Merge branch 'main' into mbg/go/semver-type

Author: mbg

.gitattributes

@@ -83,7 +83,7 @@
 /csharp/paket.main_extension.bzl linguist-generated=true
 
 # ripunzip tool
-/misc/bazel/internal/ripunzip/ripunzip-* filter=lfs diff=lfs merge=lfs -text
+/misc/ripunzip/ripunzip-* filter=lfs diff=lfs merge=lfs -text
 
 # swift prebuilt resources
 /swift/third_party/resource-dir/*.zip filter=lfs diff=lfs merge=lfs -text

MODULE.bazel

@@ -118,19 +118,19 @@ lfs_files = use_repo_rule("//misc/bazel:lfs.bzl", "lfs_files")
 
 lfs_files(
     name = "ripunzip-linux",
-    srcs = ["//misc/bazel/internal/ripunzip:ripunzip-linux"],
+    srcs = ["//misc/ripunzip:ripunzip-linux"],
     executable = True,
 )
 
 lfs_files(
     name = "ripunzip-windows",
-    srcs = ["//misc/bazel/internal/ripunzip:ripunzip-windows.exe"],
+    srcs = ["//misc/ripunzip:ripunzip-windows.exe"],
     executable = True,
 )
 
 lfs_files(
     name = "ripunzip-macos",
-    srcs = ["//misc/bazel/internal/ripunzip:ripunzip-macos"],
+    srcs = ["//misc/ripunzip:ripunzip-macos"],
     executable = True,
 )
 

cpp/ql/lib/change-notes/2024-06-14-boost-asio.md

@@ -0,0 +1,4 @@
+---
+category: minorAnalysis
+---
+* A partial model for the `Boost.Asio` network library has been added. This includes sources, sinks and summaries for certain functions in `Boost.Asio`, such as `read_until` and `write`.

cpp/ql/lib/change-notes/2024-06-14-models-as-data-yml-extensions.md

@@ -0,0 +1,4 @@
+---
+category: feature
+---
+* Data models can now be added with data extensions. In this way source, sink and summary models can be added in extension `.model.yml` files, rather than by writing classes in QL code. New models should be added in the `lib/ext` folder.

cpp/ql/lib/ext/Boost.Asio.model.yml

@@ -0,0 +1,26 @@
+extensions:
+  # partial model of the Boost::Asio network library
+extensions:
+  - addsTo:
+      pack: codeql/cpp-all
+      extensible: sourceModel
+    data: # namespace, type, subtypes, name, signature, ext, output, kind, provenance
+      - ["boost::asio", "", False, "read", "", "", "Argument[*1]", "remote", "manual"]
+      - ["boost::asio", "", False, "read_at", "", "", "Argument[*2]", "remote", "manual"]
+      - ["boost::asio", "", False, "read_until", "", "", "Argument[*1]", "remote", "manual"]
+      - ["boost::asio", "", False, "async_read", "", "", "Argument[*1]", "remote", "manual"]
+      - ["boost::asio", "", False, "async_read_at", "", "", "Argument[*2]", "remote", "manual"]
+      - ["boost::asio", "", False, "async_read_until", "", "", "Argument[*1]", "remote", "manual"]
+  - addsTo:
+      pack: codeql/cpp-all
+      extensible: sinkModel
+    data: # namespace, type, subtypes, name, signature, ext, input, kind, provenance
+      - ["boost::asio", "", False, "write", "", "", "Argument[*1]", "remote-sink", "manual"]
+      - ["boost::asio", "", False, "write_at", "", "", "Argument[*2]", "remote-sink", "manual"]
+      - ["boost::asio", "", False, "async_write", "", "", "Argument[*1]", "remote-sink", "manual"]
+      - ["boost::asio", "", False, "async_write_at", "", "", "Argument[*2]", "remote-sink", "manual"]
+  - addsTo:
+      pack: codeql/cpp-all
+      extensible: summaryModel
+    data: # namespace, type, subtypes, name, signature, ext, input, output, kind, provenance
+      - ["boost::asio", "", False, "buffer", "", "", "Argument[*0]", "ReturnValue", "taint", "manual"]

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

@@ -0,0 +1,15 @@
+extensions:
+  # Make sure that the extensible model predicates have at least one definition
+  # to avoid errors about undefined extensionals.
+  - addsTo:
+      pack: codeql/cpp-all
+      extensible: sourceModel
+    data: []
+  - addsTo:
+      pack: codeql/cpp-all
+      extensible: sinkModel
+    data: []
+  - addsTo:
+      pack: codeql/cpp-all
+      extensible: summaryModel
+    data: []

cpp/ql/lib/qlpack.yml

@@ -14,4 +14,6 @@ dependencies:
   codeql/tutorial: ${workspace}
   codeql/util: ${workspace}
   codeql/xml: ${workspace}
+dataExtensions:
+  - ext/*.model.yml
 warnOnImplicitThis: true

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

@@ -410,6 +410,10 @@ class LocalVariable extends LocalScopeVariable, @localvariable {
     or
     orphaned_variables(underlyingElement(this), unresolveElement(result))
   }
+
+  override predicate isStatic() {
+    super.isStatic() or orphaned_variables(underlyingElement(this), _)
+  }
 }
 
 /**

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

@@ -78,6 +78,7 @@ private import internal.FlowSummaryImpl
 private import internal.FlowSummaryImpl::Public
 private import internal.FlowSummaryImpl::Private
 private import internal.FlowSummaryImpl::Private::External
+private import internal.ExternalFlowExtensions as Extensions
 private import codeql.mad.ModelValidation as SharedModelVal
 private import codeql.util.Unit
 
@@ -138,6 +139,9 @@ predicate sourceModel(
     row.splitAt(";", 7) = kind
   ) and
   provenance = "manual"
+  or
+  Extensions::sourceModel(namespace, type, subtypes, name, signature, ext, output, kind, provenance,
+    _)
 }
 
 /** Holds if a sink model exists for the given parameters. */
@@ -158,6 +162,8 @@ predicate sinkModel(
     row.splitAt(";", 7) = kind
   ) and
   provenance = "manual"
+  or
+  Extensions::sinkModel(namespace, type, subtypes, name, signature, ext, input, kind, provenance, _)
 }
 
 /** Holds if a summary model exists for the given parameters. */
@@ -179,6 +185,9 @@ predicate summaryModel(
     row.splitAt(";", 8) = kind
   ) and
   provenance = "manual"
+  or
+  Extensions::summaryModel(namespace, type, subtypes, name, signature, ext, input, output, kind,
+    provenance, _)
 }
 
 private predicate relevantNamespace(string namespace) {
@@ -203,8 +212,10 @@ private predicate canonicalNamespaceLink(string namespace, string subns) {
 }
 
 /**
- * Holds if CSV framework coverage of `namespace` is `n` api endpoints of the
- * kind `(kind, part)`.
+ * Holds if MaD framework coverage of `namespace` is `n` api endpoints of the
+ * kind `(kind, part)`, and `namespaces` is the number of subnamespaces of
+ * `namespace` which have MaD framework coverage (including `namespace`
+ * itself).
  */
 predicate modelCoverage(string namespace, int namespaces, string kind, string part, int n) {
   namespaces = strictcount(string subns | canonicalNamespaceLink(namespace, subns)) and
@@ -321,10 +332,10 @@ module CsvValidation {
       or
       summaryModel(namespace, type, _, name, signature, ext, _, _, _, _) and pred = "summary"
     |
-      not namespace.regexpMatch("[a-zA-Z0-9_\\.]+") and
+      not namespace.regexpMatch("[a-zA-Z0-9_\\.:]*") and
       result = "Dubious namespace \"" + namespace + "\" in " + pred + " model."
       or
-      not type.regexpMatch("[a-zA-Z0-9_<>,\\+]+") and
+      not type.regexpMatch("[a-zA-Z0-9_<>,\\+]*") and
       result = "Dubious type \"" + type + "\" in " + pred + " model."
       or
       not name.regexpMatch("[a-zA-Z0-9_<>,]*") and

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

@@ -0,0 +1,27 @@
+/**
+ * This module provides extensible predicates for defining MaD models.
+ */
+
+/**
+ * Holds if an external source model exists for the given parameters.
+ */
+extensible predicate sourceModel(
+  string namespace, string type, boolean subtypes, string name, string signature, string ext,
+  string output, string kind, string provenance, QlBuiltins::ExtensionId madId
+);
+
+/**
+ * Holds if an external sink model exists for the given parameters.
+ */
+extensible predicate sinkModel(
+  string namespace, string type, boolean subtypes, string name, string signature, string ext,
+  string input, string kind, string provenance, QlBuiltins::ExtensionId madId
+);
+
+/**
+ * Holds if an external summary model exists for the given parameters.
+ */
+extensible predicate summaryModel(
+  string namespace, string type, boolean subtypes, string name, string signature, string ext,
+  string input, string output, string kind, string provenance, QlBuiltins::ExtensionId madId
+);