stm32 ETH_IRQHandler never hit - stm32

Im using stm32f746ZGT + LAN8472A and lwip + freertos on my custom board. I generated default ethernet project by cubemx. I set static ip.
After connect the cable Link leds seem like good. Yellow one is
stable and green one is blinky.
In debugging i can see netif is fully configured and link is up.
RXD0 and RXD1 pins which are entering mcu have signal (i can see on
oscilloscope screen).
Clocks are good (i tested with oscilloscope).
But ETH_IRQHandler never hit so it doesnt answer any questions such as ping request. (I enabled ICMP at lwIP settings). What should i check to find the source of problem

I'm stuck in an issue that seems quite similar to your. I don't have a solution yet, but thought to share something I found that maybe take you into the right direction.
What I've found is that this seems an issue of teh Micro itself. Following I'm signaling you two links that point to long disucssions regarding (I think) the same problem.
https://github.com/ARMmbed/mbed-os/issues/6262
STM Errata
STM Errata
They both seem to point out to a specific hardware vesion of the STM32F769 (Rev. A) and it seems that the thing is solved in Rev. Z......I use a lot of conditionals, I know, but until I'll be able to test a Micro with the correct revision I won't be able to confirm the problem/solution.
Hope it helps.
Lorenzo
EDIT!
While I don't have an answer yet I can confirm that on STM32F4, with the same software, the problem does not show.

Related

Re-program STM32F102 trouble

I am trying to make some custom firmware for a MIDI controller (AKAI LPD8).
There is an STM32F102R8T6 chip in the unit.
I am trying to reach it with a programmer to wipe it, but it seems to not be responsive.
Some information and thing I have tried:
The firmware that came with the unit works, so the chip is not broken
Removed the components connected to the programming pins (PA9-PA10 and PA13-PA14)
I am able to pull BOOT0 high and have it not run the main program, but I am however not able to get a life sign using either an ST-Link2(clone) connected to PA13/14, nor a USB to serial adapter connected to PA9/PA10, so I am not sure what mode it is in
The connection has been checked, and RX-TX etc is the correct way around (but also for the sake of trying it all, I reversed the connections as well...).
Tried both the STM32CubeProgrammer and stm32flash, but none connects.
I am actually not sure if AKAI have locked the chip in such a way that you cannot even do a full chip erase and use the chip for something new? The NRST pin is strangely not doing anything to the running of the firmware either when I try to pull it low.
Is there a way to reprogram these chips when they come off of a commercial product, or are they permanently locked?
Any solution/tips?
Many of the STM32 parts have "proprietary code read-out protection" (google PCROP) which but you might be lucky and they haven't enabled it in the option bytes. Read the documentation for that and the bootloader documentation and get a good idea of what you expect it to do if it is enabled and if it isn't.
If you have a scope, try watching the SWD/JTAG pins to see if there is any response from the device. (If you aren't even sure if it is in reset then scope the crystal if there is one).
If you haven't got a scope, you might be able to to verify what it is doing by seeing if it sets the pins and pull-resistors to how they would be expected to be in the bootloader mode, eg: UART TX should be high if it is enabled, even it it isn't transmitting anything. Put a strong pull-down (~1k) on there and see if it still reads high.
After hours of trying different ways of making it work (also tried the alternate mapping of the UART port), and probed the TX pin as suggested by Tom V to no avail, I have given up working on that specific chip and ordered an upgrade from the STM32F4 family instead to replace it with. A lot more power and useful peripherals.
A bit of a non-answer to the specific question. Frustrating to not have found out what was wrong (chip or approach) but being mindful of the sunk cost fallacy, I think it was best to just replace the chip with a fresh one and start development from there.

STM32H743VI jumps randomly in a nonflashable state

we are using a STM32H743VIT6 on a custom board with a JLink debugger. Out of the blue the processor jumps in a state where it isn't possible to flash the ECU anymore. The board is running but nether JMem nor our IDE (uVision) are able to access or detect the controller. Has anyone else encountered this behaviour so far? Google wasn't helpful either.
It is almost impossible to archive unless you enable RDP (which is very hard to archive if it was not the intention of the programmer).
You probably have screw-up the board design. You should have pull-up resistors on the debug lines and NRST connected to the programmer.
If you do not have NRST available simple solder the wire to the NRST, and when the programming probe connects to the uC, connect it to the GND.
If the NRST line is connected to the programmer you need to select nn the configuration "Connect under Reset"
We found the answer to our problem. Like #0___________ suggested we had an error in the board design. The NRST line had a condensator which after sufficient loading was pulling the pin up and causing the board to permanently stay in Reset until it had time to unload. After a bit of soldering the problem was solved.

STM32L4 - HAL_I2C_STATE_BUSY_RX issue

I'm having trouble with i2c communication. In my i2c bus, i have 4 boards (STM32L4 Sensortiles) and sometimes one of them blocks the bus and the others stop communicating consequently. When I reset that one that blocks the bus, all boards starting working again.
In my debugs, this bug happened when HAL_I2C_GetState(&hi2c3) returns HAL_I2C_STATE_BUSY_RX, even my others i2c functions HAL_I2C_Slave_Receive_IT and HAL_I2C_Slave_Transmit_IT returns HAL_OK:
Any ideas what Im doing wrong? Thanks.
Using libraries like HAL leads to problems like this one. Why?
HAL users feel free from knowing your hardware as magic library will do everything for you.
HAL users usually do not know how the peripheral works and how to debug it.
HAL users do not bother to go through the magic HAL function to see what the problem is.
Posting some images form debugger will not help.
What you need to do:
Get even the cheapest logic analyzer (there are $10 on ebay) and record the communication. See if the slave keeps the clock or data line low. If yes toggle the clock pin 9 times to unblock the line
If nothing helps reset the I2C peripheral using RCC register (it has to be implemented on all connected boards).

LWIP lockup on STM32

I have been having an issue in my project with LWIP. I am using a ST32F4 MCU and running with no OS. The network seems to run fine, the protocols all work, but then (usually a day or two later) the stack just stops running. It seems to happen when trying to make a new connection, but I can't confirm that because I haven't been able to locate what is causing it in the code.
Has anyone else come across this issue? I think it may be the same as this guy.
Do you call any LwIP functions from any interrupt-handlers, like UART etc?
How do you feed packets in/out of LwIP? Directly via interrupt handlers, or do you push them in from your "main-loop" ?
Lock-ups can also be signs of double free, or use-after-free of pbufs.
I also experiences that one project was unstable with wierd random locks-ups when running at the top-rated frequency of the STM32. If I clocked my STM32 at 100MHz instead of 120MHz, all my problems went away....

Atmega64 pins question. SCL/SDA same as TXD/RDX?

I'm not an Atmega pro by any means whatsoever. In fact, never worked with them. I'm trying to add an external eeprom using i2c interface to a device that uses an Atmega64 controller. From the datasheet, I see that pins 25 and 26 are labeled PD0/PD1 and are the SCL/SDA for i2c. That's fine and dandy. So then I go to the board and start tracing. I see that those go to a rocker switch with some simple circuitry that I haven't had a good look at yet, but I assumed it was something i2c compatible. But then what dawned on me was that this device has two such rockers so I expected to trace the other rocker back to the same two pins (since i2c is a bus that can have multiple slaves). Instead, they traced back to the two adjacent pins 27 and 28 labeled PD2/PD3 and RXD1/TXD1. Well, now that confused me.
I expected that I could add the memory in parallel on the SCL/SDA lines, but does what I now figure that what I discovered either means:
SCL/SDA is interchangeable with RXD1/RXD2 or
I'm confused on how SCL/SDA works and its not a bus or it can be used as a simple input?
Sadly, I'm not really sure what I'm asking. Can someone shed some light on this? Should I still be able to add the memory to SCL/SDA or probably not?
Thanks.
Steve
Nevermind, I found the answer. Its BOTH a bus and a simple input... either one or the other. In this case, its being used as an input so I can't also use it as a bus. Darn. Oh well.