After migrating to 0.22, init_psram hangs (esp32s3)

[yanshay]

I migrated an application I'm developing for esp32s3, to 0.22 .

To get it to compile it was all technicalities as in the migration instructions.
To get it to link I had to remove the link with rom_functions.x, I didn't see it documented anywhere but it wasn't found and linked without it.

When I run the code it starts and hangs in init_psram. There isn't much going on before that initialization, this is all the code that runs:

#[esp_hal_embassy::main]
async fn main(spawner: Spawner) {
    esp_println::logger::init_logger_from_env();
    info!("Application Started");

    let mut peripherals = esp_hal::init({
        let mut config = esp_hal::Config::default();
        config.cpu_clock = CpuClock::max();
        config
    });
    info!("peripherals Created");

    // IMPORTANT: PSRAM need to be initialized first, so 'Normal' allocations will use the region
    let (start, size) = psram::init_psram(peripherals.PSRAM, psram::PsramConfig::default());
    info!("Initialized psram");
    init_psram_heap(start, size);
    info!("Initialized psram heap");
...

The log output that comes out is as below, and then the program hangs and doesn't continue further (so "Initialized psram" log isn't reached).

INFO - [src/main.rs:118] Application Started
INFO - [src/main.rs:125] peripherals Created

Anything changed with psram initialization?

[bjoernQ]

Regarding PSRAM the diff doesn't show relevant changes: v0.21.1...v0.22.0

Regarding the rom-functions the migration guide chapter is here: https://github.com/esp-rs/esp-hal/blob/main/esp-wifi/MIGRATING-0.10.md#no-need-to-include-rom_functionsx-manually

[yanshay]

So any ideas what else to look for?

It's a pretty large app that has evolved over half a year through several esp-hal versions, and suddenly fails on the first few lines of code.
I could stop using log and switch to println instead, but I'm afraid it will hit me elsewhere.

[bjoernQ]

Whatever is causing the problems it's not (just) the size of the flashed binary. e.g. this

#![no_std]
#![no_main]
#![feature(asm_experimental_arch)]

use core::arch::asm;

use embassy_executor::Spawner;
use embassy_time::{Duration, Timer};
use esp_backtrace as _;
use esp_hal::prelude::*;
use esp_println::println;
use log::info;

extern crate alloc;

fn init_psram_heap(start: *mut u8, size: usize) {
    unsafe {
        esp_alloc::HEAP.add_region(esp_alloc::HeapRegion::new(
            start,
            size,
            esp_alloc::MemoryCapability::External.into(),
        ));
    }
}

#[main]
async fn main(spawner: Spawner) {
    esp_println::logger::init_logger_from_env();
    info!("Application Started");

    unsafe {
        asm!(
            "
            .rept 20000
            nop
            .endr        
        "
        );
    }

    info!("Application still alive");

    unsafe {
        asm!(
            "
            .rept 10000
            nop
            .endr        
        "
        );
    }

    foo1();

    let mut peripherals = esp_hal::init({
        let mut config = esp_hal::Config::default();
        config.cpu_clock = CpuClock::max();
        config
    });
    info!("peripherals Created");

    // IMPORTANT: PSRAM need to be initialized first, so 'Normal' allocations will use the region
    let (start, size) =
        esp_hal::psram::init_psram(peripherals.PSRAM, esp_hal::psram::PsramConfig::default());
    info!("Initialized psram");
    init_psram_heap(start, size);
    info!("Initialized psram heap");

    

    loop {
        info!("still alive");
    }
}


#[link_section = ".text"]
static FOO: [u32; 500_000] = [0u32; 500_000];



#[link_section = ".rodata"]
static FOO2: [u32; 500_000] = [0u32; 500_000];

fn foo1() {
    unsafe {
        println!("{}", FOO[1000]);
        println!("{}", FOO2[1000]);
    }

    unsafe {
        asm!(
            "
            .rept 10000
            nop
            .endr        
        "
        );
    }
    println!("still here");
}

Occupies 4mb of flash and still works fine for me

[bjoernQ]

I could stop using log and switch to println instead, but I'm afraid it will hit me elsewhere.

Yeah log not working is just a symptom of whatever is causing the problems here. There seem to be no relevant changes between 0.21.1 and 0.22.0 regarding PSPRAM

[yanshay]

Given there were no direct changes in PSRAM and from my checks neither in esp_println, and I'm using the same log crate version, so only left is the init code. Anything changed there that could affect PSRAM init behavior?

Also, dependencies could change, like the serial communication for the log, or changes in pac.

Is there any code that runs except my main at that point? due to interrupts? something the bootloader sets up maybe that could influence? It seems like so little is taking place that there is not a lot that could affect this behavior.

I switched to use println! and everything else work fine. There is a lot going on in that app, communication with esp-wifi, esp-mbedtls, mqtt, i8080, external device over I2C, json processing, and everything seems to work. If it was some memory issue or something deeper I think it would have appeared.

Only log fails in this strange way. So something must have changed.

[bjoernQ]

That's surprising! I double checked and of course a lot has changed but nothing which seems relevant for this behavior.

• esp-println didn't change
• the log crate didn't change
• PACs did change but I don't see how that could relate to logging
• init didn't change a lot from 0.21.0
• linker scripts changed, but - why is only logging affected

I would really like to understand what is happening here, but it doesn't make a lot of sense given the changes and the observed impact