fix(@typegpu/gl): Reserve all keywords defined in the GLSL ES 3.0 spec

[iwoplaza]

No description provided.

[github-actions]

pkg.pr.new

packages
Ready to be installed by your favorite package manager ⬇️

https://pkg.pr.new/software-mansion/TypeGPU/@typegpu/gl@d151ba82edee343f21aa453971fb162d3c0717e8

benchmark
view benchmark

commit
view commit

[Copilot]

Pull request overview

This PR expands @typegpu/gl’s GLSL generator identifier reservation to cover the full set of keywords (and built-in globals) from the GLSL ES 3.0 specification, preventing generated names from colliding with reserved tokens in emitted GLSL.

Changes:

• Added a reservedKeywords list (GLSL ES 3.0 keywords + built-in globals) in glslGenerator.ts.
• Updated GlslGenerator.initGenerator() to reserve all entries from that list via ctx.reserveIdentifier(...).

---

💡 Add a code-review agent skill or configure MCP servers for context-aware, tailored reviews. Learn more in the docs.

[pullfrog]

ℹ️ ВAppпру two typos inline — both inside the new keyword list.

Reviewed changes

• Replaced the one-off reserveIdentifier('gl_Position', 'global') / reserveIdentifier('sample', 'global') calls with a module-level reservedKeywords array enumerating the globals, defined keywords, and future-reserved type/qualifier words from GLSL ES 3.0, reserved in a loop inside the generator constructor.

The list goes one word too far in a couple of places (see the inline comment and nitpick).

  ^{｜ Fix all ➔ ｜ Fix 👍s ➔ ｜ View workflow run ｜ Using DeepSeek Flash (free via Pullfrog for OSS) ｜ 𝕏}

[github-actions]

Resolution Time Benchmark

---
config:
  themeVariables:
    xyChart:
      plotColorPalette: "#E63946, #3B82F6, #059669"
---
xychart
  title "Random Branching (🔴 PR | 🔵 main | 🟢 release)"
  x-axis "max depth" [1, 2, 3, 4, 5, 6, 7, 8]
  y-axis "time (ms)"
  line [1.01, 1.93, 4.38, 6.55, 7.20, 10.98, 25.09, 23.63]
  line [0.93, 1.86, 4.27, 6.53, 7.90, 12.47, 23.31, 23.28]
  line [0.95, 1.92, 4.16, 6.70, 7.96, 12.49, 23.35, 24.58]

Loading

---
config:
  themeVariables:
    xyChart:
      plotColorPalette: "#E63946, #3B82F6, #059669"
---
xychart
  title "Linear Recursion (🔴 PR | 🔵 main | 🟢 release)"
  x-axis "max depth" [1, 2, 3, 4, 5, 6, 7, 8]
  y-axis "time (ms)"
  line [0.34, 0.54, 0.71, 0.76, 1.03, 1.14, 1.33, 1.59]
  line [0.28, 0.45, 0.68, 0.76, 1.04, 1.11, 1.30, 1.50]
  line [0.31, 0.53, 0.67, 0.82, 1.03, 1.15, 1.38, 1.49]

Loading

---
config:
  themeVariables:
    xyChart:
      plotColorPalette: "#E63946, #3B82F6, #059669"
---
xychart
  title "Full Tree (🔴 PR | 🔵 main | 🟢 release)"
  x-axis "max depth" [1, 2, 3, 4, 5, 6, 7, 8]
  y-axis "time (ms)"
  line [0.87, 2.17, 3.42, 7.18, 12.24, 25.75, 54.47, 110.95]
  line [0.87, 2.05, 3.50, 7.25, 12.67, 26.88, 55.09, 111.32]
  line [0.85, 2.15, 3.37, 7.15, 12.81, 26.36, 55.32, 111.50]

Loading

[github-actions]

Bundle size comparison (import * as ... in PR vs import * as ... in target):

🟢 Decreased	➖ Unchanged	🔴 Increased	❔ Unknown
0	325	0	0

import { ... } in PR vs import * as ... in PR (is the library tree-Shakeable?):

Test	tsdown
tgpu_init.ts	263.10 kB ($${\color{green}-3.5\%}$$)
tgpu_initFromDevice.ts	262.56 kB ($${\color{green}-3.7\%}$$)
tgpu_resolve.ts	161.75 kB ($${\color{green}-40.7\%}$$)
tgpu_resolveWithContext.ts	161.68 kB ($${\color{green}-40.7\%}$$)
tgpu_bindGroupLayout.ts	62.32 kB ($${\color{green}-77.1\%}$$)
tgpu_mutableAccessor.ts	57.04 kB ($${\color{green}-79.1\%}$$)
tgpu_accessor.ts	57.04 kB ($${\color{green}-79.1\%}$$)
tgpu_privateVar.ts	55.73 kB ($${\color{green}-79.6\%}$$)
tgpu_workgroupVar.ts	55.73 kB ($${\color{green}-79.6\%}$$)
tgpu_const.ts	55.15 kB ($${\color{green}-79.8\%}$$)
tgpu_lazy.ts	54.95 kB ($${\color{green}-79.8\%}$$)
tgpu_fragmentFn.ts	39.68 kB ($${\color{green}-85.4\%}$$)
tgpu_fn.ts	39.62 kB ($${\color{green}-85.5\%}$$)
tgpu_vertexFn.ts	39.50 kB ($${\color{green}-85.5\%}$$)
tgpu_computeFn.ts	39.20 kB ($${\color{green}-85.6\%}$$)
tgpu_vertexLayout.ts	28.33 kB ($${\color{green}-89.6\%}$$)
tgpu_comptime.ts	15.93 kB ($${\color{green}-94.2\%}$$)
tgpu_unroll.ts	1.75 kB ($${\color{green}-99.4\%}$$)
tgpu_slot.ts	1.70 kB ($${\color{green}-99.4\%}$$)

If you wish to run a comparison for other, slower bundlers, run the 'Tree-shake test' from the GitHub Actions menu.

[pullfrog]

✅ No new issues found. Prior feedback addressed in d151ba82.

Reviewed changes

• Fixed the pler2DMSArray typo to sampler2DMSArray, restoring reservation of the real GLSL ES keyword (it now matches its isampler2DMSArray / usampler2DMSArray siblings).
• Removed the duplicate volatile entry.

Both items raised in the previous review are cleanly resolved; the keyword list now has each real GLSL ES 3.0 word once and correctly spelled.

  ^{｜ View workflow run ｜ Using DeepSeek Flash (free via Pullfrog for OSS) ｜ 𝕏}

[aleksanderkatan]

🔑L

[aleksanderkatan]

Function names and variable names already cannot start with gl_ (though we don't enforce this yet):

Identifiers starting with “gl_” are reserved for use by OpenGL ES, and may not be declared in a shader as either a variable or a function.

  it('...', () => {
    const gl_fn = () => {
      'use gpu';
      const gl_color = d.vec4f(1, 0, 0, 1);
      return gl_color;
    };

    const result = tgpu.resolveWithContext([gl_fn], glOptions());
    expect(result.code).toMatchInlineSnapshot(`
      "vec4 gl_fn() {
        vec4 gl_color = vec4(1, 0, 0, 1);
        return gl_color;
      }"
    `);
  });

Do we plan on banning them in the future?

[aleksanderkatan]

Please add at least one test

  it('renames invalid glsl identifiers', () => {
    const gl_InstanceID = () => {
      'use gpu';
      const gl_VertexID = 1;
      const iimage2D = 2;
      return gl_VertexID + iimage2D;
    };

    expect(tgpu.resolve([gl_InstanceID], glOptions())).toMatchInlineSnapshot(`
      "int gl_InstanceID_1() {
        int gl_VertexID_1 = 1;
        int iimage2D_1 = 2;
        return (gl_VertexID_1 + iimage2D_1);
      }"
    `);
  });