Merge pull request #12551 from github/release-prep/2.12.5

Release preparation for version 2.12.5

.git-blame-ignore-revs

@@ -0,0 +1,21 @@
+# .git-blame-ignore-revs
+# Auto-formatted Java
+730eae952139209fe9fdf598541d608f4c0c0c84
+# Auto-formatted C#
+5ad7ed49dd3de03ec6dcfcb6848758a6a987e11c
+# Auto-formatted C/C++
+ef97e539ec1971494d4bba5cafe82e00bc8217ac
+# Auto-formatted Python
+21d5fa836b3a7d020ba45e8b8168b145a9772131
+# Auto-formatted JavaScript
+8d97fe9ed327a9546ff2eaf515cf0f5214deddd9
+# Auto-formatted Ruby
+a5d229903d2f12d45f2c2c38822f1d0e7504ae7f
+# Auto-formatted Go
+08c658e66bf867090033ea096e244a93d46c0aa7
+# Auto-formatted Swift
+711d7057f79fb7d72fc3b35e010bd018f9009169
+# Auto-formatted shared ql packs
+3640b6d3a8ce9edf8e1d3ed106fe8526cf255bc0
+# Auto-formatted taint tracking files
+159d8e978c51959b380838c080d891b66e763b19

.github/actions/cache-query-compilation/action.yml

@@ -9,7 +9,7 @@ inputs:
 outputs:
   cache-dir:
     description: "The directory where the cache was stored"
-    value: ${{ steps.fill-compilation-dir.outputs.compdir }}
+    value: ${{ steps.output-compilation-dir.outputs.compdir }}
 
 runs:
   using: composite
@@ -27,7 +27,9 @@ runs:
       if: ${{ github.event_name == 'pull_request' }}
       uses: actions/cache/restore@v3
       with:
-        path: '**/.cache'
+        path: |
+          **/.cache
+          ~/.codeql/compile-cache
         key: codeql-compile-${{ inputs.key }}-pr-${{ github.sha }}
         restore-keys: |
           codeql-compile-${{ inputs.key }}-${{ github.base_ref }}-${{ env.merge_base }}
@@ -37,18 +39,111 @@ runs:
       if: ${{ github.event_name != 'pull_request' }}
       uses: actions/cache@v3
       with:
-        path: '**/.cache'
+        path: |
+          **/.cache
+          ~/.codeql/compile-cache
         key: codeql-compile-${{ inputs.key }}-${{ github.ref_name }}-${{ github.sha }} # just fill on main
         restore-keys: | # restore the latest cache if the exact cache is unavailable, to speed up compilation.
           codeql-compile-${{ inputs.key }}-${{ github.ref_name }}-
           codeql-compile-${{ inputs.key }}-main-
-    - name: Fill compilation cache directory
-      id: fill-compilation-dir
+    - name: Output-compilationdir
+      id: output-compilation-dir
       shell: bash
       run: |
-        # Move all the existing cache into another folder, so we only preserve the cache for the current queries.
-        node $GITHUB_WORKSPACE/.github/actions/cache-query-compilation/move-caches.js ${COMBINED_CACHE_DIR}
-
         echo "compdir=${COMBINED_CACHE_DIR}" >> $GITHUB_OUTPUT
       env:
         COMBINED_CACHE_DIR: ${{ runner.temp }}/compilation-dir
+    - name: Fill compilation cache directory
+      id: fill-compilation-dir
+      uses: actions/github-script@v6
+      env: 
+        COMBINED_CACHE_DIR: ${{ runner.temp }}/compilation-dir
+      with:
+        script: |
+          // # Move all the existing cache into another folder, so we only preserve the cache for the current queries.
+          // mkdir -p ${COMBINED_CACHE_DIR}
+          // rm -f **/.cache/{lock,size} # -f to avoid errors if the cache is empty.
+          // # copy the contents of the .cache folders into the combined cache folder.
+          // cp -r **/.cache/* ${COMBINED_CACHE_DIR}/ || : # ignore missing files
+          // # clean up the .cache folders
+          // rm -rf **/.cache/*
+
+          const fs = require("fs");
+          const path = require("path");
+          const os = require("os");
+
+          // the first argv is the cache folder to create.
+          const COMBINED_CACHE_DIR = process.env.COMBINED_CACHE_DIR;
+
+          function* walkCaches(dir) {
+            const files = fs.readdirSync(dir, { withFileTypes: true });
+            for (const file of files) {
+              if (file.isDirectory()) {
+                const filePath = path.join(dir, file.name);
+                yield* walkCaches(filePath);
+                if (file.name === ".cache") {
+                  yield filePath;
+                }
+              }
+            }
+          }
+
+          async function copyDir(src, dest) {
+            for await (const file of await fs.promises.readdir(src, { withFileTypes: true })) {
+              const srcPath = path.join(src, file.name);
+              const destPath = path.join(dest, file.name);
+              if (file.isDirectory()) {
+                if (!fs.existsSync(destPath)) {
+                  fs.mkdirSync(destPath);
+                }
+                await copyDir(srcPath, destPath);
+              } else {
+                await fs.promises.copyFile(srcPath, destPath);
+              }
+            }
+          }
+
+          async function main() {
+            const cacheDirs = [...walkCaches(".")];
+
+            for (const dir of cacheDirs) {
+              console.log(`Found .cache dir at ${dir}`);
+            }
+
+            const globalCacheDir = path.join(os.homedir(), ".codeql", "compile-cache");
+            if (fs.existsSync(globalCacheDir)) {
+              console.log("Found global home dir: " + globalCacheDir);
+              cacheDirs.push(globalCacheDir);
+            }
+
+            if (cacheDirs.length === 0) {
+              console.log("No cache dirs found");
+              return;
+            }
+
+            // mkdir -p ${COMBINED_CACHE_DIR}
+            fs.mkdirSync(COMBINED_CACHE_DIR, { recursive: true });
+
+            // rm -f **/.cache/{lock,size} # -f to avoid errors if the cache is empty.
+            await Promise.all(
+              cacheDirs.map((cacheDir) =>
+                (async function () {
+                  await fs.promises.rm(path.join(cacheDir, "lock"), { force: true });
+                  await fs.promises.rm(path.join(cacheDir, "size"), { force: true });
+                })()
+              )
+            );
+
+            // # copy the contents of the .cache folders into the combined cache folder.
+            // cp -r **/.cache/* ${COMBINED_CACHE_DIR}/ || : # ignore missing files
+            await Promise.all(
+              cacheDirs.map((cacheDir) => copyDir(cacheDir, COMBINED_CACHE_DIR))
+            );
+
+            // # clean up the .cache folders
+            // rm -rf **/.cache/*
+            await Promise.all(
+              cacheDirs.map((cacheDir) => fs.promises.rm(cacheDir, { recursive: true }))
+            );
+          }
+          main();

.github/actions/cache-query-compilation/move-caches.js

@@ -1,75 +0,0 @@
-// # Move all the existing cache into another folder, so we only preserve the cache for the current queries.
-// mkdir -p ${COMBINED_CACHE_DIR}
-// rm -f **/.cache/{lock,size} # -f to avoid errors if the cache is empty.
-// # copy the contents of the .cache folders into the combined cache folder.
-// cp -r **/.cache/* ${COMBINED_CACHE_DIR}/ || : # ignore missing files
-// # clean up the .cache folders
-// rm -rf **/.cache/*
-
-const fs = require("fs");
-const path = require("path");
-
-// the first argv is the cache folder to create.
-const COMBINED_CACHE_DIR = process.argv[2];
-
-function* walkCaches(dir) {
-  const files = fs.readdirSync(dir, { withFileTypes: true });
-  for (const file of files) {
-    if (file.isDirectory()) {
-      const filePath = path.join(dir, file.name);
-      yield* walkCaches(filePath);
-      if (file.name === ".cache") {
-        yield filePath;
-      }
-    }
-  }
-}
-
-async function copyDir(src, dest) {
-  for await (const file of await fs.promises.readdir(src, { withFileTypes: true })) {
-    const srcPath = path.join(src, file.name);
-    const destPath = path.join(dest, file.name);
-    if (file.isDirectory()) {
-      if (!fs.existsSync(destPath)) {
-        fs.mkdirSync(destPath);
-      }
-      await copyDir(srcPath, destPath);
-    } else {
-      await fs.promises.copyFile(srcPath, destPath);
-    }
-  }
-}
-
-async function main() {
-  const cacheDirs = [...walkCaches(".")];
-
-  for (const dir of cacheDirs) {
-    console.log(`Found .cache dir at ${dir}`);
-  }
-
-  // mkdir -p ${COMBINED_CACHE_DIR}
-  fs.mkdirSync(COMBINED_CACHE_DIR, { recursive: true });
-
-  // rm -f **/.cache/{lock,size} # -f to avoid errors if the cache is empty.
-  await Promise.all(
-    cacheDirs.map((cacheDir) =>
-      (async function () {
-        await fs.promises.rm(path.join(cacheDir, "lock"), { force: true });
-        await fs.promises.rm(path.join(cacheDir, "size"), { force: true });
-      })()
-    )
-  );
-
-  // # copy the contents of the .cache folders into the combined cache folder.
-  // cp -r **/.cache/* ${COMBINED_CACHE_DIR}/ || : # ignore missing files
-  await Promise.all(
-    cacheDirs.map((cacheDir) => copyDir(cacheDir, COMBINED_CACHE_DIR))
-  );
-
-  // # clean up the .cache folders
-  // rm -rf **/.cache/*
-  await Promise.all(
-    cacheDirs.map((cacheDir) => fs.promises.rm(cacheDir, { recursive: true }))
-  );
-}
-main();

.github/actions/find-latest-bundle/action.yml

@@ -1,26 +0,0 @@
-name: Find Latest CodeQL Bundle
-description: Finds the URL of the latest released version of the CodeQL bundle.
-outputs:
-  url:
-    description: The download URL of the latest CodeQL bundle release
-    value: ${{ steps.find-latest.outputs.url }}
-runs:
-  using: composite
-  steps:
-    - name: Find Latest Release
-      id: find-latest
-      shell: pwsh
-      run: |
-        $Latest = gh release list --repo github/codeql-action --exclude-drafts --limit 1000 |
-          ForEach-Object { $C = $_ -split "`t"; return @{ type = $C[1]; tag = $C[2]; } } |
-          Where-Object { $_.type -eq 'Latest' }
-
-        $Tag = $Latest.tag
-        if ($Tag -eq '') {
-          throw 'Failed to find latest bundle release.'
-        }
-
-        Write-Output "Latest bundle tag is '${Tag}'."
-        "url=https://github.com/github/codeql-action/releases/download/${Tag}/codeql-bundle-linux64.tar.gz" >> $env:GITHUB_OUTPUT
-      env:
-        GITHUB_TOKEN: ${{ github.token }}

.github/workflows/check-qldoc.yml

@@ -26,9 +26,8 @@ jobs:
         shell: bash
         run: |
           EXIT_CODE=0
-          # TODO: remove the swift exception from the regex when we fix generated QLdoc
           # TODO: remove the shared exception from the regex when coverage of qlpacks without dbschemes is supported
-          changed_lib_packs="$(git diff --name-only --diff-filter=ACMRT HEAD^ HEAD | { grep -Po '^(?!(swift|shared))[a-z]*/ql/lib' || true; } | sort -u)"
+          changed_lib_packs="$(git diff --name-only --diff-filter=ACMRT HEAD^ HEAD | { grep -Po '^(?!(shared))[a-z]*/ql/lib' || true; } | sort -u)"
           for pack_dir in ${changed_lib_packs}; do
             lang="${pack_dir%/ql/lib}"
             codeql generate library-doc-coverage --output="${RUNNER_TEMP}/${lang}-current.txt" --dir="${pack_dir}"

.github/workflows/codeql-analysis.yml

@@ -28,7 +28,7 @@ jobs:
 
     steps:
     - name: Setup dotnet
-      uses: actions/setup-dotnet@v2
+      uses: actions/setup-dotnet@v3
       with:
         dotnet-version: 7.0.102
 

.github/workflows/compile-queries.yml

@@ -24,14 +24,14 @@ jobs:
         with: 
           key: all-queries
       - name: check formatting
-        run: find */ql -type f \( -name "*.qll" -o -name "*.ql" \) -print0 | xargs -0 codeql query format --check-only
+        run: find */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 : ${{ github.event_name == 'pull_request' }}
         shell: bash
-        run: codeql query compile -j0 */ql/{src,examples} --keep-going --warnings=error --check-only --compilation-cache "${{ steps.query-cache.outputs.cache-dir }}"
+        run: codeql query compile -q -j0 */ql/{src,examples} --keep-going --warnings=error --check-only --compilation-cache "${{ steps.query-cache.outputs.cache-dir }}"
       - name: compile queries - full
         # do full compile if running on main - this populates the cache
         if : ${{ github.event_name != 'pull_request' }}
         shell: bash
-        run: codeql query compile -j0 */ql/{src,examples} --keep-going --warnings=error --compilation-cache "${{ steps.query-cache.outputs.cache-dir }}"
+        run: codeql query compile -q -j0 */ql/{src,examples} --keep-going --warnings=error --compilation-cache "${{ steps.query-cache.outputs.cache-dir }}"

.github/workflows/go-tests-other-os.yml

@@ -12,10 +12,10 @@ jobs:
     name: Test MacOS
     runs-on: macos-latest
     steps:
-      - name: Set up Go 1.19
+      - name: Set up Go 1.20
         uses: actions/setup-go@v3
         with:
-          go-version: 1.19
+          go-version: 1.20.0
         id: go
 
       - name: Check out code
@@ -47,10 +47,10 @@ jobs:
     name: Test Windows
     runs-on: windows-latest-xl
     steps:
-      - name: Set up Go 1.19
+      - name: Set up Go 1.20
         uses: actions/setup-go@v3
         with:
-          go-version: 1.19
+          go-version: 1.20.0
         id: go
 
       - name: Check out code

.github/workflows/go-tests.yml

@@ -20,10 +20,10 @@ jobs:
     name: Test Linux (Ubuntu)
     runs-on: ubuntu-latest-xl
     steps:
-      - name: Set up Go 1.19
+      - name: Set up Go 1.20
         uses: actions/setup-go@v3
         with:
-          go-version: 1.19
+          go-version: 1.20.0
         id: go
 
       - name: Check out code

.github/workflows/ql-for-ql-build.yml

@@ -5,13 +5,6 @@ on:
     branches: [main]
   pull_request:
     branches: [main]
-    paths:
-      - "ql/**"
-      - "**.qll"
-      - "**.ql"
-      - "**.dbscheme"
-      - "**/qlpack.yml"
-      - ".github/workflows/ql-for-ql-build.yml"
 
 env:
   CARGO_TERM_COLOR: always
@@ -22,66 +15,26 @@ jobs:
     steps:
       ### Build the queries ###
       - uses: actions/checkout@v3
-      - name: Find latest bundle
-        id: find-latest-bundle
-        uses: ./.github/actions/find-latest-bundle
+        with:
+          fetch-depth: 0
       - name: Find codeql
         id: find-codeql
-        uses: github/codeql-action/init@beae46e6b1da530ed5e9fc6a756f92433ca47ae1
+        uses: github/codeql-action/init@v2
         with:
           languages: javascript # does not matter
-          tools: ${{ steps.find-latest-bundle.outputs.url }}
-      - name: Get CodeQL version
-        id: get-codeql-version
-        run: |
-          echo "version=$("${CODEQL}" --version | head -n 1 | rev | cut -d " " -f 1 | rev)" >> $GITHUB_OUTPUT
-        shell: bash
-        env:
-          CODEQL: ${{ steps.find-codeql.outputs.codeql-path }}
       - uses: ./.github/actions/os-version
         id: os_version
-      - name: Cache entire pack
-        id: cache-pack
-        uses: actions/cache@v3
-        with:
-          path: ${{ runner.temp }}/pack
-          key: ${{ runner.os }}-${{ steps.os_version.outputs.version }}-pack-${{ hashFiles('ql/**/Cargo.lock') }}-${{ hashFiles('ql/**/*.rs') }}-${{ hashFiles('ql/**/*.ql*') }}-${{ hashFiles('ql/**/qlpack.yml') }}-${{ hashFiles('ql/ql/src/ql.dbscheme*') }}-${{ steps.get-codeql-version.outputs.version }}--${{ hashFiles('.github/workflows/ql-for-ql-build.yml') }}
-      - name: Cache queries
-        if: steps.cache-pack.outputs.cache-hit != 'true'
-        id: cache-queries
-        uses: actions/cache@v3
-        with:
-          path: ${{ runner.temp }}/queries
-          key: queries-${{ hashFiles('ql/**/*.ql*') }}-${{ hashFiles('ql/**/qlpack.yml') }}-${{ hashFiles('ql/ql/src/ql.dbscheme*') }}-${{ steps.get-codeql-version.outputs.version }}--${{ hashFiles('.github/workflows/ql-for-ql-build.yml') }}
-      - name: Build query pack
-        if: steps.cache-queries.outputs.cache-hit != 'true' && steps.cache-pack.outputs.cache-hit != 'true'
-        run: |
-          cd ql/ql/src
-          "${CODEQL}" pack create -j 16
-          mv .codeql/pack/codeql/ql/0.0.0 ${{ runner.temp }}/queries
-        env:
-          CODEQL: ${{ steps.find-codeql.outputs.codeql-path }}
-      - name: Move cache queries to pack
-        if: steps.cache-pack.outputs.cache-hit != 'true'
-        run: |
-          cp -r ${{ runner.temp }}/queries ${{ runner.temp }}/pack
-        env:
-          CODEQL: ${{ steps.find-codeql.outputs.codeql-path }}
-
       ### Build the extractor ###
       - name: Cache entire extractor
-        if: steps.cache-pack.outputs.cache-hit != 'true'
         id: cache-extractor
         uses: actions/cache@v3
         with:
           path: |
-            ql/target/release/ql-autobuilder
-            ql/target/release/ql-autobuilder.exe
-            ql/target/release/ql-extractor
-            ql/target/release/ql-extractor.exe
+            ql/extractor-pack/
+            ql/target/release/buramu
           key: ${{ runner.os }}-${{ steps.os_version.outputs.version }}-extractor-${{ hashFiles('ql/**/Cargo.lock') }}-${{ hashFiles('ql/**/*.rs') }}
       - name: Cache cargo
-        if: steps.cache-extractor.outputs.cache-hit != 'true' && steps.cache-pack.outputs.cache-hit != 'true'
+        if: steps.cache-extractor.outputs.cache-hit != 'true'
         uses: actions/cache@v3
         with:
           path: |
@@ -89,77 +42,33 @@ jobs:
             ~/.cargo/git
             ql/target
           key: ${{ runner.os }}-${{ steps.os_version.outputs.version }}-rust-cargo-${{ hashFiles('ql/**/Cargo.lock') }}
-      - name: Check formatting
-        if: steps.cache-extractor.outputs.cache-hit != 'true' && steps.cache-pack.outputs.cache-hit != 'true'
-        run: cd ql; cargo fmt --all -- --check
-      - name: Build
-        if: steps.cache-extractor.outputs.cache-hit != 'true' && steps.cache-pack.outputs.cache-hit != 'true'
-        run: cd ql; cargo build --verbose
-      - name: Run tests
-        if: steps.cache-extractor.outputs.cache-hit != 'true' && steps.cache-pack.outputs.cache-hit != 'true'
-        run: cd ql; cargo test --verbose
       - name: Release build
-        if: steps.cache-extractor.outputs.cache-hit != 'true' && steps.cache-pack.outputs.cache-hit != 'true'
-        run: cd ql; cargo build --release
-      - name: Generate dbscheme
-        if: steps.cache-extractor.outputs.cache-hit != 'true' && steps.cache-pack.outputs.cache-hit != 'true'
-        run: ql/target/release/ql-generator --dbscheme ql/ql/src/ql.dbscheme --library ql/ql/src/codeql_ql/ast/internal/TreeSitter.qll
-
-      ### Package the queries and extractor ###
-      - name: Package pack
-        if: steps.cache-pack.outputs.cache-hit != 'true'
-        run: |
-          cp -r ql/codeql-extractor.yml ql/tools ql/ql/src/ql.dbscheme.stats ${PACK}/
-          mkdir -p ${PACK}/tools/linux64
-          cp ql/target/release/ql-autobuilder  ${PACK}/tools/linux64/autobuilder
-          cp ql/target/release/ql-extractor ${PACK}/tools/linux64/extractor
-          chmod +x ${PACK}/tools/linux64/autobuilder
-          chmod +x ${PACK}/tools/linux64/extractor
+        if: steps.cache-extractor.outputs.cache-hit != 'true'
+        run: cd ql; ./scripts/create-extractor-pack.sh       
         env:
-          PACK: ${{ runner.temp }}/pack
-
-      ### Run the analysis ###
-      - name: Hack codeql-action options
+          GH_TOKEN: ${{ github.token }}   
+      - name: Cache compilation cache
+        id: query-cache
+        uses: ./.github/actions/cache-query-compilation
+        with: 
+          key: run-ql-for-ql
+      - name: Make database and analyze
         run: |
-          JSON=$(jq -nc --arg pack "${PACK}" '.database."run-queries"=["--search-path", $pack] | .resolve.queries=["--search-path", $pack] | .resolve.extractor=["--search-path", $pack] | .resolve.languages=["--search-path", $pack] | .database.init=["--search-path", $pack]')
-          echo "CODEQL_ACTION_EXTRA_OPTIONS=${JSON}" >> ${GITHUB_ENV}
-        env:
-          PACK: ${{ runner.temp }}/pack
-
-      - name: Create CodeQL config file
-        run: |
-          echo "paths-ignore:" >> ${CONF}
-          echo "  - ql/ql/test" >> ${CONF}
-          echo "  - \"*/ql/lib/upgrades/\"" >> ${CONF}
-          echo "disable-default-queries: true" >> ${CONF}
-          echo "queries:" >> ${CONF}
-          echo "  - uses: ./ql/ql/src/codeql-suites/ql-code-scanning.qls" >> ${CONF}
-          echo "Config file: "
-          cat ${CONF}
-        env:
-          CONF: ./ql-for-ql-config.yml
-      - name: Initialize CodeQL
-        uses: github/codeql-action/init@beae46e6b1da530ed5e9fc6a756f92433ca47ae1
+          ./ql/target/release/buramu | tee deprecated.blame # Add a blame file for the extractor to parse.
+          ${CODEQL} database create -l=ql --search-path ql/extractor-pack ${DB}
+          ${CODEQL} database analyze -j0 --format=sarif-latest --output=ql-for-ql.sarif ${DB} ql/ql/src/codeql-suites/ql-code-scanning.qls  --compilation-cache "${{ steps.query-cache.outputs.cache-dir }}"
+        env: 
+          CODEQL: ${{ steps.find-codeql.outputs.codeql-path }}
+          DB: ${{ runner.temp }}/DB
+          LGTM_INDEX_FILTERS: |
+            exclude:ql/ql/test
+            exclude:*/ql/lib/upgrades/
+            exclude:java/ql/integration-tests
+      - name: Upload sarif to code-scanning
+        uses: github/codeql-action/upload-sarif@v2
         with:
-          languages: ql
-          db-location: ${{ runner.temp }}/db
-          config-file: ./ql-for-ql-config.yml
-          tools: ${{ steps.find-latest-bundle.outputs.url }}
-      - name: Move pack queries
-        run: |
-          cp -r ${PACK}/queries ql/ql/src
-        env:
-          PACK: ${{ runner.temp }}/pack
-
-      - name: Perform CodeQL Analysis
-        uses: github/codeql-action/analyze@beae46e6b1da530ed5e9fc6a756f92433ca47ae1
-        with:
-          category: "ql-for-ql"
-      - name: Copy sarif file to CWD
-        run: cp ../results/ql.sarif ./ql-for-ql.sarif
-      - name: Fixup the $scema in sarif  # Until https://github.com/microsoft/sarif-vscode-extension/pull/436/ is part in a stable release
-        run: |
-          sed -i 's/\$schema.*/\$schema": "https:\/\/raw.githubusercontent.com\/oasis-tcs\/sarif-spec\/master\/Schemata\/sarif-schema-2.1.0",/' ql-for-ql.sarif
+          sarif_file: ql-for-ql.sarif
+          category: ql-for-ql
       - name: Sarif as artifact
         uses: actions/upload-artifact@v3
         with:

.github/workflows/ql-for-ql-dataset_measure.yml

@@ -25,7 +25,7 @@ jobs:
 
       - name: Find codeql
         id: find-codeql
-        uses: github/codeql-action/init@beae46e6b1da530ed5e9fc6a756f92433ca47ae1
+        uses: github/codeql-action/init@v2
         with:
           languages: javascript # does not matter
       - uses: ./.github/actions/os-version

.github/workflows/ql-for-ql-tests.yml

@@ -24,7 +24,7 @@ jobs:
       - uses: actions/checkout@v3
       - name: Find codeql
         id: find-codeql
-        uses: github/codeql-action/init@beae46e6b1da530ed5e9fc6a756f92433ca47ae1
+        uses: github/codeql-action/init@v2
         with:
           languages: javascript # does not matter
       - uses: ./.github/actions/os-version
@@ -36,6 +36,8 @@ jobs:
             ~/.cargo/git
             ql/target
           key: ${{ runner.os }}-${{ steps.os_version.outputs.version }}-qltest-cargo-${{ hashFiles('ql/rust-toolchain.toml', 'ql/**/Cargo.lock') }}
+      - name: Check formatting
+        run: cd ql; cargo fmt --all -- --check
       - name: Build extractor
         run: |
           cd ql;
@@ -67,7 +69,7 @@ jobs:
           echo "/usr/local/opt/gnu-tar/libexec/gnubin" >> $GITHUB_PATH
       - name: Find codeql
         id: find-codeql
-        uses: github/codeql-action/init@beae46e6b1da530ed5e9fc6a756f92433ca47ae1
+        uses: github/codeql-action/init@v2
         with:
           languages: javascript # does not matter
       - uses: ./.github/actions/os-version

.github/workflows/ruby-build.yml

@@ -55,12 +55,12 @@ jobs:
         id: cache-extractor
         with:
           path: |
-            ruby/target/release/ruby-autobuilder
-            ruby/target/release/ruby-autobuilder.exe
-            ruby/target/release/ruby-extractor
-            ruby/target/release/ruby-extractor.exe
-            ruby/ql/lib/codeql/ruby/ast/internal/TreeSitter.qll
-          key: ${{ runner.os }}-${{ steps.os_version.outputs.version }}-ruby-extractor-${{ hashFiles('ruby/rust-toolchain.toml', 'ruby/**/Cargo.lock') }}--${{ hashFiles('ruby/**/*.rs') }}
+            ruby/extractor/target/release/autobuilder
+            ruby/extractor/target/release/autobuilder.exe
+            ruby/extractor/target/release/extractor
+            ruby/extractor/target/release/extractor.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('ruby/extractor/**/*.rs') }}
       - uses: actions/cache@v3
         if: steps.cache-extractor.outputs.cache-hit != 'true'
         with:
@@ -68,22 +68,22 @@ jobs:
             ~/.cargo/registry
             ~/.cargo/git
             ruby/target
-          key: ${{ runner.os }}-${{ steps.os_version.outputs.version }}-ruby-rust-cargo-${{ hashFiles('ruby/rust-toolchain.toml', 'ruby/**/Cargo.lock') }}
+          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: cargo fmt --all -- --check
+        run: cd extractor && cargo fmt --all -- --check
       - name: Build
         if: steps.cache-extractor.outputs.cache-hit != 'true'
-        run: cargo build --verbose
+        run: cd extractor && cargo build --verbose
       - name: Run tests
         if: steps.cache-extractor.outputs.cache-hit != 'true'
-        run: cargo test --verbose
+        run: cd extractor && cargo test --verbose
       - name: Release build
         if: steps.cache-extractor.outputs.cache-hit != 'true'
-        run: cargo build --release
+        run: cd extractor && cargo build --release
       - name: Generate dbscheme
         if: ${{ matrix.os == 'ubuntu-latest' && steps.cache-extractor.outputs.cache-hit != 'true'}}
-        run: target/release/ruby-generator --dbscheme ql/lib/ruby.dbscheme --library ql/lib/codeql/ruby/ast/internal/TreeSitter.qll
+        run: extractor/target/release/generator --dbscheme ql/lib/ruby.dbscheme --library ql/lib/codeql/ruby/ast/internal/TreeSitter.qll
       - uses: actions/upload-artifact@v3
         if: ${{ matrix.os == 'ubuntu-latest' }}
         with:
@@ -98,10 +98,10 @@ jobs:
         with:
           name: extractor-${{ matrix.os }}
           path: |
-            ruby/target/release/ruby-autobuilder
-            ruby/target/release/ruby-autobuilder.exe
-            ruby/target/release/ruby-extractor
-            ruby/target/release/ruby-extractor.exe
+            ruby/extractor/target/release/autobuilder
+            ruby/extractor/target/release/autobuilder.exe
+            ruby/extractor/target/release/extractor
+            ruby/extractor/target/release/extractor.exe
           retention-days: 1
   compile-queries:
     runs-on: ubuntu-latest-xl
@@ -116,21 +116,22 @@ jobs:
           key: ruby-build
       - name: Build Query Pack
         run: |
-          rm -rf target/packs
-          codeql pack create ../misc/suite-helpers --output target/packs
-          codeql pack create ../shared/regex --output target/packs
-          codeql pack create ../shared/ssa --output target/packs
-          codeql pack create ../shared/tutorial --output target/packs
-          codeql pack create ql/lib --output target/packs
-          codeql pack create -j0 ql/src --output target/packs --compilation-cache "${{ steps.query-cache.outputs.cache-dir }}"
-          PACK_FOLDER=$(readlink -f target/packs/codeql/ruby-queries/*)
+          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@v3
         with:
           name: codeql-ruby-queries
           path: |
-            ruby/target/packs/*
+            ${{ runner.temp }}/query-packs/*
           retention-days: 1
 
   package:
@@ -158,12 +159,12 @@ jobs:
           mkdir -p ruby
           cp -r codeql-extractor.yml tools ql/lib/ruby.dbscheme.stats ruby/
           mkdir -p ruby/tools/{linux64,osx64,win64}
-          cp linux64/ruby-autobuilder ruby/tools/linux64/autobuilder
-          cp osx64/ruby-autobuilder ruby/tools/osx64/autobuilder
-          cp win64/ruby-autobuilder.exe ruby/tools/win64/autobuilder.exe
-          cp linux64/ruby-extractor ruby/tools/linux64/extractor
-          cp osx64/ruby-extractor ruby/tools/osx64/extractor
-          cp win64/ruby-extractor.exe ruby/tools/win64/extractor.exe
+          cp linux64/autobuilder ruby/tools/linux64/autobuilder
+          cp osx64/autobuilder ruby/tools/osx64/autobuilder
+          cp win64/autobuilder.exe ruby/tools/win64/autobuilder.exe
+          cp linux64/extractor ruby/tools/linux64/extractor
+          cp osx64/extractor ruby/tools/osx64/extractor
+          cp win64/extractor.exe ruby/tools/win64/extractor.exe
           chmod +x ruby/tools/{linux64,osx64}/{autobuilder,extractor}
           zip -rq codeql-ruby.zip ruby
       - uses: actions/upload-artifact@v3

.github/workflows/swift.yml

@@ -5,6 +5,7 @@ on:
     paths:
       - "swift/**"
       - "misc/bazel/**"
+      - "misc/codegen/**"
       - "*.bazel*"
       - .github/workflows/swift.yml
       - .github/actions/**
@@ -19,6 +20,7 @@ on:
     paths:
       - "swift/**"
       - "misc/bazel/**"
+      - "misc/codegen/**"
       - "*.bazel*"
       - .github/workflows/swift.yml
       - .github/actions/**

.pre-commit-config.yaml

@@ -53,5 +53,5 @@ repos:
         name: Run Swift code generation unit tests
         files: ^swift/codegen/.*\.py$
         language: system
-        entry: bazel test //swift/codegen/test
+        entry: bazel test //misc/codegen/test
         pass_filenames: false

CODEOWNERS

@@ -6,6 +6,7 @@
 /python/ @github/codeql-python
 /ruby/ @github/codeql-ruby
 /swift/ @github/codeql-swift
+/misc/codegen/ @github/codeql-swift
 /java/kotlin-extractor/ @github/codeql-kotlin
 /java/kotlin-explorer/ @github/codeql-kotlin
 

config/identical-files.json

@@ -1,66 +1,86 @@
 {
   "DataFlow Java/C++/C#/Go/Python/Ruby/Swift": [
+    "java/ql/lib/semmle/code/java/dataflow/internal/DataFlow.qll",
+    "cpp/ql/lib/semmle/code/cpp/dataflow/internal/DataFlow.qll",
+    "cpp/ql/lib/semmle/code/cpp/ir/dataflow/internal/DataFlow.qll",
+    "csharp/ql/lib/semmle/code/csharp/dataflow/internal/DataFlow.qll",
+    "go/ql/lib/semmle/go/dataflow/internal/DataFlow.qll",
+    "python/ql/lib/semmle/python/dataflow/new/internal/DataFlow.qll",
+    "ruby/ql/lib/codeql/ruby/dataflow/internal/DataFlow.qll",
+    "swift/ql/lib/codeql/swift/dataflow/internal/DataFlow.qll"
+  ],
+  "DataFlowImpl Java/C++/C#/Go/Python/Ruby/Swift": [
     "java/ql/lib/semmle/code/java/dataflow/internal/DataFlowImpl.qll",
+    "cpp/ql/lib/semmle/code/cpp/dataflow/internal/DataFlowImpl.qll",
+    "cpp/ql/lib/semmle/code/cpp/ir/dataflow/internal/DataFlowImpl.qll",
+    "csharp/ql/lib/semmle/code/csharp/dataflow/internal/DataFlowImpl.qll",
+    "go/ql/lib/semmle/go/dataflow/internal/DataFlowImpl.qll",
+    "python/ql/lib/semmle/python/dataflow/new/internal/DataFlowImpl.qll",
+    "ruby/ql/lib/codeql/ruby/dataflow/internal/DataFlowImpl.qll",
+    "swift/ql/lib/codeql/swift/dataflow/internal/DataFlowImpl.qll"
+  ],
+  "DataFlow Java/C++/C#/Go/Python/Ruby/Swift Legacy Configuration": [
+    "java/ql/lib/semmle/code/java/dataflow/internal/DataFlowImpl1.qll",
     "java/ql/lib/semmle/code/java/dataflow/internal/DataFlowImpl2.qll",
     "java/ql/lib/semmle/code/java/dataflow/internal/DataFlowImpl3.qll",
     "java/ql/lib/semmle/code/java/dataflow/internal/DataFlowImpl4.qll",
     "java/ql/lib/semmle/code/java/dataflow/internal/DataFlowImpl5.qll",
     "java/ql/lib/semmle/code/java/dataflow/internal/DataFlowImpl6.qll",
-    "java/ql/lib/semmle/code/java/dataflow/internal/DataFlowImplForSerializability.qll",
-    "java/ql/lib/semmle/code/java/dataflow/internal/DataFlowImplForOnActivityResult.qll",
-    "cpp/ql/lib/semmle/code/cpp/dataflow/internal/DataFlowImpl.qll",
+    "cpp/ql/lib/semmle/code/cpp/dataflow/internal/DataFlowImpl1.qll",
     "cpp/ql/lib/semmle/code/cpp/dataflow/internal/DataFlowImpl2.qll",
     "cpp/ql/lib/semmle/code/cpp/dataflow/internal/DataFlowImpl3.qll",
     "cpp/ql/lib/semmle/code/cpp/dataflow/internal/DataFlowImpl4.qll",
     "cpp/ql/lib/semmle/code/cpp/dataflow/internal/DataFlowImplLocal.qll",
-    "cpp/ql/lib/semmle/code/cpp/ir/dataflow/internal/DataFlowImpl.qll",
+    "cpp/ql/lib/semmle/code/cpp/ir/dataflow/internal/DataFlowImpl1.qll",
     "cpp/ql/lib/semmle/code/cpp/ir/dataflow/internal/DataFlowImpl2.qll",
     "cpp/ql/lib/semmle/code/cpp/ir/dataflow/internal/DataFlowImpl3.qll",
     "cpp/ql/lib/semmle/code/cpp/ir/dataflow/internal/DataFlowImpl4.qll",
-    "cpp/ql/lib/experimental/semmle/code/cpp/ir/dataflow/internal/DataFlowImpl.qll",
-    "cpp/ql/lib/experimental/semmle/code/cpp/ir/dataflow/internal/DataFlowImpl2.qll",
-    "cpp/ql/lib/experimental/semmle/code/cpp/ir/dataflow/internal/DataFlowImpl3.qll",
-    "cpp/ql/lib/experimental/semmle/code/cpp/ir/dataflow/internal/DataFlowImpl4.qll",
-    "csharp/ql/lib/semmle/code/csharp/dataflow/internal/DataFlowImpl.qll",
+    "csharp/ql/lib/semmle/code/csharp/dataflow/internal/DataFlowImpl1.qll",
     "csharp/ql/lib/semmle/code/csharp/dataflow/internal/DataFlowImpl2.qll",
     "csharp/ql/lib/semmle/code/csharp/dataflow/internal/DataFlowImpl3.qll",
     "csharp/ql/lib/semmle/code/csharp/dataflow/internal/DataFlowImpl4.qll",
     "csharp/ql/lib/semmle/code/csharp/dataflow/internal/DataFlowImpl5.qll",
     "csharp/ql/lib/semmle/code/csharp/dataflow/internal/DataFlowImplForContentDataFlow.qll",
-    "go/ql/lib/semmle/go/dataflow/internal/DataFlowImpl.qll",
+    "go/ql/lib/semmle/go/dataflow/internal/DataFlowImpl1.qll",
     "go/ql/lib/semmle/go/dataflow/internal/DataFlowImpl2.qll",
     "go/ql/lib/semmle/go/dataflow/internal/DataFlowImplForStringsNewReplacer.qll",
-    "python/ql/lib/semmle/python/dataflow/new/internal/DataFlowImpl.qll",
+    "python/ql/lib/semmle/python/dataflow/new/internal/DataFlowImpl1.qll",
     "python/ql/lib/semmle/python/dataflow/new/internal/DataFlowImpl2.qll",
     "python/ql/lib/semmle/python/dataflow/new/internal/DataFlowImpl3.qll",
     "python/ql/lib/semmle/python/dataflow/new/internal/DataFlowImpl4.qll",
     "python/ql/lib/semmle/python/dataflow/new/internal/DataFlowImplForRegExp.qll",
-    "ruby/ql/lib/codeql/ruby/dataflow/internal/DataFlowImpl.qll",
+    "ruby/ql/lib/codeql/ruby/dataflow/internal/DataFlowImpl1.qll",
     "ruby/ql/lib/codeql/ruby/dataflow/internal/DataFlowImpl2.qll",
     "ruby/ql/lib/codeql/ruby/dataflow/internal/DataFlowImplForHttpClientLibraries.qll",
     "ruby/ql/lib/codeql/ruby/dataflow/internal/DataFlowImplForPathname.qll",
-    "swift/ql/lib/codeql/swift/dataflow/internal/DataFlowImpl.qll"
+    "swift/ql/lib/codeql/swift/dataflow/internal/DataFlowImpl1.qll"
   ],
   "DataFlow Java/C++/C#/Go/Python/Ruby/Swift Common": [
     "java/ql/lib/semmle/code/java/dataflow/internal/DataFlowImplCommon.qll",
     "cpp/ql/lib/semmle/code/cpp/dataflow/internal/DataFlowImplCommon.qll",
     "cpp/ql/lib/semmle/code/cpp/ir/dataflow/internal/DataFlowImplCommon.qll",
-    "cpp/ql/lib/experimental/semmle/code/cpp/ir/dataflow/internal/DataFlowImplCommon.qll",
     "csharp/ql/lib/semmle/code/csharp/dataflow/internal/DataFlowImplCommon.qll",
     "go/ql/lib/semmle/go/dataflow/internal/DataFlowImplCommon.qll",
     "python/ql/lib/semmle/python/dataflow/new/internal/DataFlowImplCommon.qll",
     "ruby/ql/lib/codeql/ruby/dataflow/internal/DataFlowImplCommon.qll",
     "swift/ql/lib/codeql/swift/dataflow/internal/DataFlowImplCommon.qll"
   ],
-  "TaintTracking::Configuration Java/C++/C#/Go/Python/Ruby/Swift": [
+  "TaintTracking Java/C++/C#/Go/Python/Ruby/Swift": [
+    "cpp/ql/lib/semmle/code/cpp/dataflow/internal/tainttracking1/TaintTracking.qll",
+    "cpp/ql/lib/semmle/code/cpp/ir/dataflow/internal/tainttracking1/TaintTracking.qll",
+    "csharp/ql/lib/semmle/code/csharp/dataflow/internal/tainttracking1/TaintTracking.qll",
+    "go/ql/lib/semmle/go/dataflow/internal/tainttracking1/TaintTracking.qll",
+    "java/ql/lib/semmle/code/java/dataflow/internal/tainttracking1/TaintTracking.qll",
+    "python/ql/lib/semmle/python/dataflow/new/internal/tainttracking1/TaintTracking.qll",
+    "ruby/ql/lib/codeql/ruby/dataflow/internal/tainttracking1/TaintTracking.qll",
+    "swift/ql/lib/codeql/swift/dataflow/internal/tainttracking1/TaintTracking.qll"
+  ],
+  "TaintTracking Legacy Configuration Java/C++/C#/Go/Python/Ruby/Swift": [
     "cpp/ql/lib/semmle/code/cpp/dataflow/internal/tainttracking1/TaintTrackingImpl.qll",
     "cpp/ql/lib/semmle/code/cpp/dataflow/internal/tainttracking2/TaintTrackingImpl.qll",
     "cpp/ql/lib/semmle/code/cpp/ir/dataflow/internal/tainttracking1/TaintTrackingImpl.qll",
     "cpp/ql/lib/semmle/code/cpp/ir/dataflow/internal/tainttracking2/TaintTrackingImpl.qll",
     "cpp/ql/lib/semmle/code/cpp/ir/dataflow/internal/tainttracking3/TaintTrackingImpl.qll",
-    "cpp/ql/lib/experimental/semmle/code/cpp/ir/dataflow/internal/tainttracking1/TaintTrackingImpl.qll",
-    "cpp/ql/lib/experimental/semmle/code/cpp/ir/dataflow/internal/tainttracking2/TaintTrackingImpl.qll",
-    "cpp/ql/lib/experimental/semmle/code/cpp/ir/dataflow/internal/tainttracking3/TaintTrackingImpl.qll",
     "csharp/ql/lib/semmle/code/csharp/dataflow/internal/tainttracking1/TaintTrackingImpl.qll",
     "csharp/ql/lib/semmle/code/csharp/dataflow/internal/tainttracking2/TaintTrackingImpl.qll",
     "csharp/ql/lib/semmle/code/csharp/dataflow/internal/tainttracking3/TaintTrackingImpl.qll",
@@ -82,7 +102,6 @@
     "java/ql/lib/semmle/code/java/dataflow/internal/DataFlowImplConsistency.qll",
     "cpp/ql/lib/semmle/code/cpp/dataflow/internal/DataFlowImplConsistency.qll",
     "cpp/ql/lib/semmle/code/cpp/ir/dataflow/internal/DataFlowImplConsistency.qll",
-    "cpp/ql/lib/experimental/semmle/code/cpp/ir/dataflow/internal/DataFlowImplConsistency.qll",
     "csharp/ql/lib/semmle/code/csharp/dataflow/internal/DataFlowImplConsistency.qll",
     "python/ql/lib/semmle/python/dataflow/new/internal/DataFlowImplConsistency.qll",
     "ruby/ql/lib/codeql/ruby/dataflow/internal/DataFlowImplConsistency.qll",

cpp/autobuilder/Semmle.Autobuild.Cpp.Tests/BuildScripts.cs

@@ -1,5 +1,6 @@
 using Xunit;
 using Semmle.Autobuild.Shared;
+using Semmle.Util;
 using System.Collections.Generic;
 using System;
 using System.Linq;
@@ -75,6 +76,15 @@ namespace Semmle.Autobuild.Cpp.Tests
             throw new ArgumentException("Missing RunProcess " + pattern);
         }
 
+        int IBuildActions.RunProcess(string cmd, string args, string? workingDirectory, IDictionary<string, string>? env, BuildOutputHandler onOutput, BuildOutputHandler onError)
+        {
+            var ret = (this as IBuildActions).RunProcess(cmd, args, workingDirectory, env, out var stdout);
+
+            stdout.ForEach(line => onOutput(line));
+
+            return ret;
+        }
+
         public IList<string> DirectoryDeleteIn = new List<string>();
 
         void IBuildActions.DirectoryDelete(string dir, bool recursive)
@@ -131,6 +141,14 @@ namespace Semmle.Autobuild.Cpp.Tests
 
         bool IBuildActions.IsWindows() => IsWindows;
 
+        public bool IsMacOs { get; set; }
+
+        bool IBuildActions.IsMacOs() => IsMacOs;
+
+        public bool IsArm { get; set; }
+
+        bool IBuildActions.IsArm() => IsArm;
+
         string IBuildActions.PathCombine(params string[] parts)
         {
             return string.Join(IsWindows ? '\\' : '/', parts.Where(p => !string.IsNullOrWhiteSpace(p)));
@@ -176,6 +194,15 @@ namespace Semmle.Autobuild.Cpp.Tests
             if (!DownloadFiles.Contains((address, fileName)))
                 throw new ArgumentException($"Missing DownloadFile, {address}, {fileName}");
         }
+
+        public IDiagnosticsWriter CreateDiagnosticsWriter(string filename) => new TestDiagnosticWriter();
+    }
+
+    internal class TestDiagnosticWriter : IDiagnosticsWriter
+    {
+        public IList<DiagnosticMessage> Diagnostics { get; } = new List<DiagnosticMessage>();
+
+        public void AddEntry(DiagnosticMessage message) => this.Diagnostics.Add(message);
     }
 
     /// <summary>
@@ -235,6 +262,7 @@ namespace Semmle.Autobuild.Cpp.Tests
             Actions.GetEnvironmentVariable[$"CODEQL_EXTRACTOR_{codeqlUpperLanguage}_TRAP_DIR"] = "";
             Actions.GetEnvironmentVariable[$"CODEQL_EXTRACTOR_{codeqlUpperLanguage}_SOURCE_ARCHIVE_DIR"] = "";
             Actions.GetEnvironmentVariable[$"CODEQL_EXTRACTOR_{codeqlUpperLanguage}_ROOT"] = $@"C:\codeql\{codeqlUpperLanguage.ToLowerInvariant()}";
+            Actions.GetEnvironmentVariable[$"CODEQL_EXTRACTOR_{codeqlUpperLanguage}_DIAGNOSTIC_DIR"] = "";
             Actions.GetEnvironmentVariable["CODEQL_JAVA_HOME"] = @"C:\codeql\tools\java";
             Actions.GetEnvironmentVariable["CODEQL_PLATFORM"] = "win64";
             Actions.GetEnvironmentVariable["SEMMLE_DIST"] = @"C:\odasa";

cpp/autobuilder/Semmle.Autobuild.Cpp/CppAutobuilder.cs

@@ -1,4 +1,5 @@
 using Semmle.Autobuild.Shared;
+using Semmle.Util;
 
 namespace Semmle.Autobuild.Cpp
 {
@@ -21,7 +22,7 @@ namespace Semmle.Autobuild.Cpp
 
     public class CppAutobuilder : Autobuilder<CppAutobuildOptions>
     {
-        public CppAutobuilder(IBuildActions actions, CppAutobuildOptions options) : base(actions, options) { }
+        public CppAutobuilder(IBuildActions actions, CppAutobuildOptions options) : base(actions, options, new DiagnosticClassifier()) { }
 
         public override BuildScript GetBuildScript()
         {

cpp/ql/lib/CHANGELOG.md

@@ -1,3 +1,54 @@
+## 0.6.0
+
+### Breaking Changes
+
+* The `semmle.code.cpp.commons.Buffer` and `semmle.code.cpp.commons.NullTermination` libraries no longer expose `semmle.code.cpp.dataflow.DataFlow`. Please import `semmle.code.cpp.dataflow.DataFlow` directly.
+
+### Deprecated APIs
+
+* The `WriteConfig` taint tracking configuration has been deprecated. Please use `WriteFlow`.
+
+### New Features
+
+* Added support for merging two `PathGraph`s via disjoint union to allow results from multiple data flow computations in a single `path-problem` query.
+
+### Major Analysis Improvements
+
+* A new C/C++ dataflow library (`semmle.code.cpp.dataflow.new.DataFlow`) has been added.
+  The new library behaves much more like the dataflow library of other CodeQL supported
+  languages by following use-use dataflow paths instead of def-use dataflow paths.
+  The new library also better supports dataflow through indirections, and new predicates
+  such as `Node::asIndirectExpr` have been added to facilitate working with indirections.
+
+  The `semmle.code.cpp.ir.dataflow.DataFlow` library is now identical to the new
+  `semmle.code.cpp.dataflow.new.DataFlow` library.
+* The main data flow and taint tracking APIs have been changed. The old APIs
+  remain in place for now and translate to the new through a
+  backwards-compatible wrapper. If multiple configurations are in scope
+  simultaneously, then this may affect results slightly. The new API is quite
+  similar to the old, but makes use of a configuration module instead of a
+  configuration class.
+
+### Minor Analysis Improvements
+
+* Deleted the deprecated `hasGeneratedCopyConstructor` and `hasGeneratedCopyAssignmentOperator` predicates from the `Folder` class.
+* Deleted the deprecated `getPath` and `getFolder` predicates from the `XmlFile` class.
+* Deleted the deprecated `getMustlockFunction`, `getTrylockFunction`, `getLockFunction`, and `getUnlockFunction` predicates from the `MutexType` class.
+* Deleted the deprecated `getPosInBasicBlock` predicate from the `SubBasicBlock` class.
+* Deleted the deprecated `getExpr` predicate from the `PointerDereferenceExpr` class.
+* Deleted the deprecated `getUseInstruction` and `getDefinitionInstruction` predicates from the `Operand` class.
+* Deleted the deprecated `isInParameter`, `isInParameterPointer`, and `isInQualifier` predicates from the `FunctionInput` class.
+* Deleted the deprecated `isOutParameterPointer`, `isOutQualifier`, `isOutReturnValue`, and `isOutReturnPointer` predicate from the `FunctionOutput` class.
+* Deleted the deprecated 3-argument `isGuardPhi` predicate from the `RangeSsaDefinition` class.
+
+## 0.5.4
+
+No user-facing changes.
+
+## 0.5.3
+
+No user-facing changes.
+
 ## 0.5.2
 
 No user-facing changes.

cpp/ql/lib/change-notes/released/0.5.3.md

@@ -0,0 +1,3 @@
+## 0.5.3
+
+No user-facing changes.

cpp/ql/lib/change-notes/released/0.5.4.md

@@ -0,0 +1,3 @@
+## 0.5.4
+
+No user-facing changes.

cpp/ql/lib/change-notes/released/0.6.0.md

@@ -0,0 +1,42 @@
+## 0.6.0
+
+### Breaking Changes
+
+* The `semmle.code.cpp.commons.Buffer` and `semmle.code.cpp.commons.NullTermination` libraries no longer expose `semmle.code.cpp.dataflow.DataFlow`. Please import `semmle.code.cpp.dataflow.DataFlow` directly.
+
+### Deprecated APIs
+
+* The `WriteConfig` taint tracking configuration has been deprecated. Please use `WriteFlow`.
+
+### New Features
+
+* Added support for merging two `PathGraph`s via disjoint union to allow results from multiple data flow computations in a single `path-problem` query.
+
+### Major Analysis Improvements
+
+* A new C/C++ dataflow library (`semmle.code.cpp.dataflow.new.DataFlow`) has been added.
+  The new library behaves much more like the dataflow library of other CodeQL supported
+  languages by following use-use dataflow paths instead of def-use dataflow paths.
+  The new library also better supports dataflow through indirections, and new predicates
+  such as `Node::asIndirectExpr` have been added to facilitate working with indirections.
+
+  The `semmle.code.cpp.ir.dataflow.DataFlow` library is now identical to the new
+  `semmle.code.cpp.dataflow.new.DataFlow` library.
+* The main data flow and taint tracking APIs have been changed. The old APIs
+  remain in place for now and translate to the new through a
+  backwards-compatible wrapper. If multiple configurations are in scope
+  simultaneously, then this may affect results slightly. The new API is quite
+  similar to the old, but makes use of a configuration module instead of a
+  configuration class.
+
+### Minor Analysis Improvements
+
+* Deleted the deprecated `hasGeneratedCopyConstructor` and `hasGeneratedCopyAssignmentOperator` predicates from the `Folder` class.
+* Deleted the deprecated `getPath` and `getFolder` predicates from the `XmlFile` class.
+* Deleted the deprecated `getMustlockFunction`, `getTrylockFunction`, `getLockFunction`, and `getUnlockFunction` predicates from the `MutexType` class.
+* Deleted the deprecated `getPosInBasicBlock` predicate from the `SubBasicBlock` class.
+* Deleted the deprecated `getExpr` predicate from the `PointerDereferenceExpr` class.
+* Deleted the deprecated `getUseInstruction` and `getDefinitionInstruction` predicates from the `Operand` class.
+* Deleted the deprecated `isInParameter`, `isInParameterPointer`, and `isInQualifier` predicates from the `FunctionInput` class.
+* Deleted the deprecated `isOutParameterPointer`, `isOutQualifier`, `isOutReturnValue`, and `isOutReturnPointer` predicate from the `FunctionOutput` class.
+* Deleted the deprecated 3-argument `isGuardPhi` predicate from the `RangeSsaDefinition` class.

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

@@ -1,2 +1,2 @@
 ---
-lastReleaseVersion: 0.5.2
+lastReleaseVersion: 0.6.0

cpp/ql/lib/experimental/semmle/code/cpp/dataflow/ProductFlow.qll

@@ -1,5 +1,5 @@
-import experimental.semmle.code.cpp.ir.dataflow.DataFlow
-import experimental.semmle.code.cpp.ir.dataflow.DataFlow2
+import semmle.code.cpp.ir.dataflow.DataFlow
+import semmle.code.cpp.ir.dataflow.DataFlow2
 
 module ProductFlow {
   abstract class Configuration extends string {

cpp/ql/lib/experimental/semmle/code/cpp/ir/dataflow/ResolveCall.qll

@@ -1,23 +0,0 @@
-/**
- * Provides a predicate for non-contextual virtual dispatch and function
- * pointer resolution.
- */
-
-import cpp
-private import semmle.code.cpp.ir.ValueNumbering
-private import internal.DataFlowDispatch
-private import semmle.code.cpp.ir.IR
-
-/**
- * Resolve potential target function(s) for `call`.
- *
- * If `call` is a call through a function pointer (`ExprCall`) or its target is
- * a virtual member function, simple data flow analysis is performed in order
- * to identify the possible target(s).
- */
-Function resolveCall(Call call) {
-  exists(CallInstruction callInstruction |
-    callInstruction.getAst() = call and
-    result = viableCallable(callInstruction)
-  )
-}

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

@@ -1,273 +0,0 @@
-private import cpp
-private import semmle.code.cpp.ir.IR
-private import experimental.semmle.code.cpp.ir.dataflow.DataFlow
-private import experimental.semmle.code.cpp.ir.dataflow.internal.DataFlowPrivate
-private import experimental.semmle.code.cpp.ir.dataflow.internal.DataFlowUtil
-private import DataFlowImplCommon as DataFlowImplCommon
-
-/**
- * Gets a function that might be called by `call`.
- */
-cached
-Function viableCallable(CallInstruction call) {
-  DataFlowImplCommon::forceCachingInSameStage() and
-  result = call.getStaticCallTarget()
-  or
-  // If the target of the call does not have a body in the snapshot, it might
-  // be because the target is just a header declaration, and the real target
-  // will be determined at run time when the caller and callee are linked
-  // together by the operating system's dynamic linker. In case a _unique_
-  // function with the right signature is present in the database, we return
-  // that as a potential callee.
-  exists(string qualifiedName, int nparams |
-    callSignatureWithoutBody(qualifiedName, nparams, call) and
-    functionSignatureWithBody(qualifiedName, nparams, result) and
-    strictcount(Function other | functionSignatureWithBody(qualifiedName, nparams, other)) = 1
-  )
-  or
-  // Virtual dispatch
-  result = call.(VirtualDispatch::DataSensitiveCall).resolve()
-}
-
-/**
- * Provides virtual dispatch support compatible with the original
- * implementation of `semmle.code.cpp.security.TaintTracking`.
- */
-private module VirtualDispatch {
-  /** A call that may dispatch differently depending on the qualifier value. */
-  abstract class DataSensitiveCall extends DataFlowCall {
-    /**
-     * Gets the node whose value determines the target of this call. This node
-     * could be the qualifier of a virtual dispatch or the function-pointer
-     * expression in a call to a function pointer. What they have in common is
-     * that we need to find out which data flows there, and then it's up to the
-     * `resolve` predicate to stitch that information together and resolve the
-     * call.
-     */
-    abstract DataFlow::Node getDispatchValue();
-
-    /** Gets a candidate target for this call. */
-    abstract Function resolve();
-
-    /**
-     * Whether `src` can flow to this call.
-     *
-     * Searches backwards from `getDispatchValue()` to `src`. The `allowFromArg`
-     * parameter is true when the search is allowed to continue backwards into
-     * a parameter; non-recursive callers should pass `_` for `allowFromArg`.
-     */
-    predicate flowsFrom(DataFlow::Node src, boolean allowFromArg) {
-      src = this.getDispatchValue() and allowFromArg = true
-      or
-      exists(DataFlow::Node other, boolean allowOtherFromArg |
-        this.flowsFrom(other, allowOtherFromArg)
-      |
-        // Call argument
-        exists(DataFlowCall call, Position i |
-          other
-              .(DataFlow::ParameterNode)
-              .isParameterOf(pragma[only_bind_into](call).getStaticCallTarget(), i) and
-          src.(ArgumentNode).argumentOf(call, pragma[only_bind_into](pragma[only_bind_out](i)))
-        ) and
-        allowOtherFromArg = true and
-        allowFromArg = true
-        or
-        // Call return
-        exists(DataFlowCall call, ReturnKind returnKind |
-          other = getAnOutNode(call, returnKind) and
-          returnNodeWithKindAndEnclosingCallable(src, returnKind, call.getStaticCallTarget())
-        ) and
-        allowFromArg = false
-        or
-        // Local flow
-        DataFlow::localFlowStep(src, other) and
-        allowFromArg = allowOtherFromArg
-        or
-        // Flow from global variable to load.
-        exists(LoadInstruction load, GlobalOrNamespaceVariable var |
-          var = src.asVariable() and
-          other.asInstruction() = load and
-          addressOfGlobal(load.getSourceAddress(), var) and
-          // The `allowFromArg` concept doesn't play a role when `src` is a
-          // global variable, so we just set it to a single arbitrary value for
-          // performance.
-          allowFromArg = true
-        )
-        or
-        // Flow from store to global variable.
-        exists(StoreInstruction store, GlobalOrNamespaceVariable var |
-          var = other.asVariable() and
-          store = src.asInstruction() and
-          storeIntoGlobal(store, var) and
-          // Setting `allowFromArg` to `true` like in the base case means we
-          // treat a store to a global variable like the dispatch itself: flow
-          // may come from anywhere.
-          allowFromArg = true
-        )
-      )
-    }
-  }
-
-  pragma[noinline]
-  private predicate storeIntoGlobal(StoreInstruction store, GlobalOrNamespaceVariable var) {
-    addressOfGlobal(store.getDestinationAddress(), var)
-  }
-
-  /** Holds if `addressInstr` is an instruction that produces the address of `var`. */
-  private predicate addressOfGlobal(Instruction addressInstr, GlobalOrNamespaceVariable var) {
-    // Access directly to the global variable
-    addressInstr.(VariableAddressInstruction).getAstVariable() = var
-    or
-    // Access to a field on a global union
-    exists(FieldAddressInstruction fa |
-      fa = addressInstr and
-      fa.getObjectAddress().(VariableAddressInstruction).getAstVariable() = var and
-      fa.getField().getDeclaringType() instanceof Union
-    )
-  }
-
-  /**
-   * A ReturnNode with its ReturnKind and its enclosing callable.
-   *
-   * Used to fix a join ordering issue in flowsFrom.
-   */
-  pragma[noinline]
-  private predicate returnNodeWithKindAndEnclosingCallable(
-    ReturnNode node, ReturnKind kind, DataFlowCallable callable
-  ) {
-    node.getKind() = kind and
-    node.getEnclosingCallable() = callable
-  }
-
-  /** Call through a function pointer. */
-  private class DataSensitiveExprCall extends DataSensitiveCall {
-    DataSensitiveExprCall() { not exists(this.getStaticCallTarget()) }
-
-    override DataFlow::Node getDispatchValue() { result.asInstruction() = this.getCallTarget() }
-
-    override Function resolve() {
-      exists(FunctionInstruction fi |
-        this.flowsFrom(DataFlow::instructionNode(fi), _) and
-        result = fi.getFunctionSymbol()
-      ) and
-      (
-        this.getNumberOfArguments() <= result.getEffectiveNumberOfParameters() and
-        this.getNumberOfArguments() >= result.getEffectiveNumberOfParameters()
-        or
-        result.isVarargs()
-      )
-    }
-  }
-
-  /** Call to a virtual function. */
-  private class DataSensitiveOverriddenFunctionCall extends DataSensitiveCall {
-    DataSensitiveOverriddenFunctionCall() {
-      exists(this.getStaticCallTarget().(VirtualFunction).getAnOverridingFunction())
-    }
-
-    override DataFlow::Node getDispatchValue() { result.asInstruction() = this.getThisArgument() }
-
-    override MemberFunction resolve() {
-      exists(Class overridingClass |
-        this.overrideMayAffectCall(overridingClass, result) and
-        this.hasFlowFromCastFrom(overridingClass)
-      )
-    }
-
-    /**
-     * Holds if `this` is a virtual function call whose static target is
-     * overridden by `overridingFunction` in `overridingClass`.
-     */
-    pragma[noinline]
-    private predicate overrideMayAffectCall(Class overridingClass, MemberFunction overridingFunction) {
-      overridingFunction.getAnOverriddenFunction+() = this.getStaticCallTarget().(VirtualFunction) and
-      overridingFunction.getDeclaringType() = overridingClass
-    }
-
-    /**
-     * Holds if the qualifier of `this` has flow from an upcast from
-     * `derivedClass`.
-     */
-    pragma[noinline]
-    private predicate hasFlowFromCastFrom(Class derivedClass) {
-      exists(ConvertToBaseInstruction toBase |
-        this.flowsFrom(DataFlow::instructionNode(toBase), _) and
-        derivedClass = toBase.getDerivedClass()
-      )
-    }
-  }
-}
-
-/**
- * Holds if `f` is a function with a body that has name `qualifiedName` and
- * `nparams` parameter count. See `functionSignature`.
- */
-private predicate functionSignatureWithBody(string qualifiedName, int nparams, Function f) {
-  functionSignature(f, qualifiedName, nparams) and
-  exists(f.getBlock())
-}
-
-/**
- * Holds if the target of `call` is a function _with no definition_ that has
- * name `qualifiedName` and `nparams` parameter count. See `functionSignature`.
- */
-pragma[noinline]
-private predicate callSignatureWithoutBody(string qualifiedName, int nparams, CallInstruction call) {
-  exists(Function target |
-    target = call.getStaticCallTarget() and
-    not exists(target.getBlock()) and
-    functionSignature(target, qualifiedName, nparams)
-  )
-}
-
-/**
- * Holds if `f` has name `qualifiedName` and `nparams` parameter count. This is
- * an approximation of its signature for the purpose of matching functions that
- * might be the same across link targets.
- */
-private predicate functionSignature(Function f, string qualifiedName, int nparams) {
-  qualifiedName = f.getQualifiedName() and
-  nparams = f.getNumberOfParameters() and
-  not f.isStatic()
-}
-
-/**
- * Holds if the set of viable implementations that can be called by `call`
- * might be improved by knowing the call context.
- */
-predicate mayBenefitFromCallContext(CallInstruction call, Function f) {
-  mayBenefitFromCallContext(call, f, _)
-}
-
-/**
- * Holds if `call` is a call through a function pointer, and the pointer
- * value is given as the `arg`'th argument to `f`.
- */
-private predicate mayBenefitFromCallContext(
-  VirtualDispatch::DataSensitiveCall call, Function f, int arg
-) {
-  f = pragma[only_bind_out](call).getEnclosingCallable() and
-  exists(InitializeParameterInstruction init |
-    not exists(call.getStaticCallTarget()) and
-    init.getEnclosingFunction() = f and
-    call.flowsFrom(DataFlow::instructionNode(init), _) and
-    init.getParameter().getIndex() = arg
-  )
-}
-
-/**
- * Gets a viable dispatch target of `call` in the context `ctx`. This is
- * restricted to those `call`s for which a context might make a difference.
- */
-Function viableImplInCallContext(CallInstruction call, CallInstruction ctx) {
-  result = viableCallable(call) and
-  exists(int i, Function f |
-    mayBenefitFromCallContext(pragma[only_bind_into](call), f, i) and
-    f = ctx.getStaticCallTarget() and
-    result = ctx.getArgument(i).getUnconvertedResultExpression().(FunctionAccess).getTarget()
-  )
-}
-
-/** Holds if arguments at position `apos` match parameters at position `ppos`. */
-pragma[inline]
-predicate parameterMatch(ParameterPosition ppos, ArgumentPosition apos) { ppos = apos }

cpp/ql/lib/experimental/semmle/code/cpp/ir/dataflow/internal/DataFlowImplConsistency.qll

@@ -1,278 +0,0 @@
-/**
- * Provides consistency queries for checking invariants in the language-specific
- * data-flow classes and predicates.
- */
-
-private import DataFlowImplSpecific::Private
-private import DataFlowImplSpecific::Public
-private import tainttracking1.TaintTrackingParameter::Private
-private import tainttracking1.TaintTrackingParameter::Public
-
-module Consistency {
-  private newtype TConsistencyConfiguration = MkConsistencyConfiguration()
-
-  /** A class for configuring the consistency queries. */
-  class ConsistencyConfiguration extends TConsistencyConfiguration {
-    string toString() { none() }
-
-    /** Holds if `n` should be excluded from the consistency test `uniqueEnclosingCallable`. */
-    predicate uniqueEnclosingCallableExclude(Node n) { none() }
-
-    /** Holds if `n` should be excluded from the consistency test `uniqueNodeLocation`. */
-    predicate uniqueNodeLocationExclude(Node n) { none() }
-
-    /** Holds if `n` should be excluded from the consistency test `missingLocation`. */
-    predicate missingLocationExclude(Node n) { none() }
-
-    /** Holds if `n` should be excluded from the consistency test `postWithInFlow`. */
-    predicate postWithInFlowExclude(Node n) { none() }
-
-    /** Holds if `n` should be excluded from the consistency test `argHasPostUpdate`. */
-    predicate argHasPostUpdateExclude(ArgumentNode n) { none() }
-
-    /** Holds if `n` should be excluded from the consistency test `reverseRead`. */
-    predicate reverseReadExclude(Node n) { none() }
-
-    /** Holds if `n` should be excluded from the consistency test `postHasUniquePre`. */
-    predicate postHasUniquePreExclude(PostUpdateNode n) { none() }
-
-    /** Holds if `n` should be excluded from the consistency test `uniquePostUpdate`. */
-    predicate uniquePostUpdateExclude(Node n) { none() }
-
-    /** Holds if `(call, ctx)` should be excluded from the consistency test `viableImplInCallContextTooLargeExclude`. */
-    predicate viableImplInCallContextTooLargeExclude(
-      DataFlowCall call, DataFlowCall ctx, DataFlowCallable callable
-    ) {
-      none()
-    }
-
-    /** Holds if `(c, pos, p)` should be excluded from the consistency test `uniqueParameterNodeAtPosition`. */
-    predicate uniqueParameterNodeAtPositionExclude(DataFlowCallable c, ParameterPosition pos, Node p) {
-      none()
-    }
-
-    /** Holds if `(c, pos, p)` should be excluded from the consistency test `uniqueParameterNodePosition`. */
-    predicate uniqueParameterNodePositionExclude(DataFlowCallable c, ParameterPosition pos, Node p) {
-      none()
-    }
-  }
-
-  private class RelevantNode extends Node {
-    RelevantNode() {
-      this instanceof ArgumentNode or
-      this instanceof ParameterNode or
-      this instanceof ReturnNode or
-      this = getAnOutNode(_, _) or
-      simpleLocalFlowStep(this, _) or
-      simpleLocalFlowStep(_, this) or
-      jumpStep(this, _) or
-      jumpStep(_, this) or
-      storeStep(this, _, _) or
-      storeStep(_, _, this) or
-      readStep(this, _, _) or
-      readStep(_, _, this) or
-      defaultAdditionalTaintStep(this, _) or
-      defaultAdditionalTaintStep(_, this)
-    }
-  }
-
-  query predicate uniqueEnclosingCallable(Node n, string msg) {
-    exists(int c |
-      n instanceof RelevantNode and
-      c = count(nodeGetEnclosingCallable(n)) and
-      c != 1 and
-      not any(ConsistencyConfiguration conf).uniqueEnclosingCallableExclude(n) and
-      msg = "Node should have one enclosing callable but has " + c + "."
-    )
-  }
-
-  query predicate uniqueType(Node n, string msg) {
-    exists(int c |
-      n instanceof RelevantNode and
-      c = count(getNodeType(n)) and
-      c != 1 and
-      msg = "Node should have one type but has " + c + "."
-    )
-  }
-
-  query predicate uniqueNodeLocation(Node n, string msg) {
-    exists(int c |
-      c =
-        count(string filepath, int startline, int startcolumn, int endline, int endcolumn |
-          n.hasLocationInfo(filepath, startline, startcolumn, endline, endcolumn)
-        ) and
-      c != 1 and
-      not any(ConsistencyConfiguration conf).uniqueNodeLocationExclude(n) and
-      msg = "Node should have one location but has " + c + "."
-    )
-  }
-
-  query predicate missingLocation(string msg) {
-    exists(int c |
-      c =
-        strictcount(Node n |
-          not n.hasLocationInfo(_, _, _, _, _) and
-          not any(ConsistencyConfiguration conf).missingLocationExclude(n)
-        ) and
-      msg = "Nodes without location: " + c
-    )
-  }
-
-  query predicate uniqueNodeToString(Node n, string msg) {
-    exists(int c |
-      c = count(n.toString()) and
-      c != 1 and
-      msg = "Node should have one toString but has " + c + "."
-    )
-  }
-
-  query predicate missingToString(string msg) {
-    exists(int c |
-      c = strictcount(Node n | not exists(n.toString())) and
-      msg = "Nodes without toString: " + c
-    )
-  }
-
-  query predicate parameterCallable(ParameterNode p, string msg) {
-    exists(DataFlowCallable c | isParameterNode(p, c, _) and c != nodeGetEnclosingCallable(p)) and
-    msg = "Callable mismatch for parameter."
-  }
-
-  query predicate localFlowIsLocal(Node n1, Node n2, string msg) {
-    simpleLocalFlowStep(n1, n2) and
-    nodeGetEnclosingCallable(n1) != nodeGetEnclosingCallable(n2) and
-    msg = "Local flow step does not preserve enclosing callable."
-  }
-
-  query predicate readStepIsLocal(Node n1, Node n2, string msg) {
-    readStep(n1, _, n2) and
-    nodeGetEnclosingCallable(n1) != nodeGetEnclosingCallable(n2) and
-    msg = "Read step does not preserve enclosing callable."
-  }
-
-  query predicate storeStepIsLocal(Node n1, Node n2, string msg) {
-    storeStep(n1, _, n2) and
-    nodeGetEnclosingCallable(n1) != nodeGetEnclosingCallable(n2) and
-    msg = "Store step does not preserve enclosing callable."
-  }
-
-  private DataFlowType typeRepr() { result = getNodeType(_) }
-
-  query predicate compatibleTypesReflexive(DataFlowType t, string msg) {
-    t = typeRepr() and
-    not compatibleTypes(t, t) and
-    msg = "Type compatibility predicate is not reflexive."
-  }
-
-  query predicate unreachableNodeCCtx(Node n, DataFlowCall call, string msg) {
-    isUnreachableInCall(n, call) and
-    exists(DataFlowCallable c |
-      c = nodeGetEnclosingCallable(n) and
-      not viableCallable(call) = c
-    ) and
-    msg = "Call context for isUnreachableInCall is inconsistent with call graph."
-  }
-
-  query predicate localCallNodes(DataFlowCall call, Node n, string msg) {
-    (
-      n = getAnOutNode(call, _) and
-      msg = "OutNode and call does not share enclosing callable."
-      or
-      n.(ArgumentNode).argumentOf(call, _) and
-      msg = "ArgumentNode and call does not share enclosing callable."
-    ) and
-    nodeGetEnclosingCallable(n) != call.getEnclosingCallable()
-  }
-
-  // This predicate helps the compiler forget that in some languages
-  // it is impossible for a result of `getPreUpdateNode` to be an
-  // instance of `PostUpdateNode`.
-  private Node getPre(PostUpdateNode n) {
-    result = n.getPreUpdateNode()
-    or
-    none()
-  }
-
-  query predicate postIsNotPre(PostUpdateNode n, string msg) {
-    getPre(n) = n and
-    msg = "PostUpdateNode should not equal its pre-update node."
-  }
-
-  query predicate postHasUniquePre(PostUpdateNode n, string msg) {
-    not any(ConsistencyConfiguration conf).postHasUniquePreExclude(n) and
-    exists(int c |
-      c = count(n.getPreUpdateNode()) and
-      c != 1 and
-      msg = "PostUpdateNode should have one pre-update node but has " + c + "."
-    )
-  }
-
-  query predicate uniquePostUpdate(Node n, string msg) {
-    not any(ConsistencyConfiguration conf).uniquePostUpdateExclude(n) and
-    1 < strictcount(PostUpdateNode post | post.getPreUpdateNode() = n) and
-    msg = "Node has multiple PostUpdateNodes."
-  }
-
-  query predicate postIsInSameCallable(PostUpdateNode n, string msg) {
-    nodeGetEnclosingCallable(n) != nodeGetEnclosingCallable(n.getPreUpdateNode()) and
-    msg = "PostUpdateNode does not share callable with its pre-update node."
-  }
-
-  private predicate hasPost(Node n) { exists(PostUpdateNode post | post.getPreUpdateNode() = n) }
-
-  query predicate reverseRead(Node n, string msg) {
-    exists(Node n2 | readStep(n, _, n2) and hasPost(n2) and not hasPost(n)) and
-    not any(ConsistencyConfiguration conf).reverseReadExclude(n) and
-    msg = "Origin of readStep is missing a PostUpdateNode."
-  }
-
-  query predicate argHasPostUpdate(ArgumentNode n, string msg) {
-    not hasPost(n) and
-    not any(ConsistencyConfiguration c).argHasPostUpdateExclude(n) and
-    msg = "ArgumentNode is missing PostUpdateNode."
-  }
-
-  // This predicate helps the compiler forget that in some languages
-  // it is impossible for a `PostUpdateNode` to be the target of
-  // `simpleLocalFlowStep`.
-  private predicate isPostUpdateNode(Node n) { n instanceof PostUpdateNode or none() }
-
-  query predicate postWithInFlow(Node n, string msg) {
-    isPostUpdateNode(n) and
-    not clearsContent(n, _) and
-    simpleLocalFlowStep(_, n) and
-    not any(ConsistencyConfiguration c).postWithInFlowExclude(n) and
-    msg = "PostUpdateNode should not be the target of local flow."
-  }
-
-  query predicate viableImplInCallContextTooLarge(
-    DataFlowCall call, DataFlowCall ctx, DataFlowCallable callable
-  ) {
-    callable = viableImplInCallContext(call, ctx) and
-    not callable = viableCallable(call) and
-    not any(ConsistencyConfiguration c).viableImplInCallContextTooLargeExclude(call, ctx, callable)
-  }
-
-  query predicate uniqueParameterNodeAtPosition(
-    DataFlowCallable c, ParameterPosition pos, Node p, string msg
-  ) {
-    not any(ConsistencyConfiguration conf).uniqueParameterNodeAtPositionExclude(c, pos, p) and
-    isParameterNode(p, c, pos) and
-    not exists(unique(Node p0 | isParameterNode(p0, c, pos))) and
-    msg = "Parameters with overlapping positions."
-  }
-
-  query predicate uniqueParameterNodePosition(
-    DataFlowCallable c, ParameterPosition pos, Node p, string msg
-  ) {
-    not any(ConsistencyConfiguration conf).uniqueParameterNodePositionExclude(c, pos, p) and
-    isParameterNode(p, c, pos) and
-    not exists(unique(ParameterPosition pos0 | isParameterNode(p, c, pos0))) and
-    msg = "Parameter node with multiple positions."
-  }
-
-  query predicate uniqueContentApprox(Content c, string msg) {
-    not exists(unique(ContentApprox approx | approx = getContentApprox(c))) and
-    msg = "Non-unique content approximation."
-  }
-}

cpp/ql/lib/experimental/semmle/code/cpp/ir/dataflow/internal/DataFlowImplSpecific.qll

@@ -1,11 +0,0 @@
-/**
- * Provides IR-specific definitions for use in the data flow library.
- */
-module Private {
-  import DataFlowPrivate
-  import DataFlowDispatch
-}
-
-module Public {
-  import DataFlowUtil
-}

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

@@ -1,567 +0,0 @@
-private import cpp as Cpp
-private import DataFlowUtil
-private import semmle.code.cpp.ir.IR
-private import DataFlowDispatch
-private import DataFlowImplConsistency
-private import semmle.code.cpp.ir.internal.IRCppLanguage
-private import SsaInternals as Ssa
-
-/** Gets the callable in which this node occurs. */
-DataFlowCallable nodeGetEnclosingCallable(Node n) { result = n.getEnclosingCallable() }
-
-/** Holds if `p` is a `ParameterNode` of `c` with position `pos`. */
-predicate isParameterNode(ParameterNode p, DataFlowCallable c, ParameterPosition pos) {
-  p.isParameterOf(c, pos)
-}
-
-/** Holds if `arg` is an `ArgumentNode` of `c` with position `pos`. */
-predicate isArgumentNode(ArgumentNode arg, DataFlowCall c, ArgumentPosition pos) {
-  arg.argumentOf(c, pos)
-}
-
-/**
- * A data flow node that occurs as the argument of a call and is passed as-is
- * to the callable. Instance arguments (`this` pointer) and read side effects
- * on parameters are also included.
- */
-abstract class ArgumentNode extends Node {
-  /**
-   * Holds if this argument occurs at the given position in the given call.
-   * The instance argument is considered to have index `-1`.
-   */
-  abstract predicate argumentOf(DataFlowCall call, ArgumentPosition pos);
-
-  /** Gets the call in which this node is an argument. */
-  DataFlowCall getCall() { this.argumentOf(result, _) }
-}
-
-/**
- * A data flow node that occurs as the argument to a call, or an
- * implicit `this` pointer argument.
- */
-private class PrimaryArgumentNode extends ArgumentNode, OperandNode {
-  override ArgumentOperand op;
-
-  PrimaryArgumentNode() { exists(CallInstruction call | op = call.getAnArgumentOperand()) }
-
-  override predicate argumentOf(DataFlowCall call, ArgumentPosition pos) {
-    op = call.getArgumentOperand(pos.(DirectPosition).getIndex())
-  }
-
-  override string toStringImpl() { result = argumentOperandToString(op) }
-}
-
-private string argumentOperandToString(ArgumentOperand op) {
-  exists(Expr unconverted |
-    unconverted = op.getDef().getUnconvertedResultExpression() and
-    result = unconverted.toString()
-  )
-  or
-  // Certain instructions don't map to an unconverted result expression. For these cases
-  // we fall back to a simpler naming scheme. This can happen in IR-generated constructors.
-  not exists(op.getDef().getUnconvertedResultExpression()) and
-  (
-    result = "Argument " + op.(PositionalArgumentOperand).getIndex()
-    or
-    op instanceof ThisArgumentOperand and result = "Argument this"
-  )
-}
-
-private class SideEffectArgumentNode extends ArgumentNode, SideEffectOperandNode {
-  override predicate argumentOf(DataFlowCall dfCall, ArgumentPosition pos) {
-    this.getCallInstruction() = dfCall and
-    pos.(IndirectionPosition).getArgumentIndex() = this.getArgumentIndex() and
-    pos.(IndirectionPosition).getIndirectionIndex() = super.getIndirectionIndex()
-  }
-
-  override string toStringImpl() {
-    result = argumentOperandToString(this.getAddressOperand()) + " indirection"
-  }
-}
-
-/** A parameter position represented by an integer. */
-class ParameterPosition = Position;
-
-/** An argument position represented by an integer. */
-class ArgumentPosition = Position;
-
-class Position extends TPosition {
-  abstract string toString();
-}
-
-class DirectPosition extends Position, TDirectPosition {
-  int index;
-
-  DirectPosition() { this = TDirectPosition(index) }
-
-  override string toString() { if index = -1 then result = "this" else result = index.toString() }
-
-  int getIndex() { result = index }
-}
-
-class IndirectionPosition extends Position, TIndirectionPosition {
-  int argumentIndex;
-  int indirectionIndex;
-
-  IndirectionPosition() { this = TIndirectionPosition(argumentIndex, indirectionIndex) }
-
-  override string toString() {
-    if argumentIndex = -1
-    then if indirectionIndex > 0 then result = "this indirection" else result = "this"
-    else
-      if indirectionIndex > 0
-      then result = argumentIndex.toString() + " indirection"
-      else result = argumentIndex.toString()
-  }
-
-  int getArgumentIndex() { result = argumentIndex }
-
-  int getIndirectionIndex() { result = indirectionIndex }
-}
-
-newtype TPosition =
-  TDirectPosition(int index) { exists(any(CallInstruction c).getArgument(index)) } or
-  TIndirectionPosition(int argumentIndex, int indirectionIndex) {
-    hasOperandAndIndex(_, any(CallInstruction call).getArgumentOperand(argumentIndex),
-      indirectionIndex)
-  }
-
-private newtype TReturnKind =
-  TNormalReturnKind(int index) {
-    exists(IndirectReturnNode return |
-      return.getAddressOperand() = any(ReturnValueInstruction r).getReturnAddressOperand() and
-      index = return.getIndirectionIndex() - 1 // We subtract one because the return loads the value.
-    )
-  } or
-  TIndirectReturnKind(int argumentIndex, int indirectionIndex) {
-    exists(IndirectReturnNode return, ReturnIndirectionInstruction returnInd |
-      returnInd.hasIndex(argumentIndex) and
-      return.getAddressOperand() = returnInd.getSourceAddressOperand() and
-      indirectionIndex = return.getIndirectionIndex()
-    )
-  }
-
-/**
- * A return kind. A return kind describes how a value can be returned
- * from a callable. For C++, this is simply a function return.
- */
-class ReturnKind extends TReturnKind {
-  /** Gets a textual representation of this return kind. */
-  abstract string toString();
-}
-
-private class NormalReturnKind extends ReturnKind, TNormalReturnKind {
-  int index;
-
-  NormalReturnKind() { this = TNormalReturnKind(index) }
-
-  override string toString() { result = "indirect return" }
-}
-
-private class IndirectReturnKind extends ReturnKind, TIndirectReturnKind {
-  int argumentIndex;
-  int indirectionIndex;
-
-  IndirectReturnKind() { this = TIndirectReturnKind(argumentIndex, indirectionIndex) }
-
-  override string toString() { result = "indirect outparam[" + argumentIndex.toString() + "]" }
-}
-
-/** A data flow node that occurs as the result of a `ReturnStmt`. */
-class ReturnNode extends Node instanceof IndirectReturnNode {
-  /** Gets the kind of this returned value. */
-  abstract ReturnKind getKind();
-}
-
-/**
- * This predicate represents an annoying hack that we have to do. We use the
- * `ReturnIndirectionInstruction` to determine which variables need flow back
- * out of a function. However, the IR will unconditionally create those for a
- * variable passed to a function even though the variable was never updated by
- * the function. And if a function has too many `ReturnNode`s the dataflow
- * library lowers its precision for that function by disabling field flow.
- *
- * So we those eliminate `ReturnNode`s that would have otherwise been created
- * by this unconditional `ReturnIndirectionInstruction` by requiring that there
- * must exist an SSA definition of the IR variable in the function.
- */
-private predicate hasNonInitializeParameterDef(IRVariable v) {
-  exists(Ssa::Def def |
-    not def.getDefiningInstruction() instanceof InitializeParameterInstruction and
-    v = def.getSourceVariable().getBaseVariable().(Ssa::BaseIRVariable).getIRVariable()
-  )
-}
-
-class ReturnIndirectionNode extends IndirectReturnNode, ReturnNode {
-  override ReturnKind getKind() {
-    exists(int argumentIndex, ReturnIndirectionInstruction returnInd |
-      returnInd.hasIndex(argumentIndex) and
-      this.getAddressOperand() = returnInd.getSourceAddressOperand() and
-      result = TIndirectReturnKind(argumentIndex, this.getIndirectionIndex()) and
-      hasNonInitializeParameterDef(returnInd.getIRVariable())
-    )
-    or
-    this.getAddressOperand() = any(ReturnValueInstruction r).getReturnAddressOperand() and
-    result = TNormalReturnKind(this.getIndirectionIndex() - 1)
-  }
-}
-
-private Operand fullyConvertedCallStep(Operand op) {
-  not exists(getANonConversionUse(op)) and
-  exists(Instruction instr |
-    conversionFlow(op, instr, _) and
-    result = getAUse(instr)
-  )
-}
-
-/**
- * Gets the instruction that uses this operand, if the instruction is not
- * ignored for dataflow purposes.
- */
-private Instruction getUse(Operand op) {
-  result = op.getUse() and
-  not Ssa::ignoreOperand(op)
-}
-
-/** Gets a use of the instruction `instr` that is not ignored for dataflow purposes. */
-Operand getAUse(Instruction instr) {
-  result = instr.getAUse() and
-  not Ssa::ignoreOperand(result)
-}
-
-/**
- * Gets a use of `operand` that is:
- * - not ignored for dataflow purposes, and
- * - not a conversion-like instruction.
- */
-private Instruction getANonConversionUse(Operand operand) {
-  result = getUse(operand) and
-  not conversionFlow(_, result, _)
-}
-
-/**
- * Gets the operand that represents the first use of the value of `call` following
- * a sequence of conversion-like instructions.
- */
-predicate operandForfullyConvertedCall(Operand operand, CallInstruction call) {
-  exists(getANonConversionUse(operand)) and
-  (
-    operand = getAUse(call)
-    or
-    operand = fullyConvertedCallStep*(getAUse(call))
-  )
-}
-
-/**
- * Gets the instruction that represents the first use of the value of `call` following
- * a sequence of conversion-like instructions.
- *
- * This predicate only holds if there is no suitable operand (i.e., no operand of a non-
- * conversion instruction) to use to represent the value of `call` after conversions.
- */
-predicate instructionForfullyConvertedCall(Instruction instr, CallInstruction call) {
-  not operandForfullyConvertedCall(_, call) and
-  (
-    // If there is no use of the call then we pick the call instruction
-    not exists(getAUse(call)) and
-    instr = call
-    or
-    // Otherwise, flow to the first non-conversion use.
-    exists(Operand operand | operand = fullyConvertedCallStep*(getAUse(call)) |
-      instr = getANonConversionUse(operand)
-    )
-  )
-}
-
-/** Holds if `node` represents the output node for `call`. */
-private predicate simpleOutNode(Node node, CallInstruction call) {
-  operandForfullyConvertedCall(node.asOperand(), call)
-  or
-  instructionForfullyConvertedCall(node.asInstruction(), call)
-}
-
-/** A data flow node that represents the output of a call. */
-class OutNode extends Node {
-  OutNode() {
-    // Return values not hidden behind indirections
-    simpleOutNode(this, _)
-    or
-    // Return values hidden behind indirections
-    this instanceof IndirectReturnOutNode
-    or
-    // Modified arguments hidden behind indirections
-    this instanceof IndirectArgumentOutNode
-  }
-
-  /** Gets the underlying call. */
-  abstract DataFlowCall getCall();
-
-  abstract ReturnKind getReturnKind();
-}
-
-private class DirectCallOutNode extends OutNode {
-  CallInstruction call;
-
-  DirectCallOutNode() { simpleOutNode(this, call) }
-
-  override DataFlowCall getCall() { result = call }
-
-  override ReturnKind getReturnKind() { result = TNormalReturnKind(0) }
-}
-
-private class IndirectCallOutNode extends OutNode, IndirectReturnOutNode {
-  override DataFlowCall getCall() { result = this.getCallInstruction() }
-
-  override ReturnKind getReturnKind() { result = TNormalReturnKind(this.getIndirectionIndex()) }
-}
-
-private class SideEffectOutNode extends OutNode, IndirectArgumentOutNode {
-  override DataFlowCall getCall() { result = this.getCallInstruction() }
-
-  override ReturnKind getReturnKind() {
-    result = TIndirectReturnKind(this.getArgumentIndex(), this.getIndirectionIndex())
-  }
-}
-
-/**
- * Gets a node that can read the value returned from `call` with return kind
- * `kind`.
- */
-OutNode getAnOutNode(DataFlowCall call, ReturnKind kind) {
-  result.getCall() = call and
-  result.getReturnKind() = kind
-}
-
-/**
- * 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(Cpp::GlobalOrNamespaceVariable v |
-    v =
-      n1.asInstruction()
-          .(StoreInstruction)
-          .getResultAddress()
-          .(VariableAddressInstruction)
-          .getAstVariable() and
-    v = n2.asVariable()
-    or
-    v =
-      n2.asInstruction()
-          .(LoadInstruction)
-          .getSourceAddress()
-          .(VariableAddressInstruction)
-          .getAstVariable() and
-    v = n1.asVariable()
-  )
-}
-
-/**
- * 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, Content c, PostFieldUpdateNode node2) {
-  exists(int indirectionIndex1, int numberOfLoads, StoreInstruction store |
-    nodeHasInstruction(node1, store, pragma[only_bind_into](indirectionIndex1)) and
-    node2.getIndirectionIndex() = 1 and
-    numberOfLoadsFromOperand(node2.getFieldAddress(), store.getDestinationAddressOperand(),
-      numberOfLoads)
-  |
-    exists(FieldContent fc | fc = c |
-      fc.getField() = node2.getUpdatedField() and
-      fc.getIndirectionIndex() = 1 + indirectionIndex1 + numberOfLoads
-    )
-    or
-    exists(UnionContent uc | uc = c |
-      uc.getAField() = node2.getUpdatedField() and
-      uc.getIndirectionIndex() = 1 + indirectionIndex1 + numberOfLoads
-    )
-  )
-}
-
-/**
- * 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) {
-  exists(LoadInstruction load | load.getSourceAddressOperand() = operandFrom |
-    operandTo = operandFrom and ind = 0
-    or
-    numberOfLoadsFromOperand(load.getAUse(), operandTo, ind - 1)
-  )
-  or
-  exists(Operand op, Instruction instr |
-    instr = op.getDef() and
-    conversionFlow(operandFrom, instr, _) and
-    numberOfLoadsFromOperand(op, operandTo, ind)
-  )
-}
-
-/**
- * Holds if `operandFrom` flows to `operandTo` using a sequence of conversion-like
- * operations and exactly `n` `LoadInstruction` operations.
- */
-private predicate numberOfLoadsFromOperand(Operand operandFrom, Operand operandTo, int n) {
-  numberOfLoadsFromOperandRec(operandFrom, operandTo, n)
-  or
-  not any(LoadInstruction load).getSourceAddressOperand() = operandFrom and
-  not conversionFlow(operandFrom, _, _) and
-  operandFrom = operandTo and
-  n = 0
-}
-
-// Needed to join on both an operand and an index at the same time.
-pragma[noinline]
-predicate nodeHasOperand(Node node, Operand operand, int indirectionIndex) {
-  node.asOperand() = operand and indirectionIndex = 0
-  or
-  hasOperandAndIndex(node, operand, indirectionIndex)
-}
-
-// Needed to join on both an instruction and an index at the same time.
-pragma[noinline]
-predicate nodeHasInstruction(Node node, Instruction instr, int indirectionIndex) {
-  node.asInstruction() = instr and indirectionIndex = 0
-  or
-  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, Content c, Node node2) {
-  exists(FieldAddress fa1, Operand operand, int numberOfLoads, int indirectionIndex2 |
-    nodeHasOperand(node2, operand, indirectionIndex2) and
-    nodeHasOperand(node1, fa1.getObjectAddressOperand(), _) and
-    numberOfLoadsFromOperand(fa1, operand, numberOfLoads)
-  |
-    exists(FieldContent fc | fc = c |
-      fc.getField() = fa1.getField() and
-      fc.getIndirectionIndex() = indirectionIndex2 + numberOfLoads
-    )
-    or
-    exists(UnionContent uc | uc = c |
-      uc.getAField() = fa1.getField() and
-      uc.getIndirectionIndex() = indirectionIndex2 + numberOfLoads
-    )
-  )
-}
-
-/**
- * Holds if values stored inside content `c` are cleared at node `n`.
- */
-predicate clearsContent(Node n, Content c) {
-  none() // stub implementation
-}
-
-/**
- * Holds if the value that is being tracked is expected to be stored inside content `c`
- * at node `n`.
- */
-predicate expectsContent(Node n, ContentSet c) { none() }
-
-/** Gets the type of `n` used for type pruning. */
-DataFlowType getNodeType(Node n) {
-  suppressUnusedNode(n) and
-  result instanceof VoidType // stub implementation
-}
-
-/** Gets a string representation of a type returned by `getNodeType`. */
-string ppReprType(DataFlowType t) { none() } // stub implementation
-
-/**
- * Holds if `t1` and `t2` are compatible, that is, whether data can flow from
- * a node of type `t1` to a node of type `t2`.
- */
-pragma[inline]
-predicate compatibleTypes(DataFlowType t1, DataFlowType t2) {
-  any() // stub implementation
-}
-
-private predicate suppressUnusedNode(Node n) { any() }
-
-//////////////////////////////////////////////////////////////////////////////
-// Java QL library compatibility wrappers
-//////////////////////////////////////////////////////////////////////////////
-/** A node that performs a type cast. */
-class CastNode extends Node {
-  CastNode() { none() } // stub implementation
-}
-
-/**
- * A function that may contain code or a variable that may contain itself. When
- * flow crosses from one _enclosing callable_ to another, the interprocedural
- * data-flow library discards call contexts and inserts a node in the big-step
- * relation used for human-readable path explanations.
- */
-class DataFlowCallable = Cpp::Declaration;
-
-class DataFlowExpr = Expr;
-
-class DataFlowType = Type;
-
-/** A function call relevant for data flow. */
-class DataFlowCall extends CallInstruction {
-  Function getEnclosingCallable() { result = this.getEnclosingFunction() }
-}
-
-predicate isUnreachableInCall(Node n, DataFlowCall call) { none() } // stub implementation
-
-int accessPathLimit() { result = 5 }
-
-/**
- * Holds if access paths with `c` at their head always should be tracked at high
- * precision. This disables adaptive access path precision for such access paths.
- */
-predicate forceHighPrecision(Content c) { none() }
-
-/** The unit type. */
-private newtype TUnit = TMkUnit()
-
-/** The trivial type with a single element. */
-class Unit extends TUnit {
-  /** Gets a textual representation of this element. */
-  string toString() { result = "unit" }
-}
-
-/** Holds if `n` should be hidden from path explanations. */
-predicate nodeIsHidden(Node n) { n instanceof OperandNode and not n instanceof ArgumentNode }
-
-class LambdaCallKind = Unit;
-
-/** Holds if `creation` is an expression that creates a lambda of kind `kind` for `c`. */
-predicate lambdaCreation(Node creation, LambdaCallKind kind, DataFlowCallable c) { none() }
-
-/** Holds if `call` is a lambda call of kind `kind` where `receiver` is the lambda expression. */
-predicate lambdaCall(DataFlowCall call, LambdaCallKind kind, Node receiver) { none() }
-
-/** Extra data-flow steps needed for lambda flow analysis. */
-predicate additionalLambdaFlowStep(Node nodeFrom, Node nodeTo, boolean preservesValue) { none() }
-
-/**
- * Holds if flow is allowed to pass from parameter `p` and back to itself as a
- * side-effect, resulting in a summary from `p` to itself.
- *
- * One example would be to allow flow like `p.foo = p.bar;`, which is disallowed
- * by default as a heuristic.
- */
-predicate allowParameterReturnInSelf(ParameterNode p) { none() }
-
-/** An approximated `Content`. */
-class ContentApprox = Unit;
-
-/** Gets an approximated value for content `c`. */
-pragma[inline]
-ContentApprox getContentApprox(Content c) { any() }
-
-private class MyConsistencyConfiguration extends Consistency::ConsistencyConfiguration {
-  override predicate argHasPostUpdateExclude(ArgumentNode n) {
-    // The rules for whether an IR argument gets a post-update node are too
-    // complex to model here.
-    any()
-  }
-}

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

@@ -1,93 +0,0 @@
-/**
- * Provides predicates for mapping the `FunctionInput` and `FunctionOutput`
- * classes used in function models to the corresponding instructions.
- */
-
-private import semmle.code.cpp.ir.IR
-private import experimental.semmle.code.cpp.ir.dataflow.DataFlow
-private import experimental.semmle.code.cpp.ir.dataflow.internal.DataFlowUtil
-private import SsaInternals as Ssa
-
-/**
- * Gets the instruction that goes into `input` for `call`.
- */
-DataFlow::Node callInput(CallInstruction call, FunctionInput input) {
-  // An argument or qualifier
-  exists(int index |
-    result.asOperand() = call.getArgumentOperand(index) and
-    input.isParameterOrQualifierAddress(index)
-  )
-  or
-  // A value pointed to by an argument or qualifier
-  exists(int index, int indirectionIndex |
-    hasOperandAndIndex(result, call.getArgumentOperand(index), indirectionIndex) and
-    input.isParameterDerefOrQualifierObject(index, indirectionIndex)
-  )
-  or
-  exists(int ind |
-    result = getIndirectReturnOutNode(call, ind) and
-    input.isReturnValueDeref(ind)
-  )
-}
-
-/**
- * Gets the instruction that holds the `output` for `call`.
- */
-Node callOutput(CallInstruction call, FunctionOutput output) {
-  // The return value
-  result.asInstruction() = call and
-  output.isReturnValue()
-  or
-  // The side effect of a call on the value pointed to by an argument or qualifier
-  exists(int index, int indirectionIndex |
-    result.(IndirectArgumentOutNode).getArgumentIndex() = index and
-    result.(IndirectArgumentOutNode).getIndirectionIndex() = indirectionIndex and
-    result.(IndirectArgumentOutNode).getCallInstruction() = call and
-    output.isParameterDerefOrQualifierObject(index, indirectionIndex)
-  )
-  or
-  exists(int ind |
-    result = getIndirectReturnOutNode(call, ind) and
-    output.isReturnValueDeref(ind)
-  )
-}
-
-DataFlow::Node callInput(CallInstruction call, FunctionInput input, int d) {
-  exists(DataFlow::Node n | n = callInput(call, input) and d > 0 |
-    // An argument or qualifier
-    hasOperandAndIndex(result, n.asOperand(), d)
-    or
-    exists(Operand operand, int indirectionIndex |
-      // A value pointed to by an argument or qualifier
-      hasOperandAndIndex(n, operand, indirectionIndex) and
-      hasOperandAndIndex(result, operand, indirectionIndex + d)
-    )
-  )
-}
-
-private IndirectReturnOutNode getIndirectReturnOutNode(CallInstruction call, int d) {
-  result.getCallInstruction() = call and
-  result.getIndirectionIndex() = d
-}
-
-/**
- * Gets the instruction that holds the `output` for `call`.
- */
-bindingset[d]
-Node callOutput(CallInstruction call, FunctionOutput output, int d) {
-  exists(DataFlow::Node n | n = callOutput(call, output) and d > 0 |
-    // The return value
-    result = getIndirectReturnOutNode(n.asInstruction(), d)
-    or
-    // If there isn't an indirect out node for the call with indirection `d` then
-    // we conflate this with the underlying `CallInstruction`.
-    not exists(getIndirectReturnOutNode(call, d)) and
-    n.asInstruction() = result.asInstruction()
-    or
-    // The side effect of a call on the value pointed to by an argument or qualifier
-    exists(Operand operand, int indirectionIndex |
-      Ssa::outNodeHasAddressAndIndex(n, operand, indirectionIndex) and
-      Ssa::outNodeHasAddressAndIndex(result, operand, indirectionIndex + d)
-    )
-  )
-}

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

@@ -1,136 +0,0 @@
-private import cpp
-// The `ValueNumbering` library has to be imported right after `cpp` to ensure
-// that the cached IR gets the same checksum here as it does in queries that use
-// `ValueNumbering` without `DataFlow`.
-private import semmle.code.cpp.ir.ValueNumbering
-private import semmle.code.cpp.ir.IR
-private import semmle.code.cpp.ir.dataflow.DataFlow
-private import semmle.code.cpp.ir.dataflow.internal.DataFlowUtil
-private import PrintIRUtilities
-
-/**
- * Gets the local dataflow from other nodes in the same function to this node.
- */
-private string getFromFlow(DataFlow::Node useNode, int order1, int order2) {
-  exists(DataFlow::Node defNode, string prefix |
-    (
-      simpleLocalFlowStep(defNode, useNode) and prefix = ""
-      or
-      any(DataFlow::Configuration cfg).isAdditionalFlowStep(defNode, useNode) and
-      defNode.getEnclosingCallable() = useNode.getEnclosingCallable() and
-      prefix = "+"
-    ) and
-    if defNode.asInstruction() = useNode.asOperand().getAnyDef()
-    then
-      // Shorthand for flow from the def of this operand.
-      result = prefix + "def" and
-      order1 = -1 and
-      order2 = 0
-    else
-      if defNode.asOperand().getUse() = useNode.asInstruction()
-      then
-        // Shorthand for flow from an operand of this instruction
-        result = prefix + defNode.asOperand().getDumpId() and
-        order1 = -1 and
-        order2 = defNode.asOperand().getDumpSortOrder()
-      else result = prefix + nodeId(defNode, order1, order2)
-  )
-}
-
-/**
- * Gets the local dataflow from this node to other nodes in the same function.
- */
-private string getToFlow(DataFlow::Node defNode, int order1, int order2) {
-  exists(DataFlow::Node useNode, string prefix |
-    (
-      simpleLocalFlowStep(defNode, useNode) and prefix = ""
-      or
-      any(DataFlow::Configuration cfg).isAdditionalFlowStep(defNode, useNode) and
-      defNode.getEnclosingCallable() = useNode.getEnclosingCallable() and
-      prefix = "+"
-    ) and
-    if useNode.asInstruction() = defNode.asOperand().getUse()
-    then
-      // Shorthand for flow to this operand's instruction.
-      result = prefix + "result" and
-      order1 = -1 and
-      order2 = 0
-    else result = prefix + nodeId(useNode, order1, order2)
-  )
-}
-
-/**
- * Gets the properties of the dataflow node `node`.
- */
-private string getNodeProperty(DataFlow::Node node, string key) {
-  // List dataflow into and out of this node. Flow into this node is printed as `src->@`, and flow
-  // out of this node is printed as `@->dest`.
-  key = "flow" and
-  result =
-    strictconcat(string flow, boolean to, int order1, int order2 |
-      flow = getFromFlow(node, order1, order2) + "->@" and to = false
-      or
-      flow = "@->" + getToFlow(node, order1, order2) and to = true
-    |
-      flow, ", " order by to, order1, order2, flow
-    )
-  or
-  // Is this node a dataflow sink?
-  key = "sink" and
-  any(DataFlow::Configuration cfg).isSink(node) and
-  result = "true"
-  or
-  // Is this node a dataflow source?
-  key = "source" and
-  any(DataFlow::Configuration cfg).isSource(node) and
-  result = "true"
-  or
-  // Is this node a dataflow barrier, and if so, what kind?
-  key = "barrier" and
-  result =
-    strictconcat(string kind |
-      any(DataFlow::Configuration cfg).isBarrier(node) and kind = "full"
-      or
-      any(DataFlow::Configuration cfg).isBarrierIn(node) and kind = "in"
-      or
-      any(DataFlow::Configuration cfg).isBarrierOut(node) and kind = "out"
-    |
-      kind, ", "
-    )
-  or
-  // Is there partial flow from a source to this node?
-  // This property will only be emitted if partial flow is enabled by overriding
-  // `DataFlow::Configuration::explorationLimit()`.
-  key = "pflow" and
-  result =
-    strictconcat(DataFlow::PartialPathNode sourceNode, DataFlow::PartialPathNode destNode, int dist,
-      int order1, int order2 |
-      any(DataFlow::Configuration cfg).hasPartialFlow(sourceNode, destNode, dist) and
-      destNode.getNode() = node and
-      // Only print flow from a source in the same function.
-      sourceNode.getNode().getEnclosingCallable() = node.getEnclosingCallable()
-    |
-      nodeId(sourceNode.getNode(), order1, order2) + "+" + dist.toString(), ", "
-      order by
-        order1, order2, dist desc
-    )
-}
-
-/**
- * Property provider for local IR dataflow.
- */
-class LocalFlowPropertyProvider extends IRPropertyProvider {
-  override string getOperandProperty(Operand operand, string key) {
-    exists(DataFlow::Node node |
-      operand = node.asOperand() and
-      result = getNodeProperty(node, key)
-    )
-  }
-
-  override string getInstructionProperty(Instruction instruction, string key) {
-    exists(DataFlow::Node node |
-      instruction = node.asInstruction() and
-      result = getNodeProperty(node, key)
-    )
-  }
-}

cpp/ql/lib/experimental/semmle/code/cpp/ir/dataflow/internal/PrintIRStoreSteps.qll

@@ -1,33 +0,0 @@
-/**
- * Print the dataflow local store steps in IR dumps.
- */
-
-private import cpp
-// The `ValueNumbering` library has to be imported right after `cpp` to ensure
-// that the cached IR gets the same checksum here as it does in queries that use
-// `ValueNumbering` without `DataFlow`.
-private import semmle.code.cpp.ir.ValueNumbering
-private import semmle.code.cpp.ir.IR
-private import semmle.code.cpp.ir.dataflow.DataFlow
-private import semmle.code.cpp.ir.dataflow.internal.DataFlowUtil
-private import semmle.code.cpp.ir.dataflow.internal.DataFlowPrivate
-private import PrintIRUtilities
-
-/**
- * Property provider for local IR dataflow store steps.
- */
-class LocalFlowPropertyProvider extends IRPropertyProvider {
-  override string getInstructionProperty(Instruction instruction, string key) {
-    exists(DataFlow::Node objectNode, Content content |
-      key = "content[" + content.toString() + "]" and
-      instruction = objectNode.asInstruction() and
-      result =
-        strictconcat(string element, DataFlow::Node fieldNode |
-          storeStep(fieldNode, content, objectNode) and
-          element = nodeId(fieldNode, _, _)
-        |
-          element, ", "
-        )
-    )
-  }
-}

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

@@ -1,39 +0,0 @@
-/**
- * Shared utilities used when printing dataflow annotations in IR dumps.
- */
-
-private import cpp
-// The `ValueNumbering` library has to be imported right after `cpp` to ensure
-// that the cached IR gets the same checksum here as it does in queries that use
-// `ValueNumbering` without `DataFlow`.
-private import semmle.code.cpp.ir.ValueNumbering
-private import semmle.code.cpp.ir.IR
-private import semmle.code.cpp.ir.dataflow.DataFlow
-
-/**
- * Gets a short ID for an IR dataflow node.
- * - For `Instruction`s, this is just the result ID of the instruction (e.g. `m128`).
- * - For `Operand`s, this is the label of the operand, prefixed with the result ID of the
- *   instruction and a dot (e.g. `m128.left`).
- * - For `Variable`s, this is the qualified name of the variable.
- */
-string nodeId(DataFlow::Node node, int order1, int order2) {
-  exists(Instruction instruction | instruction = node.asInstruction() |
-    result = instruction.getResultId() and
-    order1 = instruction.getBlock().getDisplayIndex() and
-    order2 = instruction.getDisplayIndexInBlock()
-  )
-  or
-  exists(Operand operand, Instruction instruction |
-    operand = node.asOperand() and
-    instruction = operand.getUse()
-  |
-    result = instruction.getResultId() + "." + operand.getDumpId() and
-    order1 = instruction.getBlock().getDisplayIndex() and
-    order2 = instruction.getDisplayIndexInBlock()
-  )
-  or
-  result = "var(" + node.asVariable().getQualifiedName() + ")" and
-  order1 = 1000000 and
-  order2 = 0
-}

cpp/ql/lib/experimental/semmle/code/cpp/ir/dataflow/internal/SsaInternals.qll

@@ -1,552 +0,0 @@
-private import codeql.ssa.Ssa as SsaImplCommon
-private import semmle.code.cpp.ir.IR
-private import DataFlowUtil
-private import DataFlowImplCommon as DataFlowImplCommon
-private import semmle.code.cpp.models.interfaces.Allocation as Alloc
-private import semmle.code.cpp.models.interfaces.DataFlow as DataFlow
-private import semmle.code.cpp.ir.internal.IRCppLanguage
-private import DataFlowPrivate
-private import ssa0.SsaInternals as SsaInternals0
-import SsaInternalsCommon
-
-private module SourceVariables {
-  int getMaxIndirectionForIRVariable(IRVariable var) {
-    exists(Type type, boolean isGLValue |
-      var.getLanguageType().hasType(type, isGLValue) and
-      if isGLValue = true
-      then result = 1 + getMaxIndirectionsForType(type)
-      else result = getMaxIndirectionsForType(type)
-    )
-  }
-
-  class BaseSourceVariable = SsaInternals0::BaseSourceVariable;
-
-  class BaseIRVariable = SsaInternals0::BaseIRVariable;
-
-  class BaseCallVariable = SsaInternals0::BaseCallVariable;
-
-  cached
-  private newtype TSourceVariable =
-    TSourceIRVariable(BaseIRVariable baseVar, int ind) {
-      ind = [0 .. getMaxIndirectionForIRVariable(baseVar.getIRVariable())]
-    } or
-    TCallVariable(AllocationInstruction call, int ind) {
-      ind = [0 .. countIndirectionsForCppType(getResultLanguageType(call))]
-    }
-
-  abstract class SourceVariable extends TSourceVariable {
-    int ind;
-
-    bindingset[ind]
-    SourceVariable() { any() }
-
-    abstract string toString();
-
-    int getIndirection() { result = ind }
-
-    abstract BaseSourceVariable getBaseVariable();
-  }
-
-  class SourceIRVariable extends SourceVariable, TSourceIRVariable {
-    BaseIRVariable var;
-
-    SourceIRVariable() { this = TSourceIRVariable(var, ind) }
-
-    IRVariable getIRVariable() { result = var.getIRVariable() }
-
-    override BaseIRVariable getBaseVariable() { result.getIRVariable() = this.getIRVariable() }
-
-    override string toString() {
-      ind = 0 and
-      result = this.getIRVariable().toString()
-      or
-      ind > 0 and
-      result = this.getIRVariable().toString() + " indirection"
-    }
-  }
-
-  class CallVariable extends SourceVariable, TCallVariable {
-    AllocationInstruction call;
-
-    CallVariable() { this = TCallVariable(call, ind) }
-
-    AllocationInstruction getCall() { result = call }
-
-    override BaseCallVariable getBaseVariable() { result.getCallInstruction() = call }
-
-    override string toString() {
-      ind = 0 and
-      result = "Call"
-      or
-      ind > 0 and
-      result = "Call indirection"
-    }
-  }
-}
-
-import SourceVariables
-
-predicate hasIndirectOperand(Operand op, int indirectionIndex) {
-  exists(CppType type, int m |
-    not ignoreOperand(op) and
-    type = getLanguageType(op) and
-    m = countIndirectionsForCppType(type) and
-    indirectionIndex = [1 .. m]
-  )
-}
-
-predicate hasIndirectInstruction(Instruction instr, int indirectionIndex) {
-  exists(CppType type, int m |
-    not ignoreInstruction(instr) and
-    type = getResultLanguageType(instr) and
-    m = countIndirectionsForCppType(type) and
-    indirectionIndex = [1 .. m]
-  )
-}
-
-cached
-private newtype TDefOrUseImpl =
-  TDefImpl(Operand address, int indirectionIndex) {
-    isDef(_, _, address, _, _, indirectionIndex) and
-    // We only include the definition if the SSA pruning stage
-    // concluded that the definition is live after the write.
-    any(SsaInternals0::Def def).getAddressOperand() = address
-  } or
-  TUseImpl(Operand operand, int indirectionIndex) {
-    isUse(_, operand, _, _, indirectionIndex) and
-    not isDef(_, _, operand, _, _, _)
-  }
-
-abstract private class DefOrUseImpl extends TDefOrUseImpl {
-  /** Gets a textual representation of this element. */
-  abstract string toString();
-
-  /** Gets the block of this definition or use. */
-  abstract IRBlock getBlock();
-
-  /** Holds if this definition or use has index `index` in block `block`. */
-  abstract predicate hasIndexInBlock(IRBlock block, int index);
-
-  final predicate hasIndexInBlock(IRBlock block, int index, SourceVariable sv) {
-    this.hasIndexInBlock(block, index) and
-    sv = this.getSourceVariable()
-  }
-
-  /** Gets the location of this element. */
-  abstract Cpp::Location getLocation();
-
-  /**
-   * Gets the index (i.e., the number of loads required) of this
-   * definition or use.
-   *
-   * Note that this is _not_ the definition's (or use's) index in
-   * the enclosing basic block. To obtain this index, use
-   * `DefOrUseImpl::hasIndexInBlock/2` or `DefOrUseImpl::hasIndexInBlock/3`.
-   */
-  abstract int getIndirectionIndex();
-
-  /**
-   * Gets the instruction that computes the base of this definition or use.
-   * This is always a `VariableAddressInstruction` or an `AllocationInstruction`.
-   */
-  abstract Instruction getBase();
-
-  final BaseSourceVariable getBaseSourceVariable() {
-    exists(IRVariable var |
-      result.(BaseIRVariable).getIRVariable() = var and
-      instructionHasIRVariable(this.getBase(), var)
-    )
-    or
-    result.(BaseCallVariable).getCallInstruction() = this.getBase()
-  }
-
-  /** Gets the variable that is defined or used. */
-  final SourceVariable getSourceVariable() {
-    exists(BaseSourceVariable v, int ind |
-      sourceVariableHasBaseAndIndex(result, v, ind) and
-      defOrUseHasSourceVariable(this, v, ind)
-    )
-  }
-}
-
-pragma[noinline]
-private predicate instructionHasIRVariable(VariableAddressInstruction vai, IRVariable var) {
-  vai.getIRVariable() = var
-}
-
-private predicate defOrUseHasSourceVariable(DefOrUseImpl defOrUse, BaseSourceVariable bv, int ind) {
-  defHasSourceVariable(defOrUse, bv, ind)
-  or
-  useHasSourceVariable(defOrUse, bv, ind)
-}
-
-pragma[noinline]
-private predicate defHasSourceVariable(DefImpl def, BaseSourceVariable bv, int ind) {
-  bv = def.getBaseSourceVariable() and
-  ind = def.getIndirection()
-}
-
-pragma[noinline]
-private predicate useHasSourceVariable(UseImpl use, BaseSourceVariable bv, int ind) {
-  bv = use.getBaseSourceVariable() and
-  ind = use.getIndirection()
-}
-
-pragma[noinline]
-private predicate sourceVariableHasBaseAndIndex(SourceVariable v, BaseSourceVariable bv, int ind) {
-  v.getBaseVariable() = bv and
-  v.getIndirection() = ind
-}
-
-class DefImpl extends DefOrUseImpl, TDefImpl {
-  Operand address;
-  int ind;
-
-  DefImpl() { this = TDefImpl(address, ind) }
-
-  override Instruction getBase() { isDef(_, _, address, result, _, _) }
-
-  Operand getAddressOperand() { result = address }
-
-  int getIndirection() { isDef(_, _, address, _, result, ind) }
-
-  override int getIndirectionIndex() { result = ind }
-
-  Instruction getDefiningInstruction() { isDef(_, result, address, _, _, _) }
-
-  override string toString() { result = "DefImpl" }
-
-  override IRBlock getBlock() { result = this.getDefiningInstruction().getBlock() }
-
-  override Cpp::Location getLocation() { result = this.getDefiningInstruction().getLocation() }
-
-  final override predicate hasIndexInBlock(IRBlock block, int index) {
-    this.getDefiningInstruction() = block.getInstruction(index)
-  }
-
-  predicate isCertain() { isDef(true, _, address, _, _, ind) }
-}
-
-class UseImpl extends DefOrUseImpl, TUseImpl {
-  Operand operand;
-  int ind;
-
-  UseImpl() { this = TUseImpl(operand, ind) }
-
-  Operand getOperand() { result = operand }
-
-  override string toString() { result = "UseImpl" }
-
-  final override predicate hasIndexInBlock(IRBlock block, int index) {
-    operand.getUse() = block.getInstruction(index)
-  }
-
-  final override IRBlock getBlock() { result = operand.getUse().getBlock() }
-
-  final override Cpp::Location getLocation() { result = operand.getLocation() }
-
-  final int getIndirection() { isUse(_, operand, _, result, ind) }
-
-  override int getIndirectionIndex() { result = ind }
-
-  override Instruction getBase() { isUse(_, operand, result, _, ind) }
-
-  predicate isCertain() { isUse(true, operand, _, _, ind) }
-}
-
-/**
- * Holds if `defOrUse1` is a definition which is first read by `use`,
- * or if `defOrUse1` is a use and `use` is a next subsequent use.
- *
- * In both cases, `use` can either be an explicit use written in the
- * source file, or it can be a phi node as computed by the SSA library.
- */
-predicate adjacentDefRead(DefOrUse defOrUse1, UseOrPhi use) {
-  exists(IRBlock bb1, int i1, SourceVariable v |
-    defOrUse1.asDefOrUse().hasIndexInBlock(bb1, i1, v)
-  |
-    exists(IRBlock bb2, int i2 |
-      adjacentDefRead(_, pragma[only_bind_into](bb1), pragma[only_bind_into](i1),
-        pragma[only_bind_into](bb2), pragma[only_bind_into](i2))
-    |
-      use.asDefOrUse().(UseImpl).hasIndexInBlock(bb2, i2, v)
-    )
-    or
-    exists(PhiNode phi |
-      lastRefRedef(_, bb1, i1, phi) and
-      use.asPhi() = phi and
-      phi.getSourceVariable() = pragma[only_bind_into](v)
-    )
-  )
-}
-
-private predicate useToNode(UseOrPhi use, Node nodeTo) {
-  exists(UseImpl useImpl |
-    useImpl = use.asDefOrUse() and
-    nodeHasOperand(nodeTo, useImpl.getOperand(), useImpl.getIndirectionIndex())
-  )
-  or
-  nodeTo.(SsaPhiNode).getPhiNode() = use.asPhi()
-}
-
-pragma[noinline]
-predicate outNodeHasAddressAndIndex(
-  IndirectArgumentOutNode out, Operand address, int indirectionIndex
-) {
-  out.getAddressOperand() = address and
-  out.getIndirectionIndex() = indirectionIndex
-}
-
-private predicate defToNode(Node nodeFrom, Def def) {
-  nodeHasInstruction(nodeFrom, def.getDefiningInstruction(), def.getIndirectionIndex())
-}
-
-/**
- * INTERNAL: Do not use.
- *
- * Holds if `nodeFrom` is the node that correspond to the definition or use `defOrUse`.
- */
-predicate nodeToDefOrUse(Node nodeFrom, SsaDefOrUse defOrUse) {
-  // Node -> Def
-  defToNode(nodeFrom, defOrUse)
-  or
-  // Node -> Use
-  useToNode(defOrUse, nodeFrom)
-}
-
-/**
- * Perform a single conversion-like step from `nFrom` to `nTo`. This relation
- * only holds when there is no use-use relation out of `nTo`.
- */
-private predicate indirectConversionFlowStep(Node nFrom, Node nTo) {
-  not exists(UseOrPhi defOrUse |
-    nodeToDefOrUse(nTo, defOrUse) and
-    adjacentDefRead(defOrUse, _)
-  ) and
-  exists(Operand op1, Operand op2, int indirectionIndex, Instruction instr |
-    hasOperandAndIndex(nFrom, op1, pragma[only_bind_into](indirectionIndex)) and
-    hasOperandAndIndex(nTo, op2, pragma[only_bind_into](indirectionIndex)) and
-    instr = op2.getDef() and
-    conversionFlow(op1, instr, _)
-  )
-}
-
-/**
- * The reason for this predicate is a bit annoying:
- * We cannot mark a `PointerArithmeticInstruction` that computes an offset based on some SSA
- * variable `x` as a use of `x` since this creates taint-flow in the following example:
- * ```c
- * int x = array[source]
- * sink(*array)
- * ```
- * This is because `source` would flow from the operand of `PointerArithmeticInstruction` to the
- * result of the instruction, and into the `IndirectOperand` that represents the value of `*array`.
- * Then, via use-use flow, flow will arrive at `*array` in `sink(*array)`.
- *
- * So this predicate recurses back along conversions and `PointerArithmeticInstruction`s to find the
- * first use that has provides use-use flow, and uses that target as the target of the `nodeFrom`.
- */
-private predicate adjustForPointerArith(Node nodeFrom, UseOrPhi use) {
-  nodeFrom = any(PostUpdateNode pun).getPreUpdateNode() and
-  exists(DefOrUse defOrUse, Node adjusted |
-    indirectConversionFlowStep*(adjusted, nodeFrom) and
-    nodeToDefOrUse(adjusted, defOrUse) and
-    adjacentDefRead(defOrUse, use)
-  )
-}
-
-/** Holds if there is def-use or use-use flow from `nodeFrom` to `nodeTo`. */
-predicate ssaFlow(Node nodeFrom, Node nodeTo) {
-  // `nodeFrom = any(PostUpdateNode pun).getPreUpdateNode()` is implied by adjustedForPointerArith.
-  exists(UseOrPhi use |
-    adjustForPointerArith(nodeFrom, use) and
-    useToNode(use, nodeTo)
-  )
-  or
-  not nodeFrom = any(PostUpdateNode pun).getPreUpdateNode() and
-  exists(DefOrUse defOrUse1, UseOrPhi use |
-    nodeToDefOrUse(nodeFrom, defOrUse1) and
-    adjacentDefRead(defOrUse1, use) and
-    useToNode(use, nodeTo)
-  )
-}
-
-/** Holds if `nodeTo` receives flow from the phi node `nodeFrom`. */
-predicate fromPhiNode(SsaPhiNode nodeFrom, Node nodeTo) {
-  exists(PhiNode phi, SourceVariable sv, IRBlock bb1, int i1, UseOrPhi use |
-    phi = nodeFrom.getPhiNode() and
-    phi.definesAt(sv, bb1, i1) and
-    useToNode(use, nodeTo)
-  |
-    exists(IRBlock bb2, int i2 |
-      use.asDefOrUse().hasIndexInBlock(bb2, i2, sv) and
-      adjacentDefRead(phi, bb1, i1, bb2, i2)
-    )
-    or
-    exists(PhiNode phiTo |
-      lastRefRedef(phi, _, _, phiTo) and
-      nodeTo.(SsaPhiNode).getPhiNode() = phiTo
-    )
-  )
-}
-
-private SsaInternals0::SourceVariable getOldSourceVariable(SourceVariable v) {
-  v.getBaseVariable().(BaseIRVariable).getIRVariable() =
-    result.getBaseVariable().(SsaInternals0::BaseIRVariable).getIRVariable()
-  or
-  v.getBaseVariable().(BaseCallVariable).getCallInstruction() =
-    result.getBaseVariable().(SsaInternals0::BaseCallVariable).getCallInstruction()
-}
-
-/**
- * Holds if there is a write at index `i` in basic block `bb` to variable `v` that's
- * subsequently read (as determined by the SSA pruning stage).
- */
-private predicate variableWriteCand(IRBlock bb, int i, SourceVariable v) {
-  exists(SsaInternals0::Def def, SsaInternals0::SourceVariable v0 |
-    def.asDefOrUse().hasIndexInBlock(bb, i, v0) and
-    v0 = getOldSourceVariable(v)
-  )
-}
-
-private module SsaInput implements SsaImplCommon::InputSig {
-  import InputSigCommon
-  import SourceVariables
-
-  /**
-   * Holds if the `i`'th write in block `bb` writes to the variable `v`.
-   * `certain` is `true` if the write is guaranteed to overwrite the entire variable.
-   */
-  predicate variableWrite(IRBlock bb, int i, SourceVariable v, boolean certain) {
-    DataFlowImplCommon::forceCachingInSameStage() and
-    variableWriteCand(bb, i, v) and
-    exists(DefImpl def | def.hasIndexInBlock(bb, i, v) |
-      if def.isCertain() then certain = true else certain = false
-    )
-  }
-
-  /**
-   * Holds if the `i`'th read in block `bb` reads to the variable `v`.
-   * `certain` is `true` if the read is guaranteed. For C++, this is always the case.
-   */
-  predicate variableRead(IRBlock bb, int i, SourceVariable v, boolean certain) {
-    exists(UseImpl use | use.hasIndexInBlock(bb, i, v) |
-      if use.isCertain() then certain = true else certain = false
-    )
-  }
-}
-
-/**
- * The final SSA predicates used for dataflow purposes.
- */
-cached
-module SsaCached {
-  /**
-   * Holds if `def` is accessed at index `i1` in basic block `bb1` (either a read
-   * or a write), `def` is read at index `i2` in basic block `bb2`, and there is a
-   * path between them without any read of `def`.
-   */
-  cached
-  predicate adjacentDefRead(Definition def, IRBlock bb1, int i1, IRBlock bb2, int i2) {
-    SsaImpl::adjacentDefRead(def, bb1, i1, bb2, i2)
-  }
-
-  /**
-   * Holds if the node at index `i` in `bb` is a last reference to SSA definition
-   * `def`. The reference is last because it can reach another write `next`,
-   * without passing through another read or write.
-   */
-  cached
-  predicate lastRefRedef(Definition def, IRBlock bb, int i, Definition next) {
-    SsaImpl::lastRefRedef(def, bb, i, next)
-  }
-}
-
-cached
-private newtype TSsaDefOrUse =
-  TDefOrUse(DefOrUseImpl defOrUse) {
-    defOrUse instanceof UseImpl
-    or
-    // Like in the pruning stage, we only include definition that's live after the
-    // write as the final definitions computed by SSA.
-    exists(Definition def, SourceVariable sv, IRBlock bb, int i |
-      def.definesAt(sv, bb, i) and
-      defOrUse.(DefImpl).hasIndexInBlock(bb, i, sv)
-    )
-  } or
-  TPhi(PhiNode phi)
-
-abstract private class SsaDefOrUse extends TSsaDefOrUse {
-  string toString() { none() }
-
-  DefOrUseImpl asDefOrUse() { none() }
-
-  PhiNode asPhi() { none() }
-
-  abstract Location getLocation();
-}
-
-class DefOrUse extends TDefOrUse, SsaDefOrUse {
-  DefOrUseImpl defOrUse;
-
-  DefOrUse() { this = TDefOrUse(defOrUse) }
-
-  final override DefOrUseImpl asDefOrUse() { result = defOrUse }
-
-  final override Location getLocation() { result = defOrUse.getLocation() }
-
-  final SourceVariable getSourceVariable() { result = defOrUse.getSourceVariable() }
-
-  override string toString() { result = defOrUse.toString() }
-}
-
-class Phi extends TPhi, SsaDefOrUse {
-  PhiNode phi;
-
-  Phi() { this = TPhi(phi) }
-
-  final override PhiNode asPhi() { result = phi }
-
-  final override Location getLocation() { result = phi.getBasicBlock().getLocation() }
-
-  override string toString() { result = "Phi" }
-}
-
-class UseOrPhi extends SsaDefOrUse {
-  UseOrPhi() {
-    this.asDefOrUse() instanceof UseImpl
-    or
-    this instanceof Phi
-  }
-
-  final override Location getLocation() {
-    result = this.asDefOrUse().getLocation() or result = this.(Phi).getLocation()
-  }
-}
-
-class Def extends DefOrUse {
-  override DefImpl defOrUse;
-
-  Operand getAddressOperand() { result = defOrUse.getAddressOperand() }
-
-  Instruction getAddress() { result = this.getAddressOperand().getDef() }
-
-  /**
-   * This predicate ensures that joins go from `defOrUse` to the result
-   * instead of the other way around.
-   */
-  pragma[inline]
-  int getIndirectionIndex() {
-    pragma[only_bind_into](result) = pragma[only_bind_out](defOrUse).getIndirectionIndex()
-  }
-
-  Instruction getDefiningInstruction() { result = defOrUse.getDefiningInstruction() }
-}
-
-private module SsaImpl = SsaImplCommon::Make<SsaInput>;
-
-class PhiNode = SsaImpl::PhiNode;
-
-class Definition = SsaImpl::Definition;
-
-import SsaCached

cpp/ql/lib/experimental/semmle/code/cpp/ir/dataflow/internal/SsaInternalsCommon.qll

@@ -1,270 +0,0 @@
-import cpp as Cpp
-import semmle.code.cpp.ir.IR
-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
-
-/**
- * Holds if `operand` is an operand that is not used by the dataflow library.
- * Ignored operands are not recognizd 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 WriteSideEffectInstruction or
-    instr instanceof PhiInstruction or
-    instr instanceof ReadSideEffectInstruction or
-    instr instanceof ChiInstruction or
-    instr instanceof InitializeIndirectionInstruction
-  )
-}
-
-/**
- * Gets the C++ type of `this` in the member function `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()
-}
-
-/**
- * Gets the C++ type of the operand `operand`.
- * This is equivalent to the type of the operand's defining instruction.
- *
- * See `getResultLanguageType` for a description of this behavior.
- */
-CppType getLanguageType(Operand operand) { result = getResultLanguageType(operand.getDef()) }
-
-/**
- * Gets the maximum number of indirections a glvalue of type `type` can have.
- * For example:
- * - If `type = int`, the result is 1
- * - If `type = MyStruct`, the result is 1
- * - If `type = char*`, the result is 2
- */
-int getMaxIndirectionsForType(Type type) {
-  result = countIndirectionsForCppType(getTypeForGLValue(type))
-}
-
-/**
- * Gets the maximum number of indirections a value of type `type` can have.
- *
- * Note that this predicate is intended to be called on unspecified types
- * (i.e., `countIndirections(e.getUnspecifiedType())`).
- */
-private int countIndirections(Type t) {
-  result =
-    1 +
-      countIndirections([t.(Cpp::PointerType).getBaseType(), t.(Cpp::ReferenceType).getBaseType()])
-  or
-  not t instanceof Cpp::PointerType and
-  not t instanceof Cpp::ReferenceType and
-  result = 0
-}
-
-/**
- * Gets the maximum number of indirections a value of C++
- * type `langType` can have.
- */
-int countIndirectionsForCppType(LanguageType langType) {
-  exists(Type type | langType.hasType(type, true) |
-    result = 1 + countIndirections(type.getUnspecifiedType())
-  )
-  or
-  exists(Type type | langType.hasType(type, false) |
-    result = countIndirections(type.getUnspecifiedType())
-  )
-}
-
-/**
- * A `CallInstruction` that calls an allocation function such
- * as `malloc` or `operator new`.
- */
-class AllocationInstruction extends CallInstruction {
-  AllocationInstruction() { this.getStaticCallTarget() instanceof Cpp::AllocationFunction }
-}
-
-/**
- * Holds if `i` is a base instruction that starts a sequence of uses
- * of some variable that SSA can handle.
- *
- * This is either when `i` is a `VariableAddressInstruction` or when
- * `i` is a fresh allocation produced by an `AllocationInstruction`.
- */
-private predicate isSourceVariableBase(Instruction i) {
-  i instanceof VariableAddressInstruction or i instanceof AllocationInstruction
-}
-
-/**
- * Holds if the value pointed to by `operand` can potentially be
- * modified be the caller.
- */
-predicate isModifiableByCall(ArgumentOperand operand) {
-  exists(CallInstruction call, int index, CppType type |
-    type = getLanguageType(operand) and
-    call.getArgumentOperand(index) = operand and
-    if index = -1
-    then not call.getStaticCallTarget() instanceof Cpp::ConstMemberFunction
-    else not SideEffects::isConstPointerLike(any(Type t | type.hasType(t, _)))
-  )
-}
-
-cached
-private module Cached {
-  /**
-   * Holds if `op` is a use of an SSA variable rooted at `base` with `ind` number
-   * of indirections.
-   *
-   * `certain` is `true` if the operand is guaranteed to read the variable, and
-   * `indirectionIndex` specifies the number of loads required to read the variable.
-   */
-  cached
-  predicate isUse(boolean certain, Operand op, Instruction base, int ind, int indirectionIndex) {
-    not ignoreOperand(op) and
-    certain = true and
-    exists(LanguageType type, int m, int ind0 |
-      type = getLanguageType(op) and
-      m = countIndirectionsForCppType(type) and
-      isUseImpl(op, base, ind0) and
-      ind = ind0 + [0 .. m] and
-      indirectionIndex = ind - ind0
-    )
-  }
-
-  /**
-   * Holds if `operand` is a use of an SSA variable rooted at `base`, and the
-   * path from `base` to `operand` passes through `ind` load-like instructions.
-   */
-  private predicate isUseImpl(Operand operand, Instruction base, int ind) {
-    DataFlowImplCommon::forceCachingInSameStage() and
-    ind = 0 and
-    operand.getDef() = base and
-    isSourceVariableBase(base)
-    or
-    exists(Operand mid, Instruction instr |
-      isUseImpl(mid, base, ind) and
-      instr = operand.getDef() and
-      conversionFlow(mid, instr, false)
-    )
-    or
-    exists(int ind0 |
-      isUseImpl(operand.getDef().(LoadInstruction).getSourceAddressOperand(), base, ind0)
-      or
-      isUseImpl(operand.getDef().(InitializeParameterInstruction).getAnOperand(), base, ind0)
-    |
-      ind0 = ind - 1
-    )
-  }
-
-  /**
-   * Holds if `address` is an address of an SSA variable rooted at `base`,
-   * and `instr` is a definition of the SSA variable with `ind` number of indirections.
-   *
-   * `certain` is `true` if `instr` is guaranteed to write to the variable, and
-   * `indirectionIndex` specifies the number of loads required to read the variable
-   * after the write operation.
-   */
-  cached
-  predicate isDef(
-    boolean certain, Instruction instr, Operand address, Instruction base, int ind,
-    int indirectionIndex
-  ) {
-    certain = true and
-    exists(int ind0, CppType type, int m |
-      address =
-        [
-          instr.(StoreInstruction).getDestinationAddressOperand(),
-          instr.(InitializeParameterInstruction).getAnOperand(),
-          instr.(InitializeDynamicAllocationInstruction).getAllocationAddressOperand(),
-          instr.(UninitializedInstruction).getAnOperand()
-        ]
-    |
-      isDefImpl(address, base, ind0) and
-      type = getLanguageType(address) and
-      m = countIndirectionsForCppType(type) and
-      ind = ind0 + [1 .. m] and
-      indirectionIndex = ind - (ind0 + 1)
-    )
-  }
-
-  /**
-   * Holds if `address` is a use of an SSA variable rooted at `base`, and the
-   * path from `base` to `address` passes through `ind` load-like instructions.
-   *
-   * Note: Unlike `isUseImpl`, this predicate recurses through pointer-arithmetic
-   * instructions.
-   */
-  private predicate isDefImpl(Operand address, Instruction base, int ind) {
-    DataFlowImplCommon::forceCachingInSameStage() and
-    ind = 0 and
-    address.getDef() = base and
-    isSourceVariableBase(base)
-    or
-    exists(Operand mid, Instruction instr |
-      isDefImpl(mid, base, ind) and
-      instr = address.getDef() and
-      conversionFlow(mid, instr, _)
-    )
-    or
-    exists(int ind0 |
-      isDefImpl(address.getDef().(LoadInstruction).getSourceAddressOperand(), base, ind0)
-      or
-      isDefImpl(address.getDef().(InitializeParameterInstruction).getAnOperand(), base, ind0)
-    |
-      ind0 = ind - 1
-    )
-  }
-}
-
-import Cached
-
-/**
- * Inputs to the shared SSA library's parameterized module that is shared
- * between the SSA pruning stage, and the final SSA stage.
- */
-module InputSigCommon {
-  class BasicBlock = IRBlock;
-
-  BasicBlock getImmediateBasicBlockDominator(BasicBlock bb) { result.immediatelyDominates(bb) }
-
-  BasicBlock getABasicBlockSuccessor(BasicBlock bb) { result = bb.getASuccessor() }
-
-  class ExitBasicBlock extends IRBlock {
-    ExitBasicBlock() { this.getLastInstruction() instanceof ExitFunctionInstruction }
-  }
-}

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

@@ -1,208 +0,0 @@
-private import semmle.code.cpp.ir.IR
-private import experimental.semmle.code.cpp.ir.dataflow.DataFlow
-private import ModelUtil
-private import semmle.code.cpp.models.interfaces.DataFlow
-private import semmle.code.cpp.models.interfaces.SideEffect
-private import DataFlowUtil
-private import DataFlowPrivate
-private import semmle.code.cpp.models.Models
-
-/**
- * Holds if taint propagates from `nodeFrom` to `nodeTo` in exactly one local
- * (intra-procedural) step.
- */
-predicate localTaintStep(DataFlow::Node nodeFrom, DataFlow::Node nodeTo) {
-  DataFlow::localFlowStep(nodeFrom, nodeTo)
-  or
-  localAdditionalTaintStep(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) {
-  operandToInstructionTaintStep(nodeFrom.asOperand(), nodeTo.asInstruction())
-  or
-  modeledTaintStep(nodeFrom, nodeTo)
-  or
-  // Flow from `op` to `*op`.
-  exists(Operand operand, int indirectionIndex |
-    nodeHasOperand(nodeFrom, operand, indirectionIndex) and
-    nodeHasOperand(nodeTo, operand, indirectionIndex - 1)
-  )
-  or
-  // Flow from `instr` to `*instr`.
-  exists(Instruction instr, int indirectionIndex |
-    nodeHasInstruction(nodeFrom, instr, indirectionIndex) and
-    nodeHasInstruction(nodeTo, instr, indirectionIndex - 1)
-  )
-  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)
-  )
-}
-
-/**
- * Holds if taint propagates from `nodeFrom` to `nodeTo` in exactly one local
- * (intra-procedural) step.
- */
-private predicate operandToInstructionTaintStep(Operand opFrom, Instruction instrTo) {
-  // Taint can flow through expressions that alter the value but preserve
-  // more than one bit of it _or_ expressions that follow data through
-  // pointer indirections.
-  instrTo.getAnOperand() = opFrom and
-  (
-    instrTo instanceof ArithmeticInstruction
-    or
-    instrTo instanceof BitwiseInstruction
-    or
-    instrTo instanceof PointerArithmeticInstruction
-  )
-  or
-  // The `CopyInstruction` case is also present in non-taint data flow, but
-  // that uses `getDef` rather than `getAnyDef`. For taint, we want flow
-  // from a definition of `myStruct` to a `myStruct.myField` expression.
-  instrTo.(LoadInstruction).getSourceAddressOperand() = opFrom
-  or
-  // Unary instructions tend to preserve enough information in practice that we
-  // want taint to flow through.
-  // The exception is `FieldAddressInstruction`. Together with the rules below for
-  // `LoadInstruction`s and `ChiInstruction`s, flow through `FieldAddressInstruction`
-  // could cause flow into one field to come out an unrelated field.
-  // This would happen across function boundaries, where the IR would not be able to
-  // match loads to stores.
-  instrTo.(UnaryInstruction).getUnaryOperand() = opFrom and
-  (
-    not instrTo instanceof FieldAddressInstruction
-    or
-    instrTo.(FieldAddressInstruction).getField().getDeclaringType() instanceof Union
-  )
-}
-
-/**
- * Holds if taint may propagate from `source` to `sink` in zero or more local
- * (intra-procedural) steps.
- */
-pragma[inline]
-predicate localTaint(DataFlow::Node source, DataFlow::Node sink) { localTaintStep*(source, sink) }
-
-/**
- * Holds if taint can flow from `i1` to `i2` in zero or more
- * local (intra-procedural) steps.
- */
-pragma[inline]
-predicate localInstructionTaint(Instruction i1, Instruction i2) {
-  localTaint(DataFlow::instructionNode(i1), DataFlow::instructionNode(i2))
-}
-
-/**
- * Holds if taint can flow from `e1` to `e2` in zero or more
- * local (intra-procedural) steps.
- */
-pragma[inline]
-predicate localExprTaint(Expr e1, Expr e2) {
-  localTaint(DataFlow::exprNode(e1), DataFlow::exprNode(e2))
-}
-
-/**
- * Holds if the additional step from `src` to `sink` should be included in all
- * global taint flow configurations.
- */
-predicate defaultAdditionalTaintStep(DataFlow::Node src, DataFlow::Node sink) {
-  localAdditionalTaintStep(src, sink)
-}
-
-/**
- * Holds if default `TaintTracking::Configuration`s should allow implicit reads
- * of `c` at sinks and inputs to additional taint steps.
- */
-bindingset[node]
-predicate defaultImplicitTaintRead(DataFlow::Node node, DataFlow::Content c) { none() }
-
-/**
- * Holds if `node` should be a sanitizer in all global taint flow configurations
- * but not in local taint.
- */
-predicate defaultTaintSanitizer(DataFlow::Node node) { none() }
-
-/**
- * Holds if taint can flow from `instrIn` to `instrOut` through a call to a
- * modeled function.
- */
-predicate modeledTaintStep(DataFlow::Node nodeIn, DataFlow::Node nodeOut) {
-  // Normal taint steps
-  exists(CallInstruction call, TaintFunction func, FunctionInput modelIn, FunctionOutput modelOut |
-    call.getStaticCallTarget() = func and
-    func.hasTaintFlow(modelIn, modelOut)
-  |
-    (
-      nodeIn = callInput(call, modelIn)
-      or
-      exists(int n |
-        modelIn.isParameterDerefOrQualifierObject(n) and
-        if n = -1
-        then nodeIn = callInput(call, any(InQualifierAddress inQualifier))
-        else nodeIn = callInput(call, any(InParameter inParam | inParam.getIndex() = n))
-      )
-    ) and
-    nodeOut = callOutput(call, modelOut)
-    or
-    exists(int d |
-      nodeIn = callInput(call, modelIn, d)
-      or
-      exists(int n |
-        d = 1 and
-        modelIn.isParameterDerefOrQualifierObject(n) and
-        if n = -1
-        then nodeIn = callInput(call, any(InQualifierAddress inQualifier))
-        else nodeIn = callInput(call, any(InParameter inParam | inParam.getIndex() = n))
-      )
-    |
-      call.getStaticCallTarget() = func and
-      func.hasTaintFlow(modelIn, modelOut) and
-      nodeOut = callOutput(call, modelOut, d)
-    )
-  )
-  or
-  // Taint flow from one argument to another and data flow from an argument to a
-  // return value. This happens in functions like `strcat` and `memcpy`. We
-  // could model this flow in two separate steps, but that would add reverse
-  // flow from the write side-effect to the call instruction, which may not be
-  // desirable.
-  exists(
-    CallInstruction call, Function func, FunctionInput modelIn, OutParameterDeref modelMidOut,
-    int indexMid, InParameter modelMidIn, OutReturnValue modelOut
-  |
-    nodeIn = callInput(call, modelIn) and
-    nodeOut = callOutput(call, modelOut) and
-    call.getStaticCallTarget() = func and
-    func.(TaintFunction).hasTaintFlow(modelIn, modelMidOut) and
-    func.(DataFlowFunction).hasDataFlow(modelMidIn, modelOut) and
-    modelMidOut.isParameterDeref(indexMid) and
-    modelMidIn.isParameter(indexMid)
-  )
-  or
-  // 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,
-    FunctionInput modelIn, FunctionOutput modelOut
-  |
-    indirectArgument = callInput(call, modelIn) and
-    indirectArgument.getAddressOperand() = nodeIn.asOperand() and
-    call.getStaticCallTarget() = func and
-    (
-      func.(DataFlowFunction).hasDataFlow(modelIn, modelOut)
-      or
-      func.(TaintFunction).hasTaintFlow(modelIn, modelOut)
-    ) and
-    nodeOut = callOutput(call, modelOut)
-  )
-}

cpp/ql/lib/experimental/semmle/code/cpp/ir/dataflow/internal/ssa0/SsaInternals.qll

@@ -1,314 +0,0 @@
-/**
- * This module defines an initial SSA pruning stage that doesn't take
- * indirections into account.
- */
-
-private import codeql.ssa.Ssa as SsaImplCommon
-private import semmle.code.cpp.ir.IR
-private import experimental.semmle.code.cpp.ir.dataflow.internal.DataFlowImplCommon as DataFlowImplCommon
-private import semmle.code.cpp.models.interfaces.Allocation as Alloc
-private import semmle.code.cpp.models.interfaces.DataFlow as DataFlow
-private import semmle.code.cpp.ir.implementation.raw.internal.SideEffects as SideEffects
-private import semmle.code.cpp.ir.internal.IRCppLanguage
-private import experimental.semmle.code.cpp.ir.dataflow.internal.DataFlowPrivate
-private import experimental.semmle.code.cpp.ir.dataflow.internal.DataFlowUtil
-private import experimental.semmle.code.cpp.ir.dataflow.internal.SsaInternalsCommon
-
-private module SourceVariables {
-  newtype TBaseSourceVariable =
-    // Each IR variable gets its own source variable
-    TBaseIRVariable(IRVariable var) or
-    // Each allocation gets its own source variable
-    TBaseCallVariable(AllocationInstruction call)
-
-  abstract class BaseSourceVariable extends TBaseSourceVariable {
-    abstract string toString();
-
-    abstract DataFlowType getType();
-  }
-
-  class BaseIRVariable extends BaseSourceVariable, TBaseIRVariable {
-    IRVariable var;
-
-    IRVariable getIRVariable() { result = var }
-
-    BaseIRVariable() { this = TBaseIRVariable(var) }
-
-    override string toString() { result = var.toString() }
-
-    override DataFlowType getType() { result = var.getType() }
-  }
-
-  class BaseCallVariable extends BaseSourceVariable, TBaseCallVariable {
-    AllocationInstruction call;
-
-    BaseCallVariable() { this = TBaseCallVariable(call) }
-
-    AllocationInstruction getCallInstruction() { result = call }
-
-    override string toString() { result = call.toString() }
-
-    override DataFlowType getType() { result = call.getResultType() }
-  }
-
-  private newtype TSourceVariable =
-    TSourceIRVariable(BaseIRVariable baseVar) or
-    TCallVariable(AllocationInstruction call)
-
-  abstract class SourceVariable extends TSourceVariable {
-    abstract string toString();
-
-    abstract BaseSourceVariable getBaseVariable();
-  }
-
-  class SourceIRVariable extends SourceVariable, TSourceIRVariable {
-    BaseIRVariable var;
-
-    SourceIRVariable() { this = TSourceIRVariable(var) }
-
-    IRVariable getIRVariable() { result = var.getIRVariable() }
-
-    override BaseIRVariable getBaseVariable() { result.getIRVariable() = this.getIRVariable() }
-
-    override string toString() { result = this.getIRVariable().toString() }
-  }
-
-  class CallVariable extends SourceVariable, TCallVariable {
-    AllocationInstruction call;
-
-    CallVariable() { this = TCallVariable(call) }
-
-    AllocationInstruction getCall() { result = call }
-
-    override BaseCallVariable getBaseVariable() { result.getCallInstruction() = call }
-
-    override string toString() { result = "Call" }
-  }
-}
-
-import SourceVariables
-
-private newtype TDefOrUseImpl =
-  TDefImpl(Operand address) { isDef(_, _, address, _, _, _) } or
-  TUseImpl(Operand operand) {
-    isUse(_, operand, _, _, _) and
-    not isDef(_, _, operand, _, _, _)
-  }
-
-abstract private class DefOrUseImpl extends TDefOrUseImpl {
-  /** Gets a textual representation of this element. */
-  abstract string toString();
-
-  /** Gets the block of this definition or use. */
-  abstract IRBlock getBlock();
-
-  /** Holds if this definition or use has index `index` in block `block`. */
-  abstract predicate hasIndexInBlock(IRBlock block, int index);
-
-  final predicate hasIndexInBlock(IRBlock block, int index, SourceVariable sv) {
-    this.hasIndexInBlock(block, index) and
-    sv = this.getSourceVariable()
-  }
-
-  /** Gets the location of this element. */
-  abstract Cpp::Location getLocation();
-
-  abstract Instruction getBase();
-
-  final BaseSourceVariable getBaseSourceVariable() {
-    exists(IRVariable var |
-      result.(BaseIRVariable).getIRVariable() = var and
-      instructionHasIRVariable(this.getBase(), var)
-    )
-    or
-    result.(BaseCallVariable).getCallInstruction() = this.getBase()
-  }
-
-  /** Gets the variable that is defined or used. */
-  final SourceVariable getSourceVariable() {
-    exists(BaseSourceVariable v |
-      sourceVariableHasBaseAndIndex(result, v) and
-      defOrUseHasSourceVariable(this, v)
-    )
-  }
-}
-
-pragma[noinline]
-private predicate instructionHasIRVariable(VariableAddressInstruction vai, IRVariable var) {
-  vai.getIRVariable() = var
-}
-
-private predicate defOrUseHasSourceVariable(DefOrUseImpl defOrUse, BaseSourceVariable bv) {
-  defHasSourceVariable(defOrUse, bv)
-  or
-  useHasSourceVariable(defOrUse, bv)
-}
-
-pragma[noinline]
-private predicate defHasSourceVariable(DefImpl def, BaseSourceVariable bv) {
-  bv = def.getBaseSourceVariable()
-}
-
-pragma[noinline]
-private predicate useHasSourceVariable(UseImpl use, BaseSourceVariable bv) {
-  bv = use.getBaseSourceVariable()
-}
-
-pragma[noinline]
-private predicate sourceVariableHasBaseAndIndex(SourceVariable v, BaseSourceVariable bv) {
-  v.getBaseVariable() = bv
-}
-
-class DefImpl extends DefOrUseImpl, TDefImpl {
-  Operand address;
-
-  DefImpl() { this = TDefImpl(address) }
-
-  override Instruction getBase() { isDef(_, _, address, result, _, _) }
-
-  Operand getAddressOperand() { result = address }
-
-  Instruction getDefiningInstruction() { isDef(_, result, address, _, _, _) }
-
-  override string toString() { result = address.toString() }
-
-  override IRBlock getBlock() { result = this.getDefiningInstruction().getBlock() }
-
-  override Cpp::Location getLocation() { result = this.getDefiningInstruction().getLocation() }
-
-  final override predicate hasIndexInBlock(IRBlock block, int index) {
-    this.getDefiningInstruction() = block.getInstruction(index)
-  }
-
-  predicate isCertain() { isDef(true, _, address, _, _, _) }
-}
-
-class UseImpl extends DefOrUseImpl, TUseImpl {
-  Operand operand;
-
-  UseImpl() { this = TUseImpl(operand) }
-
-  Operand getOperand() { result = operand }
-
-  override string toString() { result = operand.toString() }
-
-  final override predicate hasIndexInBlock(IRBlock block, int index) {
-    operand.getUse() = block.getInstruction(index)
-  }
-
-  final override IRBlock getBlock() { result = operand.getUse().getBlock() }
-
-  final override Cpp::Location getLocation() { result = operand.getLocation() }
-
-  override Instruction getBase() { isUse(_, operand, result, _, _) }
-
-  predicate isCertain() { isUse(true, operand, _, _, _) }
-}
-
-private module SsaInput implements SsaImplCommon::InputSig {
-  import InputSigCommon
-  import SourceVariables
-
-  /**
-   * Holds if the `i`'th write in block `bb` writes to the variable `v`.
-   * `certain` is `true` if the write is guaranteed to overwrite the entire variable.
-   */
-  predicate variableWrite(IRBlock bb, int i, SourceVariable v, boolean certain) {
-    DataFlowImplCommon::forceCachingInSameStage() and
-    exists(DefImpl def | def.hasIndexInBlock(bb, i, v) |
-      if def.isCertain() then certain = true else certain = false
-    )
-  }
-
-  /**
-   * Holds if the `i`'th read in block `bb` reads to the variable `v`.
-   * `certain` is `true` if the read is guaranteed.
-   */
-  predicate variableRead(IRBlock bb, int i, SourceVariable v, boolean certain) {
-    exists(UseImpl use | use.hasIndexInBlock(bb, i, v) |
-      if use.isCertain() then certain = true else certain = false
-    )
-  }
-}
-
-private newtype TSsaDefOrUse =
-  TDefOrUse(DefOrUseImpl defOrUse) {
-    defOrUse instanceof UseImpl
-    or
-    // If `defOrUse` is a definition we only include it if the
-    // SSA library concludes that it's live after the write.
-    exists(Definition def, SourceVariable sv, IRBlock bb, int i |
-      def.definesAt(sv, bb, i) and
-      defOrUse.(DefImpl).hasIndexInBlock(bb, i, sv)
-    )
-  } or
-  TPhi(PhiNode phi)
-
-abstract private class SsaDefOrUse extends TSsaDefOrUse {
-  string toString() { result = "SsaDefOrUse" }
-
-  DefOrUseImpl asDefOrUse() { none() }
-
-  PhiNode asPhi() { none() }
-
-  abstract Location getLocation();
-}
-
-class DefOrUse extends TDefOrUse, SsaDefOrUse {
-  DefOrUseImpl defOrUse;
-
-  DefOrUse() { this = TDefOrUse(defOrUse) }
-
-  final override DefOrUseImpl asDefOrUse() { result = defOrUse }
-
-  final override Location getLocation() { result = defOrUse.getLocation() }
-
-  final SourceVariable getSourceVariable() { result = defOrUse.getSourceVariable() }
-}
-
-class Phi extends TPhi, SsaDefOrUse {
-  PhiNode phi;
-
-  Phi() { this = TPhi(phi) }
-
-  final override PhiNode asPhi() { result = phi }
-
-  final override Location getLocation() { result = phi.getBasicBlock().getLocation() }
-}
-
-class UseOrPhi extends SsaDefOrUse {
-  UseOrPhi() {
-    this.asDefOrUse() instanceof UseImpl
-    or
-    this instanceof Phi
-  }
-
-  final override Location getLocation() {
-    result = this.asDefOrUse().getLocation() or result = this.(Phi).getLocation()
-  }
-
-  override string toString() {
-    result = this.asDefOrUse().toString()
-    or
-    this instanceof Phi and
-    result = "Phi"
-  }
-}
-
-class Def extends DefOrUse {
-  override DefImpl defOrUse;
-
-  Operand getAddressOperand() { result = defOrUse.getAddressOperand() }
-
-  Instruction getAddress() { result = this.getAddressOperand().getDef() }
-
-  Instruction getDefiningInstruction() { result = defOrUse.getDefiningInstruction() }
-
-  override string toString() { result = this.asDefOrUse().toString() + " (def)" }
-}
-
-private module SsaImpl = SsaImplCommon::Make<SsaInput>;
-
-class PhiNode = SsaImpl::PhiNode;
-
-class Definition = SsaImpl::Definition;

cpp/ql/lib/experimental/semmle/code/cpp/ir/dataflow/internal/tainttracking1/TaintTrackingImpl.qll

@@ -1,191 +0,0 @@
-/**
- * Provides an implementation of global (interprocedural) taint tracking.
- * This file re-exports the local (intraprocedural) taint-tracking analysis
- * from `TaintTrackingParameter::Public` and adds a global analysis, mainly
- * exposed through the `Configuration` class. For some languages, this file
- * exists in several identical copies, allowing queries to use multiple
- * `Configuration` classes that depend on each other without introducing
- * mutual recursion among those configurations.
- */
-
-import TaintTrackingParameter::Public
-private import TaintTrackingParameter::Private
-
-/**
- * A configuration of interprocedural taint tracking analysis. This defines
- * sources, sinks, and any other configurable aspect of the analysis. Each
- * use of the taint tracking library must define its own unique extension of
- * this abstract class.
- *
- * A taint-tracking configuration is a special data flow configuration
- * (`DataFlow::Configuration`) that allows for flow through nodes that do not
- * necessarily preserve values but are still relevant from a taint tracking
- * perspective. (For example, string concatenation, where one of the operands
- * is tainted.)
- *
- * To create a configuration, extend this class with a subclass whose
- * characteristic predicate is a unique singleton string. For example, write
- *
- * ```ql
- * class MyAnalysisConfiguration extends TaintTracking::Configuration {
- *   MyAnalysisConfiguration() { this = "MyAnalysisConfiguration" }
- *   // Override `isSource` and `isSink`.
- *   // Optionally override `isSanitizer`.
- *   // Optionally override `isSanitizerIn`.
- *   // Optionally override `isSanitizerOut`.
- *   // Optionally override `isSanitizerGuard`.
- *   // Optionally override `isAdditionalTaintStep`.
- * }
- * ```
- *
- * Then, to query whether there is flow between some `source` and `sink`,
- * write
- *
- * ```ql
- * exists(MyAnalysisConfiguration cfg | cfg.hasFlow(source, sink))
- * ```
- *
- * Multiple configurations can coexist, but it is unsupported to depend on
- * another `TaintTracking::Configuration` or a `DataFlow::Configuration` in the
- * overridden predicates that define sources, sinks, or additional steps.
- * Instead, the dependency should go to a `TaintTracking2::Configuration` or a
- * `DataFlow2::Configuration`, `DataFlow3::Configuration`, etc.
- */
-abstract class Configuration extends DataFlow::Configuration {
-  bindingset[this]
-  Configuration() { any() }
-
-  /**
-   * Holds if `source` is a relevant taint source.
-   *
-   * The smaller this predicate is, the faster `hasFlow()` will converge.
-   */
-  // overridden to provide taint-tracking specific qldoc
-  override predicate isSource(DataFlow::Node source) { none() }
-
-  /**
-   * Holds if `source` is a relevant taint source with the given initial
-   * `state`.
-   *
-   * The smaller this predicate is, the faster `hasFlow()` will converge.
-   */
-  // overridden to provide taint-tracking specific qldoc
-  override predicate isSource(DataFlow::Node source, DataFlow::FlowState state) { none() }
-
-  /**
-   * Holds if `sink` is a relevant taint sink
-   *
-   * The smaller this predicate is, the faster `hasFlow()` will converge.
-   */
-  // overridden to provide taint-tracking specific qldoc
-  override predicate isSink(DataFlow::Node sink) { none() }
-
-  /**
-   * Holds if `sink` is a relevant taint sink accepting `state`.
-   *
-   * The smaller this predicate is, the faster `hasFlow()` will converge.
-   */
-  // overridden to provide taint-tracking specific qldoc
-  override predicate isSink(DataFlow::Node sink, DataFlow::FlowState state) { none() }
-
-  /** Holds if the node `node` is a taint sanitizer. */
-  predicate isSanitizer(DataFlow::Node node) { none() }
-
-  final override predicate isBarrier(DataFlow::Node node) {
-    this.isSanitizer(node) or
-    defaultTaintSanitizer(node)
-  }
-
-  /**
-   * Holds if the node `node` is a taint sanitizer when the flow state is
-   * `state`.
-   */
-  predicate isSanitizer(DataFlow::Node node, DataFlow::FlowState state) { none() }
-
-  final override predicate isBarrier(DataFlow::Node node, DataFlow::FlowState state) {
-    this.isSanitizer(node, state)
-  }
-
-  /** Holds if taint propagation into `node` is prohibited. */
-  predicate isSanitizerIn(DataFlow::Node node) { none() }
-
-  final override predicate isBarrierIn(DataFlow::Node node) { this.isSanitizerIn(node) }
-
-  /** Holds if taint propagation out of `node` is prohibited. */
-  predicate isSanitizerOut(DataFlow::Node node) { none() }
-
-  final override predicate isBarrierOut(DataFlow::Node node) { this.isSanitizerOut(node) }
-
-  /**
-   * DEPRECATED: Use `isSanitizer` and `BarrierGuard` module instead.
-   *
-   * Holds if taint propagation through nodes guarded by `guard` is prohibited.
-   */
-  deprecated predicate isSanitizerGuard(DataFlow::BarrierGuard guard) { none() }
-
-  deprecated final override predicate isBarrierGuard(DataFlow::BarrierGuard guard) {
-    this.isSanitizerGuard(guard)
-  }
-
-  /**
-   * DEPRECATED: Use `isSanitizer` and `BarrierGuard` module instead.
-   *
-   * Holds if taint propagation through nodes guarded by `guard` is prohibited
-   * when the flow state is `state`.
-   */
-  deprecated predicate isSanitizerGuard(DataFlow::BarrierGuard guard, DataFlow::FlowState state) {
-    none()
-  }
-
-  deprecated final override predicate isBarrierGuard(
-    DataFlow::BarrierGuard guard, DataFlow::FlowState state
-  ) {
-    this.isSanitizerGuard(guard, state)
-  }
-
-  /**
-   * Holds if taint may propagate from `node1` to `node2` in addition to the normal data-flow and taint steps.
-   */
-  predicate isAdditionalTaintStep(DataFlow::Node node1, DataFlow::Node node2) { none() }
-
-  final override predicate isAdditionalFlowStep(DataFlow::Node node1, DataFlow::Node node2) {
-    this.isAdditionalTaintStep(node1, node2) or
-    defaultAdditionalTaintStep(node1, node2)
-  }
-
-  /**
-   * Holds if taint may propagate from `node1` to `node2` in addition to the normal data-flow and taint steps.
-   * This step is only applicable in `state1` and updates the flow state to `state2`.
-   */
-  predicate isAdditionalTaintStep(
-    DataFlow::Node node1, DataFlow::FlowState state1, DataFlow::Node node2,
-    DataFlow::FlowState state2
-  ) {
-    none()
-  }
-
-  final override predicate isAdditionalFlowStep(
-    DataFlow::Node node1, DataFlow::FlowState state1, DataFlow::Node node2,
-    DataFlow::FlowState state2
-  ) {
-    this.isAdditionalTaintStep(node1, state1, node2, state2)
-  }
-
-  override predicate allowImplicitRead(DataFlow::Node node, DataFlow::ContentSet c) {
-    (
-      this.isSink(node) or
-      this.isSink(node, _) or
-      this.isAdditionalTaintStep(node, _) or
-      this.isAdditionalTaintStep(node, _, _, _)
-    ) and
-    defaultImplicitTaintRead(node, c)
-  }
-
-  /**
-   * Holds if taint may flow from `source` to `sink` for this configuration.
-   */
-  // overridden to provide taint-tracking specific qldoc
-  override predicate hasFlow(DataFlow::Node source, DataFlow::Node sink) {
-    super.hasFlow(source, sink)
-  }
-}

cpp/ql/lib/experimental/semmle/code/cpp/ir/dataflow/internal/tainttracking1/TaintTrackingParameter.qll

@@ -1,5 +0,0 @@
-import experimental.semmle.code.cpp.ir.dataflow.internal.TaintTrackingUtil as Public
-
-module Private {
-  import experimental.semmle.code.cpp.ir.dataflow.DataFlow::DataFlow as DataFlow
-}

cpp/ql/lib/experimental/semmle/code/cpp/ir/dataflow/internal/tainttracking2/TaintTrackingImpl.qll

@@ -1,191 +0,0 @@
-/**
- * Provides an implementation of global (interprocedural) taint tracking.
- * This file re-exports the local (intraprocedural) taint-tracking analysis
- * from `TaintTrackingParameter::Public` and adds a global analysis, mainly
- * exposed through the `Configuration` class. For some languages, this file
- * exists in several identical copies, allowing queries to use multiple
- * `Configuration` classes that depend on each other without introducing
- * mutual recursion among those configurations.
- */
-
-import TaintTrackingParameter::Public
-private import TaintTrackingParameter::Private
-
-/**
- * A configuration of interprocedural taint tracking analysis. This defines
- * sources, sinks, and any other configurable aspect of the analysis. Each
- * use of the taint tracking library must define its own unique extension of
- * this abstract class.
- *
- * A taint-tracking configuration is a special data flow configuration
- * (`DataFlow::Configuration`) that allows for flow through nodes that do not
- * necessarily preserve values but are still relevant from a taint tracking
- * perspective. (For example, string concatenation, where one of the operands
- * is tainted.)
- *
- * To create a configuration, extend this class with a subclass whose
- * characteristic predicate is a unique singleton string. For example, write
- *
- * ```ql
- * class MyAnalysisConfiguration extends TaintTracking::Configuration {
- *   MyAnalysisConfiguration() { this = "MyAnalysisConfiguration" }
- *   // Override `isSource` and `isSink`.
- *   // Optionally override `isSanitizer`.
- *   // Optionally override `isSanitizerIn`.
- *   // Optionally override `isSanitizerOut`.
- *   // Optionally override `isSanitizerGuard`.
- *   // Optionally override `isAdditionalTaintStep`.
- * }
- * ```
- *
- * Then, to query whether there is flow between some `source` and `sink`,
- * write
- *
- * ```ql
- * exists(MyAnalysisConfiguration cfg | cfg.hasFlow(source, sink))
- * ```
- *
- * Multiple configurations can coexist, but it is unsupported to depend on
- * another `TaintTracking::Configuration` or a `DataFlow::Configuration` in the
- * overridden predicates that define sources, sinks, or additional steps.
- * Instead, the dependency should go to a `TaintTracking2::Configuration` or a
- * `DataFlow2::Configuration`, `DataFlow3::Configuration`, etc.
- */
-abstract class Configuration extends DataFlow::Configuration {
-  bindingset[this]
-  Configuration() { any() }
-
-  /**
-   * Holds if `source` is a relevant taint source.
-   *
-   * The smaller this predicate is, the faster `hasFlow()` will converge.
-   */
-  // overridden to provide taint-tracking specific qldoc
-  override predicate isSource(DataFlow::Node source) { none() }
-
-  /**
-   * Holds if `source` is a relevant taint source with the given initial
-   * `state`.
-   *
-   * The smaller this predicate is, the faster `hasFlow()` will converge.
-   */
-  // overridden to provide taint-tracking specific qldoc
-  override predicate isSource(DataFlow::Node source, DataFlow::FlowState state) { none() }
-
-  /**
-   * Holds if `sink` is a relevant taint sink
-   *
-   * The smaller this predicate is, the faster `hasFlow()` will converge.
-   */
-  // overridden to provide taint-tracking specific qldoc
-  override predicate isSink(DataFlow::Node sink) { none() }
-
-  /**
-   * Holds if `sink` is a relevant taint sink accepting `state`.
-   *
-   * The smaller this predicate is, the faster `hasFlow()` will converge.
-   */
-  // overridden to provide taint-tracking specific qldoc
-  override predicate isSink(DataFlow::Node sink, DataFlow::FlowState state) { none() }
-
-  /** Holds if the node `node` is a taint sanitizer. */
-  predicate isSanitizer(DataFlow::Node node) { none() }
-
-  final override predicate isBarrier(DataFlow::Node node) {
-    this.isSanitizer(node) or
-    defaultTaintSanitizer(node)
-  }
-
-  /**
-   * Holds if the node `node` is a taint sanitizer when the flow state is
-   * `state`.
-   */
-  predicate isSanitizer(DataFlow::Node node, DataFlow::FlowState state) { none() }
-
-  final override predicate isBarrier(DataFlow::Node node, DataFlow::FlowState state) {
-    this.isSanitizer(node, state)
-  }
-
-  /** Holds if taint propagation into `node` is prohibited. */
-  predicate isSanitizerIn(DataFlow::Node node) { none() }
-
-  final override predicate isBarrierIn(DataFlow::Node node) { this.isSanitizerIn(node) }
-
-  /** Holds if taint propagation out of `node` is prohibited. */
-  predicate isSanitizerOut(DataFlow::Node node) { none() }
-
-  final override predicate isBarrierOut(DataFlow::Node node) { this.isSanitizerOut(node) }
-
-  /**
-   * DEPRECATED: Use `isSanitizer` and `BarrierGuard` module instead.
-   *
-   * Holds if taint propagation through nodes guarded by `guard` is prohibited.
-   */
-  deprecated predicate isSanitizerGuard(DataFlow::BarrierGuard guard) { none() }
-
-  deprecated final override predicate isBarrierGuard(DataFlow::BarrierGuard guard) {
-    this.isSanitizerGuard(guard)
-  }
-
-  /**
-   * DEPRECATED: Use `isSanitizer` and `BarrierGuard` module instead.
-   *
-   * Holds if taint propagation through nodes guarded by `guard` is prohibited
-   * when the flow state is `state`.
-   */
-  deprecated predicate isSanitizerGuard(DataFlow::BarrierGuard guard, DataFlow::FlowState state) {
-    none()
-  }
-
-  deprecated final override predicate isBarrierGuard(
-    DataFlow::BarrierGuard guard, DataFlow::FlowState state
-  ) {
-    this.isSanitizerGuard(guard, state)
-  }
-
-  /**
-   * Holds if taint may propagate from `node1` to `node2` in addition to the normal data-flow and taint steps.
-   */
-  predicate isAdditionalTaintStep(DataFlow::Node node1, DataFlow::Node node2) { none() }
-
-  final override predicate isAdditionalFlowStep(DataFlow::Node node1, DataFlow::Node node2) {
-    this.isAdditionalTaintStep(node1, node2) or
-    defaultAdditionalTaintStep(node1, node2)
-  }
-
-  /**
-   * Holds if taint may propagate from `node1` to `node2` in addition to the normal data-flow and taint steps.
-   * This step is only applicable in `state1` and updates the flow state to `state2`.
-   */
-  predicate isAdditionalTaintStep(
-    DataFlow::Node node1, DataFlow::FlowState state1, DataFlow::Node node2,
-    DataFlow::FlowState state2
-  ) {
-    none()
-  }
-
-  final override predicate isAdditionalFlowStep(
-    DataFlow::Node node1, DataFlow::FlowState state1, DataFlow::Node node2,
-    DataFlow::FlowState state2
-  ) {
-    this.isAdditionalTaintStep(node1, state1, node2, state2)
-  }
-
-  override predicate allowImplicitRead(DataFlow::Node node, DataFlow::ContentSet c) {
-    (
-      this.isSink(node) or
-      this.isSink(node, _) or
-      this.isAdditionalTaintStep(node, _) or
-      this.isAdditionalTaintStep(node, _, _, _)
-    ) and
-    defaultImplicitTaintRead(node, c)
-  }
-
-  /**
-   * Holds if taint may flow from `source` to `sink` for this configuration.
-   */
-  // overridden to provide taint-tracking specific qldoc
-  override predicate hasFlow(DataFlow::Node source, DataFlow::Node sink) {
-    super.hasFlow(source, sink)
-  }
-}

cpp/ql/lib/experimental/semmle/code/cpp/ir/dataflow/internal/tainttracking2/TaintTrackingParameter.qll

@@ -1,5 +0,0 @@
-import experimental.semmle.code.cpp.ir.dataflow.internal.TaintTrackingUtil as Public
-
-module Private {
-  import experimental.semmle.code.cpp.ir.dataflow.DataFlow2::DataFlow2 as DataFlow
-}

cpp/ql/lib/experimental/semmle/code/cpp/ir/dataflow/internal/tainttracking3/TaintTrackingImpl.qll

@@ -1,191 +0,0 @@
-/**
- * Provides an implementation of global (interprocedural) taint tracking.
- * This file re-exports the local (intraprocedural) taint-tracking analysis
- * from `TaintTrackingParameter::Public` and adds a global analysis, mainly
- * exposed through the `Configuration` class. For some languages, this file
- * exists in several identical copies, allowing queries to use multiple
- * `Configuration` classes that depend on each other without introducing
- * mutual recursion among those configurations.
- */
-
-import TaintTrackingParameter::Public
-private import TaintTrackingParameter::Private
-
-/**
- * A configuration of interprocedural taint tracking analysis. This defines
- * sources, sinks, and any other configurable aspect of the analysis. Each
- * use of the taint tracking library must define its own unique extension of
- * this abstract class.
- *
- * A taint-tracking configuration is a special data flow configuration
- * (`DataFlow::Configuration`) that allows for flow through nodes that do not
- * necessarily preserve values but are still relevant from a taint tracking
- * perspective. (For example, string concatenation, where one of the operands
- * is tainted.)
- *
- * To create a configuration, extend this class with a subclass whose
- * characteristic predicate is a unique singleton string. For example, write
- *
- * ```ql
- * class MyAnalysisConfiguration extends TaintTracking::Configuration {
- *   MyAnalysisConfiguration() { this = "MyAnalysisConfiguration" }
- *   // Override `isSource` and `isSink`.
- *   // Optionally override `isSanitizer`.
- *   // Optionally override `isSanitizerIn`.
- *   // Optionally override `isSanitizerOut`.
- *   // Optionally override `isSanitizerGuard`.
- *   // Optionally override `isAdditionalTaintStep`.
- * }
- * ```
- *
- * Then, to query whether there is flow between some `source` and `sink`,
- * write
- *
- * ```ql
- * exists(MyAnalysisConfiguration cfg | cfg.hasFlow(source, sink))
- * ```
- *
- * Multiple configurations can coexist, but it is unsupported to depend on
- * another `TaintTracking::Configuration` or a `DataFlow::Configuration` in the
- * overridden predicates that define sources, sinks, or additional steps.
- * Instead, the dependency should go to a `TaintTracking2::Configuration` or a
- * `DataFlow2::Configuration`, `DataFlow3::Configuration`, etc.
- */
-abstract class Configuration extends DataFlow::Configuration {
-  bindingset[this]
-  Configuration() { any() }
-
-  /**
-   * Holds if `source` is a relevant taint source.
-   *
-   * The smaller this predicate is, the faster `hasFlow()` will converge.
-   */
-  // overridden to provide taint-tracking specific qldoc
-  override predicate isSource(DataFlow::Node source) { none() }
-
-  /**
-   * Holds if `source` is a relevant taint source with the given initial
-   * `state`.
-   *
-   * The smaller this predicate is, the faster `hasFlow()` will converge.
-   */
-  // overridden to provide taint-tracking specific qldoc
-  override predicate isSource(DataFlow::Node source, DataFlow::FlowState state) { none() }
-
-  /**
-   * Holds if `sink` is a relevant taint sink
-   *
-   * The smaller this predicate is, the faster `hasFlow()` will converge.
-   */
-  // overridden to provide taint-tracking specific qldoc
-  override predicate isSink(DataFlow::Node sink) { none() }
-
-  /**
-   * Holds if `sink` is a relevant taint sink accepting `state`.
-   *
-   * The smaller this predicate is, the faster `hasFlow()` will converge.
-   */
-  // overridden to provide taint-tracking specific qldoc
-  override predicate isSink(DataFlow::Node sink, DataFlow::FlowState state) { none() }
-
-  /** Holds if the node `node` is a taint sanitizer. */
-  predicate isSanitizer(DataFlow::Node node) { none() }
-
-  final override predicate isBarrier(DataFlow::Node node) {
-    this.isSanitizer(node) or
-    defaultTaintSanitizer(node)
-  }
-
-  /**
-   * Holds if the node `node` is a taint sanitizer when the flow state is
-   * `state`.
-   */
-  predicate isSanitizer(DataFlow::Node node, DataFlow::FlowState state) { none() }
-
-  final override predicate isBarrier(DataFlow::Node node, DataFlow::FlowState state) {
-    this.isSanitizer(node, state)
-  }
-
-  /** Holds if taint propagation into `node` is prohibited. */
-  predicate isSanitizerIn(DataFlow::Node node) { none() }
-
-  final override predicate isBarrierIn(DataFlow::Node node) { this.isSanitizerIn(node) }
-
-  /** Holds if taint propagation out of `node` is prohibited. */
-  predicate isSanitizerOut(DataFlow::Node node) { none() }
-
-  final override predicate isBarrierOut(DataFlow::Node node) { this.isSanitizerOut(node) }
-
-  /**
-   * DEPRECATED: Use `isSanitizer` and `BarrierGuard` module instead.
-   *
-   * Holds if taint propagation through nodes guarded by `guard` is prohibited.
-   */
-  deprecated predicate isSanitizerGuard(DataFlow::BarrierGuard guard) { none() }
-
-  deprecated final override predicate isBarrierGuard(DataFlow::BarrierGuard guard) {
-    this.isSanitizerGuard(guard)
-  }
-
-  /**
-   * DEPRECATED: Use `isSanitizer` and `BarrierGuard` module instead.
-   *
-   * Holds if taint propagation through nodes guarded by `guard` is prohibited
-   * when the flow state is `state`.
-   */
-  deprecated predicate isSanitizerGuard(DataFlow::BarrierGuard guard, DataFlow::FlowState state) {
-    none()
-  }
-
-  deprecated final override predicate isBarrierGuard(
-    DataFlow::BarrierGuard guard, DataFlow::FlowState state
-  ) {
-    this.isSanitizerGuard(guard, state)
-  }
-
-  /**
-   * Holds if taint may propagate from `node1` to `node2` in addition to the normal data-flow and taint steps.
-   */
-  predicate isAdditionalTaintStep(DataFlow::Node node1, DataFlow::Node node2) { none() }
-
-  final override predicate isAdditionalFlowStep(DataFlow::Node node1, DataFlow::Node node2) {
-    this.isAdditionalTaintStep(node1, node2) or
-    defaultAdditionalTaintStep(node1, node2)
-  }
-
-  /**
-   * Holds if taint may propagate from `node1` to `node2` in addition to the normal data-flow and taint steps.
-   * This step is only applicable in `state1` and updates the flow state to `state2`.
-   */
-  predicate isAdditionalTaintStep(
-    DataFlow::Node node1, DataFlow::FlowState state1, DataFlow::Node node2,
-    DataFlow::FlowState state2
-  ) {
-    none()
-  }
-
-  final override predicate isAdditionalFlowStep(
-    DataFlow::Node node1, DataFlow::FlowState state1, DataFlow::Node node2,
-    DataFlow::FlowState state2
-  ) {
-    this.isAdditionalTaintStep(node1, state1, node2, state2)
-  }
-
-  override predicate allowImplicitRead(DataFlow::Node node, DataFlow::ContentSet c) {
-    (
-      this.isSink(node) or
-      this.isSink(node, _) or
-      this.isAdditionalTaintStep(node, _) or
-      this.isAdditionalTaintStep(node, _, _, _)
-    ) and
-    defaultImplicitTaintRead(node, c)
-  }
-
-  /**
-   * Holds if taint may flow from `source` to `sink` for this configuration.
-   */
-  // overridden to provide taint-tracking specific qldoc
-  override predicate hasFlow(DataFlow::Node source, DataFlow::Node sink) {
-    super.hasFlow(source, sink)
-  }
-}

cpp/ql/lib/experimental/semmle/code/cpp/ir/dataflow/internal/tainttracking3/TaintTrackingParameter.qll

@@ -1,5 +0,0 @@
-import experimental.semmle.code.cpp.ir.dataflow.internal.TaintTrackingUtil as Public
-
-module Private {
-  import experimental.semmle.code.cpp.ir.dataflow.DataFlow3::DataFlow3 as DataFlow
-}

cpp/ql/lib/experimental/semmle/code/cpp/security/PrivateCleartextWrite.qll

@@ -3,7 +3,7 @@
  */
 
 import cpp
-import semmle.code.cpp.dataflow.TaintTracking
+import semmle.code.cpp.ir.dataflow.TaintTracking
 import semmle.code.cpp.security.PrivateData
 import semmle.code.cpp.security.FileWrite
 import semmle.code.cpp.security.BufferWrite
@@ -36,7 +36,7 @@ module PrivateCleartextWrite {
     }
   }
 
-  class WriteConfig extends TaintTracking::Configuration {
+  deprecated class WriteConfig extends TaintTracking::Configuration {
     WriteConfig() { this = "Write configuration" }
 
     override predicate isSource(DataFlow::Node source) { source instanceof Source }
@@ -46,6 +46,16 @@ module PrivateCleartextWrite {
     override predicate isSanitizer(DataFlow::Node node) { node instanceof Sanitizer }
   }
 
+  private module WriteConfig implements DataFlow::ConfigSig {
+    predicate isSource(DataFlow::Node source) { source instanceof Source }
+
+    predicate isSink(DataFlow::Node sink) { sink instanceof Sink }
+
+    predicate isBarrier(DataFlow::Node node) { node instanceof Sanitizer }
+  }
+
+  module WriteFlow = TaintTracking::Make<WriteConfig>;
+
   class PrivateDataSource extends Source {
     PrivateDataSource() { this.getExpr() instanceof PrivateDataExpr }
   }

cpp/ql/lib/experimental/semmle/code/cpp/semantic/analysis/RangeAnalysisStage.qll

@@ -936,6 +936,15 @@ module RangeStage<DeltaSig D, BoundSig<D> Bounds, LangSig<D> LangParam, UtilSig<
     bounded(cast.getOperand(), b, delta, upper, fromBackEdge, origdelta, reason)
   }
 
+  /**
+   * Computes a normal form of `x` where -0.0 has changed to +0.0. This can be
+   * needed on the lesser side of a floating-point comparison or on both sides of
+   * a floating point equality because QL does not follow IEEE in floating-point
+   * comparisons but instead defines -0.0 to be less than and distinct from 0.0.
+   */
+  bindingset[x]
+  private float normalizeFloatUp(float x) { result = x + 0.0 }
+
   /**
    * Holds if `b + delta` is a valid bound for `e`.
    * - `upper = true`  : `e <= b + delta`
@@ -1020,6 +1029,15 @@ module RangeStage<DeltaSig D, BoundSig<D> Bounds, LangSig<D> LangParam, UtilSig<
         or
         upper = false and delta = D::fromFloat(D::toFloat(d1).minimum(D::toFloat(d2)))
       )
+      or
+      exists(SemExpr mid, D::Delta d, float f |
+        e.(SemNegateExpr).getOperand() = mid and
+        b instanceof SemZeroBound and
+        bounded(mid, b, d, upper.booleanNot(), fromBackEdge, origdelta, reason) and
+        f = normalizeFloatUp(-D::toFloat(d)) and
+        delta = D::fromFloat(f) and
+        if semPositive(e) then f >= 0 else any()
+      )
     )
   }
 

cpp/ql/lib/experimental/semmle/code/cpp/semantic/analysis/SimpleRangeAnalysis.qll

@@ -0,0 +1,132 @@
+/**
+ * Wrapper for the semantic range analysis library that mimics the
+ * interface of the simple range analysis library.
+ */
+
+private import cpp
+private import semmle.code.cpp.ir.IR
+private import experimental.semmle.code.cpp.semantic.SemanticBound
+private import experimental.semmle.code.cpp.semantic.SemanticExprSpecific
+private import RangeAnalysis
+
+/**
+ * Gets the lower bound of the expression.
+ *
+ * Note: expressions in C/C++ are often implicitly or explicitly cast to a
+ * different result type. Such casts can cause the value of the expression
+ * to overflow or to be truncated. This predicate computes the lower bound
+ * of the expression without including the effect of the casts. To compute
+ * the lower bound of the expression after all the casts have been applied,
+ * call `lowerBound` like this:
+ *
+ *    `lowerBound(expr.getFullyConverted())`
+ */
+float lowerBound(Expr expr) {
+  exists(Instruction i, SemBound b | i.getAst() = expr and b instanceof SemZeroBound |
+    semBounded(getSemanticExpr(i), b, result, false, _)
+  )
+}
+
+/**
+ * Gets the upper bound of the expression.
+ *
+ * Note: expressions in C/C++ are often implicitly or explicitly cast to a
+ * different result type. Such casts can cause the value of the expression
+ * to overflow or to be truncated. This predicate computes the upper bound
+ * of the expression without including the effect of the casts. To compute
+ * the upper bound of the expression after all the casts have been applied,
+ * call `upperBound` like this:
+ *
+ *    `upperBound(expr.getFullyConverted())`
+ */
+float upperBound(Expr expr) {
+  exists(Instruction i, SemBound b | i.getAst() = expr and b instanceof SemZeroBound |
+    semBounded(getSemanticExpr(i), b, result, true, _)
+  )
+}
+
+/**
+ * Holds if the upper bound of `expr` may have been widened. This means the
+ * upper bound is in practice likely to be overly wide.
+ */
+predicate upperBoundMayBeWidened(Expr e) { none() }
+
+/**
+ * Holds if `expr` has a provably empty range. For example:
+ *
+ *   10 < expr and expr < 5
+ *
+ * The range of an expression can only be empty if it can never be
+ * executed. For example:
+ *
+ * ```cpp
+ * if (10 < x) {
+ *   if (x < 5) {
+ *     // Unreachable code
+ *     return x; // x has an empty range: 10 < x && x < 5
+ *   }
+ * }
+ * ```
+ */
+predicate exprWithEmptyRange(Expr expr) { lowerBound(expr) > upperBound(expr) }
+
+/** Holds if the definition might overflow negatively. */
+predicate defMightOverflowNegatively(RangeSsaDefinition def, StackVariable v) { none() }
+
+/** Holds if the definition might overflow positively. */
+predicate defMightOverflowPositively(RangeSsaDefinition def, StackVariable v) { none() }
+
+/**
+ * Holds if the definition might overflow (either positively or
+ * negatively).
+ */
+predicate defMightOverflow(RangeSsaDefinition def, StackVariable v) {
+  defMightOverflowNegatively(def, v) or
+  defMightOverflowPositively(def, v)
+}
+
+/**
+ * Holds if the expression might overflow negatively. This predicate
+ * does not consider the possibility that the expression might overflow
+ * due to a conversion.
+ */
+predicate exprMightOverflowNegatively(Expr expr) { none() }
+
+/**
+ * Holds if the expression might overflow negatively. Conversions
+ * are also taken into account. For example the expression
+ * `(int16)(x+y)` might overflow due to the `(int16)` cast, rather than
+ * due to the addition.
+ */
+predicate convertedExprMightOverflowNegatively(Expr expr) {
+  exprMightOverflowNegatively(expr) or
+  convertedExprMightOverflowNegatively(expr.getConversion())
+}
+
+/**
+ * Holds if the expression might overflow positively. This predicate
+ * does not consider the possibility that the expression might overflow
+ * due to a conversion.
+ */
+predicate exprMightOverflowPositively(Expr expr) { none() }
+
+/**
+ * Holds if the expression might overflow positively. Conversions
+ * are also taken into account. For example the expression
+ * `(int16)(x+y)` might overflow due to the `(int16)` cast, rather than
+ * due to the addition.
+ */
+predicate convertedExprMightOverflowPositively(Expr expr) {
+  exprMightOverflowPositively(expr) or
+  convertedExprMightOverflowPositively(expr.getConversion())
+}
+
+/**
+ * Holds if the expression might overflow (either positively or
+ * negatively). The possibility that the expression might overflow
+ * due to an implicit or explicit cast is also considered.
+ */
+predicate convertedExprMightOverflow(Expr expr) {
+  convertedExprMightOverflowNegatively(expr) or
+  convertedExprMightOverflowPositively(expr)
+}

cpp/ql/lib/qlpack.yml

@@ -1,5 +1,5 @@
 name: codeql/cpp-all
-version: 0.5.2
+version: 0.6.0
 groups: cpp
 dbscheme: semmlecode.cpp.dbscheme
 extractor: cpp

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

@@ -227,18 +227,6 @@ class Class extends UserType {
     result = this.accessOfBaseMember(member.getDeclaringType(), member.getASpecifier())
   }
 
-  /**
-   * DEPRECATED: name changed to `hasImplicitCopyConstructor` to reflect that
-   * `= default` members are no longer included.
-   */
-  deprecated predicate hasGeneratedCopyConstructor() { this.hasImplicitCopyConstructor() }
-
-  /**
-   * DEPRECATED: name changed to `hasImplicitCopyAssignmentOperator` to
-   * reflect that `= default` members are no longer included.
-   */
-  deprecated predicate hasGeneratedCopyAssignmentOperator() { this.hasImplicitCopyConstructor() }
-
   /**
    * Holds if this class, struct or union has an implicitly-declared copy
    * constructor that is not _deleted_. This predicate is more accurate than

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

@@ -68,7 +68,9 @@ class Declaration extends Locatable, @declaration {
    * Holds if this declaration has the fully-qualified name `qualifiedName`.
    * See `getQualifiedName`.
    */
-  predicate hasQualifiedName(string qualifiedName) { this.getQualifiedName() = qualifiedName }
+  deprecated predicate hasQualifiedName(string qualifiedName) {
+    this.getQualifiedName() = qualifiedName
+  }
 
   /**
    * Holds if this declaration has a fully-qualified name with a name-space
@@ -184,7 +186,7 @@ class Declaration extends Locatable, @declaration {
   predicate hasDefinition() { exists(this.getDefinition()) }
 
   /** DEPRECATED: Use `hasDefinition` instead. */
-  predicate isDefined() { this.hasDefinition() }
+  deprecated predicate isDefined() { this.hasDefinition() }
 
   /** Gets the preferred location of this declaration, if any. */
   override Location getLocation() { none() }
@@ -617,11 +619,10 @@ private class DirectAccessHolder extends Element {
   /**
    * Like `couldAccessMember` but only contains derivations in which either
    * (5.2), (5.3) or (5.4) must be invoked. In other words, the `this`
-   * parameter is not ignored. This restriction makes it feasible to fully
-   * enumerate this predicate even on large code bases. We check for 11.4 as
-   * part of (5.3), since this further limits the number of tuples produced by
-   * this predicate.
+   * parameter is not ignored. We check for 11.4 as part of (5.3), since
+   * this further limits the number of tuples produced by this predicate.
    */
+  pragma[inline]
   predicate thisCouldAccessMember(Class memberClass, AccessSpecifier memberAccess, Class derived) {
     // Only (5.4) is recursive, and chains of invocations of (5.4) can always
     // be collapsed to one invocation by the transitivity of 11.2/4.
@@ -663,7 +664,9 @@ private class DirectAccessHolder extends Element {
     //    bypasses `p`. Then that path must be public, or we are in case 2.
     exists(AccessSpecifier public | public.hasName("public") |
       exists(Class between, Class p |
-        between.accessOfBaseMember(memberClass, memberAccess).hasName("protected") and
+        between
+            .accessOfBaseMember(pragma[only_bind_into](memberClass), memberAccess)
+            .hasName("protected") and
         this.isFriendOfOrEqualTo(p) and
         (
           // This is case 1 from above. If `p` derives privately from `between`

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

@@ -41,7 +41,7 @@ class Function extends Declaration, ControlFlowNode, AccessHolder, @function {
    * `min<int>(int, int) -> int`, and the full signature of the uninstantiated
    * template on the first line would be `min<T>(T, T) -> T`.
    */
-  string getFullSignature() {
+  deprecated string getFullSignature() {
     exists(string name, string templateArgs, string args |
       result = name + templateArgs + args + " -> " + this.getType().toString() and
       name = this.getQualifiedName() and

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

@@ -159,7 +159,8 @@ class NameQualifyingElement extends Element, @namequalifyingelement {
  * A special name-qualifying element. For example: `__super`.
  */
 library class SpecialNameQualifyingElement extends NameQualifyingElement,
-  @specialnamequalifyingelement {
+  @specialnamequalifyingelement
+{
   /** Gets the name of this special qualifying element. */
   override string getName() { specialnamequalifyingelements(underlyingElement(this), result) }
 

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

@@ -108,20 +108,6 @@ class XmlFile extends XmlParent, File {
   /** Gets the name of this XML file. */
   override string getName() { result = File.super.getAbsolutePath() }
 
-  /**
-   * DEPRECATED: Use `getAbsolutePath()` instead.
-   *
-   * Gets the path of this XML file.
-   */
-  deprecated string getPath() { result = this.getAbsolutePath() }
-
-  /**
-   * DEPRECATED: Use `getParentContainer().getAbsolutePath()` instead.
-   *
-   * Gets the path of the folder that contains this XML file.
-   */
-  deprecated string getFolder() { result = this.getParentContainer().getAbsolutePath() }
-
   /** Gets the encoding of this XML file. */
   string getEncoding() { xmlEncoding(this, result) }
 

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

@@ -12,7 +12,9 @@ predicate freeFunction(Function f, int argNum) { argNum = f.(DeallocationFunctio
  *
  * DEPRECATED: Use `DeallocationExpr` instead (this also includes `delete` expressions).
  */
-predicate freeCall(FunctionCall fc, Expr arg) { arg = fc.(DeallocationExpr).getFreedExpr() }
+deprecated predicate freeCall(FunctionCall fc, Expr arg) {
+  arg = fc.(DeallocationExpr).getFreedExpr()
+}
 
 /**
  * Is e some kind of allocation or deallocation (`new`, `alloc`, `realloc`, `delete`, `free` etc)?

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

@@ -1,5 +1,5 @@
 import cpp
-import semmle.code.cpp.dataflow.DataFlow
+private import semmle.code.cpp.ir.dataflow.DataFlow
 
 /**
  * Holds if `v` is a member variable of `c` that looks like it might be variable sized
@@ -25,10 +25,12 @@ predicate memberMayBeVarSize(Class c, MemberVariable v) {
 }
 
 /**
- * Get the size in bytes of the buffer pointed to by an expression (if this can be determined).
+ * Holds if `bufferExpr` is an allocation-like expression.
+ *
+ * This includes both actual allocations, as well as various operations that return a pointer to
+ * stack-allocated objects.
  */
-language[monotonicAggregates]
-int getBufferSize(Expr bufferExpr, Element why) {
+private int isSource(Expr bufferExpr, Element why) {
   exists(Variable bufferVar | bufferVar = bufferExpr.(VariableAccess).getTarget() |
     // buffer is a fixed size array
     result = bufferVar.getUnspecifiedType().(ArrayType).getSize() and
@@ -46,42 +48,12 @@ int getBufferSize(Expr bufferExpr, Element why) {
     ) and
     result = why.(Expr).getType().(ArrayType).getSize() and
     not exists(bufferVar.getUnspecifiedType().(ArrayType).getSize())
-    or
-    exists(Class parentClass, VariableAccess parentPtr, int bufferSize |
-      // buffer is the parentPtr->bufferVar of a 'variable size struct'
-      memberMayBeVarSize(parentClass, bufferVar) and
-      why = bufferVar and
-      parentPtr = bufferExpr.(VariableAccess).getQualifier() and
-      parentPtr.getTarget().getUnspecifiedType().(PointerType).getBaseType() = parentClass and
-      (
-        if exists(bufferVar.getType().getSize())
-        then bufferSize = bufferVar.getType().getSize()
-        else bufferSize = 0
-      ) and
-      result = getBufferSize(parentPtr, _) + bufferSize - parentClass.getSize()
-    )
   )
   or
   // buffer is a fixed size dynamic allocation
   result = bufferExpr.(AllocationExpr).getSizeBytes() and
   why = bufferExpr
   or
-  exists(DataFlow::ExprNode bufferExprNode |
-    // dataflow (all sources must be the same size)
-    bufferExprNode = DataFlow::exprNode(bufferExpr) and
-    result =
-      unique(Expr def |
-        DataFlow::localFlowStep(DataFlow::exprNode(def), bufferExprNode)
-      |
-        getBufferSize(def, _)
-      ) and
-    // find reason
-    exists(Expr def | DataFlow::localFlowStep(DataFlow::exprNode(def), bufferExprNode) |
-      why = def or
-      exists(getBufferSize(def, why))
-    )
-  )
-  or
   exists(Type bufferType |
     // buffer is the address of a variable
     why = bufferExpr.(AddressOfExpr).getAddressable() and
@@ -100,3 +72,30 @@ int getBufferSize(Expr bufferExpr, Element why) {
     result = bufferType.getSize()
   )
 }
+
+/**
+ * Get the size in bytes of the buffer pointed to by an expression (if this can be determined).
+ */
+language[monotonicAggregates]
+int getBufferSize(Expr bufferExpr, Element why) {
+  result = isSource(bufferExpr, why)
+  or
+  exists(Class parentClass, VariableAccess parentPtr, int bufferSize, Variable bufferVar |
+    bufferVar = bufferExpr.(VariableAccess).getTarget() and
+    // buffer is the parentPtr->bufferVar of a 'variable size struct'
+    memberMayBeVarSize(parentClass, bufferVar) and
+    why = bufferVar and
+    parentPtr = bufferExpr.(VariableAccess).getQualifier() and
+    parentPtr.getTarget().getUnspecifiedType().(PointerType).getBaseType() = parentClass and
+    result = getBufferSize(parentPtr, _) + bufferSize - parentClass.getSize()
+  |
+    if exists(bufferVar.getType().getSize())
+    then bufferSize = bufferVar.getType().getSize()
+    else bufferSize = 0
+  )
+  or
+  // dataflow (all sources must be the same size)
+  result = unique(Expr def | DataFlow::localExprFlowStep(def, bufferExpr) | getBufferSize(def, _)) and
+  // find reason
+  exists(Expr def | DataFlow::localExprFlowStep(def, bufferExpr) | exists(getBufferSize(def, why)))
+}

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

@@ -1,7 +1,7 @@
 import cpp
 private import semmle.code.cpp.models.interfaces.ArrayFunction
 private import semmle.code.cpp.models.implementations.Strcat
-import semmle.code.cpp.dataflow.DataFlow
+private import semmle.code.cpp.ir.dataflow.DataFlow
 
 /**
  * Holds if the expression `e` assigns something including `va` to a

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

@@ -59,26 +59,6 @@ abstract class MutexType extends Type {
    * Gets a call that unlocks any mutex of this type.
    */
   FunctionCall getUnlockAccess() { this.unlockAccess(result, _) }
-
-  /**
-   * DEPRECATED: use mustlockAccess(fc, arg) instead.
-   */
-  deprecated Function getMustlockFunction() { result = this.getMustlockAccess().getTarget() }
-
-  /**
-   * DEPRECATED: use trylockAccess(fc, arg) instead.
-   */
-  deprecated Function getTrylockFunction() { result = this.getTrylockAccess().getTarget() }
-
-  /**
-   * DEPRECATED: use lockAccess(fc, arg) instead.
-   */
-  deprecated Function getLockFunction() { result = this.getLockAccess().getTarget() }
-
-  /**
-   * DEPRECATED: use unlockAccess(fc, arg) instead.
-   */
-  deprecated Function getUnlockFunction() { result = this.getUnlockAccess().getTarget() }
 }
 
 /**

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

@@ -75,13 +75,6 @@ class SubBasicBlock extends ControlFlowNodeBase {
     )
   }
 
-  /**
-   * DEPRECATED: use `getRankInBasicBlock` instead. Note that this predicate
-   * returns a 0-based position, while `getRankInBasicBlock` returns a 1-based
-   * position.
-   */
-  deprecated int getPosInBasicBlock(BasicBlock bb) { result = this.getRankInBasicBlock(bb) - 1 }
-
   pragma[noinline]
   private int getIndexInBasicBlock(BasicBlock bb) { this = bb.getNode(result) }
 

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

@@ -19,6 +19,11 @@
 
 import cpp
 
+/**
+ * Provides classes for performing local (intra-procedural) and
+ * global (inter-procedural) data flow analyses.
+ */
 module DataFlow {
-  import semmle.code.cpp.dataflow.internal.DataFlowImpl
+  import semmle.code.cpp.dataflow.internal.DataFlow
+  import semmle.code.cpp.dataflow.internal.DataFlowImpl1
 }

cpp/ql/lib/semmle/code/cpp/dataflow/DataFlow2.qll

@@ -11,6 +11,10 @@
 
 import cpp
 
+/**
+ * Provides classes for performing local (intra-procedural) and
+ * global (inter-procedural) data flow analyses.
+ */
 module DataFlow2 {
   import semmle.code.cpp.dataflow.internal.DataFlowImpl2
 }

cpp/ql/lib/semmle/code/cpp/dataflow/DataFlow3.qll

@@ -11,6 +11,10 @@
 
 import cpp
 
+/**
+ * Provides classes for performing local (intra-procedural) and
+ * global (inter-procedural) data flow analyses.
+ */
 module DataFlow3 {
   import semmle.code.cpp.dataflow.internal.DataFlowImpl3
 }

cpp/ql/lib/semmle/code/cpp/dataflow/DataFlow4.qll

@@ -11,6 +11,10 @@
 
 import cpp
 
+/**
+ * Provides classes for performing local (intra-procedural) and
+ * global (inter-procedural) data flow analyses.
+ */
 module DataFlow4 {
   import semmle.code.cpp.dataflow.internal.DataFlowImpl4
 }

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

@@ -95,6 +95,11 @@ predicate stackPointerFlowsToUse(Expr use, Type useType, Expr source, boolean is
 cached
 private PointerType getExprPtrType(Expr use) { result = use.getUnspecifiedType() }
 
+/**
+ * Holds if `use` has type `useType` and `source` is an access to a stack variable
+ * that flows to `use`. `isLocal` is `true` if `use` is accessed via a parameter, and
+ * `false` otherwise.
+ */
 predicate stackReferenceFlowsToUse(Expr use, Type useType, Expr source, boolean isLocal) {
   // Stack variables
   exists(StackVariable var |

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

@@ -18,6 +18,11 @@
 import semmle.code.cpp.dataflow.DataFlow
 import semmle.code.cpp.dataflow.DataFlow2
 
+/**
+ * Provides classes for performing local (intra-procedural) and
+ * global (inter-procedural) taint-tracking analyses.
+ */
 module TaintTracking {
+  import semmle.code.cpp.dataflow.internal.tainttracking1.TaintTracking
   import semmle.code.cpp.dataflow.internal.tainttracking1.TaintTrackingImpl
 }

cpp/ql/lib/semmle/code/cpp/dataflow/TaintTracking2.qll

@@ -10,6 +10,11 @@
  *
  * See `semmle.code.cpp.dataflow.TaintTracking` for the full documentation.
  */
+
+/**
+ * Provides classes for performing local (intra-procedural) and
+ * global (inter-procedural) taint-tracking analyses.
+ */
 module TaintTracking2 {
   import semmle.code.cpp.dataflow.internal.tainttracking2.TaintTrackingImpl
 }

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

@@ -0,0 +1,353 @@
+/**
+ * Provides an implementation of global (interprocedural) data flow. This file
+ * re-exports the local (intraprocedural) data flow analysis from
+ * `DataFlowImplSpecific::Public` and adds a global analysis, mainly exposed
+ * through the `Make` and `MakeWithState` modules.
+ */
+
+private import DataFlowImplCommon
+private import DataFlowImplSpecific::Private
+import DataFlowImplSpecific::Public
+import DataFlowImplCommonPublic
+private import DataFlowImpl
+
+/** An input configuration for data flow. */
+signature module ConfigSig {
+  /**
+   * Holds if `source` is a relevant data flow source.
+   */
+  predicate isSource(Node source);
+
+  /**
+   * Holds if `sink` is a relevant data flow sink.
+   */
+  predicate isSink(Node sink);
+
+  /**
+   * Holds if data flow through `node` is prohibited. This completely removes
+   * `node` from the data flow graph.
+   */
+  default predicate isBarrier(Node node) { none() }
+
+  /** Holds if data flow into `node` is prohibited. */
+  default predicate isBarrierIn(Node node) { none() }
+
+  /** Holds if data flow out of `node` is prohibited. */
+  default predicate isBarrierOut(Node node) { none() }
+
+  /**
+   * Holds if data may flow from `node1` to `node2` in addition to the normal data-flow steps.
+   */
+  default predicate isAdditionalFlowStep(Node node1, Node node2) { none() }
+
+  /**
+   * Holds if an arbitrary number of implicit read steps of content `c` may be
+   * taken at `node`.
+   */
+  default predicate allowImplicitRead(Node node, ContentSet c) { none() }
+
+  /**
+   * Gets the virtual dispatch branching limit when calculating field flow.
+   * This can be overridden to a smaller value to improve performance (a
+   * value of 0 disables field flow), or a larger value to get more results.
+   */
+  default int fieldFlowBranchLimit() { result = 2 }
+
+  /**
+   * Gets a data flow configuration feature to add restrictions to the set of
+   * valid flow paths.
+   *
+   * - `FeatureHasSourceCallContext`:
+   *    Assume that sources have some existing call context to disallow
+   *    conflicting return-flow directly following the source.
+   * - `FeatureHasSinkCallContext`:
+   *    Assume that sinks have some existing call context to disallow
+   *    conflicting argument-to-parameter flow directly preceding the sink.
+   * - `FeatureEqualSourceSinkCallContext`:
+   *    Implies both of the above and additionally ensures that the entire flow
+   *    path preserves the call context.
+   *
+   * These features are generally not relevant for typical end-to-end data flow
+   * queries, but should only be used for constructing paths that need to
+   * somehow be pluggable in another path context.
+   */
+  default FlowFeature getAFeature() { none() }
+
+  /** Holds if sources should be grouped in the result of `hasFlowPath`. */
+  default predicate sourceGrouping(Node source, string sourceGroup) { none() }
+
+  /** Holds if sinks should be grouped in the result of `hasFlowPath`. */
+  default predicate sinkGrouping(Node sink, string sinkGroup) { none() }
+
+  /**
+   * Holds if hidden nodes should be included in the data flow graph.
+   *
+   * This feature should only be used for debugging or when the data flow graph
+   * is not visualized (as it is in a `path-problem` query).
+   */
+  default predicate includeHiddenNodes() { none() }
+}
+
+/** An input configuration for data flow using flow state. */
+signature module StateConfigSig {
+  bindingset[this]
+  class FlowState;
+
+  /**
+   * Holds if `source` is a relevant data flow source with the given initial
+   * `state`.
+   */
+  predicate isSource(Node source, FlowState state);
+
+  /**
+   * Holds if `sink` is a relevant data flow sink accepting `state`.
+   */
+  predicate isSink(Node sink, FlowState state);
+
+  /**
+   * Holds if data flow through `node` is prohibited. This completely removes
+   * `node` from the data flow graph.
+   */
+  default predicate isBarrier(Node node) { none() }
+
+  /**
+   * Holds if data flow through `node` is prohibited when the flow state is
+   * `state`.
+   */
+  predicate isBarrier(Node node, FlowState state);
+
+  /** Holds if data flow into `node` is prohibited. */
+  default predicate isBarrierIn(Node node) { none() }
+
+  /** Holds if data flow out of `node` is prohibited. */
+  default predicate isBarrierOut(Node node) { none() }
+
+  /**
+   * Holds if data may flow from `node1` to `node2` in addition to the normal data-flow steps.
+   */
+  default predicate isAdditionalFlowStep(Node node1, Node node2) { none() }
+
+  /**
+   * Holds if data may flow from `node1` to `node2` in addition to the normal data-flow steps.
+   * This step is only applicable in `state1` and updates the flow state to `state2`.
+   */
+  predicate isAdditionalFlowStep(Node node1, FlowState state1, Node node2, FlowState state2);
+
+  /**
+   * Holds if an arbitrary number of implicit read steps of content `c` may be
+   * taken at `node`.
+   */
+  default predicate allowImplicitRead(Node node, ContentSet c) { none() }
+
+  /**
+   * Gets the virtual dispatch branching limit when calculating field flow.
+   * This can be overridden to a smaller value to improve performance (a
+   * value of 0 disables field flow), or a larger value to get more results.
+   */
+  default int fieldFlowBranchLimit() { result = 2 }
+
+  /**
+   * Gets a data flow configuration feature to add restrictions to the set of
+   * valid flow paths.
+   *
+   * - `FeatureHasSourceCallContext`:
+   *    Assume that sources have some existing call context to disallow
+   *    conflicting return-flow directly following the source.
+   * - `FeatureHasSinkCallContext`:
+   *    Assume that sinks have some existing call context to disallow
+   *    conflicting argument-to-parameter flow directly preceding the sink.
+   * - `FeatureEqualSourceSinkCallContext`:
+   *    Implies both of the above and additionally ensures that the entire flow
+   *    path preserves the call context.
+   *
+   * These features are generally not relevant for typical end-to-end data flow
+   * queries, but should only be used for constructing paths that need to
+   * somehow be pluggable in another path context.
+   */
+  default FlowFeature getAFeature() { none() }
+
+  /** Holds if sources should be grouped in the result of `hasFlowPath`. */
+  default predicate sourceGrouping(Node source, string sourceGroup) { none() }
+
+  /** Holds if sinks should be grouped in the result of `hasFlowPath`. */
+  default predicate sinkGrouping(Node sink, string sinkGroup) { none() }
+
+  /**
+   * Holds if hidden nodes should be included in the data flow graph.
+   *
+   * This feature should only be used for debugging or when the data flow graph
+   * is not visualized (as it is in a `path-problem` query).
+   */
+  default predicate includeHiddenNodes() { none() }
+}
+
+/**
+ * Gets the exploration limit for `hasPartialFlow` and `hasPartialFlowRev`
+ * measured in approximate number of interprocedural steps.
+ */
+signature int explorationLimitSig();
+
+/**
+ * The output of a data flow computation.
+ */
+signature module DataFlowSig {
+  /**
+   * A `Node` augmented with a call context (except for sinks) and an access path.
+   * Only those `PathNode`s that are reachable from a source, and which can reach a sink, are generated.
+   */
+  class PathNode;
+
+  /**
+   * Holds if data can flow from `source` to `sink`.
+   *
+   * The corresponding paths are generated from the end-points and the graph
+   * included in the module `PathGraph`.
+   */
+  predicate hasFlowPath(PathNode source, PathNode sink);
+
+  /**
+   * Holds if data can flow from `source` to `sink`.
+   */
+  predicate hasFlow(Node source, Node sink);
+
+  /**
+   * Holds if data can flow from some source to `sink`.
+   */
+  predicate hasFlowTo(Node sink);
+
+  /**
+   * Holds if data can flow from some source to `sink`.
+   */
+  predicate hasFlowToExpr(DataFlowExpr sink);
+}
+
+/**
+ * Constructs a standard data flow computation.
+ */
+module Make<ConfigSig Config> implements DataFlowSig {
+  private module C implements FullStateConfigSig {
+    import DefaultState<Config>
+    import Config
+  }
+
+  import Impl<C>
+}
+
+/**
+ * Constructs a data flow computation using flow state.
+ */
+module MakeWithState<StateConfigSig Config> implements DataFlowSig {
+  private module C implements FullStateConfigSig {
+    import Config
+  }
+
+  import Impl<C>
+}
+
+signature class PathNodeSig {
+  /** Gets a textual representation of this element. */
+  string toString();
+
+  /**
+   * Holds if this element is at the specified location.
+   * The location spans column `startcolumn` of line `startline` to
+   * column `endcolumn` of line `endline` in file `filepath`.
+   * For more information, see
+   * [Locations](https://codeql.github.com/docs/writing-codeql-queries/providing-locations-in-codeql-queries/).
+   */
+  predicate hasLocationInfo(
+    string filepath, int startline, int startcolumn, int endline, int endcolumn
+  );
+
+  /** Gets the underlying `Node`. */
+  Node getNode();
+}
+
+signature module PathGraphSig<PathNodeSig PathNode> {
+  /** Holds if `(a,b)` is an edge in the graph of data flow path explanations. */
+  predicate edges(PathNode a, PathNode b);
+
+  /** Holds if `n` is a node in the graph of data flow path explanations. */
+  predicate nodes(PathNode n, string key, string val);
+
+  /**
+   * Holds if `(arg, par, ret, out)` forms a subpath-tuple, that is, flow through
+   * a subpath between `par` and `ret` with the connecting edges `arg -> par` and
+   * `ret -> out` is summarized as the edge `arg -> out`.
+   */
+  predicate subpaths(PathNode arg, PathNode par, PathNode ret, PathNode out);
+}
+
+/**
+ * Constructs a `PathGraph` from two `PathGraph`s by disjoint union.
+ */
+module MergePathGraph<
+  PathNodeSig PathNode1, PathNodeSig PathNode2, PathGraphSig<PathNode1> Graph1,
+  PathGraphSig<PathNode2> Graph2>
+{
+  private newtype TPathNode =
+    TPathNode1(PathNode1 p) or
+    TPathNode2(PathNode2 p)
+
+  /** A node in a graph of path explanations that is formed by disjoint union of the two given graphs. */
+  class PathNode extends TPathNode {
+    /** Gets this as a projection on the first given `PathGraph`. */
+    PathNode1 asPathNode1() { this = TPathNode1(result) }
+
+    /** Gets this as a projection on the second given `PathGraph`. */
+    PathNode2 asPathNode2() { this = TPathNode2(result) }
+
+    /** Gets a textual representation of this element. */
+    string toString() {
+      result = this.asPathNode1().toString() or
+      result = this.asPathNode2().toString()
+    }
+
+    /**
+     * Holds if this element is at the specified location.
+     * The location spans column `startcolumn` of line `startline` to
+     * column `endcolumn` of line `endline` in file `filepath`.
+     * For more information, see
+     * [Locations](https://codeql.github.com/docs/writing-codeql-queries/providing-locations-in-codeql-queries/).
+     */
+    predicate hasLocationInfo(
+      string filepath, int startline, int startcolumn, int endline, int endcolumn
+    ) {
+      this.asPathNode1().hasLocationInfo(filepath, startline, startcolumn, endline, endcolumn) or
+      this.asPathNode2().hasLocationInfo(filepath, startline, startcolumn, endline, endcolumn)
+    }
+
+    /** Gets the underlying `Node`. */
+    Node getNode() {
+      result = this.asPathNode1().getNode() or
+      result = this.asPathNode2().getNode()
+    }
+  }
+
+  /**
+   * Provides the query predicates needed to include a graph in a path-problem query.
+   */
+  module PathGraph implements PathGraphSig<PathNode> {
+    /** Holds if `(a,b)` is an edge in the graph of data flow path explanations. */
+    query predicate edges(PathNode a, PathNode b) {
+      Graph1::edges(a.asPathNode1(), b.asPathNode1()) or
+      Graph2::edges(a.asPathNode2(), b.asPathNode2())
+    }
+
+    /** Holds if `n` is a node in the graph of data flow path explanations. */
+    query predicate nodes(PathNode n, string key, string val) {
+      Graph1::nodes(n.asPathNode1(), key, val) or
+      Graph2::nodes(n.asPathNode2(), key, val)
+    }
+
+    /**
+     * Holds if `(arg, par, ret, out)` forms a subpath-tuple, that is, flow through
+     * a subpath between `par` and `ret` with the connecting edges `arg -> par` and
+     * `ret -> out` is summarized as the edge `arg -> out`.
+     */
+    query predicate subpaths(PathNode arg, PathNode par, PathNode ret, PathNode out) {
+      Graph1::subpaths(arg.asPathNode1(), par.asPathNode1(), ret.asPathNode1(), out.asPathNode1()) or
+      Graph2::subpaths(arg.asPathNode2(), par.asPathNode2(), ret.asPathNode2(), out.asPathNode2())
+    }
+  }
+}

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

@@ -1,6 +1,6 @@
 private import cpp
-private import semmle.code.cpp.dataflow.internal.DataFlowPrivate
-private import semmle.code.cpp.dataflow.internal.DataFlowUtil
+private import DataFlowPrivate
+private import DataFlowUtil
 
 /**
  * Gets a function that might be called by `call`.

cpp/ql/lib/semmle/code/cpp/dataflow/internal/DataFlowImpl1.qll

@@ -0,0 +1,396 @@
+/**
+ * DEPRECATED: Use `Make` and `MakeWithState` instead.
+ *
+ * Provides a `Configuration` class backwards-compatible interface to the data
+ * flow library.
+ */
+
+private import DataFlowImplCommon
+private import DataFlowImplSpecific::Private
+import DataFlowImplSpecific::Public
+private import DataFlowImpl
+import DataFlowImplCommonPublic
+import FlowStateString
+
+/**
+ * A configuration of interprocedural data flow analysis. This defines
+ * sources, sinks, and any other configurable aspect of the analysis. Each
+ * use of the global data flow library must define its own unique extension
+ * of this abstract class. To create a configuration, extend this class with
+ * a subclass whose characteristic predicate is a unique singleton string.
+ * For example, write
+ *
+ * ```ql
+ * class MyAnalysisConfiguration extends DataFlow::Configuration {
+ *   MyAnalysisConfiguration() { this = "MyAnalysisConfiguration" }
+ *   // Override `isSource` and `isSink`.
+ *   // Optionally override `isBarrier`.
+ *   // Optionally override `isAdditionalFlowStep`.
+ * }
+ * ```
+ * Conceptually, this defines a graph where the nodes are `DataFlow::Node`s and
+ * the edges are those data-flow steps that preserve the value of the node
+ * along with any additional edges defined by `isAdditionalFlowStep`.
+ * Specifying nodes in `isBarrier` will remove those nodes from the graph, and
+ * specifying nodes in `isBarrierIn` and/or `isBarrierOut` will remove in-going
+ * and/or out-going edges from those nodes, respectively.
+ *
+ * Then, to query whether there is flow between some `source` and `sink`,
+ * write
+ *
+ * ```ql
+ * exists(MyAnalysisConfiguration cfg | cfg.hasFlow(source, sink))
+ * ```
+ *
+ * Multiple configurations can coexist, but two classes extending
+ * `DataFlow::Configuration` should never depend on each other. One of them
+ * should instead depend on a `DataFlow2::Configuration`, a
+ * `DataFlow3::Configuration`, or a `DataFlow4::Configuration`.
+ */
+abstract class Configuration extends string {
+  bindingset[this]
+  Configuration() { any() }
+
+  /**
+   * Holds if `source` is a relevant data flow source.
+   */
+  predicate isSource(Node source) { none() }
+
+  /**
+   * Holds if `source` is a relevant data flow source with the given initial
+   * `state`.
+   */
+  predicate isSource(Node source, FlowState state) { none() }
+
+  /**
+   * Holds if `sink` is a relevant data flow sink.
+   */
+  predicate isSink(Node sink) { none() }
+
+  /**
+   * Holds if `sink` is a relevant data flow sink accepting `state`.
+   */
+  predicate isSink(Node sink, FlowState state) { none() }
+
+  /**
+   * Holds if data flow through `node` is prohibited. This completely removes
+   * `node` from the data flow graph.
+   */
+  predicate isBarrier(Node node) { none() }
+
+  /**
+   * Holds if data flow through `node` is prohibited when the flow state is
+   * `state`.
+   */
+  predicate isBarrier(Node node, FlowState state) { none() }
+
+  /** Holds if data flow into `node` is prohibited. */
+  predicate isBarrierIn(Node node) { none() }
+
+  /** Holds if data flow out of `node` is prohibited. */
+  predicate isBarrierOut(Node node) { none() }
+
+  /**
+   * DEPRECATED: Use `isBarrier` and `BarrierGuard` module instead.
+   *
+   * Holds if data flow through nodes guarded by `guard` is prohibited.
+   */
+  deprecated predicate isBarrierGuard(BarrierGuard guard) { none() }
+
+  /**
+   * DEPRECATED: Use `isBarrier` and `BarrierGuard` module instead.
+   *
+   * Holds if data flow through nodes guarded by `guard` is prohibited when
+   * the flow state is `state`
+   */
+  deprecated predicate isBarrierGuard(BarrierGuard guard, FlowState state) { none() }
+
+  /**
+   * Holds if data may flow from `node1` to `node2` in addition to the normal data-flow steps.
+   */
+  predicate isAdditionalFlowStep(Node node1, Node node2) { none() }
+
+  /**
+   * Holds if data may flow from `node1` to `node2` in addition to the normal data-flow steps.
+   * This step is only applicable in `state1` and updates the flow state to `state2`.
+   */
+  predicate isAdditionalFlowStep(Node node1, FlowState state1, Node node2, FlowState state2) {
+    none()
+  }
+
+  /**
+   * Holds if an arbitrary number of implicit read steps of content `c` may be
+   * taken at `node`.
+   */
+  predicate allowImplicitRead(Node node, ContentSet c) { none() }
+
+  /**
+   * Gets the virtual dispatch branching limit when calculating field flow.
+   * This can be overridden to a smaller value to improve performance (a
+   * value of 0 disables field flow), or a larger value to get more results.
+   */
+  int fieldFlowBranchLimit() { result = 2 }
+
+  /**
+   * Gets a data flow configuration feature to add restrictions to the set of
+   * valid flow paths.
+   *
+   * - `FeatureHasSourceCallContext`:
+   *    Assume that sources have some existing call context to disallow
+   *    conflicting return-flow directly following the source.
+   * - `FeatureHasSinkCallContext`:
+   *    Assume that sinks have some existing call context to disallow
+   *    conflicting argument-to-parameter flow directly preceding the sink.
+   * - `FeatureEqualSourceSinkCallContext`:
+   *    Implies both of the above and additionally ensures that the entire flow
+   *    path preserves the call context.
+   *
+   * These features are generally not relevant for typical end-to-end data flow
+   * queries, but should only be used for constructing paths that need to
+   * somehow be pluggable in another path context.
+   */
+  FlowFeature getAFeature() { none() }
+
+  /** Holds if sources should be grouped in the result of `hasFlowPath`. */
+  predicate sourceGrouping(Node source, string sourceGroup) { none() }
+
+  /** Holds if sinks should be grouped in the result of `hasFlowPath`. */
+  predicate sinkGrouping(Node sink, string sinkGroup) { none() }
+
+  /**
+   * Holds if data may flow from `source` to `sink` for this configuration.
+   */
+  predicate hasFlow(Node source, Node sink) { hasFlow(source, sink, this) }
+
+  /**
+   * Holds if data may flow from `source` to `sink` for this configuration.
+   *
+   * The corresponding paths are generated from the end-points and the graph
+   * included in the module `PathGraph`.
+   */
+  predicate hasFlowPath(PathNode source, PathNode sink) { hasFlowPath(source, sink, this) }
+
+  /**
+   * Holds if data may flow from some source to `sink` for this configuration.
+   */
+  predicate hasFlowTo(Node sink) { hasFlowTo(sink, this) }
+
+  /**
+   * Holds if data may flow from some source to `sink` for this configuration.
+   */
+  predicate hasFlowToExpr(DataFlowExpr sink) { this.hasFlowTo(exprNode(sink)) }
+
+  /**
+   * DEPRECATED: Use `FlowExploration<explorationLimit>` instead.
+   *
+   * Gets the exploration limit for `hasPartialFlow` and `hasPartialFlowRev`
+   * measured in approximate number of interprocedural steps.
+   */
+  deprecated int explorationLimit() { none() }
+
+  /**
+   * Holds if hidden nodes should be included in the data flow graph.
+   *
+   * This feature should only be used for debugging or when the data flow graph
+   * is not visualized (for example in a `path-problem` query).
+   */
+  predicate includeHiddenNodes() { none() }
+}
+
+/**
+ * This class exists to prevent mutual recursion between the user-overridden
+ * member predicates of `Configuration` and the rest of the data-flow library.
+ * Good performance cannot be guaranteed in the presence of such recursion, so
+ * it should be replaced by using more than one copy of the data flow library.
+ */
+abstract private class ConfigurationRecursionPrevention extends Configuration {
+  bindingset[this]
+  ConfigurationRecursionPrevention() { any() }
+
+  override predicate hasFlow(Node source, Node sink) {
+    strictcount(Node n | this.isSource(n)) < 0
+    or
+    strictcount(Node n | this.isSource(n, _)) < 0
+    or
+    strictcount(Node n | this.isSink(n)) < 0
+    or
+    strictcount(Node n | this.isSink(n, _)) < 0
+    or
+    strictcount(Node n1, Node n2 | this.isAdditionalFlowStep(n1, n2)) < 0
+    or
+    strictcount(Node n1, Node n2 | this.isAdditionalFlowStep(n1, _, n2, _)) < 0
+    or
+    super.hasFlow(source, sink)
+  }
+}
+
+/** A bridge class to access the deprecated `isBarrierGuard`. */
+private class BarrierGuardGuardedNodeBridge extends Unit {
+  abstract predicate guardedNode(Node n, Configuration config);
+
+  abstract predicate guardedNode(Node n, FlowState state, Configuration config);
+}
+
+private class BarrierGuardGuardedNode extends BarrierGuardGuardedNodeBridge {
+  deprecated override predicate guardedNode(Node n, Configuration config) {
+    exists(BarrierGuard g |
+      config.isBarrierGuard(g) and
+      n = g.getAGuardedNode()
+    )
+  }
+
+  deprecated override predicate guardedNode(Node n, FlowState state, Configuration config) {
+    exists(BarrierGuard g |
+      config.isBarrierGuard(g, state) and
+      n = g.getAGuardedNode()
+    )
+  }
+}
+
+private FlowState relevantState(Configuration config) {
+  config.isSource(_, result) or
+  config.isSink(_, result) or
+  config.isBarrier(_, result) or
+  config.isAdditionalFlowStep(_, result, _, _) or
+  config.isAdditionalFlowStep(_, _, _, result)
+}
+
+private newtype TConfigState =
+  TMkConfigState(Configuration config, FlowState state) {
+    state = relevantState(config) or state instanceof FlowStateEmpty
+  }
+
+private Configuration getConfig(TConfigState state) { state = TMkConfigState(result, _) }
+
+private FlowState getState(TConfigState state) { state = TMkConfigState(_, result) }
+
+private predicate singleConfiguration() { 1 = strictcount(Configuration c) }
+
+private module Config implements FullStateConfigSig {
+  class FlowState = TConfigState;
+
+  predicate isSource(Node source, FlowState state) {
+    getConfig(state).isSource(source, getState(state))
+    or
+    getConfig(state).isSource(source) and getState(state) instanceof FlowStateEmpty
+  }
+
+  predicate isSink(Node sink, FlowState state) {
+    getConfig(state).isSink(sink, getState(state))
+    or
+    getConfig(state).isSink(sink) and getState(state) instanceof FlowStateEmpty
+  }
+
+  predicate isBarrier(Node node) { none() }
+
+  predicate isBarrier(Node node, FlowState state) {
+    getConfig(state).isBarrier(node, getState(state)) or
+    getConfig(state).isBarrier(node) or
+    any(BarrierGuardGuardedNodeBridge b).guardedNode(node, getState(state), getConfig(state)) or
+    any(BarrierGuardGuardedNodeBridge b).guardedNode(node, getConfig(state))
+  }
+
+  predicate isBarrierIn(Node node) { any(Configuration config).isBarrierIn(node) }
+
+  predicate isBarrierOut(Node node) { any(Configuration config).isBarrierOut(node) }
+
+  predicate isAdditionalFlowStep(Node node1, Node node2) {
+    singleConfiguration() and
+    any(Configuration config).isAdditionalFlowStep(node1, node2)
+  }
+
+  predicate isAdditionalFlowStep(Node node1, FlowState state1, Node node2, FlowState state2) {
+    getConfig(state1).isAdditionalFlowStep(node1, getState(state1), node2, getState(state2)) and
+    getConfig(state2) = getConfig(state1)
+    or
+    not singleConfiguration() and
+    getConfig(state1).isAdditionalFlowStep(node1, node2) and
+    state2 = state1
+  }
+
+  predicate allowImplicitRead(Node node, ContentSet c) {
+    any(Configuration config).allowImplicitRead(node, c)
+  }
+
+  int fieldFlowBranchLimit() { result = min(any(Configuration config).fieldFlowBranchLimit()) }
+
+  FlowFeature getAFeature() { result = any(Configuration config).getAFeature() }
+
+  predicate sourceGrouping(Node source, string sourceGroup) {
+    any(Configuration config).sourceGrouping(source, sourceGroup)
+  }
+
+  predicate sinkGrouping(Node sink, string sinkGroup) {
+    any(Configuration config).sinkGrouping(sink, sinkGroup)
+  }
+
+  predicate includeHiddenNodes() { any(Configuration config).includeHiddenNodes() }
+}
+
+private import Impl<Config> as I
+import I
+
+/**
+ * A `Node` augmented with a call context (except for sinks), an access path, and a configuration.
+ * Only those `PathNode`s that are reachable from a source, and which can reach a sink, are generated.
+ */
+class PathNode instanceof I::PathNode {
+  /** Gets a textual representation of this element. */
+  final string toString() { result = super.toString() }
+
+  /**
+   * Gets a textual representation of this element, including a textual
+   * representation of the call context.
+   */
+  final string toStringWithContext() { result = super.toStringWithContext() }
+
+  /**
+   * Holds if this element is at the specified location.
+   * The location spans column `startcolumn` of line `startline` to
+   * column `endcolumn` of line `endline` in file `filepath`.
+   * For more information, see
+   * [Locations](https://codeql.github.com/docs/writing-codeql-queries/providing-locations-in-codeql-queries/).
+   */
+  final predicate hasLocationInfo(
+    string filepath, int startline, int startcolumn, int endline, int endcolumn
+  ) {
+    super.hasLocationInfo(filepath, startline, startcolumn, endline, endcolumn)
+  }
+
+  /** Gets the underlying `Node`. */
+  final Node getNode() { result = super.getNode() }
+
+  /** Gets the `FlowState` of this node. */
+  final FlowState getState() { result = getState(super.getState()) }
+
+  /** Gets the associated configuration. */
+  final Configuration getConfiguration() { result = getConfig(super.getState()) }
+
+  /** Gets a successor of this node, if any. */
+  final PathNode getASuccessor() { result = super.getASuccessor() }
+
+  /** Holds if this node is a source. */
+  final predicate isSource() { super.isSource() }
+
+  /** Holds if this node is a grouping of source nodes. */
+  final predicate isSourceGroup(string group) { super.isSourceGroup(group) }
+
+  /** Holds if this node is a grouping of sink nodes. */
+  final predicate isSinkGroup(string group) { super.isSinkGroup(group) }
+}
+
+private predicate hasFlow(Node source, Node sink, Configuration config) {
+  exists(PathNode source0, PathNode sink0 |
+    hasFlowPath(source0, sink0, config) and
+    source0.getNode() = source and
+    sink0.getNode() = sink
+  )
+}
+
+private predicate hasFlowPath(PathNode source, PathNode sink, Configuration config) {
+  hasFlowPath(source, sink) and source.getConfiguration() = config
+}
+
+private predicate hasFlowTo(Node sink, Configuration config) { hasFlow(_, sink, config) }
+
+predicate flowsTo = hasFlow/3;

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

@@ -3,15 +3,18 @@ private import DataFlowImplSpecific::Public
 import Cached
 
 module DataFlowImplCommonPublic {
-  /** A state value to track during data flow. */
-  class FlowState = string;
+  /** Provides `FlowState = string`. */
+  module FlowStateString {
+    /** A state value to track during data flow. */
+    class FlowState = string;
 
-  /**
-   * The default state, which is used when the state is unspecified for a source
-   * or a sink.
-   */
-  class FlowStateEmpty extends FlowState {
-    FlowStateEmpty() { this = "" }
+    /**
+     * The default state, which is used when the state is unspecified for a source
+     * or a sink.
+     */
+    class FlowStateEmpty extends FlowState {
+      FlowStateEmpty() { this = "" }
+    }
   }
 
   private newtype TFlowFeature =
@@ -179,6 +182,7 @@ private module LambdaFlow {
     boolean toJump, DataFlowCallOption lastCall
   ) {
     revLambdaFlow0(lambdaCall, kind, node, t, toReturn, toJump, lastCall) and
+    not expectsContent(node, _) and
     if castNode(node) or node instanceof ArgNode or node instanceof ReturnNode
     then compatibleTypes(t, getNodeDataFlowType(node))
     else any()
@@ -707,8 +711,8 @@ private module Cached {
      * Gets a viable dispatch target of `call` in the context `ctx`. This is
      * restricted to those `call`s for which a context might make a difference.
      */
-    pragma[nomagic]
-    private DataFlowCallable viableImplInCallContextExt(DataFlowCall call, DataFlowCall ctx) {
+    cached
+    DataFlowCallable viableImplInCallContextExt(DataFlowCall call, DataFlowCall ctx) {
       result = viableImplInCallContext(call, ctx) and
       result = viableCallable(call)
       or
@@ -1391,6 +1395,9 @@ class TypedContentApprox extends MkTypedContentApprox {
   /** Gets a typed content approximated by this value. */
   TypedContent getATypedContent() { result = getATypedContent(this) }
 
+  /** Gets the content. */
+  ContentApprox getContent() { result = c }
+
   /** Gets the container type. */
   DataFlowType getContainerType() { result = t }
 
@@ -1408,6 +1415,8 @@ abstract class ApproxAccessPathFront extends TApproxAccessPathFront {
 
   abstract boolean toBoolNonEmpty();
 
+  TypedContentApprox getHead() { this = TApproxFrontHead(result) }
+
   pragma[nomagic]
   TypedContent getAHead() {
     exists(TypedContentApprox cont |

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

@@ -18,6 +18,9 @@ module Consistency {
     /** Holds if `n` should be excluded from the consistency test `uniqueEnclosingCallable`. */
     predicate uniqueEnclosingCallableExclude(Node n) { none() }
 
+    /** Holds if `call` should be excluded from the consistency test `uniqueCallEnclosingCallable`. */
+    predicate uniqueCallEnclosingCallableExclude(DataFlowCall call) { none() }
+
     /** Holds if `n` should be excluded from the consistency test `uniqueNodeLocation`. */
     predicate uniqueNodeLocationExclude(Node n) { none() }
 
@@ -86,6 +89,15 @@ module Consistency {
     )
   }
 
+  query predicate uniqueCallEnclosingCallable(DataFlowCall call, string msg) {
+    exists(int c |
+      c = count(call.getEnclosingCallable()) and
+      c != 1 and
+      not any(ConsistencyConfiguration conf).uniqueCallEnclosingCallableExclude(call) and
+      msg = "Call should have one enclosing callable but has " + c + "."
+    )
+  }
+
   query predicate uniqueType(Node n, string msg) {
     exists(int c |
       n instanceof RelevantNode and

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

@@ -318,3 +318,12 @@ private class MyConsistencyConfiguration extends Consistency::ConsistencyConfigu
     // consistency alerts enough that most of them are interesting.
   }
 }
+
+/**
+ * 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
+ * when calculating the (virtual) dispatch cost.
+ *
+ * Argument `arg` is part of a path from a source to a sink, and `p` is the target parameter.
+ */
+int getAdditionalFlowIntoCallNodeTerm(ArgumentNode arg, ParameterNode p) { none() }

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

@@ -3,10 +3,10 @@
  */
 
 private import cpp
-private import semmle.code.cpp.dataflow.internal.FlowVar
+private import FlowVar
 private import semmle.code.cpp.models.interfaces.DataFlow
 private import semmle.code.cpp.controlflow.Guards
-private import semmle.code.cpp.dataflow.internal.AddressFlow
+private import AddressFlow
 
 cached
 private newtype TNode =

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

@@ -4,8 +4,8 @@
 
 import cpp
 private import semmle.code.cpp.controlflow.SSA
-private import semmle.code.cpp.dataflow.internal.SubBasicBlocks
-private import semmle.code.cpp.dataflow.internal.AddressFlow
+private import SubBasicBlocks
+private import AddressFlow
 private import semmle.code.cpp.models.implementations.Iterator
 private import semmle.code.cpp.models.interfaces.PointerWrapper
 

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

@@ -75,13 +75,6 @@ class SubBasicBlock extends ControlFlowNodeBase {
     )
   }
 
-  /**
-   * DEPRECATED: use `getRankInBasicBlock` instead. Note that this predicate
-   * returns a 0-based position, while `getRankInBasicBlock` returns a 1-based
-   * position.
-   */
-  deprecated int getPosInBasicBlock(BasicBlock bb) { result = this.getRankInBasicBlock(bb) - 1 }
-
   pragma[noinline]
   private int getIndexInBasicBlock(BasicBlock bb) { this = bb.getNode(result) }
 

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

@@ -14,7 +14,7 @@ private import semmle.code.cpp.models.interfaces.Iterator
 private import semmle.code.cpp.models.interfaces.PointerWrapper
 
 private module DataFlow {
-  import semmle.code.cpp.dataflow.internal.DataFlowUtil
+  import DataFlowUtil
 }
 
 /**

cpp/ql/lib/semmle/code/cpp/dataflow/internal/tainttracking1/TaintTracking.qll

@@ -0,0 +1,64 @@
+/**
+ * Provides classes for performing local (intra-procedural) and
+ * global (inter-procedural) taint-tracking analyses.
+ */
+
+import TaintTrackingParameter::Public
+private import TaintTrackingParameter::Private
+
+private module AddTaintDefaults<DataFlowInternal::FullStateConfigSig Config> implements
+  DataFlowInternal::FullStateConfigSig
+{
+  import Config
+
+  predicate isBarrier(DataFlow::Node node) {
+    Config::isBarrier(node) or defaultTaintSanitizer(node)
+  }
+
+  predicate isAdditionalFlowStep(DataFlow::Node node1, DataFlow::Node node2) {
+    Config::isAdditionalFlowStep(node1, node2) or
+    defaultAdditionalTaintStep(node1, node2)
+  }
+
+  predicate allowImplicitRead(DataFlow::Node node, DataFlow::ContentSet c) {
+    Config::allowImplicitRead(node, c)
+    or
+    (
+      Config::isSink(node, _) or
+      Config::isAdditionalFlowStep(node, _) or
+      Config::isAdditionalFlowStep(node, _, _, _)
+    ) and
+    defaultImplicitTaintRead(node, c)
+  }
+}
+
+/**
+ * Constructs a standard taint tracking computation.
+ */
+module Make<DataFlow::ConfigSig Config> implements DataFlow::DataFlowSig {
+  private module Config0 implements DataFlowInternal::FullStateConfigSig {
+    import DataFlowInternal::DefaultState<Config>
+    import Config
+  }
+
+  private module C implements DataFlowInternal::FullStateConfigSig {
+    import AddTaintDefaults<Config0>
+  }
+
+  import DataFlowInternal::Impl<C>
+}
+
+/**
+ * Constructs a taint tracking computation using flow state.
+ */
+module MakeWithState<DataFlow::StateConfigSig Config> implements DataFlow::DataFlowSig {
+  private module Config0 implements DataFlowInternal::FullStateConfigSig {
+    import Config
+  }
+
+  private module C implements DataFlowInternal::FullStateConfigSig {
+    import AddTaintDefaults<Config0>
+  }
+
+  import DataFlowInternal::Impl<C>
+}

cpp/ql/lib/semmle/code/cpp/dataflow/internal/tainttracking1/TaintTrackingParameter.qll

@@ -2,4 +2,5 @@ import semmle.code.cpp.dataflow.internal.TaintTrackingUtil as Public
 
 module Private {
   import semmle.code.cpp.dataflow.DataFlow::DataFlow as DataFlow
+  import semmle.code.cpp.dataflow.internal.DataFlowImpl as DataFlowInternal
 }

cpp/ql/lib/semmle/code/cpp/dataflow/new/DataFlow.qll

@@ -8,8 +8,8 @@
  * results than the AST-based library in most scenarios.
  *
  * Unless configured otherwise, _flow_ means that the exact value of
- * the source may reach the sink. We do not track flow across pointer
- * dereferences or array indexing.
+ * the source may reach the sink. To track flow where the exact value
+ * may not be preserved, import `semmle.code.cpp.dataflow.new.TaintTracking`.
  *
  * To use global (interprocedural) data flow, extend the class
  * `DataFlow::Configuration` as documented on that class. To use local
@@ -21,6 +21,11 @@
 
 import cpp
 
+/**
+ * Provides classes for performing local (intra-procedural) and
+ * global (inter-procedural) data flow analyses.
+ */
 module DataFlow {
-  import experimental.semmle.code.cpp.ir.dataflow.internal.DataFlowImpl
+  import semmle.code.cpp.ir.dataflow.internal.DataFlow
+  import semmle.code.cpp.ir.dataflow.internal.DataFlowImpl1
 }

cpp/ql/lib/semmle/code/cpp/dataflow/new/DataFlow2.qll

@@ -6,11 +6,15 @@
  * `DataFlow2::Configuration`, a `DataFlow3::Configuration`, or a
  * `DataFlow4::Configuration`.
  *
- * See `semmle.code.cpp.ir.dataflow.DataFlow` for the full documentation.
+ * See `semmle.code.cpp.dataflow.new.DataFlow` for the full documentation.
  */
 
 import cpp
 
+/**
+ * Provides classes for performing local (intra-procedural) and
+ * global (inter-procedural) data flow analyses.
+ */
 module DataFlow2 {
-  import experimental.semmle.code.cpp.ir.dataflow.internal.DataFlowImpl2
+  import semmle.code.cpp.ir.dataflow.internal.DataFlowImpl2
 }

cpp/ql/lib/semmle/code/cpp/dataflow/new/DataFlow3.qll

@@ -6,11 +6,15 @@
  * `DataFlow2::Configuration`, a `DataFlow3::Configuration`, or a
  * `DataFlow4::Configuration`.
  *
- * See `semmle.code.cpp.ir.dataflow.DataFlow` for the full documentation.
+ * See `semmle.code.cpp.dataflow.new.DataFlow` for the full documentation.
  */
 
 import cpp
 
+/**
+ * Provides classes for performing local (intra-procedural) and
+ * global (inter-procedural) data flow analyses.
+ */
 module DataFlow3 {
-  import experimental.semmle.code.cpp.ir.dataflow.internal.DataFlowImpl3
+  import semmle.code.cpp.ir.dataflow.internal.DataFlowImpl3
 }

cpp/ql/lib/semmle/code/cpp/dataflow/new/DataFlow4.qll

@@ -6,11 +6,15 @@
  * `DataFlow2::Configuration`, a `DataFlow3::Configuration`, or a
  * `DataFlow4::Configuration`.
  *
- * See `semmle.code.cpp.ir.dataflow.DataFlow` for the full documentation.
+ * See `semmle.code.cpp.dataflow.new.DataFlow` for the full documentation.
  */
 
 import cpp
 
+/**
+ * Provides classes for performing local (intra-procedural) and
+ * global (inter-procedural) data flow analyses.
+ */
 module DataFlow4 {
-  import experimental.semmle.code.cpp.ir.dataflow.internal.DataFlowImpl4
+  import semmle.code.cpp.ir.dataflow.internal.DataFlowImpl4
 }

cpp/ql/lib/semmle/code/cpp/dataflow/new/TaintTracking.qll

@@ -15,9 +15,14 @@
  * `TaintTracking::localTaintStep` with arguments of type `DataFlow::Node`.
  */
 
-import semmle.code.cpp.ir.dataflow.DataFlow
-import semmle.code.cpp.ir.dataflow.DataFlow2
+import semmle.code.cpp.dataflow.new.DataFlow
+import semmle.code.cpp.dataflow.new.DataFlow2
 
+/**
+ * Provides classes for performing local (intra-procedural) and
+ * global (inter-procedural) taint-tracking analyses.
+ */
 module TaintTracking {
-  import experimental.semmle.code.cpp.ir.dataflow.internal.tainttracking1.TaintTrackingImpl
+  import semmle.code.cpp.ir.dataflow.internal.tainttracking1.TaintTracking
+  import semmle.code.cpp.ir.dataflow.internal.tainttracking1.TaintTrackingImpl
 }

cpp/ql/lib/semmle/code/cpp/dataflow/new/TaintTracking2.qll

@@ -8,8 +8,13 @@
  * `TaintTracking::Configuration` class extends `DataFlow::Configuration`, and
  * `TaintTracking2::Configuration` extends `DataFlow2::Configuration`.
  *
- * See `semmle.code.cpp.ir.dataflow.TaintTracking` for the full documentation.
+ * See `semmle.code.cpp.dataflow.new.TaintTracking` for the full documentation.
+ */
+
+/**
+ * Provides classes for performing local (intra-procedural) and
+ * global (inter-procedural) taint-tracking analyses.
  */
 module TaintTracking2 {
-  import experimental.semmle.code.cpp.ir.dataflow.internal.tainttracking2.TaintTrackingImpl
+  import semmle.code.cpp.ir.dataflow.internal.tainttracking2.TaintTrackingImpl
 }

cpp/ql/lib/semmle/code/cpp/dataflow/new/TaintTracking3.qll

@@ -8,8 +8,13 @@
  * `TaintTracking::Configuration` class extends `DataFlow::Configuration`, and
  * `TaintTracking2::Configuration` extends `DataFlow2::Configuration`.
  *
- * See `semmle.code.cpp.ir.dataflow.TaintTracking` for the full documentation.
+ * See `semmle.code.cpp.dataflow.new.TaintTracking` for the full documentation.
+ */
+
+/**
+ * Provides classes for performing local (intra-procedural) and
+ * global (inter-procedural) taint-tracking analyses.
  */
 module TaintTracking3 {
-  import experimental.semmle.code.cpp.ir.dataflow.internal.tainttracking3.TaintTrackingImpl
+  import semmle.code.cpp.ir.dataflow.internal.tainttracking3.TaintTrackingImpl
 }

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

@@ -569,7 +569,8 @@ class BuiltInOperationBuiltInAddressOf extends UnaryOperation, BuiltInOperation,
  * ```
  */
 class BuiltInOperationIsTriviallyConstructible extends BuiltInOperation,
-  @istriviallyconstructibleexpr {
+  @istriviallyconstructibleexpr
+{
   override string toString() { result = "__is_trivially_constructible" }
 
   override string getAPrimaryQlClass() { result = "BuiltInOperationIsTriviallyConstructible" }
@@ -619,7 +620,8 @@ class BuiltInOperationIsNothrowDestructible extends BuiltInOperation, @isnothrow
  * bool v = __is_trivially_destructible(MyType);
  * ```
  */
-class BuiltInOperationIsTriviallyDestructible extends BuiltInOperation, @istriviallydestructibleexpr {
+class BuiltInOperationIsTriviallyDestructible extends BuiltInOperation, @istriviallydestructibleexpr
+{
   override string toString() { result = "__is_trivially_destructible" }
 
   override string getAPrimaryQlClass() { result = "BuiltInOperationIsTriviallyDestructible" }
@@ -738,7 +740,8 @@ class BuiltInOperationIsLiteralType extends BuiltInOperation, @isliteraltypeexpr
  * ```
  */
 class BuiltInOperationHasTrivialMoveConstructor extends BuiltInOperation,
-  @hastrivialmoveconstructorexpr {
+  @hastrivialmoveconstructorexpr
+{
   override string toString() { result = "__has_trivial_move_constructor" }
 
   override string getAPrimaryQlClass() { result = "BuiltInOperationHasTrivialMoveConstructor" }
@@ -1034,7 +1037,8 @@ class BuiltInOperationIsAggregate extends BuiltInOperation, @isaggregate {
  * ```
  */
 class BuiltInOperationHasUniqueObjectRepresentations extends BuiltInOperation,
-  @hasuniqueobjectrepresentations {
+  @hasuniqueobjectrepresentations
+{
   override string toString() { result = "__has_unique_object_representations" }
 
   override string getAPrimaryQlClass() { result = "BuiltInOperationHasUniqueObjectRepresentations" }
@@ -1107,7 +1111,8 @@ class BuiltInOperationIsLayoutCompatible extends BuiltInOperation, @islayoutcomp
  * ```
  */
 class BuiltInOperationIsPointerInterconvertibleBaseOf extends BuiltInOperation,
-  @ispointerinterconvertiblebaseof {
+  @ispointerinterconvertiblebaseof
+{
   override string toString() { result = "__is_pointer_interconvertible_base_of" }
 
   override string getAPrimaryQlClass() {

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

@@ -719,13 +719,6 @@ class ReferenceToExpr extends Conversion, @reference_to {
 class PointerDereferenceExpr extends UnaryOperation, @indirect {
   override string getAPrimaryQlClass() { result = "PointerDereferenceExpr" }
 
-  /**
-   * DEPRECATED: Use getOperand() instead.
-   *
-   * Gets the expression that is being dereferenced.
-   */
-  deprecated Expr getExpr() { result = this.getOperand() }
-
   override string getOperator() { result = "*" }
 
   override int getPrecedence() { result = 16 }

cpp/ql/lib/semmle/code/cpp/ir/dataflow/DataFlow.qll

@@ -8,8 +8,8 @@
  * results than the AST-based library in most scenarios.
  *
  * Unless configured otherwise, _flow_ means that the exact value of
- * the source may reach the sink. We do not track flow across pointer
- * dereferences or array indexing.
+ * the source may reach the sink. To track flow where the exact value
+ * may not be preserved, import `semmle.code.cpp.ir.dataflow.TaintTracking`.
  *
  * To use global (interprocedural) data flow, extend the class
  * `DataFlow::Configuration` as documented on that class. To use local
@@ -22,5 +22,6 @@
 import cpp
 
 module DataFlow {
-  import semmle.code.cpp.ir.dataflow.internal.DataFlowImpl
+  import semmle.code.cpp.ir.dataflow.internal.DataFlow
+  import semmle.code.cpp.ir.dataflow.internal.DataFlowImpl1
 }

cpp/ql/lib/semmle/code/cpp/ir/dataflow/TaintTracking.qll

@@ -19,5 +19,6 @@ import semmle.code.cpp.ir.dataflow.DataFlow
 import semmle.code.cpp.ir.dataflow.DataFlow2
 
 module TaintTracking {
+  import semmle.code.cpp.ir.dataflow.internal.tainttracking1.TaintTracking
   import semmle.code.cpp.ir.dataflow.internal.tainttracking1.TaintTrackingImpl
 }

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

@@ -0,0 +1,353 @@
+/**
+ * Provides an implementation of global (interprocedural) data flow. This file
+ * re-exports the local (intraprocedural) data flow analysis from
+ * `DataFlowImplSpecific::Public` and adds a global analysis, mainly exposed
+ * through the `Make` and `MakeWithState` modules.
+ */
+
+private import DataFlowImplCommon
+private import DataFlowImplSpecific::Private
+import DataFlowImplSpecific::Public
+import DataFlowImplCommonPublic
+private import DataFlowImpl
+
+/** An input configuration for data flow. */
+signature module ConfigSig {
+  /**
+   * Holds if `source` is a relevant data flow source.
+   */
+  predicate isSource(Node source);
+
+  /**
+   * Holds if `sink` is a relevant data flow sink.
+   */
+  predicate isSink(Node sink);
+
+  /**
+   * Holds if data flow through `node` is prohibited. This completely removes
+   * `node` from the data flow graph.
+   */
+  default predicate isBarrier(Node node) { none() }
+
+  /** Holds if data flow into `node` is prohibited. */
+  default predicate isBarrierIn(Node node) { none() }
+
+  /** Holds if data flow out of `node` is prohibited. */
+  default predicate isBarrierOut(Node node) { none() }
+
+  /**
+   * Holds if data may flow from `node1` to `node2` in addition to the normal data-flow steps.
+   */
+  default predicate isAdditionalFlowStep(Node node1, Node node2) { none() }
+
+  /**
+   * Holds if an arbitrary number of implicit read steps of content `c` may be
+   * taken at `node`.
+   */
+  default predicate allowImplicitRead(Node node, ContentSet c) { none() }
+
+  /**
+   * Gets the virtual dispatch branching limit when calculating field flow.
+   * This can be overridden to a smaller value to improve performance (a
+   * value of 0 disables field flow), or a larger value to get more results.
+   */
+  default int fieldFlowBranchLimit() { result = 2 }
+
+  /**
+   * Gets a data flow configuration feature to add restrictions to the set of
+   * valid flow paths.
+   *
+   * - `FeatureHasSourceCallContext`:
+   *    Assume that sources have some existing call context to disallow
+   *    conflicting return-flow directly following the source.
+   * - `FeatureHasSinkCallContext`:
+   *    Assume that sinks have some existing call context to disallow
+   *    conflicting argument-to-parameter flow directly preceding the sink.
+   * - `FeatureEqualSourceSinkCallContext`:
+   *    Implies both of the above and additionally ensures that the entire flow
+   *    path preserves the call context.
+   *
+   * These features are generally not relevant for typical end-to-end data flow
+   * queries, but should only be used for constructing paths that need to
+   * somehow be pluggable in another path context.
+   */
+  default FlowFeature getAFeature() { none() }
+
+  /** Holds if sources should be grouped in the result of `hasFlowPath`. */
+  default predicate sourceGrouping(Node source, string sourceGroup) { none() }
+
+  /** Holds if sinks should be grouped in the result of `hasFlowPath`. */
+  default predicate sinkGrouping(Node sink, string sinkGroup) { none() }
+
+  /**
+   * Holds if hidden nodes should be included in the data flow graph.
+   *
+   * This feature should only be used for debugging or when the data flow graph
+   * is not visualized (as it is in a `path-problem` query).
+   */
+  default predicate includeHiddenNodes() { none() }
+}
+
+/** An input configuration for data flow using flow state. */
+signature module StateConfigSig {
+  bindingset[this]
+  class FlowState;
+
+  /**
+   * Holds if `source` is a relevant data flow source with the given initial
+   * `state`.
+   */
+  predicate isSource(Node source, FlowState state);
+
+  /**
+   * Holds if `sink` is a relevant data flow sink accepting `state`.
+   */
+  predicate isSink(Node sink, FlowState state);
+
+  /**
+   * Holds if data flow through `node` is prohibited. This completely removes
+   * `node` from the data flow graph.
+   */
+  default predicate isBarrier(Node node) { none() }
+
+  /**
+   * Holds if data flow through `node` is prohibited when the flow state is
+   * `state`.
+   */
+  predicate isBarrier(Node node, FlowState state);
+
+  /** Holds if data flow into `node` is prohibited. */
+  default predicate isBarrierIn(Node node) { none() }
+
+  /** Holds if data flow out of `node` is prohibited. */
+  default predicate isBarrierOut(Node node) { none() }
+
+  /**
+   * Holds if data may flow from `node1` to `node2` in addition to the normal data-flow steps.
+   */
+  default predicate isAdditionalFlowStep(Node node1, Node node2) { none() }
+
+  /**
+   * Holds if data may flow from `node1` to `node2` in addition to the normal data-flow steps.
+   * This step is only applicable in `state1` and updates the flow state to `state2`.
+   */
+  predicate isAdditionalFlowStep(Node node1, FlowState state1, Node node2, FlowState state2);
+
+  /**
+   * Holds if an arbitrary number of implicit read steps of content `c` may be
+   * taken at `node`.
+   */
+  default predicate allowImplicitRead(Node node, ContentSet c) { none() }
+
+  /**
+   * Gets the virtual dispatch branching limit when calculating field flow.
+   * This can be overridden to a smaller value to improve performance (a
+   * value of 0 disables field flow), or a larger value to get more results.
+   */
+  default int fieldFlowBranchLimit() { result = 2 }
+
+  /**
+   * Gets a data flow configuration feature to add restrictions to the set of
+   * valid flow paths.
+   *
+   * - `FeatureHasSourceCallContext`:
+   *    Assume that sources have some existing call context to disallow
+   *    conflicting return-flow directly following the source.
+   * - `FeatureHasSinkCallContext`:
+   *    Assume that sinks have some existing call context to disallow
+   *    conflicting argument-to-parameter flow directly preceding the sink.
+   * - `FeatureEqualSourceSinkCallContext`:
+   *    Implies both of the above and additionally ensures that the entire flow
+   *    path preserves the call context.
+   *
+   * These features are generally not relevant for typical end-to-end data flow
+   * queries, but should only be used for constructing paths that need to
+   * somehow be pluggable in another path context.
+   */
+  default FlowFeature getAFeature() { none() }
+
+  /** Holds if sources should be grouped in the result of `hasFlowPath`. */
+  default predicate sourceGrouping(Node source, string sourceGroup) { none() }
+
+  /** Holds if sinks should be grouped in the result of `hasFlowPath`. */
+  default predicate sinkGrouping(Node sink, string sinkGroup) { none() }
+
+  /**
+   * Holds if hidden nodes should be included in the data flow graph.
+   *
+   * This feature should only be used for debugging or when the data flow graph
+   * is not visualized (as it is in a `path-problem` query).
+   */
+  default predicate includeHiddenNodes() { none() }
+}
+
+/**
+ * Gets the exploration limit for `hasPartialFlow` and `hasPartialFlowRev`
+ * measured in approximate number of interprocedural steps.
+ */
+signature int explorationLimitSig();
+
+/**
+ * The output of a data flow computation.
+ */
+signature module DataFlowSig {
+  /**
+   * A `Node` augmented with a call context (except for sinks) and an access path.
+   * Only those `PathNode`s that are reachable from a source, and which can reach a sink, are generated.
+   */
+  class PathNode;
+
+  /**
+   * Holds if data can flow from `source` to `sink`.
+   *
+   * The corresponding paths are generated from the end-points and the graph
+   * included in the module `PathGraph`.
+   */
+  predicate hasFlowPath(PathNode source, PathNode sink);
+
+  /**
+   * Holds if data can flow from `source` to `sink`.
+   */
+  predicate hasFlow(Node source, Node sink);
+
+  /**
+   * Holds if data can flow from some source to `sink`.
+   */
+  predicate hasFlowTo(Node sink);
+
+  /**
+   * Holds if data can flow from some source to `sink`.
+   */
+  predicate hasFlowToExpr(DataFlowExpr sink);
+}
+
+/**
+ * Constructs a standard data flow computation.
+ */
+module Make<ConfigSig Config> implements DataFlowSig {
+  private module C implements FullStateConfigSig {
+    import DefaultState<Config>
+    import Config
+  }
+
+  import Impl<C>
+}
+
+/**
+ * Constructs a data flow computation using flow state.
+ */
+module MakeWithState<StateConfigSig Config> implements DataFlowSig {
+  private module C implements FullStateConfigSig {
+    import Config
+  }
+
+  import Impl<C>
+}
+
+signature class PathNodeSig {
+  /** Gets a textual representation of this element. */
+  string toString();
+
+  /**
+   * Holds if this element is at the specified location.
+   * The location spans column `startcolumn` of line `startline` to
+   * column `endcolumn` of line `endline` in file `filepath`.
+   * For more information, see
+   * [Locations](https://codeql.github.com/docs/writing-codeql-queries/providing-locations-in-codeql-queries/).
+   */
+  predicate hasLocationInfo(
+    string filepath, int startline, int startcolumn, int endline, int endcolumn
+  );
+
+  /** Gets the underlying `Node`. */
+  Node getNode();
+}
+
+signature module PathGraphSig<PathNodeSig PathNode> {
+  /** Holds if `(a,b)` is an edge in the graph of data flow path explanations. */
+  predicate edges(PathNode a, PathNode b);
+
+  /** Holds if `n` is a node in the graph of data flow path explanations. */
+  predicate nodes(PathNode n, string key, string val);
+
+  /**
+   * Holds if `(arg, par, ret, out)` forms a subpath-tuple, that is, flow through
+   * a subpath between `par` and `ret` with the connecting edges `arg -> par` and
+   * `ret -> out` is summarized as the edge `arg -> out`.
+   */
+  predicate subpaths(PathNode arg, PathNode par, PathNode ret, PathNode out);
+}
+
+/**
+ * Constructs a `PathGraph` from two `PathGraph`s by disjoint union.
+ */
+module MergePathGraph<
+  PathNodeSig PathNode1, PathNodeSig PathNode2, PathGraphSig<PathNode1> Graph1,
+  PathGraphSig<PathNode2> Graph2>
+{
+  private newtype TPathNode =
+    TPathNode1(PathNode1 p) or
+    TPathNode2(PathNode2 p)
+
+  /** A node in a graph of path explanations that is formed by disjoint union of the two given graphs. */
+  class PathNode extends TPathNode {
+    /** Gets this as a projection on the first given `PathGraph`. */
+    PathNode1 asPathNode1() { this = TPathNode1(result) }
+
+    /** Gets this as a projection on the second given `PathGraph`. */
+    PathNode2 asPathNode2() { this = TPathNode2(result) }
+
+    /** Gets a textual representation of this element. */
+    string toString() {
+      result = this.asPathNode1().toString() or
+      result = this.asPathNode2().toString()
+    }
+
+    /**
+     * Holds if this element is at the specified location.
+     * The location spans column `startcolumn` of line `startline` to
+     * column `endcolumn` of line `endline` in file `filepath`.
+     * For more information, see
+     * [Locations](https://codeql.github.com/docs/writing-codeql-queries/providing-locations-in-codeql-queries/).
+     */
+    predicate hasLocationInfo(
+      string filepath, int startline, int startcolumn, int endline, int endcolumn
+    ) {
+      this.asPathNode1().hasLocationInfo(filepath, startline, startcolumn, endline, endcolumn) or
+      this.asPathNode2().hasLocationInfo(filepath, startline, startcolumn, endline, endcolumn)
+    }
+
+    /** Gets the underlying `Node`. */
+    Node getNode() {
+      result = this.asPathNode1().getNode() or
+      result = this.asPathNode2().getNode()
+    }
+  }
+
+  /**
+   * Provides the query predicates needed to include a graph in a path-problem query.
+   */
+  module PathGraph implements PathGraphSig<PathNode> {
+    /** Holds if `(a,b)` is an edge in the graph of data flow path explanations. */
+    query predicate edges(PathNode a, PathNode b) {
+      Graph1::edges(a.asPathNode1(), b.asPathNode1()) or
+      Graph2::edges(a.asPathNode2(), b.asPathNode2())
+    }
+
+    /** Holds if `n` is a node in the graph of data flow path explanations. */
+    query predicate nodes(PathNode n, string key, string val) {
+      Graph1::nodes(n.asPathNode1(), key, val) or
+      Graph2::nodes(n.asPathNode2(), key, val)
+    }
+
+    /**
+     * Holds if `(arg, par, ret, out)` forms a subpath-tuple, that is, flow through
+     * a subpath between `par` and `ret` with the connecting edges `arg -> par` and
+     * `ret -> out` is summarized as the edge `arg -> out`.
+     */
+    query predicate subpaths(PathNode arg, PathNode par, PathNode ret, PathNode out) {
+      Graph1::subpaths(arg.asPathNode1(), par.asPathNode1(), ret.asPathNode1(), out.asPathNode1()) or
+      Graph2::subpaths(arg.asPathNode2(), par.asPathNode2(), ret.asPathNode2(), out.asPathNode2())
+    }
+  }
+}