Crates.io | msp430 |
lib.rs | msp430 |
version | 0.4.1 |
source | src |
created_at | 2017-07-22 18:38:53.021209 |
updated_at | 2022-12-05 03:24:39.179797 |
description | Low level access to MSP430 microcontrollers |
homepage | |
repository | https://github.com/rust-embedded/msp430 |
max_upload_size | |
id | 24541 |
size | 33,023 |
msp430
Low level access to MSP430 microcontrollers
This crate is based on cortex-m crate by Jorge Aparicio (@japaric).
This crate requires a nightly rust due to the use of the new asm!
(0.3.0
and above), llvm_asm!
(0.2.2
) or old asm!
(0.2.1
and below) macros.
The below table contains compilers which are known to work:
msp430 version |
rustc compiler |
---|---|
0.3.0 |
nightly-2022-01-24 |
0.2.2 |
nightly-2020-04-22 |
0.2.1 |
nightly-2020-01-04 |
The critical-section-single-core
feature provides a critical section
implementation based upon disabling interrupts.
Critical sections by disabling interrupts are a valuable way to access I/O and shared data safely in msp430. However, rustc/LLVM does not always optimize critical sections well in terms of space. For example, sometimes rustc/LLVM will create two copies of interrupt enable assembly code when exiting a critical section that contains a branch- one copy each for branch-taken/branch-not-taken.
This crate provides three features for giving hints to rustc/LLVM for how to
optimize critical sections for size. Both critical_section::with
and
the interrupt::free
critical sections are supported, with and without the
critical-section
feature above:
outline-cs-acq
: Hint to rustc/LLVM that each critical section entry should
be a call to a single copy of an acquire
function (disable interrupts).outline-cs-rel
: Hint to rustc/LLVM that each critical section exit should
be a call to a single copy of a release
function (enable interrupts if not
in a nested critical section).outline-cs
: Convenience feature for enabling both of the above at the same
time.If saving space is a concern in your application, you should experiment with
which features provide the best size savings and balance this against the
execution overhead of the extra function calls due to outlining. The execution
overhead of enabling each hint is ~5 + 2 clock cycles for each
critical section- at least a CALL
and RET
instruction.
Licensed under either of
at your option.
Unless you explicitly state otherwise, any contribution intentionally submitted for inclusion in the work by you, as defined in the Apache-2.0 license, shall be dual licensed as above, without any additional terms or conditions.