Devlog 17 Fixing Uart
December 8, 2022
Log 17
I made a mistake when configuring the RX/TX GPIO pins. This session’s focus is on fixing that.
Fixing UART
When reviewing the datasheet for the Longan Nano’s MCU, I realized I made one mistake. The USART pins RX/TX are actually “alternate function” pins. This means I have some other bits in registers to set.
The first part to fix is in uart_init where we set some bits to enable the RCU clocks in RCU_APB2EN, we also need to enable the alternate function clock:
- li t2, (1 << 14) | (1 << 2) # set USART0EN (bit 14), PAEN (bit 2)
+ li t2, (1 << 14) | (1 << 2) | (1 << 0) # set USART0EN (bit 14), PAEN (bit 2), AFEN (bit 0)
Now if we compare the register with what we had before:
(gdb) x/1tw 0x40021000+0x18
0x40021018: 00000000000000000100000000000100
(gdb) c
Continuing.
Breakpoint 2, gpio_done () at fiveforths.s:224
224 ret
(gdb) x/1tw 0x40021000+0x18
0x40021018: 00000000000000000100000000000101
There’s that extra 1 set in the least significant bit.
Next, in gpio_init we need to configure the GPIO output as “AFIO”, so the TX pin should be set to 0b1011 (AFIO push-pull) instead of 0b0011 (GPIO push-pull):
- or t1, t1, (1 << 4) | (1 << 5) | (0 << 6) | (0 << 7) # set the bits 4,5,6,7 (output push-pull, max speed 50MHz)
+ or t1, t1, (1 << 4) | (1 << 5) | (0 << 6) | (1 << 7) # set the bits 4,5,6,7 (afio push-pull, max speed 50MHz)
Let’s compare with the previous register value:
(gdb) x/th 0x40010800+0x04
0x40010804: 0100010000110100
(gdb) c
Continuing.
Breakpoint 2, gpio_done () at fiveforths.s:224
224 ret
(gdb) x/th 0x40010800+0x04
0x40010804: 0100010010110100
Looking good! Next I think it’ll be a good idea to enable interrupts on the transmission and receive buffers. We’ll modify the uart_init to also set the RBNEIE (receive buffer), TCIE (transmission complete), TBEIE (transmission buffer) interrupt enable bits to 1:
- # enable receiver, transmitter, uart
- li t1, (1 << 2) | (1 << 3) | (1 << 13) # set REN (bit 2), TEN (bit 3), UEN (bit 13)
+ # enable receiver, transmitter, uart, interrupts
+ li t1, (1 << 2) | (1 << 3) | (1 << 5) | (1 << 6) | (1 << 7) | (1 << 13) # set REN, TEN, RBNEIE, TCIE, TBEIE, UEN bits to 1
And if we compare with the previous register values:
(gdb) x/th 0x40013800+0x0C
0x4001380c: 0010000000001100
(gdb) c
Continuing.
Breakpoint 2, gpio_done () at fiveforths.s:229
229 ret
(gdb) x/th 0x40013800+0x0C
0x4001380c: 0010000011101100
The bits 5, 6, and 7 are now also set to 1.
Closing thoughts
I enabled interrupts for the USART because I feel like it might be better to handle communication through that, instead of having a busy loop, but I’m not sure yet. In the next session I’ll try to figure out which approach is better for this Forth use-case, and I’ll try to get a test routine going to actually test the USART.