3.12 Direct memory access controller (DMA)
The direct memory access (DMA) controller is a bus master and system peripheral with single-AHB architecture.
With seven channels, it performs data transfers between memory-mapped peripherals and/or memories, to
offload the CPU.
Each channel is dedicated to managing memory access requests from one or more peripherals. The unit includes
an arbiter for handling the priority between DMA requests.
Main features of the DMA controller:
• Single-AHB master
• Peripheral-to-memory, memory-to-peripheral, memory-to-memory and peripheral-to-peripheral data
transfers
• Access, as source and destination, to on-chip memory-mapped devices such as flash memory, SRAM, and
AHB and APB peripherals
• All DMA channels independently configurable:
– Each channel is associated either with a DMA request signal coming from a peripheral, or with a
software trigger in memory-to-memory transfers. This configuration is done by software.
– Priority between the requests is programmable by software (four levels per channel: very high, high,
medium, low) and by hardware in case of equality (such as request to channel 1 has priority over
request to channel 2).
– Transfer size of source and destination are independent (byte, half-word, word), emulating packing
and unpacking. Source and destination addresses must be aligned on the data size.
– Support of transfers from/to peripherals to/from memory with circular buffer management
– Programmable number of data to be transferred: 0 to 2
16
- 1
• Generation of an interrupt request per channel. Each interrupt request originates from any of the three
DMA events: transfer complete, half transfer, or transfer error.
3.13
DMA request multiplexer (DMAMUX)
The DMAMUX request multiplexer enables routing a DMA request line between the peripherals and the DMA
controller. Each channel selects a unique DMA request line, unconditionally or synchronously with events from its
DMAMUX synchronization inputs. DMAMUX may also be used as a DMA request generator from programmable
events on its input trigger signals.
3.14
Interrupts and events
The device flexibly manages events causing interrupts of linear program execution, called exceptions. The
Cortex
®
-M0+ processor core, a nested vectored interrupt controller (NVIC) and an extended interrupt/event
controller (EXTI) are the assets contributing to handling the exceptions. Exceptions include core-internal events
such as, for example, a division by zero and, core-external events such as logical level changes on physical lines.
Exceptions result in interrupting the program flow, executing an interrupt service routine (ISR) then resuming the
original program flow.
The processor context (contents of program pointer and status registers) is stacked upon program interrupt and
unstacked upon program resume, by hardware. This avoids context stacking and unstacking in the interrupt
service routines (ISRs) by software, thus saving time, code and power. The ability to abandon and restart load-
multiple and store-multiple operations significantly increases the device’s responsiveness in processing
exceptions.
3.14.1 Nested vectored interrupt controller (NVIC)
The configurable nested vectored interrupt controller is tightly coupled with the core. It handles physical line
events associated with a non-maskable interrupt (NMI) and maskable interrupts, and Cortex
®
-M0+ exceptions. It
provides flexible priority management.
The tight coupling of the processor core with NVIC significantly reduces the latency between interrupt events and
start of corresponding interrupt service routines (ISRs). The ISR vectors are listed in a vector table, stored in the
NVIC at a base address. The vector address of an ISR to execute is hardware-built from the vector table base
address and the ISR order number used as offset.
STM32U073x8/B/C
Functional overview
DS14548 - Rev 1
page 17/145