xref: /third_party/mesa3d/src/panfrost/bifrost/valhall/ISA.xml

<!--
  Copyright (C) 2021 Collabora Ltd.

  Permission is hereby granted, free of charge, to any person obtaining a
  copy of this software and associated documentation files (the "Software"),
  to deal in the Software without restriction, including without limitation
  the rights to use, copy, modify, merge, publish, distribute, sublicense,
  and/or sell copies of the Software, and to permit persons to whom the
  Software is furnished to do so, subject to the following conditions:

  The above copyright notice and this permission notice (including the next
  paragraph) shall be included in all copies or substantial portions of the
  Software.

  THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
  IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
  FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT.  IN NO EVENT SHALL
  THE AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
  LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM,
  OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE
  SOFTWARE.
-->

<valhall>
  <lut name="Immediates">
    <desc>
      This immediates are accessible in (almost) any instruction, provided the
      immediate mode is kept to the default. They optimize for the most common
      immediate values; any immediate listed here may be used without taking up
      a uniform slot or a register. Most integer instructions can access
      separate half-words and individual bytes via swizzles on the source.
    </desc>
    <constant desc="Zero">0x00000000</constant>
    <constant desc="All ones; integer $-1$">0xFFFFFFFF</constant>
    <constant desc="Maximum integer; floating-point NaN">0x7FFFFFFF</constant>
    <constant desc="Integers $(-2, -3, -4, -5)$">0xFAFCFDFE</constant>
    <constant desc="16-bit integer $2^8$">0x01000000</constant>
    <constant desc="Multiples of 16 $(0, 32, 0, 128)$">0x80002000</constant>
    <constant desc="Multiples of 16 $(48, 80, 96, 112)$">0x70605030</constant>
    <constant desc="Multiples of 16 $(144, 160, 176, 192)$">0xC0B0A090</constant>
    <constant desc="Integers $(0, 1, 2, 3)$">0x03020100</constant>
    <constant desc="Integers $(4, 5, 6, 7)$">0x07060504</constant>
    <constant desc="Integers $(8, 9, 10, 11)$">0x0B0A0908</constant>
    <constant desc="Integers $(12, 13, 14, 15)$">0x0F0E0D0C</constant>
    <constant desc="Integers $(16, 17, 18, 19)$">0x13121110</constant>
    <constant desc="Integers $(20, 21, 22, 23)$">0x17161514</constant>
    <constant desc="Integers $(24, 25, 26, 27)$">0x1B1A1918</constant>
    <constant desc="Integers $(28, 29, 30, 31)$">0x1F1E1D1C</constant>
    <constant desc="Float $1.0$">0x3F800000</constant>
    <constant desc="Float $0.1$">0x3DCCCCCD</constant>
    <constant desc="Float $1 / \pi$">0x3EA2F983</constant>
    <constant desc="Float $\log(2)$">0x3F317218</constant>
    <constant desc="Float $\pi$">0x40490FDB</constant>
    <constant desc="Float $0.0$">0x00000000</constant>
    <constant desc="Float $65535.0 = 2^{16} - 1$">0x477FFF00</constant>
    <constant desc="Half-float $(255.0, 256.0) = (2^8 - 1, 2^8)$">0x5C005BF8</constant>
    <constant desc="Half-float $0.1 = 1 / 10$">0x2E660000</constant>
    <constant desc="Half-float $0.25 = 2^{-2}$">0x34000000</constant>
    <constant desc="Half-float $0.5 = 2^{-1}$">0x38000000</constant>
    <constant desc="Half-float $1.0 = 2^0$">0x3C000000</constant>
    <constant desc="Half-float $2.0 = 2^1$">0x40000000</constant>
    <constant desc="Half-float $4.0 = 2^2$">0x44000000</constant>
    <constant desc="Half-float $8.0 = 2^3$">0x48000000</constant>
    <constant desc="Half-float $\pi$">0x42480000</constant>
  </lut>

  <enum name="Flow">
    <desc>
      Every Valhall instruction can wait on dependency
      slots. A few special flows are available, specified in the instruction
      metadata from this enum. The `wait0126` flow is required to wait on
      dependency slot #6 and should be set on the instruction immediately
      preceding `ATEST`. The `wait` flow should be set for barriers.
      The `discard` flow only applies to fragment shaders and is used to
      terminate helper invocations, it should be set as early as possible after
      helper invocations are no longer needed as determined by data flow
      analysis. The `end` flow is used to terminate the shader, although it
      may be overloaded by the `BLEND` instruction.

      The `reconverge` flow is required on any instruction immediately
      preceding a possible change to the mask of active threads in a subgroup.
      This includes all divergent branches, but it also includes the final
      instruction at the end of any basic block where the immediate successor
      (fallthrough) is the target of a divergent branch.
    </desc>
    <value name="None" default="true">none</value>
    <value name="Wait on slot 0">wait0</value>
    <value name="Wait on slot 1">wait1</value>
    <value name="Wait on slots 0, 1">wait01</value>
    <value name="Wait on slot 2">wait2</value>
    <value name="Wait on slots 0, 2">wait02</value>
    <value name="Wait on slots 1, 2">wait12</value>
    <value name="Wait on slots 0, 1, 2">wait012</value>
    <value name="Wait on slots 0, 1, 2, 6">wait0126</value>
    <value name="Wait on slots 0, 1, 2, 6, 7">wait</value>
    <value name="Perform branch reconverge">reconverge</value>
    <reserved/>
    <reserved/>
    <value name="Terminate discarded threads">discard</value>
    <reserved/>
    <value name="Return from shader">end</value>
  </enum>

  <enum name="FAU special page 0">
    <desc>
      Situated between the immediates hard-coded in the hardware and the
      uniforms defined purely in software, Valhall has a some special
      "constants" passing through data structures. These are encoded like the
      table of immediates, as if special constant $i$ were lookup table entry
      $32 + i$.
    </desc>
    <reserved/>
    <reserved/>
    <value desc="Warp ID and warps/core - 1">warp_id</value>
    <reserved/>
    <value desc="Bounding box maximum X/Y">framebuffer_size</value>
    <value desc="ATEST datum">atest_datum</value>
    <value desc="Sample positions">sample</value>
    <reserved/>
    <value desc="Blend descriptor 0">blend_descriptor_0</value>
    <value desc="Blend descriptor 1">blend_descriptor_1</value>
    <value desc="Blend descriptor 2">blend_descriptor_2</value>
    <value desc="Blend descriptor 3">blend_descriptor_3</value>
    <value desc="Blend descriptor 4">blend_descriptor_4</value>
    <value desc="Blend descriptor 5">blend_descriptor_5</value>
    <value desc="Blend descriptor 6">blend_descriptor_6</value>
    <value desc="Blend descriptor 7">blend_descriptor_7</value>
  </enum>

  <enum name="FAU special page 1">
    <desc>
      Situated between the immediates hard-coded in the hardware and the
      uniforms defined purely in software, Valhall has a some special
      "constants" passing through data structures. These are encoded like the
      table of immediates, as if special constant $i$ were lookup table entry
      $32 + i$.
    </desc>
    <reserved/>
    <value desc="Thread local storage base pointer">thread_local_pointer</value>
    <reserved/>
    <value desc="Workgroup local storage base pointer">workgroup_local_pointer</value>
    <reserved/>
    <reserved/>
    <reserved/>
    <value desc="Shader resource table base pointer">resource_table_pointer</value>
    <reserved/>
    <reserved/>
    <reserved/>
    <reserved/>
    <reserved/>
    <reserved/>
    <reserved/>
    <reserved/>
  </enum>

  <enum name="FAU special page 3">
    <desc>
      Situated between the immediates hard-coded in the hardware and the
      uniforms defined purely in software, Valhall has a some special
      "constants" passing through data structures. These are encoded like the
      table of immediates, as if special constant $i$ were lookup table entry
      $32 + i$.
    </desc>
    <reserved/>
    <value desc="Lane ID">lane_id</value>
    <reserved/>
    <value desc="Core ID">core_id</value>
    <reserved/>
    <reserved/>
    <reserved/>
    <reserved/>
    <reserved/>
    <reserved/>
    <reserved/>
    <reserved/>
    <reserved/>
    <reserved/>
    <reserved/>
    <value desc="Program counter">program_counter</value>
  </enum>

  <enum name="Swizzles (8-bit)">
    <value default="true">b0123</value>
    <value>b3210</value>
    <value>b0101</value>
    <value>b2323</value>
    <value>b0000</value>
    <value>b1111</value>
    <value>b2222</value>
    <value>b3333</value>
    <value>b2301</value>
    <value>b1032</value>
    <value>b0011</value>
    <value>b2233</value>
    <reserved/>
    <reserved/>
    <reserved/>
    <reserved/>
  </enum>

  <enum name="Lanes (8-bit)">
    <desc>Used to select the 2 bytes for shifts of 16-bit vectors</desc>
    <value>b02</value>
    <reserved/>
    <reserved/>
    <reserved/>
    <value>b00</value>
    <value>b11</value>
    <value>b22</value>
    <value>b33</value>
    <reserved/>
    <reserved/>
    <value>b01</value>
    <value>b23</value>
    <reserved/>
    <reserved/>
    <reserved/>
    <reserved/>
  </enum>

  <enum name="Half-swizzles (8-bit)">
    <desc>
      Used to select the 2 bytes to convert for conversions from 8-bit vectors
      to 16-bit vectors
    </desc>
    <value>b00</value>
    <value>b10</value>
    <value>b20</value>
    <value>b30</value>
    <value>b01</value>
    <value>b11</value>
    <value>b21</value>
    <value>b31</value>
    <value>b02</value>
    <value>b12</value>
    <value>b22</value>
    <value>b32</value>
    <value>b03</value>
    <value>b13</value>
    <value>b23</value>
    <value>b33</value>
  </enum>

  <enum name="Swizzles (16-bit)">
    <value>h00</value> <!-- 0,2 -->
    <value>h10</value>
    <value default="true">h01</value>
    <value>h11</value>
    <value>b00</value> <!-- 0,0 -->
    <value>b20</value> <!-- 1,1 -->
    <value>b02</value> <!-- 2,2 -->
    <value>b22</value> <!-- 3,3 -->
    <value>b11</value>
    <value>b31</value>
    <value>b13</value> <!-- 0,1 -->
    <value>b33</value> <!-- 2,3 -->
    <value>b01</value>
    <value>b23</value>
    <reserved/>
    <reserved/>
  </enum>

  <enum name="Swizzles (32-bit)">
    <value default="true">none</value>
    <reserved/>
    <value>h0</value>
    <value>h1</value>
    <value>b0</value>
    <value>b1</value>
    <value>b2</value>
    <value>b3</value>
  </enum>

  <enum name="Swizzles (64-bit)">
    <value default="true">none</value>
    <reserved/>
    <value>h0</value>
    <value>h1</value>
    <value>b0</value>
    <value>b1</value>
    <value>b2</value>
    <value>b3</value>
    <value>w0</value>
    <reserved/>
    <reserved/>
    <reserved/>
    <reserved/>
    <reserved/>
    <reserved/>
    <reserved/>
  </enum>

  <enum name="Lane (8-bit)" implied="true">
    <value>b0</value>
    <value>b1</value>
    <value>b2</value>
    <value>b3</value>
  </enum>

  <enum name="Combine">
    <desc>
      Used for the lane select of `BRANCHZ`. To use an 8-bit condition, a
      separate `ICMP` is required to cast to 16-bit.
    </desc>
    <value default="true">none</value>
    <value>h0</value>
    <value>h1</value>
    <value>and</value>
    <value>lowbits</value>
  </enum>

  <enum name="Lane (16-bit)" implied="true">
    <value>h0</value>
    <value>h1</value>
  </enum>

  <enum name="Load lane (8-bit)">
    <value default="true">b0</value>
    <value>b1</value>
    <value>b2</value>
    <value>b3</value>
    <value desc="Zero-extend to 16-bit, low-half">h0</value>
    <value desc="Zero-extend to 16-bit, high-half">h1</value>
    <value desc="Zero-extend to 32-bit">w0</value>
    <value desc="Zero-extend to 32-bit">d0</value>
  </enum>

  <enum name="Load lane (16-bit)">
    <value desc="Low half" default="true">h0</value>
    <value desc="High half">h1</value>
    <value desc="Zero-extend to 32-bit">w0</value>
    <value desc="Zero-extend to 64-bit">d0</value>
    <reserved/>
    <reserved/>
    <reserved/>
    <reserved/>
  </enum>

  <enum name="Load lane (24-bit)" implied="true">
    <value default="true">identity</value>
    <reserved/>
    <reserved/>
    <reserved/>
    <reserved/>
    <reserved/>
    <reserved/>
  </enum>

  <enum name="Load lane (32-bit)">
    <value default="true">w0</value>
    <value desc="Zero-extend to 64-bit">d0</value>
    <reserved/>
    <reserved/>
    <reserved/>
    <reserved/>
    <reserved/>
    <reserved/>
  </enum>

  <enum name="Load lane (48-bit)">
    <reserved/>
    <reserved/>
    <reserved/>
    <reserved/>
    <value default="true">identity</value>
    <reserved/>
    <reserved/>
    <reserved/>
  </enum>

  <enum name="Load lane (64-bit)">
    <reserved/>
    <reserved/>
    <reserved/>
    <reserved/>
    <reserved/>
    <reserved/>
    <reserved/>
    <value default="true">identity</value>
  </enum>

  <enum name="Load lane (96-bit)">
    <reserved/>
    <reserved/>
    <reserved/>
    <reserved/>
    <reserved/>
    <reserved/>
    <value default="true">identity</value>
    <reserved/>
  </enum>

  <enum name="Load lane (128-bit)">
    <reserved/>
    <reserved/>
    <reserved/>
    <reserved/>
    <reserved/>
    <reserved/>
    <reserved/>
    <value default="true">identity</value>
  </enum>

  <enum name="Round mode">
    <desc>Corresponds to IEEE 754 rounding modes</desc>
    <value desc="Round to nearest even" default="true">rte</value>
    <value desc="Round to positive infinity">rtp</value>
    <value desc="Round to negative infinity">rtn</value>
    <value desc="Round to zero">rtz</value>
  </enum>

  <enum name="Result type">
    <desc>
      Comparison instructions like `FCMP` return a boolean but may encode this
      boolean in a variety of ways. `i1` gives a OpenGL style `0/1` boolean.
      `m1` gives a Direct3D style `0/~0` boolean. `f1` gives a floating-point
      `0.0f / 1.0f` boolean. Switching between these modes is useful to fold a
      boolean type convert into a comparison. `u1` is used internally to
      implement 64-bit comparisons.
    </desc>
    <value desc="Integer 1">i1</value>
    <value desc="Float 1">f1</value>
    <value desc="Minus 1">m1</value>
    <value desc="Low half of 64-bit compare">u1</value>
  </enum>

  <enum name="Widen">
    <value default="true">none</value>
    <value>h0</value>
    <value>h1</value>
    <reserved/>
    <reserved/>
    <reserved/>
    <reserved/>
    <reserved/>
  </enum>

  <enum name="Clamp">
    <desc>
      Clamp applied to the destination of a floating-point instruction. Note the
      clamps may be decomposed as two independent bits for `clamp_0_inf` and
      `clamp_m1_1`, with `clamp_0_1` arising as the composition of `clamp_0_inf`
      and `clamp_m1_1` in either order.

      Clamps are implemented per the SPIR-V specification:

      $$\text{clamp} \; (x, \ell, h) = \min( \max( x, \ell ), h)$$

      The min/max functions return the other operand if one operand is NaN, and
      compare $-0 &lt; +0$. That means the following identities hold for Valhall
      clamps:

      \begin{align*}
        \text{clamp}(-0.0, 0.0, 1.0) &amp; = +0.0 \\
        \text{clamp}(-\text{NaN}, 0.0, 1.0) &amp; = +0.0 \\
        \text{clamp}(\text{NaN}, 0.0, 1.0) &amp; = +0.0 \\
        &amp; \\
        \text{clamp}(-0.0, -1.0, 1.0) &amp; = -0.0 \\
        \text{clamp}(\text{NaN}, -1.0, 1.0) &amp; = -1.0 \\
        \text{clamp}(-\text{NaN}, -1.0, 1.0) &amp; = -1.0 \\
        &amp; \\
        \max(\text{NaN}, 0.0) &amp; = +0.0 \\
        \max(-\text{NaN}, 0.0) &amp; = +0.0 \\
        \max(-0.0, 0.0) &amp; = +0.0 \\
      \end{align*}

      This behaviour is consistent with the FMin/FMax/FClamp and
      NMin/NMax/NClamp rules prescribed by SPIR-V and governed by IEEE-754. As
      a consequence, substituting these clamps for equivalent minimum/maximum
      exprssions is legal even with strict floating point rules.
    </desc>
    <value default="true" desc="Identity">none</value>
    <value desc="Clamp positive">clamp_0_inf</value>
    <value desc="Clamp to $[-1, 1]$">clamp_m1_1</value>
    <value desc="Clamp to $[0, 1]$">clamp_0_1</value>
  </enum>

  <enum name="Condition">
    <desc>
      Condition code. Type must be inferred from the instruction. IEEE 754 total
      ordering only applies to floating point compares. "Not equal" and "greater
      than or less than" are distinguished by NaN behaviour conforming to
      the IEEE 754 specification.
    </desc>
    <value desc="Equal">eq</value>
    <value desc="Greater than">gt</value>
    <value desc="Greater than or equal">ge</value>
    <value desc="Not equal">ne</value>
    <value desc="Less than">lt</value>
    <value desc="Less than or equal">le</value>
    <value desc="Greater than or less than">gtlt</value>
    <value desc="Totally ordered">total</value>
  </enum>

  <enum name="Dimension">
    <desc>Texture dimension.</desc>
    <value desc="1D or buffer">1d</value>
    <value desc="2D or 2D array">2d</value>
    <value desc="3D or 3D array">3d</value>
    <value desc="Cube map or cube map array">cube</value>
  </enum>

  <enum name="LOD mode">
    <desc>Level-of-detail selection mode in a texture instruction.</desc>
    <value desc="Set to zero">zero</value>
    <value desc="Computed based on neighboring fragments">computed</value>
    <reserved/>
    <reserved/>
    <value desc="Explicitly specified in a register">explicit</value>
    <value desc="Computed based on neighboring fragments added with bias in a register">computed_bias</value>
    <value desc="Derived from a gradient descriptor in registers">grdesc</value>
    <reserved/>
  </enum>

  <enum name="Register format">
    <desc>Format of data loaded to / stored from registers for general memory access.</desc>
    <value desc="32-bit type based on descriptor format">auto</value>
    <reserved/>
    <value desc="32-bit floats">f32</value>
    <value desc="16-bit floats">f16</value>
    <value desc="32-bit signed integers">s32</value>
    <value desc="16-bit signed integers">s16</value>
    <value desc="32-bit unsigned integers">u32</value>
    <value desc="16-bit unsigned integers">u16</value>
  </enum>

  <enum name="Staging register count" implied="true">
    <value>sr0</value>
    <value>sr1</value>
    <value>sr2</value>
    <value>sr3</value>
    <value>sr4</value>
    <value>sr5</value>
    <value>sr6</value>
    <value>sr7</value>
  </enum>

  <enum name="Staging register write count" implied="true">
    <value>write1</value>
    <value>write2</value>
    <value>write3</value>
    <value>write4</value>
    <value>write5</value>
    <value>write6</value>
    <value>write7</value>
    <value>write8</value>
  </enum>

  <enum name="Write mask">
    <reserved/>
    <value>r</value>
    <value>g</value>
    <value>rg</value>
    <value>b</value>
    <value>rb</value>
    <value>gb</value>
    <value>rgb</value>
    <value>a</value>
    <value>ra</value>
    <value>ga</value>
    <value>rga</value>
    <value>ba</value>
    <value>rba</value>
    <value>gba</value>
    <value default="true">rgba</value>
  </enum>

  <enum name="Fetch component">
    <value desc="Red">gather4_r</value>
    <value desc="Green">gather4_g</value>
    <value desc="Blue">gather4_b</value>
    <value desc="Alpha">gather4_a</value>
  </enum>

  <enum name="Register type">
    <desc>Unsized type, part of a register format.</desc>
    <reserved/>
    <value name="Float">f</value>
    <value name="Unsigned">u</value>
    <value name="Signed">s</value>
  </enum>

  <enum name="Register width">
    <desc>Untyped size, part of a register format.</desc>
    <value>16</value>
    <value>32</value>
  </enum>

  <enum name="Varying texture register width">
    <desc>
      Size of results for varying texture instructions. For dual 16-bit results
      use "16-bit".
    </desc>
    <value desc="16-bit">16</value>
    <value desc="32-bit">32</value>
    <value desc="16-bit, 32-bit">16.32</value>
    <value desc="32-bit, 32-bit">32.32</value>
  </enum>

  <enum name="Vector size">
    <desc>Number of channels loaded/stored for general memory access.</desc>
    <value default="true" desc="Scalar">none</value>
    <value desc="2 channels">v2</value>
    <value desc="3 channels">v3</value>
    <value desc="4 channels">v4</value>
  </enum>

  <enum name="Slot">
    <desc>
      Dependency slot set on a message-passing instruction that writes to
      registers. Before reading the destination, a future instruction must wait
      on the specified slot. Slot #7 is for `BARRIER` instructions only.
    </desc>
    <value desc="Slot #0">slot0</value>
    <value desc="Slot #1">slot1</value>
    <value desc="Slot #2">slot2</value>
    <reserved/>
    <reserved/>
    <reserved/>
    <reserved/>
    <value desc="Slot #7">slot7</value>
  </enum>

  <enum name="Memory access">
    <desc>Memory access hint for a `LOAD` or `STORE` instruction.</desc>
    <value desc="No hint (global)" default="true">none</value>
    <value desc="Internally streaming (position output)">istream</value>
    <value desc="Externally streaming (varying output)">estream</value>
    <value desc="Force access in discarded threads (thread local storage)">force</value>
  </enum>

  <enum name="Subgroup size">
    <desc>
      Selects the effective subgroup size from subgroup operations. The hardware
      warps are sixteen threads on Valhall, but subdividing a warp may be useful
      for API requirements. In particular, derivatives may be calculated with
      quads (four threads).
    </desc>
    <value desc="Two threads">subgroup2</value>
    <value desc="Four threads">subgroup4</value>
    <value desc="Eight threads">subgroup8</value>
    <value desc="Sixteen threads" default="true">subgroup16</value>
  </enum>

  <enum name="Lane operation">
    <desc>
      Acts as a modifier on the lane specificier for a `CLPER` instruction. The
      `accumulate` mode is required for efficient subgroup reductions.
    </desc>
    <value name="No operation" default="true">none</value>
    <value name="Exclusive-or">xor</value>
    <value name="Accumulate">accumulate</value>
    <value name="Shift">shift</value>
  </enum>

  <enum name="Inactive result">
    <desc>
      Accesses to inactive lanes (due to divergence) in a subgroup is generally
      undefined in APIs. However, the results of permuting with an inactive lane
      with `CLPER.i32` are well-defined in Valhall: they return one of the
      following values, as specified in the `CLPER.i32` instructions. Sometimes
      certain values enable small optimizations.
    </desc>
    <value name="0x00000000" default="true">zero</value>
    <value name="0xFFFFFFFF">umax</value>
    <value name="0x00000001">i1</value>
    <value name="0x00010001">v2i1</value>
    <value name="0x80000000">smin</value>
    <value name="0x7FFFFFFF">smax</value>
    <value name="0x80008000">v2smin</value>
    <value name="0x7FFF7FFF">v2smax</value>
    <value name="0x80808080">v4smin</value>
    <value name="0x7F7F7F7F">v4smax</value>
    <value name="0x3F800000">f1</value>
    <value name="0x3C003C00">v2f1</value>
    <value name="0xFF800000">infn</value>
    <value name="0x7F800000">inf</value>
    <value name="0xFC00FC00">v2infn</value>
    <value name="0x7C007C00">v2inf</value>
  </enum>

  <enum name="Mux">
    <desc>
      Condition to use for a `MUX` instruction. `neg` checks the sign bit,
      `int_zero` compares to `0x00000000`, `fp_zero` compares to $\pm 0.0$ as
      an IEEE 754 float, and `bit` checks each bit separately. The `bit` mode
      acts like an imaginary `CSEL.v32u1` instruction, and implements
      `bitselect()` in OpenCL.
    </desc>
    <value desc="Negative">neg</value>
    <value desc="Integer zero" default="true">int_zero</value>
    <value desc="Floating point zero">fp_zero</value>
    <value desc="Bitwise">bit</value>
  </enum>

  <enum name="Sample mode">
    <desc>
      Varying interpolation mode, for choosing the correct sample to
      interpolate at, allowing the `sample` and `centroid` qualifiers to be
      implemented, as well as the `interpolateAt*` functions.
    </desc>
    <value desc="Center">center</value>
    <value desc="Centroid">centroid</value>
    <value desc="Sample">sample</value>
    <value desc="Explicit">explicit</value>
  </enum>

  <enum name="Update mode">
    <desc>
      The Valhall GPU maintains hidden state when interpolating varyings, to
      allow reusing sample location calculations. The update mode of a varying
      load controls this hidden state.
    </desc>
    <value desc="Store interpolation position">store</value>
    <value desc="Retrieve interpolation position">retrieve</value>
    <reserved/>
    <value desc="Clobber saved position">clobber</value>
  </enum>

  <enum name="Sample and update mode">
    <desc>
      For fused varying/texture instructions, only the following specific
      combinations of sample and update modes are permitted.
    </desc>
    <value desc="Center, store">center_store</value>
    <value desc="Centroid, store">centroid_store</value>
    <value desc="Sample, store">sample_store</value>
    <value desc="Explicit, store">explicit_store</value>
    <value desc="Center, clobber">center_clobber</value>
    <reserved/>
    <value desc="Sample, clobber">sample_clobber</value>
    <value desc="Retrieve previous state">retrieve</value>
  </enum>

  <enum name="Source format">
    <desc>
      In-memory format of varyings.

      Note: src_flat32 is only valid with 32-bit varying instructions and
      src_flat16 is only valid with 16-bit varying instructions.
    </desc>
    <value desc="Uninterpreted 32-bit values">src_flat32</value>
    <value desc="Uninterpreted 16-bit values">src_flat16</value>
    <value desc="Interpolated 32-bit floats">src_f32</value>
    <value desc="Interpolated 16-bit floats">src_f16</value>
  </enum>

  <enum name="Atomic operation">
    <desc>
      Operation performed in a general computational atomic instruction.
    </desc>
    <reserved/>
    <reserved/>
    <value desc="Add">aadd</value>
    <reserved/>
    <reserved/>
    <reserved/>
    <reserved/>
    <reserved/>
    <value desc="Signed minimum">asmin</value>
    <value desc="Signed maximum">asmax</value>
    <value desc="Unsigned minimum">aumin</value>
    <value desc="Unsigned maximum">aumax</value>
    <value desc="Bitwise and">aand</value>
    <value desc="Bitwise or">aor</value>
    <value desc="Bitwise exclusive-or">axor</value>
    <value desc="Exchange (must return the value)">axchg</value>
  </enum>

  <enum name="Atomic operation with 1">
    <desc>
      Operation performed in a computational atomic-with-1 instruction.
    </desc>
    <value desc="Increment">ainc</value>
    <value desc="Decrement">adec</value>
    <value desc="Unsigned maximum with 1">aumax1</value>
    <value desc="Signed maximum with 1">asmax1</value>
    <value desc="Set bottom bit">aor1</value>
  </enum>

  <ins name="NOP" title="No operation" dests="0" opcode="0x00" unit="CVT">
    <desc>
      Do nothing. Useful at the start of a block for waiting on slots required
      by the first actual instruction of the block, to reconcile dependencies
      after a branch. Also useful as the sole instruction of an empty shader.
    </desc>
  </ins>

  <ins name="BRANCHZ" title="Compare to zero and branch" dests="0" opcode="0x1F" unit="CVT">
    <desc>
      Branches to a specified relative offset if its source is nonzero (default)
      or if its source is zero (if `.eq` is set). The offset is 27-bits and
      sign-extended, giving an effective range of ±26-bits. The offset is
      specified in units of instructions, relative to the *next* instruction.
      Positive offsets may be interpreted as "number of instructions to skip".
      Since Valhall instructions are 8 bytes, this operates as:

      $$PC := \begin{cases} PC + 8 \cdot (\text{offset} \; + 1) &amp; \text{if} \;
      \text{src} \stackrel{?}{=} 0 \\ PC + 8 &amp; \text{otherwise} \end{cases}$$

      Used with comparison instructions to implement control flow. Tie the
      source to a nonzero constant to implement a jump. May introduce
      divergence, so generally requires `.reconverge` flow control.
    </desc>
    <src combine="true">Value to compare against zero</src>
    <imm name="offset" start="8" size="27" signed="true"/>
    <conservative/>
    <mod name="eq" start="36" size="1"/>
  </ins>

  <ins name="DISCARD.f32" title="Discard fragment" dests="0" opcode="0x20" unit="CVT">
    <desc>
      Evaluates the given condition, and if it passes, discards the current
      fragment and terminates the thread. Only valid in a **fragment** shader.
    </desc>
    <cmp/>
    <src absneg="true" swizzle="true">Left value to compare</src>
    <src absneg="true" swizzle="true">Right value to compare</src>
  </ins>

  <ins name="BRANCHZI" title="Compare to zero and branch indirect" opcode="0x2F" unit="CVT">
    <desc>
      Jump to an indirectly specified (absolute or relative) address. Used to
      jump to blend shaders at the end of a fragment shader.
    </desc>
    <src combine="true">Value to compare against zero</src>
    <src>Branch target</src>
    <conservative/>
    <mod name="eq" start="36" size="1"/>
    <mod name="absolute" start="40" size="1"/>
  </ins>

  <ins name="BARRIER" title="Execution and memory barrier" opcode="0x45" unit="NONE">
    <desc>
      General-purpose barrier. Must use slot #7. Must be paired with a
      `.wait` flow on the instruction.
    </desc>
    <slot/>
  </ins>

  <group name="CSEL" title="Floating-point conditional select" dests="1" unit="CVT">
    <ins name="CSEL.f32" opcode="0x154"/>
    <ins name="CSEL.v2f16" opcode="0x155"/>
    <desc>
      Evaluates the given condition and outputs either the true source or the
      false source.
    </desc>
    <cmp/>
    <src float="true">Left value to compare</src>
    <src float="true">Right value to compare</src>
    <src float="true">Return value if true</src>
    <src float="true">Return value if false</src>
  </group>

  <group name="CSEL" title="Integer conditional select" dests="1" unit="CVT">
    <ins name="CSEL.u32" opcode="0x150"/>
    <ins name="CSEL.v2u16" opcode="0x151"/>
    <ins name="CSEL.s32" opcode="0x158"/>
    <ins name="CSEL.v2s16" opcode="0x159"/>
    <desc>
      Evaluates the given condition and outputs either the true source or the
      false source.

      Valhall lacks integer minimum/maximum instructions. `CSEL` instructions
      with tied operands form the canonical implementations of these
      instructions. Similarly, the integer $\text{sign}$ function is canonically
      implemented with a pair of `CSEL` instructions.
    </desc>
    <cmp/>
    <src>Left value to compare</src>
    <src>Right value to compare</src>
    <src>Return value if true</src>
    <src>Return value if false</src>
  </group>

  <ins name="LD_VAR_SPECIAL" title="Load special varying" opcode="0x56" unit="V">
    <sr write="true"/>
    <sr_count/>
    <vecsize/>
    <regfmt/>
    <sample/>
    <update/>
    <slot/>
    <src/>
    <imm name="index" start="12" size="4"/> <!-- 0 for pointx, 1 for pointy, 2 for fragw, 3 for fragz -->
  </ins>

  <group name="LD_VAR_BUF_IMM" title="Load immediate varying" unit="V">
    <desc>Interpolates a given varying from hardware buffer</desc>
    <ins name="LD_VAR_BUF_IMM.f32" opcode="0x5C"/>
    <ins name="LD_VAR_BUF_IMM.f16" opcode="0x5D"/>
    <slot/>
    <vecsize/>
    <source_format/>
    <sample/>
    <update/>
    <sr write="true"/>
    <sr_count/>
    <src/>
    <imm name="index" start="16" size="8"/>
  </group>

  <group name="LD_VAR_BUF" title="Load indirect varying" unit="V">
    <desc>Interpolates a given varying from hardware buffer</desc>
    <ins name="LD_VAR_BUF.f32" opcode="0x6C"/>
    <ins name="LD_VAR_BUF.f16" opcode="0x6D"/>
    <slot/>
    <vecsize/>
    <source_format/>
    <sample/>
    <update/>
    <sr write="true"/>
    <sr_count/>
    <src/>
    <src/>
  </group>

  <ins name="LD_VAR" title="Load indirect varying" unit="V" opcode="0x64">
    <desc>Interpolates a given varying from a software buffer</desc>
    <slot/>
    <vecsize/>
    <regfmt/>
    <sample/>
    <update/>
    <sr write="true"/>
    <sr_count/>
    <src/>
    <src>Varying index and table</src>
  </ins>

  <ins name="LD_VAR_IMM" title="Load immediate varying" unit="V" opcode="0x54">
    <desc>Interpolates a given varying from a software buffer</desc>
    <slot/>
    <vecsize/>
    <regfmt/>
    <sample/>
    <update/>
    <sr write="true"/>
    <sr_count/>
    <src/>
    <imm name="table" start="8" size="4"/>
    <imm name="index" start="12" size="8"/>
  </ins>

  <ins name="LD_VAR_FLAT" title="Load indirect varying" unit="V" opcode="0x55">
    <desc>Fetches a given varying from a software buffer</desc>
    <slot/>
    <vecsize/>
    <regfmt/>
    <sr write="true"/>
    <sr_count/>
    <src>Varying index and table</src>
  </ins>

  <ins name="LD_VAR_FLAT_IMM" title="Load immediate varying" unit="V" opcode="0x41">
    <desc>Fetches a given varying from a software buffer</desc>
    <slot/>
    <vecsize/>
    <regfmt/>
    <sr write="true"/>
    <sr_count/>
    <imm name="table" start="8" size="4"/>
    <imm name="index" start="12" size="8"/>
  </ins>

  <ins name="LD_ATTR_IMM" title="Load immediate attribute" opcode="0x66" opcode2="0" unit="LS">
    <desc>
      Load `vecsize` components from the attribute descriptor at entry `index`
      of resource table `table` at index (vertex ID, instance ID), converting
      to the specified register format.
    </desc>
    <sr_count/>
    <vecsize/>
    <regfmt/>
    <slot/>
    <mod name="descriptor_type" start="128" size="1" implied="true"/>
    <sr write="true"/>
    <src>Vertex ID</src>
    <src>Instance ID</src>
    <imm name="index" start="20" size="4"/>
    <imm name="table" start="16" size="4"/>
  </ins>

  <ins name="LD_ATTR" title="Load indirect attribute" opcode="0x76" opcode2="0" unit="LS">
    <desc>
      Load `vecsize` components from the attribute descriptor at the specified
      location at index (vertex ID, instance ID), converting
      to the specified register format.

      The index must not diverge within a warp.
    </desc>
    <sr_count/>
    <vecsize/>
    <regfmt/>
    <slot/>
    <mod name="descriptor_type" start="128" size="1" implied="true"/>
    <sr write="true"/>
    <src>Vertex ID</src>
    <src>Instance ID</src>
    <src>Index and table</src>
  </ins>

  <ins name="LD_TEX_IMM" title="Load immediate texture" opcode="0x66" opcode2="1" unit="LS">
    <desc>
      Load `vecsize` components from the texture descriptor at entry `index`
      of resource table `table`, converting
      to the specified register format.
    </desc>
    <sr_count/>
    <vecsize/>
    <regfmt/>
    <slot/>
    <mod name="descriptor_type" start="128" size="1" implied="true"/>
    <sr write="true"/>
    <src>X/Y coordinates (16:16)</src>
    <src>Z/W coordinates (16:16)</src>
    <imm name="index" start="20" size="4"/>
    <imm name="table" start="16" size="4"/>
  </ins>

  <ins name="LD_TEX" title="Load indirect texture" opcode="0x76" opcode2="1" unit="LS">
    <desc>
      Load `vecsize` components from the texture descriptor at the specified
      location at index, converting
      to the specified register format.
    </desc>
    <sr_count/>
    <vecsize/>
    <regfmt/>
    <slot/>
    <mod name="descriptor_type" start="128" size="1" implied="true"/>
    <sr write="true"/>
    <src>X/Y coordinates (16:16)</src>
    <src>Z/W coordinates (16:16)</src>
    <src>Index and table</src>
  </ins>

  <ins name="LEA_ATTR_IMM" title="Load effective address of image texel" opcode="0x67" opcode2="0" unit="LS">
    <desc>
      Load the effective address of an attribute specified with the
      given immediate index. Returns three staging register: the low/high
      32-bits of the address and the internal conversion descriptor.
    </desc>
    <slot/>
    <sr_count/>
    <mod name="descriptor_type" start="128" size="1" implied="true"/>
    <sr write="true"/>
    <src>Vertex index</src>
    <src>Instance index</src>
    <imm name="table" start="16" size="4"/>
    <imm name="index" start="20" size="4"/>
  </ins>

  <ins name="LEA_ATTR" title="Load effective address of image texel" opcode="0x77" opcode2="0" unit="LS">
    <desc>
      Load the effective address of an attribute specified with the
      given index. Returns three staging register: the low/high
      32-bits of the address and the internal conversion descriptor.
    </desc>
    <vecsize/>
    <slot/>
    <sr_count/>
    <mod name="descriptor_type" start="128" size="1" implied="true"/>
    <sr write="true"/>
    <src>Vertex index</src>
    <src>Instance index</src>
    <src>Attribute index and table</src>
  </ins>

  <ins name="LEA_TEX_IMM" title="Load effective address of image texel" opcode="0x67" opcode2="1" unit="LS">
    <desc>
      Load the effective address of a texel from the image specified with the
      given immediate index. Returns three staging registers: the low/high
      32-bits of the address and the internal conversion descriptor. The format
      of the internal conversion descriptor is compatible with Bifrost but
      omits the register format, as this is specified with the ST_CVT
      instruction on Valhall.

      Coordinates are specified as 16-bit integers, packed into 32-bit sources.
    </desc>
    <slot/>
    <sr_count/>
    <mod name="descriptor_type" start="128" size="1" implied="true"/>
    <sr write="true"/>
    <src>X/Y coordinates (16:16)</src>
    <src>Z/W coordinates (16:16)</src>
    <imm name="table" start="16" size="4"/>
    <imm name="index" start="20" size="4"/>
  </ins>

  <ins name="LEA_TEX" title="Load effective address of image texel" opcode="0x77" opcode2="1" unit="LS">
    <desc>
      Load the effective address of a texel from the image specified with the
      given index. Returns three staging register: the low/high
      32-bits of the address and the internal conversion descriptor. The format
      of the internal conversion descriptor is compatible with Bifrost but
      omits the register format, as this is specified with the ST_CVT
      instruction on Valhall.

      Coordinates are specified as 16-bit integers, packed into 32-bit sources.
    </desc>
    <vecsize/>
    <slot/>
    <sr_count/>
    <mod name="descriptor_type" start="128" size="1" implied="true"/>
    <sr write="true"/>
    <src size="16">X/Y coordinates (16:16)</src>
    <src>Z/W coordinates (16:16)</src>
    <src>Index and table</src>
  </ins>

  <ins name="LD_BUFFER.i8" title="Global memory load" opcode="0x6a" opcode2="0" unit="LS">
    <desc>
      Loads a buffer descriptor. If bits 25...31 of the mode descriptor are
      all-ones, load from the buffer descriptors in the table indexed by the
      bottom byte of the mode descriptor. If they are all zeroes, load the
      contents of the buffer in the first table indexed by the bottom byte of
      the mode descriptor.
    </desc>
    <sr write="true"/>
    <sr_count/>
    <mod name="load_lane_8_bit" start="36" size="3"/>
    <mod name="unsigned" start="39" size="1"/>
    <slot/>
    <src size="32">Address to load from after adding offset</src>
    <src size="32">Mode descriptor</src>
  </ins>

  <ins name="LD_BUFFER.i16" title="Global memory load" opcode="0x6a" opcode2="1" unit="LS">
    <desc>
      Loads a buffer descriptor. If bits 25...31 of the mode descriptor are
      all-ones, load from the buffer descriptors in the table indexed by the
      bottom byte of the mode descriptor. If they are all zeroes, load the
      contents of the buffer in the first table indexed by the bottom byte of
      the mode descriptor.
    </desc>
    <sr write="true"/>
    <sr_count/>
    <mod name="load_lane_16_bit" start="36" size="3"/>
    <mod name="unsigned" start="39" size="1"/>
    <slot/>
    <src size="32">Byte offset</src>
    <src size="32">Mode descriptor</src>
  </ins>

  <ins name="LD_BUFFER.i24" title="Global memory load" opcode="0x6a" opcode2="2" unit="LS">
    <desc>
      Loads a buffer descriptor. If bits 25...31 of the mode descriptor are
      all-ones, load from the buffer descriptors in the table indexed by the
      bottom byte of the mode descriptor. If they are all zeroes, load the
      contents of the buffer in the first table indexed by the bottom byte of
      the mode descriptor.
    </desc>
    <sr write="true"/>
    <sr_count/>
    <mod name="load_lane_24_bit" start="36" size="3"/>
    <mod name="unsigned" start="39" size="1"/>
    <slot/>
    <src size="32">Byte offset</src>
    <src size="32">Mode descriptor</src>
  </ins>

  <ins name="LD_BUFFER.i32" title="Global memory load" opcode="0x6a" opcode2="3" unit="LS">
    <desc>
      Loads a buffer descriptor. If bits 25...31 of the mode descriptor are
      all-ones, load from the buffer descriptors in the table indexed by the
      bottom byte of the mode descriptor. If they are all zeroes, load the
      contents of the buffer in the first table indexed by the bottom byte of
      the mode descriptor.
    </desc>
    <sr write="true"/>
    <sr_count/>
    <mod name="load_lane_32_bit" start="36" size="3"/>
    <mod name="unsigned" start="39" size="1"/>
    <slot/>
    <src size="32">Byte offset</src>
    <src size="32">Mode descriptor</src>
  </ins>

  <ins name="LD_BUFFER.i48" title="Global memory load" opcode="0x6a" opcode2="4" unit="LS">
    <desc>
      Loads a buffer descriptor. If bits 25...31 of the mode descriptor are
      all-ones, load from the buffer descriptors in the table indexed by the
      bottom byte of the mode descriptor. If they are all zeroes, load the
      contents of the buffer in the first table indexed by the bottom byte of
      the mode descriptor.
    </desc>
    <sr write="true"/>
    <sr_count/>
    <mod name="load_lane_48_bit" start="36" size="3"/>
    <mod name="unsigned" start="39" size="1"/>
    <slot/>
    <src size="32">Byte offset</src>
    <src size="32">Mode descriptor</src>
  </ins>

  <ins name="LD_BUFFER.i64" title="Global memory load" opcode="0x6a" opcode2="5" unit="LS">
    <desc>
      Loads a buffer descriptor. If bits 25...31 of the mode descriptor are
      all-ones, load from the buffer descriptors in the table indexed by the
      bottom byte of the mode descriptor. If they are all zeroes, load the
      contents of the buffer in the first table indexed by the bottom byte of
      the mode descriptor.
    </desc>
    <sr write="true"/>
    <sr_count/>
    <mod name="load_lane_64_bit" start="36" size="3"/>
    <mod name="unsigned" start="39" size="1"/>
    <slot/>
    <src size="32">Byte offset</src>
    <src size="32">Mode descriptor</src>
  </ins>

  <ins name="LD_BUFFER.i96" title="Global memory load" opcode="0x6a" opcode2="6" unit="LS">
    <desc>
      Loads a buffer descriptor. If bits 25...31 of the mode descriptor are
      all-ones, load from the buffer descriptors in the table indexed by the
      bottom byte of the mode descriptor. If they are all zeroes, load the
      contents of the buffer in the first table indexed by the bottom byte of
      the mode descriptor.
    </desc>
    <sr write="true"/>
    <sr_count/>
    <mod name="load_lane_96_bit" start="36" size="3"/>
    <mod name="unsigned" start="39" size="1"/>
    <slot/>
    <src size="32">Byte offset</src>
    <src size="32">Mode descriptor</src>
  </ins>

  <ins name="LD_BUFFER.i128" title="Global memory load" opcode="0x6a" opcode2="7" unit="LS">
    <desc>
      Loads a buffer descriptor. If bits 25...31 of the mode descriptor are
      all-ones, load from the buffer descriptors in the table indexed by the
      bottom byte of the mode descriptor. If they are all zeroes, load the
      contents of the buffer in the first table indexed by the bottom byte of
      the mode descriptor.
    </desc>
    <sr write="true"/>
    <sr_count/>
    <mod name="load_lane_128_bit" start="36" size="3"/>
    <mod name="unsigned" start="39" size="1"/>
    <slot/>
    <src size="32">Byte offset</src>
    <src size="32">Mode descriptor</src>
  </ins>

  <ins name="LEA_BUF_IMM" title="Load buffer effective address" opcode="0x5E" unit="LS">
    <desc>
      Load effective address of a buffer with an immediate offset added.
    </desc>
    <sr write="true"/>
    <sr_count/>
    <slot/>
    <imm name="table" start="8" size="4"/>
    <imm name="index" start="12" size="8"/>
    <src>Linear ID</src>
  </ins>

  <ins name="LOAD.i8" title="Global memory load" opcode="0x60" opcode2="0" unit="LS">
    <desc>Loads from main memory</desc>
    <sr write="true"/>
    <memory_access/>
    <sr_count/>
    <mod name="load_lane_8_bit" start="36" size="3"/>
    <mod name="unsigned" start="39" size="1"/>
    <slot/>
    <src size="64">Address to load from after adding offset</src>
    <imm name="offset" start="8" size="16" signed="true"/>
  </ins>

  <ins name="LOAD.i16" title="Global memory load" opcode="0x60" opcode2="1" unit="LS">
    <desc>Loads from main memory</desc>
    <sr write="true"/>
    <memory_access/>
    <sr_count/>
    <mod name="load_lane_16_bit" start="36" size="3"/>
    <mod name="unsigned" start="39" size="1"/>
    <slot/>
    <src size="64">Address to load from after adding offset</src>
    <imm name="offset" start="8" size="16" signed="true"/>
  </ins>

  <ins name="LOAD.i24" title="Global memory load" opcode="0x60" opcode2="2" unit="LS">
    <desc>Loads from main memory</desc>
    <sr write="true"/>
    <memory_access/>
    <sr_count/>
    <mod name="load_lane_24_bit" start="36" size="3"/>
    <mod name="unsigned" start="39" size="1"/>
    <slot/>
    <src size="64">Address to load from after adding offset</src>
    <imm name="offset" start="8" size="16" signed="true"/>
  </ins>

  <ins name="LOAD.i32" title="Global memory load" opcode="0x60" opcode2="3" unit="LS">
    <desc>Loads from main memory</desc>
    <sr write="true"/>
    <memory_access/>
    <sr_count/>
    <mod name="load_lane_32_bit" start="36" size="3"/>
    <mod name="unsigned" start="39" size="1"/>
    <slot/>
    <src size="64">Address to load from after adding offset</src>
    <imm name="offset" start="8" size="16" signed="true"/>
  </ins>

  <ins name="LOAD.i48" title="Global memory load" opcode="0x60" opcode2="4" unit="LS">
    <desc>Loads from main memory</desc>
    <sr write="true"/>
    <memory_access/>
    <sr_count/>
    <mod name="load_lane_48_bit" start="36" size="3"/>
    <mod name="unsigned" start="39" size="1"/>
    <slot/>
    <src size="64">Address to load from after adding offset</src>
    <imm name="offset" start="8" size="16" signed="true"/>
  </ins>

  <ins name="LOAD.i64" title="Global memory load" opcode="0x60" opcode2="5" unit="LS">
    <desc>Loads from main memory</desc>
    <sr write="true"/>
    <memory_access/>
    <sr_count/>
    <mod name="load_lane_64_bit" start="36" size="3"/>
    <mod name="unsigned" start="39" size="1"/>
    <slot/>
    <src size="64">Address to load from after adding offset</src>
    <imm name="offset" start="8" size="16" signed="true"/>
  </ins>

  <ins name="LOAD.i96" title="Global memory load" opcode="0x60" opcode2="6" unit="LS">
    <desc>Loads from main memory</desc>
    <sr write="true"/>
    <memory_access/>
    <sr_count/>
    <mod name="load_lane_96_bit" start="36" size="3"/>
    <mod name="unsigned" start="39" size="1"/>
    <slot/>
    <src size="64">Address to load from after adding offset</src>
    <imm name="offset" start="8" size="16" signed="true"/>
  </ins>

  <ins name="LOAD.i128" title="Global memory load" opcode="0x60" opcode2="7" unit="LS">
    <desc>Loads from main memory</desc>
    <sr write="true"/>
    <memory_access/>
    <sr_count/>
    <mod name="load_lane_128_bit" start="36" size="3"/>
    <mod name="unsigned" start="39" size="1"/>
    <slot/>
    <src size="64">Address to load from after adding offset</src>
    <imm name="offset" start="8" size="16" signed="true"/>
  </ins>

  <group name="STORE" title="Global memory store" opcode="0x61" unit="LS">
    <desc>Stores to main memory</desc>
    <sr read="true"/>
    <ins name="STORE.i8" opcode2="0x0"/>
    <ins name="STORE.i16" opcode2="0x1"/>
    <ins name="STORE.i24" opcode2="0x2"/>
    <ins name="STORE.i32" opcode2="0x3"/>
    <ins name="STORE.i48" opcode2="0x4"/>
    <ins name="STORE.i64" opcode2="0x5"/>
    <ins name="STORE.i96" opcode2="0x6"/>
    <ins name="STORE.i128" opcode2="0x7"/>
    <sr_count/>
    <memory_access/>
    <slot/>
    <src size="64">Address to store to after adding offset</src>
    <imm name="offset" start="8" size="16" signed="true"/>
  </group>

  <ins name="ST_CVT" title="Store with conversion" opcode="0x71" unit="LS">
    <desc>
      Store to memory with data conversion. The address to store to is given in
      the first source, which must be a 64-bit register (a pair of 32-bit
      registers). The other source is the conversion descriptor used for the store.

      Used with LEA_TEX_IMM to implement image stores.
    </desc>
    <slot/>
    <mod name="memory_access" start="37" size="3"/>
    <vecsize/>
    <regfmt/>
    <sr read="true"/>
    <sr_count/>
    <src size="64">64-bit address to store to</src>
    <imm name="offset" start="8" size="8"/>
    <src>Internal conversion descriptor</src>
  </ins>

  <ins name="LD_TILE" title="Load from tilebuffer" opcode="0x78" unit="NONE">
    <desc>
      Loads a given render target, specified in the pixel indices descriptor, at
      a given location and sample, and convert to the format specified in the
      internal conversion descriptor. Used to implement EXT_framebuffer_fetch
      and internally in blend shaders.
    </desc>
    <sr write="true"/>
    <sr_count/>
    <vecsize/>
    <regfmt/>
    <slot/>
    <src>Pixel indices descriptor</src>
    <src>Coverage mask</src>
    <src>Conversion descriptor</src>
  </ins>

  <ins name="ST_TILE" title="Store to tilebuffer" opcode="0x79" unit="NONE">
    <desc>
      Store to given render target, specified in the pixel indices descriptor, at
      a given location and sample, and convert to the format specified in the
      internal conversion descriptor. Used internally in blend shaders.
    </desc>
    <sr read="true"/>
    <sr_count/>
    <vecsize/>
    <regfmt/>
    <slot/>
    <src>Pixel indices descriptor</src>
    <src>Coverage mask</src>
    <src>Conversion descriptor</src>
  </ins>

  <ins name="BLEND" title="Blend render target" opcode="0x7F" unit="NONE">
    <desc>
      Blends a given render target. This loads the API-specified blend state for
      the render target from the first source. Blend descriptors are available
      as special immediates. It then reads the colour to be blended from the
      first staging register, with the specified vector size and register format
      as desired. The resulting coverage mask is stored to the second set of
      staging registers.

      In the fixed-function path, `BLEND` sends the colour to the blender to be
      written to the tilebuffer. Then, if the instruction's flow control
      specifies termination, the fragment program is ended. If it does not
      specify termination, `BLEND` acts as a relative branch, branching with the
      offset specified as `target`. This allows the subsequent instructions to
      be skipped when fixed-function blending is used. Note this implicit branch
      can never introduce divergence, so `.reconverge` is not required.

      In the blend shader path, `BLEND` ignores the specified flow control and
      does not branch to the specified offset. Instead, execution continues
      normally with the next instruction. The compiler should insert code for
      calling a blend shader after the `BLEND` instruction unless it is known
      that a blend shader will never be required.

      The indirection is required to support both fixed-function and blend
      shaders efficiently and without shader variants.
    </desc>
    <sr read="true"/>
    <src size="64">Blend descriptor</src>
    <src>Sample coverage</src>
    <imm name="target" start="8" size="8"/>
    <slot/>
    <sr_count/>
    <vecsize/>
    <regfmt/>
  </ins>

  <ins name="ATEST" title="Alpha test" opcode="0x7D" unit="NONE">
    <desc>
      Does alpha-to-coverage testing, updating the sample coverage mask. ATEST
      does not do an implicit discard. It should be executed before the first
      ZS_EMIT or BLEND instruction.
    </desc>
    <sr write="true">Updated coverage mask</sr>
    <src>Input coverage mask</src>
    <src swizzle="true">Alpha value (render target 0)</src>
    <src/>
    <sr_count/>
  </ins>

  <ins name="ZS_EMIT" title="Depth/stencil write" opcode="0x7E" unit="NONE">
    <desc>
      Programatically writes out depth, stencil, or both, depending on which
      modifiers are set. Used to implement gl_FragDepth and gl_FragStencil.
    </desc>
    <mod name="z" start="25" size="1"/>
    <mod name="stencil" start="24" size="1"/>
    <sr write="true">Updated coverage mask</sr>
    <src>Depth value</src>
    <src>Stencil value</src>
    <src>Input coverage mask</src>
    <sr_count/>
    <slot/>
  </ins>

  <group name="CONVERT" title="Data conversions" dests="1" opcode="0x90" unit="CVT">
    <desc>
      Performs the given data conversion. Note that floating-point rounding is
      handled via the same hardware and therefore shares an encoding. Round mode
      is specified where it makes sense.
    </desc>

    <ins name="V2S16_TO_V2F16" opcode2="0x7"/>

    <ins name="S32_TO_F32" opcode2="0x9"/>

    <ins name="V2U16_TO_V2F16" opcode2="0x17"/>

    <ins name="U32_TO_F32" opcode2="0x19"/>

    <roundmode/>
    <src widen="true">Value to convert</src>
  </group>

  <group name="CONVERT" title="16->32 integer data conversions" dests="1" opcode="0x90" unit="CVT">
    <desc>
      Performs the given data conversion.
    </desc>

    <ins name="S16_TO_S32" opcode2="0x4"/>
    <ins name="S16_TO_F32" opcode2="0x5"/>
    <ins name="U16_TO_U32" opcode2="0x14"/>
    <ins name="U16_TO_F32" opcode2="0x15"/>

    <src swizzle="true" size="16">Value to convert</src>
  </group>

  <group name="CONVERT" title="Float-to-int data conversions" dests="1" opcode="0x90" unit="CVT">
    <desc>Performs the given data conversion.</desc>
    <ins name="F32_TO_S32" opcode2="0xC"/>
    <ins name="F32_TO_U32" opcode2="0x1C"/>
    <roundmode/>
    <src absneg="true">Value to convert</src>
  </group>

  <group name="CONVERT" title="Float-to-int data conversions" dests="1" opcode="0x90" unit="CVT">
    <desc>Performs the given data conversion.</desc>
    <ins name="V2F16_TO_V2S16" opcode2="0xE"/>
    <ins name="V2F16_TO_V2U16" opcode2="0x1E"/>
    <ins name="F16_TO_S32" opcode2="0xA"/>
    <ins name="F16_TO_U32" opcode2="0x1A"/>
    <roundmode/>
    <src swizzle="true" absneg="true" size="16">Value to convert</src>
  </group>

  <ins name="F16_TO_F32" title="16-bit float to 32-bit float conversion" dests="1" opcode="0x90" opcode2="0xB" unit="CVT">
    <desc>Converts up with the specified round mode.</desc>
    <roundmode/>
    <src lane="28" size="16" absneg="true">Value to convert</src>
  </ins>

  <group name="CONVERT" title="8-bit to 32-bit data conversions" dests="1" opcode="0x90" unit="CVT">
    <desc>
      Performs the given data conversion.
    </desc>

    <ins name="S8_TO_S32" opcode2="0x0"/>
    <ins name="S8_TO_F32" opcode2="0x1"/>

    <ins name="U8_TO_U32" opcode2="0x10"/>
    <ins name="U8_TO_F32" opcode2="0x11"/>

    <src lane="28" size="8">Value to convert</src>
  </group>

  <group name="CONVERT" title="8-bit to 16-bit data conversions" dests="1" opcode="0x90" unit="CVT">
    <desc>
      Performs the given data conversion.
    </desc>

    <ins name="V2S8_TO_V2S16" opcode2="0x2"/>
    <ins name="V2S8_TO_V2F16" opcode2="0x3"/>

    <ins name="V2U8_TO_V2U16" opcode2="0x12"/>
    <ins name="V2U8_TO_V2F16" opcode2="0x13"/>

    <src halfswizzle="true" size="8">Value to convert</src>
  </group>

  <group name="FROUND" title="Floating-point rounding" dests="1" opcode="0x90" unit="CVT">
    <desc>
      Performs the given rounding, using the convert unit.
    </desc>

    <ins name="FROUND.f32" opcode2="0xD"/>
    <ins name="FROUND.v2f16" opcode2="0xF"/>

    <roundmode/>
    <src swizzle="true" absneg="true">Value to convert</src>
  </group>

  <ins name="MOV.i32" title="Register move" dests="1" opcode="0x91" opcode2="0x0" unit="CVT">
    <desc>Canonical register-to-register move.</desc>
    <src/>
  </ins>

  <ins name="CLZ.u32" title="Count leading zeroes" dests="1" opcode="0x91" opcode2="0x4" unit="CVT">
    <desc>
      Used as a primitive for various bitwise operations.
    </desc>
    <src/>
  </ins>

  <ins name="CLZ.v2u16" title="Count leading zeroes" dests="1" opcode="0x91" opcode2="0x5" unit="CVT">
    <desc>
      Used as a primitive for various bitwise operations.
    </desc>
    <src/>
  </ins>

  <ins name="CLZ.v4u8" title="Count leading zeroes" dests="1" opcode="0x91" opcode2="0x6" unit="CVT">
    <desc>
      Used as a primitive for various bitwise operations.
    </desc>
    <src/>
  </ins>

  <ins name="IABS.s32" title="Absolute value" dests="1" opcode="0x91" opcode2="0x8" unit="CVT">
    <desc>
      64-bit abs may be constructed in 4 instructions (5 clocks) by checking the
      sign with `ICMP.s32.lt.m1 hi, 0` and negating based on the result with
      `IADD.s64` and `LSHIFT_XOR.i32` on each half.
    </desc>
    <src widen="true"/>
  </ins>

  <ins name="IABS.v2s16" title="Absolute value" dests="1" opcode="0x91" opcode2="0x9" unit="CVT">
    <src widen="true"/>
  </ins>

  <ins name="IABS.v4s8" title="Absolute value" dests="1" opcode="0x91" opcode2="0xa" unit="CVT">
    <src/>
  </ins>

  <ins name="POPCOUNT.i32" title="Population count" dests="1" opcode="0x91" opcode2="0xC" unit="SFU">
    <desc>
      Only available as 32-bit. Smaller bitsizes require explicit conversions.
      64-bit popcount may be constructed in 3 clocks by separate 32-bit
      popcounts of each half and a 32-bit add, which is guaranteed not to
      overflow.
    </desc>
    <src/>
  </ins>

  <ins name="BITREV.i32" title="Bitwise reverse" dests="1" opcode="0x91" opcode2="0xD" unit="SFU">
    <desc>
      Only available as 32-bit. Other bitsizes may be derived with swizzles.
    </desc>
    <src/>
  </ins>

  <ins name="NOT_OLD.i32" title="Bitwise complement" dests="1" opcode="0x91" opcode2="0xE" unit="SFU">
    <desc>
      For fully featured bitwise operation, see the shift opcodes.
    </desc>
    <src/>
  </ins>

  <ins name="NOT_OLD.i64" title="Bitwise complement" dests="1" opcode="0x191" opcode2="0xE" unit="SFU">
    <desc>
      For fully featured bitwise operation, see the shift opcodes.
    </desc>
    <src/>
  </ins>

  <ins name="WMASK" title="Warp mask" dests="1" opcode="0x95" unit="CVT">
    <desc>
      Returns the mask of lanes ever active within the warp (subgroup), such
      that the source is nonzero. The number of work-items in a subgroup is
      given as the popcount of this value with a nonzero input.

      An `all()` subgroup operation may be constructed as `WMASK` of the input
      compared for equality with `WMASK` of an nonzero value.

      An `any()` subgroup operation may be constructed as `WMASK` of the input
      compared against zero.
    </desc>
    <src/>
    <subgroup/>
  </ins>

  <group name="FREXP" title="Fraction/exponent extract" dests="1" opcode="0x99" unit="CVT">
    <ins name="FREXPM.f32" opcode2="0"/>
    <ins name="FREXPM.v2f16" opcode2="1"/>
    <ins name="FREXPE.f32" opcode2="2"/>
    <ins name="FREXPE.v2f16" opcode2="3"/>
    <desc>
      Breaks up the floating-point input into its fractional (mantissa) and
      exponent parts. By default, this is compatible with the `frexp()` function
      in APIs. With the log/sqrt modifiers, the floating point format is
      adjusted to be compatible with Valhall's argument reduction for logarithm
      and square root computation respectively.
    </desc>
    <mod name="sqrt" start="24" size="1"/>
    <mod name="log" start="25" size="1"/>
    <src float="true" swizzle="true"/>
  </group>

  <group name="SFU" title="Special function unit" dests="1" opcode="0x9C" unit="SFU">
    <ins name="FRCP.f32" opcode2="0"/>
    <ins name="FRCP.f16" opcode2="1"/>
    <ins name="FRSQ.f32" opcode2="2"/>
    <ins name="FRSQ.f16" opcode2="3"/>
    <ins name="FLOGD.f32" opcode2="8"/>
    <ins name="FPCLASS.f32" opcode2="10"/>
    <ins name="FPCLASS.f16" opcode2="11"/>
    <ins name="FLOG_TABLE.f32" opcode2="12"/>
    <ins name="FRCP_APPROX.f32" opcode2="14"/>
    <ins name="FRSQ_APPROX.f32" opcode2="15"/>
    <desc>
      Performs a given special function. The floating-point reciprocal (`FRCP`)
      and reciprocal square root (`FRSQ`) instructions may be freely used as-is.
      The logarithm instruction (`FLOGD.f32`) requires an argument
      reduction. See the transcendentals section for more information. Like the
      Bifrost op, `FRSQ_APPROX.f32` does an implicit `FREXPM.f32.sqrt` on the
      source.
    </desc>
    <src float="true" swizzle="true" absneg="true"/>
  </group>

  <group name="SFU" title="Special function unit" dests="1" opcode="0x9C" unit="SFU">
    <ins name="FSIN_TABLE.u6" opcode2="4"/>
    <ins name="FCOS_TABLE.u6" opcode2="5"/>
    <ins name="FSINCOS_OFFSET.u6" opcode2="6"/>
    <ins name="FEXP_TABLE.u4" opcode2="13"/>
    <desc>
      Performs a given special function. The trigonometric tables
      (`FSIN_TABLE.u6` and `FCOS_TABLE.u6`) are crude, requiring both an
      argument reduction and postprocessing.
    </desc>
    <src/>
  </group>

  <group name="FADD" title="Floating-point add" dests="1" opcode2="0" unit="FMA">
    <ins name="FADD.f32" opcode="0xA4"/>
    <ins name="FADD.v2f16" opcode="0xA5"/>
    <desc>$A + B$</desc>
    <clamp/>
    <src absneg="true" swizzle="true">A</src>
    <src absneg="true" swizzle="true">B</src>
  </group>

  <group name="FMIN" title="Floating-point minimum" dests="1" opcode2="2" unit="CVT">
    <ins name="FMIN.f32" opcode="0xA4"/>
    <ins name="FMIN.v2f16" opcode="0xA5"/>
    <desc>$\min \{ A, B \}$</desc>
    <clamp/>
    <src absneg="true" swizzle="true">A</src>
    <src absneg="true" swizzle="true">B</src>
  </group>

  <group name="FMAX" title="Floating-point maximum" dests="1" opcode2="3" unit="CVT">
    <ins name="FMAX.f32" opcode="0xA4"/>
    <ins name="FMAX.v2f16" opcode="0xA5"/>
    <desc>$\max \{ A, B \}$</desc>
    <clamp/>
    <src absneg="true" swizzle="true">A</src>
    <src absneg="true" swizzle="true">B</src>
  </group>

  <group name="V2F32_TO_V2F16" title="Vectorized floating-point conversion" dests="1" opcode2="4" unit="CVT">
    <ins name="V2F32_TO_V2F16" opcode="0xA5"/>
    <desc>
      Given a pair of 32-bit floats, output a pair of 16-bit floats packed into
      a 32-bit destination.
    </desc>
    <clamp/>
    <roundmode/>
    <src absneg="true">A</src>
    <src absneg="true">B</src>
  </group>

  <group name="LDEXP" title="Floating-point rescaling" dests="1" opcode2="6" unit="FMA">
    <ins name="LDEXP.f32" opcode="0xA4"/>
    <ins name="LDEXP.v2f16" opcode="0xA5"/>
    <desc>
      Computes $A \cdot 2^B$ by adding B to the exponent of A. Used to calculate
      various special functions, particularly base-2 exponents. Special case
      handling differs from an actual floating-point multiply, so this should
      not be used outside fixed instruction sequences.
    </desc>
    <src absneg="true" swizzle="true">A</src>
    <src/>
    <roundmode/> <!-- Also has rtna -->
    <!-- Also has infinity handling for arctan -->
  </group>

  <ins name="FEXP.f32" title="Floating-point exponent" dests="1" opcode="0xA4" opcode2="8" unit="SFU">
    <desc>
      Calculates the base-2 exponent of an argument specified as a 8:24
      fixed-point. The original argument is passed as well for correct handling
      of special cases.
    </desc>
    <clamp/>
    <src>Input as 8:24 fixed-point</src>
    <src absneg="true">Input as 32-bit float</src>
  </ins>

  <ins name="FADD_LSCALE.f32" title="Floating-point add with logarithm scale" dests="1" opcode="0xA4" opcode2="9" unit="FMA">
    <desc>
      Performs a floating-point addition specialized for logarithm computation.
    </desc>
    <clamp/>
    <src absneg="true">A</src>
    <src absneg="true">B</src>
  </ins>

  <ins name="FATAN_ASSIST.f32" title="ATAN calculation helper" dests="1" opcode="0xA4" opcode2="14" unit="SFU">
    <desc>
      Used for `atan2()` implementation. Destination is two 16-bit
      values (int and float) for the first form, and a single 32-bit float when
      `.second` is set (indicating the FATAN_TABLE.f32 instruction).
    </desc>
    <mod name="second" start="24" size="1"/>
    <src>A</src>
    <src>B</src>
  </ins>

  <group name="IADD" title="Integer addition" dests="1" opcode2="0" unit="CVT">
    <desc>
      $A + B$ with optional saturation.

      As Valhall lacks swizzle instructions, `IADD.v2i16` with zero is the
      canonical lowering for swizzles.
    </desc>
    <ins name="IADD.u32" opcode="0xA0"/>
    <ins name="IADD.v2u16" opcode="0xA1"/>
    <ins name="IADD.v4u8" opcode="0xA2"/>
    <ins name="IADD.s32" opcode="0xA8"/>
    <ins name="IADD.v2s16" opcode="0xA9"/>
    <ins name="IADD.v4s8" opcode="0x1A2"/>
    <ins name="IADD.u64" opcode="0x1A3"/>
    <ins name="IADD.s64" opcode="0x1AB"/>
    <!-- <ins name="IADD.s32" opcode="0x1A0"/> -->
    <src widen="true">A</src>
    <src widen="true">B</src>
    <saturate/>
  </group>

  <ins name="MKVEC.v2i16" title="Make 16-bit vector" dests="1" opcode="0xA1" opcode2="0x5" unit="CVT">
    <desc>Calculates $A | (B \ll 16)$. Used to implement `(ushort2)(A, B)`</desc>
    <src swizzle="true">A</src>
    <src swizzle="true">B</src>
  </ins>

  <group name="ISUB" title="Integer subtract" dests="1" opcode2="1" unit="CVT">
    <ins name="ISUB.u32" opcode="0xA0"/>
    <ins name="ISUB.v2u16" opcode="0xA1"/>
    <ins name="ISUB.v4u8" opcode="0xA2"/>
    <ins name="ISUB.s32" opcode="0xA8"/>
    <ins name="ISUB.v2s16" opcode="0xA9"/>
    <ins name="ISUB.v4s8" opcode="0x1A2"/>
    <ins name="ISUB.u64" opcode="0x1A3"/>
    <ins name="ISUB.s64" opcode="0x1AB"/>
    <desc>$A - B$ with optional saturation</desc>
    <src widen="true">A</src>
    <src widen="true">B</src>
    <saturate/>
  </group>

  <group name="SEG_ADD" title="Segment addition" dests="1" opcode2="6" unit="CVT">
    <desc>
      Similar to SHADDX, but especially used for loading offsets into
      WLS. Usually this is only required for atomic operations, which cannot
      directly use wls_pointer as an address.

      .neg indicates SEG_SUB instead.
    </desc>
    <ins name="SEG_ADD.u64" opcode="0x1A3"/>
    <mod name="neg" start="38" size="1"/>
    <mod name="preserve_null" start="39" size="1"/>
    <src>A</src>
    <src widen="true">B</src>
  </group>

  <group name="SHADDX" title="Shift, extend, and 64-bit add" dests="1" opcode2="7" unit="CVT">
    <desc>
      Sign or zero extend B to 64-bits, left-shift by `shift`, and add the
      64-bit value A. These instructions accelerate address arithmetic, but may
      be used in full generality for 64-bit integer arithmetic.
    </desc>
    <ins name="SHADDX.u64" opcode="0x1A3"/>
    <ins name="SHADDX.s64" opcode="0x1AB"/>
    <imm name="shift" start="20" size="3"/>
    <src>A</src>
    <src widen="true">B</src>
  </group>

  <group name="IMUL" title="Integer multiply" dests="1" opcode2="0x0A" unit="SFU">
    <ins name="IMUL.i32" opcode="0xA0"/>
    <ins name="IMUL.v2i16" opcode="0xA1"/>
    <ins name="IMUL.v4i8" opcode="0xA2"/>
    <ins name="IMUL.s32" opcode="0xA8"/>
    <ins name="IMUL.v2s16" opcode="0xA9"/>
    <ins name="IMUL.v4s8" opcode="0x1A2"/>
    <ins name="IMULD.u64" opcode="0x1A3"/>
    <!-- <ins name="IMUL.s32" opcode="0x1A0"/> -->
    <desc>
      $A \cdot B$ with optional saturation. Note the multipliers can only handle up to
      32-bit by 32-bit multiplies. The 64-bit "multiply" acts like IMUL.u32 but
      additionally writes the high half of the product to the high half of the
      64-bit destination. Along with IADD.u32 and IADD.u64, this allows the
      construction of a 64-bit multiply in 5 instructions (6 clocks).
    </desc>
    <src widen="true">A</src>
    <src widen="true">B</src>
    <saturate/>
  </group>

  <group name="HADD" title="Integer half-add" dests="1" opcode2="0x0B" unit="CVT">
    <ins name="HADD.u32" opcode="0xA0"/>
    <ins name="HADD.v2u16" opcode="0xA1"/>
    <ins name="HADD.v4u8" opcode="0xA2"/>
    <ins name="HADD.s32" opcode="0xA8"/>
    <ins name="HADD.v2s16" opcode="0xA9"/>
    <ins name="HADD.v4s8" opcode="0x1A2"/>
    <mod name="rhadd" start="30" size="1"/>
    <src widen="true">A</src>
    <src widen="true">B</src>
    <desc>
      $(A + B) \gg 1$ without intermediate overflow, corresponding to `hadd()` in
      OpenCL. With the `.rhadd` modifier set, it instead calculates
      $(A + B + 1) \gg 1$ corresponding to `rhadd()` in OpenCL.
    </desc>
  </group>

  <group name="CLPER" title="Cross-lane permute" dests="1" opcode2="0xF" unit="SFU">
    <ins name="CLPER.i32" opcode="0xA0"/>
    <ins name="CLPER.v2u16" opcode="0xA1"/>
    <ins name="CLPER.v4u8" opcode="0xA2"/>
    <ins name="CLPER.s32" opcode="0xA8"/>
    <ins name="CLPER.v2s16" opcode="0xA9"/>
    <ins name="CLPER.v4s8" opcode="0x1A2"/>
    <ins name="CLPER.u64" opcode="0x1A3"/>
    <ins name="CLPER.s64" opcode="0x1AB"/>
    <!-- <ins name="CLPER.s32" opcode="0x1A0"/> -->
    <desc>
      Selects the value of A in the subgroup lane given by B. This implements
      subgroup broadcasts. It may be used as a primitive for screen space
      derivatives in fragment shaders.
    </desc>
    <src>A</src>
    <src widen="true">B</src>
    <subgroup/>
    <lane_op/>
    <inactive_result/>
  </group>

  <group name="FMA" title="Fused floating-point multiply add" dests="1" unit="FMA">
    <ins name="FMA.f32" opcode="0xB2"/>
    <ins name="FMA.v2f16" opcode="0xB3"/>
    <desc>$A \cdot B + C$</desc>
    <clamp/>
    <src absneg="true" swizzle="true">A</src>
    <src absneg="true" swizzle="true">B</src>
    <src absneg="true" swizzle="true">C</src>
  </group>

  <group name="LSHIFT_AND" title="Left shift and bitwise AND" dests="1" opcode2="0x100" unit="SFU">
    <ins name="LSHIFT_AND.i32" opcode="0xB4"/>
    <ins name="LSHIFT_AND.v2i16" opcode="0xB5"/>
    <ins name="LSHIFT_AND.v4i8" opcode="0xB6"/>
    <ins name="LSHIFT_AND.i64" opcode="0x1B7"/>
    <mod name="left" start="128" size="1" implied="true"/>
    <desc>
      Left shifts its first source by a specified amount and bitwise ANDs it with the
      second source, optionally inverting the second source or the result.
    </desc>
    <not_result/>
    <src widen="true">A</src>
    <src lanes="true" size="8">shift</src>
    <src not="true">B</src>
  </group>

  <group name="RSHIFT_AND" title="Right shift and bitwise AND" dests="1" opcode2="0x000" unit="SFU">
    <ins name="RSHIFT_AND.i32" opcode="0xB4"/>
    <ins name="RSHIFT_AND.v2i16" opcode="0xB5"/>
    <ins name="RSHIFT_AND.v4i8" opcode="0xB6"/>
    <ins name="RSHIFT_AND.i64" opcode="0x1B7"/>
    <mod name="left" start="128" size="1" implied="true"/>
    <desc>
      Right shifts its first source by a specified amount and bitwise ANDs it with the
      second source, optionally inverting the second source or the result. If
      `signed` is set, the hardware performs an arithmetic right shift; otherwise,
      it performs an unsigned right shift.
    </desc>
    <mod name="signed" start="34" size="1"/>
    <not_result/>
    <src widen="true">A</src>
    <src lanes="true" size="8">shift</src>
    <src not="true">B</src>
  </group>

  <group name="LSHIFT_OR" title="Left shift and bitwise OR" dests="1" opcode2="0x101" unit="SFU">
    <ins name="LSHIFT_OR.i32" opcode="0xB4"/>
    <ins name="LSHIFT_OR.v2i16" opcode="0xB5"/>
    <ins name="LSHIFT_OR.v4i8" opcode="0xB6"/>
    <ins name="LSHIFT_OR.i64" opcode="0x1B7"/>
    <mod name="left" start="128" size="1" implied="true"/>
    <desc>
      Left shifts its first source by a specified amount and bitwise ORs it with the
      second source, optionally inverting the second source or the result.
    </desc>
    <not_result/>
    <src widen="true">A</src>
    <src lanes="true" size="8">shift</src>
    <src not="true">B</src>
  </group>

  <group name="RSHIFT_OR" title="Right shift and bitwise OR" dests="1" opcode2="0x001" unit="SFU">
    <ins name="RSHIFT_OR.i32" opcode="0xB4"/>
    <ins name="RSHIFT_OR.v2i16" opcode="0xB5"/>
    <ins name="RSHIFT_OR.v4i8" opcode="0xB6"/>
    <ins name="RSHIFT_OR.i64" opcode="0x1B7"/>
    <mod name="left" start="128" size="1" implied="true"/>
    <desc>
      Right shifts its first source by a specified amount and bitwise ORs it with the
      second source, optionally inverting the second source or the result. If
      `signed` is set, the hardware performs an arithmetic right shift; otherwise,
      it performs an unsigned right shift.
   </desc>
    <mod name="signed" start="34" size="1"/>
    <not_result/>
    <src widen="true">A</src>
    <src lanes="true" size="8">shift</src>
    <src not="true">B</src>
  </group>

  <group name="LSHIFT_XOR" title="Left shift and bitwise XOR" dests="1" opcode2="0x102" unit="SFU">
    <ins name="LSHIFT_XOR.i32" opcode="0xB4"/>
    <ins name="LSHIFT_XOR.v2i16" opcode="0xB5"/>
    <ins name="LSHIFT_XOR.v4i8" opcode="0xB6"/>
    <ins name="LSHIFT_XOR.i64" opcode="0x1B7"/>
    <mod name="left" start="128" size="1" implied="true"/>
    <desc>
      Left shifts its first source by a specified amount and bitwise XORs it with the
      second source, optionally inverting the second source or the result.
    </desc>
    <not_result/>
    <src widen="true">A</src>
    <src lanes="true" size="8">shift</src>
    <src not="true">B</src>
  </group>

  <group name="RSHIFT_XOR" title="Right shift and bitwise XOR" dests="1" opcode2="0x002" unit="SFU">
    <ins name="RSHIFT_XOR.i32" opcode="0xB4"/>
    <ins name="RSHIFT_XOR.v2i16" opcode="0xB5"/>
    <ins name="RSHIFT_XOR.v4i8" opcode="0xB6"/>
    <ins name="RSHIFT_XOR.i64" opcode="0x1B7"/>
    <mod name="left" start="128" size="1" implied="true"/>
    <desc>
      Right shifts its first source by a specified amount and bitwise XORs it with the
      second source, optionally inverting the second source or the result. If
      `signed` is set, the hardware performs an arithmetic right shift; otherwise,
      it performs an unsigned right shift.
    </desc>
    <mod name="signed" start="34" size="1"/>
    <not_result/>
    <src widen="true">A</src>
    <src lanes="true" size="8">shift</src>
    <src not="true">B</src>
  </group>

  <ins name="MUX.i32" title="Mux" dests="1" opcode="0xB8" unit="SFU">
    <desc>
      Mux between A and B based on the provided mask. The condition specified
      as the `mux` modifier is evaluated on the mask. If true, `A` is chosen,
      else `B` is chosen. The `bit` modifier acts bitwise, equivalent to
      `bitselect()` in OpenCL, so `MUX.i32.bit A, B, mask` calculates
      `(A &amp; mask) | (A &amp; ~mask)`.
    </desc>
    <mod name="mux" start="32" size="2"/>
    <src>A</src>
    <src>B</src>
    <src>Mask</src>
  </ins>

  <ins name="MUX.v2i16" title="Mux" dests="1" opcode="0xB9" unit="SFU">
    <desc>
      Mux between A and B based on the provided mask. The condition specified
      as the `mux` modifier is evaluated on the mask. If true, `A` is chosen,
      else `B` is chosen. The `bit` modifier acts bitwise, equivalent to
      `bitselect()` in OpenCL, so `MUX.i32.bit A, B, mask` calculates
      `(A &amp; mask) | (A &amp; ~mask)`.
    </desc>
    <mod name="mux" start="32" size="2"/>
    <src swizzle="true">A</src>
    <src swizzle="true">B</src>
    <src swizzle="true">Mask</src>
  </ins>

  <ins name="MUX.v4i8" title="Mux" dests="1" opcode="0xBA" unit="SFU">
    <desc>
      Mux between A and B based on the provided mask. The condition specified
      as the `mux` modifier is evaluated on the mask. If true, `A` is chosen,
      else `B` is chosen. The `bit` modifier acts bitwise, equivalent to
      `bitselect()` in OpenCL, so `MUX.i32.bit A, B, mask` calculates
      `(A &amp; mask) | (A &amp; ~mask)`.
    </desc>
    <mod name="mux" start="32" size="2"/>
    <src>A</src>
    <src>B</src>
    <src>Mask</src>
  </ins>

  <ins name="CUBE_SSEL" title="Cube S-coordinate select" dests="1" opcode="0xBC" opcode2="0" unit="SFU">
    <desc>During a cube map transform, select the S coordinate given a selected face.</desc>
    <src absneg="true">Z coordinate as 32-bit floating point</src>
    <src absneg="true">X coordinate as 32-bit floating point</src>
    <src>Cube face index</src>
  </ins>

  <ins name="CUBE_TSEL" title="Cube T-coordinate select" dests="1" opcode="0xBC" opcode2="1" unit="SFU">
    <desc>During a cube map transform, select the T coordinate given a selected face.</desc>
    <src absneg="true">Y coordinate as 32-bit floating point</src>
    <src absneg="true">Z coordinate as 32-bit floating point</src>
    <src>Cube face index</src>
  </ins>

  <ins name="MKVEC.v2i8" title="Make 8-bit vector" dests="1" opcode="0xBD" unit="CVT">
    <desc>
      Calculates $A | (B \ll 8) | (CD \ll 16)$ for 8-bit A and B and 16-bit CD.

      To implement `(uchar4) (A, B, C, D)` in full generality, use the sequence
      `MKVEC.v2i8 CD, C, D, #0; MKVEC.v2i8 out, A, B, CD`

      `MKVEC.v2i8` also allows zero extending arbitrary 8-bit lanes. For
      example, to extend `r0.b3` to `r1`, use `MKVEC.v2i8 r1, r0.b3, 0x0.b0, 0x0`.
    </desc>
    <src lane="true">A</src>
    <src lane="true">B</src>
    <src>CD</src>
  </ins>

  <ins name="CUBEFACE1" title="Cube map transform step 1" dests="1" opcode="0xC0" unit="SFU">
    <desc>Select the maximum absolute value of its arguments.</desc>
    <src absneg="true">X coordinate as 32-bit floating point</src>
    <src absneg="true">Y coordinate as 32-bit floating point</src>
    <src absneg="true">Z coordinate as 32-bit floating point</src>
  </ins>

  <ins name="CUBEFACE2" title="Cube map transform step 2" dests="1" opcode="0xC1" unit="SFU">
    <desc>Select the cube face index corresponding to the arguments.</desc>
    <src absneg="true">X coordinate as 32-bit floating point</src>
    <src absneg="true">Y coordinate as 32-bit floating point</src>
    <src absneg="true">Z coordinate as 32-bit floating point</src>
  </ins>

  <group name="IDP" title="8-bit dot product" dests="1" opcode="0xC2" unit="FMA">
    <desc>
      8-bit integer dot product between 4 channel vectors, intended for machine
      learning. Available in both unsigned and signed variants, controlling
      sign-extension/zero-extension behaviour to the final 32-bit destination.
      Saturation is available. Corresponds to the `cl_arm_integer_dot_product_*`
      family of OpenCL extensions. Not for actual use, just for completeness.
      Instead, use your platform's neural accelerator.

      For $A, B \in \{ 0, \ldots, 255 \}^4$ and $\text{Accumulator} \in
      \mathbb{Z}$, calculates $(A \cdot B) + \text{Accumulator}$ and optionally
      saturates.
    </desc>
    <ins name="IDP.v4s8" opcode2="0"/>
    <ins name="IDP.v4u8" opcode2="1"/>
    <src>A</src>
    <src>B</src>
    <src>Accumulator</src>
    <saturate/>
  </group>

  <group name="ICMP" title="Unsigned integer compare" dests="1" unit="CVT">
    <desc>
      Evaluates the given condition, do a logical and/or with the condition in
      the result source, and return in the given result type (integer
      one, integer minus one, or floating-point one). The third source is useful
      for chaining together conditions without intermediate bitwise arithmetic;
      when this is not desired, tie it to zero and use the OR combine mode (do
      not set the `.and` modifier).

      The sequence modifier `.seq` is used to construct 64-bit compares in 2
      `ICMP.u32` instructions, in conjunction with the `u1` result type on the
      low half, the `m1` result type on the high half, and the result of the low
      half comparison passed as the third source. For comparisons other than
      64-bit, do not set the `.seq` modifier and do not use the `u1` result
      type.
    </desc>
    <ins name="ICMP.u32" opcode="0xF0"/>
    <ins name="ICMP.v2u16" opcode="0xF1"/>
    <ins name="ICMP.v4u8" opcode="0xF2"/>
    <cmp/>
    <result_type/>
    <mod name="and" start="24" size="1"/>
    <mod name="seq" start="25" size="1"/>
    <src widen="true">A</src>
    <src widen="true">B</src>
    <src>C</src>
  </group>

  <group name="FCMP" title="Floating-point compare" dests="1" unit="CVT">
    <desc>
      Evaluates the given condition, do a logical and/or with the condition in
      the result source, and return in the given result type (integer
      one, integer minus one, or floating-point one). The third source is useful
      for chaining together conditions without intermediate bitwise arithmetic;
      when this is not desired, tie it to zero and use the OR combine mode (do
      not set the `.and` modifier).
    </desc>
    <ins name="FCMP.f32" opcode="0xF4"/>
    <ins name="FCMP.v2f16" opcode="0xF5"/>
    <cmp/>
    <result_type/>
    <mod name="and" start="24" size="1"/>
    <src absneg="true" swizzle="true">A</src>
    <src absneg="true" swizzle="true">B</src>
    <src>C</src>
  </group>

  <group name="ICMP" title="Signed integer compare" dests="1" unit="CVT">
    <desc>
      Evaluates the given condition, do a logical and/or with the condition in
      the result source, and return in the given result type (integer
      one, integer minus one, or floating-point one). The third source is useful
      for chaining together conditions without intermediate bitwise arithmetic;
      when this is not desired, tie it to zero and use the OR combine mode (do
      not set the `.and` modifier).

      The sequence modifier `.seq` is used to construct signed 64-bit compares
      in 1 `ICMP.u32` and 1 `ICMP.s32` instruction, in conjunction with the `u1`
      result type on the low half, the `m1` result type on the high half, and
      the result of the low half comparison passed as the third source. For
      comparisons other than 64-bit, do not set the `.seq` modifier and do not
      use the `u1` result type.
    </desc>
    <ins name="ICMP.s32" opcode="0xF8"/>
    <ins name="ICMP.v2s16" opcode="0xF9"/>
    <ins name="ICMP.v4s8" opcode="0xFA"/>
    <cmp/>
    <result_type/>
    <mod name="and" start="24" size="1"/>
    <mod name="seq" start="25" size="1"/>
    <src widen="true">A</src>
    <src widen="true">B</src>
    <src>C</src>
  </group>

  <ins name="IADD_IMM.i32" title="Integer addition with immediate" dests="1" opcode="0x110" unit="CVT">
    <desc>
      Adds an arbitrary 32-bit immediate embedded within the instruction stream.
      If no modifiers are required, this is preferred to `IADD.i32` with a
      constant accessed as a uniform. However, if the constant is available
      inline, `IADD.f32` is preferred.

      `IADD_IMM.i32` with the source tied to zero is the canonical immediate move.
    </desc>
    <src>A</src>
    <imm name="constant" start="8" size="32"/>
  </ins>

  <ins name="IADD_IMM.v2i16" title="Integer addition with immediate" dests="1" opcode="0x111" unit="CVT">
    <desc>
      Adds an arbitrary pair of 16-bit immediates embedded within the
      instruction stream. If no modifiers are required, this is preferred to
      `IADD.v2i16` with a constant accessed as a uniform. However, if the
      constant is available inline, `IADD.v2i16` is preferred. Adding only a
      single 16-bit constant requires replication of the constant.
    </desc>
    <src>A</src>
    <imm name="constant" start="8" size="32"/>
  </ins>

  <ins name="IADD_IMM.v4i8" title="Integer addition with immediate" dests="1" opcode="0x112" unit="CVT">
    <desc>
      Adds an arbitrary quad of 8-bit immediates embedded within the
      instruction stream. If no modifiers are required, this is preferred to
      `IADD.v4i8` with a constant accessed as a uniform. However, if the
      constant is available inline, `IADD.v4i8` is preferred. Adding only a
      single 8-bit constant requires replication of the constant.
    </desc>
    <src>A</src>
    <imm name="constant" start="8" size="32"/>
  </ins>

  <ins name="FADD_IMM.f32" title="Floating-point addition with immediate" dests="1" opcode="0x114" unit="FMA">
    <desc>
      Adds an arbitrary 32-bit immediate embedded within the instruction stream.
      If no modifiers are required, this is preferred to `FADD.f32` with a
      constant accessed as a uniform. However, if the constant is available
      inline, `FADD.f32` is preferred.
    </desc>
    <src>A</src>
    <imm name="constant" start="8" size="32"/>
  </ins>

  <ins name="FADD_IMM.v2f16" title="Floating-point addition with immediate" dests="1" opcode="0x115" unit="FMA">
    <desc>
      Adds an arbitrary pair of 16-bit immediates embedded within the
      instruction stream. If no modifiers are required, this is preferred to
      `FADD.v2f16` with a constant accessed as a uniform. However, if the
      constant is available inline, `FADD.v2f16` is preferred. Adding only a
      single 16-bit constant requires replication of the constant.
    </desc>
    <src float="true">A</src>
    <imm name="constant" start="8" size="32"/>
  </ins>

  <ins name="ATOM1_RETURN.i32" title="Atomic operations on memory with 1" opcode="0x69" opcode2="3" unit="LS">
    <slot/>
    <sr_count/>
    <atom_opc_1/>
    <mod name="memory_width" start="128" size="1" implied="true"/>

    <!-- Optional for ATOM1.i32, in which sr_count must be 0 -->
    <sr write="true"/>
    <src size="64">64-bit address to operate on</src>
    <imm name="offset" start="8" size="8"/>
  </ins>

  <ins name="ATOM1_RETURN.i64" title="Atomic operations on memory with 1" opcode="0x69" opcode2="5" unit="LS">
    <slot/>
    <sr_count/>
    <atom_opc_1/>
    <mod name="memory_width" start="128" size="1" implied="true"/>

    <!-- Optional for ATOM1.i64, in which sr_count must be 0 -->
    <sr write="true"/>
    <src size="64">64-bit address to operate on</src>
    <imm name="offset" start="8" size="8"/>
  </ins>

  <ins name="ATOM.i32" title="Atomic operations on memory" opcode="0x68" opcode2="3" unit="LS">
    <slot/>
    <sr_count/>
    <atom_opc/>
    <mod name="memory_width" start="128" size="1" implied="true"/>

    <sr read="true"/>
    <src size="64">64-bit address to operate on</src>
    <imm name="offset" start="8" size="8"/>
  </ins>

  <ins name="ATOM.i64" title="Atomic operations on memory" opcode="0x68" opcode2="5" unit="LS">
    <slot/>
    <sr_count/>
    <atom_opc/>
    <mod name="memory_width" start="128" size="1" implied="true"/>

    <sr read="true"/>
    <src size="64">64-bit address to operate on</src>
    <imm name="offset" start="8" size="8"/>
  </ins>

  <ins name="ATOM_RETURN.i32" title="Atomic operations on memory" opcode="0x120" opcode2="3" unit="LS">
    <slot/>
    <sr_count/>
    <sr_write_count/>

    <!-- Only valid with .xchg to implement ACMPXCHG -->
    <mod name="compare" start="26" size="1"/>

    <atom_opc/>
    <mod name="memory_width" start="128" size="1" implied="true"/>

    <sr write="true" flags="false"/>
    <sr read="true" flags="rw"/>
    <src size="64">64-bit address to operate on</src>
    <imm name="offset" start="8" size="8"/>
  </ins>

  <ins name="ATOM_RETURN.i64" title="Atomic operations on memory" opcode="0x120" opcode2="5" unit="LS">
    <slot/>
    <sr_count/>
    <sr_write_count/>
    <mod name="compare" start="26" size="1"/>
    <atom_opc/>
    <mod name="memory_width" start="128" size="1" implied="true"/>

    <sr write="true" flags="false"/>
    <sr read="true" flags="rw"/>
    <src size="64">64-bit address to operate on</src>
    <imm name="offset" start="8" size="8"/>
  </ins>

  <ins name="TEX_FETCH" title="Texel fetch" opcode="0x125" unit="T">
    <desc>Unfiltered textured instruction.</desc>
    <slot/>
    <skip/>
    <register_type/>
    <register_width/>
    <write_mask/>
    <dimension/>
    <wide_indices/>
    <array_enable/>
    <texel_offset/>

    <!-- Leave secondary_register_width as 0 -->
    <sr_count/>
    <sr_write_count/>

    <sr write="true" flags="false"/>
    <sr read="true" flags="false"/>
    <src size="64">Image to read from</src>
  </ins>

  <ins name="TEX_SINGLE" title="Texture load" opcode="0x128" unit="T">
    <desc>Ordinary texturing instruction using a sampler.</desc>
    <slot/>
    <skip/>
    <register_type/>
    <register_width/>
    <write_mask/>
    <dimension/>
    <wide_indices/>
    <array_enable/>
    <texel_offset/>
    <shadow/>
    <lod_mode/>

    <!-- Leave secondary_register_width as 0 -->
    <sr_count/>
    <sr_write_count/>

    <sr write="true" flags="false"/>
    <sr read="true" flags="false"/>
    <src size="64">Image to read from</src>
  </ins>

  <ins name="TEX_GATHER" title="Texel gather" opcode="0x129" unit="T">
    <desc>Texture gather instruction.</desc>
    <slot/>
    <skip/>
    <register_type/>
    <register_width/>
    <write_mask/>
    <dimension/>
    <wide_indices/>
    <array_enable/>
    <texel_offset/>
    <integer_coordinates/>
    <fetch_component/>
    <shadow/>

    <!-- Leave secondary_register_width as 0 -->
    <sr_count/>
    <sr_write_count/>

    <sr write="true" flags="false"/>
    <sr read="true" flags="false"/>
    <src size="64">Image to read from</src>
  </ins>

  <ins name="TEX_DUAL" title="Dual texture" opcode="0x12F" unit="T">
    <desc>Pair of texture instructions.</desc>
    <slot/>
    <skip/>
    <register_type/>
    <register_width/>
    <secondary_register_width/>
    <write_mask/>
    <dimension/>
    <wide_indices/>
    <array_enable/>
    <texel_offset/>
    <shadow/>
    <lod_mode/>

    <sr_count/>
    <sr_write_count/>

    <sr write="true" flags="false"/>
    <sr read="true" flags="false"/>
    <src size="64">Image to read from</src>
  </ins>

  <ins name="VAR_TEX_BUF_SINGLE" title="Fused varying-texturing" opcode="0x130" unit="VT">
    <desc>
      Only works for FP32 varyings. Performance characteristics are similar
      to LD_VAR_BUF_IMM_F32.v2.f32 followed by TEX, using both V and T units.
    </desc>
    <slot/>
    <skip/>
    <sample_and_update/>
    <register_type/>
    <vartex_register_width/>
    <dimension/>
    <array_enable/>
    <shadow/>
    <lod_mode/>

    <sr_write_count/>

    <sr write="true"/>
    <src size="64">Image to read from</src>
    <src>Varying offset</src>
  </ins>

  <ins name="VAR_TEX_BUF_GATHER" title="Fused varying-texturing" opcode="0x131" unit="VT">
    <desc>
      Only works for FP32 varyings. Performance characteristics are similar
      to LD_VAR_BUF_IMM_F32.v2.f32 followed by TEX, using both V and T units.
    </desc>
    <slot/>
    <skip/>
    <sample_and_update/>
    <register_type/>
    <vartex_register_width/>
    <dimension/>
    <array_enable/>
    <integer_coordinates/>
    <fetch_component/>
    <shadow/>

    <sr_write_count/>

    <sr write="true"/>
    <src size="64">Image to read from</src>
    <src>Varying offset</src>
  </ins>

  <ins name="VAR_TEX_BUF_GRADIENT" title="Fused varying-texturing" opcode="0x132" unit="VT">
    <desc>
      Only works for FP32 varyings. Performance characteristics are similar
      to LD_VAR_BUF_IMM_F32.v2.f32 followed by TEX, using both V and T units.
    </desc>
    <slot/>
    <skip/>
    <sample_and_update/>
    <register_type/>
    <vartex_register_width/>
    <dimension/>
    <array_enable/>
    <shadow/>
    <lod_bias_disable/>
    <lod_clamp_disable/>

    <sr_write_count/>

    <sr write="true"/>
    <src size="64">Image to read from</src>
    <src>Varying offset</src>
  </ins>

  <ins name="VAR_TEX_BUF_DUAL" title="Fused varying-texturing" opcode="0x137" unit="VT">
    <desc>
      Only works for FP32 varyings. Performance characteristics are similar
      to LD_VAR_BUF_IMM_F32.v2.f32 followed by TEX_DUAL, using both V and T units.
    </desc>
    <slot/>
    <skip/>
    <sample_and_update/>
    <register_type/>
    <vartex_register_width/>
    <dimension/>
    <array_enable/>
    <shadow/>
    <lod_mode/>

    <sr_write_count/>

    <sr write="true"/>
    <src size="64">Image to read from</src>
    <src>Varying offset</src>
  </ins>

  <ins name="VAR_TEX_SINGLE" title="Fused varying-texturing" opcode="0x138" unit="VT">
    <desc>
      Only works for FP32 varyings. Performance characteristics are similar
      to LD_VAR_IMM_F32.v2.f32 followed by TEX, using both V and T units.
    </desc>
    <slot/>
    <skip/>
    <sample_and_update/>
    <register_type/>
    <vartex_register_width/>
    <dimension/>
    <array_enable/>
    <shadow/>
    <lod_mode/>

    <sr_write_count/>

    <sr write="true"/>
    <src size="64">Image to read from</src>
    <src>Varying offset</src>
  </ins>

  <ins name="VAR_TEX_GATHER" title="Fused varying-texturing" opcode="0x139" unit="VT">
    <desc>
      Only works for FP32 varyings. Performance characteristics are similar
      to LD_VAR_IMM_F32.v2.f32 followed by TEX, using both V and T units.
    </desc>
    <slot/>
    <skip/>
    <sample_and_update/>
    <register_type/>
    <vartex_register_width/>
    <dimension/>
    <array_enable/>
    <integer_coordinates/>
    <fetch_component/>
    <shadow/>

    <sr_write_count/>

    <sr write="true"/>
    <src size="64">Image to read from</src>
    <src>Varying offset</src>
  </ins>

  <ins name="VAR_TEX_GRADIENT" title="Fused varying-texturing" opcode="0x13A" unit="VT">
    <desc>
      Only works for FP32 varyings. Performance characteristics are similar
      to LD_VAR_IMM_F32.v2.f32 followed by TEX, using both V and T units.
    </desc>
    <slot/>
    <skip/>
    <sample_and_update/>
    <register_type/>
    <vartex_register_width/>
    <dimension/>
    <array_enable/>
    <shadow/>
    <lod_bias_disable/>
    <lod_clamp_disable/>

    <sr_write_count/>

    <sr write="true"/>
    <src size="64">Image to read from</src>
    <src>Varying offset</src>
  </ins>

  <ins name="VAR_TEX_DUAL" title="Fused varying-texturing" opcode="0x13F" unit="VT">
    <desc>
      Only works for FP32 varyings. Performance characteristics are similar
      to LD_VAR_IMM_F32.v2.f32 followed by TEX_DUAL, using both V and T units.
    </desc>
    <slot/>
    <skip/>
    <sample_and_update/>
    <register_type/>
    <vartex_register_width/>
    <dimension/>
    <array_enable/>
    <shadow/>
    <lod_mode/>

    <sr_write_count/>

    <sr write="true"/>
    <src size="64">Image to read from</src>
    <src>Varying offset</src>
  </ins>

  <ins name="FMA_RSCALE.f32" title="Fused floating-point multiply add with exponent bias" dests="1" opcode="0x160" unit="FMA">
    <desc>
      First calculates $A \cdot B + C$ and then biases the exponent by D. Used in
      special transcendental function sequences. It should not be used for
      general code as its special case handling differs from two back-to-back
      `FMA.f32` operations. Equivalent to `FMA.f32` back-to-back with
      `LDEXP.f32`
    </desc>
    <clamp/>
    <src absneg="true">A</src>
    <src absneg="true">B</src>
    <src absneg="true">C</src>
    <src>D</src>
  </ins>

  <ins name="FMA_RSCALE_N.f32" title="Fused floating-point multiply add with exponent bias and zero override" dests="1" opcode="0x161" unit="FMA">
    <desc>
      First calculates $A \cdot B + C$ and then biases the exponent by D. If $A
      = 0$ or $B = 0$, the multiply $A \cdot B$ is treated as zero even if an
      ordinary multiply would return NaN. Used in special transcendental
      function sequences. It should not be used for general code as its special
      case handling differs from two back-to-back `FMA.f32` operations.
      Equivalent to `FMA.f32` back-to-back with `LDEXP.f32`
    </desc>
    <clamp/>
    <src absneg="true">A</src>
    <src absneg="true">B</src>
    <src absneg="true">C</src>
    <src>D</src>
  </ins>

  <ins name="FMA_RSCALE_LEFT.f32" title="Fused floating-point multiply add with exponent bias and asymmetric zero handling" dests="1" opcode="0x162" unit="FMA">
    <desc>
      First calculates $A \cdot B + C$ and then biases the exponent by D. If $A
      = 0$ or $B = 0$, the multiply is treated as $A$ even if an
      ordinary multiply would return NaN. Used in special transcendental
      function sequences. It should not be used for general code as its special
      case handling differs from two back-to-back `FMA.f32` operations.
      Equivalent to `FMA.f32` back-to-back with `LDEXP.f32`
    </desc>
    <clamp/>
    <src absneg="true">A</src>
    <src absneg="true">B</src>
    <src absneg="true">C</src>
    <src>D</src>
  </ins>

  <ins name="FMA_RSCALE_SCALE16.f32" title="Fused floating-point multiply add with 16-bit exponent bias" dests="1" opcode="0x163" unit="FMA">
    <desc>
      First calculates $A \cdot B + C$ and then biases the exponent by D,
      interpreted as a 16-bit value. Used in special transcendental function
      sequences. It should not be used for general code as its special case
      handling differs from two back-to-back `FMA.f32` operations.  Equivalent
      to `FMA.f32` back-to-back with `LDEXP.f32`
    </desc>
    <clamp/>
    <src absneg="true">A</src>
    <src absneg="true">B</src>
    <src absneg="true">C</src>
    <src>D</src>
  </ins>

</valhall>