[bug] uart issue after #883 upd

[Polaris-III]

Describe the bug
After #883 i discover that neorv32 determines any printed character as "?".
Looks like #883 uart_rx upd broke something

To Reproduce
Check "sreg" bus of "rx_engine_t" record in "neorv32_uart.vhd" file.
before #883 - sreg[9:0]
after #883 - sreg[8:0] <- where this issue happened

Expected behavior
Correct character determination

Screenshots
I push "e" button to abort autoboot sequence
Screen 1 - incorrect symbol.

Screen 2 - correct after bus fix

Environment:

• OS Ubuntu
• Xilinx xc7k325 fpga
• ft232rl as uart2usb bridge
• stock bootloader image

Hardware:

• Hardware version (1.9.8.9)

Additional context
I changed sreg back to "9 downto 0" and reimplement design - works fine

[Unike267]

Hi everyone!

I've reproduced this bug in Arty A7 35t.

As shown in the following result via CuteCom:

I hope it will be solved soon,

Cheers!

---

Edit:

Maybe the bug comes from here:

neorv32/sw/bootloader/bootloader.c

Line 395 in e2fa5c7

char c = PRINT_GETC();

Since it expects to get a char variable (8 bits) and gets other value. 🤔 That is, the problem is probably coming from the receiver.

[Unike267]

Hi,

I've fixed the issue in local changing this line:

neorv32/rtl/core/neorv32_uart.vhd

Line 289 in e2fa5c7

rx_fifo.wdata <= rx_engine.sreg(8 downto 1);

From rx_engine.sreg(8 downto 1); to rx_engine.sreg(7 downto 0);

Cheers! 😃

---

Edit:

@Polaris-III Could you test my change?

[Polaris-III]

Hi @Unike267, I'll check your solution tomorrow, thank you

[stnolting]

Seems like I have messed up the UART receiver 🙈

@Polaris-III thanks for finding this bug!

@Unike267 your fix looks great! Would you open a PR? 😉

[Unike267]

Done! 😃

When you merge it be careful with my commit description I should have written "on line 289" or "in line 289" instead of "in 289 line" sorry for that. 😅