Python: Eliminate some FPs

Two main variants: cases where `__eq__` is not defined (which inherit
`__hash__` automatically), and frozen dataclasses.

.github/dependabot.yml

@@ -45,3 +45,5 @@ updates:
     directory: "/"
     schedule:
       interval: weekly
+    exclude-paths:
+      - "misc/bazel/registry/**"

.github/workflows/compile-queries.yml

@@ -1,78 +0,0 @@
-name: "Compile all queries using the latest stable CodeQL CLI"
-
-on:
-  push:
-    branches:  # makes sure the cache gets populated - running on the branches people tend to merge into.
-      - main
-      - "rc/*"
-      - "codeql-cli-*"
-  pull_request:
-    paths:
-      - '**.ql'
-      - '**.qll'
-      - '**/qlpack.yml'
-      - '**.dbscheme'
-
-permissions:
-  contents: read
-
-jobs:
-  detect-changes:
-    if: github.repository_owner == 'github'
-    runs-on: ubuntu-latest
-    outputs:
-      languages: ${{ steps.detect.outputs.languages }}
-    steps:
-      - uses: actions/checkout@v5
-      - name: Detect changed languages
-        id: detect
-        run: |
-          if [[ "${{ github.event_name }}" == "pull_request" ]]; then
-            # For PRs, detect which languages have changes
-            changed_files=$(gh pr view ${{ github.event.pull_request.number }} --json files --jq '.files.[].path')
-            languages=()
-            for lang in actions cpp csharp go java javascript python ql ruby rust swift; do
-              if echo "$changed_files" | grep -qE "^($lang/|shared/)" ; then
-                languages+=("$lang")
-              fi
-            done
-            echo "languages=$(jq -c -n '$ARGS.positional' --args "${languages[@]}")" >> $GITHUB_OUTPUT
-          else
-            # For pushes to main/rc branches, run all languages
-            echo 'languages=["actions","cpp","csharp","go","java","javascript","python","ql","ruby","rust","swift"]' >> $GITHUB_OUTPUT
-          fi
-        env:
-          GH_TOKEN: ${{ github.token }}
-
-  compile-queries:
-    needs: detect-changes
-    if: github.repository_owner == 'github' && needs.detect-changes.outputs.languages != '[]'
-    runs-on: ubuntu-latest-xl
-    strategy:
-      fail-fast: false
-      matrix:
-        language: ${{ fromJson(needs.detect-changes.outputs.languages) }}
-
-    steps:
-      - uses: actions/checkout@v5
-      - name: Setup CodeQL
-        uses: ./.github/actions/fetch-codeql
-        with:
-          channel: 'release'
-      - name: Cache compilation cache
-        id: query-cache
-        uses: ./.github/actions/cache-query-compilation
-        with: 
-          key: ${{ matrix.language }}-queries
-      - name: check formatting
-        run: find shared ${{ matrix.language }}/ql -type f \( -name "*.qll" -o -name "*.ql" \) -print0 | xargs -0 -n 3000 -P 10 codeql query format -q --check-only
-      - name: compile queries - check-only
-        # run with --check-only if running in a PR (github.sha != main)
-        if : ${{ github.event_name == 'pull_request' }}
-        shell: bash
-        run: codeql query compile -q -j0 ${{ matrix.language }}/ql/{src,examples} --keep-going --warnings=error --check-only --compilation-cache "${{ steps.query-cache.outputs.cache-dir }}" --compilation-cache-size=500 --ram=56000
-      - name: compile queries - full
-        # do full compile if running on main - this populates the cache
-        if : ${{ github.event_name != 'pull_request' }}
-        shell: bash
-        run: codeql query compile -q -j0 ${{ matrix.language }}/ql/{src,examples} --keep-going --warnings=error --compilation-cache "${{ steps.query-cache.outputs.cache-dir }}" --compilation-cache-size=500 --ram=56000

.github/workflows/ruby-build.yml

@@ -1,236 +0,0 @@
-name: "Ruby: Build"
-
-on:
-  push:
-    paths:
-      - "ruby/**"
-      - .github/workflows/ruby-build.yml
-      - .github/actions/fetch-codeql/action.yml
-      - codeql-workspace.yml
-      - "shared/tree-sitter-extractor/**"
-    branches:
-      - main
-      - "rc/*"
-  pull_request:
-    paths:
-      - "ruby/**"
-      - .github/workflows/ruby-build.yml
-      - .github/actions/fetch-codeql/action.yml
-      - codeql-workspace.yml
-      - "shared/tree-sitter-extractor/**"
-    branches:
-      - main
-      - "rc/*"
-  workflow_dispatch:
-    inputs:
-      tag:
-        description: "Version tag to create"
-        required: false
-
-env:
-  CARGO_TERM_COLOR: always
-
-defaults:
-  run:
-    working-directory: ruby
-
-permissions:
-  contents: read
-
-jobs:
-  build:
-    strategy:
-      fail-fast: false
-      matrix:
-        os: [ubuntu-latest, macos-latest, windows-latest]
-
-    runs-on: ${{ matrix.os }}
-
-    steps:
-      - uses: actions/checkout@v5
-      - name: Install GNU tar
-        if: runner.os == 'macOS'
-        run: |
-          brew install gnu-tar
-          echo "/usr/local/opt/gnu-tar/libexec/gnubin" >> $GITHUB_PATH
-      - name: Prepare Windows
-        if: runner.os == 'Windows'
-        shell: powershell
-        run: |
-          git config --global core.longpaths true
-      - uses: ./.github/actions/os-version
-        id: os_version
-      - name: Cache entire extractor
-        uses: actions/cache@v3
-        id: cache-extractor
-        with:
-          path: |
-            target/release/codeql-extractor-ruby
-            target/release/codeql-extractor-ruby.exe
-            ruby/extractor/ql/lib/codeql/ruby/ast/internal/TreeSitter.qll
-          key: ${{ runner.os }}-${{ steps.os_version.outputs.version }}-ruby-extractor-${{ hashFiles('ruby/extractor/rust-toolchain.toml', 'ruby/extractor/Cargo.lock') }}-${{ hashFiles('shared/tree-sitter-extractor') }}-${{ hashFiles('ruby/extractor/**/*.rs') }}
-      - uses: actions/cache@v3
-        if: steps.cache-extractor.outputs.cache-hit != 'true'
-        with:
-          path: |
-            ~/.cargo/registry
-            ~/.cargo/git
-            target
-          key: ${{ runner.os }}-${{ steps.os_version.outputs.version }}-ruby-rust-cargo-${{ hashFiles('ruby/extractor/rust-toolchain.toml', 'ruby/extractor/**/Cargo.lock') }}
-      - name: Check formatting
-        if: steps.cache-extractor.outputs.cache-hit != 'true'
-        run: cd extractor && cargo fmt -- --check
-      - name: Build
-        if: steps.cache-extractor.outputs.cache-hit != 'true'
-        run: cd extractor && cargo build --verbose
-      - name: Run tests
-        if: steps.cache-extractor.outputs.cache-hit != 'true'
-        run: cd extractor && cargo test --verbose
-      - name: Release build
-        if: steps.cache-extractor.outputs.cache-hit != 'true'
-        run: cd extractor && cargo build --release
-      - name: Generate dbscheme
-        if: ${{ matrix.os == 'ubuntu-latest' && steps.cache-extractor.outputs.cache-hit != 'true'}}
-        run: ../target/release/codeql-extractor-ruby generate --dbscheme ql/lib/ruby.dbscheme --library ql/lib/codeql/ruby/ast/internal/TreeSitter.qll
-      - uses: actions/upload-artifact@v4
-        if: ${{ matrix.os == 'ubuntu-latest' }}
-        with:
-          name: ruby.dbscheme
-          path: ruby/ql/lib/ruby.dbscheme
-      - uses: actions/upload-artifact@v4
-        if: ${{ matrix.os == 'ubuntu-latest' }}
-        with:
-          name: TreeSitter.qll
-          path: ruby/ql/lib/codeql/ruby/ast/internal/TreeSitter.qll
-      - uses: actions/upload-artifact@v4
-        with:
-          name: extractor-${{ matrix.os }}
-          path: |
-            target/release/codeql-extractor-ruby
-            target/release/codeql-extractor-ruby.exe
-          retention-days: 1
-  compile-queries:
-    if: github.repository_owner == 'github'
-    runs-on: ubuntu-latest-xl
-    steps:
-      - uses: actions/checkout@v5
-      - name: Fetch CodeQL
-        uses: ./.github/actions/fetch-codeql
-      - name: Cache compilation cache
-        id: query-cache
-        uses: ./.github/actions/cache-query-compilation
-        with:
-          key: ruby-build
-      - name: Build Query Pack
-        run: |
-          PACKS=${{ runner.temp }}/query-packs
-          rm -rf $PACKS
-          codeql pack create ../misc/suite-helpers --output "$PACKS"
-          codeql pack create ../shared/regex --output "$PACKS"
-          codeql pack create ../shared/ssa --output "$PACKS"
-          codeql pack create ../shared/tutorial --output "$PACKS"
-          codeql pack create ql/lib --output "$PACKS"
-          codeql pack create -j0 ql/src --output "$PACKS" --compilation-cache "${{ steps.query-cache.outputs.cache-dir }}"
-          PACK_FOLDER=$(readlink -f "$PACKS"/codeql/ruby-queries/*)
-          codeql generate query-help --format=sarifv2.1.0 --output="${PACK_FOLDER}/rules.sarif" ql/src
-          (cd ql/src; find queries \( -name '*.qhelp' -o -name '*.rb' -o -name '*.erb' \) -exec bash -c 'mkdir -p "'"${PACK_FOLDER}"'/$(dirname "{}")"' \; -exec cp "{}" "${PACK_FOLDER}/{}" \;)
-      - uses: actions/upload-artifact@v4
-        with:
-          name: codeql-ruby-queries
-          path: |
-            ${{ runner.temp }}/query-packs/*
-          retention-days: 1
-          include-hidden-files: true
-
-  package:
-    runs-on: ubuntu-latest
-    needs: [build, compile-queries]
-    steps:
-      - uses: actions/checkout@v5
-      - uses: actions/download-artifact@v4
-        with:
-          name: ruby.dbscheme
-          path: ruby/ruby
-      - uses: actions/download-artifact@v4
-        with:
-          name: extractor-ubuntu-latest
-          path: ruby/linux64
-      - uses: actions/download-artifact@v4
-        with:
-          name: extractor-windows-latest
-          path: ruby/win64
-      - uses: actions/download-artifact@v4
-        with:
-          name: extractor-macos-latest
-          path: ruby/osx64
-      - run: |
-          mkdir -p ruby
-          cp -r codeql-extractor.yml tools ql/lib/ruby.dbscheme.stats ruby/
-          mkdir -p ruby/tools/{linux64,osx64,win64}
-          cp linux64/codeql-extractor-ruby ruby/tools/linux64/extractor
-          cp osx64/codeql-extractor-ruby ruby/tools/osx64/extractor
-          cp win64/codeql-extractor-ruby.exe ruby/tools/win64/extractor.exe
-          chmod +x ruby/tools/{linux64,osx64}/extractor
-          zip -rq codeql-ruby.zip ruby
-      - uses: actions/upload-artifact@v4
-        with:
-          name: codeql-ruby-pack
-          path: ruby/codeql-ruby.zip
-          retention-days: 1
-          include-hidden-files: true
-      - uses: actions/download-artifact@v4
-        with:
-          name: codeql-ruby-queries
-          path: ruby/qlpacks
-      - run: |
-          echo '{
-            "provide": [
-            "ruby/codeql-extractor.yml",
-            "qlpacks/*/*/*/qlpack.yml"
-            ]
-          }' > .codeqlmanifest.json
-          zip -rq codeql-ruby-bundle.zip .codeqlmanifest.json ruby qlpacks
-      - uses: actions/upload-artifact@v4
-        with:
-          name: codeql-ruby-bundle
-          path: ruby/codeql-ruby-bundle.zip
-          retention-days: 1
-          include-hidden-files: true
-
-  test:
-    defaults:
-      run:
-        working-directory: ${{ github.workspace }}
-    strategy:
-      fail-fast: false
-      matrix:
-        os: [ubuntu-latest, macos-latest, windows-latest]
-
-    runs-on: ${{ matrix.os }}
-    needs: [package]
-    steps:
-      - uses: actions/checkout@v5
-      - name: Fetch CodeQL
-        uses: ./.github/actions/fetch-codeql
-
-      - name: Download Ruby bundle
-        uses: actions/download-artifact@v4
-        with:
-          name: codeql-ruby-bundle
-          path: ${{ runner.temp }}
-      - name: Unzip Ruby bundle
-        shell: bash
-        run: unzip -q -d "${{ runner.temp }}/ruby-bundle" "${{ runner.temp }}/codeql-ruby-bundle.zip"
-
-      - name: Run QL test
-        shell: bash
-        run: |
-          codeql test run --search-path "${{ runner.temp }}/ruby-bundle" --additional-packs "${{ runner.temp }}/ruby-bundle" ruby/ql/test/library-tests/ast/constants/
-      - name: Create database
-        shell: bash
-        run: |
-          codeql database create --search-path "${{ runner.temp }}/ruby-bundle" --language ruby --source-root ruby/ql/test/library-tests/ast/constants/ ../database
-      - name: Analyze database
-        shell: bash
-        run: |
-          codeql database analyze --search-path "${{ runner.temp }}/ruby-bundle" --format=sarifv2.1.0 --output=out.sarif ../database ruby-code-scanning.qls

.github/workflows/ruby-dataset-measure.yml

@@ -1,75 +0,0 @@
-name: "Ruby: Collect database stats"
-
-on:
-  push:
-    branches:
-      - main
-      - "rc/*"
-    paths:
-      - ruby/ql/lib/ruby.dbscheme
-      - .github/workflows/ruby-dataset-measure.yml
-  pull_request:
-    branches:
-      - main
-      - "rc/*"
-    paths:
-      - ruby/ql/lib/ruby.dbscheme
-      - .github/workflows/ruby-dataset-measure.yml
-  workflow_dispatch:
-
-permissions:
-  contents: read
-
-jobs:
-  measure:
-    env:
-      CODEQL_THREADS: 4 # TODO: remove this once it's set by the CLI
-    strategy:
-      fail-fast: false
-      matrix:
-        repo: [rails/rails, discourse/discourse, spree/spree, ruby/ruby]
-    runs-on: ubuntu-latest
-    steps:
-      - uses: actions/checkout@v5
-
-      - uses: ./.github/actions/fetch-codeql
-
-      - uses: ./ruby/actions/create-extractor-pack
-
-      - name: Checkout ${{ matrix.repo }}
-        uses: actions/checkout@v5
-        with:
-          repository: ${{ matrix.repo }}
-          path: ${{ github.workspace }}/repo
-      - name: Create database
-        run: |
-          codeql database create \
-            --search-path "${{ github.workspace }}" \
-            --threads 4 \
-            --language ruby --source-root "${{ github.workspace }}/repo" \
-            "${{ runner.temp }}/database"
-      - name: Measure database
-        run: |
-          mkdir -p "stats/${{ matrix.repo }}"
-          codeql dataset measure --threads 4 --output "stats/${{ matrix.repo }}/stats.xml" "${{ runner.temp }}/database/db-ruby"
-      - uses: actions/upload-artifact@v4
-        with:
-          name: measurements-${{ hashFiles('stats/**') }}
-          path: stats
-          retention-days: 1
-
-  merge:
-    runs-on: ubuntu-latest
-    needs: measure
-    steps:
-      - uses: actions/checkout@v5
-      - uses: actions/download-artifact@v4
-        with:
-          path: stats
-      - run: |
-          python -m pip install --user lxml
-          find stats -name 'stats.xml' | sort | xargs python ruby/scripts/merge_stats.py --output ruby/ql/lib/ruby.dbscheme.stats --normalise ruby_tokeninfo
-      - uses: actions/upload-artifact@v4
-        with:
-          name: ruby.dbscheme.stats
-          path: ruby/ql/lib/ruby.dbscheme.stats

.github/workflows/ruby-qltest-rtjo.yml

@@ -1,40 +0,0 @@
-name: "Ruby: Run RTJO Language Tests"
-
-on:
-  pull_request:
-    types:
-      - opened
-      - synchronize
-      - reopened
-      - labeled
-
-env:
-  CARGO_TERM_COLOR: always
-
-defaults:
-  run:
-    working-directory: ruby
-
-permissions:
-  contents: read
-
-jobs:
-  qltest-rtjo:
-    if: "github.repository_owner == 'github' && github.event.label.name == 'Run: RTJO Language Tests'"
-    runs-on: ubuntu-latest-xl
-    strategy:
-      fail-fast: false
-    steps:
-      - uses: actions/checkout@v5
-      - uses: ./.github/actions/fetch-codeql
-      - uses: ./ruby/actions/create-extractor-pack
-      - name: Cache compilation cache
-        id: query-cache
-        uses: ./.github/actions/cache-query-compilation
-        with:
-          key: ruby-qltest
-      - name: Run QL tests
-        run: |
-          codeql test run --dynamic-join-order-mode=all --threads=0 --ram 50000 --search-path "${{ github.workspace }}" --check-databases --check-diff-informed --check-undefined-labels --check-unused-labels --check-repeated-labels --check-redefined-labels --check-use-before-definition --consistency-queries ql/consistency-queries ql/test --compilation-cache "${{ steps.query-cache.outputs.cache-dir }}"
-        env:
-          GITHUB_TOKEN: ${{ github.token }}

.github/workflows/ruby-qltest.yml

@@ -1,73 +0,0 @@
-name: "Ruby: Run QL Tests"
-
-on:
-  push:
-    paths:
-      - "ruby/**"
-      - "shared/**"
-      - .github/workflows/ruby-build.yml
-      - .github/actions/fetch-codeql/action.yml
-      - codeql-workspace.yml
-    branches:
-      - main
-      - "rc/*"
-  pull_request:
-    paths:
-      - "ruby/**"
-      - "shared/**"
-      - .github/workflows/ruby-qltest.yml
-      - .github/actions/fetch-codeql/action.yml
-      - codeql-workspace.yml
-    branches:
-      - main
-      - "rc/*"
-
-env:
-  CARGO_TERM_COLOR: always
-
-defaults:
-  run:
-    working-directory: ruby
-
-permissions:
-  contents: read
-
-jobs:
-  qlupgrade:
-    runs-on: ubuntu-latest
-    steps:
-      - uses: actions/checkout@v5
-      - uses: ./.github/actions/fetch-codeql
-      - name: Check DB upgrade scripts
-        run: |
-          echo >empty.trap
-          codeql dataset import -S ql/lib/upgrades/initial/ruby.dbscheme testdb empty.trap
-          codeql dataset upgrade testdb --additional-packs ql/lib
-          diff -q testdb/ruby.dbscheme ql/lib/ruby.dbscheme
-      - name: Check DB downgrade scripts
-        run: |
-          echo >empty.trap
-          rm -rf testdb; codeql dataset import -S ql/lib/ruby.dbscheme testdb empty.trap
-          codeql resolve upgrades --format=lines --allow-downgrades --additional-packs downgrades \
-           --dbscheme=ql/lib/ruby.dbscheme --target-dbscheme=downgrades/initial/ruby.dbscheme |
-           xargs codeql execute upgrades testdb
-          diff -q testdb/ruby.dbscheme downgrades/initial/ruby.dbscheme
-  qltest:
-    if: github.repository_owner == 'github'
-    runs-on: ubuntu-latest-xl
-    strategy:
-      fail-fast: false
-    steps:
-      - uses: actions/checkout@v5
-      - uses: ./.github/actions/fetch-codeql
-      - uses: ./ruby/actions/create-extractor-pack
-      - name: Cache compilation cache
-        id: query-cache
-        uses: ./.github/actions/cache-query-compilation
-        with:
-          key: ruby-qltest
-      - name: Run QL tests
-        run: |
-          codeql test run --threads=0 --ram 50000 --search-path "${{ github.workspace }}" --check-databases --check-diff-informed --check-undefined-labels --check-unused-labels --check-repeated-labels --check-redefined-labels --check-use-before-definition --consistency-queries ql/consistency-queries ql/test --compilation-cache "${{ steps.query-cache.outputs.cache-dir }}"
-        env:
-          GITHUB_TOKEN: ${{ github.token }}

.pre-commit-config.yaml

@@ -7,9 +7,9 @@ repos:
     rev: v3.2.0
     hooks:
       - id: trailing-whitespace
-        exclude: /test/.*$(?<!\.qlref)|.*\.patch$|.*\.qll?$
+        exclude: /test([^/]*)/.*$(?<!\.qlref)|.*\.patch$|.*\.qll?$
       - id: end-of-file-fixer
-        exclude: Cargo.lock$|/test/.*$(?<!\.qlref)|.*\.patch$|.*\.qll?$
+        exclude: Cargo.lock$|/test([^/]*)/.*$(?<!\.qlref)|.*\.patch$|.*\.qll?$
 
   - repo: https://github.com/pre-commit/mirrors-clang-format
     rev: v17.0.6

MODULE.bazel

@@ -15,23 +15,23 @@ local_path_override(
 # see https://registry.bazel.build/ for a list of available packages
 
 bazel_dep(name = "platforms", version = "1.0.0")
-bazel_dep(name = "rules_cc", version = "0.2.16")
-bazel_dep(name = "rules_go", version = "0.59.0")
-bazel_dep(name = "rules_java", version = "9.0.3")
-bazel_dep(name = "rules_pkg", version = "1.0.1")
+bazel_dep(name = "rules_cc", version = "0.2.17")
+bazel_dep(name = "rules_go", version = "0.60.0")
+bazel_dep(name = "rules_java", version = "9.6.1")
+bazel_dep(name = "rules_pkg", version = "1.2.0")
 bazel_dep(name = "rules_nodejs", version = "6.7.3")
-bazel_dep(name = "rules_python", version = "0.40.0")
-bazel_dep(name = "rules_shell", version = "0.5.0")
-bazel_dep(name = "bazel_skylib", version = "1.8.1")
-bazel_dep(name = "abseil-cpp", version = "20240116.1", repo_name = "absl")
+bazel_dep(name = "rules_python", version = "1.9.0")
+bazel_dep(name = "rules_shell", version = "0.7.1")
+bazel_dep(name = "bazel_skylib", version = "1.9.0")
+bazel_dep(name = "abseil-cpp", version = "20260107.1", repo_name = "absl")
 bazel_dep(name = "nlohmann_json", version = "3.11.3", repo_name = "json")
 bazel_dep(name = "fmt", version = "12.1.0-codeql.1")
 bazel_dep(name = "rules_kotlin", version = "2.2.2-codeql.1")
 bazel_dep(name = "gazelle", version = "0.47.0")
 bazel_dep(name = "rules_dotnet", version = "0.21.5-codeql.1")
-bazel_dep(name = "googletest", version = "1.14.0.bcr.1")
-bazel_dep(name = "rules_rust", version = "0.68.1.codeql.1")
-bazel_dep(name = "zstd", version = "1.5.5.bcr.1")
+bazel_dep(name = "googletest", version = "1.17.0.bcr.2")
+bazel_dep(name = "rules_rust", version = "0.69.0")
+bazel_dep(name = "zstd", version = "1.5.7.bcr.1")
 
 bazel_dep(name = "buildifier_prebuilt", version = "6.4.0", dev_dependency = True)
 
@@ -242,6 +242,7 @@ use_repo(
     "kotlin-compiler-2.2.0-Beta1",
     "kotlin-compiler-2.2.20-Beta2",
     "kotlin-compiler-2.3.0",
+    "kotlin-compiler-2.3.20",
     "kotlin-compiler-embeddable-1.8.0",
     "kotlin-compiler-embeddable-1.9.0-Beta",
     "kotlin-compiler-embeddable-1.9.20-Beta",
@@ -252,6 +253,7 @@ use_repo(
     "kotlin-compiler-embeddable-2.2.0-Beta1",
     "kotlin-compiler-embeddable-2.2.20-Beta2",
     "kotlin-compiler-embeddable-2.3.0",
+    "kotlin-compiler-embeddable-2.3.20",
     "kotlin-stdlib-1.8.0",
     "kotlin-stdlib-1.9.0-Beta",
     "kotlin-stdlib-1.9.20-Beta",
@@ -262,6 +264,7 @@ use_repo(
     "kotlin-stdlib-2.2.0-Beta1",
     "kotlin-stdlib-2.2.20-Beta2",
     "kotlin-stdlib-2.3.0",
+    "kotlin-stdlib-2.3.20",
 )
 
 go_sdk = use_extension("@rules_go//go:extensions.bzl", "go_sdk")

actions/ql/lib/CHANGELOG.md

@@ -1,3 +1,15 @@
+## 0.4.32
+
+No user-facing changes.
+
+## 0.4.31
+
+No user-facing changes.
+
+## 0.4.30
+
+No user-facing changes.
+
 ## 0.4.29
 
 No user-facing changes.

actions/ql/lib/change-notes/released/0.4.30.md

@@ -0,0 +1,3 @@
+## 0.4.30
+
+No user-facing changes.

actions/ql/lib/change-notes/released/0.4.31.md

@@ -0,0 +1,3 @@
+## 0.4.31
+
+No user-facing changes.

actions/ql/lib/change-notes/released/0.4.32.md

@@ -0,0 +1,3 @@
+## 0.4.32
+
+No user-facing changes.

actions/ql/lib/codeql-pack.release.yml

@@ -1,2 +1,2 @@
 ---
-lastReleaseVersion: 0.4.29
+lastReleaseVersion: 0.4.32

actions/ql/lib/qlpack.yml

@@ -1,5 +1,5 @@
 name: codeql/actions-all
-version: 0.4.30-dev
+version: 0.4.33-dev
 library: true
 warnOnImplicitThis: true
 dependencies:

actions/ql/src/CHANGELOG.md

@@ -1,3 +1,15 @@
+## 0.6.24
+
+No user-facing changes.
+
+## 0.6.23
+
+No user-facing changes.
+
+## 0.6.22
+
+No user-facing changes.
+
 ## 0.6.21
 
 No user-facing changes.

actions/ql/src/change-notes/released/0.6.22.md

@@ -0,0 +1,3 @@
+## 0.6.22
+
+No user-facing changes.

actions/ql/src/change-notes/released/0.6.23.md

@@ -0,0 +1,3 @@
+## 0.6.23
+
+No user-facing changes.

actions/ql/src/change-notes/released/0.6.24.md

@@ -0,0 +1,3 @@
+## 0.6.24
+
+No user-facing changes.

actions/ql/src/codeql-pack.release.yml

@@ -1,2 +1,2 @@
 ---
-lastReleaseVersion: 0.6.21
+lastReleaseVersion: 0.6.24

actions/ql/src/qlpack.yml

@@ -1,5 +1,5 @@
 name: codeql/actions-queries
-version: 0.6.22-dev
+version: 0.6.25-dev
 library: false
 warnOnImplicitThis: true
 groups: [actions, queries]

config/add-overlay-annotations.py

@@ -199,6 +199,7 @@ def annotate_as_appropriate(filename, lines):
     # as overlay[local?].  It is not clear that these heuristics are exactly what we want,
     # but they seem to work well enough for now (as determined by speed and accuracy numbers).
     if (filename.endswith("Test.qll") or
+        re.search(r"go/ql/lib/semmle/go/security/[^/]+[.]qll$", filename.replace(os.sep, "/")) or
         ((filename.endswith("Query.qll") or filename.endswith("Config.qll")) and
          any("implements DataFlow::ConfigSig" in line for line in lines))):
         return None

config/identical-files.json

@@ -172,10 +172,6 @@
     "cpp/ql/lib/semmle/code/cpp/ir/implementation/raw/internal/reachability/PrintDominance.qll",
     "cpp/ql/lib/semmle/code/cpp/ir/implementation/unaliased_ssa/internal/reachability/PrintDominance.qll"
   ],
-  "C# ControlFlowReachability": [
-    "csharp/ql/lib/semmle/code/csharp/dataflow/internal/ControlFlowReachability.qll",
-    "csharp/ql/lib/semmle/code/csharp/dataflow/internal/rangeanalysis/ControlFlowReachability.qll"
-  ],
   "C++ ExternalAPIs": [
     "cpp/ql/src/Security/CWE/CWE-020/ExternalAPIs.qll",
     "cpp/ql/src/Security/CWE/CWE-020/ir/ExternalAPIs.qll"

cpp/downgrades/770002bb02322e04fa25345838ce6e82af285a0b/in_trap.ql

@@ -0,0 +1,21 @@
+class Element extends @element {
+  string toString() { none() }
+}
+
+class Trap extends @trap {
+  string toString() { none() }
+}
+
+class Tag extends @tag {
+  string toString() { none() }
+}
+
+from Element e, Trap trap
+where
+  in_trap_or_tag(e, trap)
+  or
+  exists(Tag tag |
+    in_trap_or_tag(e, tag) and
+    trap_uses_tag(trap, tag)
+  )
+select e, trap

cpp/downgrades/770002bb02322e04fa25345838ce6e82af285a0b/source_file_uses_trap.ql

@@ -0,0 +1,13 @@
+class SourceFile extends @source_file {
+  string toString() { none() }
+}
+
+class Trap extends @trap {
+  string toString() { none() }
+}
+
+from SourceFile source_file, string name, Trap trap
+where
+  source_file_uses_trap(source_file, trap) and
+  source_file_name(source_file, name)
+select name, trap

cpp/downgrades/770002bb02322e04fa25345838ce6e82af285a0b/upgrade.properties

@@ -0,0 +1,8 @@
+description: Add source_file_name
+compatibility: backwards
+source_file_uses_trap.rel: run source_file_uses_trap.ql
+source_file_name.rel: delete
+tag_name.rel: delete
+trap_uses_tag.rel: delete
+in_trap.rel: run in_trap.ql
+in_trap_or_tag.rel: delete

cpp/ql/integration-tests/query-suite/cpp-code-scanning.qls.expected

@@ -52,5 +52,6 @@ ql/cpp/ql/src/Summary/LinesOfUserCode.ql
 ql/cpp/ql/src/Telemetry/CompilerErrors.ql
 ql/cpp/ql/src/Telemetry/DatabaseQuality.ql
 ql/cpp/ql/src/Telemetry/ExtractionMetrics.ql
+ql/cpp/ql/src/Telemetry/ExtractorInformation.ql
 ql/cpp/ql/src/Telemetry/MissingIncludes.ql
 ql/cpp/ql/src/Telemetry/SucceededIncludes.ql

cpp/ql/integration-tests/query-suite/cpp-security-and-quality.qls.expected

@@ -160,6 +160,7 @@ ql/cpp/ql/src/Summary/LinesOfUserCode.ql
 ql/cpp/ql/src/Telemetry/CompilerErrors.ql
 ql/cpp/ql/src/Telemetry/DatabaseQuality.ql
 ql/cpp/ql/src/Telemetry/ExtractionMetrics.ql
+ql/cpp/ql/src/Telemetry/ExtractorInformation.ql
 ql/cpp/ql/src/Telemetry/MissingIncludes.ql
 ql/cpp/ql/src/Telemetry/SucceededIncludes.ql
 ql/cpp/ql/src/jsf/4.06 Pre-Processing Directives/AV Rule 32.ql

cpp/ql/integration-tests/query-suite/cpp-security-extended.qls.expected

@@ -93,5 +93,6 @@ ql/cpp/ql/src/Summary/LinesOfUserCode.ql
 ql/cpp/ql/src/Telemetry/CompilerErrors.ql
 ql/cpp/ql/src/Telemetry/DatabaseQuality.ql
 ql/cpp/ql/src/Telemetry/ExtractionMetrics.ql
+ql/cpp/ql/src/Telemetry/ExtractorInformation.ql
 ql/cpp/ql/src/Telemetry/MissingIncludes.ql
 ql/cpp/ql/src/Telemetry/SucceededIncludes.ql

cpp/ql/lib/CHANGELOG.md

@@ -1,3 +1,17 @@
+## 8.0.3
+
+No user-facing changes.
+
+## 8.0.2
+
+No user-facing changes.
+
+## 8.0.1
+
+### Minor Analysis Improvements
+
+* Inline expectations test comments, which are of the form `// $ tag` or `// $ tag=value`, are now parsed more strictly and will not be recognized if there isn't a space after the `$` symbol.
+
 ## 8.0.0
 
 ### Breaking Changes

cpp/ql/lib/change-notes/2026-03-20-add-indirect-uninitialized-node.md

@@ -0,0 +1,4 @@
+---
+category: feature
+---
+* Added a class `IndirectUninitializedNode` to represent the indirection of an uninitialized local variable as a dataflow node.

cpp/ql/lib/change-notes/2026-03-23-indirect-parameter-nodes-and-indirect-instructions.md

@@ -0,0 +1,5 @@
+---
+category: feature
+---
+* Added a class `DataFlow::IndirectParameterNode` to represent the indirection of a parameter as a dataflow node.
+* Added a predicate `Node::asIndirectInstruction` which returns the `Instruction` that defines the indirect dataflow node, if any.

cpp/ql/lib/change-notes/2026-03-24-field-init.md

@@ -0,0 +1,5 @@
+---
+category: feature
+---
+* Added a class `ConstructorDirectFieldInit` to represent field initializations that occur in member initializer lists.
+* Added a class `ConstructorDefaultFieldInit` to represent default field initializations.

cpp/ql/lib/change-notes/2026-03-26-convert-csv-models-to-yml.md

@@ -0,0 +1,4 @@
+---
+category: breaking
+---
+* The `SourceModelCsv`, `SinkModelCsv`, and `SummaryModelCsv` classes and the associated CSV parsing infrastructure have been removed from `ExternalFlow.qll`. New models should be added as `.model.yml` files in the `ext/` directory.

cpp/ql/lib/change-notes/2026-03-30-nsdmi-dataflow.md

@@ -0,0 +1,4 @@
+---
+category: minorAnalysis
+---
+* Added dataflow through members initialized via non-static data member initialization (NSDMI).

cpp/ql/lib/change-notes/2026-03-31-http-flow-sources.md

@@ -0,0 +1,4 @@
+---
+category: minorAnalysis
+---
+* Added `HttpReceiveHttpRequest`, `HttpReceiveRequestEntityBody`, and `HttpReceiveClientCertificate` from Win32's `http.h` as remote flow sources.

cpp/ql/lib/change-notes/2026-03-31-meson.md

@@ -0,0 +1,4 @@
+---
+category: feature
+---
+* Added a subclass `MesonPrivateTestFile` of `ConfigurationTestFile` that represents files created by Meson to test the build configuration.

cpp/ql/lib/change-notes/2026-04-07-autoconf.md

@@ -0,0 +1,4 @@
+---
+category: feature
+---
+* Added a subclass `AutoconfConfigureTestFile` of `ConfigurationTestFile` that represents files created by GNU autoconf configure scripts to test the build configuration.

cpp/ql/lib/change-notes/released/8.0.1.md

@@ -0,0 +1,5 @@
+## 8.0.1
+
+### Minor Analysis Improvements
+
+* Inline expectations test comments, which are of the form `// $ tag` or `// $ tag=value`, are now parsed more strictly and will not be recognized if there isn't a space after the `$` symbol.

cpp/ql/lib/change-notes/released/8.0.2.md

@@ -0,0 +1,3 @@
+## 8.0.2
+
+No user-facing changes.

cpp/ql/lib/change-notes/released/8.0.3.md

@@ -0,0 +1,3 @@
+## 8.0.3
+
+No user-facing changes.

cpp/ql/lib/codeql-pack.release.yml

@@ -1,2 +1,2 @@
 ---
-lastReleaseVersion: 8.0.0
+lastReleaseVersion: 8.0.3

cpp/ql/lib/ext/Windows.model.yml

@@ -31,6 +31,9 @@ extensions:
       - ["", "", False, "WinHttpQueryHeadersEx", "", "", "Argument[*5]", "remote", "manual"]
       - ["", "", False, "WinHttpQueryHeadersEx", "", "", "Argument[*6]", "remote", "manual"]
       - ["", "", False, "WinHttpQueryHeadersEx", "", "", "Argument[**8]", "remote", "manual"]
+      - ["", "", False, "HttpReceiveHttpRequest", "", "", "Argument[*3]", "remote", "manual"]
+      - ["", "", False, "HttpReceiveRequestEntityBody", "", "", "Argument[*3]", "remote", "manual"]
+      - ["", "", False, "HttpReceiveClientCertificate", "", "", "Argument[*3]", "remote", "manual"]
   - addsTo:
       pack: codeql/cpp-all
       extensible: summaryModel

cpp/ql/lib/ext/ZMQ.model.yml

@@ -0,0 +1,22 @@
+# ZeroMQ networking library models
+extensions:
+  - addsTo:
+      pack: codeql/cpp-all
+      extensible: sourceModel
+    data: # namespace, type, subtypes, name, signature, ext, output, kind, provenance
+      - ["", "", False, "zmq_recv", "", "", "Argument[*1]", "remote", "manual"]
+      - ["", "", False, "zmq_recvmsg", "", "", "Argument[*1]", "remote", "manual"]
+      - ["", "", False, "zmq_msg_recv", "", "", "Argument[*0]", "remote", "manual"]
+  - addsTo:
+      pack: codeql/cpp-all
+      extensible: sinkModel
+    data: # namespace, type, subtypes, name, signature, ext, input, kind, provenance
+      - ["", "", False, "zmq_send", "", "", "Argument[*1]", "remote-sink", "manual"]
+      - ["", "", False, "zmq_sendmsg", "", "", "Argument[*1]", "remote-sink", "manual"]
+      - ["", "", False, "zmq_msg_send", "", "", "Argument[*0]", "remote-sink", "manual"]
+  - addsTo:
+      pack: codeql/cpp-all
+      extensible: summaryModel
+    data: # namespace, type, subtypes, name, signature, ext, input, output, kind, provenance
+      - ["", "", False, "zmq_msg_init_data", "", "", "Argument[*1]", "Argument[*0]", "taint", "manual"]
+      - ["", "", False, "zmq_msg_data", "", "", "Argument[*0]", "ReturnValue[*]", "taint", "manual"]

cpp/ql/lib/ext/getc.model.yml

@@ -0,0 +1,19 @@
+# Models for getc and similar character-reading functions
+extensions:
+  - addsTo:
+      pack: codeql/cpp-all
+      extensible: sourceModel
+    data: # namespace, type, subtypes, name, signature, ext, output, kind, provenance
+      - ["", "", False, "getc", "", "", "ReturnValue", "remote", "manual"]
+      - ["", "", False, "getwc", "", "", "ReturnValue", "remote", "manual"]
+      - ["", "", False, "_getc_nolock", "", "", "ReturnValue", "remote", "manual"]
+      - ["", "", False, "_getwc_nolock", "", "", "ReturnValue", "remote", "manual"]
+      - ["", "", False, "getch", "", "", "ReturnValue", "local", "manual"]
+      - ["", "", False, "_getch", "", "", "ReturnValue", "local", "manual"]
+      - ["", "", False, "_getwch", "", "", "ReturnValue", "local", "manual"]
+      - ["", "", False, "_getch_nolock", "", "", "ReturnValue", "local", "manual"]
+      - ["", "", False, "_getwch_nolock", "", "", "ReturnValue", "local", "manual"]
+      - ["", "", False, "getchar", "", "", "ReturnValue", "local", "manual"]
+      - ["", "", False, "getwchar", "", "", "ReturnValue", "local", "manual"]
+      - ["", "", False, "_getchar_nolock", "", "", "ReturnValue", "local", "manual"]
+      - ["", "", False, "_getwchar_nolock", "", "", "ReturnValue", "local", "manual"]

cpp/ql/lib/qlpack.yml

@@ -1,5 +1,5 @@
 name: codeql/cpp-all
-version: 8.0.1-dev
+version: 8.0.4-dev
 groups: cpp
 dbscheme: semmlecode.cpp.dbscheme
 extractor: cpp

cpp/ql/lib/semmle/code/cpp/ConfigurationTestFile.qll

@@ -26,3 +26,26 @@ class CmakeTryCompileFile extends ConfigurationTestFile {
     )
   }
 }
+
+/**
+ * A file created by Meson to test the system configuration.
+ */
+class MesonPrivateTestFile extends ConfigurationTestFile {
+  MesonPrivateTestFile() {
+    this.getBaseName() = "testfile.c" and
+    exists(Folder folder, Folder parent |
+      folder = this.getParentContainer() and
+      parent = folder.getParentContainer()
+    |
+      folder.getBaseName().matches("tmp%") and
+      parent.getBaseName() = "meson-private"
+    )
+  }
+}
+
+/**
+ * A file created by a GNU autoconf configure script to test the system configuration.
+ */
+class AutoconfConfigureTestFile extends ConfigurationTestFile {
+  AutoconfConfigureTestFile() { this.getBaseName().regexpMatch("conftest[0-9]*\\.c(pp)?") }
+}

cpp/ql/lib/semmle/code/cpp/Function.qll

@@ -524,6 +524,12 @@ class Function extends Declaration, ControlFlowNode, AccessHolder, @function {
       not exists(NewOrNewArrayExpr new | e = new.getAllocatorCall().getArgument(0))
     )
   }
+
+  /**
+   * Holds if this function has an ambiguous return type, meaning that zero or multiple return
+   * types for this function are present in the database (this can occur in `build-mode: none`).
+   */
+  predicate hasAmbiguousReturnType() { count(this.getType()) != 1 }
 }
 
 pragma[noinline]

cpp/ql/lib/semmle/code/cpp/commons/Printf.qll

@@ -163,12 +163,23 @@ predicate primitiveVariadicFormatter(
   )
 }
 
+/**
+ * Gets a function call whose target is a variadic formatter with the given
+ * `type`, `format` parameter index and `output` parameter index.
+ *
+ * Join-order helper for `callsVariadicFormatter`.
+ */
+pragma[nomagic]
+private predicate callsVariadicFormatterCall(FunctionCall fc, string type, int format, int output) {
+  variadicFormatter(fc.getTarget(), type, format, output)
+}
+
 private predicate callsVariadicFormatter(
   Function f, string type, int formatParamIndex, int outputParamIndex
 ) {
   // calls a variadic formatter with `formatParamIndex`, `outputParamIndex` linked
   exists(FunctionCall fc, int format, int output |
-    variadicFormatter(pragma[only_bind_into](fc.getTarget()), type, format, output) and
+    callsVariadicFormatterCall(fc, type, format, output) and
     fc.getEnclosingFunction() = f and
     fc.getArgument(format) = f.getParameter(formatParamIndex).getAnAccess() and
     fc.getArgument(output) = f.getParameter(outputParamIndex).getAnAccess()
@@ -176,7 +187,7 @@ private predicate callsVariadicFormatter(
   or
   // calls a variadic formatter with only `formatParamIndex` linked
   exists(FunctionCall fc, string calledType, int format, int output |
-    variadicFormatter(pragma[only_bind_into](fc.getTarget()), calledType, format, output) and
+    callsVariadicFormatterCall(fc, calledType, format, output) and
     fc.getEnclosingFunction() = f and
     fc.getArgument(format) = f.getParameter(formatParamIndex).getAnAccess() and
     not fc.getArgument(output) = f.getParameter(_).getAnAccess() and

cpp/ql/lib/semmle/code/cpp/controlflow/IRGuards.qll

@@ -1663,7 +1663,7 @@ private module Cached {
   private predicate compares_ge(
     ValueNumber test, Operand left, Operand right, int k, boolean isGe, GuardValue value
   ) {
-    exists(int onemk | k = 1 - onemk | compares_lt(test, right, left, onemk, isGe, value))
+    compares_lt(test, right, left, 1 - k, isGe, value)
   }
 
   /** Rearrange various simple comparisons into `left < right + k` form. */

cpp/ql/lib/semmle/code/cpp/dataflow/ExternalFlow.qll

@@ -1,9 +1,10 @@
 /**
  * INTERNAL use only. This is an experimental API subject to change without notice.
  *
- * Provides classes and predicates for dealing with flow models specified in CSV format.
+ * Provides classes and predicates for dealing with flow models specified
+ * in data extension files.
  *
- * The CSV specification has the following columns:
+ * The extensible relations have the following columns:
  * - Sources:
  *   `namespace; type; subtypes; name; signature; ext; output; kind`
  * - Sinks:
@@ -104,117 +105,9 @@ private import internal.FlowSummaryImpl::Private
 private import internal.FlowSummaryImpl::Private::External
 private import internal.ExternalFlowExtensions::Extensions as Extensions
 private import codeql.mad.ModelValidation as SharedModelVal
-private import codeql.util.Unit
 private import codeql.mad.static.ModelsAsData as SharedMaD
 
-/**
- * A unit class for adding additional source model rows.
- *
- * Extend this class to add additional source definitions.
- */
-class SourceModelCsv extends Unit {
-  /** Holds if `row` specifies a source definition. */
-  abstract predicate row(string row);
-}
-
-/**
- * A unit class for adding additional sink model rows.
- *
- * Extend this class to add additional sink definitions.
- */
-class SinkModelCsv extends Unit {
-  /** Holds if `row` specifies a sink definition. */
-  abstract predicate row(string row);
-}
-
-/**
- * A unit class for adding additional summary model rows.
- *
- * Extend this class to add additional flow summary definitions.
- */
-class SummaryModelCsv extends Unit {
-  /** Holds if `row` specifies a summary definition. */
-  abstract predicate row(string row);
-}
-
-/** Holds if `row` is a source model. */
-predicate sourceModel(string row) { any(SourceModelCsv s).row(row) }
-
-/** Holds if `row` is a sink model. */
-predicate sinkModel(string row) { any(SinkModelCsv s).row(row) }
-
-/** Holds if `row` is a summary model. */
-predicate summaryModel(string row) { any(SummaryModelCsv s).row(row) }
-
 private module MadInput implements SharedMaD::InputSig {
-  /** Holds if a source model exists for the given parameters. */
-  predicate additionalSourceModel(
-    string namespace, string type, boolean subtypes, string name, string signature, string ext,
-    string output, string kind, string provenance, string model
-  ) {
-    exists(string row |
-      sourceModel(row) and
-      row.splitAt(";", 0) = namespace and
-      row.splitAt(";", 1) = type and
-      row.splitAt(";", 2) = subtypes.toString() and
-      subtypes = [true, false] and
-      row.splitAt(";", 3) = name and
-      row.splitAt(";", 4) = signature and
-      row.splitAt(";", 5) = ext and
-      row.splitAt(";", 6) = output and
-      row.splitAt(";", 7) = kind
-    ) and
-    provenance = "manual" and
-    model = ""
-  }
-
-  /** Holds if a sink model exists for the given parameters. */
-  predicate additionalSinkModel(
-    string namespace, string type, boolean subtypes, string name, string signature, string ext,
-    string input, string kind, string provenance, string model
-  ) {
-    exists(string row |
-      sinkModel(row) and
-      row.splitAt(";", 0) = namespace and
-      row.splitAt(";", 1) = type and
-      row.splitAt(";", 2) = subtypes.toString() and
-      subtypes = [true, false] and
-      row.splitAt(";", 3) = name and
-      row.splitAt(";", 4) = signature and
-      row.splitAt(";", 5) = ext and
-      row.splitAt(";", 6) = input and
-      row.splitAt(";", 7) = kind
-    ) and
-    provenance = "manual" and
-    model = ""
-  }
-
-  /**
-   * Holds if a summary model exists for the given parameters.
-   *
-   * This predicate does not expand `@` to `*`s.
-   */
-  predicate additionalSummaryModel(
-    string namespace, string type, boolean subtypes, string name, string signature, string ext,
-    string input, string output, string kind, string provenance, string model
-  ) {
-    exists(string row |
-      summaryModel(row) and
-      row.splitAt(";", 0) = namespace and
-      row.splitAt(";", 1) = type and
-      row.splitAt(";", 2) = subtypes.toString() and
-      subtypes = [true, false] and
-      row.splitAt(";", 3) = name and
-      row.splitAt(";", 4) = signature and
-      row.splitAt(";", 5) = ext and
-      row.splitAt(";", 6) = input and
-      row.splitAt(";", 7) = output and
-      row.splitAt(";", 8) = kind
-    ) and
-    provenance = "manual" and
-    model = ""
-  }
-
   string namespaceSegmentSeparator() { result = "::" }
 }
 
@@ -250,8 +143,8 @@ predicate summaryModel(
   )
 }
 
-/** Provides a query predicate to check the CSV data for validation errors. */
-module CsvValidation {
+/** Provides a query predicate to check the data for validation errors. */
+module ModelValidation {
   private string getInvalidModelInput() {
     exists(string pred, AccessPath input, string part |
       sinkModel(_, _, _, _, _, _, input, _, _, _) and pred = "sink"
@@ -294,40 +187,6 @@ module CsvValidation {
 
   private module KindVal = SharedModelVal::KindValidation<KindValConfig>;
 
-  private string getInvalidModelSubtype() {
-    exists(string pred, string row |
-      sourceModel(row) and pred = "source"
-      or
-      sinkModel(row) and pred = "sink"
-      or
-      summaryModel(row) and pred = "summary"
-    |
-      exists(string b |
-        b = row.splitAt(";", 2) and
-        not b = ["true", "false"] and
-        result = "Invalid boolean \"" + b + "\" in " + pred + " model."
-      )
-    )
-  }
-
-  private string getInvalidModelColumnCount() {
-    exists(string pred, string row, int expect |
-      sourceModel(row) and expect = 8 and pred = "source"
-      or
-      sinkModel(row) and expect = 8 and pred = "sink"
-      or
-      summaryModel(row) and expect = 9 and pred = "summary"
-    |
-      exists(int cols |
-        cols = 1 + max(int n | exists(row.splitAt(";", n))) and
-        cols != expect and
-        result =
-          "Wrong number of columns in " + pred + " model row, expected " + expect + ", got " + cols +
-            "."
-      )
-    )
-  }
-
   private string getInvalidModelSignature() {
     exists(string pred, string namespace, string type, string name, string signature, string ext |
       sourceModel(namespace, type, _, name, signature, ext, _, _, _, _) and pred = "source"
@@ -353,12 +212,25 @@ module CsvValidation {
     )
   }
 
-  /** Holds if some row in a CSV-based flow model appears to contain typos. */
+  private string getIncorrectConstructorSummaryOutput() {
+    exists(string namespace, string type, string name, string output |
+      type = name or
+      type = name + "<" + any(string s)
+    |
+      summaryModel(namespace, type, _, name, _, _, _, output, _, _, _) and
+      output.matches("ReturnValue%") and
+      result =
+        "Constructor model for " + namespace + "." + type +
+          " should use `Argument[this]` in the output, not `ReturnValue`."
+    )
+  }
+
+  /** Holds if some row in a MaD flow model appears to contain typos. */
   query predicate invalidModelRow(string msg) {
     msg =
       [
         getInvalidModelSignature(), getInvalidModelInput(), getInvalidModelOutput(),
-        getInvalidModelSubtype(), getInvalidModelColumnCount(), KindVal::getInvalidModelKind()
+        KindVal::getInvalidModelKind(), getIncorrectConstructorSummaryOutput()
       ]
   }
 }
@@ -555,6 +427,7 @@ private Locatable getSupportedFunctionTemplateArgument(Function templateFunction
  * Normalize the `n`'th parameter of `f` by replacing template names
  * with `func:N` (where `N` is the index of the template).
  */
+pragma[nomagic]
 private string getTypeNameWithoutFunctionTemplates(Function f, int n, int remaining) {
   exists(Function templateFunction |
     templateFunction = getFullyTemplatedFunction(f) and
@@ -1011,7 +884,7 @@ private module Cached {
   }
 
   /**
-   * Holds if `node` is specified as a source with the given kind in a CSV flow
+   * Holds if `node` is specified as a source with the given kind in a MaD flow
    * model.
    */
   cached
@@ -1022,7 +895,7 @@ private module Cached {
   }
 
   /**
-   * Holds if `node` is specified as a sink with the given kind in a CSV flow
+   * Holds if `node` is specified as a sink with the given kind in a MaD flow
    * model.
    */
   cached

cpp/ql/lib/semmle/code/cpp/dataflow/internal/FlowSummaryImpl.qll

@@ -201,7 +201,7 @@ module SourceSinkInterpretationInput implements
     string toString() {
       result = this.asElement().toString()
       or
-      result = this.asNode().toString()
+      result = this.asNode().toStringImpl()
       or
       result = this.asCall().toString()
     }

cpp/ql/lib/semmle/code/cpp/exprs/Call.qll

@@ -585,12 +585,15 @@ class ConstructorDelegationInit extends ConstructorBaseInit, @ctordelegatinginit
 
 /**
  * An initialization of a member variable performed as part of a
- * constructor's explicit initializer list or implicit actions.
+ * constructor's initializer list or by default initialization.
+ *
  * In the example below, member variable `b` is being initialized by
- * constructor parameter `a`:
+ * constructor parameter `a`, and `c` is initialized by default
+ * initialization:
  * ```
  * struct S {
  *   int b;
+ *   int c = 3;
  *   S(int a): b(a) {}
  * } s(2);
  * ```
@@ -616,6 +619,28 @@ class ConstructorFieldInit extends ConstructorInit, @ctorfieldinit {
   override predicate mayBeGloballyImpure() { this.getExpr().mayBeGloballyImpure() }
 }
 
+/**
+ * An initialization of a member variable performed as part of a
+ * constructor's explicit initializer list.
+ */
+class ConstructorDirectFieldInit extends ConstructorFieldInit {
+  ConstructorDirectFieldInit() { exists(this.getChild(0)) }
+
+  override string getAPrimaryQlClass() { result = "ConstructorDirectFieldInit" }
+}
+
+/**
+ * An initialization of a member variable performed by default
+ * initialization.
+ */
+class ConstructorDefaultFieldInit extends ConstructorFieldInit {
+  ConstructorDefaultFieldInit() {
+    not exists(this.getChild(0)) and exists(this.getTarget().getInitializer())
+  }
+
+  override string getAPrimaryQlClass() { result = "ConstructorDefaultFieldInit" }
+}
+
 /**
  * A call to a destructor of a base class or field as part of a destructor's
  * compiler-generated actions.

cpp/ql/lib/semmle/code/cpp/internal/Overlay.qll

@@ -6,117 +6,67 @@ private import OverlayXml
 
 /**
  * Holds always for the overlay variant and never for the base variant.
- * This local predicate is used to define local predicates that behave
- * differently for the base and overlay variant.
  */
 overlay[local]
 predicate isOverlay() { databaseMetadata("isOverlay", "true") }
 
-overlay[local]
-private string getLocationFilePath(@location_default loc) {
-  exists(@file file | locations_default(loc, file, _, _, _, _) | files(file, result))
-}
-
 /**
- * Gets the file path for an element with a single location.
+ * Holds if the TRAP file or tag `t` is reachable from source file `sourceFile`
+ * in the base (isOverlayVariant=false) or overlay (isOverlayVariant=true) variant.
  */
 overlay[local]
-private string getSingleLocationFilePath(@element e) {
-  exists(@location_default loc |
-    var_decls(e, _, _, _, loc)
-    or
-    fun_decls(e, _, _, _, loc)
-    or
-    type_decls(e, _, loc)
-    or
-    namespace_decls(e, _, loc, _)
-    or
-    macroinvocations(e, _, loc, _)
-    or
-    preprocdirects(e, _, loc)
-    or
-    diagnostics(e, _, _, _, _, loc)
-    or
-    usings(e, _, loc, _)
-    or
-    static_asserts(e, _, _, loc, _)
-    or
-    derivations(e, _, _, _, loc)
-    or
-    frienddecls(e, _, _, loc)
-    or
-    comments(e, _, loc)
-    or
-    exprs(e, _, loc)
-    or
-    stmts(e, _, loc)
-    or
-    initialisers(e, _, _, loc)
-    or
-    attributes(e, _, _, _, loc)
-    or
-    attribute_args(e, _, _, _, loc)
-    or
-    namequalifiers(e, _, _, loc)
-    or
-    enumconstants(e, _, _, _, _, loc)
-    or
-    type_mentions(e, _, loc, _)
-    or
-    lambda_capture(e, _, _, _, _, _, loc)
-    or
-    concept_templates(e, _, loc)
-  |
-    result = getLocationFilePath(loc)
+private predicate locallyReachableTrapOrTag(
+  boolean isOverlayVariant, string sourceFile, @trap_or_tag t
+) {
+  exists(@source_file sf, @trap trap |
+    (if isOverlay() then isOverlayVariant = true else isOverlayVariant = false) and
+    source_file_uses_trap(sf, trap) and
+    source_file_name(sf, sourceFile) and
+    (t = trap or trap_uses_tag(trap, t))
   )
 }
 
 /**
- * Gets the file path for an element with potentially multiple locations.
+ * Holds if element `e` is in TRAP file or tag `t`
+ * in the base (isOverlayVariant=false) or overlay (isOverlayVariant=true) variant.
  */
 overlay[local]
-private string getMultiLocationFilePath(@element e) {
-  exists(@location_default loc |
-    var_decls(_, e, _, _, loc)
-    or
-    fun_decls(_, e, _, _, loc)
-    or
-    type_decls(_, e, loc)
-    or
-    namespace_decls(_, e, loc, _)
-  |
-    result = getLocationFilePath(loc)
-  )
-}
-
-/**
- * A local helper predicate that holds in the base variant and never in the
- * overlay variant.
- */
-overlay[local]
-private predicate isBase() { not isOverlay() }
-
-/**
- * Holds if `path` was extracted in the overlay database.
- */
-overlay[local]
-private predicate overlayHasFile(string path) {
-  isOverlay() and
-  files(_, path) and
-  path != ""
+private predicate locallyInTrapOrTag(boolean isOverlayVariant, @element e, @trap_or_tag t) {
+  (if isOverlay() then isOverlayVariant = true else isOverlayVariant = false) and
+  in_trap_or_tag(e, t)
 }
 
 /**
  * Discards an element from the base variant if:
- * - It has a single location in a file extracted in the overlay, or
- * - All of its locations are in files extracted in the overlay.
+ * - We have knowledge about what TRAP file or tag it is in (in the base).
+ * - It is not in any overlay TRAP file or tag that is reachable from an overlay source file.
+ * - For every base TRAP file or tag that contains it and is reachable from a base source file,
+ *   either the source file has changed, or the overlay has redefined the TRAP file or tag,
+ *   or the overlay runner has re-extracted the same source file.
  */
 overlay[discard_entity]
 private predicate discardElement(@element e) {
-  isBase() and
-  (
-    overlayHasFile(getSingleLocationFilePath(e))
-    or
-    forex(string path | path = getMultiLocationFilePath(e) | overlayHasFile(path))
+  // If we don't have any knowledge about what TRAP file something
+  // is in, then we don't want to discard it, so we only consider
+  // entities that are known to be in a base TRAP file or tag.
+  locallyInTrapOrTag(false, e, _) and
+  // Anything that is reachable from an overlay source file should
+  // not be discarded.
+  not exists(@trap_or_tag t | locallyInTrapOrTag(true, e, t) |
+    locallyReachableTrapOrTag(true, _, t)
+  ) and
+  // Finally, we have to make sure the base variant does not retain it.
+  // If it is reachable from a base source file, then that is
+  // sufficient unless either the base source file has changed (in
+  // particular, been deleted), or the overlay has redefined the TRAP
+  // file or tag it is in, or the overlay runner has re-extracted the same
+  // source file (e.g. because a header it includes has changed).
+  forall(@trap_or_tag t, string sourceFile |
+    locallyInTrapOrTag(false, e, t) and
+    locallyReachableTrapOrTag(false, sourceFile, t)
+  |
+    overlayChangedFiles(sourceFile) or
+    locallyReachableTrapOrTag(true, _, t) or
+    locallyReachableTrapOrTag(true, sourceFile, _)
   )
 }

cpp/ql/lib/semmle/code/cpp/ir/dataflow/internal/DataFlowDispatch.qll

@@ -238,7 +238,12 @@ private module TrackVirtualDispatch<methodDispatchSig/1 virtualDispatch0> {
 
   private import TypeTracking<Location, TtInput>::TypeTrack<qualifierSource/1>::Graph<qualifierOfVirtualCall/1>
 
-  private predicate edgePlus(PathNode n1, PathNode n2) = fastTC(edges/2)(n1, n2)
+  private predicate isSource(PathNode n) { n.isSource() }
+
+  private predicate isSink(PathNode n) { n.isSink() }
+
+  private predicate edgePlus(PathNode n1, PathNode n2) =
+    doublyBoundedFastTC(edges/2, isSource/1, isSink/1)(n1, n2)
 
   /**
    * Gets the most specific implementation of `mf` that may be called when the
@@ -255,6 +260,15 @@ private module TrackVirtualDispatch<methodDispatchSig/1 virtualDispatch0> {
     )
   }
 
+  pragma[nomagic]
+  private MemberFunction mostSpecificForSource(PathNode p1, MemberFunction mf) {
+    p1.isSource() and
+    exists(Class derived |
+      qualifierSourceImpl(p1.getNode(), derived) and
+      result = mostSpecific(mf, derived)
+    )
+  }
+
   /**
    * Gets a possible pair of end-points `(p1, p2)` where:
    * - `p1` is a derived-to-base conversion that converts from some
@@ -264,16 +278,16 @@ private module TrackVirtualDispatch<methodDispatchSig/1 virtualDispatch0> {
    * - `callable` is the most specific implementation that may be called when
    * the qualifier has type `derived`.
    */
+  bindingset[p1, p2]
+  pragma[inline_late]
   private predicate pairCand(
     PathNode p1, PathNode p2, DataFlowPrivate::DataFlowCallable callable,
     DataFlowPrivate::DataFlowCall call
   ) {
-    exists(Class derived, MemberFunction mf |
-      qualifierSourceImpl(p1.getNode(), derived) and
+    p2.isSink() and
+    exists(MemberFunction mf |
       qualifierOfVirtualCallImpl(p2.getNode(), call.asCallInstruction(), mf) and
-      p1.isSource() and
-      p2.isSink() and
-      callable.asSourceCallable() = mostSpecific(mf, derived)
+      callable.asSourceCallable() = mostSpecificForSource(p1, mf)
     )
   }
 

cpp/ql/lib/semmle/code/cpp/ir/dataflow/internal/DataFlowPrivate.qll

@@ -1,5 +1,6 @@
 private import cpp as Cpp
 private import DataFlowUtil
+private import DataFlowNodes
 private import semmle.code.cpp.ir.IR
 private import DataFlowDispatch
 private import semmle.code.cpp.ir.internal.IRCppLanguage
@@ -16,28 +17,42 @@ private import semmle.code.cpp.dataflow.ExternalFlow as External
 cached
 private module Cached {
   cached
-  module Nodes0 {
-    cached
-    newtype TIRDataFlowNode0 =
-      TInstructionNode0(Instruction i) {
-        not Ssa::ignoreInstruction(i) and
-        not exists(Operand op |
-          not Ssa::ignoreOperand(op) and i = Ssa::getIRRepresentationOfOperand(op)
-        ) and
-        // We exclude `void`-typed instructions because they cannot contain data.
-        // However, if the instruction is a glvalue, and their type is `void`, then the result
-        // type of the instruction is really `void*`, and thus we still want to have a dataflow
-        // node for it.
-        (not i.getResultType() instanceof VoidType or i.isGLValue())
-      } or
-      TMultipleUseOperandNode0(Operand op) {
-        not Ssa::ignoreOperand(op) and not exists(Ssa::getIRRepresentationOfOperand(op))
-      } or
-      TSingleUseOperandNode0(Operand op) {
-        not Ssa::ignoreOperand(op) and exists(Ssa::getIRRepresentationOfOperand(op))
-      }
+  newtype TIRDataFlowNode0 =
+    TInstructionNode0(Instruction i) {
+      not Ssa::ignoreInstruction(i) and
+      not exists(Operand op |
+        not Ssa::ignoreOperand(op) and i = Ssa::getIRRepresentationOfOperand(op)
+      ) and
+      // We exclude `void`-typed instructions because they cannot contain data.
+      // However, if the instruction is a glvalue, and their type is `void`, then the result
+      // type of the instruction is really `void*`, and thus we still want to have a dataflow
+      // node for it.
+      (not i.getResultType() instanceof VoidType or i.isGLValue())
+    } or
+    TMultipleUseOperandNode0(Operand op) {
+      not Ssa::ignoreOperand(op) and not exists(Ssa::getIRRepresentationOfOperand(op))
+    } or
+    TSingleUseOperandNode0(Operand op) {
+      not Ssa::ignoreOperand(op) and exists(Ssa::getIRRepresentationOfOperand(op))
+    }
+
+  cached
+  string toStringCached(Node n) {
+    result = toExprString(n)
+    or
+    not exists(toExprString(n)) and
+    result = n.toStringImpl()
   }
 
+  cached
+  Location getLocationCached(Node n) { result = n.getLocationImpl() }
+
+  cached
+  newtype TContentApprox =
+    TFieldApproxContent(string s) { fieldHasApproxName(_, s) } or
+    TUnionApproxContent(string s) { unionHasApproxName(_, s) } or
+    TElementApproxContent()
+
   /**
    * Gets an additional term that is added to the `join` and `branch` computations to reflect
    * an additional forward or backwards branching factor that is not taken into account
@@ -59,38 +74,174 @@ private module Cached {
       result = countNumberOfBranchesUsingParameter(switch, p)
     )
   }
-}
 
-import Cached
-private import Nodes0
+  cached
+  newtype TDataFlowCallable =
+    TSourceCallable(Cpp::Declaration decl) or
+    TSummarizedCallable(FlowSummaryImpl::Public::SummarizedCallable c)
 
-/**
- * A module for calculating the number of stars (i.e., `*`s) needed for various
- * dataflow node `toString` predicates.
- */
-module NodeStars {
-  private int getNumberOfIndirections(Node n) {
-    result = n.(RawIndirectOperand).getIndirectionIndex()
+  cached
+  newtype TDataFlowCall =
+    TNormalCall(CallInstruction call) or
+    TSummaryCall(
+      FlowSummaryImpl::Public::SummarizedCallable c, FlowSummaryImpl::Private::SummaryNode receiver
+    ) {
+      FlowSummaryImpl::Private::summaryCallbackRange(c, receiver)
+    }
+
+  /**
+   * Holds if data can flow from `node1` to `node2` in a way that loses the
+   * calling context. For example, this would happen with flow through a
+   * global or static variable.
+   */
+  cached
+  predicate jumpStep(Node n1, Node n2) {
+    exists(GlobalLikeVariable v |
+      exists(Ssa::GlobalUse globalUse |
+        v = globalUse.getVariable() and
+        n1.(FinalGlobalValue).getGlobalUse() = globalUse
+      |
+        globalUse.getIndirection() = getMinIndirectionForGlobalUse(globalUse) and
+        v = n2.asVariable()
+        or
+        v = n2.asIndirectVariable(globalUse.getIndirection())
+      )
+      or
+      exists(Ssa::GlobalDef globalDef |
+        v = globalDef.getVariable() and
+        n2.(InitialGlobalValue).getGlobalDef() = globalDef
+      |
+        globalDef.getIndirection() = getMinIndirectionForGlobalDef(globalDef) and
+        v = n1.asVariable()
+        or
+        v = n1.asIndirectVariable(globalDef.getIndirection())
+      )
+    )
     or
-    result = n.(RawIndirectInstruction).getIndirectionIndex()
-    or
-    result = n.(VariableNode).getIndirectionIndex()
-    or
-    result = n.(PostUpdateNodeImpl).getIndirectionIndex()
-    or
-    result = n.(FinalParameterNode).getIndirectionIndex()
-    or
-    result = n.(BodyLessParameterNodeImpl).getIndirectionIndex()
+    // models-as-data summarized flow
+    FlowSummaryImpl::Private::Steps::summaryJumpStep(n1.(FlowSummaryNode).getSummaryNode(),
+      n2.(FlowSummaryNode).getSummaryNode())
   }
 
   /**
-   * Gets the number of stars (i.e., `*`s) needed to produce the `toString`
-   * output for `n`.
+   * Holds if data can flow from `node1` to `node2` via an assignment to `f`.
+   * Thus, `node2` references an object with a field `f` that contains the
+   * value of `node1`.
+   *
+   * The boolean `certain` is true if the destination address does not involve
+   * any pointer arithmetic, and false otherwise.
    */
-  string stars(Node n) { result = repeatStars(getNumberOfIndirections(n)) }
+  cached
+  predicate storeStepImpl(Node node1, Content c, Node node2, boolean certain) {
+    exists(
+      PostFieldUpdateNode postFieldUpdate, int indirectionIndex1, int numberOfLoads,
+      StoreInstruction store, FieldContent fc
+    |
+      postFieldUpdate = node2 and
+      fc = c and
+      nodeHasInstruction(node1, pragma[only_bind_into](store),
+        pragma[only_bind_into](indirectionIndex1)) and
+      postFieldUpdate.getIndirectionIndex() = 1 and
+      numberOfLoadsFromOperand(postFieldUpdate.getFieldAddress(),
+        store.getDestinationAddressOperand(), numberOfLoads, certain) and
+      fc.getAField() = postFieldUpdate.getUpdatedField() and
+      getIndirectionIndexLate(fc) = 1 + indirectionIndex1 + numberOfLoads
+    )
+    or
+    // models-as-data summarized flow
+    FlowSummaryImpl::Private::Steps::summaryStoreStep(node1.(FlowSummaryNode).getSummaryNode(), c,
+      node2.(FlowSummaryNode).getSummaryNode()) and
+    certain = true
+  }
+
+  /**
+   * Holds if data can flow from `node1` to `node2` via an assignment to `f`.
+   * Thus, `node2` references an object with a field `f` that contains the
+   * value of `node1`.
+   */
+  cached
+  predicate storeStep(Node node1, ContentSet c, Node node2) { storeStepImpl(node1, c, node2, _) }
+
+  /**
+   * Holds if data can flow from `node1` to `node2` via a read of `f`.
+   * Thus, `node1` references an object with a field `f` whose value ends up in
+   * `node2`.
+   */
+  cached
+  predicate readStep(Node node1, ContentSet c, Node node2) {
+    exists(
+      FieldAddress fa1, Operand operand, int numberOfLoads, int indirectionIndex2, FieldContent fc
+    |
+      fc = c and
+      nodeHasOperand(node2, operand, indirectionIndex2) and
+      // The `1` here matches the `node2.getIndirectionIndex() = 1` conjunct
+      // in `storeStep`.
+      nodeHasOperand(node1, fa1.getObjectAddressOperand(), 1) and
+      numberOfLoadsFromOperand(fa1, operand, numberOfLoads, _) and
+      fc.getAField() = fa1.getField() and
+      getIndirectionIndexLate(fc) = indirectionIndex2 + numberOfLoads
+    )
+    or
+    // models-as-data summarized flow
+    FlowSummaryImpl::Private::Steps::summaryReadStep(node1.(FlowSummaryNode).getSummaryNode(), c,
+      node2.(FlowSummaryNode).getSummaryNode())
+  }
+
+  /**
+   * Holds if values stored inside content `c` are cleared at node `n`.
+   */
+  cached
+  predicate clearsContent(Node n, ContentSet c) {
+    n =
+      any(PostUpdateNode pun, Content d |
+        d.impliesClearOf(c) and storeStepImpl(_, d, pun, true)
+      |
+        pun
+      ).getPreUpdateNode() and
+    (
+      not exists(Operand op, Cpp::Operation p |
+        n.(IndirectOperand).hasOperandAndIndirectionIndex(op, _) and
+        (
+          p instanceof Cpp::AssignPointerAddExpr or
+          p instanceof Cpp::AssignPointerSubExpr or
+          p instanceof Cpp::CrementOperation
+        )
+      |
+        p.getAnOperand() = op.getUse().getAst()
+      )
+      or
+      forex(PostUpdateNode pun, Content d |
+        pragma[only_bind_into](d).impliesClearOf(pragma[only_bind_into](c)) and
+        storeStepImpl(_, d, pun, true) and
+        pun.getPreUpdateNode() = n
+      |
+        c.(Content).getIndirectionIndex() = d.getIndirectionIndex()
+      )
+    )
+  }
 }
 
-import NodeStars
+import Cached
+
+private int getNumberOfIndirections(Node n) {
+  result = n.(RawIndirectOperand).getIndirectionIndex()
+  or
+  result = n.(RawIndirectInstruction).getIndirectionIndex()
+  or
+  result = n.(VariableNode).getIndirectionIndex()
+  or
+  result = n.(PostUpdateNodeImpl).getIndirectionIndex()
+  or
+  result = n.(FinalParameterNode).getIndirectionIndex()
+  or
+  result = n.(BodyLessParameterNodeImpl).getIndirectionIndex()
+}
+
+/**
+ * Gets the number of stars (i.e., `*`s) needed to produce the `toString`
+ * output for `n`.
+ */
+string stars(Node n) { result = repeatStars(getNumberOfIndirections(n)) }
 
 /**
  * A cut-down `DataFlow::Node` class that does not depend on the output of SSA.
@@ -828,85 +979,10 @@ private int getMinIndirectionForGlobalDef(Ssa::GlobalDef def) {
   result = getMinIndirectionsForType(def.getUnspecifiedType())
 }
 
-/**
- * Holds if data can flow from `node1` to `node2` in a way that loses the
- * calling context. For example, this would happen with flow through a
- * global or static variable.
- */
-predicate jumpStep(Node n1, Node n2) {
-  exists(GlobalLikeVariable v |
-    exists(Ssa::GlobalUse globalUse |
-      v = globalUse.getVariable() and
-      n1.(FinalGlobalValue).getGlobalUse() = globalUse
-    |
-      globalUse.getIndirection() = getMinIndirectionForGlobalUse(globalUse) and
-      v = n2.asVariable()
-      or
-      v = n2.asIndirectVariable(globalUse.getIndirection())
-    )
-    or
-    exists(Ssa::GlobalDef globalDef |
-      v = globalDef.getVariable() and
-      n2.(InitialGlobalValue).getGlobalDef() = globalDef
-    |
-      globalDef.getIndirection() = getMinIndirectionForGlobalDef(globalDef) and
-      v = n1.asVariable()
-      or
-      v = n1.asIndirectVariable(globalDef.getIndirection())
-    )
-  )
-  or
-  // models-as-data summarized flow
-  FlowSummaryImpl::Private::Steps::summaryJumpStep(n1.(FlowSummaryNode).getSummaryNode(),
-    n2.(FlowSummaryNode).getSummaryNode())
-}
-
 bindingset[c]
 pragma[inline_late]
 private int getIndirectionIndexLate(Content c) { result = c.getIndirectionIndex() }
 
-/**
- * Holds if data can flow from `node1` to `node2` via an assignment to `f`.
- * Thus, `node2` references an object with a field `f` that contains the
- * value of `node1`.
- *
- * The boolean `certain` is true if the destination address does not involve
- * any pointer arithmetic, and false otherwise. This has to do with whether a
- * store step can be used to clear a field (see `clearsContent`).
- */
-predicate storeStepImpl(Node node1, Content c, Node node2, boolean certain) {
-  exists(
-    PostFieldUpdateNode postFieldUpdate, int indirectionIndex1, int numberOfLoads,
-    StoreInstruction store, FieldContent fc
-  |
-    postFieldUpdate = node2 and
-    fc = c and
-    nodeHasInstruction(node1, pragma[only_bind_into](store),
-      pragma[only_bind_into](indirectionIndex1)) and
-    postFieldUpdate.getIndirectionIndex() = 1 and
-    numberOfLoadsFromOperand(postFieldUpdate.getFieldAddress(),
-      store.getDestinationAddressOperand(), numberOfLoads, certain) and
-    fc.getAField() = postFieldUpdate.getUpdatedField() and
-    getIndirectionIndexLate(fc) = 1 + indirectionIndex1 + numberOfLoads
-  )
-  or
-  // models-as-data summarized flow
-  FlowSummaryImpl::Private::Steps::summaryStoreStep(node1.(FlowSummaryNode).getSummaryNode(), c,
-    node2.(FlowSummaryNode).getSummaryNode()) and
-  certain = true
-}
-
-/**
- * Holds if data can flow from `node1` to `node2` via an assignment to `f`.
- * Thus, `node2` references an object with a field `f` that contains the
- * value of `node1`.
- */
-predicate storeStep(Node node1, ContentSet c, Node node2) { storeStepImpl(node1, c, node2, _) }
-
-/**
- * Holds if `operandFrom` flows to `operandTo` using a sequence of conversion-like
- * operations and exactly `n` `LoadInstruction` operations.
- */
 private predicate numberOfLoadsFromOperandRec(
   Operand operandFrom, Operand operandTo, int ind, boolean certain
 ) {
@@ -957,63 +1033,6 @@ predicate nodeHasInstruction(Node node, Instruction instr, int indirectionIndex)
   hasInstructionAndIndex(node, instr, indirectionIndex)
 }
 
-/**
- * Holds if data can flow from `node1` to `node2` via a read of `f`.
- * Thus, `node1` references an object with a field `f` whose value ends up in
- * `node2`.
- */
-predicate readStep(Node node1, ContentSet c, Node node2) {
-  exists(
-    FieldAddress fa1, Operand operand, int numberOfLoads, int indirectionIndex2, FieldContent fc
-  |
-    fc = c and
-    nodeHasOperand(node2, operand, indirectionIndex2) and
-    // The `1` here matches the `node2.getIndirectionIndex() = 1` conjunct
-    // in `storeStep`.
-    nodeHasOperand(node1, fa1.getObjectAddressOperand(), 1) and
-    numberOfLoadsFromOperand(fa1, operand, numberOfLoads, _) and
-    fc.getAField() = fa1.getField() and
-    getIndirectionIndexLate(fc) = indirectionIndex2 + numberOfLoads
-  )
-  or
-  // models-as-data summarized flow
-  FlowSummaryImpl::Private::Steps::summaryReadStep(node1.(FlowSummaryNode).getSummaryNode(), c,
-    node2.(FlowSummaryNode).getSummaryNode())
-}
-
-/**
- * Holds if values stored inside content `c` are cleared at node `n`.
- */
-predicate clearsContent(Node n, ContentSet c) {
-  n =
-    any(PostUpdateNode pun, Content d | d.impliesClearOf(c) and storeStepImpl(_, d, pun, true) | pun)
-        .getPreUpdateNode() and
-  (
-    // The crement operations and pointer addition and subtraction self-assign. We do not
-    // want to clear the contents if it is indirectly pointed at by any of these operations,
-    // as part of the contents might still be accessible afterwards. If there is no such
-    // indirection clearing the contents is safe.
-    not exists(Operand op, Cpp::Operation p |
-      n.(IndirectOperand).hasOperandAndIndirectionIndex(op, _) and
-      (
-        p instanceof Cpp::AssignPointerAddExpr or
-        p instanceof Cpp::AssignPointerSubExpr or
-        p instanceof Cpp::CrementOperation
-      )
-    |
-      p.getAnOperand() = op.getUse().getAst()
-    )
-    or
-    forex(PostUpdateNode pun, Content d |
-      pragma[only_bind_into](d).impliesClearOf(pragma[only_bind_into](c)) and
-      storeStepImpl(_, d, pun, true) and
-      pun.getPreUpdateNode() = n
-    |
-      c.(Content).getIndirectionIndex() = d.getIndirectionIndex()
-    )
-  )
-}
-
 /**
  * Holds if the value that is being tracked is expected to be stored inside content `c`
  * at node `n`.
@@ -1046,11 +1065,6 @@ class CastNode extends Node {
   CastNode() { none() } // stub implementation
 }
 
-cached
-private newtype TDataFlowCallable =
-  TSourceCallable(Cpp::Declaration decl) or
-  TSummarizedCallable(FlowSummaryImpl::Public::SummarizedCallable c)
-
 /**
  * A callable, which may be:
  *  - a function (that may contain code)
@@ -1134,15 +1148,6 @@ class DataFlowType extends TypeFinal {
   string toString() { result = "" }
 }
 
-cached
-private newtype TDataFlowCall =
-  TNormalCall(CallInstruction call) or
-  TSummaryCall(
-    FlowSummaryImpl::Public::SummarizedCallable c, FlowSummaryImpl::Private::SummaryNode receiver
-  ) {
-    FlowSummaryImpl::Private::summaryCallbackRange(c, receiver)
-  }
-
 private predicate summarizedCallableIsManual(SummarizedCallable sc) {
   sc.asSummarizedCallable().hasManualModel()
 }
@@ -1165,7 +1170,7 @@ class DataFlowCall extends TDataFlowCall {
   /**
    * Gets the `Function` that the call targets, if this is statically known.
    */
-  Function getStaticCallSourceTarget() { none() }
+  Declaration getStaticCallSourceTarget() { none() }
 
   /**
    * Gets the target of this call. We use the following strategy for deciding
@@ -1177,7 +1182,7 @@ class DataFlowCall extends TDataFlowCall {
    * whether is it manual or generated.
    */
   final DataFlowCallable getStaticCallTarget() {
-    exists(Function target | target = this.getStaticCallSourceTarget() |
+    exists(Declaration target | target = this.getStaticCallSourceTarget() |
       // Don't use the source callable if there is a manual model for the
       // target
       not exists(SummarizedCallable sc |
@@ -1237,7 +1242,7 @@ private class NormalCall extends DataFlowCall, TNormalCall {
 
   override CallTargetOperand getCallTargetOperand() { result = call.getCallTargetOperand() }
 
-  override Function getStaticCallSourceTarget() { result = call.getStaticCallTarget() }
+  override Declaration getStaticCallSourceTarget() { result = call.getStaticCallTarget() }
 
   override ArgumentOperand getArgumentOperand(int index) { result = call.getArgumentOperand(index) }
 
@@ -1523,12 +1528,6 @@ private predicate fieldHasApproxName(Field f, string s) {
 
 private predicate unionHasApproxName(Cpp::Union u, string s) { s = u.getName().charAt(0) }
 
-cached
-private newtype TContentApprox =
-  TFieldApproxContent(string s) { fieldHasApproxName(_, s) } or
-  TUnionApproxContent(string s) { unionHasApproxName(_, s) } or
-  TElementApproxContent()
-
 /** An approximated `Content`. */
 class ContentApprox extends TContentApprox {
   string toString() { none() } // overridden in subclasses

cpp/ql/lib/semmle/code/cpp/ir/dataflow/internal/ExprNodes.qll

@@ -6,8 +6,8 @@ private import cpp
 private import semmle.code.cpp.ir.IR
 private import DataFlowUtil
 private import DataFlowPrivate
-private import semmle.code.cpp.ir.implementation.raw.internal.TranslatedExpr
-private import semmle.code.cpp.ir.implementation.raw.internal.InstructionTag
+private import DataFlowNodes
+private import semmle.code.cpp.ir.implementation.raw.internal.IRConstruction as IRConstruction
 
 cached
 private module Cached {
@@ -73,17 +73,9 @@ private module Cached {
     // a result for `getConvertedResultExpression`. We remap this here so that
     // this `ConvertInstruction` maps to the result of the expression that
     // represents the extent.
-    exists(TranslatedNonConstantAllocationSize tas |
-      result = tas.getExtent().getExpr() and
-      instr = tas.getInstruction(AllocationExtentConvertTag())
-    )
+    result = IRConstruction::Raw::getAllocationExtentConvertExpr(instr)
     or
-    // There's no instruction that returns `ParenthesisExpr`, but some queries
-    // expect this
-    exists(TranslatedTransparentConversion ttc |
-      result = ttc.getExpr().(ParenthesisExpr) and
-      instr = ttc.getResult()
-    )
+    result = IRConstruction::Raw::getTransparentConversionParenthesisExpr(instr)
     or
     // Certain expressions generate `CopyValueInstruction`s only when they
     // are needed. Examples of this include crement operations and compound
@@ -112,10 +104,10 @@ private module Cached {
     // needed, and in that case the only value that will propagate forward in
     // the program is the value that's been updated. So in those cases we just
     // use the result of `node.asDefinition()` as the result of `node.asExpr()`.
-    exists(TranslatedCoreExpr tco |
-      tco.getInstruction(_) = instr and
-      tco.producesExprResult() and
-      result = asDefinitionImpl0(instr)
+    exists(StoreInstruction store |
+      store = instr and
+      IRConstruction::Raw::instructionProducesExprResult(store) and
+      result = asDefinitionImpl0(store)
     )
     or
     // IR construction breaks an array aggregate literal `{1, 2, 3}` into a
@@ -145,18 +137,9 @@ private module Cached {
     // For an expression such as `i += 2` we pretend that the generated
     // `StoreInstruction` contains the result of the expression even though
     // this isn't totally aligned with the C/C++ standard.
-    exists(TranslatedAssignOperation tao |
-      store = tao.getInstruction(AssignmentStoreTag()) and
-      result = tao.getExpr()
-    )
+    result = IRConstruction::Raw::getAssignOperationStoreExpr(store)
     or
-    // Similarly for `i++` and `++i` we pretend that the generated
-    // `StoreInstruction` contains the result of the expression even though
-    // this isn't totally aligned with the C/C++ standard.
-    exists(TranslatedCrementOperation tco |
-      store = tco.getInstruction(CrementStoreTag()) and
-      result = tco.getExpr()
-    )
+    result = IRConstruction::Raw::getCrementOperationStoreExpr(store)
   }
 
   /**
@@ -166,11 +149,7 @@ private module Cached {
    */
   private predicate excludeAsDefinitionResult(StoreInstruction store) {
     // Exclude the store to the temporary generated by a ternary expression.
-    exists(TranslatedConditionalExpr tce |
-      store = tce.getInstruction(ConditionValueFalseStoreTag())
-      or
-      store = tce.getInstruction(ConditionValueTrueStoreTag())
-    )
+    IRConstruction::Raw::isConditionalExprTempStore(store)
   }
 
   /**

cpp/ql/lib/semmle/code/cpp/ir/dataflow/internal/ModelUtil.qll

@@ -6,6 +6,7 @@
 private import semmle.code.cpp.ir.IR
 private import semmle.code.cpp.models.interfaces.FunctionInputsAndOutputs
 private import DataFlowUtil
+private import DataFlowNodes
 private import DataFlowPrivate
 private import SsaImpl as Ssa
 

cpp/ql/lib/semmle/code/cpp/ir/dataflow/internal/PrintIRFieldFlowSteps.qll

@@ -6,6 +6,7 @@ private import cpp
 private import semmle.code.cpp.ir.IR
 private import semmle.code.cpp.ir.dataflow.internal.DataFlowUtil
 private import semmle.code.cpp.ir.dataflow.internal.DataFlowPrivate
+private import semmle.code.cpp.ir.dataflow.internal.DataFlowNodes
 private import PrintIRUtilities
 
 /** A property provider for local IR dataflow store steps. */

cpp/ql/lib/semmle/code/cpp/ir/dataflow/internal/PrintIRLocalFlow.qll

@@ -2,6 +2,7 @@ private import cpp
 private import semmle.code.cpp.ir.IR
 private import semmle.code.cpp.ir.dataflow.internal.DataFlowUtil
 private import semmle.code.cpp.ir.dataflow.internal.DataFlowPrivate
+private import semmle.code.cpp.ir.dataflow.internal.DataFlowNodes
 private import SsaImpl as Ssa
 private import PrintIRUtilities
 

cpp/ql/lib/semmle/code/cpp/ir/dataflow/internal/PrintIRUtilities.qll

@@ -6,6 +6,7 @@ private import cpp
 private import semmle.code.cpp.ir.IR
 private import semmle.code.cpp.ir.dataflow.internal.DataFlowUtil
 private import semmle.code.cpp.ir.dataflow.internal.DataFlowPrivate
+private import semmle.code.cpp.ir.dataflow.internal.DataFlowNodes
 
 private Instruction getInstruction(Node n, string stars) {
   result = [n.asInstruction(), n.(RawIndirectInstruction).getInstruction()] and

cpp/ql/lib/semmle/code/cpp/ir/dataflow/internal/SsaImpl.qll

@@ -10,8 +10,9 @@ private import semmle.code.cpp.models.interfaces.PartialFlow as PartialFlow
 private import semmle.code.cpp.models.interfaces.FunctionInputsAndOutputs as FIO
 private import semmle.code.cpp.ir.internal.IRCppLanguage
 private import semmle.code.cpp.ir.dataflow.internal.ModelUtil
-private import semmle.code.cpp.ir.implementation.raw.internal.TranslatedInitialization
+private import semmle.code.cpp.ir.implementation.raw.internal.IRConstruction as IRConstruction
 private import DataFlowPrivate
+private import DataFlowNodes
 import SsaImplCommon
 
 private module SourceVariables {
@@ -438,10 +439,7 @@ private predicate sourceVariableHasBaseAndIndex(SourceVariable v, BaseSourceVari
  * initialize `v`.
  */
 private Instruction getInitializationTargetAddress(IRVariable v) {
-  exists(TranslatedVariableInitialization init |
-    init.getIRVariable() = v and
-    result = init.getTargetAddress()
-  )
+  result = IRConstruction::Raw::getInitializationTargetAddress(v)
 }
 
 /** An initial definition of an SSA variable address. */

cpp/ql/lib/semmle/code/cpp/ir/dataflow/internal/SsaImplCommon.qll

@@ -4,75 +4,25 @@ import semmle.code.cpp.ir.internal.IRCppLanguage
 private import semmle.code.cpp.ir.implementation.raw.internal.SideEffects as SideEffects
 private import DataFlowImplCommon as DataFlowImplCommon
 private import DataFlowUtil
+private import DataFlowNodes
 private import semmle.code.cpp.models.interfaces.PointerWrapper
 private import DataFlowPrivate
 private import TypeFlow
 private import semmle.code.cpp.ir.ValueNumbering
 
 /**
- * Holds if `operand` is an operand that is not used by the dataflow library.
- * Ignored operands are not recognized as uses by SSA, and they don't have a
- * corresponding `(Indirect)OperandNode`.
- */
-predicate ignoreOperand(Operand operand) {
-  operand = any(Instruction instr | ignoreInstruction(instr)).getAnOperand() or
-  operand = any(Instruction instr | ignoreInstruction(instr)).getAUse() or
-  operand instanceof MemoryOperand
-}
-
-/**
- * Holds if `instr` is an instruction that is not used by the dataflow library.
- * Ignored instructions are not recognized as reads/writes by SSA, and they
- * don't have a corresponding `(Indirect)InstructionNode`.
- */
-predicate ignoreInstruction(Instruction instr) {
-  DataFlowImplCommon::forceCachingInSameStage() and
-  (
-    instr instanceof CallSideEffectInstruction or
-    instr instanceof CallReadSideEffectInstruction or
-    instr instanceof ExitFunctionInstruction or
-    instr instanceof EnterFunctionInstruction or
-    instr instanceof WriteSideEffectInstruction or
-    instr instanceof PhiInstruction or
-    instr instanceof ReadSideEffectInstruction or
-    instr instanceof ChiInstruction or
-    instr instanceof InitializeIndirectionInstruction or
-    instr instanceof AliasedDefinitionInstruction or
-    instr instanceof AliasedUseInstruction or
-    instr instanceof InitializeNonLocalInstruction or
-    instr instanceof ReturnIndirectionInstruction or
-    instr instanceof UninitializedGroupInstruction
-  )
-}
-
-/**
- * Gets the C++ type of `this` in the member function `f`.
+ * Gets the C++ type of `this` in an `IRFunction` generated from `f`.
  * The result is a glvalue if `isGLValue` is true, and
  * a prvalue if `isGLValue` is false.
  */
 bindingset[isGLValue]
-private CppType getThisType(Cpp::MemberFunction f, boolean isGLValue) {
-  result.hasType(f.getTypeOfThis(), isGLValue)
-}
-
-/**
- * Gets the C++ type of the instruction `i`.
- *
- * This is equivalent to `i.getResultLanguageType()` with the exception
- * of instructions that directly references a `this` IRVariable. In this
- * case, `i.getResultLanguageType()` gives an unknown type, whereas the
- * predicate gives the expected type (i.e., a potentially cv-qualified
- * type `A*` where `A` is the declaring type of the member function that
- * contains `i`).
- */
-cached
-CppType getResultLanguageType(Instruction i) {
-  if i.(VariableAddressInstruction).getIRVariable() instanceof IRThisVariable
-  then
-    if i.isGLValue()
-    then result = getThisType(i.getEnclosingFunction(), true)
-    else result = getThisType(i.getEnclosingFunction(), false)
-  else result = i.getResultLanguageType()
+private CppType getThisType(Cpp::Declaration f, boolean isGLValue) {
+  result.hasType(f.(Cpp::MemberFunction).getTypeOfThis(), isGLValue)
+  or
+  exists(Cpp::PointerType pt |
+    pt.getBaseType() = f.(Cpp::Field).getDeclaringType() and
+    result.hasType(pt, isGLValue)
+  )
 }
 
 /**
@@ -230,7 +180,8 @@ private class PointerWrapperTypeIndirection extends Indirection instanceof Point
   override predicate isAdditionalDereference(Instruction deref, Operand address) {
     exists(CallInstruction call |
       operandForFullyConvertedCall(getAUse(deref), call) and
-      this = call.getStaticCallTarget().getClassAndName(["operator*", "operator->", "get"]) and
+      this =
+        call.getStaticCallTarget().(Function).getClassAndName(["operator*", "operator->", "get"]) and
       address = call.getThisArgumentOperand()
     )
   }
@@ -249,7 +200,7 @@ private module IteratorIndirections {
 
     override predicate isAdditionalWrite(Node0Impl value, Operand address, boolean certain) {
       exists(CallInstruction call | call.getArgumentOperand(0) = value.asOperand() |
-        this = call.getStaticCallTarget().getClassAndName("operator=") and
+        this = call.getStaticCallTarget().(Function).getClassAndName("operator=") and
         address = call.getThisArgumentOperand() and
         certain = false
       )
@@ -347,10 +298,6 @@ predicate isWrite(Node0Impl value, Operand address, boolean certain) {
   )
 }
 
-predicate isAdditionalConversionFlow(Operand opFrom, Instruction instrTo) {
-  any(Indirection ind).isAdditionalConversionFlow(opFrom, instrTo)
-}
-
 newtype TBaseSourceVariable =
   // Each IR variable gets its own source variable
   TBaseIRVariable(IRVariable var) or
@@ -572,6 +519,69 @@ private class BaseCallInstruction extends BaseSourceVariableInstruction, CallIns
 
 cached
 private module Cached {
+  /**
+   * Holds if `operand` is an operand that is not used by the dataflow library.
+   * Ignored operands are not recognized as uses by SSA, and they don't have a
+   * corresponding `(Indirect)OperandNode`.
+   */
+  cached
+  predicate ignoreOperand(Operand operand) {
+    operand = any(Instruction instr | ignoreInstruction(instr)).getAnOperand() or
+    operand = any(Instruction instr | ignoreInstruction(instr)).getAUse() or
+    operand instanceof MemoryOperand
+  }
+
+  /**
+   * Holds if `instr` is an instruction that is not used by the dataflow library.
+   * Ignored instructions are not recognized as reads/writes by SSA, and they
+   * don't have a corresponding `(Indirect)InstructionNode`.
+   */
+  cached
+  predicate ignoreInstruction(Instruction instr) {
+    DataFlowImplCommon::forceCachingInSameStage() and
+    (
+      instr instanceof CallSideEffectInstruction or
+      instr instanceof CallReadSideEffectInstruction or
+      instr instanceof ExitFunctionInstruction or
+      instr instanceof EnterFunctionInstruction or
+      instr instanceof WriteSideEffectInstruction or
+      instr instanceof PhiInstruction or
+      instr instanceof ReadSideEffectInstruction or
+      instr instanceof ChiInstruction or
+      instr instanceof InitializeIndirectionInstruction or
+      instr instanceof AliasedDefinitionInstruction or
+      instr instanceof AliasedUseInstruction or
+      instr instanceof InitializeNonLocalInstruction or
+      instr instanceof ReturnIndirectionInstruction or
+      instr instanceof UninitializedGroupInstruction
+    )
+  }
+
+  cached
+  predicate isAdditionalConversionFlow(Operand opFrom, Instruction instrTo) {
+    any(Indirection ind).isAdditionalConversionFlow(opFrom, instrTo)
+  }
+
+  /**
+   * Gets the C++ type of the instruction `i`.
+   *
+   * This is equivalent to `i.getResultLanguageType()` with the exception
+   * of instructions that directly references a `this` IRVariable. In this
+   * case, `i.getResultLanguageType()` gives an unknown type, whereas the
+   * predicate gives the expected type (i.e., a potentially cv-qualified
+   * type `A*` where `A` is the declaring type of the member function that
+   * contains `i`).
+   */
+  cached
+  CppType getResultLanguageType(Instruction i) {
+    if i.(VariableAddressInstruction).getIRVariable() instanceof IRThisVariable
+    then
+      if i.isGLValue()
+      then result = getThisType(i.getEnclosingFunction(), true)
+      else result = getThisType(i.getEnclosingFunction(), false)
+    else result = i.getResultLanguageType()
+  }
+
   /** Holds if `op` is the only use of its defining instruction, and that op is used in a conversation */
   private predicate isConversion(Operand op) {
     exists(Instruction def, Operand use |

cpp/ql/lib/semmle/code/cpp/ir/dataflow/internal/TaintTrackingUtil.qll

@@ -5,64 +5,81 @@ private import semmle.code.cpp.models.interfaces.DataFlow
 private import semmle.code.cpp.models.interfaces.SideEffect
 private import DataFlowUtil
 private import DataFlowPrivate
+private import DataFlowNodes
 private import SsaImpl as Ssa
 private import semmle.code.cpp.dataflow.internal.FlowSummaryImpl as FlowSummaryImpl
 private import semmle.code.cpp.ir.dataflow.FlowSteps
 
-/**
- * Holds if taint propagates from `nodeFrom` to `nodeTo` in exactly one local
- * (intra-procedural) step. This relation is only used for local taint flow
- * (for example `TaintTracking::localTaint(source, sink)`) so it may contain
- * special cases that should only apply to local taint flow.
- */
-predicate localTaintStep(DataFlow::Node nodeFrom, DataFlow::Node nodeTo) {
-  // dataflow step
-  DataFlow::localFlowStep(nodeFrom, nodeTo)
-  or
-  // taint flow step
-  localAdditionalTaintStep(nodeFrom, nodeTo, _)
-  or
-  // models-as-data summarized flow for local data flow (i.e. special case for flow
-  // through calls to modeled functions, without relying on global dataflow to join
-  // the dots).
-  FlowSummaryImpl::Private::Steps::summaryThroughStepTaint(nodeFrom, nodeTo, _)
+cached
+private module Cached {
+  private import DataFlowImplCommon as DataFlowImplCommon
+
+  /**
+   * This predicate exists to collapse the `cached` predicates in this module with the
+   * `cached` predicates in other C/C++ dataflow files, which is then collapsed
+   * with the `cached` predicates in `DataFlowImplCommon.qll`.
+   */
+  cached
+  predicate forceCachingInSameStage() { DataFlowImplCommon::forceCachingInSameStage() }
+
+  /**
+   * Holds if taint propagates from `nodeFrom` to `nodeTo` in exactly one local
+   * (intra-procedural) step. This relation is only used for local taint flow
+   * (for example `TaintTracking::localTaint(source, sink)`) so it may contain
+   * special cases that should only apply to local taint flow.
+   */
+  cached
+  predicate localTaintStep(DataFlow::Node nodeFrom, DataFlow::Node nodeTo) {
+    // dataflow step
+    DataFlow::localFlowStep(nodeFrom, nodeTo)
+    or
+    // taint flow step
+    localAdditionalTaintStep(nodeFrom, nodeTo, _)
+    or
+    // models-as-data summarized flow for local data flow (i.e. special case for flow
+    // through calls to modeled functions, without relying on global dataflow to join
+    // the dots).
+    FlowSummaryImpl::Private::Steps::summaryThroughStepTaint(nodeFrom, nodeTo, _)
+  }
+
+  /**
+   * Holds if taint can flow in one local step from `nodeFrom` to `nodeTo` excluding
+   * local data flow steps. That is, `nodeFrom` and `nodeTo` are likely to represent
+   * different objects.
+   */
+  cached
+  predicate localAdditionalTaintStep(DataFlow::Node nodeFrom, DataFlow::Node nodeTo, string model) {
+    operandToInstructionTaintStep(nodeFrom.asOperand(), nodeTo.asInstruction()) and
+    model = ""
+    or
+    modeledTaintStep(nodeFrom, nodeTo, model)
+    or
+    // Flow from (the indirection of) an operand of a pointer arithmetic instruction to the
+    // indirection of the pointer arithmetic instruction. This provides flow from `source`
+    // in `x[source]` to the result of the associated load instruction.
+    exists(PointerArithmeticInstruction pai, int indirectionIndex |
+      nodeHasOperand(nodeFrom, pai.getAnOperand(), pragma[only_bind_into](indirectionIndex)) and
+      hasInstructionAndIndex(nodeTo, pai, indirectionIndex + 1)
+    ) and
+    model = ""
+    or
+    any(Ssa::Indirection ind).isAdditionalTaintStep(nodeFrom, nodeTo) and
+    model = ""
+    or
+    // models-as-data summarized flow
+    FlowSummaryImpl::Private::Steps::summaryLocalStep(nodeFrom.(FlowSummaryNode).getSummaryNode(),
+      nodeTo.(FlowSummaryNode).getSummaryNode(), false, model)
+    or
+    // object->field conflation for content that is a `TaintInheritingContent`.
+    exists(DataFlow::ContentSet f |
+      readStep(nodeFrom, f, nodeTo) and
+      f.getAReadContent() instanceof TaintInheritingContent
+    ) and
+    model = ""
+  }
 }
 
-/**
- * Holds if taint can flow in one local step from `nodeFrom` to `nodeTo` excluding
- * local data flow steps. That is, `nodeFrom` and `nodeTo` are likely to represent
- * different objects.
- */
-cached
-predicate localAdditionalTaintStep(DataFlow::Node nodeFrom, DataFlow::Node nodeTo, string model) {
-  operandToInstructionTaintStep(nodeFrom.asOperand(), nodeTo.asInstruction()) and
-  model = ""
-  or
-  modeledTaintStep(nodeFrom, nodeTo, model)
-  or
-  // Flow from (the indirection of) an operand of a pointer arithmetic instruction to the
-  // indirection of the pointer arithmetic instruction. This provides flow from `source`
-  // in `x[source]` to the result of the associated load instruction.
-  exists(PointerArithmeticInstruction pai, int indirectionIndex |
-    nodeHasOperand(nodeFrom, pai.getAnOperand(), pragma[only_bind_into](indirectionIndex)) and
-    hasInstructionAndIndex(nodeTo, pai, indirectionIndex + 1)
-  ) and
-  model = ""
-  or
-  any(Ssa::Indirection ind).isAdditionalTaintStep(nodeFrom, nodeTo) and
-  model = ""
-  or
-  // models-as-data summarized flow
-  FlowSummaryImpl::Private::Steps::summaryLocalStep(nodeFrom.(FlowSummaryNode).getSummaryNode(),
-    nodeTo.(FlowSummaryNode).getSummaryNode(), false, model)
-  or
-  // object->field conflation for content that is a `TaintInheritingContent`.
-  exists(DataFlow::ContentSet f |
-    readStep(nodeFrom, f, nodeTo) and
-    f.getAReadContent() instanceof TaintInheritingContent
-  ) and
-  model = ""
-}
+import Cached
 
 /**
  * Holds if taint propagates from `nodeFrom` to `nodeTo` in exactly one local
@@ -196,7 +213,7 @@ predicate modeledTaintStep(DataFlow::Node nodeIn, DataFlow::Node nodeOut, string
   // Taint flow from a pointer argument to an output, when the model specifies flow from the deref
   // to that output, but the deref is not modeled in the IR for the caller.
   exists(
-    CallInstruction call, DataFlow::SideEffectOperandNode indirectArgument, Function func,
+    CallInstruction call, SideEffectOperandNode indirectArgument, Function func,
     FunctionInput modelIn, FunctionOutput modelOut
   |
     indirectArgument = callInput(call, modelIn) and

cpp/ql/lib/semmle/code/cpp/ir/implementation/aliased_ssa/Instruction.qll

@@ -495,7 +495,7 @@ class FieldInstruction extends Instruction {
  * `FunctionAddress` instruction.
  */
 class FunctionInstruction extends Instruction {
-  Language::Function funcSymbol;
+  Language::Declaration funcSymbol;
 
   FunctionInstruction() { funcSymbol = Raw::getInstructionFunction(this) }
 
@@ -504,7 +504,7 @@ class FunctionInstruction extends Instruction {
   /**
    * Gets the function that this instruction references.
    */
-  final Language::Function getFunctionSymbol() { result = funcSymbol }
+  final Language::Declaration getFunctionSymbol() { result = funcSymbol }
 }
 
 /**
@@ -1678,7 +1678,7 @@ class CallInstruction extends Instruction {
   /**
    * Gets the `Function` that the call targets, if this is statically known.
    */
-  final Language::Function getStaticCallTarget() {
+  final Language::Declaration getStaticCallTarget() {
     result = this.getCallTarget().(FunctionAddressInstruction).getFunctionSymbol()
   }
 

cpp/ql/lib/semmle/code/cpp/ir/implementation/raw/Instruction.qll

@@ -495,7 +495,7 @@ class FieldInstruction extends Instruction {
  * `FunctionAddress` instruction.
  */
 class FunctionInstruction extends Instruction {
-  Language::Function funcSymbol;
+  Language::Declaration funcSymbol;
 
   FunctionInstruction() { funcSymbol = Raw::getInstructionFunction(this) }
 
@@ -504,7 +504,7 @@ class FunctionInstruction extends Instruction {
   /**
    * Gets the function that this instruction references.
    */
-  final Language::Function getFunctionSymbol() { result = funcSymbol }
+  final Language::Declaration getFunctionSymbol() { result = funcSymbol }
 }
 
 /**
@@ -1678,7 +1678,7 @@ class CallInstruction extends Instruction {
   /**
    * Gets the `Function` that the call targets, if this is statically known.
    */
-  final Language::Function getStaticCallTarget() {
+  final Language::Declaration getStaticCallTarget() {
     result = this.getCallTarget().(FunctionAddressInstruction).getFunctionSymbol()
   }
 

cpp/ql/lib/semmle/code/cpp/ir/implementation/raw/internal/IRConstruction.qll

@@ -15,6 +15,8 @@ private import TranslatedCall
 private import TranslatedStmt
 private import TranslatedFunction
 private import TranslatedGlobalVar
+private import TranslatedNonStaticDataMember
+private import TranslatedInitialization
 
 TranslatedElement getInstructionTranslatedElement(Instruction instruction) {
   instruction = TRawInstruction(result, _)
@@ -44,6 +46,9 @@ module Raw {
       or
       not var.isFromUninstantiatedTemplate(_) and
       var instanceof StaticInitializedStaticLocalVariable
+      or
+      not var.isFromUninstantiatedTemplate(_) and
+      var instanceof Field
     ) and
     var.hasInitializer() and
     (
@@ -63,6 +68,8 @@ module Raw {
     getTranslatedFunction(decl).hasUserVariable(var, type)
     or
     getTranslatedVarInit(decl).hasUserVariable(var, type)
+    or
+    getTranslatedFieldInit(decl).hasUserVariable(var, type)
   }
 
   cached
@@ -109,7 +116,7 @@ module Raw {
   }
 
   cached
-  Function getInstructionFunction(Instruction instruction) {
+  Declaration getInstructionFunction(Instruction instruction) {
     result =
       getInstructionTranslatedElement(instruction)
           .getInstructionFunction(getInstructionTag(instruction))
@@ -194,6 +201,89 @@ module Raw {
   Expr getInstructionUnconvertedResultExpression(Instruction instruction) {
     result = getInstructionConvertedResultExpression(instruction).getUnconverted()
   }
+
+  /**
+   * Gets the expression associated with the instruction `instr` that computes
+   * the `Convert` instruction on the extent expression of an allocation.
+   */
+  cached
+  Expr getAllocationExtentConvertExpr(Instruction instr) {
+    exists(TranslatedNonConstantAllocationSize tas |
+      instr = tas.getInstruction(AllocationExtentConvertTag()) and
+      result = tas.getExtent().getExpr()
+    )
+  }
+
+  /**
+   * Gets the `ParenthesisExpr` associated with a transparent conversion
+   * instruction, if any.
+   */
+  cached
+  ParenthesisExpr getTransparentConversionParenthesisExpr(Instruction instr) {
+    exists(TranslatedTransparentConversion ttc |
+      result = ttc.getExpr() and
+      instr = ttc.getResult()
+    )
+  }
+
+  /**
+   * Holds if `instr` belongs to a `TranslatedCoreExpr` that produces an
+   * expression result. This indicates that the instruction represents a
+   * definition whose result should be mapped back to the expression.
+   */
+  cached
+  predicate instructionProducesExprResult(Instruction instr) {
+    exists(TranslatedCoreExpr tco |
+      tco.getInstruction(_) = instr and
+      tco.producesExprResult()
+    )
+  }
+
+  /**
+   * Gets the expression associated with a `StoreInstruction` generated
+   * by an `TranslatedAssignOperation`.
+   */
+  cached
+  Expr getAssignOperationStoreExpr(StoreInstruction store) {
+    exists(TranslatedAssignOperation tao |
+      store = tao.getInstruction(AssignmentStoreTag()) and
+      result = tao.getExpr()
+    )
+  }
+
+  /**
+   * Gets the expression associated with a `StoreInstruction` generated
+   * by an `TranslatedCrementOperation`.
+   */
+  cached
+  Expr getCrementOperationStoreExpr(StoreInstruction store) {
+    exists(TranslatedCrementOperation tco |
+      store = tco.getInstruction(CrementStoreTag()) and
+      result = tco.getExpr()
+    )
+  }
+
+  /**
+   * Holds if `store` is a `StoreInstruction` that defines the temporary
+   * `IRVariable` generated as part of the translation of a ternary expression.
+   */
+  cached
+  predicate isConditionalExprTempStore(StoreInstruction store) {
+    exists(TranslatedConditionalExpr tce |
+      store = tce.getInstruction(ConditionValueFalseStoreTag())
+      or
+      store = tce.getInstruction(ConditionValueTrueStoreTag())
+    )
+  }
+
+  /** Gets the instruction that computes the address used to initialize `v`. */
+  cached
+  Instruction getInitializationTargetAddress(IRVariable v) {
+    exists(TranslatedVariableInitialization init |
+      init.getIRVariable() = v and
+      result = init.getTargetAddress()
+    )
+  }
 }
 
 class TStageInstruction = TRawInstruction or TRawUnreachedInstruction;

cpp/ql/lib/semmle/code/cpp/ir/implementation/raw/internal/SideEffects.qll

@@ -130,27 +130,31 @@ private predicate hasDefaultSideEffect(Call call, ParameterIndex i, boolean buff
 }
 
 /**
- * A `Call` or `NewOrNewArrayExpr` or `DeleteOrDeleteArrayExpr`.
+ * An expression that can have call side effects.
  *
- * All kinds of expression invoke a function as part of their evaluation. This class provides a
- * way to treat both kinds of function similarly, and to get the invoked `Function`.
+ * All kinds of expressions invoke a function as part of their evaluation. This class provides a
+ * way to treat those expressions similarly, and to get the invoked `Declaration`.
  */
-class CallOrAllocationExpr extends Expr {
-  CallOrAllocationExpr() {
+class ExprWithCallSideEffects extends Expr {
+  ExprWithCallSideEffects() {
     this instanceof Call
     or
     this instanceof NewOrNewArrayExpr
     or
     this instanceof DeleteOrDeleteArrayExpr
+    or
+    this instanceof ConstructorDefaultFieldInit
   }
 
-  /** Gets the `Function` invoked by this expression, if known. */
-  final Function getTarget() {
+  /** Gets the `Declaration` invoked by this expression, if known. */
+  final Declaration getTarget() {
     result = this.(Call).getTarget()
     or
     result = this.(NewOrNewArrayExpr).getAllocator()
     or
     result = this.(DeleteOrDeleteArrayExpr).getDeallocator()
+    or
+    result = this.(ConstructorDefaultFieldInit).getTarget()
   }
 }
 
@@ -158,7 +162,7 @@ class CallOrAllocationExpr extends Expr {
  * Returns the side effect opcode, if any, that represents any side effects not specifically modeled
  * by an argument side effect.
  */
-Opcode getCallSideEffectOpcode(CallOrAllocationExpr expr) {
+Opcode getCallSideEffectOpcode(ExprWithCallSideEffects expr) {
   not exists(expr.getTarget().(SideEffectFunction)) and result instanceof Opcode::CallSideEffect
   or
   exists(SideEffectFunction sideEffectFunction |
@@ -175,7 +179,7 @@ Opcode getCallSideEffectOpcode(CallOrAllocationExpr expr) {
 /**
  * Returns a side effect opcode for parameter index `i` of the specified call.
  *
- * This predicate will return at most two results: one read side effect, and one write side effect.
+ * This predicate will yield at most two results: one read side effect, and one write side effect.
  */
 Opcode getASideEffectOpcode(Call call, ParameterIndex i) {
   exists(boolean buffer |
@@ -228,3 +232,14 @@ Opcode getASideEffectOpcode(Call call, ParameterIndex i) {
     )
   )
 }
+
+/**
+ * Returns a side effect opcode for a default field initialization.
+ *
+ * This predicate will yield two results: one read side effect, and one write side effect.
+ */
+Opcode getDefaultFieldInitSideEffectOpcode() {
+  result instanceof Opcode::IndirectReadSideEffect
+  or
+  result instanceof Opcode::IndirectMayWriteSideEffect
+}

cpp/ql/lib/semmle/code/cpp/ir/implementation/raw/internal/TranslatedCall.qll

@@ -10,6 +10,7 @@ private import SideEffects
 private import TranslatedElement
 private import TranslatedExpr
 private import TranslatedFunction
+private import TranslatedInitialization
 private import DefaultOptions as DefaultOptions
 
 /**
@@ -348,7 +349,7 @@ class TranslatedExprCall extends TranslatedCallExpr {
 class TranslatedFunctionCall extends TranslatedCallExpr, TranslatedDirectCall {
   override FunctionCall expr;
 
-  override Function getInstructionFunction(InstructionTag tag) {
+  override Declaration getInstructionFunction(InstructionTag tag) {
     tag = CallTargetTag() and result = expr.getTarget()
   }
 
@@ -429,6 +430,9 @@ class TranslatedCallSideEffects extends TranslatedSideEffects, TTranslatedCallSi
     or
     expr instanceof DeleteOrDeleteArrayExpr and
     result = getTranslatedDeleteOrDeleteArray(expr).getInstruction(CallTag())
+    or
+    expr instanceof ConstructorDefaultFieldInit and
+    result = getTranslatedConstructorFieldInitialization(expr).getInstruction(CallTag())
   }
 }
 
@@ -504,11 +508,25 @@ abstract class TranslatedSideEffect extends TranslatedElement {
   abstract predicate sideEffectInstruction(Opcode opcode, CppType type);
 }
 
+private class CallOrDefaultFieldInit extends Expr {
+  CallOrDefaultFieldInit() {
+    this instanceof Call
+    or
+    this instanceof ConstructorDefaultFieldInit
+  }
+
+  Declaration getTarget() {
+    result = this.(Call).getTarget()
+    or
+    result = this.(ConstructorDefaultFieldInit).getTarget()
+  }
+}
+
 /**
  * The IR translation of a single argument side effect for a call.
  */
 abstract class TranslatedArgumentSideEffect extends TranslatedSideEffect {
-  Call call;
+  CallOrDefaultFieldInit callOrInit;
   int index;
   SideEffectOpcode sideEffectOpcode;
 
@@ -524,7 +542,7 @@ abstract class TranslatedArgumentSideEffect extends TranslatedSideEffect {
     result = "(read side effect for " + this.getArgString() + ")"
   }
 
-  override Call getPrimaryExpr() { result = call }
+  override Expr getPrimaryExpr() { result = callOrInit }
 
   override predicate sortOrder(int group, int indexInGroup) {
     indexInGroup = index and
@@ -586,9 +604,10 @@ abstract class TranslatedArgumentSideEffect extends TranslatedSideEffect {
     tag instanceof OnlyInstructionTag and
     operandTag instanceof BufferSizeOperandTag and
     result =
-      getTranslatedExpr(call.getArgument(call.getTarget()
-              .(SideEffectFunction)
-              .getParameterSizeIndex(index)).getFullyConverted()).getResult()
+      getTranslatedExpr(callOrInit
+            .(Call)
+            .getArgument(callOrInit.getTarget().(SideEffectFunction).getParameterSizeIndex(index))
+            .getFullyConverted()).getResult()
   }
 
   /** Holds if this side effect is a write side effect, rather than a read side effect. */
@@ -616,7 +635,7 @@ class TranslatedArgumentExprSideEffect extends TranslatedArgumentSideEffect,
   Expr arg;
 
   TranslatedArgumentExprSideEffect() {
-    this = TTranslatedArgumentExprSideEffect(call, arg, index, sideEffectOpcode)
+    this = TTranslatedArgumentExprSideEffect(callOrInit, arg, index, sideEffectOpcode)
   }
 
   final override Locatable getAst() { result = arg }
@@ -640,28 +659,31 @@ class TranslatedArgumentExprSideEffect extends TranslatedArgumentSideEffect,
  * The IR translation of an argument side effect for `*this` on a call, where there is no `Expr`
  * object that represents the `this` argument.
  *
- * The applies only to constructor calls, as the AST has exploit qualifier `Expr`s for all other
- * calls to non-static member functions.
+ * This applies to constructor calls and default field initializations, as the AST has explicit
+ * qualifier `Expr`s for all other calls to non-static member functions.
  */
-class TranslatedStructorQualifierSideEffect extends TranslatedArgumentSideEffect,
-  TTranslatedStructorQualifierSideEffect
+class TranslatedImplicitThisQualifierSideEffect extends TranslatedArgumentSideEffect,
+  TTranslatedImplicitThisQualifierSideEffect
 {
-  TranslatedStructorQualifierSideEffect() {
-    this = TTranslatedStructorQualifierSideEffect(call, sideEffectOpcode) and
+  TranslatedImplicitThisQualifierSideEffect() {
+    this = TTranslatedImplicitThisQualifierSideEffect(callOrInit, sideEffectOpcode) and
     index = -1
   }
 
-  final override Locatable getAst() { result = call }
+  final override Locatable getAst() { result = callOrInit }
 
-  final override Type getIndirectionType() { result = call.getTarget().getDeclaringType() }
+  final override Type getIndirectionType() { result = callOrInit.getTarget().getDeclaringType() }
 
   final override string getArgString() { result = "this" }
 
   final override Instruction getArgInstruction() {
     exists(TranslatedStructorCall structorCall |
-      structorCall.getExpr() = call and
+      structorCall.getExpr() = callOrInit and
       result = structorCall.getQualifierResult()
     )
+    or
+    callOrInit instanceof ConstructorDefaultFieldInit and
+    result = getTranslatedFunction(callOrInit.getEnclosingFunction()).getLoadThisInstruction()
   }
 }
 

cpp/ql/lib/semmle/code/cpp/ir/implementation/raw/internal/TranslatedCondition.qll

@@ -36,7 +36,8 @@ abstract class TranslatedCondition extends TranslatedElement {
   final override Declaration getFunction() {
     result = getEnclosingFunction(expr) or
     result = getEnclosingVariable(expr).(GlobalOrNamespaceVariable) or
-    result = getEnclosingVariable(expr).(StaticInitializedStaticLocalVariable)
+    result = getEnclosingVariable(expr).(StaticInitializedStaticLocalVariable) or
+    result = getEnclosingVariable(expr).(Field)
   }
 
   final Type getResultType() { result = expr.getUnspecifiedType() }

cpp/ql/lib/semmle/code/cpp/ir/implementation/raw/internal/TranslatedDeclarationEntry.qll

@@ -34,8 +34,11 @@ abstract class TranslatedDeclarationEntry extends TranslatedElement, TTranslated
       or
       result = entry.getDeclaration().(GlobalOrNamespaceVariable)
       or
+      result = entry.getDeclaration().(Field)
+      or
       not entry.getDeclaration() instanceof StaticInitializedStaticLocalVariable and
       not entry.getDeclaration() instanceof GlobalOrNamespaceVariable and
+      not entry.getDeclaration() instanceof Field and
       result = stmt.getEnclosingFunction()
     )
   }

cpp/ql/lib/semmle/code/cpp/ir/implementation/raw/internal/TranslatedElement.qll

@@ -767,7 +767,7 @@ newtype TTranslatedElement =
       expr = initList.getFieldExpr(field, position).getFullyConverted()
     )
     or
-    exists(ConstructorFieldInit init |
+    exists(ConstructorDirectFieldInit init |
       not ignoreExpr(init) and
       ast = init and
       field = init.getTarget() and
@@ -775,6 +775,14 @@ newtype TTranslatedElement =
       position = -1
     )
   } or
+  // The initialization of a field via a default member initializer.
+  TTranslatedDefaultFieldInitialization(Expr ast, Field field) {
+    exists(ConstructorDefaultFieldInit init |
+      not ignoreExpr(init) and
+      ast = init and
+      field = init.getTarget()
+    )
+  } or
   // The value initialization of a field due to an omitted member of an
   // initializer list.
   TTranslatedFieldValueInitialization(Expr ast, Field field) {
@@ -871,7 +879,7 @@ newtype TTranslatedElement =
   // The declaration/initialization part of a `ConditionDeclExpr`
   TTranslatedConditionDecl(ConditionDeclExpr expr) { not ignoreExpr(expr) } or
   // The side effects of a `Call`
-  TTranslatedCallSideEffects(CallOrAllocationExpr expr) {
+  TTranslatedCallSideEffects(ExprWithCallSideEffects expr) {
     not ignoreExpr(expr) and
     not ignoreSideEffects(expr)
   } or
@@ -910,15 +918,23 @@ newtype TTranslatedElement =
   } or
   // Constructor calls lack a qualifier (`this`) expression, so we need to handle the side effects
   // on `*this` without an `Expr`.
-  TTranslatedStructorQualifierSideEffect(Call call, SideEffectOpcode opcode) {
+  TTranslatedImplicitThisQualifierSideEffect(ExprWithCallSideEffects call, SideEffectOpcode opcode) {
     not ignoreExpr(call) and
     not ignoreSideEffects(call) and
-    call instanceof ConstructorCall and
-    opcode = getASideEffectOpcode(call, -1)
+    (
+      call instanceof ConstructorCall and
+      opcode = getASideEffectOpcode(call, -1)
+      or
+      call instanceof ConstructorFieldInit and
+      opcode = getDefaultFieldInitSideEffectOpcode()
+    )
   } or
   // The side effect that initializes newly-allocated memory.
   TTranslatedAllocationSideEffect(AllocationExpr expr) { not ignoreSideEffects(expr) } or
-  TTranslatedStaticStorageDurationVarInit(Variable var) { Raw::varHasIRFunc(var) } or
+  TTranslatedStaticStorageDurationVarInit(Variable var) {
+    Raw::varHasIRFunc(var) and not var instanceof Field
+  } or
+  TTranslatedNonStaticDataMemberVarInit(Field var) { Raw::varHasIRFunc(var) } or
   TTranslatedAssertionOperand(MacroInvocation mi, int index) { hasAssertionOperand(mi, index) }
 
 /**
@@ -1179,7 +1195,7 @@ abstract class TranslatedElement extends TTranslatedElement {
    * If the instruction specified by `tag` is a `FunctionInstruction`, gets the
    * `Function` for that instruction.
    */
-  Function getInstructionFunction(InstructionTag tag) { none() }
+  Declaration getInstructionFunction(InstructionTag tag) { none() }
 
   /**
    * If the instruction specified by `tag` is a `VariableInstruction`, gets the
@@ -1297,5 +1313,7 @@ abstract class TranslatedRootElement extends TranslatedElement {
     this instanceof TTranslatedFunction
     or
     this instanceof TTranslatedStaticStorageDurationVarInit
+    or
+    this instanceof TTranslatedNonStaticDataMemberVarInit
   }
 }

cpp/ql/lib/semmle/code/cpp/ir/implementation/raw/internal/TranslatedExpr.qll

@@ -14,6 +14,7 @@ private import TranslatedFunction
 private import TranslatedInitialization
 private import TranslatedStmt
 private import TranslatedGlobalVar
+private import TranslatedNonStaticDataMember
 private import IRConstruction
 import TranslatedCall
 
@@ -138,6 +139,8 @@ abstract class TranslatedExpr extends TranslatedElement {
     result = getTranslatedFunction(getEnclosingFunction(expr))
     or
     result = getTranslatedVarInit(getEnclosingVariable(expr))
+    or
+    result = getTranslatedFieldInit(getEnclosingVariable(expr))
   }
 }
 
@@ -153,7 +156,10 @@ Declaration getEnclosingDeclaration0(Expr e) {
     i.getExpr().getFullyConverted() = e and
     v = i.getDeclaration()
   |
-    if v instanceof StaticInitializedStaticLocalVariable or v instanceof GlobalOrNamespaceVariable
+    if
+      v instanceof StaticInitializedStaticLocalVariable or
+      v instanceof GlobalOrNamespaceVariable or
+      v instanceof Field
     then result = v
     else result = e.getEnclosingDeclaration()
   )
@@ -173,7 +179,10 @@ Variable getEnclosingVariable0(Expr e) {
     i.getExpr().getFullyConverted() = e and
     v = i.getDeclaration()
   |
-    if v instanceof StaticInitializedStaticLocalVariable or v instanceof GlobalOrNamespaceVariable
+    if
+      v instanceof StaticInitializedStaticLocalVariable or
+      v instanceof GlobalOrNamespaceVariable or
+      v instanceof Field
     then result = v
     else result = e.getEnclosingVariable()
   )
@@ -826,6 +835,46 @@ class TranslatedPostfixCrementOperation extends TranslatedCrementOperation {
   override Instruction getResult() { result = this.getLoadedOperand().getResult() }
 }
 
+class TranslatedParamAccessForType extends TranslatedNonConstantExpr {
+  override ParamAccessForType expr;
+
+  TranslatedParamAccessForType() {
+    // Currently only needed for this parameter accesses.
+    expr.isThisAccess()
+  }
+
+  final override Instruction getFirstInstruction(EdgeKind kind) {
+    result = this.getInstruction(OnlyInstructionTag()) and
+    kind instanceof GotoEdge
+  }
+
+  override Instruction getALastInstructionInternal() {
+    result = this.getInstruction(OnlyInstructionTag())
+  }
+
+  final override TranslatedElement getChildInternal(int id) { none() }
+
+  override Instruction getInstructionSuccessorInternal(InstructionTag tag, EdgeKind kind) {
+    tag = OnlyInstructionTag() and
+    result = this.getParent().getChildSuccessor(this, kind)
+  }
+
+  override Instruction getResult() { result = this.getInstruction(OnlyInstructionTag()) }
+
+  override predicate hasInstruction(Opcode opcode, InstructionTag tag, CppType resultType) {
+    tag = OnlyInstructionTag() and
+    opcode instanceof Opcode::CopyValue and
+    resultType = getTypeForPRValue(expr.getType())
+  }
+
+  override Instruction getInstructionRegisterOperand(InstructionTag tag, OperandTag operandTag) {
+    tag = OnlyInstructionTag() and
+    operandTag instanceof UnaryOperandTag and
+    result =
+      this.getEnclosingFunction().(TranslatedNonStaticDataMemberVarInit).getLoadThisInstruction()
+  }
+}
+
 /**
  * IR translation of an array access expression (e.g. `a[i]`). The array being accessed will either
  * be a prvalue of pointer type (possibly due to an implicit array-to-pointer conversion), or a
@@ -1215,7 +1264,7 @@ class TranslatedFunctionAccess extends TranslatedNonConstantExpr {
     resultType = this.getResultType()
   }
 
-  override Function getInstructionFunction(InstructionTag tag) {
+  override Declaration getInstructionFunction(InstructionTag tag) {
     tag = OnlyInstructionTag() and
     result = expr.getTarget()
   }
@@ -2498,7 +2547,7 @@ class TranslatedAllocatorCall extends TTranslatedAllocatorCall, TranslatedDirect
     any()
   }
 
-  override Function getInstructionFunction(InstructionTag tag) {
+  override Declaration getInstructionFunction(InstructionTag tag) {
     tag = CallTargetTag() and result = expr.getAllocator()
   }
 
@@ -2581,7 +2630,7 @@ class TranslatedDeleteOrDeleteArrayExpr extends TranslatedNonConstantExpr, Trans
     result = this.getFirstArgumentOrCallInstruction(kind)
   }
 
-  override Function getInstructionFunction(InstructionTag tag) {
+  override Declaration getInstructionFunction(InstructionTag tag) {
     tag = CallTargetTag() and result = expr.getDeallocator()
   }
 

cpp/ql/lib/semmle/code/cpp/ir/implementation/raw/internal/TranslatedInitialization.qll

@@ -148,7 +148,8 @@ abstract class TranslatedInitialization extends TranslatedElement, TTranslatedIn
   final override Declaration getFunction() {
     result = getEnclosingFunction(expr) or
     result = getEnclosingVariable(expr).(GlobalOrNamespaceVariable) or
-    result = getEnclosingVariable(expr).(StaticInitializedStaticLocalVariable)
+    result = getEnclosingVariable(expr).(StaticInitializedStaticLocalVariable) or
+    result = getEnclosingVariable(expr).(Field)
   }
 
   final override Locatable getAst() { result = expr }
@@ -514,8 +515,8 @@ TranslatedFieldInitialization getTranslatedConstructorFieldInitialization(Constr
 }
 
 /**
- * Represents the IR translation of the initialization of a field from an
- * element of an initializer list.
+ * The IR translation of the initialization of a field from an element of
+ * an initializer list.
  */
 abstract class TranslatedFieldInitialization extends TranslatedElement {
   Expr ast;
@@ -528,13 +529,11 @@ abstract class TranslatedFieldInitialization extends TranslatedElement {
   final override Declaration getFunction() {
     result = getEnclosingFunction(ast) or
     result = getEnclosingVariable(ast).(GlobalOrNamespaceVariable) or
-    result = getEnclosingVariable(ast).(StaticInitializedStaticLocalVariable)
+    result = getEnclosingVariable(ast).(StaticInitializedStaticLocalVariable) or
+    result = getEnclosingVariable(ast).(Field)
   }
 
-  final override Instruction getFirstInstruction(EdgeKind kind) {
-    result = this.getInstruction(this.getFieldAddressTag()) and
-    kind instanceof GotoEdge
-  }
+  final Field getField() { result = field }
 
   /**
    * Gets the zero-based index describing the order in which this field is to be
@@ -542,6 +541,20 @@ abstract class TranslatedFieldInitialization extends TranslatedElement {
    */
   final int getOrder() { result = field.getInitializationOrder() }
 
+  /** Gets the position in the initializer list, or `-1` if the initialization is implicit. */
+  int getPosition() { result = -1 }
+}
+
+/**
+ * The IR translation of the initialization of a field from an element of an initializer
+ * list where default initialization is not used.
+ */
+abstract class TranslatedNonDefaultFieldInitialization extends TranslatedFieldInitialization {
+  final override Instruction getFirstInstruction(EdgeKind kind) {
+    result = this.getInstruction(this.getFieldAddressTag()) and
+    kind instanceof GotoEdge
+  }
+
   override predicate hasInstruction(Opcode opcode, InstructionTag tag, CppType resultType) {
     tag = this.getFieldAddressTag() and
     opcode instanceof Opcode::FieldAddress and
@@ -559,18 +572,13 @@ abstract class TranslatedFieldInitialization extends TranslatedElement {
   }
 
   final InstructionTag getFieldAddressTag() { result = InitializerFieldAddressTag() }
-
-  final Field getField() { result = field }
-
-  /** Gets the position in the initializer list, or `-1` if the initialization is implicit. */
-  int getPosition() { result = -1 }
 }
 
 /**
- * Represents the IR translation of the initialization of a field from an
- * explicit element in an initializer list.
+ * The IR translation of the initialization of a field from an explicit element in
+ * an initializer list.
  */
-class TranslatedExplicitFieldInitialization extends TranslatedFieldInitialization,
+class TranslatedExplicitFieldInitialization extends TranslatedNonDefaultFieldInitialization,
   InitializationContext, TTranslatedExplicitFieldInitialization
 {
   Expr expr;
@@ -610,15 +618,81 @@ class TranslatedExplicitFieldInitialization extends TranslatedFieldInitializatio
   override int getPosition() { result = position }
 }
 
+/**
+ * The IR translation of the initialization of a field from an element of an initializer
+ * list where default initialization is used.
+ */
+class TranslatedDefaultFieldInitialization extends TranslatedFieldInitialization,
+  TTranslatedDefaultFieldInitialization
+{
+  TranslatedDefaultFieldInitialization() {
+    this = TTranslatedDefaultFieldInitialization(ast, field)
+  }
+
+  final override Instruction getFirstInstruction(EdgeKind kind) {
+    result = this.getInstruction(CallTargetTag()) and
+    kind instanceof GotoEdge
+  }
+
+  override Instruction getALastInstructionInternal() {
+    result = this.getSideEffects().getALastInstruction()
+  }
+
+  override TranslatedElement getLastChild() { result = this.getSideEffects() }
+
+  override Instruction getInstructionSuccessorInternal(InstructionTag tag, EdgeKind kind) {
+    tag = CallTargetTag() and
+    result = this.getInstruction(CallTag())
+    or
+    tag = CallTag() and
+    result = this.getSideEffects().getFirstInstruction(kind)
+  }
+
+  override Instruction getChildSuccessorInternal(TranslatedElement child, EdgeKind kind) {
+    child = this.getSideEffects() and
+    result = this.getParent().getChildSuccessor(this, kind)
+  }
+
+  override predicate hasInstruction(Opcode opcode, InstructionTag tag, CppType resultType) {
+    tag = CallTargetTag() and
+    opcode instanceof Opcode::FunctionAddress and
+    resultType = getFunctionGLValueType()
+    or
+    tag = CallTag() and
+    opcode instanceof Opcode::Call and
+    resultType = getVoidType()
+  }
+
+  override Instruction getInstructionRegisterOperand(InstructionTag tag, OperandTag operandTag) {
+    tag = CallTag() and
+    (
+      operandTag instanceof CallTargetOperandTag and
+      result = this.getInstruction(CallTargetTag())
+      or
+      operandTag instanceof ThisArgumentOperandTag and
+      result = getTranslatedFunction(this.getFunction()).getLoadThisInstruction()
+    )
+  }
+
+  override Declaration getInstructionFunction(InstructionTag tag) {
+    tag = CallTargetTag() and
+    result = field
+  }
+
+  override TranslatedElement getChild(int id) { id = 0 and result = this.getSideEffects() }
+
+  final TranslatedSideEffects getSideEffects() { result.getExpr() = ast }
+}
+
 private string getZeroValue(Type type) {
   if type instanceof FloatingPointType then result = "0.0" else result = "0"
 }
 
 /**
- * Represents the IR translation of the initialization of a field without a
- * corresponding element in the initializer list.
+ * The IR translation of the initialization of a field without a corresponding
+ * element in the initializer list.
  */
-class TranslatedFieldValueInitialization extends TranslatedFieldInitialization,
+class TranslatedFieldValueInitialization extends TranslatedNonDefaultFieldInitialization,
   TTranslatedFieldValueInitialization
 {
   TranslatedFieldValueInitialization() { this = TTranslatedFieldValueInitialization(ast, field) }
@@ -628,7 +702,7 @@ class TranslatedFieldValueInitialization extends TranslatedFieldInitialization,
   }
 
   override predicate hasInstruction(Opcode opcode, InstructionTag tag, CppType resultType) {
-    TranslatedFieldInitialization.super.hasInstruction(opcode, tag, resultType)
+    TranslatedNonDefaultFieldInitialization.super.hasInstruction(opcode, tag, resultType)
     or
     tag = this.getFieldDefaultValueTag() and
     opcode instanceof Opcode::Constant and
@@ -659,7 +733,8 @@ class TranslatedFieldValueInitialization extends TranslatedFieldInitialization,
   }
 
   override Instruction getInstructionRegisterOperand(InstructionTag tag, OperandTag operandTag) {
-    result = TranslatedFieldInitialization.super.getInstructionRegisterOperand(tag, operandTag)
+    result =
+      TranslatedNonDefaultFieldInitialization.super.getInstructionRegisterOperand(tag, operandTag)
     or
     tag = this.getFieldDefaultValueStoreTag() and
     (
@@ -683,8 +758,8 @@ class TranslatedFieldValueInitialization extends TranslatedFieldInitialization,
 }
 
 /**
- * Represents the IR translation of the initialization of an array element from
- * an element of an initializer list.
+ * The IR translation of the initialization of an array element from an element
+ * of an initializer list.
  */
 abstract class TranslatedElementInitialization extends TranslatedElement {
   ArrayOrVectorAggregateLiteral initList;
@@ -701,6 +776,8 @@ abstract class TranslatedElementInitialization extends TranslatedElement {
     result = getEnclosingVariable(initList).(GlobalOrNamespaceVariable)
     or
     result = getEnclosingVariable(initList).(StaticInitializedStaticLocalVariable)
+    or
+    result = getEnclosingVariable(initList).(Field)
   }
 
   final override Instruction getFirstInstruction(EdgeKind kind) {
@@ -759,8 +836,8 @@ abstract class TranslatedElementInitialization extends TranslatedElement {
 }
 
 /**
- * Represents the IR translation of the initialization of an array element from
- * an explicit element in an initializer list.
+ * The IR translation of the initialization of an array element from an explicit
+ * element in an initializer list.
  */
 class TranslatedExplicitElementInitialization extends TranslatedElementInitialization,
   TTranslatedExplicitElementInitialization, InitializationContext
@@ -808,8 +885,8 @@ class TranslatedExplicitElementInitialization extends TranslatedElementInitializ
 }
 
 /**
- * Represents the IR translation of the initialization of a range of array
- * elements without corresponding elements in the initializer list.
+ * The IR translation of the initialization of a range of array elements without
+ * corresponding elements in the initializer list.
  */
 class TranslatedElementValueInitialization extends TranslatedElementInitialization,
   TTranslatedElementValueInitialization

cpp/ql/lib/semmle/code/cpp/ir/implementation/raw/internal/TranslatedNonStaticDataMember.qll

@@ -0,0 +1,217 @@
+import semmle.code.cpp.ir.implementation.raw.internal.TranslatedElement
+private import TranslatedExpr
+private import cpp
+private import semmle.code.cpp.ir.implementation.internal.OperandTag
+private import semmle.code.cpp.ir.internal.TempVariableTag
+private import semmle.code.cpp.ir.internal.CppType
+private import TranslatedInitialization
+private import InstructionTag
+private import semmle.code.cpp.ir.internal.IRUtilities
+
+class TranslatedNonStaticDataMemberVarInit extends TranslatedRootElement,
+  TTranslatedNonStaticDataMemberVarInit, InitializationContext
+{
+  Field field;
+  Class cls;
+
+  TranslatedNonStaticDataMemberVarInit() {
+    this = TTranslatedNonStaticDataMemberVarInit(field) and
+    cls.getAMember() = field
+  }
+
+  override string toString() { result = cls.toString() + "::" + field.toString() }
+
+  final override Field getAst() { result = field }
+
+  final override Declaration getFunction() { result = field }
+
+  override Instruction getFirstInstruction(EdgeKind kind) {
+    result = this.getInstruction(EnterFunctionTag()) and
+    kind instanceof GotoEdge
+  }
+
+  override Instruction getALastInstructionInternal() {
+    result = this.getInstruction(ExitFunctionTag())
+  }
+
+  override TranslatedElement getChild(int n) {
+    n = 1 and
+    result = getTranslatedInitialization(field.getInitializer().getExpr().getFullyConverted())
+  }
+
+  override predicate hasInstruction(Opcode op, InstructionTag tag, CppType type) {
+    op instanceof Opcode::EnterFunction and
+    tag = EnterFunctionTag() and
+    type = getVoidType()
+    or
+    op instanceof Opcode::AliasedDefinition and
+    tag = AliasedDefinitionTag() and
+    type = getUnknownType()
+    or
+    op instanceof Opcode::InitializeNonLocal and
+    tag = InitializeNonLocalTag() and
+    type = getUnknownType()
+    or
+    tag = ThisAddressTag() and
+    op instanceof Opcode::VariableAddress and
+    type = getTypeForGLValue(any(UnknownType t))
+    or
+    tag = InitializerStoreTag() and
+    op instanceof Opcode::InitializeParameter and
+    type = this.getThisType()
+    or
+    tag = ThisLoadTag() and
+    op instanceof Opcode::Load and
+    type = this.getThisType()
+    or
+    tag = InitializerIndirectStoreTag() and
+    op instanceof Opcode::InitializeIndirection and
+    type = getTypeForPRValue(cls)
+    or
+    op instanceof Opcode::FieldAddress and
+    tag = InitializerFieldAddressTag() and
+    type = getTypeForGLValue(field.getType())
+    or
+    op instanceof Opcode::ReturnVoid and
+    tag = ReturnTag() and
+    type = getVoidType()
+    or
+    op instanceof Opcode::AliasedUse and
+    tag = AliasedUseTag() and
+    type = getVoidType()
+    or
+    op instanceof Opcode::ExitFunction and
+    tag = ExitFunctionTag() and
+    type = getVoidType()
+  }
+
+  override Instruction getInstructionSuccessorInternal(InstructionTag tag, EdgeKind kind) {
+    kind instanceof GotoEdge and
+    (
+      tag = EnterFunctionTag() and
+      result = this.getInstruction(AliasedDefinitionTag())
+      or
+      tag = AliasedDefinitionTag() and
+      result = this.getInstruction(InitializeNonLocalTag())
+      or
+      tag = InitializeNonLocalTag() and
+      result = this.getInstruction(ThisAddressTag())
+      or
+      tag = ThisAddressTag() and
+      result = this.getInstruction(InitializerStoreTag())
+      or
+      tag = InitializerStoreTag() and
+      result = this.getInstruction(ThisLoadTag())
+      or
+      tag = ThisLoadTag() and
+      result = this.getInstruction(InitializerIndirectStoreTag())
+      or
+      tag = InitializerIndirectStoreTag() and
+      result = this.getInstruction(InitializerFieldAddressTag())
+    )
+    or
+    tag = InitializerFieldAddressTag() and
+    result = this.getChild(1).getFirstInstruction(kind)
+    or
+    kind instanceof GotoEdge and
+    (
+      tag = ReturnTag() and
+      result = this.getInstruction(AliasedUseTag())
+      or
+      tag = AliasedUseTag() and
+      result = this.getInstruction(ExitFunctionTag())
+    )
+  }
+
+  override Instruction getChildSuccessorInternal(TranslatedElement child, EdgeKind kind) {
+    child = this.getChild(1) and
+    result = this.getInstruction(ReturnTag()) and
+    kind instanceof GotoEdge
+  }
+
+  final override CppType getInstructionMemoryOperandType(
+    InstructionTag tag, TypedOperandTag operandTag
+  ) {
+    tag = AliasedUseTag() and
+    operandTag instanceof SideEffectOperandTag and
+    result = getUnknownType()
+  }
+
+  override IRVariable getInstructionVariable(InstructionTag tag) {
+    (
+      tag = ThisAddressTag() or
+      tag = InitializerStoreTag() or
+      tag = InitializerIndirectStoreTag()
+    ) and
+    result = getIRTempVariable(field, ThisTempVar())
+  }
+
+  override Field getInstructionField(InstructionTag tag) {
+    tag = InitializerFieldAddressTag() and
+    result = field
+  }
+
+  override predicate hasTempVariable(TempVariableTag tag, CppType type) {
+    tag = ThisTempVar() and
+    type = this.getThisType()
+  }
+
+  /**
+   * Holds if this variable defines or accesses variable `var` with type `type`. This includes all
+   * parameters and local variables, plus any global variables or static data members that are
+   * directly accessed by the function.
+   */
+  final predicate hasUserVariable(Variable varUsed, CppType type) {
+    (
+      (
+        varUsed instanceof GlobalOrNamespaceVariable
+        or
+        varUsed instanceof StaticLocalVariable
+        or
+        varUsed instanceof MemberVariable and not varUsed instanceof Field
+      ) and
+      exists(VariableAccess access |
+        access.getTarget() = varUsed and
+        getEnclosingVariable(access) = field
+      )
+      or
+      field = varUsed
+      or
+      varUsed.(LocalScopeVariable).getEnclosingElement*() = field
+      or
+      varUsed.(Parameter).getCatchBlock().getEnclosingElement*() = field
+    ) and
+    type = getTypeForPRValue(getVariableType(varUsed))
+  }
+
+  override Instruction getInstructionRegisterOperand(InstructionTag tag, OperandTag operandTag) {
+    (
+      tag = InitializerStoreTag()
+      or
+      tag = ThisLoadTag()
+    ) and
+    operandTag instanceof AddressOperandTag and
+    result = this.getInstruction(ThisAddressTag())
+    or
+    (
+      tag = InitializerIndirectStoreTag() and
+      operandTag instanceof AddressOperandTag
+      or
+      tag = InitializerFieldAddressTag() and
+      operandTag instanceof UnaryOperandTag
+    ) and
+    result = this.getInstruction(ThisLoadTag())
+  }
+
+  override Instruction getTargetAddress() {
+    result = this.getInstruction(InitializerFieldAddressTag())
+  }
+
+  override Type getTargetType() { result = field.getUnspecifiedType() }
+
+  final Instruction getLoadThisInstruction() { result = this.getInstruction(ThisLoadTag()) }
+
+  private CppType getThisType() { result = getTypeForGLValue(cls) }
+}
+
+TranslatedNonStaticDataMemberVarInit getTranslatedFieldInit(Field field) { result.getAst() = field }

cpp/ql/lib/semmle/code/cpp/ir/implementation/unaliased_ssa/Instruction.qll

@@ -495,7 +495,7 @@ class FieldInstruction extends Instruction {
  * `FunctionAddress` instruction.
  */
 class FunctionInstruction extends Instruction {
-  Language::Function funcSymbol;
+  Language::Declaration funcSymbol;
 
   FunctionInstruction() { funcSymbol = Raw::getInstructionFunction(this) }
 
@@ -504,7 +504,7 @@ class FunctionInstruction extends Instruction {
   /**
    * Gets the function that this instruction references.
    */
-  final Language::Function getFunctionSymbol() { result = funcSymbol }
+  final Language::Declaration getFunctionSymbol() { result = funcSymbol }
 }
 
 /**
@@ -1678,7 +1678,7 @@ class CallInstruction extends Instruction {
   /**
    * Gets the `Function` that the call targets, if this is statically known.
    */
-  final Language::Function getStaticCallTarget() {
+  final Language::Declaration getStaticCallTarget() {
     result = this.getCallTarget().(FunctionAddressInstruction).getFunctionSymbol()
   }
 

cpp/ql/lib/semmle/code/cpp/models/Models.qll

@@ -48,7 +48,6 @@ private import implementations.SqLite3
 private import implementations.PostgreSql
 private import implementations.System
 private import implementations.StructuredExceptionHandling
-private import implementations.ZMQ
 private import implementations.Win32CommandExecution
 private import implementations.CA2AEX
 private import implementations.CComBSTR
@@ -58,3 +57,4 @@ private import implementations.CAtlFileMapping
 private import implementations.CAtlTemporaryFile
 private import implementations.CRegKey
 private import implementations.WinHttp
+private import implementations.Http

cpp/ql/lib/semmle/code/cpp/models/implementations/Gets.qll

@@ -112,21 +112,3 @@ private class GetsFunction extends DataFlowFunction, ArrayFunction, AliasFunctio
 
   override predicate hasArrayOutput(int bufParam) { bufParam = 0 }
 }
-
-/**
- * A model for `getc` and similar functions that are flow sources.
- */
-private class GetcSource extends SourceModelCsv {
-  override predicate row(string row) {
-    row =
-      [
-        ";;false;getc;;;ReturnValue;remote", ";;false;getwc;;;ReturnValue;remote",
-        ";;false;_getc_nolock;;;ReturnValue;remote", ";;false;_getwc_nolock;;;ReturnValue;remote",
-        ";;false;getch;;;ReturnValue;local", ";;false;_getch;;;ReturnValue;local",
-        ";;false;_getwch;;;ReturnValue;local", ";;false;_getch_nolock;;;ReturnValue;local",
-        ";;false;_getwch_nolock;;;ReturnValue;local", ";;false;getchar;;;ReturnValue;local",
-        ";;false;getwchar;;;ReturnValue;local", ";;false;_getchar_nolock;;;ReturnValue;local",
-        ";;false;_getwchar_nolock;;;ReturnValue;local",
-      ]
-  }
-}

cpp/ql/lib/semmle/code/cpp/models/implementations/Http.qll

@@ -0,0 +1,193 @@
+private import cpp
+private import semmle.code.cpp.ir.dataflow.FlowSteps
+private import semmle.code.cpp.dataflow.new.DataFlow
+
+private class HttpRequest extends Class {
+  HttpRequest() { this.hasGlobalName("_HTTP_REQUEST_V1") }
+}
+
+private class HttpRequestInheritingContent extends TaintInheritingContent, DataFlow::FieldContent {
+  HttpRequestInheritingContent() {
+    this.getAField().getDeclaringType() instanceof HttpRequest and
+    (
+      this.getAField().hasName("pRawUrl") and
+      this.getIndirectionIndex() = 2
+      or
+      this.getAField().hasName("CookedUrl") and
+      this.getIndirectionIndex() = 1
+      or
+      this.getAField().hasName("Headers") and
+      this.getIndirectionIndex() = 1
+      or
+      this.getAField().hasName("pEntityChunks") and
+      this.getIndirectionIndex() = 2
+      or
+      this.getAField().hasName("pSslInfo") and
+      this.getIndirectionIndex() = 2
+    )
+  }
+}
+
+private class HttpCookedUrl extends Class {
+  HttpCookedUrl() { this.hasGlobalName("_HTTP_COOKED_URL") }
+}
+
+private class HttpCookedUrlInheritingContent extends TaintInheritingContent, DataFlow::FieldContent {
+  HttpCookedUrlInheritingContent() {
+    this.getAField().getDeclaringType() instanceof HttpCookedUrl and
+    this.getAField().hasName(["pFullUrl", "pHost", "pAbsPath", "pQueryString"]) and
+    this.getIndirectionIndex() = 2
+  }
+}
+
+private class HttpRequestHeaders extends Class {
+  HttpRequestHeaders() { this.hasGlobalName("_HTTP_REQUEST_HEADERS") }
+}
+
+private class HttpRequestHeadersInheritingContent extends TaintInheritingContent,
+  DataFlow::FieldContent
+{
+  HttpRequestHeadersInheritingContent() {
+    this.getAField().getDeclaringType() instanceof HttpRequestHeaders and
+    (
+      this.getAField().hasName("KnownHeaders") and
+      this.getIndirectionIndex() = 1
+      or
+      this.getAField().hasName("pUnknownHeaders") and
+      this.getIndirectionIndex() = 2
+    )
+  }
+}
+
+private class HttpKnownHeader extends Class {
+  HttpKnownHeader() { this.hasGlobalName("_HTTP_KNOWN_HEADER") }
+}
+
+private class HttpKnownHeaderInheritingContent extends TaintInheritingContent,
+  DataFlow::FieldContent
+{
+  HttpKnownHeaderInheritingContent() {
+    this.getAField().getDeclaringType() instanceof HttpKnownHeader and
+    this.getAField().hasName("pRawValue") and
+    this.getIndirectionIndex() = 2
+  }
+}
+
+private class HttpUnknownHeader extends Class {
+  HttpUnknownHeader() { this.hasGlobalName("_HTTP_UNKNOWN_HEADER") }
+}
+
+private class HttpUnknownHeaderInheritingContent extends TaintInheritingContent,
+  DataFlow::FieldContent
+{
+  HttpUnknownHeaderInheritingContent() {
+    this.getAField().getDeclaringType() instanceof HttpUnknownHeader and
+    this.getAField().hasName(["pName", "pRawValue"]) and
+    this.getIndirectionIndex() = 2
+  }
+}
+
+private class HttpDataChunk extends Class {
+  HttpDataChunk() { this.hasGlobalName("_HTTP_DATA_CHUNK") }
+}
+
+private class HttpDataChunkInheritingContent extends TaintInheritingContent, DataFlow::FieldContent {
+  HttpDataChunkInheritingContent() {
+    this.getAField().getDeclaringType().(Union).getDeclaringType() instanceof HttpDataChunk and
+    (
+      this.getAField().hasName("FromMemory") and
+      this.getIndirectionIndex() = 1
+      or
+      this.getAField().hasName("FromFileHandle") and
+      this.getIndirectionIndex() = 1
+      or
+      this.getAField().hasName("FromFragmentCache") and
+      this.getIndirectionIndex() = 1
+      or
+      this.getAField().hasName("FromFragmentCacheEx") and
+      this.getIndirectionIndex() = 1
+      or
+      this.getAField().hasName("Trailers") and
+      this.getIndirectionIndex() = 1
+    )
+  }
+}
+
+private class FromMemory extends Class {
+  FromMemory() {
+    this.getDeclaringType().(Union).getDeclaringType() instanceof HttpDataChunk and
+    this.getAField().hasName("pBuffer")
+  }
+}
+
+private class FromMemoryInheritingContent extends TaintInheritingContent, DataFlow::FieldContent {
+  FromMemoryInheritingContent() {
+    this.getAField().getDeclaringType() instanceof FromMemory and
+    this.getAField().hasName("pBuffer") and
+    this.getIndirectionIndex() = 2
+  }
+}
+
+private class FromFileHandle extends Class {
+  FromFileHandle() {
+    this.getDeclaringType().(Union).getDeclaringType() instanceof HttpDataChunk and
+    this.getAField().hasName("FileHandle")
+  }
+}
+
+private class FromFileHandleInheritingContent extends TaintInheritingContent, DataFlow::FieldContent
+{
+  FromFileHandleInheritingContent() {
+    this.getAField().getDeclaringType() instanceof FromFileHandle and
+    this.getIndirectionIndex() = 1 and
+    this.getAField().hasName("FileHandle")
+  }
+}
+
+private class FromFragmentCacheOrCacheEx extends Class {
+  FromFragmentCacheOrCacheEx() {
+    this.getDeclaringType().(Union).getDeclaringType() instanceof HttpDataChunk and
+    this.getAField().hasName("pFragmentName")
+  }
+}
+
+private class FromFragmentCacheInheritingContent extends TaintInheritingContent,
+  DataFlow::FieldContent
+{
+  FromFragmentCacheInheritingContent() {
+    this.getAField().getDeclaringType() instanceof FromFragmentCacheOrCacheEx and
+    this.getIndirectionIndex() = 2 and
+    this.getAField().hasName("pFragmentName")
+  }
+}
+
+private class HttpSslInfo extends Class {
+  HttpSslInfo() { this.hasGlobalName("_HTTP_SSL_INFO") }
+}
+
+private class HttpSslInfoInheritingContent extends TaintInheritingContent, DataFlow::FieldContent {
+  HttpSslInfoInheritingContent() {
+    this.getAField().getDeclaringType() instanceof HttpSslInfo and
+    this.getAField().hasName(["pServerCertIssuer", "pServerCertSubject", "pClientCertInfo"]) and
+    this.getIndirectionIndex() = 2
+  }
+}
+
+private class HttpSslClientCertInfo extends Class {
+  HttpSslClientCertInfo() { this.hasGlobalName("_HTTP_SSL_CLIENT_CERT_INFO") }
+}
+
+private class HttpSslClientCertInfoInheritingContent extends TaintInheritingContent,
+  DataFlow::FieldContent
+{
+  HttpSslClientCertInfoInheritingContent() {
+    this.getAField().getDeclaringType() instanceof HttpSslClientCertInfo and
+    (
+      this.getAField().hasName("pCertEncoded") and
+      this.getIndirectionIndex() = 2
+      or
+      this.getAField().hasName("Token") and
+      this.getIndirectionIndex() = 1
+    )
+  }
+}

cpp/ql/lib/semmle/code/cpp/models/implementations/ZMQ.qll

@@ -1,45 +0,0 @@
-/**
- * Provides implementation classes modeling the ZeroMQ networking library.
- */
-
-import semmle.code.cpp.models.interfaces.FlowSource
-
-/**
- * Remote flow sources.
- */
-private class ZmqSource extends SourceModelCsv {
-  override predicate row(string row) {
-    row =
-      [
-        ";;false;zmq_recv;;;Argument[*1];remote", ";;false;zmq_recvmsg;;;Argument[*1];remote",
-        ";;false;zmq_msg_recv;;;Argument[*0];remote",
-      ]
-  }
-}
-
-/**
- * Remote flow sinks.
- */
-private class ZmqSinks extends SinkModelCsv {
-  override predicate row(string row) {
-    row =
-      [
-        ";;false;zmq_send;;;Argument[*1];remote-sink",
-        ";;false;zmq_sendmsg;;;Argument[*1];remote-sink",
-        ";;false;zmq_msg_send;;;Argument[*0];remote-sink",
-      ]
-  }
-}
-
-/**
- * Flow steps.
- */
-private class ZmqSummaries extends SummaryModelCsv {
-  override predicate row(string row) {
-    row =
-      [
-        ";;false;zmq_msg_init_data;;;Argument[*1];Argument[*0];taint",
-        ";;false;zmq_msg_data;;;Argument[*0];ReturnValue[*];taint",
-      ]
-  }
-}

cpp/ql/lib/semmlecode.cpp.dbscheme

@@ -245,6 +245,25 @@ trap_filename(
     string filename: string ref
 );
 
+/**
+ * Gives the tag name for `tag`.
+ * For debugging only.
+ */
+tag_name(
+    int tag: @tag,
+    string name: string ref
+);
+
+@trap_or_tag = @tag | @trap;
+
+/**
+ * Gives the name for the source file.
+ */
+source_file_name(
+    int sf: @source_file,
+    string name: string ref
+);
+
 /**
  * In `build-mode: none` overlay mode, indicates that `source_file`
  * (`/path/to/foo.c`) uses the TRAP file `trap_file`; i.e. it is the
@@ -252,16 +271,25 @@ trap_filename(
  * includes, or a template instantiation it transitively uses.
  */
 source_file_uses_trap(
-    string source_file: string ref,
+    int source_file: @source_file ref,
     int trap_file: @trap ref
 );
 
 /**
- * Holds if there is a definition of `element` in TRAP file `trap_file`.
+ * In `build-mode: none` overlay mode, indicates that the TRAP file
+ * `trap_file` uses tag `tag`.
  */
-in_trap(
+trap_uses_tag(
+    int trap_file: @trap ref,
+    int tag: @tag ref
+);
+
+/**
+ * Holds if there is a definition of `element` in TRAP file or tag `t`.
+ */
+in_trap_or_tag(
     int element: @element ref,
-    int trap_file: @trap ref
+    int t: @trap_or_tag ref
 );
 
 pch_uses(

cpp/ql/lib/upgrades/7e7c2f55670f8123d514cf542ccb1938118ac561/in_trap_or_tag.ql

@@ -0,0 +1,11 @@
+class Element extends @element {
+  string toString() { none() }
+}
+
+class Trap extends @trap {
+  string toString() { none() }
+}
+
+from Element e, Trap trap
+where in_trap(e, trap)
+select e, trap

cpp/ql/lib/upgrades/7e7c2f55670f8123d514cf542ccb1938118ac561/source_files.ql

@@ -0,0 +1,22 @@
+newtype TSourceFile = MkSourceFile(string name) { source_file_uses_trap(name, _) }
+
+module FreshSourceFile = QlBuiltins::NewEntity<TSourceFile>;
+
+class SourceFile extends FreshSourceFile::EntityId {
+  string toString() { none() }
+}
+
+class Trap extends @trap {
+  string toString() { none() }
+}
+
+query predicate mk_source_file_name(SourceFile source_file, string name) {
+  source_file = FreshSourceFile::map(MkSourceFile(name))
+}
+
+query predicate mk_source_file_uses_trap(SourceFile source_file, Trap trap) {
+  exists(string name |
+    source_file_uses_trap(name, trap) and
+    mk_source_file_name(source_file, name)
+  )
+}

cpp/ql/lib/upgrades/7e7c2f55670f8123d514cf542ccb1938118ac561/upgrade.properties

@@ -0,0 +1,6 @@
+description: Add source_file_name
+compatibility: backwards
+source_file_uses_trap.rel: run source_files.ql mk_source_file_uses_trap
+source_file_name.rel: run source_files.ql mk_source_file_name
+in_trap.rel: delete
+in_trap_or_tag.rel: run in_trap_or_tag.ql

cpp/ql/src/CHANGELOG.md

@@ -1,3 +1,15 @@
+## 1.5.15
+
+No user-facing changes.
+
+## 1.5.14
+
+No user-facing changes.
+
+## 1.5.13
+
+No user-facing changes.
+
 ## 1.5.12
 
 No user-facing changes.

cpp/ql/src/Diagnostics/ExtractionProblems.qll

@@ -50,7 +50,7 @@ private newtype TExtractionProblem =
 /**
  * Superclass for the extraction problem hierarchy.
  */
-class ExtractionProblem extends TExtractionProblem {
+abstract class ExtractionProblem extends TExtractionProblem {
   /** Gets the string representation of the problem. */
   string toString() { none() }
 
@@ -65,6 +65,9 @@ class ExtractionProblem extends TExtractionProblem {
 
   /** Gets the SARIF severity of this problem. */
   int getSeverity() { none() }
+
+  /** Gets the `Compilation` the problem is associated with. */
+  abstract Compilation getCompilation();
 }
 
 /**
@@ -96,6 +99,8 @@ class ExtractionUnrecoverableError extends ExtractionProblem, TCompilationFailed
     // [errors](https://docs.oasis-open.org/sarif/sarif/v2.1.0/csprd01/sarif-v2.1.0-csprd01.html#_Toc10541338).
     result = 2
   }
+
+  override Compilation getCompilation() { result = c }
 }
 
 /**
@@ -122,6 +127,8 @@ class ExtractionRecoverableWarning extends ExtractionProblem, TReportableWarning
     // [warnings](https://docs.oasis-open.org/sarif/sarif/v2.1.0/csprd01/sarif-v2.1.0-csprd01.html#_Toc10541338).
     result = 1
   }
+
+  override Compilation getCompilation() { result = err.getCompilation() }
 }
 
 /**
@@ -148,4 +155,6 @@ class ExtractionUnknownProblem extends ExtractionProblem, TUnknownProblem {
     // [warnings](https://docs.oasis-open.org/sarif/sarif/v2.1.0/csprd01/sarif-v2.1.0-csprd01.html#_Toc10541338).
     result = 1
   }
+
+  override Compilation getCompilation() { result = err.getCompilation() }
 }

cpp/ql/src/Diagnostics/ExtractionWarnings.ql

@@ -10,7 +10,9 @@ import ExtractionProblems
 
 from ExtractionProblem warning
 where
-  warning instanceof ExtractionRecoverableWarning and exists(warning.getFile().getRelativePath())
+  warning instanceof ExtractionRecoverableWarning and
+  exists(warning.getFile().getRelativePath()) and
+  not warning.getCompilation().buildModeNone()
   or
   warning instanceof ExtractionUnknownProblem
 select warning,

cpp/ql/src/Likely Bugs/Arithmetic/IntMultToLong.ql

@@ -218,7 +218,9 @@ where
   // only report if we cannot prove that the result of the
   // multiplication will be less (resp. greater) than the
   // maximum (resp. minimum) number we can compute.
-  overflows(me, t1)
+  overflows(me, t1) and
+  // exclude cases where the expression type may not have been extracted accurately
+  not me.getParent().(Call).getTarget().hasAmbiguousReturnType()
 select me,
   "Multiplication result may overflow '" + me.getType().toString() + "' before it is converted to '"
     + me.getFullyConverted().getType().toString() + "'."

cpp/ql/src/Likely Bugs/Format/WrongTypeFormatArguments.ql

@@ -168,9 +168,11 @@ where
     formatOtherArgType(ffc, n, expected, arg, actual) and
     not actual.getUnspecifiedType().(IntegralType).getSize() = sizeof_IntType()
   ) and
+  // Exclude some cases where we're less confident the result is correct / clear / valuable
   not arg.isAffectedByMacro() and
   not arg.isFromUninstantiatedTemplate(_) and
   not actual.stripType() instanceof ErroneousType and
+  not arg.getType().stripType().(RoutineType).getReturnType() instanceof ErroneousType and
   not arg.(Call).mayBeFromImplicitlyDeclaredFunction() and
   // Make sure that the format function definition is consistent
   count(ffc.getTarget().getFormatParameterIndex()) = 1

cpp/ql/src/Likely Bugs/Memory Management/AllocaInLoop.ql

@@ -15,6 +15,7 @@
 import cpp
 import semmle.code.cpp.rangeanalysis.RangeAnalysisUtils
 import semmle.code.cpp.ir.dataflow.DataFlow
+private import semmle.code.cpp.ir.dataflow.internal.DataFlowNodes
 
 /** Gets a loop that contains `e`. */
 Loop getAnEnclosingLoopOfExpr(Expr e) { result = getAnEnclosingLoopOfStmt(e.getEnclosingStmt()) }
@@ -45,9 +46,9 @@ private Expr getExpr(DataFlow::Node node) {
   or
   result = node.asOperand().getUse().getAst()
   or
-  result = node.(DataFlow::RawIndirectInstruction).getInstruction().getAst()
+  result = node.(RawIndirectInstruction).getInstruction().getAst()
   or
-  result = node.(DataFlow::RawIndirectOperand).getOperand().getUse().getAst()
+  result = node.(RawIndirectOperand).getOperand().getUse().getAst()
 }
 
 /**
@@ -208,7 +209,7 @@ class LoopWithAlloca extends Stmt {
       this.conditionRequiresInequality(va, _, _) and
       DataFlow::localFlow(result, DataFlow::exprNode(va)) and
       // Phi nodes will be preceded by nodes that represent actual definitions
-      not result instanceof DataFlow::SsaSynthNode and
+      not result instanceof SsaSynthNode and
       // A source is outside the loop if it's not inside the loop
       not exists(Expr e | e = getExpr(result) | this = getAnEnclosingLoopOfExpr(e))
     )

cpp/ql/src/Security/CWE/CWE-079/CgiXss.ql

@@ -4,7 +4,7 @@
  *              allows for a cross-site scripting vulnerability.
  * @kind path-problem
  * @problem.severity error
- * @security-severity 6.1
+ * @security-severity 7.8
  * @precision high
  * @id cpp/cgi-xss
  * @tags security

cpp/ql/src/Security/CWE/CWE-134/UncontrolledFormatString.ql

@@ -23,13 +23,31 @@ import Flow::PathGraph
 
 predicate isSource(FlowSource source, string sourceType) { sourceType = source.getSourceType() }
 
+/**
+ * Holds if `f` is a printf-like function or a (possibly nested) wrapper
+ * that forwards a format-string parameter to one.
+ *
+ * Functions that *implement* printf-like behavior (e.g. a custom
+ * `vsnprintf` variant) internally parse the caller-supplied format string
+ * and build small, bounded, local format strings such as `"%d"` or `"%ld"`
+ * for inner `sprintf` calls.  Taint that reaches those inner calls via the
+ * parsed format specifier is not exploitable, so sinks inside such
+ * functions should be excluded.
+ */
+private predicate isPrintfImplementation(Function f) {
+  f instanceof PrintfLikeFunction
+  or
+  exists(PrintfLikeFunction printf | printf.wrapperFunction(f, _, _))
+}
+
 module Config implements DataFlow::ConfigSig {
   predicate isSource(DataFlow::Node node) { isSource(node, _) }
 
   predicate isSink(DataFlow::Node node) {
     exists(PrintfLikeFunction printf |
       printf.outermostWrapperFunctionCall([node.asExpr(), node.asIndirectExpr()], _)
-    )
+    ) and
+    not isPrintfImplementation([node.asExpr(), node.asIndirectExpr()].getEnclosingFunction())
   }
 
   private predicate isArithmeticNonCharType(ArithmeticType type) {

cpp/ql/src/Security/CWE/CWE-468/SuspiciousAddWithSizeof.ql

@@ -18,7 +18,8 @@ import IncorrectPointerScalingCommon
 private predicate isCharSzPtrExpr(Expr e) {
   exists(PointerType pt | pt = e.getFullyConverted().getUnspecifiedType() |
     pt.getBaseType() instanceof CharType or
-    pt.getBaseType() instanceof VoidType
+    pt.getBaseType() instanceof VoidType or
+    pt.getBaseType() instanceof ErroneousType // this could be char / void type in a successful compilation
   )
 }
 

cpp/ql/src/Telemetry/DatabaseQuality.qll

@@ -0,0 +1,48 @@
+import cpp
+import codeql.util.ReportStats
+
+/** A file that is included in the quality statistics. */
+private class RelevantFile extends File {
+  RelevantFile() { this.fromSource() and exists(this.getRelativePath()) }
+}
+
+module CallTargetStats implements StatsSig {
+  private class RelevantCall extends Call {
+    RelevantCall() { this.getFile() instanceof RelevantFile }
+  }
+
+  // We assume that calls with an implicit target are calls that could not be
+  // resolved. This is accurate in the vast majority of cases, but is inaccurate
+  // for calls that deliberately rely on implicitly declared functions.
+  private predicate hasImplicitTarget(RelevantCall call) {
+    call.getTarget().getADeclarationEntry().isImplicit()
+  }
+
+  int getNumberOfOk() { result = count(RelevantCall call | not hasImplicitTarget(call)) }
+
+  int getNumberOfNotOk() { result = count(RelevantCall call | hasImplicitTarget(call)) }
+
+  string getOkText() { result = "calls with call target" }
+
+  string getNotOkText() { result = "calls with missing call target" }
+}
+
+private class SourceExpr extends Expr {
+  SourceExpr() { this.getFile() instanceof RelevantFile }
+}
+
+private predicate hasGoodType(Expr e) { not e.getType() instanceof ErroneousType }
+
+module ExprTypeStats implements StatsSig {
+  int getNumberOfOk() { result = count(SourceExpr e | hasGoodType(e)) }
+
+  int getNumberOfNotOk() { result = count(SourceExpr e | not hasGoodType(e)) }
+
+  string getOkText() { result = "expressions with known type" }
+
+  string getNotOkText() { result = "expressions with unknown type" }
+}
+
+module CallTargetStatsReport = ReportStats<CallTargetStats>;
+
+module ExprTypeStatsReport = ReportStats<ExprTypeStats>;