esp-idf/docs/en/api-reference/system/ulp_instruction_set.rst

{IDF_TARGET_NAME} ULP coprocessor instruction set
=================================================

This document provides details about the instructions used by {IDF_TARGET_NAME} ULP FSM coprocessor assembler.

ULP FSM coprocessor has 4 16-bit general purpose registers, labeled R0, R1, R2, R3. It also has an 8-bit counter register (stage_cnt) which can be used to implement loops. Stage count register is accessed using special instructions.

ULP coprocessor can access 8k bytes of RTC_SLOW_MEM memory region. Memory is addressed in 32-bit word units. It can also access peripheral registers in RTC_CNTL, RTC_IO, and SENS peripherals.

All instructions are 32-bit. Jump instructions, ALU instructions, peripheral register and memory access instructions are executed in 1 cycle. Instructions which work with peripherals (TSENS, ADC, I2C) take variable number of cycles, depending on peripheral operation.

The instruction syntax is case insensitive. Upper and lower case letters can be used and intermixed arbitrarily. This is true both for register names and instruction names.

.. _ulp-fsm-addressing:

Note about addressing
---------------------
{IDF_TARGET_NAME} ULP FSM coprocessor's ``JUMP``, ``ST``, ``LD`` family of instructions expect the address argument to be expressed in the following way depending on the type of address argument used:

- When the address argument is presented as a label then the instruction expects the address to be expressed as 32-bit words.

  Consider the following example program::

    entry:
            NOP
            NOP
            NOP
            NOP
    loop:
            MOVE R1, loop
            JUMP R1

  When this program is assembled and linked, address of label ``loop`` will be equal to 16 (expressed in bytes). However `JUMP` instruction expects the address stored in register ``R1`` to be expressed in 32-bit words. To account for this common use case, the assembler will convert the address of label `loop` from bytes to words, when generating the ``MOVE`` instruction. Hence, the code generated code will be equivalent to::

    0000    NOP
    0004    NOP
    0008    NOP
    000c    NOP
    0010    MOVE R1, 4
    0014    JUMP R1

- The other case is when the argument of ``MOVE`` instruction is not a label but a constant. In this case assembler will **use the value as is**, without any conversion::

            .set        val, 0x10
            MOVE        R1, val

  In this case, value loaded into ``R1`` will be ``0x10``.

  However, when an immediate value is used as an offset in ``LD`` and ``ST`` instructions, the assembler considers the address argument in bytes and converts it to 32-bit words before executing the instruction::

            ST R1, R2, 4        // offset = 4 bytes; Mem[R2 + 4 / 4] = R1

  In this case, the value in R1 is stored at the memory location pointed by [R2 + offset / 4]

  Consider the following code::

            .global array
    array:  .long 0
            .long 0
            .long 0
            .long 0

            MOVE R1, array
            MOVE R2, 0x1234
            ST R2, R1, 0      // write value of R2 into the first array element,
                              // i.e. array[0]

            ST R2, R1, 4      // write value of R2 into the second array element
                              // (4 byte offset), i.e. array[1]

            ADD R1, R1, 2     // this increments address by 2 words (8 bytes)
            ST R2, R1, 0      // write value of R2 into the third array element,
                              // i.e. array[2]

Note about instruction execution time
-------------------------------------

ULP coprocessor is clocked from RTC_FAST_CLK, which is normally derived from the internal 8MHz oscillator. Applications which need to know exact ULP clock frequency can calibrate it against the main XTAL clock::

    #include "soc/rtc.h"

    // calibrate 8M/256 clock against XTAL, get 8M/256 clock period
    uint32_t rtc_8md256_period = rtc_clk_cal(RTC_CAL_8MD256, 100);
    uint32_t rtc_fast_freq_hz = 1000000ULL * (1 << RTC_CLK_CAL_FRACT) * 256 / rtc_8md256_period;

ULP coprocessor needs certain number of clock cycles to fetch each instruction, plus certain number of cycles to execute it, depending on the instruction. See description of each instruction below for details on the execution time.

Instruction fetch time is:

- 2 clock cycles — for instructions following ALU and branch instructions.
- 4 clock cycles — in other cases.

Note that when accessing RTC memories and RTC registers, ULP coprocessor has lower priority than the main CPUs. This means that ULP coprocessor execution may be suspended while the main CPUs access same memory region as the ULP.

.. only:: esp32s2 or esp32s3

    Difference between ESP32 ULP and {IDF_TARGET_NAME} ULP Instruction sets
    -----------------------------------------------------------------------

    Compared to the ESP32 ULP FSM coprocessor, the {IDF_TARGET_NAME} ULP FSM coprocessor has an extended instruction set. The {IDF_TARGET_NAME} ULP FSM is not binary compatible with ESP32 ULP FSM,
    but a ESP32 ULP FSM  assembled program is expected to work on the {IDF_TARGET_NAME} ULP FSM after rebuilding.
    The list of the new instructions that was added to the {IDF_TARGET_NAME} ULP FSM is: ``LDL``, ``LDH``, ``STL``, ``STH``, ``ST32``, ``STO``, ``STI``, ``STI32``.


The detailed description of all instructions is presented below:

**NOP** - no operation
----------------------

**Syntax**
  **NOP**
**Operands**
  None
**Cycles**
  2 cycle to execute, 4 cycles to fetch next instruction
**Description**
  No operation is performed. Only the PC is incremented.

**Example**::

  1:    NOP


**ADD** - Add to register
-------------------------

**Syntax**
    **ADD**      *Rdst, Rsrc1, Rsrc2*

    **ADD**      *Rdst, Rsrc1, imm*


**Operands**
  - *Rdst* - Register R[0..3]
  - *Rsrc1* - Register R[0..3]
  - *Rsrc2* - Register R[0..3]
  - *Imm* - 16-bit signed value

**Cycles**
  2 cycles to execute, 4 cycles to fetch next instruction

**Description**
  The instruction adds source register to another source register or to a 16-bit signed value and stores the result in the destination register.

**Examples**::

  1:    ADD R1, R2, R3        // R1 = R2 + R3

  2:    Add R1, R2, 0x1234    // R1 = R2 + 0x1234

  3:    .set value1, 0x03     // constant value1=0x03
        Add R1, R2, value1    // R1 = R2 + value1

  4:    .global label         // declaration of variable label
        add R1, R2, label     // R1 = R2 + label
          ...
        label: nop            // definition of variable label


**SUB** - Subtract from register
--------------------------------

**Syntax**
  **SUB** *Rdst, Rsrc1, Rsrc2*

  **SUB** *Rdst, Rsrc1, imm*

**Operands**
  - *Rdst*  - Register R[0..3]
  - *Rsrc1* - Register R[0..3]
  - *Rsrc2* - Register R[0..3]
  - *Imm*   - 16-bit signed value

**Cycles**
  2 cycles to execute, 4 cycles to fetch next instruction

**Description**
  The instruction subtracts the source register from another source register or subtracts a 16-bit signed value from a source register, and stores the result to the destination register.

**Examples**::

  1:         SUB R1, R2, R3             // R1 = R2 - R3

  2:         sub R1, R2, 0x1234         // R1 = R2 - 0x1234

  3:         .set value1, 0x03          // constant value1=0x03
             SUB R1, R2, value1         // R1 = R2 - value1
  4:         .global label              // declaration of variable label
             SUB R1, R2, label          // R1 = R2 - label
               ....
    label:   nop                        // definition of variable label


**AND** - Bitwise logical AND of two operands
---------------------------------------------

**Syntax**
    **AND** *Rdst, Rsrc1, Rsrc2*

    **AND** *Rdst, Rsrc1, imm*

**Operands**
  - *Rdst* - Register R[0..3]
  - *Rsrc1* - Register R[0..3]
  - *Rsrc2* - Register R[0..3]
  - *Imm* - 16-bit signed value

**Cycles**
  2 cycles to execute, 4 cycles to fetch next instruction

**Description**
  The instruction does a bitwise logical AND of a source register and another source register or a 16-bit signed value and stores the result to the destination register.

**Examples**::

  1:        AND R1, R2, R3          // R1 = R2 & R3

  2:        AND R1, R2, 0x1234      // R1 = R2 & 0x1234

  3:        .set value1, 0x03       // constant value1=0x03
            AND R1, R2, value1      // R1 = R2 & value1

  4:        .global label           // declaration of variable label
            AND R1, R2, label       // R1 = R2 & label
                ...
    label:  nop                     // definition of variable label


**OR** - Bitwise logical OR of two operands
-------------------------------------------

**Syntax**
  **OR** *Rdst, Rsrc1, Rsrc2*

  **OR** *Rdst, Rsrc1, imm*

**Operands**
  - *Rdst* - Register R[0..3]
  - *Rsrc1* - Register R[0..3]
  - *Rsrc2* - Register R[0..3]
  - *Imm* - 16-bit signed value

**Cycles**
  2 cycles to execute, 4 cycles to fetch next instruction

**Description**
  The instruction does a bitwise logical OR of a source register and another source register or a 16-bit signed value and stores the result to the destination register.

**Examples**::

  1:       OR R1, R2, R3           // R1 = R2 || R3

  2:       OR R1, R2, 0x1234       // R1 = R2 || 0x1234

  3:       .set value1, 0x03       // constant value1=0x03
           OR R1, R2, value1       // R1 = R2 || value1

  4:       .global label           // declaration of variable label
           OR R1, R2, label        // R1 = R2 || label
           ...
    label: nop                     // definition of variable label


**LSH** - Logical Shift Left
----------------------------

**Syntax**
  **LSH** *Rdst, Rsrc1, Rsrc2*

  **LSH** *Rdst, Rsrc1, imm*

**Operands**
  - *Rdst* - Register R[0..3]
  - *Rsrc1* - Register R[0..3]
  - *Rsrc2* - Register R[0..3]
  - *Imm* - 16-bit signed value

**Cycles**
  2 cycles to execute, 4 cycles to fetch next instruction

**Description**
   The instruction does a logical shift to left of the source register by the number of bits from another source register or a 16-bit signed value and stores the result to the destination register.

**Examples**::

  1:       LSH R1, R2, R3            // R1 = R2 << R3

  2:       LSH R1, R2, 0x03          // R1 = R2 << 0x03

  3:       .set value1, 0x03         // constant value1=0x03
           LSH R1, R2, value1        // R1 = R2 << value1

  4:       .global label             // declaration of variable label
           LSH R1, R2, label         // R1 = R2 << label
           ...
    label:  nop                      // definition of variable label


**RSH** - Logical Shift Right
-----------------------------

**Syntax**
   **RSH** *Rdst, Rsrc1, Rsrc2*

   **RSH** *Rdst, Rsrc1, imm*

**Operands**
  *Rdst* - Register R[0..3]
  *Rsrc1* - Register R[0..3]
  *Rsrc2* - Register R[0..3]
  *Imm* - 16-bit signed value

**Cycles**
  2 cycles to execute, 4 cycles to fetch next instruction

**Description**
  The instruction does a logical shift to right of a source register by the number of bits from another source register or a 16-bit signed value and stores the result to the destination register.

**Examples**::

  1:        RSH R1, R2, R3              // R1 = R2 >> R3

  2:        RSH R1, R2, 0x03            // R1 = R2 >> 0x03

  3:        .set value1, 0x03           // constant value1=0x03
            RSH R1, R2, value1          // R1 = R2 >> value1

  4:        .global label               // declaration of variable label
            RSH R1, R2, label           // R1 = R2 >> label
    label:  nop                         // definition of variable label


**MOVE** – Move to register
---------------------------

**Syntax**
  **MOVE**     *Rdst, Rsrc*

  **MOVE**     *Rdst, imm*

**Operands**
  - *Rdst* – Register R[0..3]
  - *Rsrc* – Register R[0..3]
  - *Imm*  – 16-bit signed value

**Cycles**
  2 cycles to execute, 4 cycles to fetch next instruction

**Description**
   The instruction moves the value from the source register or a 16-bit signed value to the destination register.

.. note::
    Note that when a label is used as an immediate, the address of the label will be converted from bytes to words. This is because LD, ST, and JUMP instructions expect the address register value to be expressed in words rather than bytes. See the section :ref:`ulp-fsm-addressing` for more details.


**Examples**::

  1:        MOVE       R1, R2            // R1 = R2

  2:        MOVE       R1, 0x03          // R1 = 0x03

  3:        .set       value1, 0x03      // constant value1=0x03
            MOVE       R1, value1        // R1 = value1

  4:        .global     label            // declaration of label
            MOVE        R1, label        // R1 = address_of(label) / 4
            ...
    label:  nop                          // definition of label


**ST** – Store data to the memory
---------------------------------

**Syntax**
  **ST**     *Rsrc, Rdst, offset*

**Operands**
  - *Rsrc* – Register R[0..3], holds the 16-bit value to store
  - *Rdst* – Register R[0..3], address of the destination, in 32-bit words
  - *Offset* – 13-bit signed value, offset in bytes

**Cycles**
  4 cycles to execute, 4 cycles to fetch next instruction

**Description**
  The instruction stores the 16-bit value of Rsrc to the lower half-word of memory with address Rdst+offset. The upper half-word is written with the current program counter (PC) (expressed in words, shifted left by 5 bits) OR'd with Rdst (0..3)::

    Mem[Rdst + offset / 4]{31:0} = {PC[10:0], 3'b0, Rdst, Rsrc[15:0]}

  The application can use the higher 16 bits to determine which instruction in the ULP program has written any particular word into memory.

.. note::
    Note that the offset specified in bytes is converted to a 32-bit word offset before execution. See the section :ref:`ulp-fsm-addressing` for more details.

**Examples**::

  1:        ST  R1, R2, 0x12        // MEM[R2 + 0x12 / 4] = R1

  2:        .data                   // Data section definition
    Addr1:  .word     123           // Define label Addr1 16 bit
            .set      offs, 0x00    // Define constant offs
            .text                   // Text section definition
            MOVE      R1, 1         // R1 = 1
            MOVE      R2, Addr1     // R2 = Addr1
            ST        R1, R2, offs  // MEM[R2 +  0 / 4] = R1
                                    // MEM[Addr1 + 0] will be 32'h600001

.. only:: esp32s2 or esp32s3

    **STL** – Store data to the lower 16 bits of 32-bit memory
    ----------------------------------------------------------------

    **Syntax**
      **STL**    *Rsrc, Rdst, offset, Label*

    **Operands**
      - *Rsrc* – Register R[0..3], holds the 16-bit value to store
      - *Rdst* – Register R[0..3], address of the destination, in 32-bit words
      - *Offset* – 11-bit signed value, offset in bytes
      - *Label* – 2-bit user defined unsigned value

    **Cycles**
      4 cycles to execute, 4 cycles to fetch next instruction

    **Description**
      The instruction stores the 16-bit value of Rsrc to the lower half-word of the memory with address [Rdst + offset / 4]::

        Mem[Rdst + offset / 4]{15:0} = {Rsrc[15:0]}
        Mem[Rdst + offset / 4]{15:0} = {Label[1:0],Rsrc[13:0]}

      The ``ST`` and the ``STL`` commands can be used interchangeably and have been provided to maintain backward compatibility with previous versions of the ULP core.

    .. note::
        Note that the offset specified in bytes is converted to a 32-bit word offset before execution. See the section :ref:`ulp-fsm-addressing` for more details.

    **Examples**::

      1:        STL  R1, R2, 0x12           // MEM[R2 + 0x12 / 4] = R1

      2:        .data                       // Data section definition
        Addr1:  .word     123               // Define label Addr1 16 bit
                .set      offs, 0x00        // Define constant offs
                .text                       // Text section definition
                MOVE      R1, 1             // R1 = 1
                MOVE      R2, Addr1         // R2 = Addr1
                STL       R1, R2, offs      // MEM[R2 +  0 / 4] = R1
                                            // MEM[Addr1 + 0] will be 32'hxxxx0001
      3:
                MOVE      R1, 1             // R1 = 1
                STL       R1, R2, 0x12, 1   // MEM[R2 + 0x12 / 4] = 0xxxxx4001


    **STH** – Store data to the higher 16 bits of 32-bit memory
    ------------------------------------------------------------

    **Syntax**
      **STH**     *Rsrc, Rdst, offset, Label*

    **Operands**
      - *Rsrc* – Register R[0..3], holds the 16-bit value to store
      - *Rdst* – Register R[0..3], address of the destination, in 32-bit words
      - *Offset* – 11-bit signed value, offset in bytes
      - *Label* – 2-bit user defined unsigned value

    **Cycles**
      4 cycles to execute, 4 cycles to fetch next instruction

    **Description**
      The instruction stores the 16-bit value of Rsrc to the upper half-word of memory with address [Rdst + offset / 4]::

        Mem[Rdst + offset / 4]{31:16} = {Rsrc[15:0]}
        Mem[Rdst + offset / 4]{31:16} = {Label[1:0],Rsrc[13:0]}

    .. note::
        Note that the offset specified in bytes is converted to a 32-bit word offset before execution. See the section :ref:`ulp-fsm-addressing` for more details.

    **Examples**::

      1:        STH  R1, R2, 0x12           // MEM[R2 + 0x12 / 4][31:16] = R1

      2:        .data                       // Data section definition
        Addr1:  .word     123               // Define label Addr1 16 bit
                .set      offs, 0x00        // Define constant offs
                .text                       // Text section definition
                MOVE      R1, 1             // R1 = 1
                MOVE      R2, Addr1         // R2 = Addr1
                STH       R1, R2, offs      // MEM[R2 +  0 / 4] = R1
                                            // MEM[Addr1 + 0] will be 32'h0001xxxx
      3:
                MOVE      R1, 1             // R1 = 1
                STH       R1, R2, 0x12, 1   // MEM[R2 + 0x12 / 4] 0x4001xxxx


    **ST32** – Store 32-bits data to the 32-bits memory
    ---------------------------------------------------

    **Syntax**
      **ST32**     *Rsrc, Rdst, offset, Label*

    **Operands**
      - *Rsrc* – Register R[0..3], holds the 16-bit value to store
      - *Rdst* – Register R[0..3], address of the destination, in 32-bit words
      - *Offset* – 11-bit signed value, offset in bytes
      - *Label* – 2-bit user defined unsigned value

    **Cycles**
      4 cycles to execute, 4 cycles to fetch next instruction

    **Description**
      The instruction stores 11 bits of the PC value, label value and the 16-bit value of Rsrc to the 32-bit memory with address [Rdst + offset / 4]::

        Mem[Rdst + offset / 4]{31:0} = {PC[10:0],0[2:0],Label[1:0],Rsrc[15:0]}

    .. note::
        Note that the offset specified in bytes is converted to a 32-bit word offset before execution. See the section :ref:`ulp-fsm-addressing` for more details.

    **Examples**::

      1:        ST32  R1, R2, 0x12, 0       // MEM[R2 + 0x12 / 4][31:0] = {PC[10:0],0[2:0],Label[1:0],Rsrc[15:0]}

      2:        .data                       // Data section definition
        Addr1:  .word     123               // Define label Addr1 16 bit
                .set      offs, 0x00        // Define constant offs
                .text                       // Text section definition
                MOVE      R1, 1             // R1 = 1
                MOVE      R2, Addr1         // R2 = Addr1
                ST32      R1, R2, offs, 1   // MEM[R2 +  0] = {PC[10:0],0[2:0],Label[1:0],Rsrc[15:0]}
                                            // MEM[Addr1 + 0] will be 32'h00010001


    **STO** – Set offset value for auto increment operation
    -------------------------------------------------------

    **Syntax**
      **STO**     *offset*

    **Operands**
      - *Offset* – 11-bit signed value, offset in bytes

    **Cycles**
      4 cycles to execute, 4 cycles to fetch next instruction

    **Description**
      The instruction sets the 16-bit value to the offset register::

        offset = value / 4

    .. note::
        Note that the offset specified in bytes is converted to a 32-bit word offset before execution. See the section :ref:`ulp-fsm-addressing` for more details.

    **Examples**::

      1:        STO  0x12               // Offset = 0x12 / 4

      2:        .data                   // Data section definition
        Addr1:  .word     123           // Define label Addr1 16 bit
                .set      offs, 0x00    // Define constant offs
                .text                   // Text section definition
                STO       offs          // Offset = 0x00


    **STI** – Store data to the 32-bits memory with auto increment of predefined offset address
    -------------------------------------------------------------------------------------------

    **Syntax**
      **STI**     *Rsrc, Rdst, Label*

    **Operands**
      - *Rsrc* – Register R[0..3], holds the 16-bit value to store
      - *Rdst* – Register R[0..3], address of the destination, in 32-bit words
      - *Label* – 2-bit user defined unsigned value

    **Cycles**
      4 cycles to execute, 4 cycles to fetch next instruction

    **Description**
      The instruction stores the 16-bit value of Rsrc to the lower and upper half-word of memory with address [Rdst + offset / 4].
      The offset value is auto incremented when the STI instruction is called twice. Make sure to execute the ``STO`` instruction
      to set the offset value before executing the STI instruction::

        Mem[Rdst + offset / 4]{15:0/31:16} = {Rsrc[15:0]}
        Mem[Rdst + offset / 4]{15:0/31:16} = {Label[1:0],Rsrc[13:0]}

    **Examples**::

      1:        STO  4                  // Set offset to 4
                STI  R1, R2             // MEM[R2 + 4 / 4][15:0] = R1
                STI  R1, R2             // MEM[R2 + 4 / 4][31:16] = R1
                                        // offset += (1 * 4) //offset is incremented by 1 word
                STI  R1, R2             // MEM[R2 + 8 / 4][15:0] = R1
                STI  R1, R2             // MEM[R2 + 8 / 4][31:16] = R1


    **STI32** – Store 32-bits data to the 32-bits memory with auto increment of adress offset
    -----------------------------------------------------------------------------------------

    **Syntax**
      **STI32**     *Rsrc, Rdst, Label*

    **Operands**
      - *Rsrc* – Register R[0..3], holds the 16-bit value to store
      - *Rdst* – Register R[0..3], address of the destination, in 32-bit words
      - *Label* – 2-bit user defined unsigned value

    **Cycles**
      4 cycles to execute, 4 cycles to fetch next instruction

    **Description**
      The instruction stores 11 bits of the PC value, label value and the 16-bit value of Rsrc to the 32-bit memory with address [Rdst + offset / 4].
      The offset value is auto incremented each time the STI32 instruction is called. Make sure to execute the ``STO`` instruction
      to set the offset value before executing the STI32 instruction::

        Mem[Rdst + offset / 4]{31:0} = {PC[10:0],0[2:0],Label[1:0],Rsrc[15:0]}

    **Examples**::

      1:        STO    0x12
                STI32  R1, R2, 0    // MEM[R2 + 0x12 / 4][31:0] = {PC[10:0],0[2:0],Label[1:0],Rsrc[15:0]}
                                    // offset += (1 * 4) //offset is incremented by 1 word
                STI32  R1, R2, 0    // MEM[R2 + 0x16 / 4][31:0] = {PC[10:0],0[2:0],Label[1:0],Rsrc[15:0]}


**LD** – Load data from the memory
----------------------------------

**Syntax**
   **LD**      *Rdst, Rsrc, offset*

**Operands**
   - *Rdst* – Register R[0..3], destination
   - *Rsrc* – Register R[0..3], holds address of destination, in 32-bit words
   - *Offset* – 13-bit signed value, offset in bytes

**Cycles**
  4 cycles to execute, 4 cycles to fetch next instruction

**Description**
   The instruction loads the lower 16-bit half-word from memory with address [Rsrc + offset / 4] into the destination register Rdst::

     Rdst[15:0] = Mem[Rsrc + offset / 4][15:0]

.. note::
    Note that the offset specified in bytes is converted to a 32-bit word offset before execution. See the section :ref:`ulp-fsm-addressing` for more details.

**Examples**::

  1:        LD  R1, R2, 0x12            // R1 = MEM[R2 + 0x12 / 4]

  2:        .data                       // Data section definition
    Addr1:  .word     123               // Define label Addr1 16 bit
            .set      offs, 0x00        // Define constant offs
            .text                       // Text section definition
            MOVE      R1, 1             // R1 = 1
            MOVE      R2, Addr1         // R2 = Addr1 / 4 (address of label is converted into words)
            LD        R1, R2, offs      // R1 = MEM[R2 +  0]
                                        // R1 will be 123

.. only:: esp32s2 or esp32s3

    **LDL** – Load data from the lower half-word of the 32-bit memory
    ------------------------------------------------------------------

    **Syntax**
       **LDL**     *Rdst, Rsrc, offset*

    **Operands**
       - *Rdst* – Register R[0..3], destination
       - *Rsrc* – Register R[0..3], holds address of destination, in 32-bit words
       - *Offset* – 13-bit signed value, offset in bytes

    **Cycles**
      4 cycles to execute, 4 cycles to fetch next instruction

    **Description**
       The instruction loads the lower 16-bit half-word from memory with address [Rsrc + offset / 4] into the destination register Rdst::

         Rdst[15:0] = Mem[Rsrc + offset / 4][15:0]

       The ``LD`` and the ``LDL`` commands can be used interchangeably and have been provided to maintain backward compatibility with previous versions of the ULP core.

    .. note::
        Note that the offset specified in bytes is converted to a 32-bit word offset before execution. See the section :ref:`ulp-fsm-addressing` for more details.

    **Examples**::

      1:        LDL  R1, R2, 0x12           // R1 = MEM[R2 + 0x12 / 4]

      2:        .data                       // Data section definition
        Addr1:  .word     123               // Define label Addr1 16 bit
                .set      offs, 0x00        // Define constant offs
                .text                       // Text section definition
                MOVE      R1, 1             // R1 = 1
                MOVE      R2, Addr1         // R2 = Addr1 / 4 (address of label is converted into words)
                LDL       R1, R2, offs      // R1 = MEM[R2 +  0]
                                            // R1 will be 123


    **LDH** – Load data from upper half-word of the 32-bit memory
    --------------------------------------------------------------

    **Syntax**
       **LDH**     *Rdst, Rsrc, offset*

    **Operands**
       - *Rdst*  – Register R[0..3], destination
       - *Rsrc* – Register R[0..3], holds address of destination, in 32-bit words
       - *Offset* – 13-bit signed value, offset in bytes

    **Cycles**
      4 cycles to execute, 4 cycles to fetch next instruction

    **Description**
       The instruction loads the upper 16-bit half-word from memory with address [Rsrc + offset / 4] into the destination register Rdst::

         Rdst[15:0] = Mem[Rsrc + offset / 4][15:0]

    .. note::
        Note that the offset specified in bytes is converted to a 32-bit word offset before execution. See the section :ref:`ulp-fsm-addressing` for more details.

    **Examples**::

      1:        LDH  R1, R2, 0x12           // R1 = MEM[R2 + 0x12 / 4]

      2:        .data                       // Data section definition
        Addr1:  .word     0x12345678        // Define label Addr1 16 bit
                .set      offs, 0x00        // Define constant offs
                .text                       // Text section definition
                MOVE      R1, 1             // R1 = 1
                MOVE      R2, Addr1         // R2 = Addr1 / 4 (address of label is converted into words)
                LDH       R1, R2, offs      // R1 = MEM[R2 +  0]
                                            // R1 will be 0x1234


**JUMP** – Jump to an absolute address
--------------------------------------

**Syntax**
  **JUMP**        *Rdst*

  **JUMP**        *ImmAddr*

  **JUMP**        *Rdst, Condition*

  **JUMP**        *ImmAddr, Condition*


**Operands**
  - *Rdst* – Register R[0..3] containing address to jump to (expressed in 32-bit words)
  - *ImmAddr* – 13 bits address (expressed in bytes), aligned to 4 bytes
  - *Condition*:
     - EQ – jump if last ALU operation result was zero
     - OV – jump if last ALU has set overflow flag

**Cycles**
  2 cycles to execute, 2 cycles to fetch next instruction

**Description**
  The instruction makes jump to the specified address. Jump can be either unconditional or based on an ALU flag.

**Examples**::

  1:        JUMP       R1            // Jump to address in R1 (address in R1 is in 32-bit words)

  2:        JUMP       0x120, EQ     // Jump to address 0x120 (in bytes) if ALU result is zero

  3:        JUMP       label         // Jump to label
            ...
    label:  nop                      // Definition of label

  4:        .global    label         // Declaration of global label

            MOVE       R1, label     // R1 = label (value loaded into R1 is in words)
            JUMP       R1            // Jump to label
            ...
    label:  nop                      // Definition of label


**JUMPR** – Jump to a relative offset (condition based on R0)
-------------------------------------------------------------

**Syntax**
   **JUMPR**        *Step, Threshold, Condition*

**Operands**
   - *Step*          – relative shift from current position, in bytes
   - *Threshold*     – threshold value for branch condition
   - *Condition*:
      - *EQ* (equal) – jump if value in R0 == threshold
      - *LT* (less than) – jump if value in R0 < threshold
      - *LE* (less or equal) – jump if value in R0 <= threshold
      - *GT* (greater than) – jump if value in R0 > threshold
      - *GE* (greater or equal) – jump if value in R0 >= threshold

**Cycles**
  Conditions *EQ*, *GT* and *LT*: 2 cycles to execute, 2 cycles to fetch next instruction
  Conditions *LE* and *GE* are implemented in the assembler using two **JUMPR** instructions::

    // JUMPR target, threshold, LE is implemented as:

             JUMPR target, threshold, EQ
             JUMPR target, threshold, LT

    // JUMPR target, threshold, GE is implemented as:

             JUMPR target, threshold, EQ
             JUMPR target, threshold, GT

  Therefore the execution time will depend on the branches taken: either 2 cycles to execute + 2 cycles to fetch, or 4 cycles to execute + 4 cycles to fetch.

**Description**
   The instruction makes a jump to a relative address if condition is true. Condition is the result of comparison of R0 register value and the threshold value.

**Examples**::

  1:pos:    JUMPR       16, 20, GE   // Jump to address (position + 16 bytes) if value in R0 >= 20

  2:        // Down counting loop using R0 register
            MOVE        R0, 16       // load 16 into R0
    label:  SUB         R0, R0, 1    // R0--
            NOP                      // do something
            JUMPR       label, 1, GE // jump to label if R0 >= 1


**JUMPS** – Jump to a relative address (condition based on stage count)
-----------------------------------------------------------------------

**Syntax**
   **JUMPS**        *Step, Threshold, Condition*

**Operands**
   - *Step*       – relative shift from current position, in bytes
   - *Threshold*  – threshold value for branch condition
   - *Condition*:
       - *EQ* (equal) – jump if value in stage_cnt == threshold
       - *LT* (less than) –  jump if value in stage_cnt < threshold
       - *LE* (less or equal) - jump if value in stage_cnt <= threshold
       - *GT* (greater than) –  jump if value in stage_cnt > threshold
       - *GE* (greater or equal) — jump if value in stage_cnt >= threshold

**Cycles**
  2 cycles to execute, 2 cycles to fetch next instruction::

    // JUMPS target, threshold, EQ is implemented as:

             JUMPS next, threshold, LT
             JUMPS target, threshold, LE
    next:

    // JUMPS target, threshold, GT is implemented as:

             JUMPS next, threshold, LE
             JUMPS target, threshold, GE
    next:

  Therefore the execution time will depend on the branches taken: either 2 cycles to execute + 2 cycles to fetch, or 4 cycles to execute + 4 cycles to fetch.

**Description**
    The instruction makes a jump to a relative address if condition is true. Condition is the result of comparison of count register value and threshold value.

**Examples**::

   1:pos:    JUMPS     16, 20, EQ     // Jump to (position + 16 bytes) if stage_cnt == 20

   2:        // Up counting loop using stage count register
             STAGE_RST                  // set stage_cnt to 0
     label:  STAGE_INC  1               // stage_cnt++
             NOP                        // do something
             JUMPS       label, 16, LT  // jump to label if stage_cnt < 16


**STAGE_RST** – Reset stage count register
------------------------------------------
**Syntax**
     **STAGE_RST**

**Operands**
   No operands

**Description**
   The instruction sets the stage count register to 0

**Cycles**
  2 cycles to execute, 4 cycles to fetch next instruction

**Examples**::

   1:       STAGE_RST      // Reset stage count register


**STAGE_INC** – Increment stage count register
----------------------------------------------

**Syntax**
  **STAGE_INC**      *Value*

**Operands**
   - *Value* – 8 bits value

**Cycles**
  2 cycles to execute, 4 cycles to fetch next instruction

**Description**
   The instruction increments the stage count register by the given value.

**Examples**::

  1:        STAGE_INC      10          // stage_cnt += 10

  2:        // Up counting loop example:
            STAGE_RST                  // set stage_cnt to 0
    label:  STAGE_INC  1               // stage_cnt++
            NOP                        // do something
            JUMPS      label, 16, LT   // jump to label if stage_cnt < 16


**STAGE_DEC** – Decrement stage count register
----------------------------------------------

**Syntax**
  **STAGE_DEC**      *Value*

**Operands**
   - *Value* – 8 bits value

**Cycles**
  2 cycles to execute, 4 cycles to fetch next instruction

**Description**
   The instruction decrements the stage count register by the given value.

**Examples**::

  1:        STAGE_DEC      10        // stage_cnt -= 10;

  2:        // Down counting loop example
            STAGE_RST                // set stage_cnt to 0
            STAGE_INC  16            // increment stage_cnt to 16
    label:  STAGE_DEC  1             // stage_cnt--;
            NOP                      // do something
            JUMPS      label, 0, GT  // jump to label if stage_cnt > 0


**HALT** – End the program
--------------------------

**Syntax**
  **HALT**

**Operands**
  No operands

**Cycles**
  2 cycles to execute

**Description**
   The instruction halts the ULP coprocessor and restarts the ULP wakeup timer, if it is enabled.

**Examples**::

  1:       HALT      // Halt the coprocessor


**WAKE** – Wake up the chip
---------------------------

**Syntax**
   **WAKE**

**Operands**
  No operands

**Cycles**
  2 cycles to execute, 4 cycles to fetch next instruction

**Description**
  The instruction sends an interrupt from the ULP coprocessor to the RTC controller.

  - If the SoC is in deep sleep mode, and ULP wakeup is enabled, this causes the SoC to wake up.

  - If the SoC is not in deep sleep mode, and ULP interrupt bit (RTC_CNTL_ULP_CP_INT_ENA) is set in RTC_CNTL_INT_ENA_REG register, RTC interrupt will be triggered.

  Note that before using WAKE instruction, ULP program may needs to wait until RTC controller is ready to wake up the main CPU. This is indicated using RTC_CNTL_RDY_FOR_WAKEUP bit of RTC_CNTL_LOW_POWER_ST_REG register. If WAKE instruction is executed while RTC_CNTL_RDY_FOR_WAKEUP is zero, it has no effect (wake up does not occur).

**Examples**::

  1: is_rdy_for_wakeup:                   // Read RTC_CNTL_RDY_FOR_WAKEUP bit
            READ_RTC_FIELD(RTC_CNTL_LOW_POWER_ST_REG, RTC_CNTL_RDY_FOR_WAKEUP)
            AND r0, r0, 1
            JUMP is_rdy_for_wakeup, eq    // Retry until the bit is set
            WAKE                          // Trigger wake up
            REG_WR 0x006, 24, 24, 0       // Stop ULP timer (clear RTC_CNTL_ULP_CP_SLP_TIMER_EN)
            HALT                          // Stop the ULP program
            // After these instructions, SoC will wake up,
            // and ULP will not run again until started by the main program.


.. only:: esp32

    **SLEEP** – set ULP wakeup timer period
    ---------------------------------------

    **Syntax**
      **SLEEP**   *sleep_reg*

    **Operands**
       - *sleep_reg* – 0..4, selects one of ``SENS_ULP_CP_SLEEP_CYCx_REG`` registers.

    **Cycles**
      2 cycles to execute, 4 cycles to fetch next instruction

    **Description**
       The instruction selects which of the ``SENS_ULP_CP_SLEEP_CYCx_REG`` (x = 0..4) register values is to be used by the ULP wakeup timer as wakeup period. By default, the value from ``SENS_ULP_CP_SLEEP_CYC0_REG`` is used.

    **Examples**::

      1:        SLEEP     1         // Use period set in SENS_ULP_CP_SLEEP_CYC1_REG

      2:        .set sleep_reg, 4   // Set constant
                SLEEP  sleep_reg    // Use period set in SENS_ULP_CP_SLEEP_CYC4_REG


**WAIT** – wait some number of cycles
-------------------------------------

**Syntax**
   **WAIT**   *Cycles*

**Operands**
   - *Cycles* – number of cycles for wait

**Cycles**
  2 + *Cycles* cycles to execute, 4 cycles to fetch next instruction

**Description**
   The instruction delays for given number of cycles.

**Examples**::

  1:        WAIT     10         // Do nothing for 10 cycles

  2:        .set  wait_cnt, 10  // Set a constant
            WAIT  wait_cnt      // wait for 10 cycles

.. only:: not esp32

    **TSENS** – do measurement with temperature sensor
    --------------------------------------------------

    **Syntax**
      - **TSENS**   *Rdst, Wait_Delay*

    **Operands**
      - *Rdst* – Destination Register R[0..3], result will be stored to this register
      - *Wait_Delay* – number of cycles used to perform the measurement

    **Cycles**
      2 + *Wait_Delay* + 3 * TSENS_CLK to execute, 4 cycles to fetch next instruction

    **Description**
      The instruction performs measurement using TSENS and stores the result into a general purpose register.

    **Examples**::

      1:        TSENS     R1, 1000     // Measure temperature sensor for 1000 cycles,
                                      // and store result to R1


**ADC** – do measurement with ADC
---------------------------------

**Syntax**
  - **ADC**   *Rdst, Sar_sel, Mux*

  - **ADC**   *Rdst, Sar_sel, Mux, 0* — deprecated form

**Operands**
  - *Rdst* – Destination Register R[0..3], result will be stored to this register
  - *Sar_sel* – Select ADC: 0 = SARADC1, 1 = SARADC2

.. only:: esp32

  - *Mux*  - Enable ADC channel. Channel number is [Mux-1]. If the user passes Mux value 1, then ADC channel 0 gets used.

.. only:: esp32s2 or esp32s3

  - *Mux*  -  selected PAD, SARADC Pad[Mux-1] is enabled. If the user passes Mux value 1, then ADC pad 0 gets used.

**Cycles**
  ``23 + max(1, SAR_AMP_WAIT1) + max(1, SAR_AMP_WAIT2) + max(1, SAR_AMP_WAIT3) + SARx_SAMPLE_CYCLE + SARx_SAMPLE_BIT`` cycles to execute, 4 cycles to fetch next instruction

**Description**
  The instruction makes measurements from ADC.

**Examples**::

.. only:: esp32

   1:        ADC      R1, 0, 1      // Measure value using ADC1 channel 0 and store result into R1

.. only:: esp32s2 or esp32s3

   1:        ADC      R1, 0, 1      // Measure value using ADC1 pad 2 and store result into R1

.. only:: esp32

    **I2C_RD** - read single byte from I2C slave
    ----------------------------------------------

    **Syntax**
      - **I2C_RD**   *Sub_addr, High, Low, Slave_sel*

    **Operands**
      - *Sub_addr* – Address within the I2C slave to read.
      - *High*, *Low* — Define range of bits to read. Bits outside of [High, Low] range are masked.
      - *Slave_sel*  -  Index of I2C slave address to use.

    **Cycles**
      Execution time mostly depends on I2C communication time. 4 cycles to fetch next instruction.

    **Description**
      ``I2C_RD`` instruction reads one byte from I2C slave with index ``Slave_sel``. Slave address (in 7-bit format) has to be set in advance into `SENS_I2C_SLAVE_ADDRx` register field, where ``x == Slave_sel``.
      8 bits of read result is stored into `R0` register.

    **Examples**::

       1:        I2C_RD      0x10, 7, 0, 0      // Read byte from sub-address 0x10 of slave with address set in SENS_I2C_SLAVE_ADDR0


    **I2C_WR** - write single byte to I2C slave
    ----------------------------------------------

    **Syntax**
      - **I2C_WR**   *Sub_addr, Value, High, Low, Slave_sel*

    **Operands**
      - *Sub_addr* – Address within the I2C slave to write.
      - *Value* – 8-bit value to be written.
      - *High*, *Low* — Define range of bits to write. Bits outside of [High, Low] range are masked.
      - *Slave_sel*  -  Index of I2C slave address to use.

    **Cycles**
      Execution time mostly depends on I2C communication time. 4 cycles to fetch next instruction.

    **Description**
      ``I2C_WR`` instruction writes one byte to I2C slave with index ``Slave_sel``. Slave address (in 7-bit format) has to be set in advance into `SENS_I2C_SLAVE_ADDRx` register field, where ``x == Slave_sel``.

    **Examples**::

       1:        I2C_WR      0x20, 0x33, 7, 0, 1      // Write byte 0x33 to sub-address 0x20 of slave with address set in SENS_I2C_SLAVE_ADDR1.


**REG_RD** – read from peripheral register
------------------------------------------

**Syntax**
   **REG_RD**   *Addr, High, Low*

**Operands**
  - *Addr* – Register address, in 32-bit words
  - *High* – Register end bit number
  - *Low* – Register start bit number

**Cycles**
  4 cycles to execute, 4 cycles to fetch next instruction

**Description**
  The instruction reads up to 16 bits from a peripheral register into a general purpose register: ``R0 = REG[Addr][High:Low]``.

.. only:: esp32

  This instruction can access registers in RTC_CNTL, RTC_IO, SENS, and RTC_I2C peripherals. Address of the register, as seen from the ULP, can be calculated from the address of the same register on the DPORT bus as follows::

    addr_ulp = (addr_dport - DR_REG_RTCCNTL_BASE) / 4

.. only:: esp32s2 or esp32s3

  This instruction can access registers in RTC_CNTL, RTC_IO, SENS, and RTC_I2C peripherals. Address of the register, as seen from the ULP, can be calculated from the address of the same register on the PeriBUS1 as follows::

    addr_ulp = (addr_peribus1 - DR_REG_RTCCNTL_BASE) / 4

**Examples**::

  1:        REG_RD      0x120, 7, 4     // load 4 bits: R0 = {12'b0, REG[0x120][7:4]}


**REG_WR** – write to peripheral register
-----------------------------------------

**Syntax**
  **REG_WR**   *Addr, High, Low, Data*

**Operands**
  - *Addr* – Register address, in 32-bit words.
  - *High* – Register end bit number
  - *Low* – Register start bit number
  - *Data* – Value to write, 8 bits

**Cycles**
  8 cycles to execute, 4 cycles to fetch next instruction

**Description**
  The instruction writes up to 8 bits from an immediate data value into a peripheral register: ``REG[Addr][High:Low] = data``.

.. only:: esp32

  This instruction can access registers in RTC_CNTL, RTC_IO, SENS, and RTC_I2C peripherals. Address of the register, as seen from the ULP, can be calculated from the address of the same register on the DPORT bus as follows::

    addr_ulp = (addr_dport - DR_REG_RTCCNTL_BASE) / 4

.. only:: esp32s2 or esp32s3

  This instruction can access registers in RTC_CNTL, RTC_IO, SENS, and RTC_I2C peripherals. Address of the the register, as seen from the ULP, can be calculated from the address of the same register on the PeriBUS1 as follows::

    addr_ulp = (addr_peribus1 - DR_REG_RTCCNTL_BASE) / 4

**Examples**::

   1:        REG_WR      0x120, 7, 0, 0x10   // set 8 bits: REG[0x120][7:0] = 0x10


Convenience macros for peripheral registers access
--------------------------------------------------

ULP source files are passed through C preprocessor before the assembler. This allows certain macros to be used to facilitate access to peripheral registers.

Some existing macros are defined in ``soc/soc_ulp.h`` header file. These macros allow access to the fields of peripheral registers by their names.
Peripheral registers names which can be used with these macros are the ones defined in ``soc/rtc_cntl_reg.h``, ``soc/rtc_io_reg.h``, ``soc/sens_reg.h``, and ``soc/rtc_i2c_reg.h``.

READ_RTC_REG(rtc_reg, low_bit, bit_width)
  Read up to 16 bits from rtc_reg[low_bit + bit_width - 1 : low_bit] into R0. For example::

    #include "soc/soc_ulp.h"
    #include "soc/rtc_cntl_reg.h"

    /* Read 16 lower bits of RTC_CNTL_TIME0_REG into R0 */
    READ_RTC_REG(RTC_CNTL_TIME0_REG, 0, 16)

READ_RTC_FIELD(rtc_reg, field)
  Read from a field in rtc_reg into R0, up to 16 bits. For example::

    #include "soc/soc_ulp.h"
    #include "soc/sens_reg.h"

    /* Read 8-bit SENS_TSENS_OUT field of SENS_SAR_SLAVE_ADDR3_REG into R0 */
    READ_RTC_FIELD(SENS_SAR_SLAVE_ADDR3_REG, SENS_TSENS_OUT)

WRITE_RTC_REG(rtc_reg, low_bit, bit_width, value)
  Write immediate value into rtc_reg[low_bit + bit_width - 1 : low_bit], bit_width <= 8. For example::

    #include "soc/soc_ulp.h"
    #include "soc/rtc_io_reg.h"

    /* Set BIT(2) of RTC_GPIO_OUT_DATA_W1TS field in RTC_GPIO_OUT_W1TS_REG */
    WRITE_RTC_REG(RTC_GPIO_OUT_W1TS_REG, RTC_GPIO_OUT_DATA_W1TS_S + 2, 1, 1)

WRITE_RTC_FIELD(rtc_reg, field, value)
  Write immediate value into a field in rtc_reg, up to 8 bits. For example::

    #include "soc/soc_ulp.h"
    #include "soc/rtc_cntl_reg.h"

    /* Set RTC_CNTL_ULP_CP_SLP_TIMER_EN field of RTC_CNTL_STATE0_REG to 0 */
    WRITE_RTC_FIELD(RTC_CNTL_STATE0_REG, RTC_CNTL_ULP_CP_SLP_TIMER_EN, 0)