Merge branch 'main' into copilot/add-xxe-query-for-rust

This commit is contained in:
Geoffrey White
2026-03-27 18:32:18 +00:00
2113 changed files with 104338 additions and 88097 deletions

View File

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

View File

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

View File

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

View File

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

View File

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

View File

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

View File

@@ -15,23 +15,23 @@ local_path_override(
# see https://registry.bazel.build/ for a list of available packages
bazel_dep(name = "platforms", version = "1.0.0")
bazel_dep(name = "rules_cc", version = "0.2.16")
bazel_dep(name = "rules_go", version = "0.59.0")
bazel_dep(name = "rules_java", version = "9.0.3")
bazel_dep(name = "rules_pkg", version = "1.0.1")
bazel_dep(name = "rules_cc", version = "0.2.17")
bazel_dep(name = "rules_go", version = "0.60.0")
bazel_dep(name = "rules_java", version = "9.6.1")
bazel_dep(name = "rules_pkg", version = "1.2.0")
bazel_dep(name = "rules_nodejs", version = "6.7.3")
bazel_dep(name = "rules_python", version = "0.40.0")
bazel_dep(name = "rules_shell", version = "0.5.0")
bazel_dep(name = "bazel_skylib", version = "1.8.1")
bazel_dep(name = "abseil-cpp", version = "20240116.1", repo_name = "absl")
bazel_dep(name = "rules_python", version = "1.9.0")
bazel_dep(name = "rules_shell", version = "0.7.1")
bazel_dep(name = "bazel_skylib", version = "1.9.0")
bazel_dep(name = "abseil-cpp", version = "20260107.1", repo_name = "absl")
bazel_dep(name = "nlohmann_json", version = "3.11.3", repo_name = "json")
bazel_dep(name = "fmt", version = "12.1.0-codeql.1")
bazel_dep(name = "rules_kotlin", version = "2.2.2-codeql.1")
bazel_dep(name = "gazelle", version = "0.47.0")
bazel_dep(name = "rules_dotnet", version = "0.21.5-codeql.1")
bazel_dep(name = "googletest", version = "1.14.0.bcr.1")
bazel_dep(name = "rules_rust", version = "0.68.1.codeql.1")
bazel_dep(name = "zstd", version = "1.5.5.bcr.1")
bazel_dep(name = "googletest", version = "1.17.0.bcr.2")
bazel_dep(name = "rules_rust", version = "0.69.0")
bazel_dep(name = "zstd", version = "1.5.7.bcr.1")
bazel_dep(name = "buildifier_prebuilt", version = "6.4.0", dev_dependency = True)

View File

@@ -1,3 +1,11 @@
## 0.4.30
No user-facing changes.
## 0.4.29
No user-facing changes.
## 0.4.28
No user-facing changes.

View File

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

View File

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

View File

@@ -1,2 +1,2 @@
---
lastReleaseVersion: 0.4.28
lastReleaseVersion: 0.4.30

View File

@@ -1,5 +1,5 @@
name: codeql/actions-all
version: 0.4.29-dev
version: 0.4.31-dev
library: true
warnOnImplicitThis: true
dependencies:

View File

@@ -1,3 +1,11 @@
## 0.6.22
No user-facing changes.
## 0.6.21
No user-facing changes.
## 0.6.20
No user-facing changes.

View File

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

View File

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

View File

@@ -1,2 +1,2 @@
---
lastReleaseVersion: 0.6.20
lastReleaseVersion: 0.6.22

View File

@@ -1,5 +1,5 @@
name: codeql/actions-queries
version: 0.6.21-dev
version: 0.6.23-dev
library: false
warnOnImplicitThis: true
groups: [actions, queries]

View File

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

View File

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

View File

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

File diff suppressed because it is too large Load Diff

File diff suppressed because it is too large Load Diff

View File

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

View File

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

View File

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

View File

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

View File

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

View File

@@ -1,3 +1,24 @@
## 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

View File

@@ -1,4 +0,0 @@
---
category: minorAnalysis
---
* 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.

View File

@@ -1,4 +0,0 @@
---
category: fix
---
* The `allowInterproceduralFlow` predicate of must-flow data flow configurations now correctly handles direct recursion.

View File

@@ -1,4 +0,0 @@
---
category: breaking
---
* `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.

View File

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

View File

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

View File

@@ -0,0 +1,14 @@
## 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.

View File

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

View File

@@ -1,2 +1,2 @@
---
lastReleaseVersion: 7.1.1
lastReleaseVersion: 8.0.1

View File

@@ -1,5 +1,5 @@
name: codeql/cpp-all
version: 7.1.2-dev
version: 8.0.2-dev
groups: cpp
dbscheme: semmlecode.cpp.dbscheme
extractor: cpp

View File

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

View File

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

View File

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

View File

@@ -353,12 +353,26 @@ 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()
getInvalidModelSubtype(), getInvalidModelColumnCount(), KindVal::getInvalidModelKind(),
getIncorrectConstructorSummaryOutput()
]
}
}
@@ -555,6 +569,7 @@ private Locatable getSupportedFunctionTemplateArgument(Function templateFunction
* Normalize the `n`'th parameter of `f` by replacing template names
* with `func:N` (where `N` is the index of the template).
*/
pragma[nomagic]
private string getTypeNameWithoutFunctionTemplates(Function f, int n, int remaining) {
exists(Function templateFunction |
templateFunction = getFullyTemplatedFunction(f) and

View File

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

View File

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

View File

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

File diff suppressed because it is too large Load Diff

View File

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

View File

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

View File

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

View File

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

View File

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

View File

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

View File

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

View File

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

View File

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

View File

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

View File

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

File diff suppressed because it is too large Load Diff

View File

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

File diff suppressed because it is too large Load Diff

View File

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

View File

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

View File

@@ -1,3 +1,11 @@
## 1.5.13
No user-facing changes.
## 1.5.12
No user-facing changes.
## 1.5.11
No user-facing changes.

View File

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

View File

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

View File

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

View File

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

View File

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

View File

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

View File

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

View File

@@ -0,0 +1,28 @@
/**
* @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) or
ExprTypeStatsReport::numberOfOk(key, value) or
ExprTypeStatsReport::numberOfNotOk(key, value) or
ExprTypeStatsReport::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

View File

@@ -0,0 +1,4 @@
---
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.

View File

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

View File

@@ -0,0 +1,4 @@
---
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.

View File

@@ -0,0 +1,4 @@
---
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.

View File

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

View File

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

View File

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

View File

@@ -1,2 +1,2 @@
---
lastReleaseVersion: 1.5.11
lastReleaseVersion: 1.5.13

View File

@@ -1,5 +1,5 @@
name: codeql/cpp-queries
version: 1.5.12-dev
version: 1.5.14-dev
groups:
- cpp
- queries

View File

@@ -8,6 +8,7 @@ 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
@@ -403,7 +404,7 @@ private module SinkModelGeneratorInput implements SinkModelGeneratorInputSig {
}
predicate apiSource(DataFlow::Node source) {
DataFlowPrivate::nodeHasOperand(source, any(DataFlow::FieldAddress fa), 1)
DataFlowPrivate::nodeHasOperand(source, any(DataFlowNodes::FieldAddress fa), 1)
or
source instanceof DataFlow::ParameterNode
}
@@ -416,7 +417,7 @@ private module SinkModelGeneratorInput implements SinkModelGeneratorInputSig {
result = "Argument[" + DataFlow::repeatStars(indirectionIndex) + argumentIndex + "]"
)
or
DataFlowPrivate::nodeHasOperand(source, any(DataFlow::FieldAddress fa), 1) and
DataFlowPrivate::nodeHasOperand(source, any(DataFlowNodes::FieldAddress fa), 1) and
result = qualifierString()
}

View File

@@ -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()

View File

@@ -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
}
};

View File

@@ -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;

View File

@@ -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
}

View File

@@ -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

View File

@@ -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);

View File

@@ -15,7 +15,10 @@ predicate instructionGuardChecks(IRGuardCondition gc, Instruction checked, boole
module BarrierGuard = DataFlow::InstructionBarrierGuard<instructionGuardChecks/3>;
predicate indirectBarrierGuard(DataFlow::Node node, string s) {
node = BarrierGuard::getAnIndirectBarrierNode(_) and
// 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
if node.isGLValue()
then s = "glval<" + node.getType().toString().replaceAll(" ", "") + ">"
else s = node.getType().toString().replaceAll(" ", "")

View File

@@ -1,6 +1,7 @@
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"

View File

@@ -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
}
}

View File

@@ -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
}
}

View File

@@ -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>
}

View File

@@ -10,24 +10,24 @@ struct S {
void unique_ptr_init(S s) {
unique_ptr<S> p(new S); // MISSING: $ussa=dynamic{1}
int i = (*p).x; //$ MISSING: ussa=dynamic{1}[0..4)<int>
*p = s; //$ MISSING: ussa=dynamic{1}[0..4)<S>
int i = (*p).x; // $ MISSING: ussa=dynamic{1}[0..4)<int>
*p = s; // $ MISSING: ussa=dynamic{1}[0..4)<S>
unique_ptr<S> q = std::move(p);
*(q.get()) = s; //$ MISSING: ussa=dynamic{1}[0..4)<S>
*(q.get()) = s; // $ MISSING: ussa=dynamic{1}[0..4)<S>
shared_ptr<S> t(std::move(q));
t->x = 5; //$ MISSING: ussa=dynamic{1}[0..4)<int>
*t = s; //$ MISSING: ussa=dynamic{1}[0..4)<S>
*(t.get()) = s; //$ MISSING: ussa=dynamic{1}[0..4)<S>
t->x = 5; // $ MISSING: ussa=dynamic{1}[0..4)<int>
*t = s; // $ MISSING: ussa=dynamic{1}[0..4)<S>
*(t.get()) = s; // $ MISSING: ussa=dynamic{1}[0..4)<S>
}
void shared_ptr_init(S s) {
shared_ptr<S> p(new S); //$ MISSING: ussa=dynamic{1}
int i = (*p).x; //$ MISSING: ussa=dynamic{1}[0..4)<int>
*p = s; //$ MISSING: ussa=dynamic{1}[0..4)<S>
shared_ptr<S> p(new S); // $ MISSING: ussa=dynamic{1}
int i = (*p).x; // $ MISSING: ussa=dynamic{1}[0..4)<int>
*p = s; // $ MISSING: ussa=dynamic{1}[0..4)<S>
shared_ptr<S> q = std::move(p);
*(q.get()) = s; //$ MISSING: ussa=dynamic{1}[0..4)<S>
*(q.get()) = s; // $ MISSING: ussa=dynamic{1}[0..4)<S>
shared_ptr<S> t(q);
t->x = 5; //$ MISSING: ussa=dynamic{1}[0..4)<int>
*t = s; //$ MISSING: ussa=dynamic{1}[0..4)<S>
*(t.get()) = s; //$ MISSING: ussa=dynamic{1}[0..4)<S>
t->x = 5; // $ MISSING: ussa=dynamic{1}[0..4)<int>
*t = s; // $ MISSING: ussa=dynamic{1}[0..4)<S>
*(t.get()) = s; // $ MISSING: ussa=dynamic{1}[0..4)<S>
}

View File

@@ -46,7 +46,7 @@ int test4() {
}
range(total); // $ MISSING: range=>=0
range(i); // $ range===2
range(total + i); // $ range="<=Phi: i+2" MISSING: range===i+2 range=>=2 range=>=i+0
range(total + i); // $ range="<=Phi: i+2" MISSING: range===i+2 range=>=2 range=>=i+0
return total + i;
}
@@ -210,7 +210,7 @@ int test14(int x) {
int x3 = (int)(unsigned int)x;
range(x3);
char c0 = x;
range(c0);
range(c0);
unsigned short s0 = x;
range(s0);
range(x0 + x1 + x2 + x3 + c0 + s0); // $ overflow=+ overflow=+-
@@ -218,7 +218,7 @@ int test14(int x) {
}
long long test15(long long x) {
return (x > 0 && (range(x), x == (int)x)) ? // $ range=>=1
return (x > 0 && (range(x), x == (int)x)) ? // $ range=>=1
(range(x), x) : // $ range=>=1
(range(x), -1);
}
@@ -228,7 +228,7 @@ int test_unary(int a) {
int total = 0;
if (3 <= a && a <= 11) {
range(a); // $ range=<=11 range=>=3
range(a); // $ range=<=11 range=>=3
int b = +a;
range(b); // $ range=<=11 range=>=3
int c = -a;
@@ -384,7 +384,7 @@ int test_mult02(int a, int b) {
total += r;
range(total); // $ range=">=Phi: 0-143" range=">=Phi: 0-286"
}
range(total); // $range=">=Phi: 0-143" range=">=Phi: 0-286"
range(total); // $ range=">=Phi: 0-143" range=">=Phi: 0-286"
return total;
}
@@ -467,7 +467,7 @@ int test_mult04(int a, int b) {
range(a); // $ range=<=0 range=>=-17
range(b); // $ range=<=0 range=>=-13
int r = a*b; // 0 .. 221
range(r); // $ range=<=221 range=>=0
range(r); // $ range=<=221 range=>=0
total += r;
range(total); // $ range="<=Phi: - ...+221"
}
@@ -1030,7 +1030,7 @@ void test_negate_signed(int s) {
}
}
// By setting the guard after the use in another guard we
// By setting the guard after the use in another guard we
// don't get the useful information
void test_guard_after_use(int pos, int size, int offset) {
if (pos + offset >= size) { // $ overflow=+-
@@ -1040,12 +1040,12 @@ void test_guard_after_use(int pos, int size, int offset) {
return;
}
range(pos + 1); // $ overflow=+ range="==InitializeParameter: pos+1" MISSING: range="<=InitializeParameter: size-1"
}
}
int cond();
// This is basically what we get when we have a loop that calls
// This is basically what we get when we have a loop that calls
// realloc in some iterations
void alloc_in_loop(int origLen) {
if (origLen <= 10) {
@@ -1066,12 +1066,12 @@ void alloc_in_loop(int origLen) {
}
}
// This came from a case where it handled the leftovers before an unrolled loop
// This came from a case where it handled the leftovers before an unrolled loop
void mask_at_start(int len) {
if (len < 0) {
return;
}
int leftOver = len & 63;
int leftOver = len & 63;
for (int i = 0; i < leftOver; i++) {
range(i); // $ range=<=62 range=>=0 range="<=Store: ... & ... | Store: leftOver-1" range="<=InitializeParameter: len-1"
}

View File

@@ -1,14 +1,14 @@
void Complex(void) {
_Complex float cf; //$irtype=cfloat8
_Complex double cd; //$irtype=cfloat16
_Complex long double cld; //$irtype=cfloat32
_Complex float cf; // $ irtype=cfloat8
_Complex double cd; // $ irtype=cfloat16
_Complex long double cld; // $ irtype=cfloat32
// _Complex __float128 cf128;
}
void Imaginary(void) {
_Imaginary float jf; //$irtype=ifloat4
_Imaginary double jd; //$irtype=ifloat8
_Imaginary long double jld; //$irtype=ifloat16
_Imaginary float jf; // $ irtype=ifloat4
_Imaginary double jd; // $ irtype=ifloat8
_Imaginary long double jld; // $ irtype=ifloat16
// _Imaginary __float128 jf128;
}

View File

@@ -22,44 +22,44 @@ enum class ScopedE {
};
void IRTypes() {
char c; //$irtype=int1
signed char sc; //$irtype=int1
unsigned char uc; //$irtype=uint1
short s; //$irtype=int2
signed short ss; //$irtype=int2
unsigned short us; //$irtype=uint2
int i; //$irtype=int4
signed int si; //$irtype=int4
unsigned int ui; //$irtype=uint4
long l; //$irtype=int8
signed long sl; //$irtype=int8
unsigned long ul; //$irtype=uint8
long long ll; //$irtype=int8
signed long long sll; //$irtype=int8
unsigned long long ull; //$irtype=uint8
bool b; //$irtype=bool1
float f; //$irtype=float4
double d; //$irtype=float8
long double ld; //$irtype=float16
__float128 f128; //$irtype=float16
char c; // $ irtype=int1
signed char sc; // $ irtype=int1
unsigned char uc; // $ irtype=uint1
short s; // $ irtype=int2
signed short ss; // $ irtype=int2
unsigned short us; // $ irtype=uint2
int i; // $ irtype=int4
signed int si; // $ irtype=int4
unsigned int ui; // $ irtype=uint4
long l; // $ irtype=int8
signed long sl; // $ irtype=int8
unsigned long ul; // $ irtype=uint8
long long ll; // $ irtype=int8
signed long long sll; // $ irtype=int8
unsigned long long ull; // $ irtype=uint8
bool b; // $ irtype=bool1
float f; // $ irtype=float4
double d; // $ irtype=float8
long double ld; // $ irtype=float16
__float128 f128; // $ irtype=float16
wchar_t wc; //$irtype=uint4
// char8_t c8; //$irtype=uint1
char16_t c16; //$irtype=uint2
char32_t c32; //$irtype=uint4
wchar_t wc; // $ irtype=uint4
// char8_t c8; // $ irtype=uint1
char16_t c16; // $ irtype=uint2
char32_t c32; // $ irtype=uint4
int* pi; //$irtype=addr8
int& ri = i; //$irtype=addr8
void (*pfn)() = nullptr; //$irtype=func8
void (&rfn)() = IRTypes; //$irtype=func8
int* pi; // $ irtype=addr8
int& ri = i; // $ irtype=addr8
void (*pfn)() = nullptr; // $ irtype=func8
void (&rfn)() = IRTypes; // $ irtype=func8
A s_a; //$irtype=opaque4{A}
B s_b; //$irtype=opaque16{B}
A s_a; // $ irtype=opaque4{A}
B s_b; // $ irtype=opaque16{B}
E e; //$irtype=uint4
ScopedE se; //$irtype=uint4
E e; // $ irtype=uint4
ScopedE se; // $ irtype=uint4
B a_b[10]; //$irtype=opaque160{B[10]}
B a_b[10]; // $ irtype=opaque160{B[10]}
}
// semmle-extractor-options: -std=c++17 --clang

View File

@@ -0,0 +1,28 @@
// semmle-extractor-options: --expect_errors
void test_float_double1(float f, double d) {
float r1 = f * f; // GOOD
float r2 = f * d; // GOOD
double r3 = f * f; // BAD
double r4 = f * d; // GOOD
float f1 = fabsf(f * f); // GOOD
float f2 = fabsf(f * d); // GOOD
double f3 = fabs(f * f); // BAD [NOT DETECTED]
double f4 = fabs(f * d); // GOOD
}
double fabs(double f);
float fabsf(float f);
void test_float_double2(float f, double d) {
float r1 = f * f; // GOOD
float r2 = f * d; // GOOD
double r3 = f * f; // BAD
double r4 = f * d; // GOOD
float f1 = fabsf(f * f); // GOOD
float f2 = fabsf(f * d); // GOOD
double f3 = fabs(f * f); // BAD [NOT DETECTED]
double f4 = fabs(f * d); // GOOD
}

View File

@@ -1,3 +1,5 @@
| Buildless.c:6:17:6:21 | ... * ... | Multiplication result may overflow 'float' before it is converted to 'double'. |
| Buildless.c:21:17:21:21 | ... * ... | Multiplication result may overflow 'float' before it is converted to 'double'. |
| IntMultToLong.c:4:10:4:14 | ... * ... | Multiplication result may overflow 'int' before it is converted to 'long long'. |
| IntMultToLong.c:7:16:7:20 | ... * ... | Multiplication result may overflow 'int' before it is converted to 'long long'. |
| IntMultToLong.c:18:19:18:23 | ... * ... | Multiplication result may overflow 'float' before it is converted to 'double'. |

View File

@@ -1 +1,3 @@
| second.cpp:26:18:26:39 | ... - ... | This format specifier for type 'int' does not match the argument type 'long'. |
| second.cpp:29:18:29:39 | ... - ... | This format specifier for type 'unsigned int' does not match the argument type 'long'. |
| tests.c:7:18:7:18 | 1 | This format specifier for type 'char *' does not match the argument type 'int'. |

Some files were not shown because too many files have changed in this diff Show More