dap: fix consecutive AP read handling and improve USB pipelining

USB/Ring buffer improvements:
- Implement ring buffer pattern matching original debugprobe for proper
  USB pipelining support (DAP_PACKET_COUNT=2)
- Add QueueCommands/ExecuteCommands support for atomic command batching
- Add comprehensive USB packet tracing for debugging

Debug shell enhancements:
- Add 'dap stats' command showing USB, ring buffer, and SWD statistics
- Add 'dap trace' command for protocol tracing
- Add 'dap dump' command for USB packet trace analysis
- Add 'dap reset' command to clear statistics

SWD/PIO improvements:
- Improve SWD turnaround and data phase configuration
- Add dormant wakeup and JTAG-to-SWD sequences for ADIv5.2+ targets
- Various PIO diagnostics enhancements

Author: vjardin

include/dap.h

@@ -66,3 +66,23 @@ uint32_t dap_vendor_command(const uint8_t *request, uint8_t *response);
  * @return Number of idle cycles
  */
 uint8_t dap_get_idle_cycles(void);
+
+/**
+ * @brief Get configured SWD turnaround period
+ *
+ * Returns the turnaround period (1-4 clock cycles) as configured
+ * by DAP_SWD_Configure command.
+ *
+ * @return Turnaround period in clock cycles
+ */
+uint8_t dap_get_swd_turnaround(void);
+
+/**
+ * @brief Get configured SWD data phase flag
+ *
+ * Returns whether data phase should be generated on WAIT/FAULT,
+ * as configured by DAP_SWD_Configure command.
+ *
+ * @return 1 if data phase should be generated, 0 otherwise
+ */
+uint8_t dap_get_swd_data_phase(void);

include/pio_swd.h

@@ -42,73 +42,84 @@
 
 #else
 /*
- * Pre-assembled PIO instructions (fallback when pioasm not available)
+ * Pre-assembled PIO instructions for probe.pio (standard version)
+ *
+ * This is the basic SWD probe program from the original debugprobe.
+ * It uses 1-bit optional side-set for SWCLK control only.
+ * Direction control is via PIO pindirs, not an external OE pin.
+ *
+ * Pin layout (consecutive from PROBE_PIN_OFFSET=12):
+ *   - GPIO12: SWCLK - Clock (side-set controlled)
+ *   - GPIO13: SWDI  - Data input from target (separate pin)
+ *   - GPIO14: SWDIO - Data output to target
+ *
+ * The level shifter has auto-direction sensing - no OE control needed.
  *
  * .program probe
  * .side_set 1 opt
  *
  * public write_cmd:
- * public turnaround_cmd:
+ * public turnaround_cmd:              ; Alias of write (offset 0)
  *     pull
- * write_bitloop:
- *     out pins, 1             [1]  side 0x0
- *     jmp x-- write_bitloop   [1]  side 0x1
+ * write_bitloop:                      ; offset 1
+ *     out pins, 1             [1]  side 0x0   ; Data output on negedge
+ *     jmp x-- write_bitloop   [1]  side 0x1   ; Captured by target on posedge
  *
  * .wrap_target
- * public get_next_cmd:
- *     pull                         side 0x0
- *     out x, 8
- *     out pindirs, 1
- *     out pc, 5
+ * public get_next_cmd:                ; offset 3
+ *     pull                         side 0x0   ; SWCLK initially low
+ *     out x, 8                                ; Get bit count
+ *     out pindirs, 1                          ; Set SWDIO direction
+ *     out pc, 5                               ; Go to command routine
  *
- * read_bitloop:
+ * read_bitloop:                       ; offset 7
  *     nop
- * public read_cmd:
- *     in pins, 1              [1]  side 0x1
+ * public read_cmd:                    ; offset 8
+ *     in pins, 1              [1]  side 0x1   ; Data captured on posedge
  *     jmp x-- read_bitloop         side 0x0
  *     push
- * .wrap
+ * .wrap                               ; offset 10
  */
 
 static const uint16_t pio_swd_program_instructions[] = {
-    /* 0: pull                       - write_cmd, turnaround_cmd */
+    /* 0: write_cmd/turnaround_cmd - pull */
     0x80a0,
-    /* 1: out pins, 1 [1] side 0x0   - write_bitloop */
-    0x6101,
-    /* 2: jmp x-- 1 [1] side 0x1     - continue write loop */
-    0x1541,
-    /* 3: pull side 0x0              - get_next_cmd (wrap_target) */
+    /* 1: out pins, 1 [1] side 0x0 */
+    0x7101,
+    /* 2: jmp x-- 1 [1] side 0x1 */
+    0x1941,
+    /* 3: get_next_cmd (wrap_target) - pull side 0x0 */
     0x90a0,
-    /* 4: out x, 8                   - get bit count */
+    /* 4: out x, 8 */
     0x6028,
-    /* 5: out pindirs, 1             - set SWDIO direction */
+    /* 5: out pindirs, 1 */
     0x6081,
-    /* 6: out pc, 5                  - jump to command routine */
+    /* 6: out pc, 5 */
     0x60a5,
-    /* 7: nop                        - read_bitloop (mov y, y) */
+    /* 7: read_bitloop - nop */
     0xa042,
-    /* 8: in pins, 1 [1] side 0x1    - read_cmd */
-    0x5501,
-    /* 9: jmp x-- 7 side 0x0         - continue read loop */
+    /* 8: read_cmd - in pins, 1 [1] side 0x1 */
+    0x5901,
+    /* 9: jmp x-- 7 side 0x0 */
     0x1047,
-    /* 10: push                      */
+    /* 10: push (wrap) */
     0x8020,
 };
 
 static const pio_program_t pio_swd_program = {
     .instructions = pio_swd_program_instructions,
     .length = ARRAY_SIZE(pio_swd_program_instructions),
-    .origin = -1,
+    .origin = 0,   /* MUST be 0 - jmp instructions have hardcoded absolute targets */
     .pio_version = 0,
 };
 
 #define PIO_SWD_PROGRAM_LENGTH  ARRAY_SIZE(pio_swd_program_instructions)
 #define PIO_SWD_WRAP_TARGET     3   /* get_next_cmd */
 #define PIO_SWD_WRAP            10  /* after push */
 
-/* Entry points (offsets) */
+/* Entry points (offsets) - turnaround_cmd = write_cmd (same entry) */
 #define PIO_SWD_OFFSET_WRITE_CMD      0
-#define PIO_SWD_OFFSET_TURNAROUND_CMD 0
+#define PIO_SWD_OFFSET_TURNAROUND_CMD 0   /* Alias of write_cmd */
 #define PIO_SWD_OFFSET_GET_NEXT_CMD   3
 #define PIO_SWD_OFFSET_READ_CMD       8
 

include/probe.h

@@ -205,6 +205,13 @@ void probe_get_pin_info(int *swclk_pin, int *swdio_pin, int *reset_pin);
  */
 bool probe_is_pio_enabled(void);
 
+/**
+ * @brief Enable/disable PIO acceleration (for debugging)
+ *
+ * @param enable true to use PIO, false for GPIO bit-bang
+ */
+void probe_set_pio_enabled(bool enable);
+
 /**
  * @brief Get system clock frequency
  *

include/probe_config.h

@@ -19,8 +19,10 @@
 /* Debug Probe Hardware Pin Assignments */
 #if defined(CONFIG_BOARD_RPI_PICO) || defined(CONFIG_BOARD_RPI_PICO_W)
 /* Pico board configuration (DEBUG_ON_PICO mode) */
+#define PROBE_PIN_OFFSET        2     /* Base pin for PIO consecutive pins */
 #define PROBE_SWCLK_PIN         2
 #define PROBE_SWDIO_PIN         3
+#define PROBE_SWDI_PIN          PROBE_SWDIO_PIN  /* Same pin on Pico (no separate input) */
 #define PROBE_SWDIR_PIN         (-1)  /* Direction control (not used on Pico) */
 #define PROBE_RESET_PIN         6     /* Target reset */
 
@@ -44,11 +46,25 @@
 #define PROBE_IO_OEN            (-1)
 
 #else
-/* Debug Probe Hardware configuration */
-#define PROBE_SWCLK_PIN         11
-#define PROBE_SWDIO_PIN         12
-#define PROBE_SWDIR_PIN         13
-#define PROBE_RESET_PIN         14
+/* Debug Probe Hardware configuration
+ *
+ * From original board_debug_probe_config.h - the Debug Probe uses:
+ *   - SWCLK (GPIO12): Clock output
+ *   - SWDI (GPIO13): Data INPUT from target (separate pin via level shifter)
+ *   - SWDIO (GPIO14): Data OUTPUT to target (via level shifter)
+ *
+ * The level shifter has auto-direction sensing - no OE or direction control.
+ * The input and output paths are separate, which is why SWDI != SWDIO.
+ *
+ * Pin offset for PIO: 12 (consecutive pins for SWCLK=12, then either SWDI or SWDIO)
+ * Note: The probe.pio program (not probe_oen.pio) is used since there's no SWDIOEN.
+ */
+#define PROBE_PIN_OFFSET        12    /* Base pin for PIO consecutive pins */
+#define PROBE_SWCLK_PIN         12    /* SWCLK output */
+#define PROBE_SWDI_PIN          13    /* SWDI - Data input from target */
+#define PROBE_SWDIO_PIN         14    /* SWDIO - Data output to target */
+#define PROBE_SWDIR_PIN         (-1)  /* No direction control */
+#define PROBE_RESET_PIN         (-1)  /* No reset pin on Debug Probe */
 
 /* JTAG pins for Debug Probe (TDI/TDO not available on hardware) */
 #define PROBE_TCK_PIN           PROBE_SWCLK_PIN  /* TCK = SWCLK */
@@ -79,8 +95,8 @@
 /* Legacy alias for backward compatibility */
 #define PROBE_DAP_LED           PROBE_DAP_RUNNING_LED
 
-/* Output enable for level shifters */
-#define PROBE_IO_OEN            10
+/* No level shifter output enable on Debug Probe - auto-direction sensing */
+#define PROBE_IO_OEN            (-1)
 #endif
 
 /* JTAG availability check */
@@ -91,7 +107,10 @@
 /* USB Configuration */
 #define CFG_TUD_HID_EP_BUFSIZE  64
 #define DAP_PACKET_SIZE         64
-#define DAP_PACKET_COUNT        4
+/* Packet count for USB buffering.
+ * Must be at least 2 for ring buffer logic to work (modulo N-1 check).
+ * Value of 2 matches original debugprobe and enables pyocd pipelining. */
+#define DAP_PACKET_COUNT        2
 
 /* UART Configuration */
 #ifndef CONFIG_DEBUGPROBE_UART_BAUDRATE