panic! in embassy-usb-synopsys-otg on second serial connection (BUT only if defmt-trace level debugging not enabled)

[paulc]

I had a strange issue running the usb_serial example on an STM32F401 where the serial device would work fine for the first connection from the host, but if I disconnected the terminal session and then re-connected the driver would panic with the following message.

95.600067 [panic_probe::print_defmt::print] lib.rs:104  ERROR panicked at /Users/paulc/Sync/Development/embedded-rust/embassy/embassy-usb-synopsys-otg/src/lib.rs:46:9:
assertion failed: ep_num < ep_count

Very strangely, if I enabled the defmt feature in embassy-stm32 crate and set DEFMT_LOG=trace the problem would go away. As an aside, enabling the defmt feature in the embassy-usb-synopsys-otg results in lots of compile errors with defmt::assert! (see #3492) (constant function errors) which makes debugging a bit tricky.

It looks like the driver is getting an unexpected response from let status = r.grxstsp().read() (status.epnum() was 12 rather than 1 when the error showed up in my testing).

The patch below fixes this by replacing the assert!(ep_num < ep_count) with a test which skips the packet rather than panic-ing.

I suspect that there is a more fundamental issue here but but this does at least fix the panic! issue.

diff --git a/embassy-usb-synopsys-otg/src/lib.rs b/embassy-usb-synopsys-otg/src/lib.rs
index f90403936..3197817ee 100644
--- a/embassy-usb-synopsys-otg/src/lib.rs
+++ b/embassy-usb-synopsys-otg/src/lib.rs
@@ -43,7 +43,16 @@ pub unsafe fn on_interrupt<const MAX_EP_COUNT: usize>(
         let ep_num = status.epnum() as usize;
         let len = status.bcnt() as usize;
 
-        assert!(ep_num < ep_count);
+        if ep_num >= ep_count {
+            // Removed `assert!(ep_num < ep_count);` as this fails for second serial connection UNLESS
+            // defmt-trace level debugging is enabled in embassy-stm32 (note that the defmt feature
+            // doesnt work in embassy-usb-synopsys-otg as defmt::assert! generates multiple errors)
+            //
+            // I'm guessing this may be some sort of timing issue so instead of panic-ing skip the
+            // invalid packet :‑(
+            error!("Skipping Invalid Packet (ep_num >= ep_count): ep_num: {} ep_count: {}", ep_num, ep_count);
+            continue;
+        }
 
         match status.pktstsd() {
             vals::Pktstsd::SETUP_DATA_RX => {

[vDorst]

I see these panics to with this #3212 (comment) usb audio. But every reset, pressing the reset butten, the device/software behaves differently.

[votrungchi]

I think it related my issue #3459