Initial plan

.bazelrc

@@ -11,8 +11,6 @@ build --compilation_mode opt
 common --override_module=semmle_code=%workspace%/misc/bazel/semmle_code_stub
 
 build --repo_env=CC=clang --repo_env=CXX=clang++
-# Disable Android SDK auto-detection (we don't use it, and rules_android has Bazel 9 compatibility issues)
-build --repo_env=ANDROID_HOME=
 
 # print test output, like sembuild does.
 # Set to `errors` if this is too verbose.
@@ -36,7 +34,7 @@ common --@rules_dotnet//dotnet/settings:strict_deps=false
 common --@rules_rust//rust/toolchain/channel=nightly
 
 # Reduce this eventually to empty, once we've fixed all our usages of java, and https://github.com/bazel-contrib/rules_go/issues/4193 is fixed
-common --incompatible_autoload_externally="+@rules_cc,+@rules_java,+@rules_shell"
+common --incompatible_autoload_externally="+@rules_java,+@rules_shell"
 
 build --java_language_version=17
 build --tool_java_language_version=17

.bazelversion

@@ -1 +1 @@
-9.0.0
+8.4.2

.github/dependabot.yml

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

.github/workflows/compile-queries.yml

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

MODULE.bazel

@@ -15,23 +15,21 @@ local_path_override(
 # see https://registry.bazel.build/ for a list of available packages
 
 bazel_dep(name = "platforms", version = "1.0.0")
-bazel_dep(name = "rules_cc", version = "0.2.17")
-bazel_dep(name = "rules_go", version = "0.60.0")
-bazel_dep(name = "rules_java", version = "9.6.1")
-bazel_dep(name = "rules_pkg", version = "1.2.0")
-bazel_dep(name = "rules_nodejs", version = "6.7.3")
-bazel_dep(name = "rules_python", version = "1.9.0")
-bazel_dep(name = "rules_shell", version = "0.6.1")
-bazel_dep(name = "bazel_skylib", version = "1.9.0")
-bazel_dep(name = "abseil-cpp", version = "20260107.1", repo_name = "absl")
+bazel_dep(name = "rules_go", version = "0.56.1")
+bazel_dep(name = "rules_pkg", version = "1.0.1")
+bazel_dep(name = "rules_nodejs", version = "6.2.0-codeql.1")
+bazel_dep(name = "rules_python", version = "0.40.0")
+bazel_dep(name = "rules_shell", version = "0.5.0")
+bazel_dep(name = "bazel_skylib", version = "1.8.1")
+bazel_dep(name = "abseil-cpp", version = "20240116.1", repo_name = "absl")
 bazel_dep(name = "nlohmann_json", version = "3.11.3", repo_name = "json")
 bazel_dep(name = "fmt", version = "12.1.0-codeql.1")
-bazel_dep(name = "rules_kotlin", version = "2.2.2-codeql.1")
-bazel_dep(name = "gazelle", version = "0.47.0")
+bazel_dep(name = "rules_kotlin", version = "2.2.0-codeql.1")
+bazel_dep(name = "gazelle", version = "0.40.0")
 bazel_dep(name = "rules_dotnet", version = "0.21.5-codeql.1")
-bazel_dep(name = "googletest", version = "1.17.0.bcr.2")
-bazel_dep(name = "rules_rust", version = "0.69.0")
-bazel_dep(name = "zstd", version = "1.5.7.bcr.1")
+bazel_dep(name = "googletest", version = "1.14.0.bcr.1")
+bazel_dep(name = "rules_rust", version = "0.66.0")
+bazel_dep(name = "zstd", version = "1.5.5.bcr.1")
 
 bazel_dep(name = "buildifier_prebuilt", version = "6.4.0", dev_dependency = True)
 
@@ -43,7 +41,7 @@ RUST_EDITION = "2024"
 # a nightly toolchain is required to enable experimental_use_cc_common_link, which we require internally
 # we prefer to run the same version as internally, even if experimental_use_cc_common_link is not really
 # required in this repo
-RUST_VERSION = "nightly/2026-01-22"
+RUST_VERSION = "nightly/2025-08-01"
 
 rust = use_extension("@rules_rust//rust:extensions.bzl", "rust")
 rust.toolchain(
@@ -55,26 +53,26 @@ rust.toolchain(
     ],
     # generated by buildutils-internal/scripts/fill-rust-sha256s.py (internal repo)
     sha256s = {
-        "2026-01-22/rustc-nightly-x86_64-unknown-linux-gnu.tar.xz": "88db619323cc1321630d124efa51ed02fabc5e020f08cfa0eda2c0ac1afbe69a",
-        "2026-01-22/rustc-nightly-x86_64-apple-darwin.tar.xz": "08484da3fa38db56f93629aeabdc0ae9ff8ed9704c0792d35259cbc849b3f54c",
-        "2026-01-22/rustc-nightly-aarch64-apple-darwin.tar.xz": "a39c0b21b7058e364ea1bd43144e42e4bf1efade036b2e82455f2afce194ee81",
-        "2026-01-22/rustc-nightly-x86_64-pc-windows-msvc.tar.xz": "d00248ee9850dbb6932b2578e32ff74fc7c429854c1aa071066ca31b65385a3b",
-        "2026-01-22/clippy-nightly-x86_64-unknown-linux-gnu.tar.xz": "70656a0ce994ffff16d5a35a7b170a0acd41e9bb54a589c96ed45bf97b094a4d",
-        "2026-01-22/clippy-nightly-x86_64-apple-darwin.tar.xz": "fe242519fa961522734733009705aec3c2d9a20cc57291f2aa614e5e6262c88f",
-        "2026-01-22/clippy-nightly-aarch64-apple-darwin.tar.xz": "38bb226363ec97c9722edf966cd58774a683e19fd2ff2a6030094445d51e06f9",
-        "2026-01-22/clippy-nightly-x86_64-pc-windows-msvc.tar.xz": "6da9b4470beea67abfebf046f141eee0d2a8db7c7a9e4e2294478734fd477228",
-        "2026-01-22/cargo-nightly-x86_64-unknown-linux-gnu.tar.xz": "99004e9d10c43a01499642f53bb3184d41137a95d65bfb217098840a9e79e892",
-        "2026-01-22/cargo-nightly-x86_64-apple-darwin.tar.xz": "6e021394cf8d8400ac6cfdfcef24e4d74f988e91eb8028b36de3a64ce3502990",
-        "2026-01-22/cargo-nightly-aarch64-apple-darwin.tar.xz": "4b2494cb69ab64132cddbc411a38ea9f1105e54d6f986e43168d54f79510c673",
-        "2026-01-22/cargo-nightly-x86_64-pc-windows-msvc.tar.xz": "c36613cf57407212d10d37b76e49a60ff42336e953cdff9e177283f530a83fc1",
-        "2026-01-22/llvm-tools-nightly-x86_64-unknown-linux-gnu.tar.xz": "0b123c5027dbd833aae6845ffe9bd07d309bf798746a7176aadaea68fbcbd05d",
-        "2026-01-22/llvm-tools-nightly-x86_64-apple-darwin.tar.xz": "a47864491ad5619158c950ab7570fb6e487d5117338585c27334d45824b406d8",
-        "2026-01-22/llvm-tools-nightly-aarch64-apple-darwin.tar.xz": "db9bc826d6e2e7e914505d50157682e516ceb90357e83d77abddc32c2d962f41",
-        "2026-01-22/llvm-tools-nightly-x86_64-pc-windows-msvc.tar.xz": "ffaa406932b2fe62e01dad61cf4ed34860a5d2a6f9306ca340d79e630d930039",
-        "2026-01-22/rust-std-nightly-x86_64-unknown-linux-gnu.tar.xz": "e9c0d5e06e18a4b509391b3088f29293e310cdc8ccc865be8fa3f09733326925",
-        "2026-01-22/rust-std-nightly-x86_64-apple-darwin.tar.xz": "25d75995cee679a4828ca9fe48c5a31a67c3b0846018440ef912e5a6208f53f6",
-        "2026-01-22/rust-std-nightly-aarch64-apple-darwin.tar.xz": "e4132bf3f2eed4684c86756a02315bcf481c23e675e3e25630fc604c9cb4594c",
-        "2026-01-22/rust-std-nightly-x86_64-pc-windows-msvc.tar.xz": "961bb535ef95ae8a5fa4e224cb94aff190f155c45a9bcf7a53e184b024aa41b1",
+        "2025-08-01/rustc-nightly-x86_64-unknown-linux-gnu.tar.xz": "9bbeaf5d3fc7247d31463a9083aa251c995cc50662c8219e7a2254d76a72a9a4",
+        "2025-08-01/rustc-nightly-x86_64-apple-darwin.tar.xz": "c9ea539a8eff0d5d162701f99f9e1aabe14dd0dfb420d62362817a5d09219de7",
+        "2025-08-01/rustc-nightly-aarch64-apple-darwin.tar.xz": "ae83feebbc39cfd982e4ecc8297731fe79c185173aee138467b334c5404b3773",
+        "2025-08-01/rustc-nightly-x86_64-pc-windows-msvc.tar.xz": "9f170c30d802a349be60cf52ec46260802093cb1013ad667fc0d528b7b10152f",
+        "2025-08-01/clippy-nightly-x86_64-unknown-linux-gnu.tar.xz": "9ae5f3cd8f557c4f6df522597c69d14398cf604cfaed2b83e767c4b77a7eaaf6",
+        "2025-08-01/clippy-nightly-x86_64-apple-darwin.tar.xz": "983cb9ee0b6b968188e04ab2d33743d54764b2681ce565e1b3f2b9135c696a3e",
+        "2025-08-01/clippy-nightly-aarch64-apple-darwin.tar.xz": "ed2219dbc49d088225e1b7c5c4390fa295066e071fddaa2714018f6bb39ddbf0",
+        "2025-08-01/clippy-nightly-x86_64-pc-windows-msvc.tar.xz": "911f40ab5cbdd686f40e00965271fe47c4805513a308ed01f30eafb25b448a50",
+        "2025-08-01/cargo-nightly-x86_64-unknown-linux-gnu.tar.xz": "106463c284e48e4904c717471eeec2be5cc83a9d2cae8d6e948b52438cad2e69",
+        "2025-08-01/cargo-nightly-x86_64-apple-darwin.tar.xz": "6ad35c40efc41a8c531ea43235058347b6902d98a9693bf0aed7fc16d5590cef",
+        "2025-08-01/cargo-nightly-aarch64-apple-darwin.tar.xz": "dd28c365e9d298abc3154c797720ad36a0058f131265c9978b4c8e4e37012c8a",
+        "2025-08-01/cargo-nightly-x86_64-pc-windows-msvc.tar.xz": "7b431286e12d6b3834b038f078389a00cac73f351e8c3152b2504a3c06420b3b",
+        "2025-08-01/llvm-tools-nightly-x86_64-unknown-linux-gnu.tar.xz": "e342e305d7927cc288d386983b2bc253cfad3776b113386e903d0b302648ef47",
+        "2025-08-01/llvm-tools-nightly-x86_64-apple-darwin.tar.xz": "e44dd3506524d85c37b3a54bcc91d01378fd2c590b2db5c5974d12f05c1b84d1",
+        "2025-08-01/llvm-tools-nightly-aarch64-apple-darwin.tar.xz": "0c1b5f46dd81be4a9227b10283a0fcaa39c14fea7e81aea6fd6d9887ff6cdc41",
+        "2025-08-01/llvm-tools-nightly-x86_64-pc-windows-msvc.tar.xz": "423e5fd11406adccbc31b8456ceb7375ce055cdf45e90d2c3babeb2d7f58383f",
+        "2025-08-01/rust-std-nightly-x86_64-unknown-linux-gnu.tar.xz": "3c0ceb46a252647a1d4c7116d9ccae684fa5e42aaf3296419febd2c962c3b41d",
+        "2025-08-01/rust-std-nightly-x86_64-apple-darwin.tar.xz": "3be416003cab10f767390a753d1d16ae4d26c7421c03c98992cf1943e5b0efe8",
+        "2025-08-01/rust-std-nightly-aarch64-apple-darwin.tar.xz": "4046ac0ef951cb056b5028a399124f60999fa37792eab69d008d8d7965f389b4",
+        "2025-08-01/rust-std-nightly-x86_64-pc-windows-msvc.tar.xz": "191ed9d8603c3a4fe5a7bbbc2feb72049078dae2df3d3b7d5dedf3abbf823e6e",
     },
     versions = [RUST_VERSION],
 )
@@ -190,15 +188,6 @@ pip.parse(
 )
 use_repo(pip, "codegen_deps")
 
-python = use_extension("@rules_python//python/extensions:python.bzl", "python")
-python.toolchain(
-    is_default = True,
-    python_version = "3.12",
-)
-use_repo(python, "python_3_12", "python_versions")
-
-register_toolchains("@python_versions//3.12:all")
-
 swift_deps = use_extension("//swift/third_party:load.bzl", "swift_deps")
 
 # following list can be kept in sync with `bazel mod tidy`

README.md

@@ -29,7 +29,3 @@ You can install the [CodeQL for Visual Studio Code](https://marketplace.visualst
 ### Tasks
 
 The `.vscode/tasks.json` file defines custom tasks specific to working in this repository. To invoke one of these tasks, select the `Terminal | Run Task...` menu option, and then select the desired task from the dropdown. You can also invoke the `Tasks: Run Task` command from the command palette.
-
-
-
-test

actions/ql/lib/CHANGELOG.md

@@ -1,15 +1,3 @@
-## 0.4.30
-
-No user-facing changes.
-
-## 0.4.29
-
-No user-facing changes.
-
-## 0.4.28
-
-No user-facing changes.
-
 ## 0.4.27
 
 ### Bug Fixes

actions/ql/lib/change-notes/released/0.4.28.md

@@ -1,3 +0,0 @@
-## 0.4.28
-
-No user-facing changes.

actions/ql/lib/change-notes/released/0.4.29.md

@@ -1,3 +0,0 @@
-## 0.4.29
-
-No user-facing changes.

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

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

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

@@ -1,2 +1,2 @@
 ---
-lastReleaseVersion: 0.4.30
+lastReleaseVersion: 0.4.27

actions/ql/lib/qlpack.yml

@@ -1,5 +1,5 @@
 name: codeql/actions-all
-version: 0.4.31-dev
+version: 0.4.28-dev
 library: true
 warnOnImplicitThis: true
 dependencies:

actions/ql/src/CHANGELOG.md

@@ -1,15 +1,3 @@
-## 0.6.22
-
-No user-facing changes.
-
-## 0.6.21
-
-No user-facing changes.
-
-## 0.6.20
-
-No user-facing changes.
-
 ## 0.6.19
 
 No user-facing changes.

actions/ql/src/change-notes/released/0.6.20.md

@@ -1,3 +0,0 @@
-## 0.6.20
-
-No user-facing changes.

actions/ql/src/change-notes/released/0.6.21.md

@@ -1,3 +0,0 @@
-## 0.6.21
-
-No user-facing changes.

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

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

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

@@ -1,2 +1,2 @@
 ---
-lastReleaseVersion: 0.6.22
+lastReleaseVersion: 0.6.19

actions/ql/src/qlpack.yml

@@ -1,5 +1,5 @@
 name: codeql/actions-queries
-version: 0.6.23-dev
+version: 0.6.20-dev
 library: false
 warnOnImplicitThis: true
 groups: [actions, queries]

config/add-overlay-annotations.py

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

config/identical-files.json

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

cpp/downgrades/770002bb02322e04fa25345838ce6e82af285a0b/in_trap.ql

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

cpp/downgrades/770002bb02322e04fa25345838ce6e82af285a0b/source_file_uses_trap.ql

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

cpp/downgrades/770002bb02322e04fa25345838ce6e82af285a0b/upgrade.properties

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

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

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

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

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

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

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

cpp/ql/lib/CHANGELOG.md

@@ -1,30 +1,3 @@
-## 8.0.1
-
-### Minor Analysis Improvements
-
-* Inline expectations test comments, which are of the form `// $ tag` or `// $ tag=value`, are now parsed more strictly and will not be recognized if there isn't a space after the `$` symbol.
-
-## 8.0.0
-
-### Breaking Changes
-
-* CodeQL version 2.24.2 accidentally introduced a syntactical breaking change to `BarrierGuard<...>::getAnIndirectBarrierNode` and `InstructionBarrierGuard<...>::getAnIndirectBarrierNode`. These breaking changes have now been reverted so that the original code compiles again.
-* `MustFlow`, the inter-procedural must-flow data flow analysis library, has been re-worked to use parameterized modules. Like in the case of data flow and taint tracking, instead of extending the `MustFlowConfiguration` class, the user should now implement a module with the `MustFlow::ConfigSig` signature, and instantiate the `MustFlow::Global` parameterized module with the implemented module.
-
-### Minor Analysis Improvements
-
-* Refactored the "Year field changed using an arithmetic operation without checking for leap year" query (`cpp/leap-year/unchecked-after-arithmetic-year-modification`) to address large numbers of false positive results.
-
-### Bug Fixes
-
-* The `allowInterproceduralFlow` predicate of must-flow data flow configurations now correctly handles direct recursion.
-
-## 7.1.1
-
-### Minor Analysis Improvements
-
-* Added remote flow source models for the `winhttp.h` windows header and the Azure SDK core library for C/C++.
-
 ## 7.1.0
 
 ### New Features

cpp/ql/lib/change-notes/2026-02-03-windows-remote-flow-sources.md

@@ -0,0 +1,4 @@
+---
+category: minorAnalysis
+---
+* Added remote flow source models for the `winhttp.h` windows header and the Azure SDK core library for C/C++.

cpp/ql/lib/change-notes/released/7.1.1.md

@@ -1,5 +0,0 @@
-## 7.1.1
-
-### Minor Analysis Improvements
-
-* Added remote flow source models for the `winhttp.h` windows header and the Azure SDK core library for C/C++.

cpp/ql/lib/change-notes/released/8.0.0.md

@@ -1,14 +0,0 @@
-## 8.0.0
-
-### Breaking Changes
-
-* CodeQL version 2.24.2 accidentally introduced a syntactical breaking change to `BarrierGuard<...>::getAnIndirectBarrierNode` and `InstructionBarrierGuard<...>::getAnIndirectBarrierNode`. These breaking changes have now been reverted so that the original code compiles again.
-* `MustFlow`, the inter-procedural must-flow data flow analysis library, has been re-worked to use parameterized modules. Like in the case of data flow and taint tracking, instead of extending the `MustFlowConfiguration` class, the user should now implement a module with the `MustFlow::ConfigSig` signature, and instantiate the `MustFlow::Global` parameterized module with the implemented module.
-
-### Minor Analysis Improvements
-
-* Refactored the "Year field changed using an arithmetic operation without checking for leap year" query (`cpp/leap-year/unchecked-after-arithmetic-year-modification`) to address large numbers of false positive results.
-
-### Bug Fixes
-
-* The `allowInterproceduralFlow` predicate of must-flow data flow configurations now correctly handles direct recursion.

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

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

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

@@ -1,2 +1,2 @@
 ---
-lastReleaseVersion: 8.0.1
+lastReleaseVersion: 7.1.0

cpp/ql/lib/qlpack.yml

@@ -1,5 +1,5 @@
 name: codeql/cpp-all
-version: 8.0.2-dev
+version: 7.1.1-dev
 groups: cpp
 dbscheme: semmlecode.cpp.dbscheme
 extractor: cpp

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

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

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

@@ -14,9 +14,7 @@ class PackedTimeType extends Type {
   }
 }
 
-private predicate timeType(string typeName) {
-  typeName = ["_SYSTEMTIME", "SYSTEMTIME", "tm", "TIME_FIELDS", "_TIME_FIELDS", "PTIME_FIELDS"]
-}
+private predicate timeType(string typeName) { typeName = ["_SYSTEMTIME", "SYSTEMTIME", "tm"] }
 
 /**
  * A type that is used to represent times and dates in an 'unpacked' form, that is,
@@ -97,24 +95,3 @@ class StructTmMonthFieldAccess extends MonthFieldAccess {
 class StructTmYearFieldAccess extends YearFieldAccess {
   StructTmYearFieldAccess() { this.getTarget().getName() = "tm_year" }
 }
-
-/**
- * A `DayFieldAccess` for the `TIME_FIELDS` struct.
- */
-class TimeFieldsDayFieldAccess extends DayFieldAccess {
-  TimeFieldsDayFieldAccess() { this.getTarget().getName() = "Day" }
-}
-
-/**
- * A `MonthFieldAccess` for the `TIME_FIELDS` struct.
- */
-class TimeFieldsMonthFieldAccess extends MonthFieldAccess {
-  TimeFieldsMonthFieldAccess() { this.getTarget().getName() = "Month" }
-}
-
-/**
- * A `YearFieldAccess` for the `TIME_FIELDS` struct.
- */
-class TimeFieldsYearFieldAccess extends YearFieldAccess {
-  TimeFieldsYearFieldAccess() { this.getTarget().getName() = "Year" }
-}

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

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

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

@@ -353,26 +353,12 @@ module CsvValidation {
     )
   }
 
-  private string getIncorrectConstructorSummaryOutput() {
-    exists(string namespace, string type, string name, string output |
-      type = name or
-      type = name + "<" + any(string s)
-    |
-      summaryModel(namespace, type, _, name, _, _, _, output, _, _, _) and
-      output.matches("ReturnValue%") and
-      result =
-        "Constructor model for " + namespace + "." + type +
-          " should use `Argument[this]` in the output, not `ReturnValue`."
-    )
-  }
-
   /** Holds if some row in a CSV-based flow model appears to contain typos. */
   query predicate invalidModelRow(string msg) {
     msg =
       [
         getInvalidModelSignature(), getInvalidModelInput(), getInvalidModelOutput(),
-        getInvalidModelSubtype(), getInvalidModelColumnCount(), KindVal::getInvalidModelKind(),
-        getIncorrectConstructorSummaryOutput()
+        getInvalidModelSubtype(), getInvalidModelColumnCount(), KindVal::getInvalidModelKind()
       ]
   }
 }
@@ -569,7 +555,6 @@ private Locatable getSupportedFunctionTemplateArgument(Function templateFunction
  * Normalize the `n`'th parameter of `f` by replacing template names
  * with `func:N` (where `N` is the index of the template).
  */
-pragma[nomagic]
 private string getTypeNameWithoutFunctionTemplates(Function f, int n, int remaining) {
   exists(Function templateFunction |
     templateFunction = getFullyTemplatedFunction(f) and

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

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

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

@@ -6,67 +6,85 @@ private import OverlayXml
 
 /**
  * Holds always for the overlay variant and never for the base variant.
+ * This local predicate is used to define local predicates that behave
+ * differently for the base and overlay variant.
  */
 overlay[local]
 predicate isOverlay() { databaseMetadata("isOverlay", "true") }
 
+overlay[local]
+private string getLocationFilePath(@location_default loc) {
+  exists(@file file | locations_default(loc, file, _, _, _, _) | files(file, result))
+}
+
 /**
- * Holds if the TRAP file or tag `t` is reachable from source file `sourceFile`
- * in the base (isOverlayVariant=false) or overlay (isOverlayVariant=true) variant.
+ * Gets the file path for an element with a single location.
  */
 overlay[local]
-private predicate locallyReachableTrapOrTag(
-  boolean isOverlayVariant, string sourceFile, @trap_or_tag t
-) {
-  exists(@source_file sf, @trap trap |
-    (if isOverlay() then isOverlayVariant = true else isOverlayVariant = false) and
-    source_file_uses_trap(sf, trap) and
-    source_file_name(sf, sourceFile) and
-    (t = trap or trap_uses_tag(trap, t))
+private string getSingleLocationFilePath(@element e) {
+  exists(@location_default loc |
+    var_decls(e, _, _, _, loc)
+    or
+    fun_decls(e, _, _, _, loc)
+    or
+    type_decls(e, _, loc)
+    or
+    namespace_decls(e, _, loc, _)
+    or
+    macroinvocations(e, _, loc, _)
+    or
+    preprocdirects(e, _, loc)
+  |
+    result = getLocationFilePath(loc)
   )
 }
 
 /**
- * Holds if element `e` is in TRAP file or tag `t`
- * in the base (isOverlayVariant=false) or overlay (isOverlayVariant=true) variant.
+ * Gets the file path for an element with potentially multiple locations.
  */
 overlay[local]
-private predicate locallyInTrapOrTag(boolean isOverlayVariant, @element e, @trap_or_tag t) {
-  (if isOverlay() then isOverlayVariant = true else isOverlayVariant = false) and
-  in_trap_or_tag(e, t)
+private string getMultiLocationFilePath(@element e) {
+  exists(@location_default loc |
+    var_decls(_, e, _, _, loc)
+    or
+    fun_decls(_, e, _, _, loc)
+    or
+    type_decls(_, e, loc)
+    or
+    namespace_decls(_, e, loc, _)
+  |
+    result = getLocationFilePath(loc)
+  )
+}
+
+/**
+ * A local helper predicate that holds in the base variant and never in the
+ * overlay variant.
+ */
+overlay[local]
+private predicate isBase() { not isOverlay() }
+
+/**
+ * Holds if `path` was extracted in the overlay database.
+ */
+overlay[local]
+private predicate overlayHasFile(string path) {
+  isOverlay() and
+  files(_, path) and
+  path != ""
 }
 
 /**
  * Discards an element from the base variant if:
- * - We have knowledge about what TRAP file or tag it is in (in the base).
- * - It is not in any overlay TRAP file or tag that is reachable from an overlay source file.
- * - For every base TRAP file or tag that contains it and is reachable from a base source file,
- *   either the source file has changed, or the overlay has redefined the TRAP file or tag,
- *   or the overlay runner has re-extracted the same source file.
+ * - It has a single location in a file extracted in the overlay, or
+ * - All of its locations are in files extracted in the overlay.
  */
 overlay[discard_entity]
 private predicate discardElement(@element e) {
-  // If we don't have any knowledge about what TRAP file something
-  // is in, then we don't want to discard it, so we only consider
-  // entities that are known to be in a base TRAP file or tag.
-  locallyInTrapOrTag(false, e, _) and
-  // Anything that is reachable from an overlay source file should
-  // not be discarded.
-  not exists(@trap_or_tag t | locallyInTrapOrTag(true, e, t) |
-    locallyReachableTrapOrTag(true, _, t)
-  ) and
-  // Finally, we have to make sure the base variant does not retain it.
-  // If it is reachable from a base source file, then that is
-  // sufficient unless either the base source file has changed (in
-  // particular, been deleted), or the overlay has redefined the TRAP
-  // file or tag it is in, or the overlay runner has re-extracted the same
-  // source file (e.g. because a header it includes has changed).
-  forall(@trap_or_tag t, string sourceFile |
-    locallyInTrapOrTag(false, e, t) and
-    locallyReachableTrapOrTag(false, sourceFile, t)
-  |
-    overlayChangedFiles(sourceFile) or
-    locallyReachableTrapOrTag(true, _, t) or
-    locallyReachableTrapOrTag(true, sourceFile, _)
+  isBase() and
+  (
+    overlayHasFile(getSingleLocationFilePath(e))
+    or
+    forex(string path | path = getMultiLocationFilePath(e) | overlayHasFile(path))
   )
 }

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

@@ -8,145 +8,83 @@ private import cpp
 private import semmle.code.cpp.ir.IR
 
 /**
- * Provides an inter-procedural must-flow data flow analysis.
+ * A configuration of a data flow analysis that performs must-flow analysis. This is different
+ * from `DataFlow.qll` which performs may-flow analysis (i.e., it finds paths where the source _may_
+ * flow to the sink).
+ *
+ * Like in `DataFlow.qll`, each use of the `MustFlow.qll` 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 and override `isSource`, `isSink` (and
+ * `isAdditionalFlowStep` if additional steps are required).
  */
-module MustFlow {
-  /**
-   * An input configuration of a data flow analysis that performs must-flow analysis. This is different
-   * from `DataFlow.qll` which performs may-flow analysis (i.e., it finds paths where the source _may_
-   * flow to the sink).
-   */
-  signature module ConfigSig {
-    /**
-     * Holds if `source` is a relevant data flow source.
-     */
-    predicate isSource(Instruction source);
-
-    /**
-     * Holds if `sink` is a relevant data flow sink.
-     */
-    predicate isSink(Operand sink);
-
-    /**
-     * Holds if data flow through `instr` is prohibited.
-     */
-    default predicate isBarrier(Instruction instr) { none() }
-
-    /**
-     * Holds if the additional flow step from `node1` to `node2` must be taken
-     * into account in the analysis.
-     */
-    default predicate isAdditionalFlowStep(Operand node1, Instruction node2) { none() }
-
-    /** Holds if this configuration allows flow from arguments to parameters. */
-    default predicate allowInterproceduralFlow() { any() }
-  }
+abstract class MustFlowConfiguration extends string {
+  bindingset[this]
+  MustFlowConfiguration() { any() }
 
   /**
-   * Constructs a global must-flow computation.
+   * Holds if `source` is a relevant data flow source.
    */
-  module Global<ConfigSig Config> {
-    import Config
+  abstract predicate isSource(Instruction source);
 
-    /**
-     * Holds if data must flow from `source` to `sink`.
-     *
-     * The corresponding paths are generated from the end-points and the graph
-     * included in the module `PathGraph`.
-     */
-    predicate flowPath(PathNode source, PathSink sink) {
-      isSource(source.getInstruction()) and
-      source.getASuccessor*() = sink
-    }
+  /**
+   * Holds if `sink` is a relevant data flow sink.
+   */
+  abstract predicate isSink(Operand sink);
 
-    /** Holds if `node` flows from a source. */
-    pragma[nomagic]
-    private predicate flowsFromSource(Instruction node) {
-      not isBarrier(node) and
-      (
-        isSource(node)
-        or
-        exists(Instruction mid |
-          step(mid, node) and
-          flowsFromSource(mid)
-        )
-      )
-    }
+  /**
+   * Holds if data flow through `instr` is prohibited.
+   */
+  predicate isBarrier(Instruction instr) { none() }
 
-    /** Holds if `node` flows to a sink. */
-    pragma[nomagic]
-    private predicate flowsToSink(Instruction node) {
-      flowsFromSource(node) and
-      (
-        isSink(node.getAUse())
-        or
-        exists(Instruction mid |
-          step(node, mid) and
-          flowsToSink(mid)
-        )
-      )
-    }
+  /**
+   * Holds if the additional flow step from `node1` to `node2` must be taken
+   * into account in the analysis.
+   */
+  predicate isAdditionalFlowStep(Operand node1, Instruction node2) { none() }
 
-    /** Holds if `nodeFrom` flows to `nodeTo`. */
-    private predicate step(Instruction nodeFrom, Instruction nodeTo) {
-      Cached::localStep(nodeFrom, nodeTo)
-      or
-      allowInterproceduralFlow() and
-      Cached::flowThroughCallable(nodeFrom, nodeTo)
-      or
-      isAdditionalFlowStep(nodeFrom.getAUse(), nodeTo)
-    }
+  /** Holds if this configuration allows flow from arguments to parameters. */
+  predicate allowInterproceduralFlow() { any() }
 
-    private newtype TLocalPathNode =
-      MkLocalPathNode(Instruction n) {
-        flowsToSink(n) and
-        (
-          isSource(n)
-          or
-          exists(PathNode mid | step(mid.getInstruction(), n))
-        )
-      }
-
-    /** A `Node` that is in a path from a source to a sink. */
-    class PathNode extends TLocalPathNode {
-      Instruction n;
-
-      PathNode() { this = MkLocalPathNode(n) }
-
-      /** Gets the underlying node. */
-      Instruction getInstruction() { result = n }
-
-      /** Gets a textual representation of this node. */
-      string toString() { result = n.getAst().toString() }
-
-      /** Gets the location of this element. */
-      Location getLocation() { result = n.getLocation() }
-
-      /** Gets a successor node, if any. */
-      PathNode getASuccessor() { step(this.getInstruction(), result.getInstruction()) }
-    }
-
-    private class PathSink extends PathNode {
-      PathSink() { isSink(this.getInstruction().getAUse()) }
-    }
-
-    /**
-     * Provides the query predicates needed to include a graph in a path-problem query.
-     */
-    module PathGraph {
-      private predicate reach(PathNode n) { n instanceof PathSink or reach(n.getASuccessor()) }
-
-      /** Holds if `(a,b)` is an edge in the graph of data flow path explanations. */
-      query predicate edges(PathNode a, PathNode b) { a.getASuccessor() = b and reach(b) }
-
-      /** Holds if `n` is a node in the graph of data flow path explanations. */
-      query predicate nodes(PathNode n, string key, string val) {
-        reach(n) and key = "semmle.label" and val = n.toString()
-      }
-    }
+  /**
+   * Holds if data must 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`.
+   */
+  final predicate hasFlowPath(MustFlowPathNode source, MustFlowPathSink sink) {
+    this.isSource(source.getInstruction()) and
+    source.getASuccessor*() = sink
   }
 }
 
+/** Holds if `node` flows from a source. */
+pragma[nomagic]
+private predicate flowsFromSource(Instruction node, MustFlowConfiguration config) {
+  not config.isBarrier(node) and
+  (
+    config.isSource(node)
+    or
+    exists(Instruction mid |
+      step(mid, node, config) and
+      flowsFromSource(mid, pragma[only_bind_into](config))
+    )
+  )
+}
+
+/** Holds if `node` flows to a sink. */
+pragma[nomagic]
+private predicate flowsToSink(Instruction node, MustFlowConfiguration config) {
+  flowsFromSource(node, pragma[only_bind_into](config)) and
+  (
+    config.isSink(node.getAUse())
+    or
+    exists(Instruction mid |
+      step(node, mid, config) and
+      flowsToSink(mid, pragma[only_bind_into](config))
+    )
+  )
+}
+
 cached
 private module Cached {
   /** Holds if `p` is the `n`'th parameter of the non-virtual function `f`. */
@@ -164,7 +102,7 @@ private module Cached {
     not f.isVirtual() and
     call.getPositionalArgument(n) = instr and
     f = call.getStaticCallTarget() and
-    isEnclosingNonVirtualFunctionInitializeParameter(init, f) and
+    getEnclosingNonVirtualFunctionInitializeParameter(init, f) and
     init.getParameter().getIndex() = pragma[only_bind_into](pragma[only_bind_out](n))
   }
 
@@ -173,7 +111,7 @@ private module Cached {
    * corresponding initialization instruction that receives the value of `instr` in `f`.
    */
   pragma[noinline]
-  private predicate isPositionalArgumentInitParam(
+  private predicate getPositionalArgumentInitParam(
     CallInstruction call, Instruction instr, InitializeParameterInstruction init, Function f
   ) {
     exists(int n |
@@ -188,18 +126,18 @@ private module Cached {
    * `instr` in `f`.
    */
   pragma[noinline]
-  private predicate isThisArgumentInitParam(
+  private predicate getThisArgumentInitParam(
     CallInstruction call, Instruction instr, InitializeParameterInstruction init, Function f
   ) {
     not f.isVirtual() and
     call.getStaticCallTarget() = f and
-    isEnclosingNonVirtualFunctionInitializeParameter(init, f) and
+    getEnclosingNonVirtualFunctionInitializeParameter(init, f) and
     call.getThisArgument() = instr and
     init.getIRVariable() instanceof IRThisVariable
   }
 
   /** Holds if `f` is the enclosing non-virtual function of `init`. */
-  private predicate isEnclosingNonVirtualFunctionInitializeParameter(
+  private predicate getEnclosingNonVirtualFunctionInitializeParameter(
     InitializeParameterInstruction init, Function f
   ) {
     not f.isVirtual() and
@@ -207,7 +145,7 @@ private module Cached {
   }
 
   /** Holds if `f` is the enclosing non-virtual function of `init`. */
-  private predicate isEnclosingNonVirtualFunctionInitializeIndirection(
+  private predicate getEnclosingNonVirtualFunctionInitializeIndirection(
     InitializeIndirectionInstruction init, Function f
   ) {
     not f.isVirtual() and
@@ -215,16 +153,15 @@ private module Cached {
   }
 
   /**
-   * Holds if `argument` is an argument (or argument indirection) to a call, and
-   * `parameter` is the corresponding initialization instruction in the call target.
+   * Holds if `instr` is an argument (or argument indirection) to a call, and
+   * `succ` is the corresponding initialization instruction in the call target.
    */
-  cached
-  predicate flowThroughCallable(Instruction argument, Instruction parameter) {
+  private predicate flowThroughCallable(Instruction argument, Instruction parameter) {
     // Flow from an argument to a parameter
     exists(CallInstruction call, InitializeParameterInstruction init | init = parameter |
-      isPositionalArgumentInitParam(call, argument, init, call.getStaticCallTarget())
+      getPositionalArgumentInitParam(call, argument, init, call.getStaticCallTarget())
       or
-      isThisArgumentInitParam(call, argument, init, call.getStaticCallTarget())
+      getThisArgumentInitParam(call, argument, init, call.getStaticCallTarget())
     )
     or
     // Flow from argument indirection to parameter indirection
@@ -233,7 +170,7 @@ private module Cached {
     |
       init = parameter and
       read.getPrimaryInstruction() = call and
-      isEnclosingNonVirtualFunctionInitializeIndirection(init, call.getStaticCallTarget())
+      getEnclosingNonVirtualFunctionInitializeIndirection(init, call.getStaticCallTarget())
     |
       exists(int n |
         read.getSideEffectOperand().getAnyDef() = argument and
@@ -268,10 +205,92 @@ private module Cached {
   }
 
   cached
-  predicate localStep(Instruction nodeFrom, Instruction nodeTo) {
+  predicate step(Instruction nodeFrom, Instruction nodeTo) {
     exists(Operand mid |
       instructionToOperandStep(nodeFrom, mid) and
       operandToInstructionStep(mid, nodeTo)
     )
+    or
+    flowThroughCallable(nodeFrom, nodeTo)
+  }
+}
+
+/**
+ * Gets the enclosing callable of `n`. Unlike `n.getEnclosingCallable()`, this
+ * predicate ensures that joins go from `n` to the result instead of the other
+ * way around.
+ */
+pragma[inline]
+private IRFunction getEnclosingCallable(Instruction n) {
+  pragma[only_bind_into](result) = pragma[only_bind_out](n).getEnclosingIRFunction()
+}
+
+/** Holds if `nodeFrom` flows to `nodeTo`. */
+private predicate step(Instruction nodeFrom, Instruction nodeTo, MustFlowConfiguration config) {
+  exists(config) and
+  Cached::step(pragma[only_bind_into](nodeFrom), pragma[only_bind_into](nodeTo)) and
+  (
+    config.allowInterproceduralFlow()
+    or
+    getEnclosingCallable(nodeFrom) = getEnclosingCallable(nodeTo)
+  )
+  or
+  config.isAdditionalFlowStep(nodeFrom.getAUse(), nodeTo)
+}
+
+private newtype TLocalPathNode =
+  MkLocalPathNode(Instruction n, MustFlowConfiguration config) {
+    flowsToSink(n, config) and
+    (
+      config.isSource(n)
+      or
+      exists(MustFlowPathNode mid | step(mid.getInstruction(), n, config))
+    )
+  }
+
+/** A `Node` that is in a path from a source to a sink. */
+class MustFlowPathNode extends TLocalPathNode {
+  Instruction n;
+
+  MustFlowPathNode() { this = MkLocalPathNode(n, _) }
+
+  /** Gets the underlying node. */
+  Instruction getInstruction() { result = n }
+
+  /** Gets a textual representation of this node. */
+  string toString() { result = n.getAst().toString() }
+
+  /** Gets the location of this element. */
+  Location getLocation() { result = n.getLocation() }
+
+  /** Gets a successor node, if any. */
+  MustFlowPathNode getASuccessor() {
+    step(this.getInstruction(), result.getInstruction(), this.getConfiguration())
+  }
+
+  /** Gets the associated configuration. */
+  MustFlowConfiguration getConfiguration() { this = MkLocalPathNode(_, result) }
+}
+
+private class MustFlowPathSink extends MustFlowPathNode {
+  MustFlowPathSink() { this.getConfiguration().isSink(this.getInstruction().getAUse()) }
+}
+
+/**
+ * Provides the query predicates needed to include a graph in a path-problem query.
+ */
+module PathGraph {
+  private predicate reach(MustFlowPathNode n) {
+    n instanceof MustFlowPathSink or reach(n.getASuccessor())
+  }
+
+  /** Holds if `(a,b)` is an edge in the graph of data flow path explanations. */
+  query predicate edges(MustFlowPathNode a, MustFlowPathNode b) {
+    a.getASuccessor() = b and reach(b)
+  }
+
+  /** Holds if `n` is a node in the graph of data flow path explanations. */
+  query predicate nodes(MustFlowPathNode n, string key, string val) {
+    reach(n) and key = "semmle.label" and val = n.toString()
   }
 }

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

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

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

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

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

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

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

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

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

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

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

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

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

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

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

@@ -4,12 +4,47 @@ import semmle.code.cpp.ir.internal.IRCppLanguage
 private import semmle.code.cpp.ir.implementation.raw.internal.SideEffects as SideEffects
 private import DataFlowImplCommon as DataFlowImplCommon
 private import DataFlowUtil
-private import DataFlowNodes
 private import semmle.code.cpp.models.interfaces.PointerWrapper
 private import DataFlowPrivate
 private import TypeFlow
 private import semmle.code.cpp.ir.ValueNumbering
 
+/**
+ * Holds if `operand` is an operand that is not used by the dataflow library.
+ * Ignored operands are not recognized as uses by SSA, and they don't have a
+ * corresponding `(Indirect)OperandNode`.
+ */
+predicate ignoreOperand(Operand operand) {
+  operand = any(Instruction instr | ignoreInstruction(instr)).getAnOperand() or
+  operand = any(Instruction instr | ignoreInstruction(instr)).getAUse() or
+  operand instanceof MemoryOperand
+}
+
+/**
+ * Holds if `instr` is an instruction that is not used by the dataflow library.
+ * Ignored instructions are not recognized as reads/writes by SSA, and they
+ * don't have a corresponding `(Indirect)InstructionNode`.
+ */
+predicate ignoreInstruction(Instruction instr) {
+  DataFlowImplCommon::forceCachingInSameStage() and
+  (
+    instr instanceof CallSideEffectInstruction or
+    instr instanceof CallReadSideEffectInstruction or
+    instr instanceof ExitFunctionInstruction or
+    instr instanceof EnterFunctionInstruction or
+    instr instanceof WriteSideEffectInstruction or
+    instr instanceof PhiInstruction or
+    instr instanceof ReadSideEffectInstruction or
+    instr instanceof ChiInstruction or
+    instr instanceof InitializeIndirectionInstruction or
+    instr instanceof AliasedDefinitionInstruction or
+    instr instanceof AliasedUseInstruction or
+    instr instanceof InitializeNonLocalInstruction or
+    instr instanceof ReturnIndirectionInstruction or
+    instr instanceof UninitializedGroupInstruction
+  )
+}
+
 /**
  * Gets the C++ type of `this` in the member function `f`.
  * The result is a glvalue if `isGLValue` is true, and
@@ -20,6 +55,26 @@ 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.
@@ -292,6 +347,10 @@ predicate isWrite(Node0Impl value, Operand address, boolean certain) {
   )
 }
 
+predicate isAdditionalConversionFlow(Operand opFrom, Instruction instrTo) {
+  any(Indirection ind).isAdditionalConversionFlow(opFrom, instrTo)
+}
+
 newtype TBaseSourceVariable =
   // Each IR variable gets its own source variable
   TBaseIRVariable(IRVariable var) or
@@ -513,69 +572,6 @@ private class BaseCallInstruction extends BaseSourceVariableInstruction, CallIns
 
 cached
 private module Cached {
-  /**
-   * Holds if `operand` is an operand that is not used by the dataflow library.
-   * Ignored operands are not recognized as uses by SSA, and they don't have a
-   * corresponding `(Indirect)OperandNode`.
-   */
-  cached
-  predicate ignoreOperand(Operand operand) {
-    operand = any(Instruction instr | ignoreInstruction(instr)).getAnOperand() or
-    operand = any(Instruction instr | ignoreInstruction(instr)).getAUse() or
-    operand instanceof MemoryOperand
-  }
-
-  /**
-   * Holds if `instr` is an instruction that is not used by the dataflow library.
-   * Ignored instructions are not recognized as reads/writes by SSA, and they
-   * don't have a corresponding `(Indirect)InstructionNode`.
-   */
-  cached
-  predicate ignoreInstruction(Instruction instr) {
-    DataFlowImplCommon::forceCachingInSameStage() and
-    (
-      instr instanceof CallSideEffectInstruction or
-      instr instanceof CallReadSideEffectInstruction or
-      instr instanceof ExitFunctionInstruction or
-      instr instanceof EnterFunctionInstruction or
-      instr instanceof WriteSideEffectInstruction or
-      instr instanceof PhiInstruction or
-      instr instanceof ReadSideEffectInstruction or
-      instr instanceof ChiInstruction or
-      instr instanceof InitializeIndirectionInstruction or
-      instr instanceof AliasedDefinitionInstruction or
-      instr instanceof AliasedUseInstruction or
-      instr instanceof InitializeNonLocalInstruction or
-      instr instanceof ReturnIndirectionInstruction or
-      instr instanceof UninitializedGroupInstruction
-    )
-  }
-
-  cached
-  predicate isAdditionalConversionFlow(Operand opFrom, Instruction instrTo) {
-    any(Indirection ind).isAdditionalConversionFlow(opFrom, instrTo)
-  }
-
-  /**
-   * Gets the C++ type of the instruction `i`.
-   *
-   * This is equivalent to `i.getResultLanguageType()` with the exception
-   * of instructions that directly references a `this` IRVariable. In this
-   * case, `i.getResultLanguageType()` gives an unknown type, whereas the
-   * predicate gives the expected type (i.e., a potentially cv-qualified
-   * type `A*` where `A` is the declaring type of the member function that
-   * contains `i`).
-   */
-  cached
-  CppType getResultLanguageType(Instruction i) {
-    if i.(VariableAddressInstruction).getIRVariable() instanceof IRThisVariable
-    then
-      if i.isGLValue()
-      then result = getThisType(i.getEnclosingFunction(), true)
-      else result = getThisType(i.getEnclosingFunction(), false)
-    else result = i.getResultLanguageType()
-  }
-
   /** Holds if `op` is the only use of its defining instruction, and that op is used in a conversation */
   private predicate isConversion(Operand op) {
     exists(Instruction def, Operand use |

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

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

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

@@ -15,7 +15,6 @@ private import TranslatedCall
 private import TranslatedStmt
 private import TranslatedFunction
 private import TranslatedGlobalVar
-private import TranslatedInitialization
 
 TranslatedElement getInstructionTranslatedElement(Instruction instruction) {
   instruction = TRawInstruction(result, _)
@@ -195,89 +194,6 @@ module Raw {
   Expr getInstructionUnconvertedResultExpression(Instruction instruction) {
     result = getInstructionConvertedResultExpression(instruction).getUnconverted()
   }
-
-  /**
-   * Gets the expression associated with the instruction `instr` that computes
-   * the `Convert` instruction on the extent expression of an allocation.
-   */
-  cached
-  Expr getAllocationExtentConvertExpr(Instruction instr) {
-    exists(TranslatedNonConstantAllocationSize tas |
-      instr = tas.getInstruction(AllocationExtentConvertTag()) and
-      result = tas.getExtent().getExpr()
-    )
-  }
-
-  /**
-   * Gets the `ParenthesisExpr` associated with a transparent conversion
-   * instruction, if any.
-   */
-  cached
-  ParenthesisExpr getTransparentConversionParenthesisExpr(Instruction instr) {
-    exists(TranslatedTransparentConversion ttc |
-      result = ttc.getExpr() and
-      instr = ttc.getResult()
-    )
-  }
-
-  /**
-   * Holds if `instr` belongs to a `TranslatedCoreExpr` that produces an
-   * expression result. This indicates that the instruction represents a
-   * definition whose result should be mapped back to the expression.
-   */
-  cached
-  predicate instructionProducesExprResult(Instruction instr) {
-    exists(TranslatedCoreExpr tco |
-      tco.getInstruction(_) = instr and
-      tco.producesExprResult()
-    )
-  }
-
-  /**
-   * Gets the expression associated with a `StoreInstruction` generated
-   * by an `TranslatedAssignOperation`.
-   */
-  cached
-  Expr getAssignOperationStoreExpr(StoreInstruction store) {
-    exists(TranslatedAssignOperation tao |
-      store = tao.getInstruction(AssignmentStoreTag()) and
-      result = tao.getExpr()
-    )
-  }
-
-  /**
-   * Gets the expression associated with a `StoreInstruction` generated
-   * by an `TranslatedCrementOperation`.
-   */
-  cached
-  Expr getCrementOperationStoreExpr(StoreInstruction store) {
-    exists(TranslatedCrementOperation tco |
-      store = tco.getInstruction(CrementStoreTag()) and
-      result = tco.getExpr()
-    )
-  }
-
-  /**
-   * Holds if `store` is a `StoreInstruction` that defines the temporary
-   * `IRVariable` generated as part of the translation of a ternary expression.
-   */
-  cached
-  predicate isConditionalExprTempStore(StoreInstruction store) {
-    exists(TranslatedConditionalExpr tce |
-      store = tce.getInstruction(ConditionValueFalseStoreTag())
-      or
-      store = tce.getInstruction(ConditionValueTrueStoreTag())
-    )
-  }
-
-  /** Gets the instruction that computes the address used to initialize `v`. */
-  cached
-  Instruction getInitializationTargetAddress(IRVariable v) {
-    exists(TranslatedVariableInitialization init |
-      init.getIRVariable() = v and
-      result = init.getTargetAddress()
-    )
-  }
 }
 
 class TStageInstruction = TRawInstruction or TRawUnreachedInstruction;

cpp/ql/lib/semmle/code/cpp/rangeanalysis/RangeAnalysisUtils.qll

@@ -404,7 +404,7 @@ predicate cmpWithLinearBound(
  * For example, if `t` is a signed 32-bit type then holds if `lb` is
  * `-2^31` and `ub` is `2^31 - 1`.
  */
-private predicate typeBounds0(ArithmeticType t, float lb, float ub) {
+private predicate typeBounds(ArithmeticType t, float lb, float ub) {
   exists(IntegralType integralType, float limit |
     integralType = t and limit = 2.pow(8 * integralType.getSize())
   |
@@ -423,42 +423,6 @@ private predicate typeBounds0(ArithmeticType t, float lb, float ub) {
   t instanceof FloatingPointType and lb = -(1.0 / 0.0) and ub = 1.0 / 0.0
 }
 
-/**
- * Gets the underlying type for an enumeration `e`.
- *
- * If the enumeration does not have an explicit type we approximate it using
- * the following rules:
- * - The result type is always `signed`, and
- * - if the largest value fits in an `int` the result is `int`. Otherwise, the
- * result is `long`.
- */
-private IntegralType getUnderlyingTypeForEnum(Enum e) {
-  result = e.getExplicitUnderlyingType()
-  or
-  not e.hasExplicitUnderlyingType() and
-  result.isSigned() and
-  exists(IntType intType |
-    if max(e.getAnEnumConstant().getValue().toFloat()) >= 2.pow(8 * intType.getSize() - 1)
-    then result instanceof LongType
-    else result = intType
-  )
-}
-
-/**
- * Holds if `lb` and `ub` are the lower and upper bounds of the unspecified
- * type `t`.
- *
- * For example, if `t` is a signed 32-bit type then holds if `lb` is
- * `-2^31` and `ub` is `2^31 - 1`.
- *
- * Unlike `typeBounds0`, this predicate also handles `Enum` types.
- */
-private predicate typeBounds(Type t, float lb, float ub) {
-  typeBounds0(t, lb, ub)
-  or
-  typeBounds0(getUnderlyingTypeForEnum(t), lb, ub)
-}
-
 private Type stripReference(Type t) {
   if t instanceof ReferenceType then result = t.(ReferenceType).getBaseType() else result = t
 }

cpp/ql/lib/semmle/code/cpp/rangeanalysis/SimpleRangeAnalysis.qll

@@ -512,8 +512,8 @@ private module BoundsEstimate {
    */
   float getBoundsLimit() {
     // This limit is arbitrary, but low enough that it prevents timeouts on
-    // specific observed customer databases (and in the tests).
-    result = 2.0.pow(29)
+    // specific observed customer databases (and the in the tests).
+    result = 2.0.pow(40)
   }
 
   /** Gets the maximum number of bounds possible for `t` when widening is used. */
@@ -552,47 +552,34 @@ private module BoundsEstimate {
   private float nrOfBoundsPhiGuard(RangeSsaDefinition def, StackVariable v) {
     // If we have
     //
-    //   if (x < c) { e1 } else { e2 }
-    //   e3
-    //
-    // then `{ e1 }` and `{ e2 }` are both guard phi nodes guarded by `x < c`.
-    // The range analysis propagates bounds on `x` into both branches, filtered
-    // by the condition. In this case all lower bounds flow to `{ e1 }` and only
-    // lower bounds that are smaller than `c` flow to `{ e2 }`.
-    //
-    // The largest number of bounds possible for `e3` is the number of bounds on `x` plus
-    // one. This happens when all bounds flow from `x` to `e1` to `e3` and the
-    // bound `c` can flow to `e2` to `e3`.
-    //
-    // We want to optimize our bounds estimate for `e3`, as that is the estimate
-    // that can continue propagating forward. We don't know how the existing
-    // bounds will be split between the different branches. That depends on
-    // whether the range analysis is tracking lower bounds or upper bounds, and
-    // on the meaning of the condition.
-    //
-    // As a heuristic we divide the number of bounds on `x` by 2 to "average"
-    // the effect of the condition and add 1 to account for the bound from the
-    // condition itself. This will approximate estimates inside the branches,
-    // but will give a good estimate after the branches are merged.
-    //
-    // This also handles cases such as this one
-    //
     //   if (x < c) { e1 }
-    //   e3
+    //   e2
     //
-    // where `e3` is both a guard phi node (guarded by `x < c`) and a normal
-    // phi node (control is merged after the `if` statement). Here half of the
-    // bounds flow into the branch and then to `e3` as a normal phi node and the
-    // "other" half flow from the condition to `e3` as a guard phi node.
-    exists(float varBounds |
-      // If there's different `access`es, then they refer to the same
-      // variable with the same lower bounds. Hence adding these guards makes no
-      // sense (the implementation will take the union, but they'll be removed by
-      // deduplication). Hence we use `max` as an approximation.
-      varBounds =
-        max(VariableAccess access | isGuardPhiWithBound(def, v, access) | nrOfBoundsExpr(access)) and
-      result = (varBounds + 1) / 2
-    )
+    // then `e2` is both a guard phi node (guarded by `x < c`) and a normal
+    // phi node (control is merged after the `if` statement).
+    //
+    // Assume `x` has `n` bounds. Then `n` bounds are propagated to the guard
+    // phi node `{ e1 }` and, since `{ e1 }` is input to `e2` as a normal phi
+    // node, `n` bounds are propagated to `e2`. If we also propagate the `n`
+    // bounds to `e2` as a guard phi node, then we square the number of
+    // bounds.
+    //
+    // However in practice `x < c` is going to cut down the number of bounds:
+    // The tracked bounds can't flow to both branches as that would require
+    // them to simultaneously be greater and smaller than `c`. To approximate
+    // this better, the contribution from a guard phi node that is also a
+    // normal phi node is 1.
+    exists(def.getAPhiInput(v)) and
+    isGuardPhiWithBound(def, v, _) and
+    result = 1
+    or
+    not exists(def.getAPhiInput(v)) and
+    // If there's different `access`es, then they refer to the same variable
+    // with the same lower bounds. Hence adding these guards make no sense (the
+    // implementation will take the union, but they'll be removed by
+    // deduplication). Hence we use `max` as an approximation.
+    result =
+      max(VariableAccess access | isGuardPhiWithBound(def, v, access) | nrOfBoundsExpr(access))
     or
     def.isPhiNode(v) and
     not isGuardPhiWithBound(def, v, _) and
@@ -2193,16 +2180,6 @@ module SimpleRangeAnalysisInternal {
 
   /** Gets the estimate of the number of bounds for `e`. */
   float estimateNrOfBounds(Expr e) { result = BoundsEstimate::nrOfBoundsExpr(e) }
-
-  /** Counts the numbers of lower bounds that are computed internally for `e`. */
-  float countNrOfLowerBounds(Expr e) {
-    result = strictcount(float lb | lb = getLowerBoundsImpl(e) | lb)
-  }
-
-  /** Counts the numbers of upper bounds that are computed internally for `e`. */
-  float countNrOfUpperBounds(Expr e) {
-    result = strictcount(float ub | ub = getUpperBoundsImpl(e) | ub)
-  }
 }
 
 /** Provides predicates for debugging the simple range analysis library. */
@@ -2231,7 +2208,7 @@ private module Debug {
    */
   predicate countGetLowerBoundsImpl(Expr e, int n) {
     e = getRelevantLocatable() and
-    n = SimpleRangeAnalysisInternal::countNrOfLowerBounds(e)
+    n = strictcount(float lb | lb = getLowerBoundsImpl(e) | lb)
   }
 
   float debugNrOfBounds(Expr e) {

cpp/ql/lib/semmlecode.cpp.dbscheme

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

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

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

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

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

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

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

cpp/ql/src/CHANGELOG.md

@@ -1,15 +1,3 @@
-## 1.5.13
-
-No user-facing changes.
-
-## 1.5.12
-
-No user-facing changes.
-
-## 1.5.11
-
-No user-facing changes.
-
 ## 1.5.10
 
 No user-facing changes.

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

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

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

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

cpp/ql/src/Likely Bugs/Leap Year/LeapYear.qll

@@ -308,37 +308,3 @@ private module PossibleYearArithmeticOperationCheckConfig implements DataFlow::C
 
 module PossibleYearArithmeticOperationCheckFlow =
   TaintTracking::Global<PossibleYearArithmeticOperationCheckConfig>;
-
-/**
- * A time conversion function where either
- * 1) an incorrect leap year date would result in an error that can be checked from the return value or
- * 2) an incorrect leap year date is auto corrected (no checks required)
- */
-class TimeConversionFunction extends Function {
-  boolean autoLeapYearCorrecting;
-
-  TimeConversionFunction() {
-    autoLeapYearCorrecting = false and
-    (
-      this.getName() =
-        [
-          "FileTimeToSystemTime", "SystemTimeToFileTime", "SystemTimeToTzSpecificLocalTime",
-          "SystemTimeToTzSpecificLocalTimeEx", "TzSpecificLocalTimeToSystemTime",
-          "TzSpecificLocalTimeToSystemTimeEx", "RtlLocalTimeToSystemTime",
-          "RtlTimeToSecondsSince1970", "_mkgmtime", "SetSystemTime", "VarUdateFromDate", "from_tm"
-        ]
-      or
-      // Matches all forms of GetDateFormat, e.g. GetDateFormatA/W/Ex
-      this.getName().matches("GetDateFormat%")
-    )
-    or
-    autoLeapYearCorrecting = true and
-    this.getName() =
-      ["mktime", "_mktime32", "_mktime64", "SystemTimeToVariantTime", "VariantTimeToSystemTime"]
-  }
-
-  /**
-   * Holds if the function is expected to auto convert a bad leap year date.
-   */
-  predicate isAutoLeapYearCorrecting() { autoLeapYearCorrecting = true }
-}

cpp/ql/src/Likely Bugs/Leap Year/UncheckedLeapYearAfterYearModification.ql

@@ -1,7 +1,7 @@
 /**
  * @name Year field changed using an arithmetic operation without checking for leap year
  * @description A field that represents a year is being modified by an arithmetic operation, but no proper check for leap years can be detected afterwards.
- * @kind path-problem
+ * @kind problem
  * @problem.severity warning
  * @id cpp/leap-year/unchecked-after-arithmetic-year-modification
  * @precision medium
@@ -11,844 +11,49 @@
 
 import cpp
 import LeapYear
-import semmle.code.cpp.controlflow.IRGuards
 
-/**
- * Functions whose operations should never be considered a
- * source or sink of a dangerous leap year operation.
- * The general concept is to add conversion functions
- * that convert one time type to another. Often
- * other ignorable operation heuristics will filter these,
- * but some cases, the simplest approach is to simply filter
- * the function entirely.
- * Note that flow through these functions should still be allowed
- * we just cannot start or end flow from an operation to a
- * year assignment in one of these functions.
- */
-class IgnorableFunction extends Function {
-  IgnorableFunction() {
-    // arithmetic in known time conversion functions may look like dangerous operations
-    // we assume all known time conversion functions are safe.
-    this instanceof TimeConversionFunction
-    or
-    // Helper utility in postgres with string time conversions
-    this.getName() = "DecodeISO8601Interval"
-    or
-    // helper utility for date conversions in qtbase
-    this.getName() = "adjacentDay"
-    or
-    // Windows API function that does timezone conversions
-    this.getName().matches("%SystemTimeToTzSpecificLocalTime%")
-    or
-    // Windows APIs that do time conversions
-    this.getName().matches("%localtime%\\_s%")
-    or
-    // Windows APIs that do time conversions
-    this.getName().matches("%SpecificLocalTimeToSystemTime%")
-    or
-    // postgres function for diffing timestamps, date for leap year
-    // is not applicable.
-    this.getName().toLowerCase().matches("%timestamp%age%")
-    or
-    // Reading byte streams often involves operations of some base, but that's
-    // not a real source of leap year issues.
-    this.getName().toLowerCase().matches("%read%bytes%")
-    or
-    // A postgres function for local time conversions
-    // conversion operations (from one time structure to another) are generally ignorable
-    this.getName() = "localsub"
-    or
-    // Indication of a calendar not applicable to
-    // gregorian leap year, e.g., Hijri, Persian, Hebrew
-    this.getName().toLowerCase().matches("%hijri%")
-    or
-    this.getFile().getBaseName().toLowerCase().matches("%hijri%")
-    or
-    this.getName().toLowerCase().matches("%persian%")
-    or
-    this.getFile().getBaseName().toLowerCase().matches("%persian%")
-    or
-    this.getName().toLowerCase().matches("%hebrew%")
-    or
-    this.getFile().getBaseName().toLowerCase().matches("%hebrew%")
-    or
-    // misc. from string/char converters heuristic
-    this.getName()
-        .toLowerCase()
-        .matches(["%char%to%", "%string%to%", "%from%char%", "%from%string%"])
-    or
-    // boost's gregorian.cpp has year manipulations that are checked in complex ways.
-    // ignore the entire file as a source or sink.
-    this.getFile().getAbsolutePath().toLowerCase().matches("%boost%gregorian.cpp%")
-  }
-}
-
-/**
- * The set of expressions which are ignorable; either because they seem to not be part of a year mutation,
- * or because they seem to be a conversion pattern of mapping date scalars.
- */
-abstract class IgnorableOperation extends Expr { }
-
-class IgnorableExprRem extends IgnorableOperation instanceof RemExpr { }
-
-/**
- * An operation with 10, 100, 1000, 10000 as an operand is often a sign of conversion
- * or atoi.
- */
-class IgnorableExpr10MultipleComponent extends IgnorableOperation {
-  IgnorableExpr10MultipleComponent() {
-    this.(Operation).getAnOperand().getValue().toInt() in [10, 100, 1000, 10000]
-    or
-    exists(AssignOperation a | a.getRValue() = this |
-      a.getRValue().getValue().toInt() in [10, 100, 1000, 10000]
-    )
-  }
-}
-
-/**
- * An operation involving a sub expression with char literal `48`, ignore as a likely string conversion. For example: `X - '0'`
- */
-class IgnorableExpr48Mapping extends IgnorableOperation {
-  IgnorableExpr48Mapping() {
-    this.(SubExpr).getRightOperand().getValue().toInt() = 48
-    or
-    exists(AssignSubExpr e | e.getRValue() = this | e.getRValue().getValue().toInt() = 48)
-  }
-}
-
-/**
- * A binary or arithmetic operation whereby one of the components is textual or a string.
- */
-class IgnorableCharLiteralArithmetic extends IgnorableOperation {
-  IgnorableCharLiteralArithmetic() {
-    this.(BinaryArithmeticOperation).getAnOperand() instanceof TextLiteral
-    or
-    this instanceof TextLiteral and
-    any(AssignArithmeticOperation arith).getRValue() = this
-  }
-}
-
-/**
- * Constants often used in date conversions (from one date data type to another)
- * Numerous examples exist, like 1900 or 2000 that convert years from one
- * representation to another.
- * Also '0' is sometimes observed as an atoi style conversion.
- */
-bindingset[c]
-predicate isLikelyConversionConstant(int c) {
-  exists(int i | i = c.abs() |
-    i =
-      [
-        146097, // days in 400-year Gregorian cycle
-        36524, // days in 100-year Gregorian subcycle
-        1461, // days in 4-year cycle (incl. 1 leap)
-        32044, // Fliegel-van Flandern JDN epoch shift
-        1721425, // JDN of 0001-01-01 (Gregorian)
-        1721119, // alt epoch offset
-        2400000, // MJD -> JDN conversion
-        2400001, // alt MJD -> JDN conversion
-        2141, // fixed-point month/day extraction
-        65536, // observed in some conversions
-        7834, // observed in some conversions
-        256, // observed in some conversions
-        292275056, // qdatetime.h Qt Core year range first year constant
-        292278994, // qdatetime.h Qt Core year range last year constant
-        1601, // Windows FILETIME epoch start year
-        1970, // Unix epoch start year
-        70, // Unix epoch start year short form
-        1899, // Observed in uses with 1900 to address off by one scenarios
-        1900, // Used when converting a 2 digit year
-        2000, // Used when converting a 2 digit year
-        1400, // Hijri base year, used when converting a 2 digit year
-        1980, // FAT filesystem epoch start year
-        227013, // constant observed for Hirji year conversion, and Hirji years are not applicable for gregorian leap year
-        10631, // constant observed for Hirji year conversion, and Hirji years are not applicable for gregorian leap year,
-        80, // 1980/01/01 is the start of the epoch on DOS
-        0
-      ]
-  )
-}
-
-/**
- * An `isLikelyConversionConstant` constant indicates conversion that is ignorable, e.g.,
- * julian to gregorian conversion or conversions from linux time structs
- * that start at 1900, etc.
- */
-class IgnorableConstantArithmetic extends IgnorableOperation {
-  IgnorableConstantArithmetic() {
-    exists(int i | isLikelyConversionConstant(i) |
-      this.(Operation).getAnOperand().getValue().toInt() = i
-      or
-      exists(AssignArithmeticOperation a | this = a.getRValue() |
-        a.getRValue().getValue().toInt() = i
-      )
-    )
-  }
-}
-
-// If a unary minus assume it is some sort of conversion
-class IgnorableUnaryMinus extends IgnorableOperation {
-  IgnorableUnaryMinus() {
-    this instanceof UnaryMinusExpr
-    or
-    this.(Operation).getAnOperand() instanceof UnaryMinusExpr
-  }
-}
-
-/**
- * An argument to a function is ignorable if the function that is called is an ignored function
- */
-class OperationAsArgToIgnorableFunction extends IgnorableOperation {
-  OperationAsArgToIgnorableFunction() {
-    exists(Call c |
-      c.getAnArgument().getAChild*() = this and
-      c.getTarget() instanceof IgnorableFunction
-    )
-  }
-}
-
-/**
- * A binary operation on two literals means the result is constant/known
- * and the operation is basically ignorable (it's not a real operation but
- * probably one visual simplicity what it means).
- */
-class ConstantBinaryArithmeticOperation extends IgnorableOperation, BinaryArithmeticOperation {
-  ConstantBinaryArithmeticOperation() {
-    this.getLeftOperand() instanceof Literal and
-    this.getRightOperand() instanceof Literal
-  }
-}
-
-class IgnorableBinaryBitwiseOperation extends IgnorableOperation instanceof BinaryBitwiseOperation {
-}
-
-class IgnorableUnaryBitwiseOperation extends IgnorableOperation instanceof UnaryBitwiseOperation { }
-
-class IgnorableAssignmentBitwiseOperation extends IgnorableOperation instanceof AssignBitwiseOperation
-{ }
-
-/**
- * An arithmetic operation where one of the operands is a pointer or char type, ignore it
- */
-class IgnorablePointerOrCharArithmetic extends IgnorableOperation {
-  IgnorablePointerOrCharArithmetic() {
-    this instanceof BinaryArithmeticOperation and
-    exists(Expr op | op = this.(BinaryArithmeticOperation).getAnOperand() |
-      op.getUnspecifiedType() instanceof PointerType
-      or
-      op.getUnspecifiedType() instanceof CharType
-      or
-      // Operations on calls to functions that accept char or char*
-      op.(Call).getAnArgument().getUnspecifiedType().stripType() instanceof CharType
-      or
-      // Operations on calls to functions named like "strlen", "wcslen", etc
-      // NOTE: workaround for cases where the wchar_t type is not a char, but an unsigned short
-      // unclear if there is a best way to filter cases like these out based on type info.
-      op.(Call).getTarget().getName().matches("%len%")
-    )
-    or
-    exists(AssignArithmeticOperation a | a.getRValue() = this |
-      exists(Expr op | op = a.getAnOperand() |
-        op.getUnspecifiedType() instanceof PointerType
-        or
-        op.getUnspecifiedType() instanceof CharType
-        or
-        // Operations on calls to functions that accept char or char*
-        op.(Call).getAnArgument().getUnspecifiedType().stripType() instanceof CharType
-      )
-      or
-      // Operations on calls to functions named like "strlen", "wcslen", etc
-      // for example `strlen(foo) + bar`
-      this.(BinaryArithmeticOperation).getAnOperand().(Call).getTarget().getName().matches("%len%")
-    )
-  }
-}
-
-/**
- * Holds for an expression that is an add or similar operation that could flow to a Year field.
- */
-predicate isOperationSourceCandidate(Expr e) {
-  not e instanceof IgnorableOperation and
-  exists(Function f |
-    f = e.getEnclosingFunction() and
-    not f instanceof IgnorableFunction
-  ) and
-  (
-    e instanceof SubExpr
-    or
-    e instanceof AddExpr
-    or
-    e instanceof CrementOperation
-    or
-    e instanceof AssignSubExpr
-    or
-    e instanceof AssignAddExpr
-  )
-}
-
-/**
- * A data flow that tracks an ignorable operation (such as a bitwise operation) to an operation source, so we may disqualify it.
- */
-module IgnorableOperationToOperationSourceCandidateConfig implements DataFlow::ConfigSig {
-  predicate isSource(DataFlow::Node n) { n.asExpr() instanceof IgnorableOperation }
-
-  predicate isSink(DataFlow::Node n) { isOperationSourceCandidate(n.asExpr()) }
-
-  // looking for sources and sinks in the same function
-  DataFlow::FlowFeature getAFeature() {
-    result instanceof DataFlow::FeatureEqualSourceSinkCallContext
-  }
-}
-
-module IgnorableOperationToOperationSourceCandidateFlow =
-  TaintTracking::Global<IgnorableOperationToOperationSourceCandidateConfig>;
-
-/**
- * The set of all expressions which is a candidate expression and also does not flow from to to some ignorable expression (eg. bitwise op)
- * ```
- * a = something <<< 2;
- * myDate.year = a + 1;        // invalid
- * ...
- * a = someDate.year + 1;
- * myDate.year = a;            // valid
- * ```
- */
-class OperationSource extends Expr {
-  OperationSource() {
-    isOperationSourceCandidate(this) and
-    // If the candidate came from an ignorable operation, ignore the candidate
-    // NOTE: we cannot easily flow the candidate to an ignorable operation as that can
-    // be tricky in practice, e.g., a mod operation on a year would be part of a leap year check
-    // but a mod operation ending in a year is more indicative of something to ignore (a conversion)
-    not exists(IgnorableOperationToOperationSourceCandidateFlow::PathNode sink |
-      sink.getNode().asExpr() = this and
-      sink.isSink()
-    )
-  }
-}
-
-class YearFieldAssignmentNode extends DataFlow::Node {
-  YearFieldAccess access;
-
-  YearFieldAssignmentNode() {
-    exists(Function f |
-      f = this.getEnclosingCallable().getUnderlyingCallable() and not f instanceof IgnorableFunction
-    ) and
-    (
-      this.asDefinition().(Assignment).getLValue() = access
-      or
-      this.asDefinition().(CrementOperation).getOperand() = access
-      or
-      exists(Call c | c.getAnArgument() = access and this.asDefiningArgument() = access)
-      or
-      exists(Call c, AddressOfExpr aoe |
-        c.getAnArgument() = aoe and
-        aoe.getOperand() = access and
-        this.asDefiningArgument() = aoe
-      )
-    )
-  }
-
-  YearFieldAccess getYearFieldAccess() { result = access }
-}
-
-/**
- * A DataFlow configuration for identifying flows from an identified source
- * to the Year field of a date object.
- */
-module OperationToYearAssignmentConfig implements DataFlow::ConfigSig {
-  predicate isSource(DataFlow::Node n) { n.asExpr() instanceof OperationSource }
-
-  predicate isSink(DataFlow::Node n) {
-    n instanceof YearFieldAssignmentNode and
-    not isYearModifiedWithCheck(n) and
-    not isControlledByMonthEqualityCheckNonFebruary(n.asExpr())
-  }
-
-  predicate isBarrier(DataFlow::Node n) {
-    exists(ArrayExpr arr | arr.getArrayOffset() = n.asExpr())
-    or
-    n.getType().getUnspecifiedType() instanceof PointerType
-    or
-    n.getType().getUnspecifiedType() instanceof CharType
-    or
-    // If a type resembles "string" ignore flow (likely string conversion, currently ignored)
-    n.getType().getUnspecifiedType().stripType().getName().toLowerCase().matches("%string%")
-    or
-    n.asExpr() instanceof IgnorableOperation
-    or
-    // Flowing into variables that indicate likely non-gregorian years are barriers
-    // e.g., names similar to hijri, persian, lunar, chinese, hebrew, etc.
-    exists(Variable v |
-      v.getName()
-          .toLowerCase()
-          .matches(["%hijri%", "%persian%", "%lunar%", "%chinese%", "%hebrew%"]) and
-      v.getAnAccess() = [n.asIndirectExpr(), n.asExpr()]
-    )
-    or
-    isLeapYearCheckSink(n)
-    or
-    // this is a bit of a hack to address cases where a year is normalized and checked, but the
-    // normalized year is never itself assigned to the final year struct
-    //    isLeapYear(getCivilYear(year))
-    //    struct.year = year
-    // This is assuming a user would have done this all on one line though.
-    // setting a variable for the conversion and passing that separately would be more difficult to track
-    // considering this approach good enough for current observed false positives
-    exists(Expr arg |
-      isLeapYearCheckCall(_, arg) and arg.getAChild*() = [n.asExpr(), n.asIndirectExpr()]
-    )
-    or
-    // If as the flow progresses, the value holding a dangerous operation result
-    // is apparently being passed by address to some function, it is more than likely
-    // intended to be modified, and therefore, the definition is killed.
-    exists(Call c | c.getAnArgument().(AddressOfExpr).getAnOperand() = n.asIndirectExpr())
-  }
-
-  /** Block flow out of an operation source to get the "closest" operation to the sink */
-  predicate isBarrierIn(DataFlow::Node n) { isSource(n) }
-
-  predicate isBarrierOut(DataFlow::Node n) { isSink(n) }
-}
-
-module OperationToYearAssignmentFlow = TaintTracking::Global<OperationToYearAssignmentConfig>;
-
-predicate isLeapYearCheckSink(DataFlow::Node sink) {
-  exists(LeapYearGuardCondition lgc |
-    lgc.checkedYearAccess() = [sink.asExpr(), sink.asIndirectExpr()]
-  )
-  or
-  isLeapYearCheckCall(_, [sink.asExpr(), sink.asIndirectExpr()])
-}
-
-predicate yearAssignmentToCheckCommonSteps(DataFlow::Node node1, DataFlow::Node node2) {
-  // flow from a YearFieldAccess to the qualifier
-  node2.asExpr() = node1.asExpr().(YearFieldAccess).getQualifier*()
-  or
-  // getting the 'access' can be tricky at definitions (assignments especially)
-  // as dataflow uses asDefinition not asExpr.
-  // the YearFieldAssignmentNode holds the access in these cases
-  node1.(YearFieldAssignmentNode).getYearFieldAccess().getQualifier() = node2.asExpr()
-  or
-  // flow from a year access qualifier to a year field
-  exists(YearFieldAccess yfa | node2.asExpr() = yfa and node1.asExpr() = yfa.getQualifier())
-  or
-  node1.(YearFieldAssignmentNode).getYearFieldAccess().getQualifier() = node2.asExpr()
-  or
-  // Pass through any intermediate struct
-  exists(Assignment a |
-    a.getRValue() = node1.asExpr() and
-    node2.asExpr() = a.getLValue().(YearFieldAccess).getQualifier*()
-  )
-  or
-  // in cases of t.year = x and the value of x is checked, but the year t.year isn't directly checked
-  // flow from a year assignment node to an RHS if it is an assignment
-  // e.g.,
-  //    t.year = x;
-  //    isLeapYear(x);
-  //    --> at this point there is no flow of t.year to a check, but only its raw value
-  // To detect the flow of 'x' to the isLeapYear check,
-  // flow from t.year to 'x' (at assignment, t.year = x, flow to the RHS to track use-use flow of x)
-  exists(YearFieldAssignmentNode yfan |
-    node1 = yfan and
-    node2.asExpr() = yfan.asDefinition().(Assignment).getRValue()
-  )
-}
-
-/**
- * A flow configuration from a Year field access to some Leap year check or guard
- */
-module YearAssignmentToLeapYearCheckConfig implements DataFlow::ConfigSig {
-  predicate isSource(DataFlow::Node source) { source instanceof YearFieldAssignmentNode }
-
-  predicate isSink(DataFlow::Node sink) { isLeapYearCheckSink(sink) }
-
-  predicate isAdditionalFlowStep(DataFlow::Node node1, DataFlow::Node node2) {
-    yearAssignmentToCheckCommonSteps(node1, node2)
-  }
-
-  /**
-   * Enforcing the check must occur in the same call context as the source,
-   * i.e., do not return from the source function and check in a caller.
-   */
-  DataFlow::FlowFeature getAFeature() { result instanceof DataFlow::FeatureHasSourceCallContext }
-}
-
-module YearAssignmentToLeapYearCheckFlow =
-  TaintTracking::Global<YearAssignmentToLeapYearCheckConfig>;
-
-/** Does there exist a flow from the given YearFieldAccess to a Leap Year check or guard? */
-predicate isYearModifiedWithCheck(YearFieldAssignmentNode n) {
-  exists(YearAssignmentToLeapYearCheckFlow::PathNode src |
-    src.isSource() and
-    src.getNode() = n
-  )
-  or
-  // If the time flows to a time conversion whose value/result is checked,
-  // assume the leap year is being handled.
-  exists(YearAssignmentToCheckedTimeConversionFlow::PathNode timeQualSrc |
-    timeQualSrc.isSource() and
-    timeQualSrc.getNode() = n
-  )
-}
-
-/**
- * An expression which checks the value of a Month field `a->month == 1`.
- */
-class MonthEqualityCheck extends EqualityOperation {
-  MonthEqualityCheck() { this.getAnOperand() instanceof MonthFieldAccess }
-
-  Expr getExprCompared() {
-    exists(Expr e |
-      e = this.getAnOperand() and
-      not e instanceof MonthFieldAccess and
-      result = e
-    )
-  }
-}
-
-final class FinalMonthEqualityCheck = MonthEqualityCheck;
-
-class MonthEqualityCheckGuard extends GuardCondition, FinalMonthEqualityCheck { }
-
-/**
- * Verifies if the expression is guarded by a check on the Month property of a date struct, that is NOT February.
- */
-bindingset[e]
-pragma[inline_late]
-predicate isControlledByMonthEqualityCheckNonFebruary(Expr e) {
-  exists(MonthEqualityCheckGuard monthGuard, Expr compared |
-    monthGuard.controls(e.getBasicBlock(), true) and
-    compared = monthGuard.getExprCompared() and
-    not compared.getValue().toInt() = 2
-  )
-}
-
-/**
- * Flow from a year field access to a time conversion function
- * that auto converts feb29 in non-leap year, or through a conversion function that doesn't
- * auto convert to a sanity check guard of the result for error conditions.
- */
-module YearAssignmentToCheckedTimeConversionConfig implements DataFlow::StateConfigSig {
-  // Flow state tracks if flow goes through a known time conversion function
-  // see `TimeConversionFunction`.
-  // A valid check with a time conversion function is either the case:
-  // 1) the year flows into a time conversion function, and the time conversion function's result is checked or
-  // 2) the year flows into a time conversion function that auto corrects for leap year, so no check is necessary.
-  class FlowState = boolean;
-
-  predicate isSource(DataFlow::Node source, FlowState state) {
-    source instanceof YearFieldAssignmentNode and
-    state = false
-  }
-
-  predicate isSink(DataFlow::Node sink, FlowState state) {
-    // Case 1: Flow through a time conversion function that requires a check,
-    // and we have arrived at a guard, implying the result was checked for possible error, including leap year error.
-    // state = true indicates the flow went through a time conversion function
-    state = true and
-    (
-      exists(IfStmt ifs | ifs.getCondition().getAChild*() = [sink.asExpr(), sink.asIndirectExpr()])
-      or
-      exists(ConditionalExpr ce |
-        ce.getCondition().getAChild*() = [sink.asExpr(), sink.asIndirectExpr()]
-      )
-      or
-      exists(Loop l | l.getCondition().getAChild*() = [sink.asExpr(), sink.asIndirectExpr()])
-    )
-    or
-    // Case 2: Flow through a time conversion function that auto corrects for leap year, so no check is necessary.
-    // state true or false, as flowing through a time conversion function is not necessary in this instance.
-    state in [true, false] and
-    exists(Call c, TimeConversionFunction f |
-      f.isAutoLeapYearCorrecting() and
-      c.getTarget() = f and
-      c.getAnArgument().getAChild*() = [sink.asExpr(), sink.asIndirectExpr()]
-    )
-  }
-
-  predicate isAdditionalFlowStep(
-    DataFlow::Node node1, FlowState state1, DataFlow::Node node2, FlowState state2
-  ) {
-    state1 in [true, false] and
-    state2 = true and
-    exists(Call c |
-      c.getTarget() instanceof TimeConversionFunction and
-      c.getAnArgument().getAChild*() = [node1.asExpr(), node1.asIndirectExpr()] and
-      node2.asExpr() = c
-    )
-  }
-
-  predicate isAdditionalFlowStep(DataFlow::Node node1, DataFlow::Node node2) {
-    yearAssignmentToCheckCommonSteps(node1, node2)
-  }
-
-  DataFlow::FlowFeature getAFeature() { result instanceof DataFlow::FeatureHasSourceCallContext }
-}
-
-module YearAssignmentToCheckedTimeConversionFlow =
-  DataFlow::GlobalWithState<YearAssignmentToCheckedTimeConversionConfig>;
-
-/**
- * Finds flow from a parameter of a function to a leap year check.
- * This is necessary to handle for scenarios like this:
- *
- *    year = DANGEROUS_OP // source
- *    isLeap = isLeapYear(year);
- *    // logic based on isLeap
- *    struct.year = year; // sink
- *
- * In this case, we may flow a dangerous op to a year assignment, failing
- * to barrier the flow through a leap year check, as the leap year check
- * is nested, and dataflow does not progress down into the check and out.
- * Instead, the point of this flow is to detect isLeapYear's argument
- * is checked for leap year, making the isLeapYear call a barrier for
- * the dangerous flow if we flow through the parameter identified to
- * be checked.
- */
-module ParameterToLeapYearCheckConfig implements DataFlow::ConfigSig {
-  predicate isSource(DataFlow::Node source) { exists(source.asParameter()) }
-
-  predicate isSink(DataFlow::Node sink) {
-    exists(LeapYearGuardCondition lgc |
-      lgc.checkedYearAccess() = [sink.asExpr(), sink.asIndirectExpr()]
-    )
-  }
-
-  predicate isAdditionalFlowStep(DataFlow::Node node1, DataFlow::Node node2) {
-    // flow from a YearFieldAccess to the qualifier
-    node2.asExpr() = node1.asExpr().(YearFieldAccess).getQualifier*()
-    or
-    // flow from a year access qualifier to a year field
-    exists(YearFieldAccess yfa | node2.asExpr() = yfa and node1.asExpr() = yfa.getQualifier())
-  }
-
-  /**
-   * Enforcing the check must occur in the same call context as the source,
-   * i.e., do not return from the source function and check in a caller.
-   */
-  DataFlow::FlowFeature getAFeature() { result instanceof DataFlow::FeatureHasSourceCallContext }
-}
-
-// NOTE: I do not believe taint flow is necessary here as we should
-// be flowing directyly from some parameter to a leap year check.
-module ParameterToLeapYearCheckFlow = DataFlow::Global<ParameterToLeapYearCheckConfig>;
-
-predicate isLeapYearCheckCall(Call c, Expr arg) {
-  exists(ParameterToLeapYearCheckFlow::PathNode src, Function f, int i |
-    src.isSource() and
-    f.getParameter(i) = src.getNode().asParameter() and
-    c.getTarget() = f and
-    c.getArgument(i) = arg
-  )
-}
-
-class LeapYearGuardCondition extends GuardCondition {
-  Expr yearSinkDiv4;
-  Expr yearSinkDiv100;
-  Expr yearSinkDiv400;
-
-  LeapYearGuardCondition() {
-    exists(
-      LogicalAndExpr andExpr, LogicalOrExpr orExpr, GuardCondition div4Check,
-      GuardCondition div100Check, GuardCondition div400Check, GuardValue gv
-    |
-      // canonical case:
-      // form: `(year % 4 == 0) && (year % 100 != 0 || year % 400 == 0)`
-      //    `!((year % 4 == 0) && (year % 100 != 0 || year % 400 == 0))`
-      //    `!(year % 4) && (year % 100 || !(year % 400))`
-      // Also accepting `((year & 3) == 0) && (year % 100 != 0 || year % 400 == 0)`
-      // and `(year % 4 == 0) && (year % 100 > 0 || year % 400 == 0)`
-      this = andExpr and
-      andExpr.hasOperands(div4Check, orExpr) and
-      orExpr.hasOperands(div100Check, div400Check) and
-      (
-        // year % 4 == 0
-        exists(RemExpr e |
-          div4Check.comparesEq(e, 0, true, gv) and
-          e.getRightOperand().getValue().toInt() = 4 and
-          yearSinkDiv4 = e.getLeftOperand()
-        )
-        or
-        // year & 3 == 0
-        exists(BitwiseAndExpr e |
-          div4Check.comparesEq(e, 0, true, gv) and
-          e.getRightOperand().getValue().toInt() = 3 and
-          yearSinkDiv4 = e.getLeftOperand()
-        )
-      ) and
-      exists(RemExpr e |
-        // year % 100 != 0 or year % 100 > 0
-        (
-          div100Check.comparesEq(e, 0, false, gv) or
-          div100Check.comparesLt(e, 1, false, gv)
-        ) and
-        e.getRightOperand().getValue().toInt() = 100 and
-        yearSinkDiv100 = e.getLeftOperand()
-      ) and
-      // year % 400 == 0
-      exists(RemExpr e |
-        div400Check.comparesEq(e, 0, true, gv) and
-        e.getRightOperand().getValue().toInt() = 400 and
-        yearSinkDiv400 = e.getLeftOperand()
-      )
-      or
-      // Inverted logic case:
-      //  `year % 4 != 0 || (year % 100 == 0 && year % 400 != 0)`
-      // or `year & 3 != 0 || (year % 100 == 0 && year % 400 != 0)`
-      // also accepting `year % 4 > 0 || (year % 100 == 0 && year % 400 > 0)`
-      this = orExpr and
-      orExpr.hasOperands(div4Check, andExpr) and
-      andExpr.hasOperands(div100Check, div400Check) and
-      (
-        // year % 4 != 0 or year % 4 > 0
-        exists(RemExpr e |
-          (
-            div4Check.comparesEq(e, 0, false, gv)
-            or
-            div4Check.comparesLt(e, 1, false, gv)
-          ) and
-          e.getRightOperand().getValue().toInt() = 4 and
-          yearSinkDiv4 = e.getLeftOperand()
-        )
-        or
-        // year & 3 != 0
-        exists(BitwiseAndExpr e |
-          div4Check.comparesEq(e, 0, false, gv) and
-          e.getRightOperand().getValue().toInt() = 3 and
-          yearSinkDiv4 = e.getLeftOperand()
-        )
-      ) and
-      // year % 100 == 0
-      exists(RemExpr e |
-        div100Check.comparesEq(e, 0, true, gv) and
-        e.getRightOperand().getValue().toInt() = 100 and
-        yearSinkDiv100 = e.getLeftOperand()
-      ) and
-      // year % 400 != 0 or year % 400 > 0
-      exists(RemExpr e |
-        (
-          div400Check.comparesEq(e, 0, false, gv)
-          or
-          div400Check.comparesLt(e, 1, false, gv)
-        ) and
-        e.getRightOperand().getValue().toInt() = 400 and
-        yearSinkDiv400 = e.getLeftOperand()
-      )
-    )
-  }
-
-  Expr getYearSinkDiv4() { result = yearSinkDiv4 }
-
-  Expr getYearSinkDiv100() { result = yearSinkDiv100 }
-
-  Expr getYearSinkDiv400() { result = yearSinkDiv400 }
-
-  /**
-   * Gets the variable access that is used in all 3 components of the leap year check
-   * e.g., see getYearSinkDiv4/100/400..
-   * If a field access is used, the qualifier and the field access are both returned
-   * in checked condition.
-   * NOTE: if the year is not checked using the same access in all 3 components, no result is returned.
-   * The typical case observed is a consistent variable access is used. If not, this may indicate a bug.
-   * We could check more accurately with a dataflow analysis, but this is likely sufficient for now.
-   */
-  VariableAccess checkedYearAccess() {
-    exists(Variable var |
-      (
-        this.getYearSinkDiv4().getAChild*() = var.getAnAccess() and
-        this.getYearSinkDiv100().getAChild*() = var.getAnAccess() and
-        this.getYearSinkDiv400().getAChild*() = var.getAnAccess() and
-        result = var.getAnAccess() and
-        (
-          result = this.getYearSinkDiv4().getAChild*() or
-          result = this.getYearSinkDiv100().getAChild*() or
-          result = this.getYearSinkDiv400().getAChild*()
-        )
-      )
-    )
-  }
-}
-
-/**
- * A difficult case to detect is if a year modification is tied to a month or day modification
- * and the month or day is safe for leap year.
- *    e.g.,
- *          year++;
- *          month = 1;
- *          // alternative: day = 15;
- *        ... values eventually used in the same time struct
- * If this is even more challenging if the struct the values end up in are not
- * local (set inter-procedurally).
- * This configuration looks for constants 1-31 flowing to a month or day assignment.
- * It is assumed a user of this flow will check if the month/day source and month/day sink
- * are in the same basic blocks as a year modification source and a year modification sink.
- * It is also assumed a user will check if the constant source is a value that is ignorable
- * e.g., if it is 2 and the sink is a month assignment, then it isn't ignorable or
- * if the value is < 27 and is a day assignment, it is likely ignorable
- *
- * Obviously this does not handle all conditions (e.g., the month set in another block).
- * It is meant to capture the most common cases of false positives.
- */
-module CandidateConstantToDayOrMonthAssignmentConfig implements DataFlow::ConfigSig {
-  predicate isSource(DataFlow::Node source) {
-    source.asExpr().getValue().toInt() in [1 .. 31] and
-    (
-      exists(Assignment a | a.getRValue() = source.asExpr())
-      or
-      exists(Call c | c.getAnArgument() = source.asExpr())
-    )
-  }
-
-  predicate isSink(DataFlow::Node sink) {
-    exists(Assignment a |
-      (a.getLValue() instanceof MonthFieldAccess or a.getLValue() instanceof DayFieldAccess) and
-      a.getRValue() = sink.asExpr()
-    )
-  }
-}
-
-// NOTE: only data flow here (no taint tracking) as we want the exact
-// constant flowing to the month assignment
-module CandidateConstantToDayOrMonthAssignmentFlow =
-  DataFlow::Global<CandidateConstantToDayOrMonthAssignmentConfig>;
-
-/**
- * Holds if value the assignment `a` resolves to (`dayOrMonthValSrcExpr`) doesn't represent February,
- * and/or if it represents a day, is a 'safe' day (meaning the 27th or prior).
- */
-bindingset[dayOrMonthValSrcExpr]
-predicate isSafeValueForAssignmentOfMonthOrDayValue(Assignment a, Expr dayOrMonthValSrcExpr) {
-  a.getLValue() instanceof MonthFieldAccess and
-  dayOrMonthValSrcExpr.getValue().toInt() != 2
-  or
-  a.getLValue() instanceof DayFieldAccess and
-  dayOrMonthValSrcExpr.getValue().toInt() <= 27
-}
-
-import OperationToYearAssignmentFlow::PathGraph
-
-from OperationToYearAssignmentFlow::PathNode src, OperationToYearAssignmentFlow::PathNode sink
+from Variable var, LeapYearFieldAccess yfa
 where
-  OperationToYearAssignmentFlow::flowPath(src, sink) and
-  // Check if a month is set in the same block as the year operation source
-  // and the month value would indicate its set to any other month than february.
-  // Finds if the source year node is in the same block as a source month block
-  // and if the same for the sinks.
-  not exists(DataFlow::Node dayOrMonthValSrc, DataFlow::Node dayOrMonthValSink, Assignment a |
-    CandidateConstantToDayOrMonthAssignmentFlow::flow(dayOrMonthValSrc, dayOrMonthValSink) and
-    a.getRValue() = dayOrMonthValSink.asExpr() and
-    dayOrMonthValSink.getBasicBlock() = sink.getNode().getBasicBlock() and
-    exists(IRBlock dayOrMonthValBB |
-      dayOrMonthValBB = dayOrMonthValSrc.getBasicBlock() and
-      // The source of the day is set in the same block as the source for the year
-      // or the source for the day is set in the same block as the sink for the year
-      dayOrMonthValBB in [
-          src.getNode().getBasicBlock(),
-          sink.getNode().getBasicBlock()
-        ]
-    ) and
-    isSafeValueForAssignmentOfMonthOrDayValue(a, dayOrMonthValSrc.asExpr())
+  exists(VariableAccess va |
+    yfa.getQualifier() = va and
+    var.getAnAccess() = va and
+    // The year is modified with an arithmetic operation. Avoid values that are likely false positives
+    yfa.isModifiedByArithmeticOperationNotForNormalization() and
+    // Avoid false positives
+    not (
+      // If there is a local check for leap year after the modification
+      exists(LeapYearFieldAccess yfacheck |
+        yfacheck.getQualifier() = var.getAnAccess() and
+        yfacheck.isUsedInCorrectLeapYearCheck() and
+        yfacheck.getBasicBlock() = yfa.getBasicBlock().getASuccessor*()
+      )
+      or
+      // If there is a data flow from the variable that was modified to a function that seems to check for leap year
+      exists(VariableAccess source, ChecksForLeapYearFunctionCall fc |
+        source = var.getAnAccess() and
+        LeapYearCheckFlow::flow(DataFlow::exprNode(source), DataFlow::exprNode(fc.getAnArgument()))
+      )
+      or
+      // If there is a data flow from the field that was modified to a function that seems to check for leap year
+      exists(VariableAccess vacheck, YearFieldAccess yfacheck, ChecksForLeapYearFunctionCall fc |
+        vacheck = var.getAnAccess() and
+        yfacheck.getQualifier() = vacheck and
+        LeapYearCheckFlow::flow(DataFlow::exprNode(yfacheck), DataFlow::exprNode(fc.getAnArgument()))
+      )
+      or
+      // If there is a successor or predecessor that sets the month = 1
+      exists(MonthFieldAccess mfa, AssignExpr ae |
+        mfa.getQualifier() = var.getAnAccess() and
+        mfa.isModified() and
+        (
+          mfa.getBasicBlock() = yfa.getBasicBlock().getASuccessor*() or
+          yfa.getBasicBlock() = mfa.getBasicBlock().getASuccessor+()
+        ) and
+        ae = mfa.getEnclosingElement() and
+        ae.getAnOperand().getValue().toInt() = 1
+      )
+    )
   )
-select sink, src, sink,
-  "Year field has been modified, but no appropriate check for LeapYear was found."
+select yfa,
+  "Field $@ on variable $@ has been modified, but no appropriate check for LeapYear was found.",
+  yfa.getTarget(), yfa.getTarget().toString(), var, var.toString()

cpp/ql/src/Likely Bugs/Leap Year/UncheckedReturnValueForTimeFunctions.ql

@@ -44,9 +44,23 @@ class SafeTimeGatheringFunction extends Function {
   }
 }
 
+/**
+ * This list of APIs should check for the return value to detect problems during the conversion.
+ */
+class TimeConversionFunction extends Function {
+  TimeConversionFunction() {
+    this.getQualifiedName() =
+      [
+        "FileTimeToSystemTime", "SystemTimeToFileTime", "SystemTimeToTzSpecificLocalTime",
+        "SystemTimeToTzSpecificLocalTimeEx", "TzSpecificLocalTimeToSystemTime",
+        "TzSpecificLocalTimeToSystemTimeEx", "RtlLocalTimeToSystemTime",
+        "RtlTimeToSecondsSince1970", "_mkgmtime"
+      ]
+  }
+}
+
 from FunctionCall fcall, TimeConversionFunction trf, Variable var
 where
-  not trf.isAutoLeapYearCorrecting() and
   fcall = trf.getACallToThisFunction() and
   fcall instanceof ExprInVoidContext and
   var.getUnderlyingType() instanceof UnpackedTimeType and

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

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

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

@@ -16,15 +16,17 @@
 import cpp
 import semmle.code.cpp.ir.IR
 import semmle.code.cpp.ir.dataflow.MustFlow
-import ReturnStackAllocatedMemory::PathGraph
+import PathGraph
 
 /** Holds if `f` has a name that we interpret as evidence of intentionally returning the value of the stack pointer. */
 predicate intentionallyReturnsStackPointer(Function f) {
   f.getName().toLowerCase().matches(["%stack%", "%sp%"])
 }
 
-module ReturnStackAllocatedMemoryConfig implements MustFlow::ConfigSig {
-  predicate isSource(Instruction source) {
+class ReturnStackAllocatedMemoryConfig extends MustFlowConfiguration {
+  ReturnStackAllocatedMemoryConfig() { this = "ReturnStackAllocatedMemoryConfig" }
+
+  override predicate isSource(Instruction source) {
     exists(Function func |
       // Rule out FPs caused by extraction errors.
       not func.hasErrors() and
@@ -48,7 +50,7 @@ module ReturnStackAllocatedMemoryConfig implements MustFlow::ConfigSig {
     )
   }
 
-  predicate isSink(Operand sink) {
+  override predicate isSink(Operand sink) {
     // Holds if `sink` is a node that represents the `StoreInstruction` that is subsequently used in
     // a `ReturnValueInstruction`.
     // We use the `StoreInstruction` instead of the instruction that defines the
@@ -70,7 +72,7 @@ module ReturnStackAllocatedMemoryConfig implements MustFlow::ConfigSig {
   //   int* px = id(&x);
   //  }
   //   ```
-  predicate allowInterproceduralFlow() { none() }
+  override predicate allowInterproceduralFlow() { none() }
 
   /**
    * This configuration intentionally conflates addresses of fields and their object, and pointer offsets
@@ -85,22 +87,20 @@ module ReturnStackAllocatedMemoryConfig implements MustFlow::ConfigSig {
    * }
    * ```
    */
-  predicate isAdditionalFlowStep(Operand node1, Instruction node2) {
+  override predicate isAdditionalFlowStep(Operand node1, Instruction node2) {
     node2.(FieldAddressInstruction).getObjectAddressOperand() = node1
     or
     node2.(PointerOffsetInstruction).getLeftOperand() = node1
   }
 
-  predicate isBarrier(Instruction n) { n.getResultType() instanceof ErroneousType }
+  override predicate isBarrier(Instruction n) { n.getResultType() instanceof ErroneousType }
 }
 
-module ReturnStackAllocatedMemory = MustFlow::Global<ReturnStackAllocatedMemoryConfig>;
-
 from
-  ReturnStackAllocatedMemory::PathNode source, ReturnStackAllocatedMemory::PathNode sink,
-  Instruction instr
+  MustFlowPathNode source, MustFlowPathNode sink, Instruction instr,
+  ReturnStackAllocatedMemoryConfig conf
 where
-  ReturnStackAllocatedMemory::flowPath(pragma[only_bind_into](source), pragma[only_bind_into](sink)) and
+  conf.hasFlowPath(pragma[only_bind_into](source), pragma[only_bind_into](sink)) and
   source.getInstruction() = instr
 select sink.getInstruction(), source, sink, "May return stack-allocated memory from $@.",
   instr.getAst(), instr.getAst().toString()

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

@@ -15,7 +15,7 @@
 import cpp
 import semmle.code.cpp.ir.IR
 import semmle.code.cpp.ir.dataflow.MustFlow
-import UninitializedLocal::PathGraph
+import PathGraph
 
 /**
  * Auxiliary predicate: Types that don't require initialization
@@ -70,26 +70,25 @@ predicate isSinkImpl(Instruction sink, VariableAccess va) {
   )
 }
 
-module UninitializedLocalConfig implements MustFlow::ConfigSig {
-  predicate isSource(Instruction source) {
+class MustFlow extends MustFlowConfiguration {
+  MustFlow() { this = "MustFlow" }
+
+  override predicate isSource(Instruction source) {
     source instanceof UninitializedInstruction and
     exists(Type t | t = source.getResultType() | not allocatedType(t))
   }
 
-  predicate isSink(Operand sink) { isSinkImpl(sink.getDef(), _) }
+  override predicate isSink(Operand sink) { isSinkImpl(sink.getDef(), _) }
 
-  predicate allowInterproceduralFlow() { none() }
+  override predicate allowInterproceduralFlow() { none() }
 
-  predicate isBarrier(Instruction instr) { instr instanceof ChiInstruction }
+  override predicate isBarrier(Instruction instr) { instr instanceof ChiInstruction }
 }
 
-module UninitializedLocal = MustFlow::Global<UninitializedLocalConfig>;
-
 from
-  VariableAccess va, LocalVariable v, UninitializedLocal::PathNode source,
-  UninitializedLocal::PathNode sink
+  VariableAccess va, LocalVariable v, MustFlow conf, MustFlowPathNode source, MustFlowPathNode sink
 where
-  UninitializedLocal::flowPath(source, sink) and
+  conf.hasFlowPath(source, sink) and
   isSinkImpl(sink.getInstruction(), va) and
   v = va.getTarget()
 select va, source, sink, "The variable $@ may not be initialized at this access.", v, v.getName()

cpp/ql/src/Likely Bugs/OO/UnsafeUseOfThis.ql

@@ -17,16 +17,16 @@
 import cpp
 import semmle.code.cpp.ir.IR
 import semmle.code.cpp.ir.dataflow.MustFlow
-import UnsafeUseOfThis::PathGraph
+import PathGraph
 
-module UnsafeUseOfThisConfig implements MustFlow::ConfigSig {
-  predicate isSource(Instruction source) { isSource(source, _, _) }
+class UnsafeUseOfThisConfig extends MustFlowConfiguration {
+  UnsafeUseOfThisConfig() { this = "UnsafeUseOfThisConfig" }
 
-  predicate isSink(Operand sink) { isSink(sink, _) }
+  override predicate isSource(Instruction source) { isSource(source, _, _) }
+
+  override predicate isSink(Operand sink) { isSink(sink, _) }
 }
 
-module UnsafeUseOfThis = MustFlow::Global<UnsafeUseOfThisConfig>;
-
 /** Holds if `sink` is a `this` pointer used by the call instruction `call`. */
 predicate isSink(Operand sink, CallInstruction call) {
   exists(PureVirtualFunction func |
@@ -66,17 +66,19 @@ predicate isSource(InitializeParameterInstruction source, string msg, Class c) {
  * - `msg` is a string describing whether `source` is from a constructor or destructor.
  */
 predicate flows(
-  UnsafeUseOfThis::PathNode source, string msg, Class sourceClass, UnsafeUseOfThis::PathNode sink,
+  MustFlowPathNode source, string msg, Class sourceClass, MustFlowPathNode sink,
   CallInstruction call
 ) {
-  UnsafeUseOfThis::flowPath(source, sink) and
-  isSource(source.getInstruction(), msg, sourceClass) and
-  isSink(sink.getInstruction().getAUse(), call)
+  exists(UnsafeUseOfThisConfig conf |
+    conf.hasFlowPath(source, sink) and
+    isSource(source.getInstruction(), msg, sourceClass) and
+    isSink(sink.getInstruction().getAUse(), call)
+  )
 }
 
 from
-  UnsafeUseOfThis::PathNode source, UnsafeUseOfThis::PathNode sink, CallInstruction call,
-  string msg, Class sourceClass
+  MustFlowPathNode source, MustFlowPathNode sink, CallInstruction call, string msg,
+  Class sourceClass
 where
   flows(source, msg, sourceClass, sink, call) and
   // Only raise an alert if there is no override of the pure virtual function in any base class.

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

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

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

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

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

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

cpp/ql/src/Telemetry/DatabaseQuality.qll

@@ -1,25 +0,0 @@
-import cpp
-import codeql.util.ReportStats
-
-module CallTargetStats implements StatsSig {
-  private class RelevantCall extends Call {
-    RelevantCall() { this.getFile() = any(File f | f.fromSource() and exists(f.getRelativePath())) }
-  }
-
-  // We assume that calls with an implicit target are calls that could not be
-  // resolved. This is accurate in the vast majority of cases, but is inaccurate
-  // for calls that deliberately rely on implicitly declared functions.
-  private predicate hasImplicitTarget(RelevantCall call) {
-    call.getTarget().getADeclarationEntry().isImplicit()
-  }
-
-  int getNumberOfOk() { result = count(RelevantCall call | not hasImplicitTarget(call)) }
-
-  int getNumberOfNotOk() { result = count(RelevantCall call | hasImplicitTarget(call)) }
-
-  string getOkText() { result = "calls with call target" }
-
-  string getNotOkText() { result = "calls with missing call target" }
-}
-
-module CallTargetStatsReport = ReportStats<CallTargetStats>;

cpp/ql/src/Telemetry/ExtractorInformation.ql

@@ -1,25 +0,0 @@
-/**
- * @name C/C++ extraction information
- * @description Information about the extraction for a C/C++ database
- * @kind metric
- * @tags summary telemetry
- * @id cpp/telemetry/extraction-information
- */
-
-import cpp
-import DatabaseQuality
-
-from string key, float value
-where
-  (
-    CallTargetStatsReport::numberOfOk(key, value) or
-    CallTargetStatsReport::numberOfNotOk(key, value) or
-    CallTargetStatsReport::percentageOfOk(key, value)
-  ) and
-  /* Infinity */
-  value != 1.0 / 0.0 and
-  /* -Infinity */
-  value != -1.0 / 0.0 and
-  /* NaN */
-  value != 0.0 / 0.0
-select key, value

cpp/ql/src/change-notes/2026-03-11-integer-multiplication-cast-to-long.md

@@ -1,4 +0,0 @@
----
-category: minorAnalysis
----
-* Fixed an issue with the "Multiplication result converted to larger type" (`cpp/integer-multiplication-cast-to-long`) query causing false positive results in `build-mode: none` databases.

cpp/ql/src/change-notes/2026-03-13-adjust-xss-and-log-injection-severity.md

@@ -1,4 +0,0 @@
----
-category: queryMetadata
----
-* The `@security-severity` metadata of `cpp/cgi-xss` has been increased from 6.1 (medium) to 7.8 (high).

cpp/ql/src/change-notes/2026-03-16-wrong-type-format-argument.md

@@ -1,4 +0,0 @@
----
-category: minorAnalysis
----
-* Fixed an issue with the "Wrong type of arguments to formatting function" (`cpp/wrong-type-format-argument`) query causing false positive results in `build-mode: none` databases.

cpp/ql/src/change-notes/2026-03-19-suspicious-add-sizeof.md

@@ -1,4 +0,0 @@
----
-category: minorAnalysis
----
-* Fixed an issue with the "Suspicious add with sizeof" (`cpp/suspicious-add-sizeof`) query causing false positive results in `build-mode: none` databases.

cpp/ql/src/change-notes/2026-03-19-tainted-format-string.md

@@ -1,4 +0,0 @@
----
-category: minorAnalysis
----
-* Fixed an issue with the "Uncontrolled format string" (`cpp/tainted-format-string`) query involving certain kinds of formatting function implementations.

cpp/ql/src/change-notes/released/1.5.11.md

@@ -1,3 +0,0 @@
-## 1.5.11
-
-No user-facing changes.

cpp/ql/src/change-notes/released/1.5.12.md

@@ -1,3 +0,0 @@
-## 1.5.12
-
-No user-facing changes.

cpp/ql/src/change-notes/released/1.5.13.md

@@ -1,3 +0,0 @@
-## 1.5.13
-
-No user-facing changes.

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

@@ -1,2 +1,2 @@
 ---
-lastReleaseVersion: 1.5.13
+lastReleaseVersion: 1.5.10

cpp/ql/src/qlpack.yml

@@ -1,5 +1,5 @@
 name: codeql/cpp-queries
-version: 1.5.14-dev
+version: 1.5.11-dev
 groups:
   - cpp
   - queries

cpp/ql/src/utils/modelgenerator/internal/CaptureModels.qll

@@ -8,7 +8,6 @@ private import semmle.code.cpp.dataflow.ExternalFlow as ExternalFlow
 private import semmle.code.cpp.ir.dataflow.internal.DataFlowImplCommon as DataFlowImplCommon
 private import semmle.code.cpp.ir.dataflow.internal.DataFlowImplSpecific
 private import semmle.code.cpp.ir.dataflow.internal.DataFlowPrivate as DataFlowPrivate
-private import semmle.code.cpp.ir.dataflow.internal.DataFlowNodes as DataFlowNodes
 private import semmle.code.cpp.dataflow.internal.FlowSummaryImpl as FlowSummaryImpl
 private import semmle.code.cpp.ir.dataflow.internal.TaintTrackingImplSpecific
 private import semmle.code.cpp.dataflow.new.TaintTracking as Tt
@@ -404,7 +403,7 @@ private module SinkModelGeneratorInput implements SinkModelGeneratorInputSig {
   }
 
   predicate apiSource(DataFlow::Node source) {
-    DataFlowPrivate::nodeHasOperand(source, any(DataFlowNodes::FieldAddress fa), 1)
+    DataFlowPrivate::nodeHasOperand(source, any(DataFlow::FieldAddress fa), 1)
     or
     source instanceof DataFlow::ParameterNode
   }
@@ -417,7 +416,7 @@ private module SinkModelGeneratorInput implements SinkModelGeneratorInputSig {
       result = "Argument[" + DataFlow::repeatStars(indirectionIndex) + argumentIndex + "]"
     )
     or
-    DataFlowPrivate::nodeHasOperand(source, any(DataFlowNodes::FieldAddress fa), 1) and
+    DataFlowPrivate::nodeHasOperand(source, any(DataFlow::FieldAddress fa), 1) and
     result = qualifierString()
   }
 

cpp/ql/test/library-tests/dataflow/fields/A.cpp

@@ -46,7 +46,7 @@ public:
   {
     C *c = new C();
     B *b = B::make(c);
-    sink(b->c); // $ ast,ir
+    sink(b->c); // $ast,ir
   }
 
   void f2()

cpp/ql/test/library-tests/dataflow/fields/C.cpp

@@ -26,9 +26,9 @@ public:
 
   void func()
   {
-    sink(s1); // $ ast,ir
+    sink(s1); // $ast,ir
     sink(s2); // $ MISSING: ast,ir
-    sink(s3); // $ ast,ir
+    sink(s3); // $ast,ir
     sink(s4); // $ MISSING: ast,ir
   }
 };

cpp/ql/test/library-tests/dataflow/fields/D.cpp

@@ -19,7 +19,7 @@ public:
   };
 
   static void sinkWrap(Box2* b2) {
-    sink(b2->getBox1()->getElem()); // $ ast,ir=28:15 ast,ir=35:15 ast,ir=42:15 ast,ir=49:15
+    sink(b2->getBox1()->getElem()); // $ast,ir=28:15 ast,ir=35:15 ast,ir=42:15 ast,ir=49:15
   }
 
   Box2* boxfield;

cpp/ql/test/library-tests/dataflow/fields/by_reference.cpp

@@ -48,25 +48,25 @@ struct S {
 void test_setDirectly() {
   S s;
   s.setDirectly(user_input());
-  sink(s.getDirectly()); // $ ast ir
+  sink(s.getDirectly()); // $ast ir
 }
 
 void test_setIndirectly() {
   S s;
   s.setIndirectly(user_input());
-  sink(s.getIndirectly()); // $ ast ir
+  sink(s.getIndirectly()); // $ast ir
 }
 
 void test_setThroughNonMember() {
   S s;
   s.setThroughNonMember(user_input());
-  sink(s.getThroughNonMember()); // $ ast ir
+  sink(s.getThroughNonMember()); // $ast ir
 }
 
 void test_nonMemberSetA() {
   S s;
   nonMemberSetA(&s, user_input());
-  sink(nonMemberGetA(&s)); // $ ast,ir
+  sink(nonMemberGetA(&s)); // $ast,ir
 }
 
 ////////////////////
@@ -112,7 +112,7 @@ void test_outer_with_ptr(Outer *pouter) {
   sink(outer.a); // $ ast,ir
 
   sink(pouter->inner_nested.a); // $ ast,ir
-  sink(pouter->inner_ptr->a); // $ ast,ir
+  sink(pouter->inner_ptr->a); // $ast,ir
   sink(pouter->a); // $ ast,ir
 }
 

cpp/ql/test/library-tests/dataflow/fields/simple.cpp

@@ -64,7 +64,7 @@ void single_field_test()
     A a;
     a.i = user_input();
     A a2 = a;
-    sink(a2.i); // $ ast,ir
+    sink(a2.i); //$ ast,ir
 }
 
 struct C {
@@ -81,7 +81,7 @@ struct C2
 
     void m() {
         f2.f1 = user_input();
-        sink(getf2f1()); // $ ast,ir
+        sink(getf2f1()); //$ ast,ir
     }
 };
 
@@ -91,7 +91,7 @@ void single_field_test_typedef(A_typedef a)
 {
     a.i = user_input();
     A_typedef a2 = a;
-    sink(a2.i); // $ ast,ir
+    sink(a2.i); //$ ast,ir
 }
 
 namespace TestAdditionalCallTargets {
@@ -168,4 +168,4 @@ void test_union_with_two_instantiations_of_different_sizes() {
   sink(u_int.y); // $ MISSING: ir
 }
 
-} // namespace Simple
+} // namespace Simple

cpp/ql/test/library-tests/dataflow/fields/struct_init.c

@@ -12,14 +12,14 @@ struct Outer {
 };
 
 void absink(struct AB *ab) {
-  sink(ab->a); // $ ast,ir=20:20 ast,ir=27:7 ast,ir=40:20
+  sink(ab->a); //$ ast,ir=20:20 ast,ir=27:7 ast,ir=40:20
   sink(ab->b); // no flow
 }
 
 int struct_init(void) {
   struct AB ab = { user_input(), 0 };
 
-  sink(ab.a); // $ ast,ir
+  sink(ab.a); //$ ast,ir
   sink(ab.b); // no flow
   absink(&ab);
 
@@ -28,9 +28,9 @@ int struct_init(void) {
     &ab,
   };
 
-  sink(outer.nestedAB.a); // $ ast,ir
+  sink(outer.nestedAB.a); //$ ast,ir
   sink(outer.nestedAB.b); // no flow
-  sink(outer.pointerAB->a); // $ ast,ir
+  sink(outer.pointerAB->a); //$ ast,ir
   sink(outer.pointerAB->b); // no flow
 
   absink(&outer.nestedAB);

cpp/ql/test/library-tests/dataflow/ir-barrier-guards/test.ql

@@ -15,10 +15,7 @@ predicate instructionGuardChecks(IRGuardCondition gc, Instruction checked, boole
 module BarrierGuard = DataFlow::InstructionBarrierGuard<instructionGuardChecks/3>;
 
 predicate indirectBarrierGuard(DataFlow::Node node, string s) {
-  // This any(...) could technically be removed, but it helps us verify that we don't
-  // accidentially change the API of this predicate (for instance, by having
-  // the column be a unit parameter).
-  node = BarrierGuard::getAnIndirectBarrierNode(any(int indirectionIndex)) and
+  node = BarrierGuard::getAnIndirectBarrierNode(_) and
   if node.isGLValue()
   then s = "glval<" + node.getType().toString().replaceAll(" ", "") + ">"
   else s = node.getType().toString().replaceAll(" ", "")

cpp/ql/test/library-tests/dataflow/models-as-data/FlowSummaryNode.ql

@@ -1,7 +1,6 @@
 import testModels
 private import semmle.code.cpp.ir.dataflow.internal.DataFlowPrivate
 private import semmle.code.cpp.ir.dataflow.internal.DataFlowUtil
-private import semmle.code.cpp.ir.dataflow.internal.DataFlowNodes
 
 string describe(DataFlow::Node n) {
   n instanceof ParameterNode and result = "ParameterNode"

cpp/ql/test/library-tests/dataflow/models-as-data/tests.cpp

@@ -75,7 +75,7 @@ void test_sources() {
 	int e = localMadSource();
 	sink(e); // $ ir
 
-	sink(MyNamespace::namespaceLocalMadSource()); // $ ir
+	sink(MyNamespace::namespaceLocalMadSource()); // $: ir
 	sink(MyNamespace::namespaceLocalMadSourceVar); // $ ir
 	sink(MyNamespace::MyNamespace2::namespace2LocalMadSource()); // $ ir
 	sink(MyNamespace::localMadSource()); // $ (the MyNamespace version of this function is not a source)
@@ -475,4 +475,4 @@ void test_receive_array() {
 	int array[10] = {x};
 	int y = receive_array(array);
 	sink(y); // $ ir
-}
+}

cpp/ql/test/library-tests/dataflow/taint-tests/taint.cpp

@@ -450,7 +450,7 @@ void test_qualifiers()
 	b.member = source();
 	sink(b); // $ ir MISSING: ast
 	sink(b.member); // $ ast,ir
-	sink(b.getMember()); // $ MISSING: ir ast
+	sink(b.getMember()); // $  MISSING: ir ast
 
 	c = new MyClass2(0);
 
@@ -865,4 +865,4 @@ void test_iconv(size_t size) {
 	size_t size_out;
 	iconv(0, &s, &size, &p, &size_out);
 	sink(*p); // $ ast,ir
-}
+}

cpp/ql/test/library-tests/ir/points_to/points_to.cpp

@@ -24,64 +24,64 @@ struct DerivedVI : virtual Base1 {
 };
 
 void Locals() {
-  Point pt = {  // $ ussa=pt
-    1,  // $ ussa=pt[0..4)<int>
-    2  // $ ussa=pt[4..8)<int>
+  Point pt = {  //$ussa=pt
+    1,  //$ussa=pt[0..4)<int>
+    2  //$ussa=pt[4..8)<int>
   };
-  int i = pt.x;  // $ ussa=pt[0..4)<int>
-  i = pt.y;  // $ ussa=pt[4..8)<int>
+  int i = pt.x;  //$ussa=pt[0..4)<int>
+  i = pt.y;  //$ussa=pt[4..8)<int>
   int* p = &pt.x;
-  i = *p;  // $ ussa=pt[0..4)<int>
+  i = *p;  //$ussa=pt[0..4)<int>
   p = &pt.y;
-  i = *p;  // $ ussa=pt[4..8)<int>
+  i = *p;  //$ussa=pt[4..8)<int>
 }
 
 void PointsTo(
-  int a,         // $ raw=a
-  Point& b,      // $ raw=b ussa=*b
-  Point* c,      // $ raw=c ussa=*c
-  int* d,        // $ raw=d ussa=*d
-  DerivedSI* e,  // $ raw=e ussa=*e
-  DerivedMI* f,  // $ raw=f ussa=*f
-  DerivedVI* g   // $ raw=g ussa=*g
+  int a,         //$raw=a
+  Point& b,      //$raw=b ussa=*b
+  Point* c,      //$raw=c ussa=*c
+  int* d,        //$raw=d ussa=*d
+  DerivedSI* e,  //$raw=e ussa=*e
+  DerivedMI* f,  //$raw=f ussa=*f
+  DerivedVI* g   //$raw=g ussa=*g
 ) {
 
-  int i = a;  // $ raw=a
-  i = *&a;  // $ raw=a
-  i = *(&a + 0);  // $ raw=a
-  i = b.x;  // $ raw=b ussa=*b[0..4)<int>
-  i = b.y;  // $ raw=b ussa=*b[4..8)<int>
-  i = c->x;  // $ raw=c ussa=*c[0..4)<int>
-  i = c->y;  // $ raw=c ussa=*c[4..8)<int>
-  i = *d;  // $ raw=d ussa=*d[0..4)<int>
-  i = *(d + 0);  // $ raw=d ussa=*d[0..4)<int>
-  i = d[5];  // $ raw=d ussa=*d[20..24)<int>
-  i = 5[d];  // $ raw=d ussa=*d[20..24)<int>
-  i = d[a];  // $ raw=d raw=a ussa=*d[?..?)<int>
-  i = a[d];  // $ raw=d raw=a ussa=*d[?..?)<int>
+  int i = a;  //$raw=a
+  i = *&a;  //$raw=a
+  i = *(&a + 0);  //$raw=a
+  i = b.x;  //$raw=b ussa=*b[0..4)<int>
+  i = b.y;  //$raw=b ussa=*b[4..8)<int>
+  i = c->x;  //$raw=c ussa=*c[0..4)<int>
+  i = c->y;  //$raw=c ussa=*c[4..8)<int>
+  i = *d;  //$raw=d ussa=*d[0..4)<int>
+  i = *(d + 0);  //$raw=d ussa=*d[0..4)<int>
+  i = d[5];  //$raw=d ussa=*d[20..24)<int>
+  i = 5[d];  //$raw=d ussa=*d[20..24)<int>
+  i = d[a];  //$raw=d raw=a ussa=*d[?..?)<int>
+  i = a[d];  //$raw=d raw=a ussa=*d[?..?)<int>
 
-  int* p = &b.x;  // $ raw=b
-  i = *p;  // $ ussa=*b[0..4)<int>
-  p = &b.y;  // $ raw=b
-  i = *p;  // $ ussa=*b[4..8)<int>
-  p = &c->x;  // $ raw=c
-  i = *p;  // $ ussa=*c[0..4)<int>
-  p = &c->y;  // $ raw=c
-  i = *p;  // $ ussa=*c[4..8)<int>
-  p = &d[5];  // $ raw=d
-  i = *p;  // $ ussa=*d[20..24)<int>
-  p = &d[a];  // $ raw=d raw=a
-  i = *p;  // $ ussa=*d[?..?)<int>
+  int* p = &b.x;  //$raw=b
+  i = *p;  //$ussa=*b[0..4)<int>
+  p = &b.y;  //$raw=b
+  i = *p;  //$ussa=*b[4..8)<int>
+  p = &c->x;  //$raw=c
+  i = *p;  //$ussa=*c[0..4)<int>
+  p = &c->y;  //$raw=c
+  i = *p;  //$ussa=*c[4..8)<int>
+  p = &d[5];  //$raw=d
+  i = *p;  //$ussa=*d[20..24)<int>
+  p = &d[a];  //$raw=d raw=a
+  i = *p;  //$ussa=*d[?..?)<int>
 
-  Point* q = &c[a];  // $ raw=c raw=a
-  i = q->x;  // $ ussa=*c[?..?)<int>
-  i = q->y;  // $ ussa=*c[?..?)<int>
+  Point* q = &c[a];  //$raw=c raw=a
+  i = q->x;  //$ussa=*c[?..?)<int>
+  i = q->y;  //$ussa=*c[?..?)<int>
 
-  i = e->b1;  // $ raw=e ussa=*e[0..4)<int>
-  i = e->dsi;  // $ raw=e ussa=*e[4..8)<int>
-  i = f->b1;  // $ raw=f ussa=*f[0..4)<int>
-  i = f->b2;  // $ raw=f ussa=*f[4..8)<int>
-  i = f->dmi;  // $ raw=f ussa=*f[8..12)<int>
-  i = g->b1;  // $ raw=g ussa=*g[?..?)<int>
-  i = g->dvi;  // $ raw=g ussa=*g[8..12)<int>
-}
+  i = e->b1;  //$raw=e ussa=*e[0..4)<int>
+  i = e->dsi;  //$raw=e ussa=*e[4..8)<int>
+  i = f->b1;  //$raw=f ussa=*f[0..4)<int>
+  i = f->b2;  //$raw=f ussa=*f[4..8)<int>
+  i = f->dmi;  //$raw=f ussa=*f[8..12)<int>
+  i = g->b1;  //$raw=g ussa=*g[?..?)<int>
+  i = g->dvi;  //$raw=g ussa=*g[8..12)<int>
+}