pull in improved API to loader

This commit is contained in:
bunnie 2024-01-27 16:53:15 +08:00
parent 74219d58f5
commit 474d42ff87
2 changed files with 18 additions and 31 deletions

View file

@ -17,9 +17,7 @@ utralib = { version = "0.1.24", optional = true, default-features = false }
armv7 = { git = "https://github.com/Foundation-Devices/armv7.git", branch = "update", optional = true }
atsama5d27 = { git = "https://github.com/Foundation-Devices/atsama5d27.git", branch = "master", optional = true }
sha2-loader = { path = "./sha2-loader", default-features = false, optional = true }
cramium-hal = { path = "../libs/cramium-hal", optional = true, default_features = false, features = [
"baremetal",
] }
cramium-hal = { path = "../libs/cramium-hal", optional = true, default_features = false }
xous-pl230 = { path = "../libs/xous-pl230", optional = true, features = [
"tests",
"pio",

View file

@ -118,17 +118,16 @@ pub fn early_init() {
let baudrate: u32 = 115200;
let freq: u32 = 100_000_000;
// the address of the UART buffer is "hard-allocated" at an offset one page from the top of
// IFRAM0. This is a convention that must be respected by the UDMA UART library implementation
// for things to work.
let uart_buf_addr = utralib::HW_IFRAM0_MEM + utralib::HW_IFRAM0_MEM_LEN - 4096;
let mut udma_uart = unsafe {
// safety: this is safe to call, because we set up clock and events prior to calling new.
udma::Uart::new(utra::udma_uart_0::HW_UDMA_UART_0_BASE, baudrate, freq)
};
let tx_buf = unsafe {
// safety: it's safe only because we are manually tracking the allocations in IFRAM0. Yuck!
core::slice::from_raw_parts_mut(
(utralib::HW_IFRAM0_MEM + utralib::HW_IFRAM0_MEM_LEN - 4096) as *mut u8,
4096,
)
udma::Uart::get_handle(utra::udma_uart_0::HW_UDMA_UART_0_BASE, uart_buf_addr, uart_buf_addr)
};
udma_uart.set_baud(baudrate, freq);
// Board bring-up: send characters to confirm the UART is configured & ready to go for the logging crate!
// The "boot gutter" also has a role to pause the system in "real mode" before VM is mapped in Xous
// makes things a little bit cleaner for JTAG ops, it seems.
@ -148,35 +147,25 @@ pub fn early_init() {
xous_pl230::pl230_tests::units::basic_tests(&mut pl230);
xous_pl230::pl230_tests::units::pio_test(&mut pl230);
// Confirm that Rx can work while also adding a boot wait point.
let rx_buf = unsafe {
// safety: it's safe only because we are manually tracking the allocations in IFRAM0. Yuck!
core::slice::from_raw_parts_mut(
(utralib::HW_IFRAM0_MEM + utralib::HW_IFRAM0_MEM_LEN - 8192) as *mut u8,
1,
)
};
const BANNER: &'static str = "\n\rKeep pressing keys to continue boot...\r\n";
tx_buf[..BANNER.len()].copy_from_slice(BANNER.as_bytes());
udma_uart.write(&tx_buf[..BANNER.len()]);
udma_uart.write(BANNER.as_bytes());
// space for one character, plus appending CRLF for the return
let mut rx_buf = [0u8; 3];
// receive characters -- print them back. just to prove that this works. no other reason than that.
for _ in 0..4 {
udma_uart.read(rx_buf);
udma_uart.read(&mut rx_buf[..1]);
const DBG_MSG: &'static str = "Got: ";
tx_buf[..DBG_MSG.len()].copy_from_slice(DBG_MSG.as_bytes());
udma_uart.write(&tx_buf[..DBG_MSG.len()]);
tx_buf[0] = rx_buf[0];
udma_uart.write(&tx_buf[..1]);
tx_buf[0] = '\n' as u32 as u8;
tx_buf[1] = '\r' as u32 as u8;
udma_uart.write(&tx_buf[..2]);
udma_uart.write(&DBG_MSG.as_bytes());
rx_buf[1] = '\n' as u32 as u8;
rx_buf[2] = '\r' as u32 as u8;
udma_uart.write(&rx_buf);
}
}
const ONWARD: &'static str = "\n\rBooting!\n\r";
tx_buf[..ONWARD.len()].copy_from_slice(ONWARD.as_bytes());
udma_uart.write(&tx_buf[..ONWARD.len()]);
udma_uart.write(&ONWARD.as_bytes());
}
#[cfg(feature = "platform-tests")]