Files
codebase-memory-mcp/.github/workflows/_memwaste.yml
Martin Vogel 1d58385b51 refactor(mem): route this branch's allocations through the memory core
The memory-core linter is a ratchet: raw malloc/calloc/realloc/free/strdup
sites may not grow, because allocations belong on one accounted path
(src/foundation/mem_core.h) rather than spread across hundreds of call
sites. This branch's own work had added 35 raw sites across nine files --
the parallel k8s prep, the configlink key matching, the worker pool's
counted path, the Go shared stdlib registry, the spill namespace copy,
the route service array and the namespace-name arrays.

All of them now use cbm_alloc/cbm_calloc/cbm_realloc/cbm_free/
cbm_mem_strdup. One was a latent bug rather than a style point:
pass_k8s freed a buffer that k8s_read_file had allocated with raw malloc,
so converting only the free would have mismatched the allocator -- the
read function and every free of its result are converted together.

The Windows free-space path widens its argument through the memory core
instead of cbm_utf8_to_wide(), which allocates with raw malloc.

Two files improved enough to tighten their baselines, which the ratchet
requires so the gain cannot be quietly given back: sqlite_writer.c 92->74
(the index-cell arena), pass_k8s.c 13->6 and registry.c 19->17.

Signed-off-by: Martin Vogel <martin.vogel.tech@gmail.com>
2026-09-18 23:27:23 +02:00

184 lines
8.0 KiB
YAML

# Reusable: the waste sanitizer (memory + CPU waste), REPORT-ONLY.
#
# What runs (scripts/memwaste.sh, memwaste-scaling.py, memwaste-ratchet.py):
# mode=pr Linux: the event lane (memory waste + CPU work) indexing this
# repository at the checked-out commit, and the CPU scaling lane
# (src/ replicated 2x against 4x, per-function counter ratios).
# mode=full the same, plus the event lane on macOS and Windows and the access
# lane (read/write verdicts per allocation) on Linux.
#
# The corpus is the repository itself: pinned by the commit, no network, and a
# real mix of C, C++, Python, JavaScript and vendored code.
#
# REPORT-ONLY: every job uploads its report and prints the ratchet comparison
# against scripts/memwaste-baseline.json, but nothing here can turn a PR red --
# the ratchet runs without --enforce and the PR's ci-ok gate does not depend on
# these jobs. A job fails only when the MEASUREMENT is unsound (a table
# overflowed, frees the layer never saw): that is a defect of the tool, and a
# red job is how it gets noticed. Switching the ratchet to --enforce is a
# separate, deliberate change once the numbers have been stable across reruns.
name: Memwaste
on:
workflow_call:
inputs:
mode:
description: "pr (Linux event + scaling) or full (all platforms, access lane)"
type: string
default: pr
permissions:
contents: read
jobs:
event-linux:
runs-on: ubuntu-latest
timeout-minutes: 60
steps:
- uses: actions/checkout@3d3c42e5aac5ba805825da76410c181273ba90b1 # v7.0.1
- name: Install deps
run: sudo apt-get update && sudo apt-get install -y zlib1g-dev python3 clang llvm
- name: Event lane on this repository
run: |
# `cmd | tee` reports TEE's status: without this a failing lane
# (a missing exec bit, a build error) passes the step and only
# surfaces as the ratchet not finding its JSON.
set -o pipefail
mkdir -p "$RUNNER_TEMP/memwaste-event"
scripts/memwaste.sh "$GITHUB_WORKSPACE" --top 40 --out "$RUNNER_TEMP/memwaste-event" \
| tee "$RUNNER_TEMP/memwaste-event/report.txt"
- name: Ratchet (report-only)
run: python3 scripts/memwaste-ratchet.py --out "$RUNNER_TEMP/memwaste-event" --venue ${{ inputs.mode }}-linux
- name: Upload report
if: ${{ always() }}
uses: actions/upload-artifact@043fb46d1a93c77aae656e7c1c64a875d1fc6a0a # v7.0.1
with:
name: memwaste-event-linux-${{ inputs.mode }}
path: |
${{ runner.temp }}/memwaste-event/report.txt
${{ runner.temp }}/memwaste-event/memwaste.json
${{ runner.temp }}/memwaste-event/index.json
if-no-files-found: warn
scaling-linux:
runs-on: ubuntu-latest
timeout-minutes: 60
steps:
- uses: actions/checkout@3d3c42e5aac5ba805825da76410c181273ba90b1 # v7.0.1
- name: Install deps
run: sudo apt-get update && sudo apt-get install -y zlib1g-dev python3 clang llvm
- name: CPU scaling lane (src/ at 2x against 4x)
# Report-only: the script exits 1 when it flags a super-linear function.
continue-on-error: true
run: |
# `cmd | tee` reports TEE's status: without this a failing lane
# (a missing exec bit, a build error) passes the step and only
# surfaces as the ratchet not finding its JSON.
set -o pipefail
mkdir -p "$RUNNER_TEMP/memwaste-scaling"
python3 scripts/memwaste-scaling.py --corpus src --replicas 2 \
--out "$RUNNER_TEMP/memwaste-scaling" | tee "$RUNNER_TEMP/memwaste-scaling/report.txt"
- name: Upload report
if: ${{ always() }}
uses: actions/upload-artifact@043fb46d1a93c77aae656e7c1c64a875d1fc6a0a # v7.0.1
with:
name: memwaste-scaling-linux-${{ inputs.mode }}
path: ${{ runner.temp }}/memwaste-scaling/report.txt
if-no-files-found: warn
access-linux:
if: ${{ inputs.mode == 'full' }}
runs-on: ubuntu-latest
timeout-minutes: 120
steps:
- uses: actions/checkout@3d3c42e5aac5ba805825da76410c181273ba90b1 # v7.0.1
- name: Install deps
run: sudo apt-get update && sudo apt-get install -y zlib1g-dev python3 clang llvm
- name: Access lane on this repository
run: |
# `cmd | tee` reports TEE's status: without this a failing lane
# (a missing exec bit, a build error) passes the step and only
# surfaces as the ratchet not finding its JSON.
set -o pipefail
mkdir -p "$RUNNER_TEMP/memwaste-access"
scripts/memwaste.sh "$GITHUB_WORKSPACE" --lane access --top 40 --sort untouched \
--out "$RUNNER_TEMP/memwaste-access" | tee "$RUNNER_TEMP/memwaste-access/report.txt"
- name: Upload report
if: ${{ always() }}
uses: actions/upload-artifact@043fb46d1a93c77aae656e7c1c64a875d1fc6a0a # v7.0.1
with:
name: memwaste-access-linux-${{ inputs.mode }}
path: |
${{ runner.temp }}/memwaste-access/report.txt
${{ runner.temp }}/memwaste-access/memwaste.json
if-no-files-found: warn
event-macos:
if: ${{ inputs.mode == 'full' }}
runs-on: macos-14
timeout-minutes: 60
steps:
- uses: actions/checkout@3d3c42e5aac5ba805825da76410c181273ba90b1 # v7.0.1
- name: Event lane on this repository
run: |
# `cmd | tee` reports TEE's status: without this a failing lane
# (a missing exec bit, a build error) passes the step and only
# surfaces as the ratchet not finding its JSON.
set -o pipefail
mkdir -p "$RUNNER_TEMP/memwaste-event"
scripts/memwaste.sh "$GITHUB_WORKSPACE" --top 40 --out "$RUNNER_TEMP/memwaste-event" \
| tee "$RUNNER_TEMP/memwaste-event/report.txt"
- name: Ratchet (report-only)
run: python3 scripts/memwaste-ratchet.py --out "$RUNNER_TEMP/memwaste-event" --venue full-macos
- name: Upload report
if: ${{ always() }}
uses: actions/upload-artifact@043fb46d1a93c77aae656e7c1c64a875d1fc6a0a # v7.0.1
with:
name: memwaste-event-macos-${{ inputs.mode }}
path: |
${{ runner.temp }}/memwaste-event/report.txt
${{ runner.temp }}/memwaste-event/memwaste.json
if-no-files-found: warn
event-windows:
if: ${{ inputs.mode == 'full' }}
runs-on: windows-latest
timeout-minutes: 90
steps:
- uses: actions/checkout@3d3c42e5aac5ba805825da76410c181273ba90b1 # v7.0.1
- uses: msys2/setup-msys2@66cd2cce69caa17b53920067426061ca1de3a884 # v2
with:
msystem: CLANG64
path-type: inherit
install: >-
mingw-w64-clang-x86_64-clang
mingw-w64-clang-x86_64-llvm
mingw-w64-clang-x86_64-zlib
mingw-w64-clang-x86_64-python3
make
git
coreutils
- name: Event lane on this repository
shell: msys2 {0}
run: |
# `cmd | tee` reports TEE's status: without this a failing lane
# (a missing exec bit, a build error) passes the step and only
# surfaces as the ratchet not finding its JSON.
set -o pipefail
out="$(cygpath -u "$RUNNER_TEMP")/memwaste-event"
mkdir -p "$out"
CC=clang CXX=clang++ scripts/memwaste.sh "$(cygpath -u "$GITHUB_WORKSPACE")" --top 40 --out "$out" \
| tee "$out/report.txt"
- name: Ratchet (report-only)
shell: msys2 {0}
run: python3 scripts/memwaste-ratchet.py --out "$(cygpath -u "$RUNNER_TEMP")/memwaste-event" --venue full-windows
- name: Upload report
if: ${{ always() }}
uses: actions/upload-artifact@043fb46d1a93c77aae656e7c1c64a875d1fc6a0a # v7.0.1
with:
name: memwaste-event-windows-${{ inputs.mode }}
path: |
${{ runner.temp }}/memwaste-event/report.txt
${{ runner.temp }}/memwaste-event/memwaste.json
if-no-files-found: warn