Commit Graph

10804 Commits

Author SHA1 Message Date
Mathias Vorreiter Pedersen
84a61d1e02 C++: No need for 'matches'. 2023-03-09 15:36:26 +00:00
Mathias Vorreiter Pedersen
59402eb754 Merge pull request #12462 from MathiasVP/disable-std-order-in-fwd-flow-stage-1
DataFlow: Disable standard order in `Stage1::fwdFlow`
2023-03-09 15:30:05 +00:00
Mathias Vorreiter Pedersen
2931e5dea8 C++: Reduce duplication by blocking flow into sources (since we'll already be considering flow starting at those sources) and out of sinks (since we'll already be alerting on this sink if it's relevant). 2023-03-09 14:59:13 +00:00
Mathias Vorreiter Pedersen
03ba7ea851 C++: Move the weird global property 'not sqlite_encryption_used()' from the sink definition to the source definition. The dataflow library starts tracking flow from the sources, so it's better to to rule out the entire database in the source definition than in the sink definition. 2023-03-09 14:59:13 +00:00
Mathias Vorreiter Pedersen
7819a7d2bc C++: Severely restrict the set of sinks in 'cpp/cleartext-storage-database'. This reduces the number of sinks considered on the 'sysown/proxysql' from > 62000 sinks to ~1000 sinks. 2023-03-09 14:59:13 +00:00
Asger F
6e744093e2 Merge pull request #12398 from github/post-release-prep/codeql-cli-2.12.4
Post-release preparation for codeql-cli-2.12.4
2023-03-09 15:38:21 +01:00
Jeroen Ketema
de97ae38dc C++: C++: Use getAUse in getIRRepresentationOfIndirectOperand 2023-03-09 15:15:00 +01:00
Mathias Vorreiter Pedersen
1f77f77153 DataFlow: Sync identical files. 2023-03-09 10:41:15 +00:00
Mathias Vorreiter Pedersen
c7b41ca470 C++: Disable standard order for 'fwdFlow' in stage 1 of dataflow. 2023-03-09 10:41:06 +00:00
Jeroen Ketema
e5ce27fbf9 C++: Fix a number of instances where a dataflow node did not have a location 2023-03-09 10:32:30 +01:00
Jeroen Ketema
31fa230c0d C++: Remove toStringImpl predicate that is overridden in every subclass 2023-03-09 10:29:17 +01:00
Jeroen Ketema
4ee13a3234 C++: Remove getLocationImpl predicate that is always overridden
Also remove the predicate referred to in ints implementation, as it is no
longer used.
2023-03-09 10:27:15 +01:00
Jeroen Ketema
791f5913d2 C++: Fix multiple getLocation consistency error in use-use dataflow 2023-03-09 10:25:02 +01:00
Mathias Vorreiter Pedersen
f19f7967c2 C++: Fix join order.
Before (I stopped midway):

```
(72s) Tuple counts for _#Class#bacd9b46::Class::getADerivedClass#0#dispredPlus#ff_#Declaration#4bfb53be::AccessHolder::getE__#antijoin_rhs#1/3@fb0627h8 after 1m4s:
  ...

  20000     ~0%       {5} r28 = r26 UNION r27
  224367484 ~7%       {9} r29 = JOIN r28 WITH Class#bacd9b46::Class::accessOfBaseMember#2#dispred#ffff_1023#join_rhs ON FIRST 1 OUTPUT Rhs.3, "protected", Lhs.1 'arg0', Lhs.2 'arg1', Lhs.3 'arg2', Lhs.0, Lhs.4, Rhs.1, Rhs.2

  111914129 ~0%       {7} r30 = JOIN r29 WITH specifiers ON FIRST 2 OUTPUT Lhs.6, Lhs.2 'arg0', Lhs.3 'arg1', Lhs.4 'arg2', Lhs.5, Lhs.7, Lhs.8

  123503367 ~0%       {8} r31 = JOIN r30 WITH Declaration#4bfb53be::DirectAccessHolder::isFriendOfOrEqualTo#1#dispred#ff ON FIRST 1 OUTPUT Lhs.3 'arg2', Rhs.1, Lhs.1 'arg0', Lhs.2 'arg1', Lhs.4, Lhs.0, Lhs.5, Lhs.6
  331748250 ~0%       {10} r32 = JOIN r31 WITH Class#bacd9b46::Class::accessOfBaseMember#2#dispred#ffff ON FIRST 2 OUTPUT Lhs.2 'arg0', Lhs.3 'arg1', Lhs.0 'arg2', Lhs.4, Lhs.5, Lhs.6, Lhs.7, Lhs.1, Rhs.2, Rhs.3
  331748250 ~0%       {10} r33 = SELECT r32 ON In.8 = In.9
  331748250 ~2%       {9} r34 = SCAN r33 OUTPUT In.7, In.5, In.8, In.0 'arg0', In.1 'arg1', In.2 'arg2', In.3, In.4, In.6
  38000     ~4%       {10} r35 = JOIN r34 WITH Class#bacd9b46::Class::accessOfBaseMember#2#dispred#ffff ON FIRST 3 OUTPUT Rhs.3, Lhs.3 'arg0', Lhs.4 'arg1', Lhs.5 'arg2', Lhs.6, Lhs.7, Lhs.1, Lhs.8, Lhs.0, Lhs.2
  37500     ~0%       {11} r36 = JOIN r35 WITH specifiers ON FIRST 1 OUTPUT Lhs.1 'arg0', Lhs.2 'arg1', Lhs.3 'arg2', Lhs.4, Lhs.5, Lhs.6, Lhs.7, Lhs.8, Lhs.9, Lhs.0, Rhs.1
  28973     ~0%       {11} r37 = SELECT r36 ON In.10 >= "protected"
  28973     ~98%      {6} r38 = SCAN r37 OUTPUT In.8, "public", In.0 'arg0', In.1 'arg1', In.2 'arg2', In.6

  111913629 ~6%       {7} r39 = JOIN r29 WITH specifiers ON FIRST 2 OUTPUT Lhs.6, Lhs.4 'arg2', Lhs.2 'arg0', Lhs.3 'arg1', Lhs.5, Lhs.7, Lhs.8
  110582830 ~0%       {8} r40 = JOIN r39 WITH Declaration#4bfb53be::DirectAccessHolder::isFriendOfOrEqualTo#1#dispred#ff ON FIRST 2 OUTPUT Lhs.1 'arg2', Lhs.5, Lhs.6, Lhs.2 'arg0', Lhs.3 'arg1', Lhs.1 'arg2', Lhs.4, Lhs.0

  123503367 ~0%       {8} r41 = JOIN r30 WITH Declaration#4bfb53be::DirectAccessHolder::isFriendOfOrEqualTo#1#dispred#ff ON FIRST 1 OUTPUT Rhs.1, Lhs.3 'arg2', Lhs.1 'arg0', Lhs.2 'arg1', Lhs.4, Lhs.0, Lhs.5, Lhs.6
  0         ~0%       {8} r42 = JOIN r41 WITH #Class#bacd9b46::Class::getADerivedClass#0#dispredPlus#ff ON FIRST 2 OUTPUT Lhs.0, Lhs.6, Lhs.7, Lhs.2 'arg0', Lhs.3 'arg1', Lhs.1 'arg2', Lhs.4, Lhs.5

  110582830 ~0%       {8} r43 = r40 UNION r42
  15000     ~6%       {8} r44 = JOIN r43 WITH Class#bacd9b46::Class::accessOfBaseMember#2#dispred#ffff ON FIRST 3 OUTPUT Lhs.5 'arg2', Lhs.1, Lhs.3 'arg0', Lhs.4 'arg1', Lhs.6, Lhs.7, Lhs.2, Lhs.0
  ...
```

After:

```
Tuple counts for _#Class#bacd9b46::Class::getADerivedClass#0#dispredPlus#ff_#Declaration#4bfb53be::AccessHolder::getE__#antijoin_rhs#1/3@997a3ai9 after 744ms:
  ...

  78600   ~8%       {6} r29 = r26 UNION r28
  437816  ~0%       {9} r30 = JOIN r29 WITH Class#bacd9b46::Class::accessOfBaseMember#2#dispred#ffff ON FIRST 1 OUTPUT Lhs.1 'arg0', Lhs.2 'arg1', Lhs.0 'arg2', Lhs.3, Lhs.4, Lhs.5, Rhs.1, Rhs.2, Rhs.3
  430928  ~0%       {9} r31 = SELECT r30 ON In.7 = In.8
  430928  ~0%       {7} r32 = SCAN r31 OUTPUT In.5, In.6, In.0 'arg0', In.1 'arg1', In.2 'arg2', In.3, In.7
  1096333 ~0%       {7} r33 = JOIN r32 WITH Class#bacd9b46::Class::accessOfBaseMember#2#dispred#ffff ON FIRST 2 OUTPUT Lhs.1, Lhs.5, Rhs.2, Lhs.2 'arg0', Lhs.3 'arg1', Lhs.4 'arg2', Lhs.6
  777970  ~0%       {8} r34 = JOIN r33 WITH Class#bacd9b46::Class::accessOfBaseMember#2#dispred#ffff ON FIRST 3 OUTPUT Lhs.0, Lhs.1, Lhs.2, Rhs.3, Lhs.3 'arg0', Lhs.4 'arg1', Lhs.5 'arg2', Lhs.6

  334217  ~0%       {6} r35 = JOIN r14 WITH Declaration#4bfb53be::DirectAccessHolder::isFriendOfOrEqualTo#1#dispred#ff ON FIRST 1 OUTPUT Lhs.3 'arg2', Rhs.1, Lhs.1 'arg0', Lhs.2 'arg1', Lhs.4, Lhs.0
  235623  ~0%       {8} r36 = JOIN r35 WITH Class#bacd9b46::Class::accessOfBaseMember#2#dispred#ffff ON FIRST 2 OUTPUT Lhs.2 'arg0', Lhs.3 'arg1', Lhs.0 'arg2', Lhs.4, Lhs.5, Lhs.1, Rhs.2, Rhs.3
  235623  ~0%       {8} r37 = SELECT r36 ON In.6 = In.7
  235623  ~0%       {7} r38 = SCAN r37 OUTPUT In.5, In.6, In.0 'arg0', In.1 'arg1', In.2 'arg2', In.3, In.4
  437303  ~0%       {9} r39 = JOIN r38 WITH Class#bacd9b46::Class::accessOfBaseMember#2#dispred#ffff_0213#join_rhs ON FIRST 2 OUTPUT Rhs.3, Lhs.2 'arg0', Lhs.3 'arg1', Lhs.4 'arg2', Lhs.5, Lhs.6, Lhs.0, Lhs.1, Rhs.2
  437303  ~4%       {10} r40 = JOIN r39 WITH specifiers ON FIRST 1 OUTPUT Lhs.1 'arg0', Lhs.2 'arg1', Lhs.3 'arg2', Lhs.4, Lhs.5, Lhs.6, Lhs.7, Lhs.8, Lhs.0, Rhs.1
  352102  ~1%       {10} r41 = SELECT r40 ON In.9 >= "protected"
  352102  ~0%       {6} r42 = SCAN r41 OUTPUT In.7, In.3, In.0 'arg0', In.1 'arg1', In.2 'arg2', In.6
  775332  ~0%       {8} r43 = JOIN r42 WITH Class#bacd9b46::Class::accessOfBaseMember#2#dispred#ffff ON FIRST 2 OUTPUT Lhs.0, Lhs.1, Rhs.2, Rhs.3, Lhs.2 'arg0', Lhs.3 'arg1', Lhs.4 'arg2', Lhs.5

  1553302 ~51%      {8} r44 = r34 UNION r43
  1553302 ~152%     {7} r45 = JOIN r44 WITH Class#bacd9b46::Class::accessOfBaseMember#2#dispred#ffff ON FIRST 4 OUTPUT Lhs.7, "public", Lhs.4 'arg0', Lhs.5 'arg1', Lhs.6 'arg2', Lhs.2, Lhs.3
  ...
```
2023-03-09 09:23:56 +00:00
Jeroen Ketema
55da3257bf C++: Fix multiple toString consistency error in use-use dataflow 2023-03-09 10:08:25 +01:00
Jeroen Ketema
a2248cb5a0 C++: Use getAUse in getIRRepresentationOfOperand 2023-03-09 09:45:54 +01:00
Mathias Vorreiter Pedersen
540ce1f0db Contrary to what the QLDoc says, this predicate was way too large to be
evaluated on the 'quick-lint/quick-lint-js' project.

Before:
```
Most expensive predicates for completed query RuleOfTwo.ql:
        time  | evals |   max @ iter | predicate
        ------|-------|--------------|----------
        25m9s |       |              | Declaration#4bfb53be::DirectAccessHolder::thisCouldAccessMember#3#dispred#ffff@8a38e2tm
        17m1s |       |              | Declaration#4bfb53be::DirectAccessHolder::thisCouldAccessMember#3#dispred#fffb@0796c497
         3.5s |   130 | 116ms @ 3    | Declaration#4bfb53be::DirectAccessHolder::thisCanAccessClassTrans#fff@926a68j9
         3.3s |       |              | Declaration#4bfb53be::DirectAccessHolder::thisCouldAccessMember#3#dispred#fffb_1230#join_rhs@25e9ffj8
         1.7s |     3 |  1.7s @ 1    | Element#496c7fc2::ElementBase::toString#0#dispred#ff@fcd81c49
         1.3s |       |              | Declaration#4bfb53be::DirectAccessHolder::thisCouldAccessMember#3#dispred#fffb_0132#join_rhs@9c2065t1
         1.3s |       |              | Declaration#4bfb53be::DirectAccessHolder::thisCouldAccessMember#3#dispred#ffff_0132#join_rhs@672330eh
         1.1s |       |              | Declaration#4bfb53be::DirectAccessHolder::thisCanAccessClassTrans#fff_102#join_rhs@f7d5464o
        829ms |   336 |  85ms @ 6    | Enclosing#c50c5fbf::exprEnclosingElement#1#ff@e34d9wq1
        615ms |       |              | Expr#ef463c5d::Expr::getType#ff@e265e79q
```

After:
```
Most expensive predicates for completed query RuleOfTwo.ql:
        time  | evals |  max @ iter | predicate
        ------|-------|-------------|----------
        11.8s |       |             | _#Class#bacd9b46::Class::getADerivedClass#0#dispredPlus#ff_#Declaration#4bfb53be::AccessHolder::getE__#antijoin_rhs#1@fb0627h8
        4.8s |       |             | _#Class#bacd9b46::Class::getADerivedClass#0#dispredPlus#ff_#Declaration#4bfb53be::AccessHolder::getE__#antijoin_rhs#4@c43dbeia
        3.8s |       |             | _#Class#bacd9b46::Class::getADerivedClass#0#dispredPlus#ff_#Declaration#4bfb53be::AccessHolder::getE__#antijoin_rhs#3@313e5963
        3.4s |   130 | 93ms @ 3    | Declaration#4bfb53be::DirectAccessHolder::thisCanAccessClassTrans#fff@a0289bfg
        1.5s |     3 | 1.5s @ 1    | Element#496c7fc2::ElementBase::toString#0#dispred#ff@fcd81c49
        806ms |       |             | Declaration#4bfb53be::DirectAccessHolder::thisCanAccessClassTrans#fff_021#join_rhs@cc1b76s7
        721ms |   336 | 61ms @ 5    | Enclosing#c50c5fbf::exprEnclosingElement#1#ff@e34d9wq1
        489ms |       |             | Expr#ef463c5d::Expr::getType#ff@e265e79q
        337ms |   130 | 62ms @ 5    | Class#bacd9b46::Class::accessOfBaseMemberMulti#ffff@0165b0dr
        329ms |       |             | Variable#7a968d4e::ParameterDeclarationEntry::getAnonymousParameterDescription#0#dispred#ff@0f12bdvq
        211ms |       |             | exprs_10#join_rhs@5481143i
```
2023-03-08 17:44:19 +00:00
Robert Marsh
b941d54f1f C++ Move RangeAnalysis to work around shadowing 2023-03-08 11:32:37 -05:00
Robert Marsh
50fac3060c C++: split RA into constant and relative phases 2023-03-08 11:32:36 -05:00
Jeroen Ketema
30cbc91092 C++: Update XXE XML query with DataFlow::ConfigSig 2023-03-08 15:04:53 +01:00
Jeroen Ketema
6f2407412e C++: Update some dataflow tests to use DataFlow::ConfigSig 2023-03-08 15:04:53 +01:00
Jeroen Ketema
8253f2d343 C++: Update UnsafeDaclSecurityDescriptor with DataFlow::ConfigSig 2023-03-08 15:04:53 +01:00
Jeroen Ketema
7fe1a9431c C++: Update PotentiallyExposedSystemData with DataFlow::ConfigSig 2023-03-08 15:04:53 +01:00
Jeroen Ketema
53aa34bdd3 C++: Update UnsafeCreateProcessCall with DataFlow::ConfigSig 2023-03-08 15:04:53 +01:00
Jeroen Ketema
af612a12de C++: Update TlsSettingsMisconfiguration with DataFlow::ConfigSig 2023-03-08 15:04:52 +01:00
Jeroen Ketema
4363a8ea30 C++: Update leap year queries with DataFlow::ConfigSig 2023-03-08 15:04:52 +01:00
Jeroen Ketema
e65ba13da4 C++: Update NonConstantFormat with DataFlow::ConfigSig 2023-03-08 15:04:52 +01:00
Jeroen Ketema
661160a98e C++: Update PrivateCleartextWrite with DataFlow::ConfigSig 2023-03-08 15:04:45 +01:00
Mathias Vorreiter Pedersen
a247a8b3ea Merge branch 'main' into mathiasvp/replace-ast-with-ir-use-usedataflow 2023-03-08 13:35:35 +00:00
Mathias Vorreiter Pedersen
e68bb53a6b Merge pull request #12435 from jketema/more-config
C++: Convert a number of data flow based queries to use `ConfigSig`
2023-03-08 13:25:54 +00:00
Mathias Vorreiter Pedersen
619266d04b C++: Fix floating point imprecision. 2023-03-08 13:24:01 +00:00
Mathias Vorreiter Pedersen
ce0f2b1788 C++: Accept test changes. 2023-03-08 13:23:58 +00:00
Mathias Vorreiter Pedersen
b0cb65403b C++: Add range analysis for unary minus. 2023-03-08 13:23:18 +00:00
Mathias Vorreiter Pedersen
4cb5bea2c6 C++: Add simple negation test cases. 2023-03-08 13:23:05 +00:00
Mathias Vorreiter Pedersen
c84d88f5aa Merge pull request #12429 from MathiasVP/actually-implement-language-specific-flow-into-call-node-cand1
C++: Implement `getAdditionalFlowIntoCallNodeTerm`
2023-03-08 11:58:56 +00:00
Mathias Vorreiter Pedersen
8308c661b4 Merge pull request #12432 from MathiasVP/fix-ir-uninitialized-node
C++: Fix `asUninitialized`
2023-03-08 10:03:46 +00:00
Mathias Vorreiter Pedersen
5a6b94eda2 C++: Respond to PR reviews. 2023-03-08 09:38:56 +00:00
Mathias Vorreiter Pedersen
263b208282 C++: Disambiguate a test annotation. 2023-03-08 09:07:07 +00:00
Jeroen Ketema
5391b13db9 C++: Make dataflow configuration modules private in qll files 2023-03-08 09:18:09 +01:00
Jeroen Ketema
0f8a12f3ac C++: Add change note for deprecated data flow configurations in qll files 2023-03-08 09:00:43 +01:00
Jeroen Ketema
13bdd9c0c6 C++: Fix query compliation
Apparently some queries we skipped in the testing I did locally.
2023-03-07 19:16:10 +01:00
Jeroen Ketema
57c5d5f2c7 C++: Add QLDoc on configuration modules where the original class had one 2023-03-07 19:01:05 +01:00
Jeroen Ketema
0c39d1e5ca C++: Fix query formatting 2023-03-07 18:55:58 +01:00
Jeroen Ketema
2eb2e11ef7 C++: Fix query compilation 2023-03-07 18:53:07 +01:00
Jeroen Ketema
fb57914751 C++: Convert a number of data flow based queries to use ConfigSig 2023-03-07 18:21:52 +01:00
Mathias Vorreiter Pedersen
cc0b8bbebb Merge pull request #12430 from MathiasVP/no-to-string-on-state-in-cast-array-pointer-arith
C++: Convert `cpp/upcast-array-pointer-arithmetic` to the new API
2023-03-07 16:48:15 +00:00
Mathias Vorreiter Pedersen
eea02e1ac1 C++: Accept test changes. 2023-03-07 16:18:43 +00:00
Mathias Vorreiter Pedersen
a39a6ea648 C++: Don't use indirect instructions for asUnitialized. 2023-03-07 16:18:27 +00:00
Mathias Vorreiter Pedersen
ce02de48a0 C++: Fix Code Scanning error. 2023-03-07 14:40:36 +00:00
Mathias Vorreiter Pedersen
f2b311a008 C++: We don't need to check type equivalence at the end anymore: the dataflow state now precisely tracks the types. 2023-03-07 14:31:11 +00:00