Compare commits

..

467 Commits

Author SHA1 Message Date
Edoardo Pirovano
33ea91042a Restore queries to old state 2022-11-04 10:09:46 +00:00
Asger F
8502939b65 Merge pull request #11081 from asgerf/ql/dependency-paths
QL4QL: Add DependencyPath.ql query
2022-11-03 10:17:45 +01:00
Mathias Vorreiter Pedersen
01f3150a70 Merge pull request #11092 from hvitved/swift/avoid-deprecated-ssa-predicates
Swift: Avoid calls to deprecated SSA predicates
2022-11-03 09:03:47 +00:00
Asger F
fbcdb53d72 QL: Add option to follow 'cached' dependencies 2022-11-03 09:43:56 +01:00
Tom Hvitved
4e3fcc3235 Swift: Avoid calls to deprecated SSA predicates 2022-11-03 09:03:20 +01:00
Dave Bartolomeo
499f20f6e8 Merge pull request #11004 from dbartol/dbartol/use-workspace-versions 2022-11-02 20:02:48 -04:00
Tom Hvitved
46631d6eaf Merge pull request #10931 from hvitved/ruby/fix-flow-into-phis
Ruby: Fix flow steps into phi nodes
2022-11-02 21:07:06 +01:00
Chris Smowton
8eee450c65 Merge pull request #11064 from smowton/smowton/fix/kotlin-inherited-defaults
Kotlin: handle default parameter values inherited from an overridden function
2022-11-02 18:13:14 +00:00
Dave Bartolomeo
15be488c53 Fix typo 2022-11-02 12:40:55 -04:00
Dave Bartolomeo
a475e5758d Merge remote-tracking branch 'upstream/main' into dbartol/use-workspace-versions 2022-11-02 12:38:03 -04:00
Dave Bartolomeo
7cad4b7918 Revert changes to ATM, which isn't in the workspace 2022-11-02 12:37:30 -04:00
Chris Smowton
197be69425 Merge pull request #11069 from smowton/smowton/fix/kotlin-vararg-constructor-defaults
Kotlin vararg constructors: don't call a default-parameter constructor just because there are no varargs
2022-11-02 16:08:45 +00:00
Tamás Vajk
b59cb778ef Merge pull request #11077 from tamasvajk/kotlin-code-quality
Kotlin: Remove code duplication
2022-11-02 16:00:55 +01:00
Asger F
bac573bbed QL4QL: Add DependencyPath.ql query 2022-11-02 15:46:12 +01:00
Ian Lynagh
c2339b1203 Merge pull request #11050 from igfoo/igfoo/compiler_arguments
Kotlin: Run compiler_arguments test on all platforms
2022-11-02 14:07:36 +00:00
Ian Lynagh
285ff54853 Merge pull request #11052 from igfoo/igfoo/default-parameter-mad-flow
Kotlin: Run default-parameter-mad-flow on all platforms
2022-11-02 14:07:27 +00:00
Ian Lynagh
60fe5d6428 Merge pull request #11054 from igfoo/igfoo/gradle_kotlinx_serialization
Kotlin: Reunite the parts of gradle_kotlinx_serialization
2022-11-02 14:07:10 +00:00
Chris Smowton
cad268476c Accept test changes re: method source-locs 2022-11-02 14:00:14 +00:00
Chris Smowton
d704795d90 Use versioned all-overrides function 2022-11-02 13:57:53 +00:00
Chris Smowton
d700fddfdd Kotlin: handle default parameter values inherited from an overridden function 2022-11-02 13:57:53 +00:00
Jean Helie
37869e816b Merge pull request #11067 from github/jhelie/add-atm-model-integration-tests-hello-world
ATM: add hello world workflow for development of model integration te…
2022-11-02 14:20:51 +01:00
Tamas Vajk
1d3f4826a9 Fix failing internal checkß 2022-11-02 13:57:11 +01:00
Jean Helie
45320d91d1 Merge pull request #11065 from github/jhelie/add-check-to-run-atm-queries
ATM: add PR check running boosted queries
2022-11-02 12:28:40 +01:00
Chris Smowton
39520e54ea Restore enum class exclusion from constructor default parameters 2022-11-02 11:28:13 +00:00
Tamas Vajk
145e2093f3 Kotlin: Remove code duplication 2022-11-02 11:46:33 +01:00
Tom Hvitved
f603d96f48 Merge pull request #11074 from github/revert-10576-ssa/consistency-queries
Revert "SSA: Turn consistency predicates into `query` predicates"
2022-11-02 11:29:42 +01:00
Tom Hvitved
f3e7d8778c Merge pull request #11075 from hvitved/ruby/avoid-ssa-deprecated-predicates
Ruby: Avoid calls to deprecated SSA predicates
2022-11-02 11:29:22 +01:00
Jean Helie
c4b750002e Merge branch 'main' into jhelie/add-atm-model-integration-tests-hello-world 2022-11-02 10:09:56 +01:00
Jean Helie
e8549a413b rename workflow 2022-11-02 10:09:06 +01:00
Tom Hvitved
2d5b9c12a6 Ruby: Avoid calls to deprecated SSA predicates 2022-11-02 09:37:28 +01:00
Tony Torralba
759ffc4743 Merge pull request #11027 from atorralba/atorralba/swift/webview-js-native-bridge-sources
Swift: WebView JS-native bridge sources
2022-11-02 09:32:57 +01:00
Tamás Vajk
79aba19dde Merge pull request #10951 from tamasvajk/kotlin-fix-startoffset
Kotlin: Fix location (start position) of method calls
2022-11-02 09:20:32 +01:00
Tom Hvitved
780ea72b3b Revert "SSA: Turn consistency predicates into query predicates" 2022-11-02 09:11:45 +01:00
Tom Hvitved
28b7ab7fbe Merge pull request #11066 from hvitved/ssa/deprecate-no-uncertain-reads-predicates 2022-11-01 22:06:08 +01:00
Jean Helie
3d1f75221b address @henrymercer's comments 2022-11-01 20:57:58 +01:00
Jean Helie
44bf8184fe ATM: add PR check running the boosted queries 2022-11-01 20:57:58 +01:00
Jean Helie
ce1092c33d add test repo 2022-11-01 20:57:58 +01:00
Chris Smowton
4910bf12e9 Kotlin vararg constructors: don't call a default-parameter constructor just because there are no varargs
This is the constructor case of https://github.com/github/codeql/pull/10909
2022-11-01 19:28:56 +00:00
Dave Bartolomeo
9d5e5e3ee7 ${workspace} all the things 2022-11-01 13:29:05 -04:00
Dave Bartolomeo
49c4c554c4 Merge from main 2022-11-01 13:22:40 -04:00
Tom Hvitved
ee9163aa40 Ruby: Fix flow steps into phi nodes
- Add missing flow from post-update nodes into phi nodes.
- Prevent flow from reads into phi nodes when use-use flow is prohibited.
2022-11-01 16:33:06 +01:00
Tom Hvitved
a191edfbd5 Ruby: Add data flow tests that illustrate problems with flow into SSA phi nodes 2022-11-01 16:32:46 +01:00
Tom Hvitved
640b0ce093 SSA: Deprecate *NoUncertainReads predicates 2022-11-01 15:21:32 +01:00
Jean Helie
1d6db8db9a ATM: add hello world workflow for development of model integration tests workflow 2022-11-01 14:51:11 +01:00
Tom Hvitved
e8f9429b92 Merge pull request #10917 from hvitved/ruby/singleton-call-sensitivity
Ruby: Call-context sensitivity for singleton method calls
2022-11-01 14:13:26 +01:00
Geoffrey White
84c754e007 Merge pull request #11062 from geoffw0/rename
Swift: Rename ECB-Encryption directory
2022-11-01 12:59:53 +00:00
Arthur Baars
aba87a139d Merge pull request #10668 from aibaars/ruby-deps
Ruby: update dependencies
2022-11-01 13:55:42 +01:00
Tom Hvitved
f623ea0a55 Merge pull request #10576 from hvitved/ssa/consistency-queries
SSA: Turn consistency predicates into `query` predicates
2022-11-01 13:07:33 +01:00
Arthur Baars
8b39059d3a Update .github/workflows/ruby-build.yml 2022-11-01 10:49:12 +01:00
Geoffrey White
a1edd65542 Merge pull request #11034 from geoffw0/global
Swift: Add and use AbstractFunctionDecl.hasGlobalName predicate.
2022-11-01 09:27:26 +00:00
Geoffrey White
c3577b2256 Swift: Rename test directory. 2022-11-01 09:21:50 +00:00
Geoffrey White
7d80c5c7f7 Swift: Rename query directory. 2022-11-01 09:21:10 +00:00
Tom Hvitved
4edef874d6 SSA: Turn consistency predicates into query predicates 2022-11-01 10:01:56 +01:00
erik-krogh
84a7fddd95 remove explicit versions in lock files, as the dependencies are all installed locally 2022-11-01 09:09:26 +01:00
Erik Krogh Kristensen
ff2a5e8c27 Merge pull request #10986 from erik-krogh/tsPerf
JS: push more context into load/store steps from the exploratory flow-analysis
2022-11-01 09:03:24 +01:00
Erik Krogh Kristensen
994c033c62 Merge pull request #11049 from erik-krogh/noCrossTalk
QL: prevent some cross-talk between modules
2022-10-31 21:41:00 +01:00
Tamas Vajk
91972d1d1f Kotlin: Fix location (start position) of method calls 2022-10-31 21:08:15 +01:00
Jeroen Ketema
222c9a6357 Merge pull request #11048 from jketema/add-constant-test
C++: Add `strcpy` test for `cpp/non-constant-format`
2022-10-31 18:01:13 +01:00
Ian Lynagh
caa56c9cad Kotlin: compiler_arguments test: Normalise path separators in output
Otherwise we get different output on Windows vs Posix.
2022-10-31 16:31:08 +00:00
Jeroen Ketema
b43cbf7f95 Update cpp/ql/test/query-tests/Likely Bugs/Format/NonConstantFormat/test.cpp
Co-authored-by: Mathias Vorreiter Pedersen <mathiasvp@github.com>
2022-10-31 17:03:29 +01:00
Ian Lynagh
067704a59a Kotlin: default-parameter-mad-flow test: Make kotlinc call work on Windows 2022-10-31 15:57:34 +00:00
Ian Lynagh
2783668092 Kotlin: compiler_arguments test: Fix running gradle on Windows 2022-10-31 15:29:23 +00:00
erik-krogh
86e81f523c add explicit this 2022-10-31 15:56:01 +01:00
Ian Lynagh
6b5deee9a8 Kotlin: Reunite the parts of gradle_kotlinx_serialization 2022-10-31 14:55:58 +00:00
Ian Lynagh
7c4df8c81f Kotlin: Run default-parameter-mad-flow on all platforms 2022-10-31 14:51:15 +00:00
erik-krogh
7d0beeafad QL: prevent some cross-talk between modules 2022-10-31 15:51:05 +01:00
Ian Lynagh
b1d9f9f9d5 Kotlin: Run compiler_arguments test on all platforms 2022-10-31 14:46:46 +00:00
Jeroen Ketema
abe9258943 C++: Add strcpy test for cpp/non-constant-format 2022-10-31 15:29:17 +01:00
Jeroen Ketema
83afc2a0ad C++: Add strcpy prototype to test 2022-10-31 15:25:35 +01:00
Ian Lynagh
21600c612d Merge pull request #11037 from igfoo/igfoo/log
Kotlin: Integration tests: Allow \ as a path separator in logs test
2022-10-31 14:03:57 +00:00
Tamás Vajk
e356720c73 Merge pull request #11014 from tamasvajk/kotlin-for-loop-var
Kotlin: Resugar `for` loops
2022-10-31 14:48:21 +01:00
Chris Smowton
d959630991 Merge pull request #11040 from smowton/smowton/fix/inherited-method-implementing-collection-interface
Kotlin: fix method types when an inherited method implements a collection type
2022-10-31 12:40:28 +00:00
Ian Lynagh
a53c2104d1 Kotlin: Integration tests: Allow \ as a path separator in logs test 2022-10-31 11:24:39 +00:00
Tony Torralba
b62ede1544 Fix issue in JsExportedSource
Model the source as an access to the tainted field, instead of the field itself (which didn't work)
2022-10-31 12:08:03 +01:00
Chris Smowton
c11d63e4d2 Merge pull request #11015 from smowton/smowton/fix/go-cleartext-logging-exclude-protobuf-getters
Go: exclude protobuf read steps from cleartext-logging query
2022-10-31 10:43:52 +00:00
Geoffrey White
ca586b4f3d Merge remote-tracking branch 'upstream/main' into global 2022-10-31 10:28:29 +00:00
Geoffrey White
0dd8f574a7 Swift: Redesign as a FreeFunctionDecl class + add some qldoc. 2022-10-31 10:24:12 +00:00
Geoffrey White
c161bb5e95 Merge pull request #11035 from geoffw0/simplify2
Swift: Simplify some more QL
2022-10-31 09:50:55 +00:00
Rasmus Wriedt Larsen
ead0844174 Merge pull request #10998 from RasmusWL/essa-use-use-test
Python: Add failing ESSA use-use test
2022-10-31 10:38:26 +01:00
Tamas Vajk
4cd0f1ca66 Apply code review findings 2022-10-31 08:43:53 +01:00
Harry Maclean
3f403f0f87 Merge pull request #10700 from hmac/activesupport
Ruby: Model some ActiveSupport methods
2022-10-31 11:50:44 +13:00
Chris Smowton
b370497f96 Avoid split overrides 2022-10-29 18:23:45 +01:00
Chris Smowton
3573e211cc Correct test expectations 2022-10-29 11:40:58 +01:00
Chris Smowton
b6e4f472d1 Remove unnecessary import 2022-10-29 11:40:57 +01:00
Chris Smowton
6d321e0151 Add change note 2022-10-29 11:40:57 +01:00
Chris Smowton
5c66d87ed6 gofmt 2022-10-29 11:40:57 +01:00
Chris Smowton
0c6c135967 Go: exclude protobuf read steps from cleartext-logging query
This query already treats structs differently to usual: it includes field -> whole struct taint steps, but explicitly excludes struct -> field steps. This means that a logging framework sinking an entire struct with a tainted field yields an alert, but we don't get FPs caused by writing field `x` but then reading field `y`.

However, protobuf messages have a special treatment, with taint usually associated with the whole struct and getter methods propagating that taint out. Suppressing these getter method steps specifically for the cleartext-logging query mirrors its treatment of structs in general and avoids this sort of field-mismatch FP.

On the downside we will miss same-field propagation like `m.field = password; Log(m.GetField())` if we don't have source code for the implementation of `m`. However this is hopefully unusual since the typical use of protobufs is to serialize and deserialize, rather than using the struct as a general-purpose datastructure.
2022-10-29 11:40:57 +01:00
Chris Smowton
8266a22332 Kotlin: fix method types when an inherited method implements a collection type
In this circumstance the compiler seems to generate a specialised version of the implementing function with its argument type replaced by the interface-implementing child class' type parameter. However it stores a back-pointer to the real declared function, which we should use as the call target.
2022-10-29 11:29:04 +01:00
Dave Bartolomeo
85790fcade Merge pull request #10964 from smowton/smowton/admin/modernise-qlpacks
qlpacks: libraryPathDependencies -> dependencies
2022-10-28 16:44:22 -04:00
Chris Smowton
d9744c81b7 Merge pull request #11017 from smowton/smowton/fix/kotlin-wildcard-suppression-annotation
Kotlin: fix wildcard suppression where the annotation applies to a parent type/argument.
2022-10-28 18:33:07 +01:00
Ian Lynagh
84427e132e Kotlin: Move the logs test to all-platforms 2022-10-28 17:56:41 +01:00
Geoffrey White
f122005aaf Swift: Simplify out some variables. 2022-10-28 17:26:17 +01:00
Tony Torralba
2402504a4c Add missing SummaryPostUpdateNode 2022-10-28 18:24:17 +02:00
Geoffrey White
b4d939a620 Swift: Correct a comment. 2022-10-28 17:11:24 +01:00
Chris Smowton
f9e811bddf Legacy support qlpacks: continue using libraryPathDependencies; add a comment noting this is obsolete. 2022-10-28 16:47:30 +01:00
Chris Smowton
1914a114a2 Merge pull request #11018 from smowton/smowton/fix/kotlin-extension-specialisation
Kotlin: specialise extension receivers the same as other function parameters
2022-10-28 16:15:41 +01:00
Chris Smowton
d6e2f5f4a8 Use ?.not() to negate a nullable boolean 2022-10-28 16:13:55 +01:00
Chris Smowton
1e1c9f639c Avoid Kotlin 1.5+ function firstNotNullOfOrNull 2022-10-28 16:13:55 +01:00
Chris Smowton
24f87ac963 Kotlin: fix wildcard suppression where the annotation applies to a parent type/argument.
In the process I also fix the missed case where suppression can be switched off using a parameterized annotation.
2022-10-28 16:13:55 +01:00
Ian Lynagh
2796d60d79 Merge pull request #11019 from igfoo/igfoo/win_integ
Kotlin: Get some integration tests running on Windows
2022-10-28 16:12:15 +01:00
Chris Smowton
5ad5cdce47 Swift integration-test runner: use --additional-packs 2022-10-28 16:07:38 +01:00
Chris Smowton
ee63e60bb7 qlpacks: libraryPathDependencies -> dependencies 2022-10-28 16:07:36 +01:00
Geoffrey White
648c2d09f9 Swift: Simplify InsecureTLS.ql. 2022-10-28 15:56:03 +01:00
Tony Torralba
baf7986cfa Rework types exported through JSContext
Better model the JSExport protocol logic
2022-10-28 15:56:05 +02:00
Rasmus Wriedt Larsen
a04c78ab94 Python: Apply suggestions from code review
Co-authored-by: yoff <lerchedahl@gmail.com>
2022-10-28 15:31:42 +02:00
Ian Lynagh
49425e6c2a Kotlin: Integration tests: Make a couple more posix-only for now 2022-10-28 13:59:36 +01:00
Geoffrey White
cf9c3afc86 Swift: Add and use AbstractFunctionDecl.hasGlobalName predicate. 2022-10-28 13:57:24 +01:00
Tamas Vajk
d745381ebe Remove unneeded consistency test output 2022-10-28 14:56:25 +02:00
Tamas Vajk
803a97df7f Kotlin: Resugar for loops with tuples as loop variables 2022-10-28 14:55:50 +02:00
Tamas Vajk
841340b266 Kotlin: Resugar for loops 2022-10-28 14:55:50 +02:00
Tamas Vajk
1e3060598f Kotlin: Add for loop tests 2022-10-28 14:55:50 +02:00
Tamás Vajk
caf9ac50d9 Merge pull request #11026 from tamasvajk/kotlin-remove-kotlin-java-eq-test
Kotlin: Remove `javaEquivalent` consistency query
2022-10-28 14:08:53 +02:00
Chris Smowton
366410ee9e Fix incorrect parameter ordering 2022-10-28 12:58:23 +01:00
Mathias Vorreiter Pedersen
0a3d0c4f56 Merge pull request #11031 from geoffw0/simplify
Swift: Simplify queries using MethodDecl.hasQualifiedName
2022-10-28 13:58:08 +02:00
Tony Torralba
48b0cc0229 Add models for JSContext and JSValue 2022-10-28 13:01:25 +02:00
Tony Torralba
81701547b2 Add taint sources for WKScriptMessage
This is what contains externally-provided data in Webview JS-native bridges
2022-10-28 12:58:27 +02:00
Mathias Vorreiter Pedersen
142e50008e Merge pull request #10967 from MathiasVP/fix-swift-summary
Swift: Fix flow out of summarized callables
2022-10-28 12:57:52 +02:00
Geoffrey White
368f37a27e Swift: And another. 2022-10-28 11:46:27 +01:00
Geoffrey White
1f3ed1cec7 Merge remote-tracking branch 'upstream/main' into simplify 2022-10-28 11:42:05 +01:00
Geoffrey White
6fca350714 Use MethodDecl.hasQualifiedName. 2022-10-28 11:41:42 +01:00
AlexDenisov
ce441ade63 Merge pull request #11028 from github/redsun82/swift-filesystem
Swift: fix remapping
2022-10-28 12:11:26 +02:00
Mathias Vorreiter Pedersen
062a0abceb Swift: Fix flow out of summarized callables. 2022-10-28 12:09:05 +02:00
Ian Lynagh
f387eb21eb Kotlin: Integration tests: Add a qlpack.yml 2022-10-28 10:53:45 +01:00
Ian Lynagh
382c08e3cd Kotlin: Fix some integrations tests on Windows 2022-10-28 10:53:45 +01:00
Ian Lynagh
15d5369bdd Kotlin: Run some integration tests on Windows too 2022-10-28 10:53:45 +01:00
Tamás Vajk
8bc46d5e56 Merge pull request #11025 from tamasvajk/kotlin-fix-external-location
Kotlin: Fix external location in integration test
2022-10-28 11:33:25 +02:00
Paolo Tranquilli
a87495226a Swift: fix remapping
With the change to `std::filesystem` some path concatenations were
translated to appending, which is not the same. In case rhs is absolute
`lhs / rhs == rhs`, while concatenating treats `rhs` as if it was
relative. The same behaviour can be obtained in `std::filesystem` by
using `lhs / rhs.relative_path()`.
2022-10-28 11:16:49 +02:00
Tamas Vajk
7ceadb0df0 Kotlin: Remove javaEquivalent consistency query
The `javaEquivalent` consistency query is no longer needed, as the `diags` query is now a superset of it.
2022-10-28 10:44:42 +02:00
Rasmus Wriedt Larsen
8628ff5e52 Merge pull request #10999 from RasmusWL/inline-fail-tag
InlineExpectationsTest: Fail if missing `getARelevantTag`
2022-10-28 10:35:49 +02:00
Mathias Vorreiter Pedersen
95a54f79d8 Merge pull request #10938 from geoffw0/printfprecision
C++: Fix printf.qll bug
2022-10-28 10:33:58 +02:00
Tamas Vajk
99880c980c Kotlin: Fix external location in integration test 2022-10-28 10:24:14 +02:00
Jeroen Ketema
4ca0838815 Merge pull request #11009 from RasmusWL/dataflow-label
Misc: Add automatic `DataFlow Library` label
2022-10-28 09:58:33 +02:00
Mathias Vorreiter Pedersen
22cdeec3fb Merge branch 'main' into printfprecision 2022-10-28 09:29:29 +02:00
Erik Krogh Kristensen
93fb2930c8 Merge pull request #10968 from erik-krogh/fixRbCode
RB: fix rb/code-injection
2022-10-28 09:14:14 +02:00
Harry Maclean
368ce69198 Fix qldoc formatting 2022-10-28 11:31:55 +13:00
Harry Maclean
9df8edcb1c Ruby: fix formatting 2022-10-28 11:31:55 +13:00
Harry Maclean
cd34686967 Ruby: Document flow summary for Hash#extract! 2022-10-28 11:31:55 +13:00
Harry Maclean
ca7b48c3d5 Add change note 2022-10-28 11:31:55 +13:00
Harry Maclean
5e781f24b6 Ruby: Remove duplicate test
This is already tested in hash-flow.
2022-10-28 11:31:55 +13:00
Harry Maclean
4ec527a9ea Ruby: Explain difference between flow tests
The type-tracking flow tests document the difference in sensitivity
between type-tracking and dataflow, so failures in that test are
expected.
2022-10-28 11:31:55 +13:00
Harry Maclean
6e8446b6ae Fix tests 2022-10-28 11:31:55 +13:00
Harry Maclean
ef260db76e Fix singleton set literal 2022-10-28 11:31:55 +13:00
Harry Maclean
71d703f2a5 Ruby: Add ActiveSupport extensions 2022-10-28 11:31:55 +13:00
Harry Maclean
cb37a0e835 Ruby: Add summaries for Hash#deep_merge(!) 2022-10-28 11:31:55 +13:00
Harry Maclean
3dea1d6a60 Ruby: Add flow summary for Hash#except! 2022-10-28 11:31:55 +13:00
Harry Maclean
0454642220 Ruby: Model deep_dup and presence 2022-10-28 11:31:55 +13:00
Harry Maclean
9f260853ac Ruby: Model more ActiveSupport string extensions 2022-10-28 11:31:55 +13:00
Harry Maclean
b389d50943 Ruby: Identify safe_constantize 2022-10-28 11:31:54 +13:00
Rasmus Wriedt Larsen
2c7570e971 Apply suggestions from code review
Co-authored-by: Jeroen Ketema <93738568+jketema@users.noreply.github.com>
2022-10-27 22:16:48 +02:00
Geoffrey White
ca279f4073 Merge pull request #10996 from geoffw0/methods
Swift: Add MethodDecl.hasQualifiedName
2022-10-27 19:18:48 +01:00
Chris Smowton
45a4cd89a6 Kotlin: specialise extension receivers the same as other function parameters
This arises when a generic class extends one of its parameters; for example, `class G<T> { val T.v; get() = 1 }`, where specialisation `G<List>` should generate a method specialisation `getV(List)`.
2022-10-27 18:31:19 +01:00
Geoffrey White
3507ea3f2a Swift: Autoformat. 2022-10-27 17:37:21 +01:00
Henry Mercer
19b7e9ebc7 Merge pull request #10997 from github/henrymercer/go-extract-file-locations
Go: Extract locations of successfully extracted files
2022-10-27 16:12:15 +01:00
Chris Smowton
5ef99ca5bd Merge pull request #11003 from smowton/smowton/fix/reintroduce-pointless-wildcards
Kotlin: reintroduce pointless wildcards when a Java declaration explicitly uses them
2022-10-27 16:06:21 +01:00
Ian Lynagh
6533e2ea5c Merge pull request #10976 from igfoo/igfoo/version
Kotlin: Ignore tags when comparing versions
2022-10-27 15:50:39 +01:00
Ian Lynagh
b4242dd913 Merge pull request #11012 from igfoo/igfoo/mkdir
Kotlin: Fix integration tests on Mac
2022-10-27 15:39:10 +01:00
Rasmus Wriedt Larsen
b840e8efb8 Java: Remove MISSING: XssSink annotations from text/plain responses in JaxWs 2022-10-27 15:55:14 +02:00
Erik Krogh Kristensen
bbdda9ef70 Merge pull request #10727 from erik-krogh/js-last-msg
JS: fix some more style-guide violations in the alert-messages
2022-10-27 15:48:12 +02:00
Geoffrey White
d0f45180ab Swift: cache some predicated. 2022-10-27 14:17:51 +01:00
Rasmus Wriedt Larsen
cee9139a0d Java: Correctly annotate missing XSS sinks in JaxWs modeling 2022-10-27 15:17:17 +02:00
Ian Lynagh
f237360d81 Merge pull request #11011 from igfoo/igfoo/modules
Kotlin: Handle /modules/... paths specially too
2022-10-27 13:48:32 +01:00
Ian Lynagh
6c232f95bc Kotlin: Fix integration tests on Mac 2022-10-27 13:48:04 +01:00
Paolo Tranquilli
507e3b35ad Merge pull request #10987 from github/redsun82/swift-filesystem
Swift: use `std::filesystem` and `picoSHA2`
2022-10-27 14:14:37 +02:00
Taus
503cc560cf Merge pull request #10943 from bananabr/main
Javascript/Python: Tokens built from predictable UUIDs
2022-10-27 14:12:34 +02:00
Rasmus Wriedt Larsen
adf109b624 Merge branch 'main' into inline-fail-tag 2022-10-27 13:42:32 +02:00
Ian Lynagh
bafa80667c Kotlin: Handle /modules/... paths specially too
On Windows, we don't want a C: prefix on these either.
2022-10-27 12:24:28 +01:00
Ian Lynagh
6d77b34323 Merge pull request #11000 from igfoo/igfoo/unknown-binary-location
Kotlin: Handle /!unknown-binary-location/... paths specially on Windows
2022-10-27 12:09:32 +01:00
Jeroen Ketema
1d7efd8e82 Merge pull request #10905 from jsoref/spelling-code-scanning-product
Spelling code scanning product
2022-10-27 12:55:37 +02:00
Erik Krogh Kristensen
cecb498bf3 Merge pull request #10984 from tyage/add-next-js-source
JS: Add Next.js parameters as source
2022-10-27 10:36:12 +02:00
Erik Krogh Kristensen
71f29f037a Merge pull request #10988 from erik-krogh/passwrd
JS: remove some FPs in `js/password-in-configuration-file`
2022-10-27 10:34:56 +02:00
Tamás Vajk
a428ab5f73 Merge pull request #11006 from tamasvajk/kotlin-fix-test-1
Kotlin: fix test to expect diagnostic
2022-10-27 10:34:24 +02:00
Paolo Tranquilli
09f549ab38 Merge pull request #11007 from github/redsun82/cmake-generator
Swift: fix cmake generator on Linux
2022-10-27 09:52:55 +02:00
Rasmus Wriedt Larsen
4079223151 Misc: Add automatic DataFlow Library label
Would just be nice for filtering PRs
2022-10-27 09:40:15 +02:00
Paolo Tranquilli
09a51ecdd5 Swift: fix cmake generator on Linux 2022-10-27 09:32:05 +02:00
Paolo Tranquilli
3fca25310f Swift: fix copy option 2022-10-27 09:29:05 +02:00
Paolo Tranquilli
22db4932ee Swift: add overwrite_existing to source archiving 2022-10-27 09:26:57 +02:00
Tamas Vajk
1727fcb845 Kotlin: fix test to expect diagnostic 2022-10-27 09:14:23 +02:00
Rasmus Wriedt Larsen
6d43db43dd Ruby: Fix tag missing from getARelevantTag 2022-10-27 09:12:06 +02:00
Rasmus Wriedt Larsen
977792070a Java: Fix tag missing from getARelevantTag 2022-10-27 09:11:24 +02:00
Rasmus Wriedt Larsen
fc7eb5b4fc InlineExpectationsTest: sync 2022-10-27 09:02:28 +02:00
Rasmus Wriedt Larsen
dbd84b2d37 InlineExpectationsTest: Add quote around missing tag
To aid with quickly scanning where the missing tag is. I just had to do
this myself looking over some test failures, and it all just blurred
into each other in the logs.

see https://github.com/github/codeql/actions/runs/3332266045/jobs/5512944867#step:5:467
2022-10-27 09:02:28 +02:00
Tamás Vajk
f1fcb64e94 Merge pull request #10992 from tamasvajk/kotlin-unused-extension
Kotlin: do not report on unused `object` extension parameters
2022-10-27 08:50:33 +02:00
tyage
c22f9443f2 Refactoring Next.js parameter
Co-authored-by: Erik Krogh Kristensen <erik-krogh@github.com>
2022-10-27 10:28:51 +09:00
tyage
e8b751ae17 Update javascript/ql/src/change-notes/2022-10-26-nextjs-params.md
Co-authored-by: Erik Krogh Kristensen <erik-krogh@github.com>
2022-10-27 10:24:08 +09:00
tyage
ac27307a2b Update javascript/ql/lib/semmle/javascript/frameworks/Next.qll
Co-authored-by: Erik Krogh Kristensen <erik-krogh@github.com>
2022-10-27 10:23:59 +09:00
tyage
54050bf1b6 update test result XssWithAdditionalSources 2022-10-27 10:23:37 +09:00
Harry Maclean
bdb143cf83 Merge pull request #10913 from thiggy1342/expand-ruby-ssrf-sinks-faraday-connection-new
Ruby: Add Faraday::Connection.new as sink for SSRF query
2022-10-27 10:33:44 +13:00
Dave Bartolomeo
23b572e9b7 Use ${workspace} for intra-workspace dependencies
Now that the released CLI supports replacement variables in dependency version ranges, we can now mark our published library packs as depending on whatever version of their dependency is in our workspace, without having to manually bump the dependency version every release.

Note that when the packs are published, the dependencies in the published pack file are rewritten to have the correct specific version.
2022-10-26 16:40:01 -04:00
erik-krogh
2ace10b294 bump the version of the shared pack in the QL-for-QL qlpack.yml file 2022-10-26 22:16:42 +02:00
Chris Smowton
28b6e263ec Kotlin: reintroduce pointless wildcards when a Java declaration explicitly uses them
For example, Java code might use `HasOutVariance<? extends String>`, or `HasInVariance<? super Object>`, both of which are needless wildcards and which the Kotlin extractor would previously have refused to reintroduce due to their not specifying a larger type than their bound. However this led to inconsistency with Java extraction, which
extracts the type as it appears in source.

This seems to particularly happen with generated code, e.g. the output of the Kotlin protobuf compiler.
2022-10-26 20:05:27 +01:00
Daniel Santos
63c71b7d09 Merge branch 'main' into main 2022-10-26 14:05:26 -05:00
Ian Lynagh
0a470b0864 Kotlin: Handle /!unknown-binary-location/... paths specially on Windows
The standard code wants to normalise it to C:/!unknown-binary-location/...
which is particularly annoying for cross-platform test output.
2022-10-26 19:20:32 +01:00
Henry Mercer
c1984ea35f Go: Update expected output 2022-10-26 19:11:21 +01:00
Daniel Santos
64da2cec50 removed unnecessary getACall and fixed formatting 2022-10-26 12:02:55 -05:00
Rasmus Wriedt Larsen
5e9897d150 InlineExpectationsTest: sync 2022-10-26 18:21:13 +02:00
Rasmus Wriedt Larsen
76e84ef63a InlineExpectationsTest: Fail if missing getARelevantTag 2022-10-26 18:20:37 +02:00
Rasmus Wriedt Larsen
bfe9aa1225 InlineExpectationsTest: Add test showing what happens if you leave out getARelevantTag 2022-10-26 18:00:03 +02:00
Rasmus Wriedt Larsen
b3f29b0a53 Python: Add failing ESSA use-use test
I initially created this as a dataflow test, but then realized it could
just be an ESSA test. I cound't find any existing ESSA tests though :|
so created a new dir for it.
2022-10-26 17:49:33 +02:00
Geoffrey White
a32b08f56a Swift: remove redundant line. 2022-10-26 16:39:33 +01:00
Geoffrey White
e981a28b0f Swift: autoformat test. 2022-10-26 16:32:52 +01:00
Henry Mercer
b0b321a16f Go: Standardise formatting 2022-10-26 16:31:08 +01:00
Henry Mercer
4bc8529490 Go: Extract locations of successfully extracted files
Switch the successfully extracted files query to the `location, message` results format so that we get rich location information when exporting the results of this query to SARIF.  Previously the query used the `message` results format, which meant the interpreted results lacked a location.
2022-10-26 16:28:02 +01:00
Geoffrey White
0b3408b1f6 Swift: Fix typo. 2022-10-26 16:24:25 +01:00
Geoffrey White
5d21c51deb Swift: use hasQualifiedName in UnsafeWebViewFetch.ql. 2022-10-26 16:12:29 +01:00
Geoffrey White
0d41d4e90c Swift: for consistancy, lets have a simple hasName function as well. 2022-10-26 16:11:01 +01:00
Geoffrey White
b24a27d4ae Swift: Add hasQualifiedName methods and tests. 2022-10-26 16:03:49 +01:00
Chris Smowton
fac383a3ac Merge pull request #10974 from smowton/smowton/fix/dont-translate-tochar
Kotlin: don't try to call nonexistent `j.l.Number.toChar`
2022-10-26 14:18:03 +01:00
Tamas Vajk
9cc7a30a75 Kotlin: do not report on unused object extension parameters 2022-10-26 15:06:51 +02:00
Tamas Vajk
fbcf7ea669 Kotlin: Add test case for unused extension parameters 2022-10-26 15:05:59 +02:00
Asger F
c9dfba344a Merge pull request #10925 from asgerf/ql/navigate-doc
Docs: Mention new navigation commands
2022-10-26 14:29:42 +02:00
Ian Lynagh
37c40c58d2 Merge pull request #10959 from igfoo/igfoo/diags
Java/Kotlin: Add a diagnostics consistency query
2022-10-26 13:07:01 +01:00
Paolo Tranquilli
521e6235b5 Swift: use std::filesystem and picoSHA2
This replaces usages of `llvm::fs` and string manipulation with
`std::filesystem`, also replacing `std::string` with
`std::filesystem::path` where it made sense.

Moreover MD5 hashing used in macOS file remapping was replaced by
SHA256 hashing using a small header-only SHA256 C++ library with an
MIT license, https://github.com/okdshin/PicoSHA2.

File contents hashing was relocated to the newly created `file` library
for later planned reuse.
2022-10-26 13:23:44 +02:00
Ian Lynagh
dd7ec499df Kotlin: Ignore tags when comparing versions
We thought that 1.7.20-Beta > 1.7.20, and so tried to use 1.7.0's
extractor with 1.7.20.
2022-10-26 12:21:55 +01:00
erik-krogh
0f9b4334cc remove some FPs in js/password-in-configuration-file 2022-10-26 11:51:56 +02:00
Paolo Tranquilli
e422a4eef9 Swift: move TargetFile to a separate lib 2022-10-26 10:54:51 +02:00
erik-krogh
21e7e27e1f push more context into load/store steps from the exploratory flow-analysis 2022-10-26 10:52:47 +02:00
Erik Krogh Kristensen
52cd200ca0 Merge pull request #10985 from asgerf/js/reaches-return-escape
JS: Do not track returned values out of the enclosing function
2022-10-26 10:52:11 +02:00
Tony Torralba
924995d9e1 Merge pull request #10977 from github/workflow/coverage/update
Update CSV framework coverage reports
2022-10-26 09:51:17 +02:00
Asger F
414bd40c41 JS: Do not track returned values out of the enclosing function 2022-10-26 09:29:49 +02:00
Paolo Tranquilli
a3234503b8 Merge pull request #10983 from github/redsun82/swift-third-party
Swift: move libraries from `tools` to `third_party`
2022-10-26 08:59:50 +02:00
Mathias Vorreiter Pedersen
58b6c45d27 Merge pull request #10958 from geoffw0/comma
C++: Fix performance issue on cpp/comma-before-misleading-indentation
2022-10-26 08:29:18 +02:00
tyage
7a19744cf2 add change note 2022-10-26 15:17:50 +09:00
tyage
95dca7c3ed update comment 2022-10-26 15:13:59 +09:00
tyage
09f8ca8cc0 add query in comment 2022-10-26 15:13:03 +09:00
tyage
232893aafa make query parameters in ServerSideProps and next/router
as a RemoteFlowSource
2022-10-26 14:41:07 +09:00
Paolo Tranquilli
c8788bb5cd Swift: move libraries from tools to third_party 2022-10-26 07:05:56 +02:00
tyage
1f4fc7fc2d add params, query to test 2022-10-26 10:53:11 +09:00
tyage
06925681b0 add test for context.params 2022-10-26 10:53:11 +09:00
github-actions[bot]
5454f9a738 Add changed framework coverage reports 2022-10-26 00:20:29 +00:00
Daniel Santos
f7ace6f801 Update javascript/ql/src/experimental/Security/CWE-340/TokenBuiltFromUUID.ql
Co-authored-by: Erik Krogh Kristensen <erik-krogh@github.com>
2022-10-25 14:27:03 -05:00
Geoffrey White
1e8b4bdd6f Merge pull request #10973 from geoffw0/comment
Swift: Fix UrlRemoteFlowSource name clash
2022-10-25 18:51:51 +01:00
thiggy1342
9c1fbfd330 Merge branch 'main' into expand-ruby-ssrf-sinks-faraday-connection-new 2022-10-25 13:09:17 -04:00
Chris Smowton
004f4be5fb Kotlin: don't try to call nonexistent j.l.Number.toChar
Previously we thought this could be callable because Kotlin's view of `j.l.Integer` inherits `k.Number` which defines `toChar`.
2022-10-25 17:09:05 +01:00
Geoffrey White
53fa91f8ba Swift: Add comment. 2022-10-25 16:51:57 +01:00
Daniel Santos
feece6f7b4 Merge branch 'github:main' into main 2022-10-25 10:43:20 -05:00
Geoffrey White
a67bd4d903 Swift: Fix name clash. 2022-10-25 16:40:27 +01:00
Ian Lynagh
4050801a17 Kotlin: Autoformat query 2022-10-25 16:26:12 +01:00
Ian Lynagh
52cfc33576 Kotlin: Accept test changes 2022-10-25 16:26:12 +01:00
Ian Lynagh
63b64e4daa Kotlin: Test tweaks for the diags consistency query 2022-10-25 16:26:11 +01:00
Ian Lynagh
caf7ebc634 Java/Kotlin: Add a diagnostic consistency query 2022-10-25 16:26:11 +01:00
Ian Lynagh
185d43a7b0 Kotlin: Turn warnings into trace messages
This is normal behaviour, nothing to be concerned about.
2022-10-25 16:26:11 +01:00
Daniel Santos
5b080481aa TokenBuiltFromUuid formatting 2022-10-25 09:51:48 -05:00
Daniel Santos
b8d60edb49 TokenBuiltFromUuid isAdditionalTaintStep refactor 2022-10-25 09:51:07 -05:00
Daniel Santos
375edf7455 TokenAssignmentValueSink refactor 2022-10-25 09:50:04 -05:00
thiggy1342
3659eaa780 add markdown file extension 2022-10-25 10:13:19 -04:00
Geoffrey White
257748d82b C++: Rename predicate. 2022-10-25 14:52:22 +01:00
Geoffrey White
3d025ea77e Merge pull request #10903 from geoffw0/review
Swift: Add some summary queries.
2022-10-25 14:47:09 +01:00
yo-h
01a67adb49 Merge pull request #10738 from github/yo-h-patch-1
Java: update framework list
2022-10-25 09:42:18 -04:00
Tamás Vajk
3264bbc1db Merge pull request #10962 from tamasvajk/kotlin-unreachable-catch
Kotlin: Exclude .kt files from `java/unreachable-catch-clause`
2022-10-25 15:01:25 +02:00
Tamás Vajk
7013663d13 Merge pull request #10881 from tamasvajk/kotlin-constant-expr
Kotlin: Exclude constructs in serialization constructors from `java/evaluation-to-constant`
2022-10-25 15:00:58 +02:00
erik-krogh
e8dce25cc2 fix rb/code-injection 2022-10-25 14:44:23 +02:00
Geoffrey White
b59f01f968 Swift: Use UnknownFile. 2022-10-25 13:44:13 +01:00
Mathias Vorreiter Pedersen
6a7bcd384a Merge pull request #10939 from rdmarsh2/rdmarsh2/cpp/modulus-analysis-comments
C++: additional comments for modulus analysis
2022-10-25 14:29:54 +02:00
Tamas Vajk
80fa45fd8e Fix expected fest file after rebase 2022-10-25 13:52:25 +02:00
Tamas Vajk
f1e6b756e3 Add integration test with constant expression in a generated constructor 2022-10-25 13:52:25 +02:00
Tamas Vajk
eaa04b72f1 Apply code review findings 2022-10-25 13:49:54 +02:00
Tamas Vajk
78c23c2657 Kotlin: Exclude constructs in serialization constructors from java/evaluation-to-constant 2022-10-25 13:49:54 +02:00
Tamás Vajk
30fc6acb19 Merge pull request #10961 from tamasvajk/kotlin-abstract-collection-cast
Kotlin: Improve `java/abstract-to-concrete-cast` to handle `when` branches
2022-10-25 13:27:19 +02:00
Henry Mercer
7e2c06de80 Merge pull request #10963 from github/codeql-ci/js/ml-powered-pack-release-0.3.6
JS: Bump version numbers of ML-powered packs after 0.3.6 release
2022-10-25 12:12:15 +01:00
yoff
9d542f1be9 Merge pull request #10887 from Sim4n6/TarSlipImprov
Python: Add TarSlip Improv query
2022-10-25 13:02:52 +02:00
Chris Smowton
b9f4856d47 Merge pull request #10876 from smowton/smowton/feature/kotlin-default-method-auto-mad
Java models-as-data: infer Kotlin $default models from that of its parent function
2022-10-25 11:58:54 +01:00
Paolo Tranquilli
a4258ea390 Merge pull request #10953 from github/redsun82/cmake-generator
Bazel/CMake: make cmake runnable from outside the workspace
2022-10-25 12:47:10 +02:00
Chris Smowton
c4ba644dfd Merge pull request #10952 from smowton/smowton/fix/java-interface-redeclares-tostring
Kotlin: extract interface redeclarations of `Object` methods
2022-10-25 11:29:10 +01:00
Henry Mercer
1dc14bcaee Merge branch 'main' into codeql-ci/js/ml-powered-pack-release-0.3.6 2022-10-25 10:54:08 +01:00
github-actions[bot]
caf3a098c8 JS: Bump version of ML-powered library and query packs to 0.3.7 2022-10-25 09:12:00 +00:00
Tamas Vajk
0e4287e378 Kotlin: Exclude .kt files from java/unreachable-catch-clause 2022-10-25 11:06:51 +02:00
Tamas Vajk
4b042f9770 Kotlin: Add test cases for java/unreachable-catch-clause 2022-10-25 11:06:15 +02:00
github-actions[bot]
5d100c8036 JS: Bump patch version of ML-powered library and query packs 2022-10-25 09:00:40 +00:00
Tamas Vajk
a0490f454b Kotlin: Improve java/abstract-to-concrete-cast to handle when branches 2022-10-25 10:17:47 +02:00
Tamas Vajk
690d6517d7 Kotlin: Add abstract to concrete type cast guarded by when 2022-10-25 10:16:40 +02:00
Daniel Santos
3051903037 Merge branch 'github:main' into main 2022-10-24 15:47:11 -05:00
Philip Ginsbach
b9f1cc5c6f Merge pull request #10929 from github/ginsbach/TypeSignatureDocumentation
documentation for type signature members
2022-10-24 20:41:25 +01:00
Chris Smowton
d171decad7 Accept test changes
All of java.util.{Collection,List,Map} redeclare `boolean equals(Object)` in order to add documentation, as a side-effect creating a real symbol that can be used as a dispatch target.
2022-10-24 19:49:29 +01:00
Chris Smowton
843f847960 Merge pull request #10921 from smowton/smowton/fix/ignore-enhanced-nullability
Kotlin: ignore enhanced nullability when extracting primitive types
2022-10-24 19:43:04 +01:00
Chris Smowton
7a0bded2ac Kotlin: support argument-range specifications for $default methods 2022-10-24 19:31:03 +01:00
Chris Smowton
8d10b1b77b Convert test to inline-expectation test 2022-10-24 19:15:35 +01:00
Paolo Tranquilli
89ca7e26fe Merge pull request #10955 from github/redsun82/swift-fix-missing-bodies
Swift: fix missing extraction of function bodies in SPM builds
2022-10-24 20:00:02 +02:00
Geoffrey White
6f77e14aef C++: Fix rare performance issue on cpp/comma-before-misleading-indentation. 2022-10-24 18:21:10 +01:00
Erik Krogh Kristensen
ef5132b0ae Merge pull request #10883 from erik-krogh/codeSink
RB: don't flag code-injection for dynamic loading where an attacker only controls a substring
2022-10-24 18:59:36 +02:00
Daniel Santos
5ab068a3cc Update python/ql/src/experimental/Security/CWE-340/TokenBuiltFromUUID.ql
Co-authored-by: Taus <tausbn@github.com>
2022-10-24 11:55:21 -05:00
Daniel Santos
be8780742b Update python/ql/src/experimental/Security/CWE-340/TokenBuiltFromUUID.ql
You are totally right! I just scanned the module's document and assumed it would implement it all. Pasting the documentation here for future reference https://docs.python.org/3/library/uuid.html?highlight=uuid#uuid.UUID.

Co-authored-by: Taus <tausbn@github.com>
2022-10-24 11:49:17 -05:00
Paolo Tranquilli
d419749eb2 Swift: fix missing extraction of function bodies in SPM builds
For some reason `-experimental-skip-non-inlinable-function-bodies-without-types`
is passed to the frontend, which will skip extraction of most bodies.

By suppressing that option the problem goes away.
2022-10-24 17:11:13 +02:00
Paolo Tranquilli
6651c9447e Swift: failing test for extracting function bodies 2022-10-24 17:10:38 +02:00
Daniel Santos
a2ad924376 Minor formatting fixes 2022-10-24 09:38:17 -05:00
thiggy1342
952ad6ea46 Merge branch 'main' into expand-ruby-ssrf-sinks-faraday-connection-new 2022-10-24 09:52:24 -04:00
Tony Torralba
2148e8be4d Merge pull request #10892 from atorralba/atorralba/swift/customurlschemes
Swift: Add a new Custom URL Scheme source
2022-10-24 15:33:27 +02:00
Tony Torralba
30f5fb6d83 Update expectations after merge 2022-10-24 14:24:13 +02:00
Paolo Tranquilli
1866a98c77 Bazel/CMake: fix typo in README.md 2022-10-24 14:19:11 +02:00
Erik Krogh Kristensen
5ff98cd80e Merge pull request #10888 from erik-krogh/glob
Ruby: add model for Dir.glob and other Dir methods
2022-10-24 14:17:37 +02:00
Paolo Tranquilli
f49f6430a1 Bazel/CMake: make cmake runnable from outside the workspace
Also added a small `README.md` file.
2022-10-24 14:15:45 +02:00
Asger F
bcfe4ece6f Merge pull request #10918 from asgerf/rb/constant-compound-assignment
Ruby: handle compound constant-assignment
2022-10-24 14:07:28 +02:00
Asger F
cac2e2e2e4 Merge pull request #10928 from asgerf/rb/assumed-global-const
Ruby: assume some global constants are defined
2022-10-24 14:06:34 +02:00
Chris Smowton
c6f4742f29 Kotlin: extract interface redeclarations of Object methods
Due to a probable compiler bug (?) the redeclaration looks like a fake symbol, leading to Java dispatching against a declaration that Kotlin doesn't believe exists.
2022-10-24 12:45:07 +01:00
Paolo Tranquilli
22adf21dd3 Merge pull request #10912 from jketema/templ-func-prototype
C++: Update test result after extractor changes
2022-10-24 13:44:02 +02:00
Tamás Vajk
1d2087b92a Merge pull request #10949 from tamasvajk/kotlin-underscore-var
Kotlin: exclude Kotlin files from `java/underscore-identifier`
2022-10-24 13:32:49 +02:00
Asger F
0ffb0f6d4d Ruby: constant lookup is unaffected by blocks 2022-10-24 13:07:21 +02:00
Chris Smowton
86e99c497d Merge pull request #10930 from smowton/smowton/fix/external-property-overloads
Kotlin: give external extension properties with matching name and file distinct trap filenames
2022-10-24 11:32:37 +01:00
erik-krogh
07d90b34df use instanceof in DirPathAccess 2022-10-24 12:05:26 +02:00
Erik Krogh Kristensen
669b0c35fe fix qldoc
Co-authored-by: Nick Rolfe <nickrolfe@github.com>
2022-10-24 12:05:26 +02:00
erik-krogh
85cd7f9121 add model for Dir.glob and other Dir methods 2022-10-24 12:05:26 +02:00
Tony Torralba
f523fbc9d0 Merge branch 'main' into atorralba/swift/customurlschemes 2022-10-24 11:41:50 +02:00
Tony Torralba
3973e1ce04 Update swift/ql/test/library-tests/dataflow/dataflow/FlowConfig.qll
Co-authored-by: Geoffrey White <40627776+geoffw0@users.noreply.github.com>
2022-10-24 11:37:51 +02:00
Chris Smowton
88c4a2f6e2 Merge pull request #10936 from smowton/smowton/fix/internal-constructor-called-from-java
Kotlin: make internal constructors' trap labels consistent with the Java extractor
2022-10-24 09:37:00 +01:00
Paolo Tranquilli
90d4861b70 Merge pull request #10875 from github/redsun82/swift-codegen-doc
Swift: add infrastructure for documenting generated code
2022-10-24 10:14:57 +02:00
Tony Torralba
80f7d58fae Add missing tests for not-quite-working flow steps 2022-10-24 09:37:22 +02:00
Paolo Tranquilli
fd226c51c1 Merge pull request #10924 from github/redsun82/swift-fix-qltest-failure-reporting
Swift: add qltest tests and fix its failure reporting
2022-10-24 09:34:12 +02:00
Paolo Tranquilli
15e5faf5b6 Merge branch 'main' into templ-func-prototype 2022-10-24 09:19:46 +02:00
Tamas Vajk
d585839b7e Kotlin: exclude Kotlin files from java/underscore-identifier 2022-10-24 09:05:28 +02:00
Tamas Vajk
0192ae8331 Kotlin: Add test case for variables named underscore 2022-10-24 09:04:54 +02:00
Daniel Santos
066ffb7520 Tokens built from predictable UUIDs 2022-10-22 11:15:43 -05:00
Arthur Baars
b3855b089a Ruby: some more tests 2022-10-22 14:15:29 +02:00
ALJI Mohamed
92a3846102 Fix query to omit sinks within std lib files 2022-10-22 09:35:55 +01:00
ALJI Mohamed
fdbed2a019 Add expected test results without considering inStdLib files. 2022-10-22 09:34:57 +01:00
Arthur Baars
ccaa12998d Ruby: desugar compound constant-assignments 2022-10-22 01:11:35 +02:00
ALJI Mohamed
0f44268038 Add expected test results 2022-10-21 22:14:55 +01:00
ALJI Mohamed
7d60f1f1c8 Modified the QL ref file and add TarSlip examples 2022-10-21 22:14:00 +01:00
ALJI Mohamed
7319052495 Delete the examples/ 2022-10-21 21:47:00 +01:00
ALJI Mohamed
31a6fb4181 Add TarSlip qlref for query-tests 2022-10-21 21:28:20 +01:00
Sim4n6
925f9d09e5 Update python/ql/src/experimental/Security/CWE-022bis/TarSlipImprov.ql
Co-authored-by: yoff <lerchedahl@gmail.com>
2022-10-21 21:06:51 +01:00
Nick Rolfe
e5663574fe Merge pull request #10935 from github/nickrolfe/taint-step 2022-10-21 19:28:23 +01:00
Geoffrey White
8a8b1aff7f Swift: Restrict expressions count to expressions with locations. 2022-10-21 18:57:15 +01:00
Geoffrey White
3215295d06 Swift: simpkify SummaryStats.ql description. 2022-10-21 18:48:08 +01:00
Geoffrey White
dfe336cd33 C++: Autoformat tests. 2022-10-21 18:36:12 +01:00
Geoffrey White
2f849b4e77 C++: Change note. 2022-10-21 17:53:11 +01:00
Robert Marsh
88708d015c C++: additional comments for modulus analysis 2022-10-21 12:50:41 -04:00
Geoffrey White
0d030d2b13 C++: Fix FormatLiteral.getMaxConvertedLength bug. 2022-10-21 17:29:55 +01:00
Geoffrey White
06e86accac C++: Add a few cases to the formatLiteral test. 2022-10-21 17:29:28 +01:00
Geoffrey White
c8bf0d03a5 C++: Add formatAttribute test. 2022-10-21 17:29:27 +01:00
Geoffrey White
1376385abb C++: Add formatLiteral test. 2022-10-21 17:29:26 +01:00
Chris Smowton
00800017fd Kotlin: make internal constructors' trap labels consistent with the Java extractor
Previously we accidentally named these something like <init>$main, which is a name-mangling the Kotlin compiler applies to internal methods but not to constructors, which look to Java just like regular public constructors.
2022-10-21 16:48:37 +01:00
Nick Rolfe
9fb436e22b Ruby: add change note for localTaintStep fix 2022-10-21 16:33:29 +01:00
Nick Rolfe
269c27757d Ruby: include value-preserving flow in localTaintStep 2022-10-21 16:17:11 +01:00
Nick Rolfe
5319216c18 Ruby: add test of TaintTracking::localFlowStep 2022-10-21 16:04:04 +01:00
Jean Helie
88c6453fa6 Merge pull request #10934 from github/jhelie/add-hello-world-workflow-atm-queries
ATM: add hello world version of workflow checking queries run
2022-10-21 16:58:44 +02:00
Jean Helie
c0593c945b ATM: add hello world version of workflow checking queries run 2022-10-21 16:37:43 +02:00
Paolo Tranquilli
6bd09b1858 Merge branch 'main' into redsun82/swift-codegen-doc 2022-10-21 15:31:52 +02:00
Chris Smowton
42d6968c20 Kotlin: give external extension properties with matching name and file distinct trap filenames 2022-10-21 14:28:53 +01:00
Paolo Tranquilli
408968a417 Swift: fix swift compilation in QL tests 2022-10-21 15:20:38 +02:00
Asger F
84ae17dcbb Ruby: ensure Object is a transitive superclass 2022-10-21 15:18:59 +02:00
Philip Ginsbach
0dc7123ded documentation for type signature members 2022-10-21 13:42:12 +01:00
Chris Smowton
5e28e5a170 Merge pull request #10909 from smowton/smowton/fix/kotlin-varargs-dataflow
Kotlin: Fix varargs dataflow, and varargs default handling
2022-10-21 13:32:34 +01:00
Paolo Tranquilli
bd62f2be0e Merge branch 'main' into redsun82/swift-fix-qltest-failure-reporting 2022-10-21 14:07:19 +02:00
Paolo Tranquilli
04f6debb88 Swift: fix bazel packaging 2022-10-21 13:42:24 +02:00
Asger F
23697dba26 Docs: Mention new navigation commands
Document the changes in this PR:
  https://github.com/github/vscode-codeql/pull/1568
2022-10-21 13:33:08 +02:00
Chris Smowton
b80bf4a73e Account for route to IrSimpleType.kotlinType changing as of v1.7.0 2022-10-21 11:55:15 +01:00
Paolo Tranquilli
dbdf6ea489 Swift: fix qltest failure reporting
`qltest.sh` was not exiting with a failure when the extractor was
failing.
2022-10-21 12:54:09 +02:00
Paolo Tranquilli
cf7a5f877b Swift: add qltest.sh tests 2022-10-21 12:54:09 +02:00
Jonas Jensen
7a8c9e7644 Merge pull request #10919 from kaspersv/kaspersv/document-equiv-rel-module
QL language spec: Document built-in equivalence relation module
2022-10-21 12:47:29 +02:00
Ian Lynagh
2e6d6e1538 Merge pull request #10894 from igfoo/igfoo/psi
Kotlin: Refactor PSI handling
2022-10-21 11:43:49 +01:00
Chris Smowton
1fe9e8457f Kotlin: Fix varargs dataflow, and varargs default handling
Dataflow requires accounting for the fact that the varargs parameter isn't necessarily last in the parameter list in a couple more places. Default handling just requires that if the only null parameter is the varargs argument, and it has no default value, then no $default method is required-- the caller is expected to simply pass nothing (at QL
/ source level) or an empty array (at JVM level).
2022-10-21 11:14:41 +01:00
Kasper Svendsen
b29ed3b85a Address reviewer comments from @jbj 2022-10-21 12:08:51 +02:00
Arthur Baars
a56ed88db2 Merge pull request #10920 from github/post-release-prep/codeql-cli-2.11.2
Post-release preparation for codeql-cli-2.11.2
2022-10-21 11:58:12 +02:00
Chris Smowton
7889d9cffa Kotlin: ignore enhanced nullability when extracting primitive types
Otherwise we'll mistake `@NotNull Integer` for `int` and similar, causing a mismatch vs. Java signatures.
2022-10-21 10:55:26 +01:00
Tom Hvitved
4422327c00 Ruby: Call-context sensitivity for singleton method calls 2022-10-21 11:48:25 +02:00
Asger F
3fd2b9ad7b Ruby: add a comment
This would have saved me some time
2022-10-21 11:44:12 +02:00
Asger F
ee7970afcb Ruby: treat String as a builtin 2022-10-21 11:44:11 +02:00
Asger F
db58e3357b Ruby: allow speculative container qname resolution 2022-10-21 11:44:11 +02:00
Kasper Svendsen
925fd2eb45 Accept reviewer reformulation
Co-authored-by: Anders Schack-Mulligen <aschackmull@users.noreply.github.com>
2022-10-21 11:34:48 +02:00
Kasper Svendsen
6fe0de8a9e Accept auto-format suggestion
Co-authored-by: Anders Schack-Mulligen <aschackmull@users.noreply.github.com>
2022-10-21 11:32:37 +02:00
Kasper Svendsen
7faea53c18 QL language spec: Document built-in equivalence relation module 2022-10-21 11:03:01 +02:00
Rasmus Wriedt Larsen
8e8fb3d34f Merge pull request #10911 from RasmusWL/location-debug
Python: add `debug based on location` snippet
2022-10-21 10:59:51 +02:00
Tony Torralba
7a43bdbf05 Apply suggestions from code review
Co-authored-by: Geoffrey White <40627776+geoffw0@users.noreply.github.com>
2022-10-21 10:20:01 +02:00
Tamás Vajk
352c20b0c8 Merge pull request #10885 from tamasvajk/kotlin-const-loop-cond-fp
Kotlin: Add test case for false positive with modified captured variable
2022-10-21 10:08:02 +02:00
github-actions[bot]
be7693283b Post-release preparation for codeql-cli-2.11.2 2022-10-21 08:07:17 +00:00
Tom Hvitved
6feff7e3ed Ruby: Add more data-flow call sensitivity tests 2022-10-21 09:36:34 +02:00
Tom Hvitved
cf35299d08 Merge pull request #10910 from hvitved/ruby/call-graph-refactor
Ruby: Refactor call graph logic for singleton methods
2022-10-21 09:36:13 +02:00
Asger F
d26b0892cf Ruby: also add an AST test 2022-10-21 09:23:21 +02:00
Asger F
038bdecad7 Ruby: add test with compound assignment to a constant 2022-10-21 09:20:03 +02:00
Tamas Vajk
9d1af76c02 Add more test cases 2022-10-21 08:51:11 +02:00
Tamas Vajk
7559d3095f Revert "Kotlin: Exclude captured variables from constant loop condition check"
This reverts commit 3e476f96bd.
2022-10-21 08:38:30 +02:00
Tom Hvitved
db699ae314 Ruby: Refactor call graph logic for singleton methods 2022-10-21 07:27:41 +02:00
thiggy1342
4e5c1f210d Update ruby/ql/lib/change-notes/2022-10-20-expand-faraday-model-for-ssrf-sink
Co-authored-by: Rahul Zhade <rzhade3@users.noreply.github.com>
2022-10-20 17:33:17 -04:00
thiggy1342
ffd596b295 Merge branch 'main' into expand-ruby-ssrf-sinks-faraday-connection-new 2022-10-20 17:12:08 -04:00
Chris Smowton
ac013f9d19 Merge pull request #10889 from smowton/smowton/fix/enum-entry-class-warning
Kotlin: Don't warn on extracting an enum-entry class
2022-10-20 22:08:29 +01:00
thiggy1342
244a3329e0 Merge branch 'main' into expand-ruby-ssrf-sinks-faraday-connection-new 2022-10-20 16:37:57 -04:00
thiggy1342
4c3e3e442a Add Faraday::Connection.new as sink for SSRF query 2022-10-20 20:32:08 +00:00
Jeroen Ketema
4b5674af32 C++: Update test result after extractor changes 2022-10-20 22:18:32 +02:00
Rasmus Wriedt Larsen
ad915e2698 Python: add debug based on location snippet 2022-10-20 21:20:24 +02:00
Geoffrey White
138643519c Merge pull request #10757 from geoffw0/sqlinject
Swift: Query for SQL injection
2022-10-20 18:55:38 +01:00
Geoffrey White
661106c1a0 Apply suggestions from code review
Co-authored-by: Ben Ahmady <32935794+subatoi@users.noreply.github.com>
2022-10-20 17:54:40 +01:00
Chuan-kai Lin
2e9c8c759c Merge pull request #10907 from cklin/document-assume-small-delta
QL language spec: pragma[assume_small_delta]
2022-10-20 09:00:45 -07:00
Chris Smowton
f2749a8878 Don't warn on extracting an enum-entry class 2022-10-20 16:09:45 +01:00
Chuan-kai Lin
9df725901b QL language spec: pragma[assume_small_delta] 2022-10-20 07:30:02 -07:00
Asger F
8c2c28dd56 Ruby: add test showing missing superclass edge 2022-10-20 15:56:58 +02:00
Ian Lynagh
9bc0c98b8e Kotlin: Update logs test 2022-10-20 14:18:31 +01:00
Josh Soref
397b724da1 spelling: triggered
Signed-off-by: Josh Soref <2119212+jsoref@users.noreply.github.com>
2022-10-20 08:21:02 -04:00
Josh Soref
c224231497 spelling: specify
Signed-off-by: Josh Soref <2119212+jsoref@users.noreply.github.com>
2022-10-20 08:21:02 -04:00
Josh Soref
f7d78486d1 spelling: repositories
Signed-off-by: Josh Soref <2119212+jsoref@users.noreply.github.com>
2022-10-20 08:21:02 -04:00
Josh Soref
6767554e19 spelling: recognition
Signed-off-by: Josh Soref <2119212+jsoref@users.noreply.github.com>
2022-10-20 08:21:02 -04:00
Josh Soref
9ed6d97f96 spelling: recommended
Signed-off-by: Josh Soref <2119212+jsoref@users.noreply.github.com>
2022-10-20 08:21:02 -04:00
Josh Soref
f0d27c0257 spelling: properties
Signed-off-by: Josh Soref <2119212+jsoref@users.noreply.github.com>
2022-10-20 08:21:02 -04:00
Josh Soref
485543c2b5 spelling: precondition
Signed-off-by: Josh Soref <2119212+jsoref@users.noreply.github.com>
2022-10-20 08:21:01 -04:00
Josh Soref
ff1b3208ae spelling: official
Signed-off-by: Josh Soref <2119212+jsoref@users.noreply.github.com>
2022-10-20 08:18:23 -04:00
Josh Soref
ff6676e59b spelling: normalize
Signed-off-by: Josh Soref <2119212+jsoref@users.noreply.github.com>
2022-10-20 08:18:23 -04:00
Josh Soref
e4321f07a0 spelling: mimic
Signed-off-by: Josh Soref <2119212+jsoref@users.noreply.github.com>
2022-10-20 08:18:23 -04:00
Josh Soref
1600825679 spelling: implicit
Signed-off-by: Josh Soref <2119212+jsoref@users.noreply.github.com>
2022-10-20 08:18:23 -04:00
Josh Soref
f5ff522a50 spelling: implicitly
Signed-off-by: Josh Soref <2119212+jsoref@users.noreply.github.com>
2022-10-20 08:18:23 -04:00
Josh Soref
c5c9f4d746 spelling: dependencies
Signed-off-by: Josh Soref <2119212+jsoref@users.noreply.github.com>
2022-10-20 08:18:23 -04:00
Josh Soref
474aef438b spelling: connection
Signed-off-by: Josh Soref <2119212+jsoref@users.noreply.github.com>
2022-10-20 08:18:23 -04:00
Josh Soref
4c6454971f spelling: compound
Signed-off-by: Josh Soref <2119212+jsoref@users.noreply.github.com>
2022-10-20 08:18:23 -04:00
Josh Soref
8ff24bc3b9 spelling: additional
Signed-off-by: Josh Soref <2119212+jsoref@users.noreply.github.com>
2022-10-20 08:18:23 -04:00
Geoffrey White
5a3577679d Swift: Improve metadata. 2022-10-20 12:44:56 +01:00
Geoffrey White
adeef309f3 Swift: Add some queries to help examine databases. 2022-10-20 12:34:07 +01:00
Ian Lynagh
291330c7e1 Kotlin: Small code simplification 2022-10-20 12:29:48 +01:00
Paolo Tranquilli
7144383505 Swift: fix british spelling of behaviour 2022-10-20 11:43:46 +02:00
Paolo Tranquilli
f9df8a645f Swift: fix potential name conflict on schema class fields 2022-10-20 11:30:05 +02:00
Paolo Tranquilli
8813aea893 Swift: allow default class doc name to be set for properties 2022-10-20 11:23:13 +02:00
Paolo Tranquilli
c22a7e1c81 Swift: rename _DocnameModifier to _DocModifier 2022-10-20 11:05:55 +02:00
Paolo Tranquilli
37b405f134 Swift: add generated docs for predicates 2022-10-20 11:05:01 +02:00
Paolo Tranquilli
22bd10132f Swift: insert blank line between doc and desc 2022-10-20 10:49:26 +02:00
erik-krogh
bb8bcd4643 fix typo 2022-10-20 10:48:02 +02:00
Paolo Tranquilli
b65f49bd50 Swift: document introducer_int 2022-10-20 10:46:12 +02:00
Paolo Tranquilli
7b181a2de0 Swift: change doc of Immediate property getters 2022-10-20 10:39:37 +02:00
Paolo Tranquilli
6830c2f355 Swift: enhance property docs 2022-10-20 10:35:47 +02:00
erik-krogh
24916f8538 rename runsImmediately to runsArbitraryCode 2022-10-20 10:10:11 +02:00
Paolo Tranquilli
9abaa5c0b3 Swift: rename doc_name with doc in properties 2022-10-20 08:59:08 +02:00
Paolo Tranquilli
492d5aec78 Swift: rename doc to description in properties 2022-10-20 08:57:41 +02:00
Geoffrey White
5b1e138300 Swift: Another qhelp edit. 2022-10-19 20:49:26 +01:00
Geoffrey White
495f744cd3 Swift: Attempt to address qhelp suggestions. 2022-10-19 20:44:27 +01:00
Geoffrey White
05d9c7b892 Swift: More 'an SQL' -> 'a SQL'. 2022-10-19 19:44:59 +01:00
Geoffrey White
83dc6d1564 Apply suggestions from code review
Co-authored-by: Ben Ahmady <32935794+subatoi@users.noreply.github.com>
2022-10-19 19:42:35 +01:00
Ian Lynagh
74a4061508 Kotlin: Refactor PSI handling
We were giving warnings about comments, when we were actually trying to
populate numlines.
2022-10-19 18:02:24 +01:00
Tony Torralba
c2a2d6b379 Fix LaunchOptionsUrlVarDecl
Update test expectations
2022-10-19 17:42:28 +02:00
ALJI Mohamed
9163cbec09 Restrict the reach for an additional taint step 2022-10-19 16:08:49 +01:00
ALJI Mohamed
25a7fcffc0 Add an additional taint step 2022-10-19 16:01:34 +01:00
Chris Smowton
4da480ecc0 Accept test changes resulting from correctly mapping extension methods' default proxies 2022-10-19 15:56:17 +01:00
Tony Torralba
e2c9240973 Add a new Custom URL Scheme source
Also adds a couple of data flow steps to model flow through `?` expressions.
2022-10-19 16:55:14 +02:00
ALJI Mohamed
d6fa745279 Add TarSlip Improv query 2022-10-19 14:01:40 +01:00
Tamas Vajk
3e476f96bd Kotlin: Exclude captured variables from constant loop condition check 2022-10-19 15:01:17 +02:00
Tamas Vajk
0bc57410a0 Kotlin: Add FP test case for constant loop condition 2022-10-19 14:19:49 +02:00
erik-krogh
3dd89bb7bf remove duplicate alerts due to multiple states reaching the same sink 2022-10-19 13:19:18 +02:00
erik-krogh
226bd1f321 add flow-state support to sanitizers in code-execution, and use that to refactor the string-concatenation-sanitizer 2022-10-19 13:06:54 +02:00
erik-krogh
3e51f6fa8e use flow-states to remove FPs related to an attacker only controlling a substring in code-injection 2022-10-19 13:00:44 +02:00
erik-krogh
2a72e89090 add a runsImmediately predicate to CodeExecution (name chosen by Copilot) 2022-10-19 12:30:47 +02:00
Paolo Tranquilli
861377f650 Swift: property doc tweaks 2022-10-19 11:40:05 +02:00
erik-krogh
d77b31672d add failing test for safe-ish uses of Object.send 2022-10-19 11:27:08 +02:00
erik-krogh
cb33d5aeff add test for .send(..) in code-injection 2022-10-19 11:25:30 +02:00
Chris Smowton
b148e3168f Java models-as-data: infer Kotlin $default models from that of its parent function 2022-10-18 18:17:08 +01:00
Geoffrey White
027b71381a Swift: annotate all cases. 2022-10-18 16:38:02 +01:00
Paolo Tranquilli
65fd9cbf9c Swift: docname and desc examples 2022-10-18 17:05:19 +02:00
Paolo Tranquilli
35c1d311c5 Swift: add doc name override 2022-10-18 17:04:51 +02:00
Paolo Tranquilli
8de7df9c21 Swift: add auto-generated docs for getters 2022-10-18 17:04:51 +02:00
Paolo Tranquilli
4d87abed0e Swift: generate docname in qlgen 2022-10-18 17:04:51 +02:00
Paolo Tranquilli
5f7fa6f915 Swift: generate class docs
Python docstrings in `schema.py` are now added to the generated classes.

As an example, a docstring is added to `Expr`.
2022-10-18 17:04:51 +02:00
Paolo Tranquilli
f41fd81965 Swift: add docstring parsing 2022-10-18 16:54:26 +02:00
Geoffrey White
9767064310 Swift: Fix bug for sqlite3_prepare_v3. 2022-10-17 13:40:35 +01:00
Geoffrey White
1221cbaee7 Swift: Updated results after merge with main. 2022-10-17 13:35:46 +01:00
Geoffrey White
13018150ed Merge branch 'main' into sqlinject 2022-10-17 13:30:14 +01:00
Geoffrey White
8eccae1cdd Swift: Fix the qhelp. 2022-10-14 15:31:53 +01:00
Geoffrey White
f96e4eb87e Swift: One more go at getting the query message how ql-for-ql wants it. 2022-10-14 15:28:14 +01:00
Geoffrey White
227b10adf6 Swift: Qhelp. 2022-10-14 15:18:47 +01:00
Geoffrey White
24c6bb4c52 Swift: More modern (?) phrasing. 2022-10-14 14:41:02 +01:00
Geoffrey White
3da3a278ab Swift: Query metadata. 2022-10-14 14:31:38 +01:00
Geoffrey White
76ff593cc5 Swift: Bring it all together into a query. 2022-10-13 16:06:44 +01:00
Geoffrey White
7d78df25bf Swift: Define SQL sinks. 2022-10-13 15:50:57 +01:00
Geoffrey White
ce5631e7cb Swift: Complete the rename. 2022-10-13 15:22:36 +01:00
Geoffrey White
12cb099376 Swift: Rename to match other languages (except Java). 2022-10-13 15:21:39 +01:00
Geoffrey White
398b2a392f Swift: Add more test variants. 2022-10-13 15:13:29 +01:00
Geoffrey White
5496b11153 Swift: Update tests based on feedback. 2022-10-12 14:52:14 +01:00
Geoffrey White
4258147edf Swift: Test SQL injection via the SQLite.swift library. 2022-10-10 17:40:22 +01:00
Geoffrey White
964c92418c Swift: Test SQL injection via the C API. 2022-10-10 17:40:22 +01:00
Geoffrey White
bcab9d8e7c Swift: Add framework for SQL Injection query. 2022-10-10 17:25:08 +01:00
Tamás Vajk
e060ac71bb Change Kotlin stdlib identifier 2022-10-10 13:50:33 +02:00
yo-h
213c5bdab6 Update frameworks.csv 2022-10-07 17:45:39 -04:00
erik-krogh
3e06e201c9 add change-note 2022-10-07 13:45:30 +02:00
erik-krogh
368f84785b fix some more style-guide violations in the alert-messages 2022-10-07 11:22:22 +02:00
Arthur Baars
ae7e6ef701 Ruby: update dependencies 2022-10-04 13:44:22 +02:00
10109 changed files with 322623 additions and 868660 deletions

8
.github/labeler.yml vendored
View File

@@ -43,3 +43,11 @@ documentation:
"QL-for-QL":
- ql/**/*
- .github/workflows/ql-for-ql*
# Since these are all shared files that need to be synced, just pick _one_ copy of each.
"DataFlow Library":
- "java/ql/lib/semmle/code/java/dataflow/internal/DataFlowImpl.qll"
- "java/ql/lib/semmle/code/java/dataflow/internal/DataFlowImplCommon.qll"
- "java/ql/lib/semmle/code/java/dataflow/internal/tainttracking1/TaintTrackingImpl.qll"
- "java/ql/lib/semmle/code/java/dataflow/internal/DataFlowImplConsistency.qll"
- "java/ql/lib/semmle/code/java/dataflow/internal/FlowSummaryImpl.qll"

View File

@@ -0,0 +1,56 @@
name: ATM Check Queries Run
env:
DB_PATH: test_db
ATM_MODEL_PACK: javascript/ql/experimental/adaptivethreatmodeling/src
QUERY_SUITE: codeql-suites/javascript-atm-code-scanning.qls
on:
pull_request:
paths:
- ".github/workflows/atm-check-queries-run.yml"
- "javascript/ql/experimental/adaptivethreatmodeling/**"
workflow_dispatch:
jobs:
run-atm-queries:
runs-on: ubuntu-latest
steps:
- uses: actions/checkout@v3
- name: Install CodeQL CLI
env:
GITHUB_TOKEN: ${{ secrets.GITHUB_TOKEN }}
run: |
gh extensions install github/gh-codeql
gh codeql download
- name: Install ATM model pack
env:
GITHUB_TOKEN: ${{ secrets.GITHUB_TOKEN }}
run: |
set -exu
# Install ATM model pack
gh codeql pack install ${ATM_MODEL_PACK}
# Retrieve model checksum
model_checksum=$(gh codeql resolve extensions ${ATM_MODEL_PACK}/${QUERY_SUITE} | jq -r '.models[0].checksum')
# Trust the model so that we can use it in the ATM boosted queries
mkdir -p "$HOME/.config/codeql"
echo "--insecurely-execute-ml-model-checksums ${model_checksum}" >> "$HOME/.config/codeql/config"
- name: Create test DB
env:
GITHUB_TOKEN: ${{ secrets.GITHUB_TOKEN }}
run: |
gh codeql database create ${RUNNER_TEMP}/${DB_PATH} --source-root config/atm/ --language javascript
- name: Run ATM query suite
env:
GITHUB_TOKEN: ${{ secrets.GITHUB_TOKEN }}
run: |
gh codeql database run-queries -vv -- ${RUNNER_TEMP}/${DB_PATH} ${ATM_MODEL_PACK}/${QUERY_SUITE}

View File

@@ -0,0 +1,12 @@
name: ATM Model Integration Tests
on:
workflow_dispatch:
jobs:
hello-world:
runs-on: ubuntu-latest
steps:
- name: foo
run: echo "Hello world"

View File

@@ -96,8 +96,8 @@ jobs:
- name: Build Query Pack
run: |
codeql pack create ../shared/ssa --output target/packs
codeql pack create ../misc/suite-helpers --output target/packs
codeql pack create ql/lib --output target/packs
codeql pack install ql/src
codeql pack create ql/src --output target/packs
PACK_FOLDER=$(readlink -f target/packs/codeql/ruby-queries/*)
codeql generate query-help --format=sarifv2.1.0 --output="${PACK_FOLDER}/rules.sarif" ql/src
@@ -202,7 +202,7 @@ jobs:
echo 'name: sample-tests
version: 0.0.0
dependencies:
codeql/ruby-all: 0.0.1
codeql/ruby-all: "*"
extractor: ruby
tests: .
' > qlpack.yml

View File

@@ -23,12 +23,23 @@ jobs:
- uses: ./.github/actions/fetch-codeql
- name: Check QL formatting
run: find ql "(" -name "*.ql" -or -name "*.qll" ")" -print0 | xargs -0 codeql query format --check-only
qltest-test:
runs-on: ubuntu-latest
steps:
- uses: actions/checkout@v3
- uses: bazelbuild/setup-bazelisk@v2
- uses: actions/setup-python@v4
with:
python-version-file: 'swift/.python-version'
- name: Test qltest.sh
run: |
bazel test //swift/tools/test/qltest
qltest:
runs-on: ${{ matrix.os }}
strategy:
fail-fast: false
matrix:
os : [ubuntu-20.04, macos-latest]
os: [ ubuntu-20.04, macos-latest ]
steps:
- uses: actions/checkout@v3
- uses: ./.github/actions/fetch-codeql

View File

@@ -52,7 +52,7 @@
| Unneeded defensive code | More true positive and fewer false positive results | This query now recognizes additional defensive code patterns. |
| Unsafe dynamic method access | Fewer false positive results | This query no longer flags concatenated strings as unsafe method names. |
| Unused parameter | Fewer false positive results | This query no longer flags parameters with leading underscore. |
| Unused variable, import, function or class | Fewer false positive results | This query now flags fewer variables that are implictly used by JSX elements. It no longer flags variables with a leading underscore and variables in dead code. |
| Unused variable, import, function or class | Fewer false positive results | This query now flags fewer variables that are implicitly used by JSX elements. It no longer flags variables with a leading underscore and variables in dead code. |
| Unvalidated dynamic method call | More true positive results | This query now flags concatenated strings as unvalidated method names in more cases. |
| Useless assignment to property. | Fewer false positive results | This query now treats assignments with complex right-hand sides correctly. |
| Useless conditional | Fewer results | Additional defensive coding patterns are now ignored. |

View File

@@ -19,7 +19,7 @@ The following changes in version 1.23 affect C/C++ analysis in all applications.
| Hard-coded Japanese era start date in call (`cpp/japanese-era/constructor-or-method-with-exact-era-date`) | Deprecated | This query has been deprecated. Use the new combined query Hard-coded Japanese era start date (`cpp/japanese-era/exact-era-date`) instead. |
| Hard-coded Japanese era start date in struct (`cpp/japanese-era/struct-with-exact-era-date`) | Deprecated | This query has been deprecated. Use the new combined query Hard-coded Japanese era start date (`cpp/japanese-era/exact-era-date`) instead. |
| Hard-coded Japanese era start date (`cpp/japanese-era/exact-era-date`) | More correct results | This query now checks for the beginning date of the Reiwa era (1st May 2019). |
| Non-constant format string (`cpp/non-constant-format`) | Fewer false positive results | Fixed false positive results triggrered by mismatching declarations of a formatting function. |
| Non-constant format string (`cpp/non-constant-format`) | Fewer false positive results | Fixed false positive results triggered by mismatching declarations of a formatting function. |
| Sign check of bitwise operation (`cpp/bitwise-sign-check`) | Fewer false positive results | Results involving `>=` or `<=` are no longer reported. |
| Too few arguments to formatting function (`cpp/wrong-number-format-arguments`) | Fewer false positive results | Fixed false positive results triggered by mismatching declarations of a formatting function. |
| Too many arguments to formatting function (`cpp/too-many-format-arguments`) | Fewer false positive results | Fixed false positive results triggered by mismatching declarations of a formatting function. |

View File

@@ -91,7 +91,7 @@
## Changes to libraries
* The predicates `RegExpTerm.getSuccessor` and `RegExpTerm.getPredecessor` have been changed to reflect textual, not operational, matching order. This only makes a difference in lookbehind assertions, which are operationally matched backwards. Previously, `getSuccessor` would mimick this, so in an assertion `(?<=ab)` the term `b` would be considered the predecessor, not the successor, of `a`. Textually, however, `a` is still matched before `b`, and this is the order we now follow.
* The predicates `RegExpTerm.getSuccessor` and `RegExpTerm.getPredecessor` have been changed to reflect textual, not operational, matching order. This only makes a difference in lookbehind assertions, which are operationally matched backwards. Previously, `getSuccessor` would mimic this, so in an assertion `(?<=ab)` the term `b` would be considered the predecessor, not the successor, of `a`. Textually, however, `a` is still matched before `b`, and this is the order we now follow.
* An extensible model of the `EventEmitter` pattern has been implemented.
* Taint-tracking configurations now interact differently with the `data` flow label, which may affect queries
that combine taint-tracking and flow labels.

View File

@@ -4,6 +4,7 @@ provide:
- "*/ql/test/qlpack.yml"
- "*/ql/examples/qlpack.yml"
- "*/ql/consistency-queries/qlpack.yml"
- "*/upgrades/qlpack.yml"
- "shared/*/qlpack.yml"
- "cpp/ql/test/query-tests/Security/CWE/CWE-190/semmle/tainted/qlpack.yml"
- "go/ql/config/legacy-support/qlpack.yml"
@@ -25,9 +26,3 @@ provide:
- "ruby/extractor-pack/codeql-extractor.yml"
- "swift/extractor-pack/codeql-extractor.yml"
- "ql/extractor-pack/codeql-extractor.ym"
versionPolicies:
default:
requireChangeNotes: true
committedPrereleaseSuffix: dev
committedVersion: nextPatchRelease

View File

@@ -0,0 +1,21 @@
const mongoose = require('mongoose');
Logger = require('./logger').Logger;
Note = require('./models/note').Note;
(async () => {
if (process.argv.length != 5) {
Logger.log("Creates a private note. Usage: node add-note.js <token> <title> <body>")
return;
}
// Open the default mongoose connection
await mongoose.connect('mongodb://localhost:27017/notes', { useFindAndModify: false });
const [userToken, title, body] = process.argv.slice(2);
await Note.create({ title, body, userToken });
Logger.log(`Created private note with title ${title} and body ${body} belonging to user with token ${userToken}.`);
await mongoose.connection.close();
})();

View File

@@ -0,0 +1,68 @@
const bodyParser = require('body-parser');
const express = require('express');
const mongoose = require('mongoose');
const notesApi = require('./notes-api');
const usersApi = require('./users-api');
const addSampleData = module.exports.addSampleData = async () => {
const [userA, userB] = await User.create([
{
name: "A",
token: "tokenA"
},
{
name: "B",
token: "tokenB"
}
]);
await Note.create([
{
title: "Public note belonging to A",
body: "This is a public note belonging to A",
isPublic: true,
ownerToken: userA.token
},
{
title: "Public note belonging to B",
body: "This is a public note belonging to B",
isPublic: true,
ownerToken: userB.token
},
{
title: "Private note belonging to A",
body: "This is a private note belonging to A",
ownerToken: userA.token
},
{
title: "Private note belonging to B",
body: "This is a private note belonging to B",
ownerToken: userB.token
}
]);
}
module.exports.startApp = async () => {
// Open the default mongoose connection
await mongoose.connect('mongodb://mongo:27017/notes', { useFindAndModify: false });
// Drop contents of DB
mongoose.connection.dropDatabase();
// Add some sample data
await addSampleData();
const app = express();
app.use(bodyParser.json());
app.use(bodyParser.urlencoded());
app.get('/', async (_req, res) => {
res.send('Hello World');
});
app.use('/api/notes', notesApi.router);
app.use('/api/users', usersApi.router);
app.listen(3000);
Logger.log('Express started on port 3000');
};

View File

@@ -0,0 +1,7 @@
const startApp = require('./app').startApp;
Logger = require('./logger').Logger;
Note = require('./models/note').Note;
User = require('./models/user').User;
startApp();

View File

@@ -0,0 +1,5 @@
module.exports.Logger = class {
log(message, ...objs) {
console.log(message, objs);
}
};

View File

@@ -0,0 +1,8 @@
const mongoose = require('mongoose');
module.exports.Note = mongoose.model('Note', new mongoose.Schema({
title: String,
body: String,
ownerToken: String,
isPublic: Boolean
}));

View File

@@ -0,0 +1,6 @@
const mongoose = require('mongoose');
module.exports.User = mongoose.model('User', new mongoose.Schema({
name: String,
token: String
}));

View File

@@ -0,0 +1,44 @@
const express = require('express')
const router = module.exports.router = express.Router();
function serializeNote(note) {
return {
title: note.title,
body: note.body
};
}
router.post('/find', async (req, res) => {
const notes = await Note.find({
ownerToken: req.body.token
}).exec();
res.json({
notes: notes.map(serializeNote)
});
});
router.get('/findPublic', async (_req, res) => {
const notes = await Note.find({
isPublic: true
}).exec();
res.json({
notes: notes.map(serializeNote)
});
});
router.post('/findVisible', async (req, res) => {
const notes = await Note.find({
$or: [
{
isPublic: true
},
{
ownerToken: req.body.token
}
]
}).exec();
res.json({
notes: notes.map(serializeNote)
});
});

View File

@@ -0,0 +1,37 @@
const mongoose = require('mongoose');
Logger = require('./logger').Logger;
Note = require('./models/note').Note;
User = require('./models/user').User;
(async () => {
if (process.argv.length != 3) {
Logger.log("Outputs all notes visible to a user. Usage: node read-notes.js <token>")
return;
}
// Open the default mongoose connection
await mongoose.connect('mongodb://localhost:27017/notes', { useFindAndModify: false });
const ownerToken = process.argv[2];
const user = await User.findOne({
token: ownerToken
}).exec();
const notes = await Note.find({
$or: [
{ isPublic: true },
{ ownerToken }
]
}).exec();
notes.map(note => {
Logger.log("Title:" + note.title);
Logger.log("By:" + user.name);
Logger.log("Body:" + note.body);
Logger.log();
});
await mongoose.connection.close();
})();

View File

@@ -0,0 +1,25 @@
const express = require('express')
Logger = require('./logger').Logger;
const router = module.exports.router = express.Router();
router.post('/updateName', async (req, res) => {
Logger.log("/updateName called with new name", req.body.name);
await User.findOneAndUpdate({
token: req.body.token
}, {
name: req.body.name
}).exec();
res.json({
name: req.body.name
});
});
router.post('/getName', async (req, res) => {
const user = await User.findOne({
token: req.body.token
}).exec();
res.json({
name: user.name
});
});

View File

@@ -1,6 +1,4 @@
name: codeql/cpp-examples
groups:
- cpp
- examples
version: 0.0.2
dependencies:
codeql/cpp-all: "*"

View File

@@ -1,15 +0,0 @@
package(default_visibility = ["//cpp:__pkg__"])
load("@rules_pkg//:mappings.bzl", "pkg_files")
pkg_files(
name = "dbscheme",
srcs = ["semmlecode.cpp.dbscheme"],
prefix = "cpp",
)
pkg_files(
name = "dbscheme-stats",
srcs = ["semmlecode.cpp.dbscheme.stats"],
prefix = "cpp",
)

View File

@@ -1,180 +0,0 @@
## 0.4.2
No user-facing changes.
## 0.4.1
No user-facing changes.
## 0.4.0
### Deprecated APIs
* Some classes/modules with upper-case acronyms in their name have been renamed to follow our style-guide.
The old name still exists as a deprecated alias.
### New Features
* Added subclasses of `BuiltInOperations` for `__is_same`, `__is_function`, `__is_layout_compatible`, `__is_pointer_interconvertible_base_of`, `__is_array`, `__array_rank`, `__array_extent`, `__is_arithmetic`, `__is_complete_type`, `__is_compound`, `__is_const`, `__is_floating_point`, `__is_fundamental`, `__is_integral`, `__is_lvalue_reference`, `__is_member_function_pointer`, `__is_member_object_pointer`, `__is_member_pointer`, `__is_object`, `__is_pointer`, `__is_reference`, `__is_rvalue_reference`, `__is_scalar`, `__is_signed`, `__is_unsigned`, `__is_void`, and `__is_volatile`.
### Bug Fixes
* Fixed an issue in the taint tracking analysis where implicit reads were not allowed by default in sinks or additional taint steps that used flow states.
## 0.3.5
## 0.3.4
### Deprecated APIs
* Many classes/predicates/modules with upper-case acronyms in their name have been renamed to follow our style-guide.
The old name still exists as a deprecated alias.
### New Features
* Added support for getting the link targets of global and namespace variables.
* Added a `BlockAssignExpr` class, which models a `memcpy`-like operation used in compiler generated copy/move constructors and assignment operations.
### Minor Analysis Improvements
* All deprecated predicates/classes/modules that have been deprecated for over a year have been deleted.
## 0.3.3
### New Features
* Added a predicate `getValueConstant` to `AttributeArgument` that yields the argument value as an `Expr` when the value is a constant expression.
* A new class predicate `MustFlowConfiguration::allowInterproceduralFlow` has been added to the `semmle.code.cpp.ir.dataflow.MustFlow` library. The new predicate can be overridden to disable interprocedural flow.
* Added subclasses of `BuiltInOperations` for `__builtin_bit_cast`, `__builtin_shuffle`, `__has_unique_object_representations`, `__is_aggregate`, and `__is_assignable`.
### Major Analysis Improvements
* The IR dataflow library now includes flow through global variables. This enables new findings in many scenarios.
## 0.3.2
### Bug Fixes
* Under certain circumstances a variable declaration that is not also a definition could be associated with a `Variable` that did not have the definition as a `VariableDeclarationEntry`. This is now fixed, and a unique `Variable` will exist that has both the declaration and the definition as a `VariableDeclarationEntry`.
## 0.3.1
### Minor Analysis Improvements
* `AnalysedExpr::isNullCheck` and `AnalysedExpr::isValidCheck` have been updated to handle variable accesses on the left-hand side of the C++ logical "and", and variable declarations in conditions.
## 0.3.0
### Deprecated APIs
* The `BarrierGuard` class has been deprecated. Such barriers and sanitizers can now instead be created using the new `BarrierGuard` parameterized module.
### Bug Fixes
* `UserType.getADeclarationEntry()` now yields all forward declarations when the user type is a `class`, `struct`, or `union`.
## 0.2.3
### New Features
* An `isBraced` predicate was added to the `Initializer` class which holds when a C++ braced initializer was used in the initialization.
## 0.2.2
### Deprecated APIs
* The `AnalysedString` class in the `StringAnalysis` module has been replaced with `AnalyzedString`, to follow our style guide. The old name still exists as a deprecated alias.
### New Features
* A `getInitialization` predicate was added to the `ConstexprIfStmt`, `IfStmt`, and `SwitchStmt` classes that yields the C++17-style initializer of the `if` or `switch` statement when it exists.
## 0.2.1
## 0.2.0
### Breaking Changes
* The signature of `allowImplicitRead` on `DataFlow::Configuration` and `TaintTracking::Configuration` has changed from `allowImplicitRead(DataFlow::Node node, DataFlow::Content c)` to `allowImplicitRead(DataFlow::Node node, DataFlow::ContentSet c)`.
### Minor Analysis Improvements
* More Windows pool allocation functions are now detected as `AllocationFunction`s.
* The `semmle.code.cpp.commons.Buffer` library has been enhanced to handle array members of classes that do not specify a size.
## 0.1.0
### Breaking Changes
* The recently added flow-state versions of `isBarrierIn`, `isBarrierOut`, `isSanitizerIn`, and `isSanitizerOut` in the data flow and taint tracking libraries have been removed.
### New Features
* A new library `semmle.code.cpp.security.PrivateData` has been added. The new library heuristically detects variables and functions dealing with sensitive private data, such as e-mail addresses and credit card numbers.
### Minor Analysis Improvements
* The `semmle.code.cpp.security.SensitiveExprs` library has been enhanced with some additional rules for detecting credentials.
## 0.0.13
## 0.0.12
### Breaking Changes
* The flow state variants of `isBarrier` and `isAdditionalFlowStep` are no longer exposed in the taint tracking library. The `isSanitizer` and `isAdditionalTaintStep` predicates should be used instead.
### Deprecated APIs
* Many classes/predicates/modules that had upper-case acronyms have been renamed to follow our style-guide.
The old name still exists as a deprecated alias.
### New Features
* The data flow and taint tracking libraries have been extended with versions of `isBarrierIn`, `isBarrierOut`, and `isBarrierGuard`, respectively `isSanitizerIn`, `isSanitizerOut`, and `isSanitizerGuard`, that support flow states.
### Minor Analysis Improvements
* `DefaultOptions::exits` now holds for C11 functions with the `_Noreturn` or `noreturn` specifier.
* `hasImplicitCopyConstructor` and `hasImplicitCopyAssignmentOperator` now correctly handle implicitly-deleted operators in templates.
* All deprecated predicates/classes/modules that have been deprecated for over a year have been deleted.
## 0.0.11
### Minor Analysis Improvements
* Many queries now support structured bindings, as structured bindings are now handled in the IR translation.
## 0.0.10
### New Features
* Added a `isStructuredBinding` predicate to the `Variable` class which holds when the variable is declared as part of a structured binding declaration.
## 0.0.9
## 0.0.8
### Deprecated APIs
* The `codeql/cpp-upgrades` CodeQL pack has been removed. All upgrades scripts have been merged into the `codeql/cpp-all` CodeQL pack.
### Minor Analysis Improvements
* `FormatLiteral::getMaxConvertedLength` now uses range analysis to provide a
more accurate length for integers formatted with `%x`
## 0.0.7
## 0.0.6
## 0.0.5
## 0.0.4
### New Features
* The QL library `semmle.code.cpp.commons.Exclusions` now contains a predicate
`isFromSystemMacroDefinition` for identifying code that originates from a
macro outside the project being analyzed.

View File

@@ -54,13 +54,11 @@ class Options extends string {
*
* By default, this holds for `exit`, `_exit`, `abort`, `__assert_fail`,
* `longjmp`, `__builtin_unreachable` and any function with a
* `noreturn` attribute or specifier.
* `noreturn` attribute.
*/
predicate exits(Function f) {
f.getAnAttribute().hasName("noreturn")
or
f.getASpecifier().hasName("noreturn")
or
f.hasGlobalOrStdName([
"exit", "_exit", "abort", "__assert_fail", "longjmp", "__builtin_unreachable"
])
@@ -75,7 +73,7 @@ class Options extends string {
* __assume(0);
* ```
* (note that in this case if the hint is wrong and the expression is reached at
* runtime, the program's behavior is undefined)
* runtime, the program's behaviour is undefined)
*/
predicate exprExits(Expr e) {
e.(AssumeExpr).getChild(0).(CompileTimeConstantInt).getIntValue() = 0 or

View File

@@ -39,7 +39,7 @@ class CustomOptions extends Options {
*
* By default, this holds for `exit`, `_exit`, `abort`, `__assert_fail`,
* `longjmp`, `error`, `__builtin_unreachable` and any function with a
* `noreturn` attribute or specifier.
* `noreturn` attribute.
*/
override predicate exits(Function f) { Options.super.exits(f) }
@@ -50,7 +50,7 @@ class CustomOptions extends Options {
* __assume(0);
* ```
* (note that in this case if the hint is wrong and the expression is reached at
* runtime, the program's behavior is undefined)
* runtime, the program's behaviour is undefined)
*/
override predicate exprExits(Expr e) { Options.super.exprExits(e) }

View File

@@ -1,5 +0,0 @@
## 0.0.10
### New Features
* Added a `isStructuredBinding` predicate to the `Variable` class which holds when the variable is declared as part of a structured binding declaration.

View File

@@ -1,5 +0,0 @@
## 0.0.11
### Minor Analysis Improvements
* Many queries now support structured bindings, as structured bindings are now handled in the IR translation.

View File

@@ -1,20 +0,0 @@
## 0.0.12
### Breaking Changes
* The flow state variants of `isBarrier` and `isAdditionalFlowStep` are no longer exposed in the taint tracking library. The `isSanitizer` and `isAdditionalTaintStep` predicates should be used instead.
### Deprecated APIs
* Many classes/predicates/modules that had upper-case acronyms have been renamed to follow our style-guide.
The old name still exists as a deprecated alias.
### New Features
* The data flow and taint tracking libraries have been extended with versions of `isBarrierIn`, `isBarrierOut`, and `isBarrierGuard`, respectively `isSanitizerIn`, `isSanitizerOut`, and `isSanitizerGuard`, that support flow states.
### Minor Analysis Improvements
* `DefaultOptions::exits` now holds for C11 functions with the `_Noreturn` or `noreturn` specifier.
* `hasImplicitCopyConstructor` and `hasImplicitCopyAssignmentOperator` now correctly handle implicitly-deleted operators in templates.
* All deprecated predicates/classes/modules that have been deprecated for over a year have been deleted.

View File

@@ -1 +0,0 @@
## 0.0.13

View File

@@ -1,7 +0,0 @@
## 0.0.4
### New Features
* The QL library `semmle.code.cpp.commons.Exclusions` now contains a predicate
`isFromSystemMacroDefinition` for identifying code that originates from a
macro outside the project being analyzed.

View File

@@ -1 +0,0 @@
## 0.0.5

View File

@@ -1 +0,0 @@
## 0.0.6

View File

@@ -1 +0,0 @@
## 0.0.7

View File

@@ -1,10 +0,0 @@
## 0.0.8
### Deprecated APIs
* The `codeql/cpp-upgrades` CodeQL pack has been removed. All upgrades scripts have been merged into the `codeql/cpp-all` CodeQL pack.
### Minor Analysis Improvements
* `FormatLiteral::getMaxConvertedLength` now uses range analysis to provide a
more accurate length for integers formatted with `%x`

View File

@@ -1,2 +0,0 @@
## 0.0.9

View File

@@ -1,13 +0,0 @@
## 0.1.0
### Breaking Changes
* The recently added flow-state versions of `isBarrierIn`, `isBarrierOut`, `isSanitizerIn`, and `isSanitizerOut` in the data flow and taint tracking libraries have been removed.
### New Features
* A new library `semmle.code.cpp.security.PrivateData` has been added. The new library heuristically detects variables and functions dealing with sensitive private data, such as e-mail addresses and credit card numbers.
### Minor Analysis Improvements
* The `semmle.code.cpp.security.SensitiveExprs` library has been enhanced with some additional rules for detecting credentials.

View File

@@ -1,10 +0,0 @@
## 0.2.0
### Breaking Changes
* The signature of `allowImplicitRead` on `DataFlow::Configuration` and `TaintTracking::Configuration` has changed from `allowImplicitRead(DataFlow::Node node, DataFlow::Content c)` to `allowImplicitRead(DataFlow::Node node, DataFlow::ContentSet c)`.
### Minor Analysis Improvements
* More Windows pool allocation functions are now detected as `AllocationFunction`s.
* The `semmle.code.cpp.commons.Buffer` library has been enhanced to handle array members of classes that do not specify a size.

View File

@@ -1 +0,0 @@
## 0.2.1

View File

@@ -1,9 +0,0 @@
## 0.2.2
### Deprecated APIs
* The `AnalysedString` class in the `StringAnalysis` module has been replaced with `AnalyzedString`, to follow our style guide. The old name still exists as a deprecated alias.
### New Features
* A `getInitialization` predicate was added to the `ConstexprIfStmt`, `IfStmt`, and `SwitchStmt` classes that yields the C++17-style initializer of the `if` or `switch` statement when it exists.

View File

@@ -1,5 +0,0 @@
## 0.2.3
### New Features
* An `isBraced` predicate was added to the `Initializer` class which holds when a C++ braced initializer was used in the initialization.

View File

@@ -1,9 +0,0 @@
## 0.3.0
### Deprecated APIs
* The `BarrierGuard` class has been deprecated. Such barriers and sanitizers can now instead be created using the new `BarrierGuard` parameterized module.
### Bug Fixes
* `UserType.getADeclarationEntry()` now yields all forward declarations when the user type is a `class`, `struct`, or `union`.

View File

@@ -1,5 +0,0 @@
## 0.3.1
### Minor Analysis Improvements
* `AnalysedExpr::isNullCheck` and `AnalysedExpr::isValidCheck` have been updated to handle variable accesses on the left-hand side of the C++ logical "and", and variable declarations in conditions.

View File

@@ -1,5 +0,0 @@
## 0.3.2
### Bug Fixes
* Under certain circumstances a variable declaration that is not also a definition could be associated with a `Variable` that did not have the definition as a `VariableDeclarationEntry`. This is now fixed, and a unique `Variable` will exist that has both the declaration and the definition as a `VariableDeclarationEntry`.

View File

@@ -1,11 +0,0 @@
## 0.3.3
### New Features
* Added a predicate `getValueConstant` to `AttributeArgument` that yields the argument value as an `Expr` when the value is a constant expression.
* A new class predicate `MustFlowConfiguration::allowInterproceduralFlow` has been added to the `semmle.code.cpp.ir.dataflow.MustFlow` library. The new predicate can be overridden to disable interprocedural flow.
* Added subclasses of `BuiltInOperations` for `__builtin_bit_cast`, `__builtin_shuffle`, `__has_unique_object_representations`, `__is_aggregate`, and `__is_assignable`.
### Major Analysis Improvements
* The IR dataflow library now includes flow through global variables. This enables new findings in many scenarios.

View File

@@ -1,15 +0,0 @@
## 0.3.4
### Deprecated APIs
* Many classes/predicates/modules with upper-case acronyms in their name have been renamed to follow our style-guide.
The old name still exists as a deprecated alias.
### New Features
* Added support for getting the link targets of global and namespace variables.
* Added a `BlockAssignExpr` class, which models a `memcpy`-like operation used in compiler generated copy/move constructors and assignment operations.
### Minor Analysis Improvements
* All deprecated predicates/classes/modules that have been deprecated for over a year have been deleted.

View File

@@ -1 +0,0 @@
## 0.3.5

View File

@@ -1,14 +0,0 @@
## 0.4.0
### Deprecated APIs
* Some classes/modules with upper-case acronyms in their name have been renamed to follow our style-guide.
The old name still exists as a deprecated alias.
### New Features
* Added subclasses of `BuiltInOperations` for `__is_same`, `__is_function`, `__is_layout_compatible`, `__is_pointer_interconvertible_base_of`, `__is_array`, `__array_rank`, `__array_extent`, `__is_arithmetic`, `__is_complete_type`, `__is_compound`, `__is_const`, `__is_floating_point`, `__is_fundamental`, `__is_integral`, `__is_lvalue_reference`, `__is_member_function_pointer`, `__is_member_object_pointer`, `__is_member_pointer`, `__is_object`, `__is_pointer`, `__is_reference`, `__is_rvalue_reference`, `__is_scalar`, `__is_signed`, `__is_unsigned`, `__is_void`, and `__is_volatile`.
### Bug Fixes
* Fixed an issue in the taint tracking analysis where implicit reads were not allowed by default in sinks or additional taint steps that used flow states.

View File

@@ -1,3 +0,0 @@
## 0.4.1
No user-facing changes.

View File

@@ -1,3 +0,0 @@
## 0.4.2
No user-facing changes.

View File

@@ -1,2 +0,0 @@
---
lastReleaseVersion: 0.4.2

View File

@@ -69,4 +69,6 @@ import semmle.code.cpp.Comments
import semmle.code.cpp.Preprocessor
import semmle.code.cpp.Iteration
import semmle.code.cpp.NameQualifiers
import semmle.code.cpp.ObjectiveC
import semmle.code.cpp.exprs.ObjectiveC
import DefaultOptions

View File

@@ -1,301 +0,0 @@
import experimental.semmle.code.cpp.ir.dataflow.DataFlow
import experimental.semmle.code.cpp.ir.dataflow.DataFlow2
module ProductFlow {
abstract class Configuration extends string {
bindingset[this]
Configuration() { any() }
/**
* Holds if `(source1, source2)` is a relevant data flow source.
*
* `source1` and `source2` must belong to the same callable.
*/
predicate isSourcePair(DataFlow::Node source1, DataFlow::Node source2) { none() }
/**
* Holds if `(source1, source2)` is a relevant data flow source with initial states `state1`
* and `state2`, respectively.
*
* `source1` and `source2` must belong to the same callable.
*/
predicate isSourcePair(
DataFlow::Node source1, DataFlow::FlowState state1, DataFlow::Node source2,
DataFlow::FlowState state2
) {
state1 = "" and
state2 = "" and
this.isSourcePair(source1, source2)
}
/**
* Holds if `(sink1, sink2)` is a relevant data flow sink.
*
* `sink1` and `sink2` must belong to the same callable.
*/
predicate isSinkPair(DataFlow::Node sink1, DataFlow::Node sink2) { none() }
/**
* Holds if `(sink1, sink2)` is a relevant data flow sink with final states `state1`
* and `state2`, respectively.
*
* `sink1` and `sink2` must belong to the same callable.
*/
predicate isSinkPair(
DataFlow::Node sink1, DataFlow::FlowState state1, DataFlow::Node sink2,
DataFlow::FlowState state2
) {
state1 = "" and
state2 = "" and
this.isSinkPair(sink1, sink2)
}
/**
* Holds if data flow through `node` is prohibited through the first projection of the product
* dataflow graph when the flow state is `state`.
*/
predicate isBarrier1(DataFlow::Node node, DataFlow::FlowState state) {
this.isBarrier1(node) and state = ""
}
/**
* Holds if data flow through `node` is prohibited through the second projection of the product
* dataflow graph when the flow state is `state`.
*/
predicate isBarrier2(DataFlow::Node node, DataFlow::FlowState state) {
this.isBarrier2(node) and state = ""
}
/**
* Holds if data flow through `node` is prohibited through the first projection of the product
* dataflow graph.
*/
predicate isBarrier1(DataFlow::Node node) { none() }
/**
* Holds if data flow through `node` is prohibited through the second projection of the product
* dataflow graph.
*/
predicate isBarrier2(DataFlow::Node node) { none() }
/**
* Holds if data flow out of `node` is prohibited in the first projection of the product
* dataflow graph.
*/
predicate isBarrierOut1(DataFlow::Node node) { none() }
/**
* Holds if data flow out of `node` is prohibited in the second projection of the product
* dataflow graph.
*/
predicate isBarrierOut2(DataFlow::Node node) { none() }
/*
* Holds if data may flow from `node1` to `node2` in addition to the normal data-flow steps in
* the first projection of the product dataflow graph.
*/
predicate isAdditionalFlowStep1(DataFlow::Node node1, DataFlow::Node node2) { none() }
/**
* Holds if data may flow from `node1` to `node2` in addition to the normal data-flow steps in
* the first projection of the product dataflow graph.
*
* This step is only applicable in `state1` and updates the flow state to `state2`.
*/
predicate isAdditionalFlowStep1(
DataFlow::Node node1, DataFlow::FlowState state1, DataFlow::Node node2,
DataFlow::FlowState state2
) {
state1 instanceof DataFlow::FlowStateEmpty and
state2 instanceof DataFlow::FlowStateEmpty and
this.isAdditionalFlowStep1(node1, node2)
}
/**
* Holds if data may flow from `node1` to `node2` in addition to the normal data-flow steps in
* the second projection of the product dataflow graph.
*/
predicate isAdditionalFlowStep2(DataFlow::Node node1, DataFlow::Node node2) { none() }
/**
* Holds if data may flow from `node1` to `node2` in addition to the normal data-flow steps in
* the second projection of the product dataflow graph.
*
* This step is only applicable in `state1` and updates the flow state to `state2`.
*/
predicate isAdditionalFlowStep2(
DataFlow::Node node1, DataFlow::FlowState state1, DataFlow::Node node2,
DataFlow::FlowState state2
) {
state1 instanceof DataFlow::FlowStateEmpty and
state2 instanceof DataFlow::FlowStateEmpty and
this.isAdditionalFlowStep2(node1, node2)
}
/**
* Holds if data flow into `node` is prohibited in the first projection of the product
* dataflow graph.
*/
predicate isBarrierIn1(DataFlow::Node node) { none() }
/**
* Holds if data flow into `node` is prohibited in the second projection of the product
* dataflow graph.
*/
predicate isBarrierIn2(DataFlow::Node node) { none() }
predicate hasFlowPath(
DataFlow::PathNode source1, DataFlow2::PathNode source2, DataFlow::PathNode sink1,
DataFlow2::PathNode sink2
) {
reachable(this, source1, source2, sink1, sink2)
}
}
private import Internal
module Internal {
class Conf1 extends DataFlow::Configuration {
Conf1() { this = "Conf1" }
override predicate isSource(DataFlow::Node source, DataFlow::FlowState state) {
exists(Configuration conf | conf.isSourcePair(source, state, _, _))
}
override predicate isSink(DataFlow::Node sink, DataFlow::FlowState state) {
exists(Configuration conf | conf.isSinkPair(sink, state, _, _))
}
override predicate isBarrier(DataFlow::Node node, DataFlow::FlowState state) {
exists(Configuration conf | conf.isBarrier1(node, state))
}
override predicate isBarrierOut(DataFlow::Node node) {
exists(Configuration conf | conf.isBarrierOut1(node))
}
override predicate isAdditionalFlowStep(
DataFlow::Node node1, DataFlow::FlowState state1, DataFlow::Node node2,
DataFlow::FlowState state2
) {
exists(Configuration conf | conf.isAdditionalFlowStep1(node1, state1, node2, state2))
}
override predicate isBarrierIn(DataFlow::Node node) {
exists(Configuration conf | conf.isBarrierIn1(node))
}
}
class Conf2 extends DataFlow2::Configuration {
Conf2() { this = "Conf2" }
override predicate isSource(DataFlow::Node source, DataFlow::FlowState state) {
exists(Configuration conf, DataFlow::PathNode source1 |
conf.isSourcePair(source1.getNode(), source1.getState(), source, state) and
any(Conf1 c).hasFlowPath(source1, _)
)
}
override predicate isSink(DataFlow::Node sink, DataFlow::FlowState state) {
exists(Configuration conf, DataFlow::PathNode sink1 |
conf.isSinkPair(sink1.getNode(), sink1.getState(), sink, state) and
any(Conf1 c).hasFlowPath(_, sink1)
)
}
override predicate isBarrier(DataFlow::Node node, DataFlow::FlowState state) {
exists(Configuration conf | conf.isBarrier2(node, state))
}
override predicate isBarrierOut(DataFlow::Node node) {
exists(Configuration conf | conf.isBarrierOut2(node))
}
override predicate isAdditionalFlowStep(
DataFlow::Node node1, DataFlow::FlowState state1, DataFlow::Node node2,
DataFlow::FlowState state2
) {
exists(Configuration conf | conf.isAdditionalFlowStep2(node1, state1, node2, state2))
}
override predicate isBarrierIn(DataFlow::Node node) {
exists(Configuration conf | conf.isBarrierIn2(node))
}
}
}
pragma[nomagic]
private predicate reachableInterprocEntry(
Configuration conf, DataFlow::PathNode source1, DataFlow2::PathNode source2,
DataFlow::PathNode node1, DataFlow2::PathNode node2
) {
conf.isSourcePair(node1.getNode(), node1.getState(), node2.getNode(), node2.getState()) and
node1 = source1 and
node2 = source2
or
exists(
DataFlow::PathNode midEntry1, DataFlow2::PathNode midEntry2, DataFlow::PathNode midExit1,
DataFlow2::PathNode midExit2
|
reachableInterprocEntry(conf, source1, source2, midEntry1, midEntry2) and
interprocEdgePair(midExit1, midExit2, node1, node2) and
localPathStep1*(midEntry1, midExit1) and
localPathStep2*(midEntry2, midExit2)
)
}
private predicate localPathStep1(DataFlow::PathNode pred, DataFlow::PathNode succ) {
DataFlow::PathGraph::edges(pred, succ) and
pragma[only_bind_out](pred.getNode().getEnclosingCallable()) =
pragma[only_bind_out](succ.getNode().getEnclosingCallable())
}
private predicate localPathStep2(DataFlow2::PathNode pred, DataFlow2::PathNode succ) {
DataFlow2::PathGraph::edges(pred, succ) and
pragma[only_bind_out](pred.getNode().getEnclosingCallable()) =
pragma[only_bind_out](succ.getNode().getEnclosingCallable())
}
pragma[nomagic]
private predicate interprocEdge1(
Declaration predDecl, Declaration succDecl, DataFlow::PathNode pred1, DataFlow::PathNode succ1
) {
DataFlow::PathGraph::edges(pred1, succ1) and
predDecl != succDecl and
pred1.getNode().getEnclosingCallable() = predDecl and
succ1.getNode().getEnclosingCallable() = succDecl
}
pragma[nomagic]
private predicate interprocEdge2(
Declaration predDecl, Declaration succDecl, DataFlow2::PathNode pred2, DataFlow2::PathNode succ2
) {
DataFlow2::PathGraph::edges(pred2, succ2) and
predDecl != succDecl and
pred2.getNode().getEnclosingCallable() = predDecl and
succ2.getNode().getEnclosingCallable() = succDecl
}
private predicate interprocEdgePair(
DataFlow::PathNode pred1, DataFlow2::PathNode pred2, DataFlow::PathNode succ1,
DataFlow2::PathNode succ2
) {
exists(Declaration predDecl, Declaration succDecl |
interprocEdge1(predDecl, succDecl, pred1, succ1) and
interprocEdge2(predDecl, succDecl, pred2, succ2)
)
}
private predicate reachable(
Configuration conf, DataFlow::PathNode source1, DataFlow2::PathNode source2,
DataFlow::PathNode sink1, DataFlow2::PathNode sink2
) {
exists(DataFlow::PathNode mid1, DataFlow2::PathNode mid2 |
reachableInterprocEntry(conf, source1, source2, mid1, mid2) and
conf.isSinkPair(sink1.getNode(), sink1.getState(), sink2.getNode(), sink2.getState()) and
localPathStep1*(mid1, sink1) and
localPathStep2*(mid2, sink2)
)
}
}

View File

@@ -1,26 +0,0 @@
/**
* Provides a library for local (intra-procedural) and global (inter-procedural)
* data flow analysis: deciding whether data can flow from a _source_ to a
* _sink_. This library differs from the one in `semmle.code.cpp.dataflow` in that
* this library uses the IR (Intermediate Representation) library, which provides
* a more precise semantic representation of the program, whereas the other dataflow
* library uses the more syntax-oriented ASTs. This library should provide more accurate
* results than the AST-based library in most scenarios.
*
* Unless configured otherwise, _flow_ means that the exact value of
* the source may reach the sink. We do not track flow across pointer
* dereferences or array indexing.
*
* To use global (interprocedural) data flow, extend the class
* `DataFlow::Configuration` as documented on that class. To use local
* (intraprocedural) data flow between expressions, call
* `DataFlow::localExprFlow`. For more general cases of local data flow, call
* `DataFlow::localFlow` or `DataFlow::localFlowStep` with arguments of type
* `DataFlow::Node`.
*/
import cpp
module DataFlow {
import experimental.semmle.code.cpp.ir.dataflow.internal.DataFlowImpl
}

View File

@@ -1,16 +0,0 @@
/**
* Provides a `DataFlow2` module, which is a copy of the `DataFlow` module. Use
* this class when data-flow configurations must depend on each other. Two
* classes extending `DataFlow::Configuration` should never depend on each
* other, but one of them should instead depend on a
* `DataFlow2::Configuration`, a `DataFlow3::Configuration`, or a
* `DataFlow4::Configuration`.
*
* See `semmle.code.cpp.ir.dataflow.DataFlow` for the full documentation.
*/
import cpp
module DataFlow2 {
import experimental.semmle.code.cpp.ir.dataflow.internal.DataFlowImpl2
}

View File

@@ -1,16 +0,0 @@
/**
* Provides a `DataFlow3` module, which is a copy of the `DataFlow` module. Use
* this class when data-flow configurations must depend on each other. Two
* classes extending `DataFlow::Configuration` should never depend on each
* other, but one of them should instead depend on a
* `DataFlow2::Configuration`, a `DataFlow3::Configuration`, or a
* `DataFlow4::Configuration`.
*
* See `semmle.code.cpp.ir.dataflow.DataFlow` for the full documentation.
*/
import cpp
module DataFlow3 {
import experimental.semmle.code.cpp.ir.dataflow.internal.DataFlowImpl3
}

View File

@@ -1,16 +0,0 @@
/**
* Provides a `DataFlow4` module, which is a copy of the `DataFlow` module. Use
* this class when data-flow configurations must depend on each other. Two
* classes extending `DataFlow::Configuration` should never depend on each
* other, but one of them should instead depend on a
* `DataFlow2::Configuration`, a `DataFlow3::Configuration`, or a
* `DataFlow4::Configuration`.
*
* See `semmle.code.cpp.ir.dataflow.DataFlow` for the full documentation.
*/
import cpp
module DataFlow4 {
import experimental.semmle.code.cpp.ir.dataflow.internal.DataFlowImpl4
}

View File

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

View File

@@ -1,23 +0,0 @@
/**
* Provides classes for performing local (intra-procedural) and
* global (inter-procedural) taint-tracking analyses.
*
* We define _taint propagation_ informally to mean that a substantial part of
* the information from the source is preserved at the sink. For example, taint
* propagates from `x` to `x + 100`, but it does not propagate from `x` to `x >
* 100` since we consider a single bit of information to be too little.
*
* To use global (interprocedural) taint tracking, extend the class
* `TaintTracking::Configuration` as documented on that class. To use local
* (intraprocedural) taint tracking between expressions, call
* `TaintTracking::localExprTaint`. For more general cases of local taint
* tracking, call `TaintTracking::localTaint` or
* `TaintTracking::localTaintStep` with arguments of type `DataFlow::Node`.
*/
import semmle.code.cpp.ir.dataflow.DataFlow
import semmle.code.cpp.ir.dataflow.DataFlow2
module TaintTracking {
import experimental.semmle.code.cpp.ir.dataflow.internal.tainttracking1.TaintTrackingImpl
}

View File

@@ -1,15 +0,0 @@
/**
* Provides a `TaintTracking2` module, which is a copy of the `TaintTracking`
* module. Use this class when data-flow configurations or taint-tracking
* configurations must depend on each other. Two classes extending
* `DataFlow::Configuration` should never depend on each other, but one of them
* should instead depend on a `DataFlow2::Configuration`, a
* `DataFlow3::Configuration`, or a `DataFlow4::Configuration`. The
* `TaintTracking::Configuration` class extends `DataFlow::Configuration`, and
* `TaintTracking2::Configuration` extends `DataFlow2::Configuration`.
*
* See `semmle.code.cpp.ir.dataflow.TaintTracking` for the full documentation.
*/
module TaintTracking2 {
import experimental.semmle.code.cpp.ir.dataflow.internal.tainttracking2.TaintTrackingImpl
}

View File

@@ -1,15 +0,0 @@
/**
* Provides a `TaintTracking3` module, which is a copy of the `TaintTracking`
* module. Use this class when data-flow configurations or taint-tracking
* configurations must depend on each other. Two classes extending
* `DataFlow::Configuration` should never depend on each other, but one of them
* should instead depend on a `DataFlow2::Configuration`, a
* `DataFlow3::Configuration`, or a `DataFlow4::Configuration`. The
* `TaintTracking::Configuration` class extends `DataFlow::Configuration`, and
* `TaintTracking2::Configuration` extends `DataFlow2::Configuration`.
*
* See `semmle.code.cpp.ir.dataflow.TaintTracking` for the full documentation.
*/
module TaintTracking3 {
import experimental.semmle.code.cpp.ir.dataflow.internal.tainttracking3.TaintTrackingImpl
}

View File

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

View File

@@ -1,235 +0,0 @@
/**
* Provides consistency queries for checking invariants in the language-specific
* data-flow classes and predicates.
*/
private import DataFlowImplSpecific::Private
private import DataFlowImplSpecific::Public
private import tainttracking1.TaintTrackingParameter::Private
private import tainttracking1.TaintTrackingParameter::Public
module Consistency {
private newtype TConsistencyConfiguration = MkConsistencyConfiguration()
/** A class for configuring the consistency queries. */
class ConsistencyConfiguration extends TConsistencyConfiguration {
string toString() { none() }
/** Holds if `n` should be excluded from the consistency test `uniqueEnclosingCallable`. */
predicate uniqueEnclosingCallableExclude(Node n) { none() }
/** Holds if `n` should be excluded from the consistency test `uniqueNodeLocation`. */
predicate uniqueNodeLocationExclude(Node n) { none() }
/** Holds if `n` should be excluded from the consistency test `missingLocation`. */
predicate missingLocationExclude(Node n) { none() }
/** Holds if `n` should be excluded from the consistency test `postWithInFlow`. */
predicate postWithInFlowExclude(Node n) { none() }
/** Holds if `n` should be excluded from the consistency test `argHasPostUpdate`. */
predicate argHasPostUpdateExclude(ArgumentNode n) { none() }
/** Holds if `n` should be excluded from the consistency test `reverseRead`. */
predicate reverseReadExclude(Node n) { none() }
/** Holds if `n` should be excluded from the consistency test `postHasUniquePre`. */
predicate postHasUniquePreExclude(PostUpdateNode n) { none() }
/** Holds if `n` should be excluded from the consistency test `uniquePostUpdate`. */
predicate uniquePostUpdateExclude(Node n) { none() }
/** Holds if `(call, ctx)` should be excluded from the consistency test `viableImplInCallContextTooLargeExclude`. */
predicate viableImplInCallContextTooLargeExclude(
DataFlowCall call, DataFlowCall ctx, DataFlowCallable callable
) {
none()
}
}
private class RelevantNode extends Node {
RelevantNode() {
this instanceof ArgumentNode or
this instanceof ParameterNode or
this instanceof ReturnNode or
this = getAnOutNode(_, _) or
simpleLocalFlowStep(this, _) or
simpleLocalFlowStep(_, this) or
jumpStep(this, _) or
jumpStep(_, this) or
storeStep(this, _, _) or
storeStep(_, _, this) or
readStep(this, _, _) or
readStep(_, _, this) or
defaultAdditionalTaintStep(this, _) or
defaultAdditionalTaintStep(_, this)
}
}
query predicate uniqueEnclosingCallable(Node n, string msg) {
exists(int c |
n instanceof RelevantNode and
c = count(nodeGetEnclosingCallable(n)) and
c != 1 and
not any(ConsistencyConfiguration conf).uniqueEnclosingCallableExclude(n) and
msg = "Node should have one enclosing callable but has " + c + "."
)
}
query predicate uniqueType(Node n, string msg) {
exists(int c |
n instanceof RelevantNode and
c = count(getNodeType(n)) and
c != 1 and
msg = "Node should have one type but has " + c + "."
)
}
query predicate uniqueNodeLocation(Node n, string msg) {
exists(int c |
c =
count(string filepath, int startline, int startcolumn, int endline, int endcolumn |
n.hasLocationInfo(filepath, startline, startcolumn, endline, endcolumn)
) and
c != 1 and
not any(ConsistencyConfiguration conf).uniqueNodeLocationExclude(n) and
msg = "Node should have one location but has " + c + "."
)
}
query predicate missingLocation(string msg) {
exists(int c |
c =
strictcount(Node n |
not exists(string filepath, int startline, int startcolumn, int endline, int endcolumn |
n.hasLocationInfo(filepath, startline, startcolumn, endline, endcolumn)
) and
not any(ConsistencyConfiguration conf).missingLocationExclude(n)
) and
msg = "Nodes without location: " + c
)
}
query predicate uniqueNodeToString(Node n, string msg) {
exists(int c |
c = count(n.toString()) and
c != 1 and
msg = "Node should have one toString but has " + c + "."
)
}
query predicate missingToString(string msg) {
exists(int c |
c = strictcount(Node n | not exists(n.toString())) and
msg = "Nodes without toString: " + c
)
}
query predicate parameterCallable(ParameterNode p, string msg) {
exists(DataFlowCallable c | isParameterNode(p, c, _) and c != nodeGetEnclosingCallable(p)) and
msg = "Callable mismatch for parameter."
}
query predicate localFlowIsLocal(Node n1, Node n2, string msg) {
simpleLocalFlowStep(n1, n2) and
nodeGetEnclosingCallable(n1) != nodeGetEnclosingCallable(n2) and
msg = "Local flow step does not preserve enclosing callable."
}
private DataFlowType typeRepr() { result = getNodeType(_) }
query predicate compatibleTypesReflexive(DataFlowType t, string msg) {
t = typeRepr() and
not compatibleTypes(t, t) and
msg = "Type compatibility predicate is not reflexive."
}
query predicate unreachableNodeCCtx(Node n, DataFlowCall call, string msg) {
isUnreachableInCall(n, call) and
exists(DataFlowCallable c |
c = nodeGetEnclosingCallable(n) and
not viableCallable(call) = c
) and
msg = "Call context for isUnreachableInCall is inconsistent with call graph."
}
query predicate localCallNodes(DataFlowCall call, Node n, string msg) {
(
n = getAnOutNode(call, _) and
msg = "OutNode and call does not share enclosing callable."
or
n.(ArgumentNode).argumentOf(call, _) and
msg = "ArgumentNode and call does not share enclosing callable."
) and
nodeGetEnclosingCallable(n) != call.getEnclosingCallable()
}
// This predicate helps the compiler forget that in some languages
// it is impossible for a result of `getPreUpdateNode` to be an
// instance of `PostUpdateNode`.
private Node getPre(PostUpdateNode n) {
result = n.getPreUpdateNode()
or
none()
}
query predicate postIsNotPre(PostUpdateNode n, string msg) {
getPre(n) = n and
msg = "PostUpdateNode should not equal its pre-update node."
}
query predicate postHasUniquePre(PostUpdateNode n, string msg) {
not any(ConsistencyConfiguration conf).postHasUniquePreExclude(n) and
exists(int c |
c = count(n.getPreUpdateNode()) and
c != 1 and
msg = "PostUpdateNode should have one pre-update node but has " + c + "."
)
}
query predicate uniquePostUpdate(Node n, string msg) {
not any(ConsistencyConfiguration conf).uniquePostUpdateExclude(n) and
1 < strictcount(PostUpdateNode post | post.getPreUpdateNode() = n) and
msg = "Node has multiple PostUpdateNodes."
}
query predicate postIsInSameCallable(PostUpdateNode n, string msg) {
nodeGetEnclosingCallable(n) != nodeGetEnclosingCallable(n.getPreUpdateNode()) and
msg = "PostUpdateNode does not share callable with its pre-update node."
}
private predicate hasPost(Node n) { exists(PostUpdateNode post | post.getPreUpdateNode() = n) }
query predicate reverseRead(Node n, string msg) {
exists(Node n2 | readStep(n, _, n2) and hasPost(n2) and not hasPost(n)) and
not any(ConsistencyConfiguration conf).reverseReadExclude(n) and
msg = "Origin of readStep is missing a PostUpdateNode."
}
query predicate argHasPostUpdate(ArgumentNode n, string msg) {
not hasPost(n) and
not any(ConsistencyConfiguration c).argHasPostUpdateExclude(n) and
msg = "ArgumentNode is missing PostUpdateNode."
}
// This predicate helps the compiler forget that in some languages
// it is impossible for a `PostUpdateNode` to be the target of
// `simpleLocalFlowStep`.
private predicate isPostUpdateNode(Node n) { n instanceof PostUpdateNode or none() }
query predicate postWithInFlow(Node n, string msg) {
isPostUpdateNode(n) and
not clearsContent(n, _) and
simpleLocalFlowStep(_, n) and
not any(ConsistencyConfiguration c).postWithInFlowExclude(n) and
msg = "PostUpdateNode should not be the target of local flow."
}
query predicate viableImplInCallContextTooLarge(
DataFlowCall call, DataFlowCall ctx, DataFlowCallable callable
) {
callable = viableImplInCallContext(call, ctx) and
not callable = viableCallable(call) and
not any(ConsistencyConfiguration c).viableImplInCallContextTooLargeExclude(call, ctx, callable)
}
}

View File

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

View File

@@ -1,560 +0,0 @@
private import cpp as Cpp
private import DataFlowUtil
private import semmle.code.cpp.ir.IR
private import DataFlowDispatch
private import DataFlowImplConsistency
private import semmle.code.cpp.ir.internal.IRCppLanguage
private import SsaInternals as Ssa
/** Gets the callable in which this node occurs. */
DataFlowCallable nodeGetEnclosingCallable(Node n) { result = n.getEnclosingCallable() }
/** Holds if `p` is a `ParameterNode` of `c` with position `pos`. */
predicate isParameterNode(ParameterNode p, DataFlowCallable c, ParameterPosition pos) {
p.isParameterOf(c, pos)
}
/** Holds if `arg` is an `ArgumentNode` of `c` with position `pos`. */
predicate isArgumentNode(ArgumentNode arg, DataFlowCall c, ArgumentPosition pos) {
arg.argumentOf(c, pos)
}
/**
* A data flow node that occurs as the argument of a call and is passed as-is
* to the callable. Instance arguments (`this` pointer) and read side effects
* on parameters are also included.
*/
abstract class ArgumentNode extends Node {
/**
* Holds if this argument occurs at the given position in the given call.
* The instance argument is considered to have index `-1`.
*/
abstract predicate argumentOf(DataFlowCall call, ArgumentPosition pos);
/** Gets the call in which this node is an argument. */
DataFlowCall getCall() { this.argumentOf(result, _) }
}
/**
* A data flow node that occurs as the argument to a call, or an
* implicit `this` pointer argument.
*/
private class PrimaryArgumentNode extends ArgumentNode, OperandNode {
override ArgumentOperand op;
PrimaryArgumentNode() { exists(CallInstruction call | op = call.getAnArgumentOperand()) }
override predicate argumentOf(DataFlowCall call, ArgumentPosition pos) {
op = call.getArgumentOperand(pos.(DirectPosition).getIndex())
}
override string toStringImpl() { result = argumentOperandToString(op) }
}
private string argumentOperandToString(ArgumentOperand op) {
exists(Expr unconverted |
unconverted = op.getDef().getUnconvertedResultExpression() and
result = unconverted.toString()
)
or
// Certain instructions don't map to an unconverted result expression. For these cases
// we fall back to a simpler naming scheme. This can happen in IR-generated constructors.
not exists(op.getDef().getUnconvertedResultExpression()) and
(
result = "Argument " + op.(PositionalArgumentOperand).getIndex()
or
op instanceof ThisArgumentOperand and result = "Argument this"
)
}
private class SideEffectArgumentNode extends ArgumentNode, SideEffectOperandNode {
override predicate argumentOf(DataFlowCall dfCall, ArgumentPosition pos) {
this.getCallInstruction() = dfCall and
pos.(IndirectionPosition).getArgumentIndex() = this.getArgumentIndex() and
pos.(IndirectionPosition).getIndirectionIndex() = super.getIndirectionIndex()
}
override string toStringImpl() {
result = argumentOperandToString(this.getAddressOperand()) + " indirection"
}
}
/** A parameter position represented by an integer. */
class ParameterPosition = Position;
/** An argument position represented by an integer. */
class ArgumentPosition = Position;
class Position extends TPosition {
abstract string toString();
}
class DirectPosition extends Position, TDirectPosition {
int index;
DirectPosition() { this = TDirectPosition(index) }
override string toString() { if index = -1 then result = "this" else result = index.toString() }
int getIndex() { result = index }
}
class IndirectionPosition extends Position, TIndirectionPosition {
int argumentIndex;
int indirectionIndex;
IndirectionPosition() { this = TIndirectionPosition(argumentIndex, indirectionIndex) }
override string toString() {
if argumentIndex = -1
then if indirectionIndex > 0 then result = "this indirection" else result = "this"
else
if indirectionIndex > 0
then result = argumentIndex.toString() + " indirection"
else result = argumentIndex.toString()
}
int getArgumentIndex() { result = argumentIndex }
int getIndirectionIndex() { result = indirectionIndex }
}
newtype TPosition =
TDirectPosition(int index) { exists(any(CallInstruction c).getArgument(index)) } or
TIndirectionPosition(int argumentIndex, int indirectionIndex) {
hasOperandAndIndex(_, any(CallInstruction call).getArgumentOperand(argumentIndex),
indirectionIndex)
}
private newtype TReturnKind =
TNormalReturnKind(int index) {
exists(IndirectReturnNode return |
return.getAddressOperand() = any(ReturnValueInstruction r).getReturnAddressOperand() and
index = return.getIndirectionIndex() - 1 // We subtract one because the return loads the value.
)
} or
TIndirectReturnKind(int argumentIndex, int indirectionIndex) {
exists(IndirectReturnNode return, ReturnIndirectionInstruction returnInd |
returnInd.hasIndex(argumentIndex) and
return.getAddressOperand() = returnInd.getSourceAddressOperand() and
indirectionIndex = return.getIndirectionIndex()
)
}
/**
* A return kind. A return kind describes how a value can be returned
* from a callable. For C++, this is simply a function return.
*/
class ReturnKind extends TReturnKind {
/** Gets a textual representation of this return kind. */
abstract string toString();
}
private class NormalReturnKind extends ReturnKind, TNormalReturnKind {
int index;
NormalReturnKind() { this = TNormalReturnKind(index) }
override string toString() { result = "indirect return" }
}
private class IndirectReturnKind extends ReturnKind, TIndirectReturnKind {
int argumentIndex;
int indirectionIndex;
IndirectReturnKind() { this = TIndirectReturnKind(argumentIndex, indirectionIndex) }
override string toString() { result = "indirect outparam[" + argumentIndex.toString() + "]" }
}
/** A data flow node that occurs as the result of a `ReturnStmt`. */
class ReturnNode extends Node instanceof IndirectReturnNode {
/** Gets the kind of this returned value. */
abstract ReturnKind getKind();
}
/**
* This predicate represents an annoying hack that we have to do. We use the
* `ReturnIndirectionInstruction` to determine which variables need flow back
* out of a function. However, the IR will unconditionally create those for a
* variable passed to a function even though the variable was never updated by
* the function. And if a function has too many `ReturnNode`s the dataflow
* library lowers its precision for that function by disabling field flow.
*
* So we those eliminate `ReturnNode`s that would have otherwise been created
* by this unconditional `ReturnIndirectionInstruction` by requiring that there
* must exist an SSA definition of the IR variable in the function.
*/
private predicate hasNonInitializeParameterDef(IRVariable v) {
exists(Ssa::Def def |
not def.getDefiningInstruction() instanceof InitializeParameterInstruction and
v = def.getSourceVariable().getBaseVariable().(Ssa::BaseIRVariable).getIRVariable()
)
}
class ReturnIndirectionNode extends IndirectReturnNode, ReturnNode {
override ReturnKind getKind() {
exists(int argumentIndex, ReturnIndirectionInstruction returnInd |
returnInd.hasIndex(argumentIndex) and
this.getAddressOperand() = returnInd.getSourceAddressOperand() and
result = TIndirectReturnKind(argumentIndex, this.getIndirectionIndex()) and
hasNonInitializeParameterDef(returnInd.getIRVariable())
)
or
this.getAddressOperand() = any(ReturnValueInstruction r).getReturnAddressOperand() and
result = TNormalReturnKind(this.getIndirectionIndex() - 1)
}
}
private Operand fullyConvertedCallStep(Operand op) {
not exists(getANonConversionUse(op)) and
exists(Instruction instr |
conversionFlow(op, instr, _) and
result = getAUse(instr)
)
}
/**
* Gets the instruction that uses this operand, if the instruction is not
* ignored for dataflow purposes.
*/
private Instruction getUse(Operand op) {
result = op.getUse() and
not Ssa::ignoreOperand(op)
}
/** Gets a use of the instruction `instr` that is not ignored for dataflow purposes. */
Operand getAUse(Instruction instr) {
result = instr.getAUse() and
not Ssa::ignoreOperand(result)
}
/**
* Gets a use of `operand` that is:
* - not ignored for dataflow purposes, and
* - not a conversion-like instruction.
*/
private Instruction getANonConversionUse(Operand operand) {
result = getUse(operand) and
not conversionFlow(_, result, _)
}
/**
* Gets the operand that represents the first use of the value of `call` following
* a sequence of conversion-like instructions.
*/
predicate operandForfullyConvertedCall(Operand operand, CallInstruction call) {
exists(getANonConversionUse(operand)) and
(
operand = getAUse(call)
or
operand = fullyConvertedCallStep*(getAUse(call))
)
}
/**
* Gets the instruction that represents the first use of the value of `call` following
* a sequence of conversion-like instructions.
*
* This predicate only holds if there is no suitable operand (i.e., no operand of a non-
* conversion instruction) to use to represent the value of `call` after conversions.
*/
predicate instructionForfullyConvertedCall(Instruction instr, CallInstruction call) {
not operandForfullyConvertedCall(_, call) and
(
// If there is no use of the call then we pick the call instruction
not exists(getAUse(call)) and
instr = call
or
// Otherwise, flow to the first non-conversion use.
exists(Operand operand | operand = fullyConvertedCallStep*(getAUse(call)) |
instr = getANonConversionUse(operand)
)
)
}
/** Holds if `node` represents the output node for `call`. */
private predicate simpleOutNode(Node node, CallInstruction call) {
operandForfullyConvertedCall(node.asOperand(), call)
or
instructionForfullyConvertedCall(node.asInstruction(), call)
}
/** A data flow node that represents the output of a call. */
class OutNode extends Node {
OutNode() {
// Return values not hidden behind indirections
simpleOutNode(this, _)
or
// Return values hidden behind indirections
this instanceof IndirectReturnOutNode
or
// Modified arguments hidden behind indirections
this instanceof IndirectArgumentOutNode
}
/** Gets the underlying call. */
abstract DataFlowCall getCall();
abstract ReturnKind getReturnKind();
}
private class DirectCallOutNode extends OutNode {
CallInstruction call;
DirectCallOutNode() { simpleOutNode(this, call) }
override DataFlowCall getCall() { result = call }
override ReturnKind getReturnKind() { result = TNormalReturnKind(0) }
}
private class IndirectCallOutNode extends OutNode, IndirectReturnOutNode {
override DataFlowCall getCall() { result = this.getCallInstruction() }
override ReturnKind getReturnKind() { result = TNormalReturnKind(this.getIndirectionIndex()) }
}
private class SideEffectOutNode extends OutNode, IndirectArgumentOutNode {
override DataFlowCall getCall() { result = this.getCallInstruction() }
override ReturnKind getReturnKind() {
result = TIndirectReturnKind(this.getArgumentIndex(), this.getIndirectionIndex())
}
}
/**
* Gets a node that can read the value returned from `call` with return kind
* `kind`.
*/
OutNode getAnOutNode(DataFlowCall call, ReturnKind kind) {
result.getCall() = call and
result.getReturnKind() = kind
}
/**
* Holds if data can flow from `node1` to `node2` in a way that loses the
* calling context. For example, this would happen with flow through a
* global or static variable.
*/
predicate jumpStep(Node n1, Node n2) {
exists(Cpp::GlobalOrNamespaceVariable v |
v =
n1.asInstruction()
.(StoreInstruction)
.getResultAddress()
.(VariableAddressInstruction)
.getAstVariable() and
v = n2.asVariable()
or
v =
n2.asInstruction()
.(LoadInstruction)
.getSourceAddress()
.(VariableAddressInstruction)
.getAstVariable() and
v = n1.asVariable()
)
}
/**
* Holds if data can flow from `node1` to `node2` via an assignment to `f`.
* Thus, `node2` references an object with a field `f` that contains the
* value of `node1`.
*/
predicate storeStep(Node node1, Content c, PostFieldUpdateNode node2) {
exists(int indirectionIndex1, int numberOfLoads, StoreInstruction store |
nodeHasInstruction(node1, store, pragma[only_bind_into](indirectionIndex1)) and
node2.getIndirectionIndex() = 1 and
numberOfLoadsFromOperand(node2.getFieldAddress(), store.getDestinationAddressOperand(),
numberOfLoads)
|
exists(FieldContent fc | fc = c |
fc.getField() = node2.getUpdatedField() and
fc.getIndirectionIndex() = 1 + indirectionIndex1 + numberOfLoads
)
or
exists(UnionContent uc | uc = c |
uc.getAField() = node2.getUpdatedField() and
uc.getIndirectionIndex() = 1 + indirectionIndex1 + numberOfLoads
)
)
}
/**
* Holds if `operandFrom` flows to `operandTo` using a sequence of conversion-like
* operations and exactly `n` `LoadInstruction` operations.
*/
private predicate numberOfLoadsFromOperandRec(Operand operandFrom, Operand operandTo, int ind) {
exists(LoadInstruction load | load.getSourceAddressOperand() = operandFrom |
operandTo = operandFrom and ind = 0
or
numberOfLoadsFromOperand(load.getAUse(), operandTo, ind - 1)
)
or
exists(Operand op, Instruction instr |
instr = op.getDef() and
conversionFlow(operandFrom, instr, _) and
numberOfLoadsFromOperand(op, operandTo, ind)
)
}
/**
* Holds if `operandFrom` flows to `operandTo` using a sequence of conversion-like
* operations and exactly `n` `LoadInstruction` operations.
*/
private predicate numberOfLoadsFromOperand(Operand operandFrom, Operand operandTo, int n) {
numberOfLoadsFromOperandRec(operandFrom, operandTo, n)
or
not any(LoadInstruction load).getSourceAddressOperand() = operandFrom and
not conversionFlow(operandFrom, _, _) and
operandFrom = operandTo and
n = 0
}
// Needed to join on both an operand and an index at the same time.
pragma[noinline]
predicate nodeHasOperand(Node node, Operand operand, int indirectionIndex) {
node.asOperand() = operand and indirectionIndex = 0
or
hasOperandAndIndex(node, operand, indirectionIndex)
}
// Needed to join on both an instruction and an index at the same time.
pragma[noinline]
predicate nodeHasInstruction(Node node, Instruction instr, int indirectionIndex) {
node.asInstruction() = instr and indirectionIndex = 0
or
hasInstructionAndIndex(node, instr, indirectionIndex)
}
/**
* Holds if data can flow from `node1` to `node2` via a read of `f`.
* Thus, `node1` references an object with a field `f` whose value ends up in
* `node2`.
*/
predicate readStep(Node node1, Content c, Node node2) {
exists(FieldAddress fa1, Operand operand, int numberOfLoads, int indirectionIndex2 |
nodeHasOperand(node2, operand, indirectionIndex2) and
nodeHasOperand(node1, fa1.getObjectAddressOperand(), _) and
numberOfLoadsFromOperand(fa1, operand, numberOfLoads)
|
exists(FieldContent fc | fc = c |
fc.getField() = fa1.getField() and
fc.getIndirectionIndex() = indirectionIndex2 + numberOfLoads
)
or
exists(UnionContent uc | uc = c |
uc.getAField() = fa1.getField() and
uc.getIndirectionIndex() = indirectionIndex2 + numberOfLoads
)
)
}
/**
* Holds if values stored inside content `c` are cleared at node `n`.
*/
predicate clearsContent(Node n, Content c) {
none() // stub implementation
}
/**
* Holds if the value that is being tracked is expected to be stored inside content `c`
* at node `n`.
*/
predicate expectsContent(Node n, ContentSet c) { none() }
/** Gets the type of `n` used for type pruning. */
DataFlowType getNodeType(Node n) {
suppressUnusedNode(n) and
result instanceof VoidType // stub implementation
}
/** Gets a string representation of a type returned by `getNodeType`. */
string ppReprType(DataFlowType t) { none() } // stub implementation
/**
* Holds if `t1` and `t2` are compatible, that is, whether data can flow from
* a node of type `t1` to a node of type `t2`.
*/
pragma[inline]
predicate compatibleTypes(DataFlowType t1, DataFlowType t2) {
any() // stub implementation
}
private predicate suppressUnusedNode(Node n) { any() }
//////////////////////////////////////////////////////////////////////////////
// Java QL library compatibility wrappers
//////////////////////////////////////////////////////////////////////////////
/** A node that performs a type cast. */
class CastNode extends Node {
CastNode() { none() } // stub implementation
}
/**
* A function that may contain code or a variable that may contain itself. When
* flow crosses from one _enclosing callable_ to another, the interprocedural
* data-flow library discards call contexts and inserts a node in the big-step
* relation used for human-readable path explanations.
*/
class DataFlowCallable = Cpp::Declaration;
class DataFlowExpr = Expr;
class DataFlowType = Type;
/** A function call relevant for data flow. */
class DataFlowCall extends CallInstruction {
Function getEnclosingCallable() { result = this.getEnclosingFunction() }
}
predicate isUnreachableInCall(Node n, DataFlowCall call) { none() } // stub implementation
int accessPathLimit() { result = 5 }
/**
* Holds if access paths with `c` at their head always should be tracked at high
* precision. This disables adaptive access path precision for such access paths.
*/
predicate forceHighPrecision(Content c) { none() }
/** The unit type. */
private newtype TUnit = TMkUnit()
/** The trivial type with a single element. */
class Unit extends TUnit {
/** Gets a textual representation of this element. */
string toString() { result = "unit" }
}
/** Holds if `n` should be hidden from path explanations. */
predicate nodeIsHidden(Node n) { n instanceof OperandNode and not n instanceof ArgumentNode }
class LambdaCallKind = Unit;
/** Holds if `creation` is an expression that creates a lambda of kind `kind` for `c`. */
predicate lambdaCreation(Node creation, LambdaCallKind kind, DataFlowCallable c) { none() }
/** Holds if `call` is a lambda call of kind `kind` where `receiver` is the lambda expression. */
predicate lambdaCall(DataFlowCall call, LambdaCallKind kind, Node receiver) { none() }
/** Extra data-flow steps needed for lambda flow analysis. */
predicate additionalLambdaFlowStep(Node nodeFrom, Node nodeTo, boolean preservesValue) { none() }
/**
* Holds if flow is allowed to pass from parameter `p` and back to itself as a
* side-effect, resulting in a summary from `p` to itself.
*
* One example would be to allow flow like `p.foo = p.bar;`, which is disallowed
* by default as a heuristic.
*/
predicate allowParameterReturnInSelf(ParameterNode p) { none() }
private class MyConsistencyConfiguration extends Consistency::ConsistencyConfiguration {
override predicate argHasPostUpdateExclude(ArgumentNode n) {
// The rules for whether an IR argument gets a post-update node are too
// complex to model here.
any()
}
}

View File

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

View File

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

View File

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

View File

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

View File

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

View File

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

View File

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

View File

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

View File

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

View File

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

View File

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

View File

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

View File

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

View File

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

View File

@@ -37,7 +37,7 @@ abstract class SimpleRangeAnalysisDefinition extends RangeSsaDefinition {
* dependencies. Without this information, range analysis might work for
* simple cases but will go into infinite loops on complex code.
*
* For example, when modeling the definition by reference in a call to an
* For example, when modelling the definition by reference in a call to an
* overloaded `operator=`, written as `v = e`, the definition of `(this, v)`
* depends on `e`.
*/

View File

@@ -28,10 +28,6 @@ private newtype TBound =
i.(LoadInstruction).getSourceAddress() instanceof FieldAddressInstruction
or
i.getAUse() instanceof ArgumentOperand
or
i instanceof PointerArithmeticInstruction
or
i.getAUse() instanceof AddressOperand
)
}
@@ -77,7 +73,7 @@ class ValueNumberBound extends Bound, TBoundValueNumber {
this = TBoundValueNumber(valueNumber(result)) and delta = 0
}
override string toString() { result = "ValueNumberBound" }
override string toString() { result = vn.getExampleInstruction().toString() }
override Location getLocation() { result = vn.getLocation() }

View File

@@ -5,7 +5,7 @@
* `Instruction` level), and then using the array length analysis and the range
* analysis together to prove that some of these pointer dereferences are safe.
*
* The analysis is soundy, i.e. it is sound if no undefined behavior is present
* The analysis is soundy, i.e. it is sound if no undefined behaviour is present
* in the program.
* Furthermore, it crucially depends on the soundiness of the range analysis and
* the array length analysis.

View File

@@ -4,7 +4,7 @@
import cpp
import semmle.code.cpp.dataflow.TaintTracking
import semmle.code.cpp.security.PrivateData
import experimental.semmle.code.cpp.security.PrivateData
import semmle.code.cpp.security.FileWrite
import semmle.code.cpp.security.BufferWrite

View File

@@ -0,0 +1,52 @@
/**
* Provides classes and predicates for identifying private data and functions for security.
*
* 'Private' data in general is anything that would compromise user privacy if exposed. This
* library tries to guess where private data may either be stored in a variable or produced by a
* function.
*
* This library is not concerned with credentials. See `SensitiveActions` for expressions related
* to credentials.
*/
import cpp
/** A string for `match` that identifies strings that look like they represent private data. */
private string privateNames() {
result =
[
// Inspired by the list on https://cwe.mitre.org/data/definitions/359.html
// Government identifiers, such as Social Security Numbers
"%social%security%number%",
// Contact information, such as home addresses and telephone numbers
"%postcode%", "%zipcode%",
// result = "%telephone%" or
// Geographic location - where the user is (or was)
"%latitude%", "%longitude%",
// Financial data - such as credit card numbers, salary, bank accounts, and debts
"%creditcard%", "%salary%", "%bankaccount%",
// Communications - e-mail addresses, private e-mail messages, SMS text messages, chat logs, etc.
// result = "%email%" or
// result = "%mobile%" or
"%employer%",
// Health - medical conditions, insurance status, prescription records
"%medical%"
]
}
/** An expression that might contain private data. */
abstract class PrivateDataExpr extends Expr { }
/** A functiond call that might produce private data. */
class PrivateFunctionCall extends PrivateDataExpr, FunctionCall {
PrivateFunctionCall() {
exists(string s | this.getTarget().getName().toLowerCase() = s | s.matches(privateNames()))
}
}
/** An access to a variable that might contain private data. */
class PrivateVariableAccess extends PrivateDataExpr, VariableAccess {
PrivateVariableAccess() {
exists(string s | this.getTarget().getName().toLowerCase() = s | s.matches(privateNames()))
}
}

View File

@@ -1,7 +0,0 @@
import SemanticExpr
import SemanticBound
import SemanticSSA
import SemanticGuard
import SemanticCFG
import SemanticType
import SemanticOpcode

View File

@@ -1,42 +0,0 @@
/**
* Semantic wrapper around the language-specific bounds library.
*/
private import SemanticExpr
private import SemanticExprSpecific::SemanticExprConfig as Specific
private import SemanticSSA
/**
* A valid base for an expression bound.
*
* Can be either a variable (`SemSsaBound`) or zero (`SemZeroBound`).
*/
class SemBound instanceof Specific::Bound {
final string toString() { result = super.toString() }
final SemExpr getExpr(int delta) { result = Specific::getBoundExpr(this, delta) }
}
/**
* A bound that is a constant zero.
*/
class SemZeroBound extends SemBound {
SemZeroBound() { Specific::zeroBound(this) }
}
/**
* A bound that is an SSA definition.
*/
class SemSsaBound extends SemBound {
/**
* The variables whose value is used as the bound.
*
* Can be multi-valued in some implementations. If so, all variables will be equivalent.
*/
SemSsaVariable var;
SemSsaBound() { Specific::ssaBound(this, var) }
/** Gets a variable whose value is used as the bound. */
final SemSsaVariable getAVariable() { result = var }
}

View File

@@ -1,22 +0,0 @@
/**
* Semantic interface to the control flow graph.
*/
private import Semantic
private import SemanticExprSpecific::SemanticExprConfig as Specific
/**
* A basic block in the control-flow graph.
*/
class SemBasicBlock extends Specific::BasicBlock {
/** Holds if this block (transitively) dominates `otherblock`. */
final predicate bbDominates(SemBasicBlock otherBlock) { Specific::bbDominates(this, otherBlock) }
/** Holds if this block has dominance information. */
final predicate hasDominanceInformation() { Specific::hasDominanceInformation(this) }
/** Gets an expression that is evaluated in this basic block. */
final SemExpr getAnExpr() { result.getBasicBlock() = this }
final int getUniqueId() { result = Specific::getBasicBlockUniqueId(this) }
}

View File

@@ -1,309 +0,0 @@
/**
* Semantic interface for expressions.
*/
private import Semantic
private import SemanticExprSpecific::SemanticExprConfig as Specific
/**
* An language-neutral expression.
*
* The expression computes a value of type `getSemType()`. The actual computation is determined by
* the expression's opcode (`getOpcode()`).
*/
class SemExpr instanceof Specific::Expr {
final string toString() { result = super.toString() }
final Specific::Location getLocation() { result = super.getLocation() }
Opcode getOpcode() { result instanceof Opcode::Unknown }
SemType getSemType() { result = Specific::getUnknownExprType(this) }
final SemBasicBlock getBasicBlock() { result = Specific::getExprBasicBlock(this) }
}
/** An expression with an opcode other than `Unknown`. */
abstract private class SemKnownExpr extends SemExpr {
Opcode opcode;
SemType type;
final override Opcode getOpcode() { result = opcode }
final override SemType getSemType() { result = type }
}
/** An expression that returns a literal value. */
class SemLiteralExpr extends SemKnownExpr {
SemLiteralExpr() {
Specific::integerLiteral(this, type, _) and opcode instanceof Opcode::Constant
or
Specific::largeIntegerLiteral(this, type, _) and opcode instanceof Opcode::Constant
or
Specific::booleanLiteral(this, type, _) and opcode instanceof Opcode::Constant
or
Specific::floatingPointLiteral(this, type, _) and opcode instanceof Opcode::Constant
or
Specific::nullLiteral(this, type) and opcode instanceof Opcode::Constant
or
Specific::stringLiteral(this, type, _) and opcode instanceof Opcode::StringConstant
}
}
/** An expression that returns a numeric literal value. */
class SemNumericLiteralExpr extends SemLiteralExpr {
SemNumericLiteralExpr() {
Specific::integerLiteral(this, _, _)
or
Specific::largeIntegerLiteral(this, _, _)
or
Specific::floatingPointLiteral(this, _, _)
}
/**
* Gets an approximation of the value of the literal, as a `float`.
*
* If the value can be precisely represented as a `float`, the result will be exact. If the actual
* value cannot be precisely represented (for example, it is an integer with more than 53
* significant bits), then the result is an approximation.
*/
float getApproximateFloatValue() { none() }
}
/** An expression that returns an integer literal value. */
class SemIntegerLiteralExpr extends SemNumericLiteralExpr {
SemIntegerLiteralExpr() {
Specific::integerLiteral(this, _, _)
or
Specific::largeIntegerLiteral(this, _, _)
}
/**
* Gets the value of the literal, if it can be represented as an `int`.
*
* If the value is outside the range of an `int`, use `getApproximateFloatValue()` to get a value
* that is equal to the actual integer value, within rounding error.
*/
final int getIntValue() { Specific::integerLiteral(this, _, result) }
final override float getApproximateFloatValue() {
result = getIntValue()
or
Specific::largeIntegerLiteral(this, _, result)
}
}
/**
* An expression that returns a floating-point literal value.
*/
class SemFloatingPointLiteralExpr extends SemNumericLiteralExpr {
float value;
SemFloatingPointLiteralExpr() { Specific::floatingPointLiteral(this, _, value) }
final override float getApproximateFloatValue() { result = value }
/** Gets the value of the literal. */
final float getFloatValue() { result = value }
}
/**
* An expression that consumes two operands.
*/
class SemBinaryExpr extends SemKnownExpr {
SemExpr leftOperand;
SemExpr rightOperand;
SemBinaryExpr() { Specific::binaryExpr(this, opcode, type, leftOperand, rightOperand) }
/** Gets the left operand. */
final SemExpr getLeftOperand() { result = leftOperand }
/** Gets the right operand. */
final SemExpr getRightOperand() { result = rightOperand }
/** Holds if `a` and `b` are the two operands, in either order. */
final predicate hasOperands(SemExpr a, SemExpr b) {
a = getLeftOperand() and b = getRightOperand()
or
a = getRightOperand() and b = getLeftOperand()
}
/** Gets the two operands. */
final SemExpr getAnOperand() { result = getLeftOperand() or result = getRightOperand() }
}
/** An expression that performs and ordered comparison of two operands. */
class SemRelationalExpr extends SemBinaryExpr {
SemRelationalExpr() {
opcode instanceof Opcode::CompareLT
or
opcode instanceof Opcode::CompareLE
or
opcode instanceof Opcode::CompareGT
or
opcode instanceof Opcode::CompareGE
}
/**
* Get the operand that will be less than the other operand if the result of the comparison is
* `true`.
*
* For `x < y` or `x <= y`, this will return `x`.
* For `x > y` or `x >= y`, this will return `y`.`
*/
final SemExpr getLesserOperand() {
if opcode instanceof Opcode::CompareLT or opcode instanceof Opcode::CompareLE
then result = getLeftOperand()
else result = getRightOperand()
}
/**
* Get the operand that will be greater than the other operand if the result of the comparison is
* `true`.
*
* For `x < y` or `x <= y`, this will return `y`.
* For `x > y` or `x >= y`, this will return `x`.`
*/
final SemExpr getGreaterOperand() {
if opcode instanceof Opcode::CompareGT or opcode instanceof Opcode::CompareGE
then result = getLeftOperand()
else result = getRightOperand()
}
/** Holds if this comparison returns `false` if the two operands are equal. */
final predicate isStrict() {
opcode instanceof Opcode::CompareLT or opcode instanceof Opcode::CompareGT
}
}
class SemAddExpr extends SemBinaryExpr {
SemAddExpr() { opcode instanceof Opcode::Add or opcode instanceof Opcode::PointerAdd }
}
class SemSubExpr extends SemBinaryExpr {
SemSubExpr() { opcode instanceof Opcode::Sub or opcode instanceof Opcode::PointerSub }
}
class SemMulExpr extends SemBinaryExpr {
SemMulExpr() { opcode instanceof Opcode::Mul }
}
class SemDivExpr extends SemBinaryExpr {
SemDivExpr() { opcode instanceof Opcode::Div }
}
class SemRemExpr extends SemBinaryExpr {
SemRemExpr() { opcode instanceof Opcode::Rem }
}
class SemShiftLeftExpr extends SemBinaryExpr {
SemShiftLeftExpr() { opcode instanceof Opcode::ShiftLeft }
}
class SemShiftRightExpr extends SemBinaryExpr {
SemShiftRightExpr() { opcode instanceof Opcode::ShiftRight }
}
class SemShiftRightUnsignedExpr extends SemBinaryExpr {
SemShiftRightUnsignedExpr() { opcode instanceof Opcode::ShiftRightUnsigned }
}
class SemBitAndExpr extends SemBinaryExpr {
SemBitAndExpr() { opcode instanceof Opcode::BitAnd }
}
class SemBitOrExpr extends SemBinaryExpr {
SemBitOrExpr() { opcode instanceof Opcode::BitOr }
}
class SemBitXorExpr extends SemBinaryExpr {
SemBitXorExpr() { opcode instanceof Opcode::BitXor }
}
class SemUnaryExpr extends SemKnownExpr {
SemExpr operand;
SemUnaryExpr() { Specific::unaryExpr(this, opcode, type, operand) }
final SemExpr getOperand() { result = operand }
}
class SemBoxExpr extends SemUnaryExpr {
SemBoxExpr() { opcode instanceof Opcode::Box }
}
class SemUnboxExpr extends SemUnaryExpr {
SemUnboxExpr() { opcode instanceof Opcode::Unbox }
}
class SemConvertExpr extends SemUnaryExpr {
SemConvertExpr() { opcode instanceof Opcode::Convert }
}
class SemCopyValueExpr extends SemUnaryExpr {
SemCopyValueExpr() { opcode instanceof Opcode::CopyValue }
}
class SemNegateExpr extends SemUnaryExpr {
SemNegateExpr() { opcode instanceof Opcode::Negate }
}
class SemBitComplementExpr extends SemUnaryExpr {
SemBitComplementExpr() { opcode instanceof Opcode::BitComplement }
}
class SemLogicalNotExpr extends SemUnaryExpr {
SemLogicalNotExpr() { opcode instanceof Opcode::LogicalNot }
}
class SemAddOneExpr extends SemUnaryExpr {
SemAddOneExpr() { opcode instanceof Opcode::AddOne }
}
class SemSubOneExpr extends SemUnaryExpr {
SemSubOneExpr() { opcode instanceof Opcode::SubOne }
}
private class SemNullaryExpr extends SemKnownExpr {
SemNullaryExpr() { Specific::nullaryExpr(this, opcode, type) }
}
class SemInitializeParameterExpr extends SemNullaryExpr {
SemInitializeParameterExpr() { opcode instanceof Opcode::InitializeParameter }
}
class SemLoadExpr extends SemNullaryExpr {
SemLoadExpr() { opcode instanceof Opcode::Load }
final SemSsaVariable getDef() { result.getAUse() = this }
}
class SemSsaLoadExpr extends SemLoadExpr {
SemSsaLoadExpr() { exists(getDef()) }
}
class SemNonSsaLoadExpr extends SemLoadExpr {
SemNonSsaLoadExpr() { not exists(getDef()) }
}
class SemStoreExpr extends SemUnaryExpr {
SemStoreExpr() { opcode instanceof Opcode::Store }
}
class SemConditionalExpr extends SemKnownExpr {
SemExpr condition;
SemExpr trueResult;
SemExpr falseResult;
SemConditionalExpr() {
opcode instanceof Opcode::Conditional and
Specific::conditionalExpr(this, type, condition, trueResult, falseResult)
}
final SemExpr getBranchExpr(boolean branch) {
branch = true and result = trueResult
or
branch = false and result = falseResult
}
}

View File

@@ -1,360 +0,0 @@
/**
* C++-specific implementation of the semantic interface.
*/
private import cpp as Cpp
private import semmle.code.cpp.ir.IR as IR
private import Semantic
private import experimental.semmle.code.cpp.rangeanalysis.Bound as IRBound
private import semmle.code.cpp.controlflow.IRGuards as IRGuards
private import semmle.code.cpp.ir.ValueNumbering
module SemanticExprConfig {
class Location = Cpp::Location;
class Expr = IR::Instruction;
SemBasicBlock getExprBasicBlock(Expr e) { result = getSemanticBasicBlock(e.getBlock()) }
private predicate anyConstantExpr(Expr expr, SemType type, string value) {
exists(IR::ConstantInstruction instr | instr = expr |
type = getSemanticType(instr.getResultIRType()) and
value = instr.getValue()
)
}
predicate integerLiteral(Expr expr, SemIntegerType type, int value) {
exists(string valueString |
anyConstantExpr(expr, type, valueString) and
value = valueString.toInt()
)
}
predicate largeIntegerLiteral(Expr expr, SemIntegerType type, float approximateFloatValue) {
exists(string valueString |
anyConstantExpr(expr, type, valueString) and
not exists(valueString.toInt()) and
approximateFloatValue = valueString.toFloat()
)
}
predicate floatingPointLiteral(Expr expr, SemFloatingPointType type, float value) {
exists(string valueString |
anyConstantExpr(expr, type, valueString) and value = valueString.toFloat()
)
}
predicate booleanLiteral(Expr expr, SemBooleanType type, boolean value) {
exists(string valueString |
anyConstantExpr(expr, type, valueString) and
(
valueString = "true" and value = true
or
valueString = "false" and value = false
)
)
}
predicate nullLiteral(Expr expr, SemAddressType type) { anyConstantExpr(expr, type, _) }
predicate stringLiteral(Expr expr, SemType type, string value) {
anyConstantExpr(expr, type, value) and expr instanceof IR::StringConstantInstruction
}
predicate binaryExpr(Expr expr, Opcode opcode, SemType type, Expr leftOperand, Expr rightOperand) {
exists(IR::BinaryInstruction instr | instr = expr |
type = getSemanticType(instr.getResultIRType()) and
leftOperand = instr.getLeft() and
rightOperand = instr.getRight() and
// REVIEW: Merge the two `Opcode` types.
opcode.toString() = instr.getOpcode().toString()
)
}
predicate unaryExpr(Expr expr, Opcode opcode, SemType type, Expr operand) {
type = getSemanticType(expr.getResultIRType()) and
(
exists(IR::UnaryInstruction instr | instr = expr |
operand = instr.getUnary() and
// REVIEW: Merge the two operand types.
opcode.toString() = instr.getOpcode().toString()
)
or
exists(IR::StoreInstruction instr | instr = expr |
operand = instr.getSourceValue() and
opcode instanceof Opcode::Store
)
)
}
predicate nullaryExpr(Expr expr, Opcode opcode, SemType type) {
type = getSemanticType(expr.getResultIRType()) and
(
expr instanceof IR::LoadInstruction and opcode instanceof Opcode::Load
or
expr instanceof IR::InitializeParameterInstruction and
opcode instanceof Opcode::InitializeParameter
)
}
predicate conditionalExpr(
Expr expr, SemType type, Expr condition, Expr trueResult, Expr falseResult
) {
none()
}
SemType getUnknownExprType(Expr expr) { result = getSemanticType(expr.getResultIRType()) }
class BasicBlock = IR::IRBlock;
predicate bbDominates(BasicBlock dominator, BasicBlock dominated) {
dominator.dominates(dominated)
}
predicate hasDominanceInformation(BasicBlock block) { any() }
private predicate id(Cpp::Locatable x, Cpp::Locatable y) { x = y }
private predicate idOf(Cpp::Locatable x, int y) = equivalenceRelation(id/2)(x, y)
int getBasicBlockUniqueId(BasicBlock block) { idOf(block.getFirstInstruction().getAst(), result) }
newtype TSsaVariable =
TSsaInstruction(IR::Instruction instr) { instr.hasMemoryResult() } or
TSsaOperand(IR::Operand op) { op.isDefinitionInexact() } or
TSsaPointerArithmeticGuard(IR::PointerArithmeticInstruction instr) {
exists(Guard g, IR::Operand use | use = instr.getAUse() |
g.comparesLt(use, _, _, _, _) or
g.comparesLt(_, use, _, _, _) or
g.comparesEq(use, _, _, _, _) or
g.comparesEq(_, use, _, _, _)
)
}
class SsaVariable extends TSsaVariable {
string toString() { none() }
Location getLocation() { none() }
IR::Instruction asInstruction() { none() }
IR::PointerArithmeticInstruction asPointerArithGuard() { none() }
IR::Operand asOperand() { none() }
}
class SsaInstructionVariable extends SsaVariable, TSsaInstruction {
IR::Instruction instr;
SsaInstructionVariable() { this = TSsaInstruction(instr) }
final override string toString() { result = instr.toString() }
final override Location getLocation() { result = instr.getLocation() }
final override IR::Instruction asInstruction() { result = instr }
}
class SsaPointerArithmeticGuard extends SsaVariable, TSsaPointerArithmeticGuard {
IR::PointerArithmeticInstruction instr;
SsaPointerArithmeticGuard() { this = TSsaPointerArithmeticGuard(instr) }
final override string toString() { result = instr.toString() }
final override Location getLocation() { result = instr.getLocation() }
final override IR::PointerArithmeticInstruction asPointerArithGuard() { result = instr }
}
class SsaOperand extends SsaVariable, TSsaOperand {
IR::Operand op;
SsaOperand() { this = TSsaOperand(op) }
final override string toString() { result = op.toString() }
final override Location getLocation() { result = op.getLocation() }
final override IR::Operand asOperand() { result = op }
}
predicate explicitUpdate(SsaVariable v, Expr sourceExpr) { v.asInstruction() = sourceExpr }
predicate phi(SsaVariable v) { v.asInstruction() instanceof IR::PhiInstruction }
SsaVariable getAPhiInput(SsaVariable v) {
exists(IR::PhiInstruction instr | v.asInstruction() = instr |
result.asInstruction() = instr.getAnInput()
or
result.asOperand() = instr.getAnInputOperand()
)
}
Expr getAUse(SsaVariable v) {
result.(IR::LoadInstruction).getSourceValue() = v.asInstruction()
or
result = valueNumber(v.asPointerArithGuard()).getAnInstruction()
}
SemType getSsaVariableType(SsaVariable v) {
result = getSemanticType(v.asInstruction().getResultIRType())
}
BasicBlock getSsaVariableBasicBlock(SsaVariable v) {
result = v.asInstruction().getBlock()
or
result = v.asOperand().getUse().getBlock()
}
private newtype TReadPosition =
TReadPositionBlock(IR::IRBlock block) or
TReadPositionPhiInputEdge(IR::IRBlock pred, IR::IRBlock succ) {
exists(IR::PhiInputOperand input |
pred = input.getPredecessorBlock() and
succ = input.getUse().getBlock()
)
}
class SsaReadPosition extends TReadPosition {
string toString() { none() }
Location getLocation() { none() }
predicate hasRead(SsaVariable v) { none() }
}
private class SsaReadPositionBlock extends SsaReadPosition, TReadPositionBlock {
IR::IRBlock block;
SsaReadPositionBlock() { this = TReadPositionBlock(block) }
final override string toString() { result = block.toString() }
final override Location getLocation() { result = block.getLocation() }
final override predicate hasRead(SsaVariable v) {
exists(IR::Operand operand |
operand.getDef() = v.asInstruction() or
operand.getDef() = valueNumber(v.asPointerArithGuard()).getAnInstruction()
|
not operand instanceof IR::PhiInputOperand and
operand.getUse().getBlock() = block
)
}
}
private class SsaReadPositionPhiInputEdge extends SsaReadPosition, TReadPositionPhiInputEdge {
IR::IRBlock pred;
IR::IRBlock succ;
SsaReadPositionPhiInputEdge() { this = TReadPositionPhiInputEdge(pred, succ) }
final override string toString() { result = pred.toString() + "->" + succ.toString() }
final override Location getLocation() { result = succ.getLocation() }
final override predicate hasRead(SsaVariable v) {
exists(IR::PhiInputOperand operand |
operand.getDef() = v.asInstruction() or
operand.getDef() = valueNumber(v.asPointerArithGuard()).getAnInstruction()
|
operand.getPredecessorBlock() = pred and
operand.getUse().getBlock() = succ
)
}
}
predicate hasReadOfSsaVariable(SsaReadPosition pos, SsaVariable v) { pos.hasRead(v) }
predicate readBlock(SsaReadPosition pos, BasicBlock block) { pos = TReadPositionBlock(block) }
predicate phiInputEdge(SsaReadPosition pos, BasicBlock origBlock, BasicBlock phiBlock) {
pos = TReadPositionPhiInputEdge(origBlock, phiBlock)
}
predicate phiInput(SsaReadPosition pos, SsaVariable phi, SsaVariable input) {
exists(IR::PhiInputOperand operand |
pos = TReadPositionPhiInputEdge(operand.getPredecessorBlock(), operand.getUse().getBlock())
|
phi.asInstruction() = operand.getUse() and
(
input.asInstruction() = operand.getDef()
or
input.asOperand() = operand
)
)
}
class Bound instanceof IRBound::Bound {
string toString() { result = super.toString() }
final Location getLocation() { result = super.getLocation() }
}
private class ValueNumberBound extends Bound {
IRBound::ValueNumberBound bound;
ValueNumberBound() { bound = this }
override string toString() { result = bound.toString() }
}
predicate zeroBound(Bound bound) { bound instanceof IRBound::ZeroBound }
predicate ssaBound(Bound bound, SsaVariable v) {
v.asInstruction() = bound.(IRBound::ValueNumberBound).getValueNumber().getAnInstruction()
}
Expr getBoundExpr(Bound bound, int delta) {
result = bound.(IRBound::Bound).getInstruction(delta)
}
class Guard = IRGuards::IRGuardCondition;
predicate guard(Guard guard, BasicBlock block) { block = guard.getBlock() }
Expr getGuardAsExpr(Guard guard) { result = guard }
predicate equalityGuard(Guard guard, Expr e1, Expr e2, boolean polarity) {
guard.comparesEq(e1.getAUse(), e2.getAUse(), 0, true, polarity)
}
predicate guardDirectlyControlsBlock(Guard guard, BasicBlock controlled, boolean branch) {
guard.controls(controlled, branch)
}
predicate guardHasBranchEdge(Guard guard, BasicBlock bb1, BasicBlock bb2, boolean branch) {
guard.controlsEdge(bb1, bb2, branch)
}
Guard comparisonGuard(Expr e) { result = e }
predicate implies_v2(Guard g1, boolean b1, Guard g2, boolean b2) {
none() // TODO
}
}
SemExpr getSemanticExpr(IR::Instruction instr) { result = instr }
IR::Instruction getCppInstruction(SemExpr e) { e = result }
SemBasicBlock getSemanticBasicBlock(IR::IRBlock block) { result = block }
IR::IRBlock getCppBasicBlock(SemBasicBlock block) { block = result }
SemSsaVariable getSemanticSsaVariable(IR::Instruction instr) {
result.(SemanticExprConfig::SsaVariable).asInstruction() = instr
}
IR::Instruction getCppSsaVariableInstruction(SemSsaVariable var) {
var.(SemanticExprConfig::SsaVariable).asInstruction() = result
}
SemBound getSemanticBound(IRBound::Bound bound) { result = bound }
IRBound::Bound getCppBound(SemBound bound) { bound = result }
SemGuard getSemanticGuard(IRGuards::IRGuardCondition guard) { result = guard }
IRGuards::IRGuardCondition getCppGuard(SemGuard guard) { guard = result }

View File

@@ -1,65 +0,0 @@
/**
* Semantic interface to the guards library.
*/
private import Semantic
private import SemanticExprSpecific::SemanticExprConfig as Specific
class SemGuard instanceof Specific::Guard {
SemBasicBlock block;
SemGuard() { Specific::guard(this, block) }
final string toString() { result = super.toString() }
final Specific::Location getLocation() { result = super.getLocation() }
final predicate isEquality(SemExpr e1, SemExpr e2, boolean polarity) {
Specific::equalityGuard(this, e1, e2, polarity)
}
final predicate directlyControls(SemBasicBlock controlled, boolean branch) {
Specific::guardDirectlyControlsBlock(this, controlled, branch)
}
final predicate hasBranchEdge(SemBasicBlock bb1, SemBasicBlock bb2, boolean branch) {
Specific::guardHasBranchEdge(this, bb1, bb2, branch)
}
final SemBasicBlock getBasicBlock() { result = block }
final SemExpr asExpr() { result = Specific::getGuardAsExpr(this) }
}
predicate semImplies_v2(SemGuard g1, boolean b1, SemGuard g2, boolean b2) {
Specific::implies_v2(g1, b1, g2, b2)
}
/**
* Holds if `guard` directly controls the position `controlled` with the
* value `testIsTrue`.
*/
predicate semGuardDirectlyControlsSsaRead(
SemGuard guard, SemSsaReadPosition controlled, boolean testIsTrue
) {
guard.directlyControls(controlled.(SemSsaReadPositionBlock).getBlock(), testIsTrue)
or
exists(SemSsaReadPositionPhiInputEdge controlledEdge | controlledEdge = controlled |
guard.directlyControls(controlledEdge.getOrigBlock(), testIsTrue) or
guard.hasBranchEdge(controlledEdge.getOrigBlock(), controlledEdge.getPhiBlock(), testIsTrue)
)
}
/**
* Holds if `guard` controls the position `controlled` with the value `testIsTrue`.
*/
predicate semGuardControlsSsaRead(SemGuard guard, SemSsaReadPosition controlled, boolean testIsTrue) {
semGuardDirectlyControlsSsaRead(guard, controlled, testIsTrue)
or
exists(SemGuard guard0, boolean testIsTrue0 |
semImplies_v2(guard0, testIsTrue0, guard, testIsTrue) and
semGuardControlsSsaRead(guard0, controlled, testIsTrue0)
)
}
SemGuard semGetComparisonGuard(SemRelationalExpr e) { result = Specific::comparisonGuard(e) }

View File

@@ -1,187 +0,0 @@
/**
* Definitions of all possible opcodes for `SemExpr`.
*/
private newtype TOpcode =
TInitializeParameter() or
TCopyValue() or
TLoad() or
TStore() or
TAdd() or
TSub() or
TMul() or
TDiv() or
TRem() or
TNegate() or
TShiftLeft() or
TShiftRight() or
TShiftRightUnsigned() or // TODO: Based on type
TBitAnd() or
TBitOr() or
TBitXor() or
TBitComplement() or
TLogicalNot() or
TCompareEQ() or
TCompareNE() or
TCompareLT() or
TCompareGT() or
TCompareLE() or
TCompareGE() or
TPointerAdd() or
TPointerSub() or
TPointerDiff() or
TConvert() or
TConstant() or
TStringConstant() or
TAddOne() or // TODO: Combine with `TAdd`
TSubOne() or // TODO: Combine with `TSub`
TConditional() or // TODO: Represent as flow
TCall() or
TBox() or
TUnbox() or
TUnknown()
class Opcode extends TOpcode {
string toString() { result = "???" }
}
module Opcode {
class InitializeParameter extends Opcode, TInitializeParameter {
override string toString() { result = "InitializeParameter" }
}
class CopyValue extends Opcode, TCopyValue {
override string toString() { result = "CopyValue" }
}
class Load extends Opcode, TLoad {
override string toString() { result = "Load" }
}
class Store extends Opcode, TStore {
override string toString() { result = "Store" }
}
class Add extends Opcode, TAdd {
override string toString() { result = "Add" }
}
class PointerAdd extends Opcode, TPointerAdd {
override string toString() { result = "PointerAdd" }
}
class Sub extends Opcode, TSub {
override string toString() { result = "Sub" }
}
class PointerSub extends Opcode, TPointerSub {
override string toString() { result = "PointerSub" }
}
class Mul extends Opcode, TMul {
override string toString() { result = "Mul" }
}
class Div extends Opcode, TDiv {
override string toString() { result = "Div" }
}
class Rem extends Opcode, TRem {
override string toString() { result = "Rem" }
}
class Negate extends Opcode, TNegate {
override string toString() { result = "Negate" }
}
class ShiftLeft extends Opcode, TShiftLeft {
override string toString() { result = "ShiftLeft" }
}
class ShiftRight extends Opcode, TShiftRight {
override string toString() { result = "ShiftRight" }
}
class ShiftRightUnsigned extends Opcode, TShiftRightUnsigned {
override string toString() { result = "ShiftRightUnsigned" }
}
class BitAnd extends Opcode, TBitAnd {
override string toString() { result = "BitAnd" }
}
class BitOr extends Opcode, TBitOr {
override string toString() { result = "BitOr" }
}
class BitXor extends Opcode, TBitXor {
override string toString() { result = "BitXor" }
}
class BitComplement extends Opcode, TBitComplement {
override string toString() { result = "BitComplement" }
}
class LogicalNot extends Opcode, TLogicalNot {
override string toString() { result = "LogicalNot" }
}
class CompareEQ extends Opcode, TCompareEQ {
override string toString() { result = "CompareEQ" }
}
class CompareNE extends Opcode, TCompareNE {
override string toString() { result = "CompareNE" }
}
class CompareLT extends Opcode, TCompareLT {
override string toString() { result = "CompareLT" }
}
class CompareLE extends Opcode, TCompareLE {
override string toString() { result = "CompareLE" }
}
class CompareGT extends Opcode, TCompareGT {
override string toString() { result = "CompareGT" }
}
class CompareGE extends Opcode, TCompareGE {
override string toString() { result = "CompareGE" }
}
class Convert extends Opcode, TConvert {
override string toString() { result = "Convert" }
}
class AddOne extends Opcode, TAddOne {
override string toString() { result = "AddOne" }
}
class SubOne extends Opcode, TSubOne {
override string toString() { result = "SubOne" }
}
class Conditional extends Opcode, TConditional {
override string toString() { result = "Conditional" }
}
class Constant extends Opcode, TConstant {
override string toString() { result = "Constant" }
}
class StringConstant extends Opcode, TStringConstant {
override string toString() { result = "StringConstant" }
}
class Box extends Opcode, TBox {
override string toString() { result = "Box" }
}
class Unbox extends Opcode, TUnbox {
override string toString() { result = "Unbox" }
}
class Unknown extends Opcode, TUnknown {
override string toString() { result = "Unknown" }
}
}

View File

@@ -1,75 +0,0 @@
/**
* Semantic interface to the SSA library.
*/
private import Semantic
private import SemanticExprSpecific::SemanticExprConfig as Specific
class SemSsaVariable instanceof Specific::SsaVariable {
final string toString() { result = super.toString() }
final Specific::Location getLocation() { result = super.getLocation() }
final SemExpr getAUse() { result = Specific::getAUse(this) }
final SemType getType() { result = Specific::getSsaVariableType(this) }
final SemBasicBlock getBasicBlock() { result = Specific::getSsaVariableBasicBlock(this) }
}
class SemSsaExplicitUpdate extends SemSsaVariable {
SemExpr sourceExpr;
SemSsaExplicitUpdate() { Specific::explicitUpdate(this, sourceExpr) }
final SemExpr getSourceExpr() { result = sourceExpr }
}
class SemSsaPhiNode extends SemSsaVariable {
SemSsaPhiNode() { Specific::phi(this) }
final SemSsaVariable getAPhiInput() { result = Specific::getAPhiInput(this) }
}
class SemSsaReadPosition instanceof Specific::SsaReadPosition {
final string toString() { result = super.toString() }
final Specific::Location getLocation() { result = super.getLocation() }
final predicate hasReadOfVar(SemSsaVariable var) { Specific::hasReadOfSsaVariable(this, var) }
}
class SemSsaReadPositionPhiInputEdge extends SemSsaReadPosition {
SemBasicBlock origBlock;
SemBasicBlock phiBlock;
SemSsaReadPositionPhiInputEdge() { Specific::phiInputEdge(this, origBlock, phiBlock) }
predicate phiInput(SemSsaPhiNode phi, SemSsaVariable inp) { Specific::phiInput(this, phi, inp) }
SemBasicBlock getOrigBlock() { result = origBlock }
SemBasicBlock getPhiBlock() { result = phiBlock }
}
class SemSsaReadPositionBlock extends SemSsaReadPosition {
SemBasicBlock block;
SemSsaReadPositionBlock() { Specific::readBlock(this, block) }
SemBasicBlock getBlock() { result = block }
SemExpr getAnExpr() { result = getBlock().getAnExpr() }
}
/**
* Holds if `inp` is an input to `phi` along a back edge.
*/
predicate semBackEdge(SemSsaPhiNode phi, SemSsaVariable inp, SemSsaReadPositionPhiInputEdge edge) {
edge.phiInput(phi, inp) and
// Conservatively assume that every edge is a back edge if we don't have dominance information.
(
phi.getBasicBlock().bbDominates(edge.getOrigBlock()) or
not edge.getOrigBlock().hasDominanceInformation()
)
}

View File

@@ -1,301 +0,0 @@
/**
* Minimal, language-neutral type system for semantic analysis.
*/
private import SemanticTypeSpecific as Specific
class LanguageType = Specific::Type;
cached
private newtype TSemType =
TSemVoidType() { Specific::voidType(_) } or
TSemUnknownType() { Specific::unknownType(_) } or
TSemErrorType() { Specific::errorType(_) } or
TSemBooleanType(int byteSize) { Specific::booleanType(_, byteSize) } or
TSemIntegerType(int byteSize, boolean signed) { Specific::integerType(_, byteSize, signed) } or
TSemFloatingPointType(int byteSize) { Specific::floatingPointType(_, byteSize) } or
TSemAddressType(int byteSize) { Specific::addressType(_, byteSize) } or
TSemFunctionAddressType(int byteSize) { Specific::functionAddressType(_, byteSize) } or
TSemOpaqueType(int byteSize, Specific::OpaqueTypeTag tag) {
Specific::opaqueType(_, byteSize, tag)
}
/**
* The language-neutral type of a semantic expression,
* The interface to `SemType` and its subclasses is the same across all languages for which the IR
* is supported, so analyses that expect to be used for multiple languages should generally use
* `SemType` rather than a language-specific type.
*
* Many types from the language-specific type system will map to a single canonical `SemType`. Two
* types that map to the same `SemType` are considered equivalent by semantic analysis. As an
* example, in C++, all pointer types map to the same instance of `SemAddressType`.
*/
class SemType extends TSemType {
/** Gets a textual representation of this type. */
string toString() { none() }
/**
* Gets a string that uniquely identifies this `SemType`. This string is often the same as the
* result of `SemType.toString()`, but for some types it may be more verbose to ensure uniqueness.
*/
string getIdentityString() { result = toString() }
/**
* Gets the size of the type, in bytes, if known.
*
* This will hold for all `SemType` objects except `SemUnknownType` and `SemErrorType`.
*/
// This predicate is overridden with `pragma[noinline]` in every leaf subclass.
// This allows callers to ask for things like _the_ floating-point type of
// size 4 without getting a join that first finds all types of size 4 and
// _then_ restricts them to floating-point types.
int getByteSize() { none() }
}
/**
* An unknown type. Generally used to represent results and operands that access an unknown set of
* memory locations, such as the side effects of a function call.
*/
class SemUnknownType extends SemType, TSemUnknownType {
final override string toString() { result = "unknown" }
final override int getByteSize() { none() }
}
/**
* A void type, which has no values. Used to represent the result type of an expression that does
* not produce a result.
*/
class SemVoidType extends SemType, TSemVoidType {
final override string toString() { result = "void" }
final override int getByteSize() { result = 0 }
}
/**
* An error type. Used when an error in the source code prevents the extractor from determining the
* proper type.
*/
class SemErrorType extends SemType, TSemErrorType {
final override string toString() { result = "error" }
final override int getByteSize() { result = 0 }
}
private class SemSizedType extends SemType {
int byteSize;
SemSizedType() {
this = TSemBooleanType(byteSize) or
this = TSemIntegerType(byteSize, _) or
this = TSemFloatingPointType(byteSize) or
this = TSemAddressType(byteSize) or
this = TSemFunctionAddressType(byteSize) or
this = TSemOpaqueType(byteSize, _)
}
// Don't override `getByteSize()` here. The optimizer seems to generate better code when this is
// overridden only in the leaf classes.
}
/**
* A Boolean type, which can hold the values `true` (non-zero) or `false` (zero).
*/
class SemBooleanType extends SemSizedType, TSemBooleanType {
final override string toString() { result = "bool" + byteSize.toString() }
pragma[noinline]
final override int getByteSize() { result = byteSize }
}
/**
* A numeric type. This includes `SemSignedIntegerType`, `SemUnsignedIntegerType`, and
* `SemFloatingPointType`.
*/
class SemNumericType extends SemSizedType {
SemNumericType() {
this = TSemIntegerType(byteSize, _) or
this = TSemFloatingPointType(byteSize)
}
// Don't override `getByteSize()` here. The optimizer seems to generate better code when this is
// overridden only in the leaf classes.
}
/**
* An integer type. This includes `SemSignedIntegerType` and `SemUnsignedIntegerType`.
*/
class SemIntegerType extends SemNumericType {
boolean signed;
SemIntegerType() { this = TSemIntegerType(byteSize, signed) }
/** Holds if this integer type is signed. */
final predicate isSigned() { signed = true }
/** Holds if this integer type is unsigned. */
final predicate isUnsigned() { not isSigned() }
// Don't override `getByteSize()` here. The optimizer seems to generate better code when this is
// overridden only in the leaf classes.
}
/**
* A signed two's-complement integer. Also used to represent enums whose underlying type is a signed
* integer, as well as character types whose representation is signed.
*/
class SemSignedIntegerType extends SemIntegerType {
SemSignedIntegerType() { signed = true }
final override string toString() { result = "int" + byteSize.toString() }
pragma[noinline]
final override int getByteSize() { result = byteSize }
}
/**
* An unsigned two's-complement integer. Also used to represent enums whose underlying type is an
* unsigned integer, as well as character types whose representation is unsigned.
*/
class SemUnsignedIntegerType extends SemIntegerType {
SemUnsignedIntegerType() { signed = false }
final override string toString() { result = "uint" + byteSize.toString() }
pragma[noinline]
final override int getByteSize() { result = byteSize }
}
/**
* A floating-point type.
*/
class SemFloatingPointType extends SemNumericType, TSemFloatingPointType {
final override string toString() { result = "float" + byteSize.toString() }
pragma[noinline]
final override int getByteSize() { result = byteSize }
}
/**
* An address type, representing the memory address of data. Used to represent pointers, references,
* and lvalues, include those that are garbage collected.
*
* The address of a function is represented by the separate `SemFunctionAddressType`.
*/
class SemAddressType extends SemSizedType, TSemAddressType {
final override string toString() { result = "addr" + byteSize.toString() }
pragma[noinline]
final override int getByteSize() { result = byteSize }
}
/**
* An address type, representing the memory address of code. Used to represent function pointers,
* function references, and the target of a direct function call.
*/
class SemFunctionAddressType extends SemSizedType, TSemFunctionAddressType {
final override string toString() { result = "func" + byteSize.toString() }
pragma[noinline]
final override int getByteSize() { result = byteSize }
}
/**
* A type with known size that does not fit any of the other kinds of type. Used to represent
* classes, structs, unions, fixed-size arrays, pointers-to-member, and more.
*/
class SemOpaqueType extends SemSizedType, TSemOpaqueType {
Specific::OpaqueTypeTag tag;
SemOpaqueType() { this = TSemOpaqueType(byteSize, tag) }
final override string toString() {
result = "opaque" + byteSize.toString() + "{" + tag.toString() + "}"
}
final override string getIdentityString() {
result = "opaque" + byteSize.toString() + "{" + Specific::getOpaqueTagIdentityString(tag) + "}"
}
/**
* Gets the "tag" that differentiates this type from other incompatible opaque types that have the
* same size.
*/
final Specific::OpaqueTypeTag getTag() { result = tag }
pragma[noinline]
final override int getByteSize() { result = byteSize }
}
cached
SemType getSemanticType(Specific::Type type) {
exists(int byteSize |
Specific::booleanType(type, byteSize) and result = TSemBooleanType(byteSize)
or
exists(boolean signed |
Specific::integerType(type, byteSize, signed) and
result = TSemIntegerType(byteSize, signed)
)
or
Specific::floatingPointType(type, byteSize) and result = TSemFloatingPointType(byteSize)
or
Specific::addressType(type, byteSize) and result = TSemAddressType(byteSize)
or
Specific::functionAddressType(type, byteSize) and result = TSemFunctionAddressType(byteSize)
or
exists(Specific::OpaqueTypeTag tag |
Specific::opaqueType(type, byteSize, tag) and result = TSemOpaqueType(byteSize, tag)
)
)
or
Specific::errorType(type) and result = TSemErrorType()
or
Specific::unknownType(type) and result = TSemUnknownType()
}
/**
* Holds if the conversion from `fromType` to `toType` can never overflow or underflow.
*/
predicate conversionCannotOverflow(SemNumericType fromType, SemNumericType toType) {
// Identity cast
fromType = toType
or
// Treat any cast to an FP type as safe. It can lose precision, but not overflow.
toType instanceof SemFloatingPointType and fromType = any(SemNumericType n)
or
exists(SemIntegerType fromInteger, SemIntegerType toInteger, int fromSize, int toSize |
fromInteger = fromType and
toInteger = toType and
fromSize = fromInteger.getByteSize() and
toSize = toInteger.getByteSize()
|
// Conversion to a larger type. Safe unless converting signed -> unsigned.
fromSize < toSize and
(
toInteger.isSigned()
or
not fromInteger.isSigned()
)
)
}
/**
* INTERNAL: Do not use.
* Query predicates used to check invariants that should hold for all `SemType` objects.
*/
module SemTypeConsistency {
/**
* Holds if the type has no result for `getSemanticType()`.
*/
query predicate missingSemType(Specific::Type type, string message) {
not exists(getSemanticType(type)) and
message = "`Type` does not have a corresponding `SemType`."
}
/**
* Holds if the type has more than one result for `getSemanticType()`.
*/
query predicate multipleSemTypes(Specific::Type type, string message) {
strictcount(getSemanticType(type)) > 1 and
message =
"`Type` " + type + " has multiple `SemType`s: " +
concat(getSemanticType(type).toString(), ", ")
}
}

View File

@@ -1,43 +0,0 @@
/**
* C++-specific implementation of the semantic type system.
*/
private import semmle.code.cpp.ir.IR as IR
private import cpp as Cpp
private import semmle.code.cpp.ir.internal.IRCppLanguage as Language
class Type = IR::IRType;
class OpaqueTypeTag = Language::OpaqueTypeTag;
predicate voidType(Type type) { type instanceof IR::IRVoidType }
predicate errorType(Type type) { type instanceof IR::IRErrorType }
predicate unknownType(Type type) { type instanceof IR::IRUnknownType }
predicate booleanType(Type type, int byteSize) { byteSize = type.(IR::IRBooleanType).getByteSize() }
predicate integerType(Type type, int byteSize, boolean signed) {
byteSize = type.(IR::IRSignedIntegerType).getByteSize() and signed = true
or
byteSize = type.(IR::IRUnsignedIntegerType).getByteSize() and signed = false
}
predicate floatingPointType(Type type, int byteSize) {
byteSize = type.(IR::IRFloatingPointType).getByteSize()
}
predicate addressType(Type type, int byteSize) { byteSize = type.(IR::IRAddressType).getByteSize() }
predicate functionAddressType(Type type, int byteSize) {
byteSize = type.(IR::IRFunctionAddressType).getByteSize()
}
predicate opaqueType(Type type, int byteSize, OpaqueTypeTag tag) {
exists(IR::IROpaqueType opaque | opaque = type |
byteSize = opaque.getByteSize() and tag = opaque.getTag()
)
}
predicate getOpaqueTagIdentityString = Language::getOpaqueTagIdentityString/1;

View File

@@ -1,31 +0,0 @@
/**
* Simple constant analysis using the Semantic interface.
*/
private import experimental.semmle.code.cpp.semantic.Semantic
private import ConstantAnalysisSpecific as Specific
/** An expression that always has the same integer value. */
pragma[nomagic]
private predicate constantIntegerExpr(SemExpr e, int val) {
// An integer literal
e.(SemIntegerLiteralExpr).getIntValue() = val
or
// Copy of another constant
exists(SemSsaExplicitUpdate v, SemExpr src |
e = v.getAUse() and
src = v.getSourceExpr() and
constantIntegerExpr(src, val)
)
or
// Language-specific enhancements
val = Specific::getIntConstantValue(e)
}
/** An expression that always has the same integer value. */
class SemConstantIntegerExpr extends SemExpr {
SemConstantIntegerExpr() { constantIntegerExpr(this, _) }
/** Gets the integer value of this expression. */
int getIntValue() { constantIntegerExpr(this, result) }
}

View File

@@ -1,10 +0,0 @@
/**
* C++-specific implementation of constant analysis.
*/
private import experimental.semmle.code.cpp.semantic.Semantic
/**
* Gets the constant integer value of the specified expression, if any.
*/
int getIntConstantValue(SemExpr expr) { none() }

View File

@@ -1,311 +0,0 @@
/**
* Provides inferences of the form: `e` equals `b + v` modulo `m` where `e` is
* an expression, `b` is a `Bound` (typically zero or the value of an SSA
* variable), and `v` is an integer in the range `[0 .. m-1]`.
*/
private import ModulusAnalysisSpecific::Private
private import experimental.semmle.code.cpp.semantic.Semantic
private import ConstantAnalysis
private import RangeUtils
/**
* Holds if `e + delta` equals `v` at `pos`.
*/
private predicate valueFlowStepSsa(SemSsaVariable v, SemSsaReadPosition pos, SemExpr e, int delta) {
semSsaUpdateStep(v, e, delta) and pos.hasReadOfVar(v)
or
exists(SemGuard guard, boolean testIsTrue |
pos.hasReadOfVar(v) and
guard = semEqFlowCond(v, e, delta, true, testIsTrue) and
semGuardDirectlyControlsSsaRead(guard, pos, testIsTrue)
)
}
/**
* Holds if `add` is the addition of `larg` and `rarg`, neither of which are
* `ConstantIntegerExpr`s.
*/
private predicate nonConstAddition(SemExpr add, SemExpr larg, SemExpr rarg) {
exists(SemAddExpr a | a = add |
larg = a.getLeftOperand() and
rarg = a.getRightOperand()
) and
not larg instanceof SemConstantIntegerExpr and
not rarg instanceof SemConstantIntegerExpr
}
/**
* Holds if `sub` is the subtraction of `larg` and `rarg`, where `rarg` is not
* a `ConstantIntegerExpr`.
*/
private predicate nonConstSubtraction(SemExpr sub, SemExpr larg, SemExpr rarg) {
exists(SemSubExpr s | s = sub |
larg = s.getLeftOperand() and
rarg = s.getRightOperand()
) and
not rarg instanceof SemConstantIntegerExpr
}
/** Gets an expression that is the remainder modulo `mod` of `arg`. */
private SemExpr modExpr(SemExpr arg, int mod) {
exists(SemRemExpr rem |
result = rem and
arg = rem.getLeftOperand() and
rem.getRightOperand().(SemConstantIntegerExpr).getIntValue() = mod and
mod >= 2
)
or
exists(SemConstantIntegerExpr c |
mod = 2.pow([1 .. 30]) and
c.getIntValue() = mod - 1 and
result.(SemBitAndExpr).hasOperands(arg, c)
)
}
/**
* Gets a guard that tests whether `v` is congruent with `val` modulo `mod` on
* its `testIsTrue` branch.
*/
private SemGuard moduloCheck(SemSsaVariable v, int val, int mod, boolean testIsTrue) {
exists(SemExpr rem, SemConstantIntegerExpr c, int r, boolean polarity |
result.isEquality(rem, c, polarity) and
c.getIntValue() = r and
rem = modExpr(v.getAUse(), mod) and
(
testIsTrue = polarity and val = r
or
testIsTrue = polarity.booleanNot() and
mod = 2 and
val = 1 - r and
(r = 0 or r = 1)
)
)
}
/**
* Holds if a guard ensures that `v` at `pos` is congruent with `val` modulo `mod`.
*/
private predicate moduloGuardedRead(SemSsaVariable v, SemSsaReadPosition pos, int val, int mod) {
exists(SemGuard guard, boolean testIsTrue |
pos.hasReadOfVar(v) and
guard = moduloCheck(v, val, mod, testIsTrue) and
semGuardControlsSsaRead(guard, pos, testIsTrue)
)
}
/** Holds if `factor` is a power of 2 that divides `mask`. */
bindingset[mask]
private predicate andmaskFactor(int mask, int factor) {
mask % factor = 0 and
factor = 2.pow([1 .. 30])
}
/** Holds if `e` is evenly divisible by `factor`. */
private predicate evenlyDivisibleExpr(SemExpr e, int factor) {
exists(SemConstantIntegerExpr c, int k | k = c.getIntValue() |
e.(SemMulExpr).getAnOperand() = c and factor = k.abs() and factor >= 2
or
e.(SemShiftLeftExpr).getRightOperand() = c and factor = 2.pow(k) and k > 0
or
e.(SemBitAndExpr).getAnOperand() = c and factor = max(int f | andmaskFactor(k, f))
)
}
/**
* Holds if `rix` is the number of input edges to `phi`.
*/
private predicate maxPhiInputRank(SemSsaPhiNode phi, int rix) {
rix = max(int r | rankedPhiInput(phi, _, _, r))
}
/**
* Gets the remainder of `val` modulo `mod`.
*
* For `mod = 0` the result equals `val` and for `mod > 1` the result is within
* the range `[0 .. mod-1]`.
*/
bindingset[val, mod]
private int remainder(int val, int mod) {
mod = 0 and result = val
or
mod > 1 and result = ((val % mod) + mod) % mod
}
/**
* Holds if `inp` is an input to `phi` and equals `phi` modulo `mod` along `edge`.
*/
private predicate phiSelfModulus(
SemSsaPhiNode phi, SemSsaVariable inp, SemSsaReadPositionPhiInputEdge edge, int mod
) {
exists(SemSsaBound phibound, int v, int m |
edge.phiInput(phi, inp) and
phibound.getAVariable() = phi and
ssaModulus(inp, edge, phibound, v, m) and
mod = m.gcd(v) and
mod != 1
)
}
/**
* Holds if `b + val` modulo `mod` is a candidate congruence class for `phi`.
*/
private predicate phiModulusInit(SemSsaPhiNode phi, SemBound b, int val, int mod) {
exists(SemSsaVariable inp, SemSsaReadPositionPhiInputEdge edge |
edge.phiInput(phi, inp) and
ssaModulus(inp, edge, b, val, mod)
)
}
/**
* Holds if all inputs to `phi` numbered `1` to `rix` are equal to `b + val` modulo `mod`.
*/
pragma[nomagic]
private predicate phiModulusRankStep(SemSsaPhiNode phi, SemBound b, int val, int mod, int rix) {
rix = 0 and
phiModulusInit(phi, b, val, mod)
or
exists(SemSsaVariable inp, SemSsaReadPositionPhiInputEdge edge, int v1, int m1 |
mod != 1 and
val = remainder(v1, mod)
|
exists(int v2, int m2 |
rankedPhiInput(pragma[only_bind_out](phi), inp, edge, rix) and
phiModulusRankStep(phi, b, v1, m1, rix - 1) and
ssaModulus(inp, edge, b, v2, m2) and
mod = m1.gcd(m2).gcd(v1 - v2)
)
or
exists(int m2 |
rankedPhiInput(phi, inp, edge, rix) and
phiModulusRankStep(phi, b, v1, m1, rix - 1) and
phiSelfModulus(phi, inp, edge, m2) and
mod = m1.gcd(m2)
)
)
}
/**
* Holds if `phi` is equal to `b + val` modulo `mod`.
*/
private predicate phiModulus(SemSsaPhiNode phi, SemBound b, int val, int mod) {
exists(int r |
maxPhiInputRank(phi, r) and
phiModulusRankStep(phi, b, val, mod, r)
)
}
/**
* Holds if `v` at `pos` is equal to `b + val` modulo `mod`.
*/
private predicate ssaModulus(SemSsaVariable v, SemSsaReadPosition pos, SemBound b, int val, int mod) {
phiModulus(v, b, val, mod) and pos.hasReadOfVar(v)
or
b.(SemSsaBound).getAVariable() = v and pos.hasReadOfVar(v) and val = 0 and mod = 0
or
exists(SemExpr e, int val0, int delta |
semExprModulus(e, b, val0, mod) and
valueFlowStepSsa(v, pos, e, delta) and
val = remainder(val0 + delta, mod)
)
or
moduloGuardedRead(v, pos, val, mod) and b instanceof SemZeroBound
}
/**
* Holds if `e` is equal to `b + val` modulo `mod`.
*
* There are two cases for the modulus:
* - `mod = 0`: The equality `e = b + val` is an ordinary equality.
* - `mod > 1`: `val` lies within the range `[0 .. mod-1]`.
*/
cached
predicate semExprModulus(SemExpr e, SemBound b, int val, int mod) {
not ignoreExprModulus(e) and
(
e = b.getExpr(val) and mod = 0
or
evenlyDivisibleExpr(e, mod) and
val = 0 and
b instanceof SemZeroBound
or
exists(SemSsaVariable v, SemSsaReadPositionBlock bb |
ssaModulus(v, bb, b, val, mod) and
e = v.getAUse() and
bb.getAnExpr() = e
)
or
exists(SemExpr mid, int val0, int delta |
semExprModulus(mid, b, val0, mod) and
semValueFlowStep(e, mid, delta) and
val = remainder(val0 + delta, mod)
)
or
exists(SemConditionalExpr cond, int v1, int v2, int m1, int m2 |
cond = e and
condExprBranchModulus(cond, true, b, v1, m1) and
condExprBranchModulus(cond, false, b, v2, m2) and
mod = m1.gcd(m2).gcd(v1 - v2) and
mod != 1 and
val = remainder(v1, mod)
)
or
exists(SemBound b1, SemBound b2, int v1, int v2, int m1, int m2 |
addModulus(e, true, b1, v1, m1) and
addModulus(e, false, b2, v2, m2) and
mod = m1.gcd(m2) and
mod != 1 and
val = remainder(v1 + v2, mod)
|
b = b1 and b2 instanceof SemZeroBound
or
b = b2 and b1 instanceof SemZeroBound
)
or
exists(int v1, int v2, int m1, int m2 |
subModulus(e, true, b, v1, m1) and
subModulus(e, false, any(SemZeroBound zb), v2, m2) and
mod = m1.gcd(m2) and
mod != 1 and
val = remainder(v1 - v2, mod)
)
)
}
private predicate condExprBranchModulus(
SemConditionalExpr cond, boolean branch, SemBound b, int val, int mod
) {
semExprModulus(cond.getBranchExpr(branch), b, val, mod)
}
private predicate addModulus(SemExpr add, boolean isLeft, SemBound b, int val, int mod) {
exists(SemExpr larg, SemExpr rarg | nonConstAddition(add, larg, rarg) |
semExprModulus(larg, b, val, mod) and isLeft = true
or
semExprModulus(rarg, b, val, mod) and isLeft = false
)
}
private predicate subModulus(SemExpr sub, boolean isLeft, SemBound b, int val, int mod) {
exists(SemExpr larg, SemExpr rarg | nonConstSubtraction(sub, larg, rarg) |
semExprModulus(larg, b, val, mod) and isLeft = true
or
semExprModulus(rarg, b, val, mod) and isLeft = false
)
}
/**
* Holds if `inp` is an input to `phi` along `edge` and this input has index `r`
* in an arbitrary 1-based numbering of the input edges to `phi`.
*/
private predicate rankedPhiInput(
SemSsaPhiNode phi, SemSsaVariable inp, SemSsaReadPositionPhiInputEdge edge, int r
) {
edge.phiInput(phi, inp) and
edge =
rank[r](SemSsaReadPositionPhiInputEdge e |
e.phiInput(phi, _)
|
e order by e.getOrigBlock().getUniqueId()
)
}

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