uint32 isr = 0;
/*
* Helper functions
*/
// This is our DMA interrupt handler.
void rx_dma_irq(void) {
irq_fired = 1;
isr = dma_get_isr_bits(USART_DMA_DEV, USART_RX_DMA_TUBE);
}
// Configure the USART receiver for use with DMA:
// 1. Turn it on.
// 2. Set the "DMA request on RX" bit in USART_CR3 (USART_CR3_DMAR).
void setup_usart(void) {
serial->begin(BAUD);
usart_dev *serial_dev = serial->c_dev();
serial_dev->regs->CR3 = USART_CR3_DMAR;
}
// Set up our dma_tube_config structure. (We could have done this
// above, when we declared tube_config, but having this function makes
// it easier to explain what's going on).
void setup_tube_config(void) {
// We're receiving from the USART data register. serial->c_dev()
// returns a pointer to the libmaple usart_dev for that serial
// port, so this is a pointer to its data register.
tube_config.tube_src = &serial->c_dev()->regs->DR;
// We're only interested in the bottom 8 bits of that data register.
tube_config.tube_src_size = DMA_SIZE_8BITS;
// We're storing to rx_buf.
tube_config.tube_dst = rx_buf;
// rx_buf is a char array, and a "char" takes up 8 bits on STM32.
tube_config.tube_dst_size = DMA_SIZE_8BITS;
// Only fill BUF_SIZE - 1 characters, to leave a null byte at the end.
tube_config.tube_nr_xfers = BUF_SIZE - 1;
// Flags:
// - DMA_CFG_DST_INC so we start at the beginning of rx_buf and
// fill towards the end.
// - DMA_CFG_CIRC so we go back to the beginning and start over when
// rx_buf fills up.
// - DMA_CFG_CMPLT_IE to turn on interrupts on transfer completion.
tube_config.tube_flags = DMA_CFG_DST_INC | DMA_CFG_CIRC | DMA_CFG_CMPLT_IE;
// Target data: none. It's important to set this to NULL if you
// don't have any special (microcontroller-specific) configuration
// in mind, which we don't.
tube_config.target_data = NULL;
// DMA request source.
tube_config.tube_req_src = USART_DMA_REQ_SRC;
}
// Configure the DMA controller to serve DMA requests from the USART.
void setup_dma_xfer(void) {
// First, turn it on.
dma_init(USART_DMA_DEV);
// Next, configure it by calling dma_tube_cfg(), and check to make
// sure it succeeded. DMA tubes have many restrictions on their
// configuration, and there are configurations which work on some
// types of STM32 but not others. libmaple tries hard to make
// things just work, but checking the return status is important!
int status = dma_tube_cfg(USART_DMA_DEV, USART_RX_DMA_TUBE, &tube_config);
ASSERT(status == DMA_TUBE_CFG_SUCCESS);
// Now we'll perform any other configuration we want. For this
// example, we attach an interrupt handler.
dma_attach_interrupt(USART_DMA_DEV, USART_RX_DMA_TUBE, rx_dma_irq);
// Turn on the DMA tube. It will now begin serving requests.
dma_enable(USART_DMA_DEV, USART_RX_DMA_TUBE);
}
/*
* setup() and loop()
*/
void setup(void) {
pinMode(BOARD_LED_PIN, OUTPUT);
setup_tube_config();
setup_dma_xfer();
setup_usart();
}
void loop(void) {
toggleLED();
delay(100);
// See if the interrupt handler got called since the last time we
// checked.
if (irq_fired) {
serial->println("** IRQ **");
// Notice how the interrupt status register (ISR) bits show
// transfer complete _and_ half-complete here, but the ISR
// bits we print next will be zero. That's because the
// variable "isr" gets set _inside_ rx_dma_irq(). After it
// exits, libmaple cleans up by clearing the tube's ISR
// bits. (If it didn't, and we forgot to, the interrupt would
// repeatedly fire forever.)
serial->print("ISR bits: 0x");
serial->println(isr, HEX);
irq_fired = 0;
}
// Print the ISR bits.
//
// Notice that the "transfer half-complete" ISR flag gets set when
// we reach the rx_buf half-way point. This is true even though we
// don't tell the DMA controller to interrupt us on a
// half-complete transfer. That is, the ISR bits get set at the
// right times no matter what; we just don't get interrupted
// unless we asked. (If an error or other problem occurs, the
// relevant ISR bits will get set in the same way).
serial->print("[");
serial->print(millis());
serial->print("]\tISR bits: 0x");
uint8 isr_bits = dma_get_isr_bits(USART_DMA_DEV, USART_RX_DMA_TUBE);
serial->print(isr_bits, HEX);
// Print the contents of rx_buf. If you keep typing after it fills
// up, the new characters will overwrite the old ones, thanks to
// DMA_CIRC_MODE.
serial->print("\tCharacter buffer contents: '");
serial->print(rx_buf);
serial->println("'");
if (isr_bits == 0x7) {
serial->println("** Clearing ISR bits.");
dma_clear_isr_bits(USART_DMA_DEV, USART_RX_DMA_TUBE);
}
}
// ------- init() and main() --------------------------------------------------
// Force init to be called *first*, i.e. before static object allocation.
// Otherwise, statically allocated objects that need libmaple may fail.
__attribute__((constructor)) void premain() {
init();
}
int main(void) {
setup();
while (true) {
loop();
}
return 0;
}