FilesExpand file tree

main

/

linux-x86-render-smoke.yml

Copy path

View runs

Blame

More file actions

Blame

More file actions

Latest commit

 

History

History

History

140 lines (127 loc) · 5.62 KB

main

/

linux-x86-render-smoke.yml

Copy path

Top

File metadata and controls

• Code
• Blame

140 lines (127 loc) · 5.62 KB

Raw

Copy raw file

Download raw file

Open symbols panel

Edit and raw actions

1

2

3

4

5

6

7

8

9

10

11

12

13

14

15

16

17

18

19

20

21

22

23

24

25

26

27

28

29

30

31

32

33

34

35

36

37

38

39

40

41

42

43

44

45

46

47

48

49

50

51

52

53

54

55

56

57

58

59

60

61

62

63

64

65

66

67

68

69

70

71

72

73

74

75

76

77

78

79

80

81

82

83

84

85

86

87

88

89

90

91

92

93

94

95

96

97

98

99

100

101

102

103

104

105

106

107

108

109

110

111

112

113

114

115

116

117

118

119

120

121

122

123

124

125

126

127

128

129

130

131

132

133

134

135

136

137

138

139

140

name: Linux x86 Render Smoke

# Proves the real WASM viewer renders on Linux (#247).

#

# On Linux the viewer runs under WebKitGTK, where WebGL2 is unavailable on an

# OffscreenCanvas, so the worker posts frames for the main thread to draw. That

# path had no CI coverage at all: the wheel smoke that renders a sketch is

# enabled only in macos-arm-build.yml. #247 therefore reached a release branch

# and could only be found by hand.

#

# This lane compiles a sketch and renders it in the shipped viewer under Xvfb

# with software GL, then requires the screenshot to contain lit pixels

# (ci/check_render.py). A viewer that loads and draws nothing fails here, which

# is exactly how #247 and #250 presented.

#

# It renders tests/fixtures/wasm_no_screenmap: a plain strip, which is also the

# #250 regression (its layout only exists after the first frame). A heavier

# sketch is deliberately not used — a 16x16 grid through llvmpipe on a 2-core

# runner starves the page's event loop and the harness times out.

on:

push:

branches: [main]

pull_request:

types: [opened, synchronize, reopened, labeled, unlabeled]

workflow_dispatch:

inputs:

candidate_sha:

description: Exact 40-character commit SHA to validate

required: true

type: string

# One live run per workflow per ref: a new push to a branch or PR cancels the

# superseded run instead of queuing behind it in the shared account-wide runner

# pool. Runs on main get a unique group, so they are never cancelled or coalesced.

concurrency:

group: ${{ github.workflow }}-${{ github.ref == 'refs/heads/main' && github.run_id || github.ref }}

cancel-in-progress: ${{ github.ref != 'refs/heads/main' }}

jobs:

render-smoke:

if: github.event_name == 'workflow_dispatch' || (github.event_name == 'pull_request' && contains(github.event.pull_request.labels.*.name, 'ci-full'))

runs-on: ubuntu-24.04

timeout-minutes: 45

defaults:

run:

shell: bash

env:

PYTHONIOENCODING: utf-8

steps:

- uses: actions/checkout@v4

with:

ref: ${{ github.event_name == 'workflow_dispatch' && inputs.candidate_sha || github.event_name == 'pull_request' && github.event.pull_request.head.sha || github.sha }}

- name: Verify checkout SHA

env:

EXPECTED_SHA: ${{ github.event_name == 'workflow_dispatch' && inputs.candidate_sha || github.event_name == 'pull_request' && github.event.pull_request.head.sha || github.sha }}

run: |

[[ "$EXPECTED_SHA" =~ ^[0-9a-f]{40}$ ]]

if [[ "$GITHUB_EVENT_NAME" == workflow_dispatch ]]; then

test "$EXPECTED_SHA" = "$GITHUB_SHA"

fi

test "$(git rev-parse HEAD)" = "$EXPECTED_SHA"

- uses: astral-sh/setup-uv@v5

with:

enable-cache: true

python-version: "3.11.9"

- name: Install Python

run: uv python install "$UV_PYTHON"

# Before setup-soldr, so its dependency cook can find GLib (#262).

- name: Install system dependencies

run: |

sudo apt-get update -qq

sudo apt-get install -y -qq libgtk-3-dev libwebkit2gtk-4.1-dev libayatana-appindicator3-dev librsvg2-dev

# The viewer needs a display, and the runner has no GPU.

sudo apt-get install -y -qq xvfb libgl1-mesa-dri

- uses: zackees/setup-soldr@v0 # v0.9.82 or later (floating major)

env:

GITHUB_TOKEN: ${{ github.token }}

with:

# Same cache namespace as the other ubuntu-24.04 debug builds.

cache-preset: foundation

cache-key-suffix: ubuntu-24.04-native

linker: platform-default

toolchain-file: rust-toolchain.toml

prebuild-deps-flags: ""

- name: Build the fastled binary

run: soldr cargo build --bin fastled

- name: Start Xvfb

run: |

Xvfb :99 -screen 0 1280x1024x24 >"$RUNNER_TEMP/xvfb.log" 2>&1 &

# Wait for the server's socket rather than a probe tool, so this needs

# nothing beyond xvfb itself.

for _ in $(seq 1 30); do

if [ -S /tmp/.X11-unix/X99 ]; then break; fi

sleep 1

done

test -S /tmp/.X11-unix/X99 || { cat "$RUNNER_TEMP/xvfb.log"; exit 1; }

- name: Render a sketch in the shipped viewer

env:

DISPLAY: ":99"

GDK_BACKEND: x11

LIBGL_ALWAYS_SOFTWARE: "1"

WEBKIT_DISABLE_COMPOSITING_MODE: "1"

run: |

mkdir -p "$RUNNER_TEMP/render"

cp -R tests/fixtures/wasm_no_screenmap/. "$RUNNER_TEMP/render/"

./target/debug/fastled "$RUNNER_TEMP/render" \

--test \

--test-wait-secs=2 \

--test-timeout-secs=600 \

--test-ready-timeout-secs=180 \

--test-exit-on-error \

--test-screenshot="$RUNNER_TEMP/render.png" \

--test-log="$RUNNER_TEMP/viewer.log"

- name: Require a real render, not just a viewer that loaded

run: uv run --no-project python ci/check_render.py "$RUNNER_TEMP/render.png"

# The late-screenmap recovery is what makes this sketch renderable at all

# (#250); assert it took the intended path rather than some other one.

- name: Check the late screenmap recovery

run: grep -F "[fastled] late screenmap recovered" "$RUNNER_TEMP/viewer.log"

- name: Upload the render and the viewer log

if: always()

uses: actions/upload-artifact@v4

with:

name: linux-render-smoke

path: |

${{ runner.temp }}/render.png

${{ runner.temp }}/viewer.log

${{ runner.temp }}/xvfb.log

if-no-files-found: warn

retention-days: 7