No description
Find a file
lei.chen 5b9d727086 Fix two flaky test failures: commit sweep timeout + array stats identity
Controller suite flake (user-reported, -j parallel runs): the commit
engine's async timeout sweep armed continuation deadlines with the
hardcoded DEFAULT_SYNC_TIMEOUT_MS (500ms) instead of the configured
sync_timeout_ms, while the blocking persist_commit_barrier honors the
config. Fixtures configuring longer or shorter sync timeouts saw the
sweep resolve -ETIMEDOUT at 500ms before their explicit
truncation/second-call steps, flipping "Expected -17/-2 Was 0/-110"
assertions (test_commit_resolve.c:1458/:1616). commit_submit_async and
commit_sweep_worker now both use p->config.sync_timeout_ms with
DEFAULT_SYNC_TIMEOUT_MS only as the zeroed-config fallback (mirroring
persist_commit_barrier's timeout_ms == 0 convention), and the sweep
WARN reports the effective value.

Apiserver flake (found during full-build verification):
deriveArrayJobStats took the array identity from the first element
seen over store.Jobs.List(), but List() walks a Go map whose order is
randomized per call -- when element rows disagree (the pinned test
case), the stats identity flipped run to run (~12% of -count=1000
runs). The identity carrier is now the lowest Unique_JobID (the array
head, element 1), tracked via arrayAgg.carrierUID; other rows
contribute counters only. Counters were already order-independent and
are unchanged. The test now pins the head-carries-identity contract
explicitly.

Also fixes a latent parse-error trap in the test: a string literal
ending a line (C-style implicit concatenation) triggers Go's automatic
semicolon insertion -- the next line's argument list parses as
"missing ',' before newline". String concatenation across lines must
keep the '+' at the end of the line.

Verified: controller suite 1342/0 (three runs incl. -j 7 parallel
stress); sweep WARNs now match fixture configs (300ms fixture warns at
300, longer-config fixtures emit none); go test -run
TestDeriveArrayJobStats -count=1000 all green (was 118-130 failures);
full ./tools/xce-build.sh all green end to end.

Co-Authored-By: Claude Code <noreply@anthropic.com>
2026-10-22 02:51:42 +00:00
.claude build: recalibrate asan per-job memory budget 1536MB -> 512MB 2026-10-21 15:38:46 +00:00
.kiro docs(spec): project-config: a built-in default project always exists 2026-10-13 07:17:22 +00:00
docs docs: mark P1 PEND-reason cluster verification done (070e9992) 2026-09-01 15:26:52 +00:00
issues chore(scheduler-allocator-runner): validation passed; new_requirements item 4 resolved (task 5.2) 2026-09-30 10:11:10 +00:00
raft chore: checkpoint all prior work for clean state 2026-06-04 13:04:16 +00:00
scripts build: remove nix dependency layer -- apt system libs + meson wrap subprojects 2026-10-20 18:04:22 +00:00
src Fix two flaky test failures: commit sweep timeout + array stats identity 2026-10-22 02:51:42 +00:00
.clang-format reformat code and job array 2025-11-07 16:14:27 +08:00
.gitignore build: remove nix dependency layer -- apt system libs + meson wrap subprojects 2026-10-20 18:04:22 +00:00
.mcp.json chore(mcp): use XCE_API_KEY env var in xce-scheduler MCP config 2026-08-12 08:41:19 +00:00
.opencode.json checkpoint: bug-851 root cause + raft TLA+ conformance spec family 2026-06-21 14:02:15 +00:00
CLAUDE.md docs(steering): resource decision and question log (PR-1..3 proposals, R/P/N items) 2026-10-12 04:00:47 +00:00
DEPS.md build(packaging): bundle ttyd prebuilt; record license-compliance notes (DEPS.md) 2026-10-21 03:33:10 +00:00
gaps.md build: remove nix dependency layer -- apt system libs + meson wrap subprojects 2026-10-20 18:04:22 +00:00
Makefile build: remove nix dependency layer -- apt system libs + meson wrap subprojects 2026-10-20 18:04:22 +00:00
README.md docs(readme): replace nix build instructions with the de-nix meson flow 2026-10-21 03:44:43 +00:00

1. commit 规则

  • 每个人 fork 自己的 branch, 从自己的 branch 做 PR,review 完再 commit

2. code format

2.1. C

推荐使用 clang-format,公共配置放在源码里面 xce/.clang-format

(理论上直接使用源码里面 xce/.clang-format 文件格式不会变)

2.1.1. vscode 可以在配置中加上如下内容

"C_Cpp.clang_format_style": "WebKit",
"C_Cpp.clang_format_fallbackStyle": "WebKit",

2.1.2. vim & emacs

clang-format -style="{BasedOnStyle: webkit}" -i XX.c

2.2. golang

使用默认 gofmt

3. build environment and build (example of ubuntu)

构建不依赖 nix:直接在 Ubuntu(24.04 基线)上用 apt 系统库 + meson wrap subprojects 构建。完整的依赖清单(工具链 + 11 个 apt 库的版本下限 + 6 个随仓库 提交的 wrap subproject 源码闭包)见 DEPS.md。

3.1 build environment

一次性安装构建工具链与系统库(与 CI 的 scripts/ci-ubuntu.sh 完全一致):

sudo apt-get update
# 工具链
sudo apt-get install -y build-essential pkg-config ninja-build \
    python3 python3-pip python3-venv git
# meson >= 1.5(Ubuntu 24.04 自带 1.3,pip 原地升级)
sudo apt-get install -y meson
meson --version | grep -qE '^1\.[5-9]' || pip3 install --break-system-packages 'meson>=1.5'
# cap'n proto(schema 编译器 + pycapnp 链接的 C++ 运行时)
sudo apt-get install -y capnproto libcapnp-dev
# C 库(DEPS.md "apt 可直接安装" 表)
sudo apt-get install -y libuv1-dev libssl-dev libedit-dev libsqlite3-dev \
    libhwloc-dev libcgroup-dev libtirpc-dev libcyaml-dev libyaml-dev \
    zlib1g-dev libsystemd-dev
# Go(apiserver/xcecli 依赖已全部 vendored,构建离线)
sudo apt-get install -y golang-go

其余依赖(c-capnproto、librtti、struct2json、tinylog、unity、packcc)不在系统里装—— 它们以 meson wrap subproject 的形式提交在 src/subprojects/(源码闭包),meson 配置时自动拉取/构建,无需手动处理。

3.2 checkout source code

git clone ssh://forgejo@gitea.shenrs.eu/rshen/xce.git
cd xce

(参考第 1 节,fork 自己的 branch 再提交 PR)

3.3 build (for daily develop)

日常开发直接用 src/tools/xce-build.sh(若在 Claude Code 中开发,对应 /xce-build skill)。脚本开头有依赖预检(缺工具/缺库时报出具体的 apt 包名,而不是构建中途的 晦涩报错),然后执行 meson setup / meson compile,asan profile 下还会自动跑测试:

cd xce/src

tools/xce-build.sh              # 全部组件 + 全部测试 (asan profile, 输出到 xce.build/)
tools/xce-build.sh controller   # 只构建/测试 controller (lib + controller + xce_test)
tools/xce-build.sh worker       # 只构建 worker
tools/xce-build.sh scheduler    # 只构建/测试 scheduler
tools/xce-build.sh apiserver    # 只构建/测试 apiserver + xcecli (Go)

支持三种 profile(tools/xce-build.sh [component] [profile]):

  • asan(默认):debug + AddressSanitizer,输出目录 xce.build/,会运行测试
  • debug:-O0 -g3 无 sanitizer,输出目录 build.dbg/,仅编译,用于 gdb 调试
  • release:-O3 无 sanitizer,输出目录 build.release/,仅编译,用于 soak/性能测试

注意:单组件构建默认会先 rm -rf 对应的 build 目录再重新配置,若同一目录下已有其他 组件的产物会被清掉;需要在同一目录下累积多个组件时加 --incremental(增量 reconfigure,不清空目录;capnp 生成目标存在未修的依赖边缺失,增量构建若结果可疑 请退回全量重建,见脚本内 bug-1565 注释)。

也可以手动执行底层的 meson 命令(xce-build.sh 是对下面步骤的封装):

cd xce/src

# generate build configuration (debug + address sanitizer)
meson setup xce.build --buildtype=debug -Db_sanitize=address

# build
meson compile -C xce.build

# The generated files will be in xce.build/

3.3.1 更新 Go 依赖(vendor 维护)

Go 模块(frontend/apiserver、frontend/xcecli、 autoscaler/providers/baidu/src/bcctool)的依赖以 vendor/ 目录的形式提交进仓库, 构建时 tools/go_build 使用 GOFLAGS=-mod=vendor、GOPROXY=off 完全离线编译。

修改了某个模块的 go.mod / go.sum(新增、升级、删除依赖)后,重新生成对应模块 的 vendor/ 并提交(普通环境即可,无需任何特殊 shell):

cd xce/src/frontend/apiserver   # 其他两个模块同理
go mod tidy
go mod vendor
git add go.mod go.sum vendor

3.4 package (tarball 分发)

产品分发形态是自包含 tarball(daemon + CLI + $ORIGIN RPATH 的 bundled 非基础库

  • systemd unit 样例 + configs/schema + ttyd),打包规则见 src/tools/xce-package.sh 头部注释与 DEPS.md "运行期 bundle 规则":
cd xce/src

tools/xce-build.sh all release        # 先出 release 构建(build.release/)
tools/xce-package.sh                  # 默认从 build.release/ 打到 dist/
# 或显式指定: tools/xce-package.sh build.release dist

产物:dist/xce-<version>-<commit>-<fingerprint>-linux-x86_64.tar.gz。目标机上:

tar -xzf xce-*.tar.gz && cd xce-*/
sudo ./install.sh                      # 默认装到 /opt/xce,PREFIX=... 可覆盖

3.5 CI

scripts/ci-ubuntu.sh 在裸 Ubuntu 24.04 容器里跑完整流水线(装依赖 → asan 全量 构建+测试 → release 构建 → 打 tarball smoke),与客户环境同构:

bash scripts/ci-ubuntu.sh

4. misc

4.1. Windows remap ctrl and alt key

create a reg file, e.g. remapkey.reg

Windows Registry Editor Version 5.00
[HKEY_LOCAL_MACHINE\SYSTEM\CurrentControlSet\Control\Keyboard Layout]
"Scancode Map"=hex:00,00,00,00,00,00,00,00,05,00,00,00,38,00,1d,00,1d,00,38,00,38,e0,1d,e0,1d,e0,38,e0,00,00,00,00

then double click remapkey.reg to import it. Restart OS.