Add BWM_CMD_GET_UART_BAUD definition
Added a new command to get the current UART baud rate from the ESP. Signed-off-by: Niel Nielsen <nieldk@gmail.com>
This commit is contained in:
1 parent
a3b4767ff6
commit
b211a92a3a
1 file changed
+7
-2
@@ -51,6 +51,7 @@
|
||||
// SET is a HOST_CMD carrying u32 LE baud; the ESP replies with a SLAVE_RESP
|
||||
// echoing this cmd (len 0) at the OLD baud, then commits to the new baud.
|
||||
#define BWM_CMD_SET_UART_BAUD 1011
|
||||
#define BWM_CMD_GET_UART_BAUD 1009 // read back the ESP's live baud (negotiation verify)
|
||||
// Flow control (ack window) - ARM-side only, no BWM firmware change required.
|
||||
// The ESP already replies to every forward frame with a SLAVE_RESP echoing
|
||||
// cmd=SEND_FORWARD_DATA, and it sends that ack only *after* app_ble_send() has
|
||||
@@ -66,8 +67,12 @@
|
||||
// keep enough headroom to ride out delayed acks. Only ~1 response is ever
|
||||
// really in flight during an upload, so this does not risk an ESP overrun.
|
||||
#define BWM_FC_WINDOW 16 // max un-acked forward frames in flight
|
||||
#ifndef BWM_FC_ACK_TIMEOUT_SPINS
|
||||
#define BWM_FC_ACK_TIMEOUT_SPINS 200000 // safety valve: proceed if an ack is lost (avoid hard hang)
|
||||
#ifndef BWM_FC_ACK_TIMEOUT_MS
|
||||
// Hard cap (ms) on how long a forward write may block the main loop waiting for
|
||||
// acks. A spin COUNT was unbounded in wall-clock time and could hang the main
|
||||
// loop long enough that the client gives up and the device looks dead (USB still
|
||||
// enumerates on interrupts). Time-bounded => the main loop is always serviced.
|
||||
#define BWM_FC_ACK_TIMEOUT_MS 50 // safety valve: proceed if acks stall, never hard-hang
|
||||
#endif // safety valve: give up waiting for credit (avoid hard hang)
|
||||
|
||||
#define BWM_CRC16_POLY 0x1021
|
||||
|
||||
Reference in new issue
Block a user