M-Labs/artiq-zynq
10
1
Fork 20
Code Issues 13 Pull Requests Releases Wiki Activity

use csr::virtual_leds for SFP0..3 LED indication #244

Merged
sb10q merged 10 commits from morgan/artiq-zynq:feature into master 2023-08-28 16:08:11 +08:00
Conversation 0 Commits 10 Files Changed 4 +5 -20
2 changed files with 5 additions and 20 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files
Showing only changes of commit cfe53a57a3 - Show all commits

21
src/runtime/src/main.rs
View File

@ -104,29 +104,14 @@ async fn report_async_rtio_errors() {
}
}
#[cfg(all(feature = "target_kasli_soc", has_drtio))]
static mut LAST_VIRTUAL_LED_STATUS: u8 = 0;
#[cfg(all(feature = "target_kasli_soc", has_drtio))]
#[cfg(feature = "target_kasli_soc")]
async fn io_expanders_service(
i2c_bus: RefCell<&mut libboard_zynq::i2c::I2c>,
io_expander0: RefCell<io_expander::IoExpander>,
morgan marked this conversation as resolved Outdated
sb10q commented 2023-08-04 15:41:03 +08:00
Outdated
Review
Copy Link

The only unsafe line is the next one, no?

The only unsafe line is the next one, no?
morgan commented 2023-08-04 15:48:54 +08:00
Outdated
Review
Copy Link

LAST_VIRTUAL_LED_STATUS is a static mut. So it need an unsafe too

```LAST_VIRTUAL_LED_STATUS``` is a static mut. So it need an unsafe too
sb10q commented 2023-08-13 14:48:09 +08:00
Outdated
Review
Copy Link

I don't think this should be gated on has_drtio.

I don't think this should be gated on has_drtio.
io_expander1: RefCell<io_expander::IoExpander>,
) {
morgan marked this conversation as resolved Outdated
mwojcik commented 2023-08-01 16:35:29 +08:00
Outdated
Review
Copy Link

Could save a CSR call by getting the value before the if

Could save a CSR call by getting the value before the ``if``
loop {
let _ = block_async!((|| -> nb::Result<(), Void> {
unsafe {
let current = pl::csr::virtual_leds::status_read();
if current != LAST_VIRTUAL_LED_STATUS {
LAST_VIRTUAL_LED_STATUS = current;
Ok(())
} else {
Err(nb::Error::WouldBlock)
}
}
})())
.await;
task::r#yield().await;
io_expander0
.borrow_mut()
.service(&mut i2c_bus.borrow_mut())
sb10q commented 2023-08-13 14:50:38 +08:00
Outdated
Review
Copy Link

I don't think this entire block_async is any useful. It also duplicates logic and breaks abstraction layers. Anyway just remove it.

I don't think this entire block_async is any useful. It also duplicates logic and breaks abstraction layers. Anyway just remove it.
morgan commented 2023-08-15 11:28:26 +08:00
Outdated
Review
Copy Link

The block_async prevent "io_expanders_service" from locking up the whole cpu. As it blocks the loop from running over and over again.

The block_async prevent "io_expanders_service" from locking up the whole cpu. As it blocks the loop from running over and over again.
@ -192,7 +177,7 @@ pub fn main_core0() {
task::spawn(report_async_rtio_errors());
#[cfg(all(feature = "target_kasli_soc", has_drtio))]
#[cfg(feature = "target_kasli_soc")]
task::spawn(io_expanders_service(
RefCell::new(i2c_bus),
RefCell::new(io_expander0),

4
src/satman/src/main.rs
View File

@ -671,7 +671,7 @@ pub extern "C" fn main_core0() -> i32 {
for mut rep in repeaters.iter_mut() {
rep.service(&routing_table, rank, &mut timer);
}
#[cfg(all(feature = "target_kasli_soc", has_drtio))]
#[cfg(feature = "target_kasli_soc")]
{
io_expander0
.service(&mut i2c)
@ -717,7 +717,7 @@ pub extern "C" fn main_core0() -> i32 {
for mut rep in repeaters.iter_mut() {
rep.service(&routing_table, rank, &mut timer);
}
#[cfg(all(feature = "target_kasli_soc", has_drtio))]
#[cfg(feature = "target_kasli_soc")]
{
io_expander0
.service(&mut i2c)