add-uint-support

Ajouter la prise en charge des types entiers non signés (uint) aux opérateurs PyTorch en mettant à jour les macros AT_DISPATCH. Utiliser lors de l'ajout de la prise en charge des types uint16, uint32, uint64 à…

npx skills add https://github.com/pytorch/pytorch --skill add-uint-support

Add Unsigned Integer (uint) Support to Operators

This skill helps add support for unsigned integer types (uint16, uint32, uint64) to PyTorch operators by updating their AT_DISPATCH macros.

When to use this skill

Use this skill when:

  • Adding uint16, uint32, or uint64 support to an operator
  • User mentions "unsigned types", "uint support", "barebones unsigned types"
  • Enabling support for kUInt16, kUInt32, kUInt64 in kernels
  • Working with operator implementations that need expanded type coverage

Quick reference

Add unsigned types to existing dispatch:

// Before
AT_DISPATCH_V2(dtype, "op", AT_WRAP([&]() {
  kernel<scalar_t>();
}), AT_EXPAND(AT_ALL_TYPES));

// After (method 1: add unsigned types explicitly)
AT_DISPATCH_V2(dtype, "op", AT_WRAP([&]() {
  kernel<scalar_t>();
}), AT_EXPAND(AT_ALL_TYPES), AT_EXPAND(AT_BAREBONES_UNSIGNED_TYPES));

// After (method 2: use V2 integral types if AT_INTEGRAL_TYPES present)
AT_DISPATCH_V2(dtype, "op", AT_WRAP([&]() {
  kernel<scalar_t>();
}), AT_EXPAND(AT_INTEGRAL_TYPES_V2), AT_EXPAND(AT_FLOATING_TYPES));

Type group reference

Unsigned type groups:

  • AT_BAREBONES_UNSIGNED_TYPES: kUInt16, kUInt32, kUInt64
  • AT_INTEGRAL_TYPES_V2: AT_INTEGRAL_TYPES + AT_BAREBONES_UNSIGNED_TYPES

Relationship:

AT_INTEGRAL_TYPES          // kByte, kChar, kInt, kLong, kShort
AT_BAREBONES_UNSIGNED_TYPES  // kUInt16, kUInt32, kUInt64
AT_INTEGRAL_TYPES_V2       // INTEGRAL_TYPES + BAREBONES_UNSIGNED_TYPES

Instructions

Step 1: Determine if conversion to V2 is needed

Check if the file uses AT_DISPATCH_V2:

If using old AT_DISPATCH:

  • First convert to AT_DISPATCH_V2 using the at-dispatch-v2 skill
  • Then proceed with adding uint support

If already using AT_DISPATCH_V2:

  • Proceed directly to Step 2

Step 2: Analyze the current dispatch macro

Identify what type groups are currently in use:

AT_DISPATCH_V2(dtype, "op", AT_WRAP([&]() {
  // body
}), AT_EXPAND(AT_ALL_TYPES), kHalf, kBFloat16);
    ^^^^^^^^^^^^^^^^^^^^^^^^^
    Current type coverage

Common patterns:

  • AT_EXPAND(AT_ALL_TYPES) → includes AT_INTEGRAL_TYPES + AT_FLOATING_TYPES
  • AT_EXPAND(AT_INTEGRAL_TYPES) → signed integers only
  • AT_EXPAND(AT_FLOATING_TYPES) → floating point types

Step 3: Choose the uint addition method

Two approaches:

Method 1: Add AT_BAREBONES_UNSIGNED_TYPES explicitly

  • Use when: You want to be explicit about adding uint support
  • Add AT_EXPAND(AT_BAREBONES_UNSIGNED_TYPES) to the type list

Method 2: Substitute AT_INTEGRAL_TYPES with AT_INTEGRAL_TYPES_V2

  • Use when: The dispatch already uses AT_EXPAND(AT_INTEGRAL_TYPES)
  • More concise: replaces one type group with its superset
  • Only applicable if AT_INTEGRAL_TYPES is present

Step 4: Apply the transformation

Method 1 example:

// Before
AT_DISPATCH_V2(
    dtype,
    "min_values_cuda",
    AT_WRAP([&]() {
      kernel_impl<scalar_t>(iter);
    }),
    AT_EXPAND(AT_ALL_TYPES),
    kBFloat16, kHalf, kBool
);

// After (add unsigned types)
AT_DISPATCH_V2(
    dtype,
    "min_values_cuda",
    AT_WRAP([&]() {
      kernel_impl<scalar_t>(iter);
    }),
    AT_EXPAND(AT_ALL_TYPES),
    AT_EXPAND(AT_BAREBONES_UNSIGNED_TYPES),
    kBFloat16, kHalf, kBool
);

Method 2 example:

// Before
AT_DISPATCH_V2(
    dtype,
    "integral_op",
    AT_WRAP([&]() {
      kernel<scalar_t>();
    }),
    AT_EXPAND(AT_INTEGRAL_TYPES)
);

// After (substitute with V2)
AT_DISPATCH_V2(
    dtype,
    "integral_op",
    AT_WRAP([&]() {
      kernel<scalar_t>();
    }),
    AT_EXPAND(AT_INTEGRAL_TYPES_V2)
);

Step 5: Handle AT_ALL_TYPES vs individual type groups

If the dispatch uses AT_EXPAND(AT_ALL_TYPES):

  • AT_ALL_TYPES = AT_INTEGRAL_TYPES + AT_FLOATING_TYPES
  • To add uint: add AT_EXPAND(AT_BAREBONES_UNSIGNED_TYPES) to the list

If the dispatch separately lists INTEGRAL and FLOATING:

// Before
AT_EXPAND(AT_INTEGRAL_TYPES), AT_EXPAND(AT_FLOATING_TYPES)

// After (Method 2 preferred)
AT_EXPAND(AT_INTEGRAL_TYPES_V2), AT_EXPAND(AT_FLOATING_TYPES)

Step 6: Verify all dispatch sites

Check the file for ALL dispatch macros that need uint support:

  • Some operators have multiple dispatch sites (CPU, CUDA, different functions)
  • Apply the transformation consistently across all sites
  • Ensure each gets the same type coverage updates

Step 7: Validate the changes

Check that:

  • AT_DISPATCH_V2 format is used (not old AT_DISPATCH)
  • Unsigned types are added via one of the two methods
  • All relevant dispatch sites in the file are updated
  • Type groups use AT_EXPAND()
  • Arguments are properly formatted and comma-separated

Common patterns

Pattern 1: AT_ALL_TYPES + extras

// Before
AT_DISPATCH_V2(dtype, "op", AT_WRAP([&]() {
  kernel<scalar_t>();
}), AT_EXPAND(AT_ALL_TYPES), kHalf, kBFloat16);

// After
AT_DISPATCH_V2(dtype, "op", AT_WRAP([&]() {
  kernel<scalar_t>();
}), AT_EXPAND(AT_ALL_TYPES), AT_EXPAND(AT_BAREBONES_UNSIGNED_TYPES), kHalf, kBFloat16);

Pattern 2: Separate INTEGRAL + FLOATING

// Before
AT_DISPATCH_V2(dtype, "op", AT_WRAP([&]() {
  kernel<scalar_t>();
}), AT_EXPAND(AT_INTEGRAL_TYPES), AT_EXPAND(AT_FLOATING_TYPES));

// After
AT_DISPATCH_V2(dtype, "op", AT_WRAP([&]() {
  kernel<scalar_t>();
}), AT_EXPAND(AT_INTEGRAL_TYPES_V2), AT_EXPAND(AT_FLOATING_TYPES));

Pattern 3: Old dispatch needs conversion first

// Before (needs v2 conversion first)
AT_DISPATCH_ALL_TYPES_AND2(kHalf, kBFloat16, dtype, "op", [&]() {
  kernel<scalar_t>();
});

// After v2 conversion
AT_DISPATCH_V2(dtype, "op", AT_WRAP([&]() {
  kernel<scalar_t>();
}), AT_EXPAND(AT_ALL_TYPES), kHalf, kBFloat16);

// After adding uint support
AT_DISPATCH_V2(dtype, "op", AT_WRAP([&]() {
  kernel<scalar_t>();
}), AT_EXPAND(AT_ALL_TYPES), AT_EXPAND(AT_BAREBONES_UNSIGNED_TYPES), kHalf, kBFloat16);

Multiple dispatch sites example

For a file with multiple functions:

void min_values_kernel_cuda(TensorIterator& iter) {
  AT_DISPATCH_V2(iter.dtype(), "min_values_cuda", AT_WRAP([&]() {
    impl<scalar_t>(iter);
  }), AT_EXPAND(AT_ALL_TYPES), AT_EXPAND(AT_BAREBONES_UNSIGNED_TYPES), kBFloat16, kHalf);
  //                           ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
  //                           Added uint support
}

void min_launch_kernel(TensorIterator &iter) {
  AT_DISPATCH_V2(iter.input_dtype(), "min_cuda", AT_WRAP([&]() {
    gpu_reduce_kernel<scalar_t>(iter);
  }), AT_EXPAND(AT_ALL_TYPES), AT_EXPAND(AT_BAREBONES_UNSIGNED_TYPES), kBFloat16, kHalf);
  //                           ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
  //                           Added uint support here too
}

Decision tree

Use this decision tree to determine the approach:

Is the file using AT_DISPATCH_V2?
├─ No → Use at-dispatch-v2 skill first, then continue
└─ Yes
   └─ Does it use AT_EXPAND(AT_INTEGRAL_TYPES)?
      ├─ Yes → Replace with AT_EXPAND(AT_INTEGRAL_TYPES_V2)
      └─ No → Add AT_EXPAND(AT_BAREBONES_UNSIGNED_TYPES) to type list

Edge cases

Case 1: Dispatch with only floating types

If the operator only supports floating point types, don't add uint support:

// Leave as-is - floating point only operator
AT_DISPATCH_V2(dtype, "float_op", AT_WRAP([&]() {
  kernel<scalar_t>();
}), AT_EXPAND(AT_FLOATING_TYPES), kHalf);

Case 2: Complex types present

Unsigned types work alongside complex types:

AT_DISPATCH_V2(dtype, "op", AT_WRAP([&]() {
  kernel<scalar_t>();
}), AT_EXPAND(AT_ALL_TYPES),
    AT_EXPAND(AT_BAREBONES_UNSIGNED_TYPES),
    AT_EXPAND(AT_COMPLEX_TYPES),
    kHalf, kBFloat16);

Case 3: Already has uint support

Check if uint types are already present:

  • If AT_INTEGRAL_TYPES_V2 is used → already has uint support
  • If AT_BAREBONES_UNSIGNED_TYPES is already in list → already has uint support
  • Skip the file if uint support is already present

Workflow

When asked to add uint support:

  1. Read the target file
  2. Check if using AT_DISPATCH_V2:
    • If not → use at-dispatch-v2 skill first
  3. Identify all dispatch macro sites
  4. For each dispatch:
    • Analyze current type groups
    • Choose method (add BAREBONES_UNSIGNED or upgrade to V2)
    • Apply transformation with Edit tool
  5. Show the user the changes
  6. Explain what was modified

Important notes

  • Always check if v2 conversion is needed first
  • Apply changes consistently across all dispatch sites in the file
  • Method 2 (AT_INTEGRAL_TYPES_V2) is cleaner when applicable
  • Method 1 (explicit AT_BAREBONES_UNSIGNED_TYPES) is more explicit
  • Unsigned types are: kUInt16, kUInt32, kUInt64 (not kByte which is uint8)
  • Some operators may not semantically support unsigned types - use judgment

Testing

After adding uint support, the operator should accept uint16, uint32, and uint64 tensors. The user is responsible for functional testing.

Plus de skills de pytorch

zephyr
pytorch
Construire et configurer ExecuTorch en tant que module Zephyr RTOS pour cartes embarquées. Utiliser lors de la configuration d’un espace de travail Zephyr avec ET, de l’ajout de support de carte (superpositions,…
aoti-debug
pytorch
Déboguer les erreurs et plantages d'AOTInductor (AOTI). Utiliser en cas de segfaults AOTI, d'erreurs de non-correspondance de périphérique, d'échecs de chargement de constantes ou d'erreurs d'exécution provenant de…
skill-writer
pytorch
Guide pour créer des Agent Skills bien structurées pour Claude Code avec les bonnes pratiques et la validation. Couvre l'ensemble du cycle de vie d'une Skill : cadrage, structure des fichiers, validation du frontmatter YAML, organisation du contenu et procédures de test. Impose des règles de nommage strictes (minuscules, traits d'union, 64 caractères max) et des exigences de description (déclencheurs spécifiques, types de fichiers, clauses "quoi" et "quand"). Fournit des modèles pour les motifs courants, notamment les Skills en lecture seule, les Skills basées sur des scripts et les Skills multi-fichiers avec...
triaging-issues
pytorch
Trie les problèmes GitHub en les acheminant vers les équipes d'astreinte, en appliquant des étiquettes et en fermant les questions. À utiliser lors du traitement de nouveaux problèmes PyTorch ou lorsqu'on vous demande de trier un…
wheel-size-analyzer
pytorch
Analyser les tailles des wheels PyTorch nightly sur une plage de dates à l’aide de l’API des artefacts de GitHub Actions. À utiliser pour suivre les changements de taille des binaires, identifier la taille des wheels…
release-go-live-binary-build-matrix
pytorch
Mettre à jour tools/scripts/generate_binary_build_matrix.py lorsqu'une version de PyTorch est mise en ligne. Avance CURRENT_STABLE_VERSION vers la nouvelle version stable, promeut la…
pr-review
pytorch
Examiner les demandes de tirage PyTorch pour la qualité du code, la couverture des tests, la sécurité et la compatibilité ascendante. Utiliser lors de l'examen des PR, lorsqu'on vous demande de réviser des modifications de code,…
qualcomm
pytorch
Construire, tester ou développer le backend QNN (Qualcomm AI Engine Direct). Utiliser lorsque vous travaillez sur backends/qualcomm/, en construisant QNN (utiliser…