pghs975uc
/
infer_clone
1
0
Fork 0
Code Issues Pull Requests Releases Wiki Activity
You can not select more than 25 topics Topics must start with a letter or number, can include dashes ('-') and can be up to 35 characters long.
Branches Tags
${ item.name }
Create tag ${ searchTerm }
Create branch ${ searchTerm }
from 'e1f4da320f'
${ noResults }
infer_clone/infer/tests/codetoanalyze/java/performance/issues.exp

224 lines
61 KiB
Raw Normal View History Unescape

[RFC][java] kill regexp-based line number fixer Summary: Old versions of sawja/javalib got the line numbers slightly wrong. The workaround was to do a regexp search in the source file for the right line. My understanding is that this is no longer necessary. This diff removes it. Reviewed By: jvillard Differential Revision: D19033415 fbshipit-source-id: 2da19d66d
5 years ago
codetoanalyze/java/performance/AndroidXCollectionTest.java, AndroidXCollectionTest.iterate_over_sparsearraycompat_linear(androidx.collection.SparseArrayCompat):void, 0, EXPENSIVE_EXECUTION_TIME, no_bucket, ERROR, [with estimated cost 5 + 5 ⋅ sa.length, O(sa.length), degree = 1,{sa.length},Loop at line 12]
codetoanalyze/java/performance/Array.java, codetoanalyze.java.performance.Array.array_access_overrun_bad():void, 3, BUFFER_OVERRUN_L2, no_bucket, ERROR, [<Offset trace>,Assignment,<Length trace>,Array declaration,Array access: Offset: [2, 8] Size: 8]
codetoanalyze/java/performance/Array.java, codetoanalyze.java.performance.Array.array_access_weird_ok(long[],int):void, 0, EXPENSIVE_EXECUTION_TIME, no_bucket, ERROR, [with estimated cost 2 + 13 ⋅ length, O(length), degree = 1,{length},Loop at line 30]
codetoanalyze/java/performance/Array.java, codetoanalyze.java.performance.Array.binary_search_log(java.lang.String[]):int, 0, EXPENSIVE_EXECUTION_TIME, no_bucket, ERROR, [with estimated cost 2 + log(arr.length), O(log(arr.length)), degree = 0 + 1⋅log,{arr.length},Modeled call to Arrays.binarySearch]
codetoanalyze/java/performance/Array.java, codetoanalyze.java.performance.Array.bsearch_log(int):void, 1, EXPENSIVE_EXECUTION_TIME, no_bucket, ERROR, [with estimated cost 5 + log(i), O(log(i)), degree = 0 + 1⋅log,{i},Modeled call to Arrays.binarySearch]
codetoanalyze/java/performance/Array.java, codetoanalyze.java.performance.Array.call_gen_and_iter_types(int):void, 0, EXPENSIVE_EXECUTION_TIME, no_bucket, ERROR, [with estimated cost 13 + 5 ⋅ x, O(x), degree = 1,{x},call to java.lang.String[] Array.gen_and_iter_types(int),Loop at line 76]
[inferbo] Weak update on array contents Summary: The `represents_multiple_values` flag was adopted to decide whether updating abstract value strongly or weakly. However, the flag was included in the `Val` domain, which is strange, because it is a property of abstract locations, rather than abstract values. This makes the behavior of memory update function depend on the abstract value to update, making its code complicated. This diff detach the `represents_multiple_values` flag from the `Value` domain, thus the memory update does not depend on the abstract value. Since this is a refactoring, I believe the diff should not make many semantic changes. Reviewed By: ezgicicek Differential Revision: D16441734 fbshipit-source-id: 4c10779d7
5 years ago
codetoanalyze/java/performance/Array.java, codetoanalyze.java.performance.Array.call_gen_and_iter_types_linear_FP(int,int):void, 0, INFINITE_EXECUTION_TIME, no_bucket, ERROR, [Unbounded loop,Loop at line 88]
[RFC][java] kill regexp-based line number fixer Summary: Old versions of sawja/javalib got the line numbers slightly wrong. The workaround was to do a regexp search in the source file for the right line. My understanding is that this is no longer necessary. This diff removes it. Reviewed By: jvillard Differential Revision: D19033415 fbshipit-source-id: 2da19d66d
5 years ago
codetoanalyze/java/performance/Array.java, codetoanalyze.java.performance.Array.call_gen_and_iter_types_linear_FP(int,int):void, 2, INTEGER_OVERFLOW_L5, no_bucket, ERROR, [<LHS trace>,Assignment,Binary operation: ([0, +oo] + 1):signed32]
codetoanalyze/java/performance/Array.java, codetoanalyze.java.performance.Array.copyOf_linear(java.lang.String[]):void, 0, EXPENSIVE_EXECUTION_TIME, no_bucket, ERROR, [with estimated cost 4 + arr.length, O(arr.length), degree = 1,{arr.length},Modeled call to Arrays.copyOf]
codetoanalyze/java/performance/Array.java, codetoanalyze.java.performance.Array.fill_big_constant():void, 1, EXPENSIVE_EXECUTION_TIME, no_bucket, ERROR, [with estimated cost 307, O(1), degree = 0]
codetoanalyze/java/performance/Array.java, codetoanalyze.java.performance.Array.fill_linear(java.lang.String[]):void, 0, EXPENSIVE_EXECUTION_TIME, no_bucket, ERROR, [with estimated cost 2 + arr.length, O(arr.length), degree = 1,{arr.length},Modeled call to Arrays.fill]
codetoanalyze/java/performance/Array.java, codetoanalyze.java.performance.Array.gen_and_iter_types(int):java.lang.String[], 1, EXPENSIVE_EXECUTION_TIME, no_bucket, ERROR, [with estimated cost 5 + 5 ⋅ length, O(length), degree = 1,{length},Loop at line 76]
codetoanalyze/java/performance/Array.java, codetoanalyze.java.performance.Array.init_array_linear():void, 1, EXPENSIVE_EXECUTION_TIME, no_bucket, ERROR, [with estimated cost 2564, O(1), degree = 0]
codetoanalyze/java/performance/Array.java, codetoanalyze.java.performance.Array.sort_array_nlogn(int):void, 1, EXPENSIVE_EXECUTION_TIME, no_bucket, ERROR, [with estimated cost 5 + size × log(size), O(size × log(size)), degree = 1 + 1⋅log,{size},Modeled call to Arrays.sort,{size},Modeled call to Arrays.sort]
codetoanalyze/java/performance/Array.java, codetoanalyze.java.performance.Array.toArray_linear(java.util.ArrayList):void, 0, EXPENSIVE_EXECUTION_TIME, no_bucket, ERROR, [with estimated cost 2 + 6 ⋅ list.length + 3 ⋅ (list.length + 1), O(list.length), degree = 1,{list.length + 1},Loop at line 92,{list.length},Loop at line 92]
codetoanalyze/java/performance/ArrayCost.java, ArrayCost.ArrayCost(int[]):void, 3, EXPENSIVE_EXECUTION_TIME, no_bucket, ERROR, [with estimated cost 5 + 5 ⋅ mag.length, O(mag.length), degree = 1,{mag.length},Loop at line 15]
codetoanalyze/java/performance/ArrayCost.java, ArrayCost.isPowOfTwo_FP(int):boolean, 3, EXPENSIVE_EXECUTION_TIME, no_bucket, ERROR, [with estimated cost 883, O(1), degree = 0]
codetoanalyze/java/performance/ArrayListTest.java, ArrayListTest.arraylist_add3_overrun_bad():void, 4, BUFFER_OVERRUN_L1, no_bucket, ERROR, [<Length trace>,Array declaration,Through,Through,Through,Array access: Offset added: 4 Size: 3]
codetoanalyze/java/performance/ArrayListTest.java, ArrayListTest.arraylist_addAll_bad():void, 9, BUFFER_OVERRUN_L1, no_bucket, ERROR, [<Length trace>,Array declaration,Through,Through,Through,Array access: Offset added: 5 Size: 4]
codetoanalyze/java/performance/ArrayListTest.java, ArrayListTest.arraylist_add_in_nested_loop_constant():void, 0, EXPENSIVE_EXECUTION_TIME, no_bucket, ERROR, [with estimated cost 1772, O(1), degree = 0]
codetoanalyze/java/performance/ArrayListTest.java, ArrayListTest.arraylist_empty_overrun_bad():void, 1, BUFFER_OVERRUN_L1, no_bucket, ERROR, [<Length trace>,Array declaration,Array access: Offset added: 1 Size: 0]
codetoanalyze/java/performance/ArrayListTest.java, ArrayListTest.arraylist_empty_underrun_bad():void, 1, BUFFER_OVERRUN_L1, no_bucket, ERROR, [<Length trace>,Array declaration,Array access: Offset added: -1 Size: 0]
codetoanalyze/java/performance/ArrayListTest.java, ArrayListTest.arraylist_get_overrun_bad():void, 2, BUFFER_OVERRUN_L1, no_bucket, ERROR, [<Length trace>,Array declaration,Through,Array access: Offset: 2 Size: 1]
codetoanalyze/java/performance/ArrayListTest.java, ArrayListTest.arraylist_get_underrun_bad():void, 1, BUFFER_OVERRUN_L1, no_bucket, ERROR, [<Length trace>,Array declaration,Array access: Offset: 0 Size: 0]
codetoanalyze/java/performance/ArrayListTest.java, ArrayListTest.arraylist_remove_bad():void, 4, BUFFER_OVERRUN_L1, no_bucket, ERROR, [<Length trace>,Array declaration,Through,Through,Through,Array access: Offset: 1 Size: 1]
codetoanalyze/java/performance/ArrayListTest.java, ArrayListTest.arraylist_remove_in_loop_Good():void, 4, EXPENSIVE_EXECUTION_TIME, no_bucket, ERROR, [with estimated cost 211, O(1), degree = 0]
codetoanalyze/java/performance/ArrayListTest.java, ArrayListTest.arraylist_remove_in_loop_Good():void, 5, BUFFER_OVERRUN_L5, no_bucket, ERROR, [<Offset trace>,Assignment,<Length trace>,Array declaration,Array access: Offset: [0, 9] Size: [0, 10]]
codetoanalyze/java/performance/ArrayListTest.java, ArrayListTest.arraylist_remove_overrun_bad():void, 2, BUFFER_OVERRUN_L1, no_bucket, ERROR, [<Length trace>,Array declaration,Through,Array access: Offset: 1 Size: 1]
codetoanalyze/java/performance/ArrayListTest.java, ArrayListTest.arraylist_set_overrun_bad():void, 2, BUFFER_OVERRUN_L1, no_bucket, ERROR, [<Length trace>,Array declaration,Through,Array access: Offset: 1 Size: 1]
codetoanalyze/java/performance/ArrayListTest.java, ArrayListTest.arraylist_set_underrun_bad():void, 1, BUFFER_OVERRUN_L1, no_bucket, ERROR, [<Length trace>,Array declaration,Array access: Offset: 0 Size: 0]
codetoanalyze/java/performance/ArrayListTest.java, ArrayListTest.boolean_control_var_linear():void, 0, EXPENSIVE_EXECUTION_TIME, no_bucket, ERROR, [with estimated cost 2 + 19 ⋅ this.arr.length + 4 ⋅ (this.arr.length + 1), O(this.arr.length), degree = 1,{this.arr.length + 1},Loop at line 300,{this.arr.length},Loop at line 300]
codetoanalyze/java/performance/ArrayListTest.java, ArrayListTest.call_init_with_put_linear(java.util.ArrayList):void, 0, EXPENSIVE_EXECUTION_TIME, no_bucket, ERROR, [with estimated cost 13 + 13 ⋅ a.length + 3 ⋅ (a.length + 1), O(a.length), degree = 1,{a.length + 1},call to HashMap ArrayListTest.init_with_put_linear(ArrayList),Loop at line 309,{a.length},call to HashMap ArrayListTest.init_with_put_linear(ArrayList),Loop at line 309]
codetoanalyze/java/performance/ArrayListTest.java, ArrayListTest.call_sortArrayList(java.util.ArrayList):void, 0, EXPENSIVE_EXECUTION_TIME, no_bucket, ERROR, [with estimated cost 4 + list.length × log(list.length), O(list.length × log(list.length)), degree = 1 + 1⋅log,{list.length},call to void ArrayListTest.sortArrayList(ArrayList),Modeled call to Collections.sort,{list.length},call to void ArrayListTest.sortArrayList(ArrayList),Modeled call to Collections.sort]
codetoanalyze/java/performance/ArrayListTest.java, ArrayListTest.constructor_add_all(java.util.ArrayList,java.util.ArrayList):void, 3, EXPENSIVE_EXECUTION_TIME, no_bucket, ERROR, [with estimated cost 10 + 5 ⋅ (l.length + list.length) + 3 ⋅ (l.length + list.length + 1), O((l.length + list.length)), degree = 1,{l.length + list.length + 1},Loop at line 250,{l.length + list.length},Loop at line 250]
codetoanalyze/java/performance/ArrayListTest.java, ArrayListTest.constructor_add_all_sym(java.util.ArrayList,java.util.ArrayList):void, 3, EXPENSIVE_EXECUTION_TIME, no_bucket, ERROR, [with estimated cost 10 + 5 ⋅ (l.length + list.length) + 3 ⋅ (l.length + list.length + 1), O((l.length + list.length)), degree = 1,{l.length + list.length + 1},Loop at line 257,{l.length + list.length},Loop at line 257]
codetoanalyze/java/performance/ArrayListTest.java, ArrayListTest.constructor_linear(java.util.ArrayList):void, 1, EXPENSIVE_EXECUTION_TIME, no_bucket, ERROR, [with estimated cost 6 + 5 ⋅ list.length + 3 ⋅ (list.length + 1), O(list.length), degree = 1,{list.length + 1},Loop at line 232,{list.length},Loop at line 232]
codetoanalyze/java/performance/ArrayListTest.java, ArrayListTest.constructor_modify(java.util.ArrayList):void, 5, EXPENSIVE_EXECUTION_TIME, no_bucket, ERROR, [with estimated cost 18 + 5 ⋅ (list.length + 4) + 3 ⋅ (list.length + 5), O(list.length), degree = 1,{list.length + 5},Loop at line 243,{list.length + 4},Loop at line 243]
codetoanalyze/java/performance/ArrayListTest.java, ArrayListTest.empty_list_constant(int):void, 1, CONDITION_ALWAYS_FALSE, no_bucket, WARNING, [Here]
codetoanalyze/java/performance/ArrayListTest.java, ArrayListTest.init_with_put_linear(java.util.ArrayList):java.util.HashMap, 1, EXPENSIVE_EXECUTION_TIME, no_bucket, ERROR, [with estimated cost 7 + 13 ⋅ a.length + 3 ⋅ (a.length + 1), O(a.length), degree = 1,{a.length + 1},Loop at line 309,{a.length},Loop at line 309]
codetoanalyze/java/performance/ArrayListTest.java, ArrayListTest.iterate_over_arraylist(java.util.ArrayList):void, 0, EXPENSIVE_EXECUTION_TIME, no_bucket, ERROR, [with estimated cost 5 + 5 ⋅ list.length, O(list.length), degree = 1,{list.length},Loop at line 15]
codetoanalyze/java/performance/ArrayListTest.java, ArrayListTest.iterate_over_arraylist_shortcut_FP(java.util.ArrayList):boolean, 0, EXPENSIVE_EXECUTION_TIME, no_bucket, ERROR, [with estimated cost 4 + 11 ⋅ list.length + 2 ⋅ list.length × (11-max(10, list.elements)) + 3 ⋅ (list.length + 1) × (11-max(10, list.elements)), O(list.length × (-list.elements + 11)), degree = 2,{11-max(10, list.elements)},Loop at line 198,{list.length + 1},Loop at line 198,{11-max(10, list.elements)},Loop at line 198,{list.length},Loop at line 198]
codetoanalyze/java/performance/ArrayListTest.java, ArrayListTest.iterate_over_arraylist_with_inner(java.util.ArrayList):void, 2, EXPENSIVE_EXECUTION_TIME, no_bucket, ERROR, [with estimated cost 7 + 11 ⋅ list1.length + 3 ⋅ (list1.length + 1), O(list1.length), degree = 1,{list1.length + 1},Loop at line 189,{list1.length},Loop at line 189]
codetoanalyze/java/performance/ArrayListTest.java, ArrayListTest.iterate_over_local_arraylist(java.util.ArrayList):void, 1, EXPENSIVE_EXECUTION_TIME, no_bucket, ERROR, [with estimated cost 7 + 5 ⋅ list.length, O(list.length), degree = 1,{list.length},Loop at line 20]
codetoanalyze/java/performance/ArrayListTest.java, ArrayListTest.iterate_while_has_next(java.util.ArrayList):void, 2, EXPENSIVE_EXECUTION_TIME, no_bucket, ERROR, [with estimated cost 4 + 10 ⋅ list.length + 3 ⋅ (list.length + 1), O(list.length), degree = 1,{list.length + 1},Loop at line 181,{list.length},Loop at line 181]
codetoanalyze/java/performance/ArrayListTest.java, ArrayListTest.iterate_with_iterator(java.util.ArrayList):void, 0, EXPENSIVE_EXECUTION_TIME, no_bucket, ERROR, [with estimated cost 4 + 8 ⋅ list.length + 3 ⋅ (list.length + 1), O(list.length), degree = 1,{list.length + 1},Loop at line 175,{list.length},Loop at line 175]
codetoanalyze/java/performance/ArrayListTest.java, ArrayListTest.json_array_constructor_linear(java.util.ArrayList):void, 1, EXPENSIVE_EXECUTION_TIME, no_bucket, ERROR, [with estimated cost 6 + 5 ⋅ arr.length + 3 ⋅ (arr.length + 1), O(arr.length), degree = 1,{arr.length + 1},Loop at line 277,{arr.length},Loop at line 277]
codetoanalyze/java/performance/ArrayListTest.java, ArrayListTest.linear(int,java.util.ArrayList):void, 0, EXPENSIVE_EXECUTION_TIME, no_bucket, ERROR, [with estimated cost 4 + 9 ⋅ (-i + a.length + 1), O((-i + a.length)), degree = 1,{-i + a.length + 1},Loop at line 284]
codetoanalyze/java/performance/ArrayListTest.java, ArrayListTest.max_linear(java.util.ArrayList):Person, 0, EXPENSIVE_EXECUTION_TIME, no_bucket, ERROR, [with estimated cost 8 + people.length, O(people.length), degree = 1,{people.length},Modeled call to Collections.max]
codetoanalyze/java/performance/ArrayListTest.java, ArrayListTest.remove_string_from_list(java.lang.String):boolean, 1, EXPENSIVE_EXECUTION_TIME, no_bucket, ERROR, [with estimated cost 5 + 12 ⋅ this.list.length + 3 ⋅ (this.list.length + 1), O(this.list.length), degree = 1,{this.list.length + 1},Loop at line 221,{this.list.length},Loop at line 221]
codetoanalyze/java/performance/ArrayListTest.java, ArrayListTest.sortArrayList(java.util.ArrayList):void, 0, EXPENSIVE_EXECUTION_TIME, no_bucket, ERROR, [with estimated cost 2 + list.length × log(list.length), O(list.length × log(list.length)), degree = 1 + 1⋅log,{list.length},Modeled call to Collections.sort,{list.length},Modeled call to Collections.sort]
codetoanalyze/java/performance/ArrayListTest.java, ArrayListTest.sort_comparator_nlogn(java.util.ArrayList):void, 0, EXPENSIVE_EXECUTION_TIME, no_bucket, ERROR, [with estimated cost 8 + people.length × log(people.length), O(people.length × log(people.length)), degree = 1 + 1⋅log,{people.length},Modeled call to Collections.sort,{people.length},Modeled call to Collections.sort]
codetoanalyze/java/performance/ArrayListTest.java, ArrayListTest.substitute_array_block_linear(java.util.ArrayList,java.util.ArrayList):void, 6, EXPENSIVE_EXECUTION_TIME, no_bucket, ERROR, [with estimated cost 33 + 5 ⋅ (b.length + a.length), O((b.length + a.length)), degree = 1,{b.length + a.length},call to void ArrayListTest.iterate_over_arraylist(ArrayList),Loop at line 15]
codetoanalyze/java/performance/ArrayListTest.java, ArrayListTest.substitute_array_block_linear(java.util.ArrayList,java.util.ArrayList):void, 6, INTEGER_OVERFLOW_L5, no_bucket, ERROR, [Parameter `b.elements[*]`,Call,Parameter `a.elements[*]`,Through,Through,Call,<LHS trace>,Assignment,Binary operation: ([0, +oo] + 1):signed32 by call to `void ArrayListTest.iterate_over_arraylist(ArrayList)` ]
codetoanalyze/java/performance/Break.java, codetoanalyze.java.performance.Break.break_constant(int):int, 0, EXPENSIVE_EXECUTION_TIME, no_bucket, ERROR, [with estimated cost 10 + 7 ⋅ p, O(p), degree = 1,{p},call to int Break.break_loop(int,int),Loop at line 12]
codetoanalyze/java/performance/Break.java, codetoanalyze.java.performance.Break.break_loop(int,int):int, 0, EXPENSIVE_EXECUTION_TIME, no_bucket, ERROR, [with estimated cost 2 + 7 ⋅ p, O(p), degree = 1,{p},Loop at line 12]
codetoanalyze/java/performance/Break.java, codetoanalyze.java.performance.Break.break_outer_loop_MaybeInfinite(int,int):void, 2, EXPENSIVE_EXECUTION_TIME, no_bucket, ERROR, [with estimated cost 2 + 4 ⋅ maxI + 3 ⋅ maxI × (min(12, maxJ)) + 5 ⋅ maxI × (12-max(0, maxJ)) + 5 ⋅ (min(11, maxI)) × (min(11, maxJ)), O(maxI × maxJ), degree = 2,{min(11, maxJ)},Loop at line 37,{min(11, maxI)},Loop at line 35,{12-max(0, maxJ)},Loop at line 35,{min(12, maxJ)},Loop at line 37,{maxI},Loop at line 35]
codetoanalyze/java/performance/CantHandle.java, CantHandle.quadratic(int):void, 0, EXPENSIVE_EXECUTION_TIME, no_bucket, ERROR, [with estimated cost 2 + 6 ⋅ x², O(x²), degree = 2,{x},Loop at line 31,{x},Loop at line 31]
[cost] Refactor cost issue types and enable detecting allocation complexity increase on cold start Summary: - Add allocation costs to `costs-report.json` and enable diffing over allocation costs. - Also, let's be more consistent and modular in naming our cost issues. - introduce a generic issue type `X_TIME_COMPLEXITY_INCREASE` where `X` can be one of the cost kinds. If the function is on the cold start, issue can have the `COLD_START` suffix. Similarly for infinite/zero/expensive calls. - Change `PERFORMANCE_VARIATION` -> `EXECUTION_TIME_COMPLEXITY_INCREASE` - Add new issue type for `ALLOCATION_COMPLEXITY_INCREASE_COLD_START` which will be enabled by default - Refactor cost issues to be more modular and succinct. This also makes addition of a new cost kind very easy by adding the kind into the `enabled_cost_kinds` list in `CostKind.ml` Reviewed By: mbouaziz Differential Revision: D15822681 fbshipit-source-id: cf89ece59
6 years ago
codetoanalyze/java/performance/CantHandle.java, CantHandle.square_root_FP(int):void, 0, INFINITE_EXECUTION_TIME, no_bucket, ERROR, [Unbounded loop,Loop at line 17]
[RFC][java] kill regexp-based line number fixer Summary: Old versions of sawja/javalib got the line numbers slightly wrong. The workaround was to do a regexp search in the source file for the right line. My understanding is that this is no longer necessary. This diff removes it. Reviewed By: jvillard Differential Revision: D19033415 fbshipit-source-id: 2da19d66d
5 years ago
codetoanalyze/java/performance/CantHandle.java, CantHandle.square_root_FP(int):void, 1, INTEGER_OVERFLOW_L5, no_bucket, ERROR, [<LHS trace>,Assignment,<RHS trace>,Assignment,Binary operation: ([0, +oo] × [0, +oo]):signed32]
codetoanalyze/java/performance/CantHandle.java, CantHandle.square_root_FP(int):void, 2, INTEGER_OVERFLOW_L5, no_bucket, ERROR, [<LHS trace>,Assignment,Binary operation: ([0, +oo] + 1):signed32]
[cost] Refactor cost issue types and enable detecting allocation complexity increase on cold start Summary: - Add allocation costs to `costs-report.json` and enable diffing over allocation costs. - Also, let's be more consistent and modular in naming our cost issues. - introduce a generic issue type `X_TIME_COMPLEXITY_INCREASE` where `X` can be one of the cost kinds. If the function is on the cold start, issue can have the `COLD_START` suffix. Similarly for infinite/zero/expensive calls. - Change `PERFORMANCE_VARIATION` -> `EXECUTION_TIME_COMPLEXITY_INCREASE` - Add new issue type for `ALLOCATION_COMPLEXITY_INCREASE_COLD_START` which will be enabled by default - Refactor cost issues to be more modular and succinct. This also makes addition of a new cost kind very easy by adding the kind into the `enabled_cost_kinds` list in `CostKind.ml` Reviewed By: mbouaziz Differential Revision: D15822681 fbshipit-source-id: cf89ece59
6 years ago
codetoanalyze/java/performance/CantHandle.java, CantHandle.square_root_variant_FP(int):void, 0, INFINITE_EXECUTION_TIME, no_bucket, ERROR, [Unbounded loop,Loop at line 25]
[RFC][java] kill regexp-based line number fixer Summary: Old versions of sawja/javalib got the line numbers slightly wrong. The workaround was to do a regexp search in the source file for the right line. My understanding is that this is no longer necessary. This diff removes it. Reviewed By: jvillard Differential Revision: D19033415 fbshipit-source-id: 2da19d66d
5 years ago
codetoanalyze/java/performance/CantHandle.java, CantHandle.square_root_variant_FP(int):void, 2, INTEGER_OVERFLOW_L5, no_bucket, ERROR, [<LHS trace>,Assignment,Binary operation: ([0, +oo] + 1):signed32]
codetoanalyze/java/performance/CollectionTest.java, CollectionTest.ensure_call(CollectionTest$MyCollection):void, 0, EXPENSIVE_EXECUTION_TIME, no_bucket, ERROR, [with estimated cost 11 + 5 ⋅ list.length, O(list.length), degree = 1,{list.length},call to void CollectionTest.iterate_over_mycollection(CollectionTest$MyCollection),Loop at line 17]
codetoanalyze/java/performance/CollectionTest.java, CollectionTest.iterate_over_call_quad(int,CollectionTest$MyCollection):void, 0, EXPENSIVE_EXECUTION_TIME, no_bucket, ERROR, [with estimated cost 4 + 18 ⋅ list.length + 5 ⋅ list.length × list.length + 3 ⋅ (list.length + 1), O(list.length × list.length), degree = 2,{list.length + 1},Loop at line 48,{list.length},call to void CollectionTest.iterate_over_mycollection(CollectionTest$MyCollection),Loop at line 17,{list.length},Loop at line 48]
codetoanalyze/java/performance/CollectionTest.java, CollectionTest.iterate_over_mycollection(CollectionTest$MyCollection):void, 0, EXPENSIVE_EXECUTION_TIME, no_bucket, ERROR, [with estimated cost 5 + 5 ⋅ list.length, O(list.length), degree = 1,{list.length},Loop at line 17]
codetoanalyze/java/performance/CollectionTest.java, CollectionTest.iterate_over_mycollection_quad_FN(java.util.concurrent.ConcurrentLinkedQueue):void, 0, EXPENSIVE_EXECUTION_TIME, no_bucket, ERROR, [with estimated cost 4 + 18 ⋅ mSubscribers.length + 5 ⋅ mSubscribers.length × mSubscribers.elements.*.length.ub + 3 ⋅ (mSubscribers.length + 1), O(mSubscribers.length × mSubscribers.elements.*.length.ub), degree = 2,{mSubscribers.length + 1},Loop at line 29,{mSubscribers.elements.*.length.ub},call to void CollectionTest.iterate_over_mycollection(CollectionTest$MyCollection),Loop at line 17,{mSubscribers.length},Loop at line 29]
codetoanalyze/java/performance/CollectionTest.java, CollectionTest.iterate_over_some_java_collection(java.util.concurrent.ConcurrentLinkedQueue):void, 0, EXPENSIVE_EXECUTION_TIME, no_bucket, ERROR, [with estimated cost 4 + 8 ⋅ mSubscribers.length + 3 ⋅ (mSubscribers.length + 1), O(mSubscribers.length), degree = 1,{mSubscribers.length + 1},Loop at line 22,{mSubscribers.length},Loop at line 22]
codetoanalyze/java/performance/CollectionTest.java, CollectionTest.loop_over_call(int,CollectionTest$MyCollection):void, 0, EXPENSIVE_EXECUTION_TIME, no_bucket, ERROR, [with estimated cost 2 + 15 ⋅ size + 5 ⋅ size × list.length, O(size × list.length), degree = 2,{list.length},call to void CollectionTest.iterate_over_mycollection(CollectionTest$MyCollection),Loop at line 17,{size},Loop at line 41]
codetoanalyze/java/performance/CollectionTest.java, CollectionTest.nested_iterator_qubic(int,CollectionTest$MyCollection,CollectionTest$MyCollection):void, 0, EXPENSIVE_EXECUTION_TIME, no_bucket, ERROR, [with estimated cost 4 + 13 ⋅ list1.length + 5 ⋅ list1.length × list1.length × list2.length + 5 ⋅ list1.length × list1.length × list2.length + 28 ⋅ list1.length × list2.length + 3 ⋅ list1.length × (list2.length + 1) + 3 ⋅ (list1.length + 1), O(list1.length × list1.length × list2.length), degree = 3,{list1.length + 1},Loop at line 55,{list2.length + 1},Loop at line 56,{list2.length},Loop at line 56,{list2.length},Loop at line 56,{list1.length},call to void CollectionTest.iterate_over_mycollection(CollectionTest$MyCollection),Loop at line 17,{list2.length},Loop at line 56,{list1.length},call to void CollectionTest.iterate_over_mycollection(CollectionTest$MyCollection),Loop at line 17,{list1.length},Loop at line 55]
codetoanalyze/java/performance/CollectionTest.java, CollectionTest.sparse_array_linear(android.util.SparseArray):void, 0, EXPENSIVE_EXECUTION_TIME, no_bucket, ERROR, [with estimated cost 2 + 5 ⋅ arr.length + 3 ⋅ (arr.length + 1), O(arr.length), degree = 1,{arr.length + 1},Loop at line 64,{arr.length},Loop at line 64]
codetoanalyze/java/performance/CollectionsTest.java, CollectionsTest.binary_search_log(java.util.List):int, 0, EXPENSIVE_EXECUTION_TIME, no_bucket, ERROR, [with estimated cost 2 + log(list.length), O(log(list.length)), degree = 0 + 1⋅log,{list.length},Modeled call to Collections.binarySearch]
codetoanalyze/java/performance/CollectionsTest.java, CollectionsTest.copy_linear(java.util.List,java.util.List):void, 0, EXPENSIVE_EXECUTION_TIME, no_bucket, ERROR, [with estimated cost 3 + list_to.length, O(list_to.length), degree = 1,{list_to.length},Modeled call to Collections.copy]
codetoanalyze/java/performance/CollectionsTest.java, CollectionsTest.emptyList_constant():void, 0, CONDITION_ALWAYS_FALSE, no_bucket, WARNING, [Here]
codetoanalyze/java/performance/CollectionsTest.java, CollectionsTest.emptySet_zero():void, 1, CONDITION_ALWAYS_FALSE, no_bucket, WARNING, [Here]
codetoanalyze/java/performance/CollectionsTest.java, CollectionsTest.fill_linear(java.util.List,java.lang.String):void, 0, EXPENSIVE_EXECUTION_TIME, no_bucket, ERROR, [with estimated cost 3 + list.length, O(list.length), degree = 1,{list.length},Modeled call to Collections.fill]
codetoanalyze/java/performance/CollectionsTest.java, CollectionsTest.globalEmptyList_constant():void, 0, CONDITION_ALWAYS_FALSE, no_bucket, WARNING, [Here]
codetoanalyze/java/performance/CollectionsTest.java, CollectionsTest.reverse_linear(java.util.List):void, 0, EXPENSIVE_EXECUTION_TIME, no_bucket, ERROR, [with estimated cost 2 + list.length, O(list.length), degree = 1,{list.length},Modeled call to Collections.reverse]
codetoanalyze/java/performance/CollectionsTest.java, CollectionsTest.shuffle_linear(java.util.List,java.util.Random):void, 0, EXPENSIVE_EXECUTION_TIME, no_bucket, ERROR, [with estimated cost 3 + list.length, O(list.length), degree = 1,{list.length},Modeled call to Collections.shuffle]
codetoanalyze/java/performance/CollectionsTest.java, CollectionsTest.unmodifiable_linear(java.util.List):void, 1, EXPENSIVE_EXECUTION_TIME, no_bucket, ERROR, [with estimated cost 5 + 5 ⋅ list.length + 3 ⋅ (list.length + 1), O(list.length), degree = 1,{list.length + 1},Loop at line 53,{list.length},Loop at line 53]
codetoanalyze/java/performance/CollectionsTest.java, CollectionsTest.unmodifiable_map(java.util.Map):void, 0, EXPENSIVE_EXECUTION_TIME, no_bucket, ERROR, [with estimated cost 10 + 8 ⋅ map.length + 3 ⋅ (map.length + 1), O(map.length), degree = 1,{map.length + 1},Loop at line 57,{map.length},Loop at line 57]
codetoanalyze/java/performance/CollectionsTest.java, CollectionsTest.unmodifiable_set(java.util.Set):void, 0, EXPENSIVE_EXECUTION_TIME, no_bucket, ERROR, [with estimated cost 7 + 8 ⋅ set.length + 3 ⋅ (set.length + 1), O(set.length), degree = 1,{set.length + 1},Loop at line 61,{set.length},Loop at line 61]
codetoanalyze/java/performance/Compound_loop.java, codetoanalyze.java.performance.Compound_loop.compound_while(int):int, 2, CONDITION_ALWAYS_TRUE, no_bucket, WARNING, [Here]
codetoanalyze/java/performance/Compound_loop.java, codetoanalyze.java.performance.Compound_loop.compound_while(int):int, 2, EXPENSIVE_EXECUTION_TIME, no_bucket, ERROR, [with estimated cost 4 + 5 ⋅ m + 2 ⋅ (1+max(0, m)), O(m), degree = 1,{1+max(0, m)},Loop at line 15,{m},Loop at line 15]
codetoanalyze/java/performance/Compound_loop.java, codetoanalyze.java.performance.Compound_loop.nested_while_and_or(int):int, 2, CONDITION_ALWAYS_TRUE, no_bucket, WARNING, [Here]
codetoanalyze/java/performance/Compound_loop.java, codetoanalyze.java.performance.Compound_loop.nested_while_and_or(int):int, 2, CONDITION_ALWAYS_TRUE, no_bucket, WARNING, [Here]
[inferbo] Add traces to Conditions always true/false and Unreachable code Reviewed By: ezgicicek Differential Revision: D13082665 fbshipit-source-id: bb0e4cbf3
6 years ago
codetoanalyze/java/performance/Compound_loop.java, codetoanalyze.java.performance.Compound_loop.nested_while_and_or(int):int, 3, CONDITION_ALWAYS_TRUE, no_bucket, WARNING, [Here]
codetoanalyze/java/performance/Compound_loop.java, codetoanalyze.java.performance.Compound_loop.nested_while_and_or(int):int, 3, CONDITION_ALWAYS_TRUE, no_bucket, WARNING, [Here]
[cost] Refactor cost issue types and enable detecting allocation complexity increase on cold start Summary: - Add allocation costs to `costs-report.json` and enable diffing over allocation costs. - Also, let's be more consistent and modular in naming our cost issues. - introduce a generic issue type `X_TIME_COMPLEXITY_INCREASE` where `X` can be one of the cost kinds. If the function is on the cold start, issue can have the `COLD_START` suffix. Similarly for infinite/zero/expensive calls. - Change `PERFORMANCE_VARIATION` -> `EXECUTION_TIME_COMPLEXITY_INCREASE` - Add new issue type for `ALLOCATION_COMPLEXITY_INCREASE_COLD_START` which will be enabled by default - Refactor cost issues to be more modular and succinct. This also makes addition of a new cost kind very easy by adding the kind into the `enabled_cost_kinds` list in `CostKind.ml` Reviewed By: mbouaziz Differential Revision: D15822681 fbshipit-source-id: cf89ece59
6 years ago
codetoanalyze/java/performance/Compound_loop.java, codetoanalyze.java.performance.Compound_loop.while_and_or(int):void, 0, INFINITE_EXECUTION_TIME, no_bucket, ERROR, [Unbounded loop,Loop at line 24]
[RFC][java] kill regexp-based line number fixer Summary: Old versions of sawja/javalib got the line numbers slightly wrong. The workaround was to do a regexp search in the source file for the right line. My understanding is that this is no longer necessary. This diff removes it. Reviewed By: jvillard Differential Revision: D19033415 fbshipit-source-id: 2da19d66d
5 years ago
codetoanalyze/java/performance/Compound_loop.java, codetoanalyze.java.performance.Compound_loop.while_and_or(int):void, 1, CONDITION_ALWAYS_TRUE, no_bucket, WARNING, [Here]
codetoanalyze/java/performance/Compound_loop.java, codetoanalyze.java.performance.Compound_loop.while_and_or(int):void, 2, INTEGER_OVERFLOW_L5, no_bucket, ERROR, [<LHS trace>,Assignment,Binary operation: ([0, +oo] + 1):signed32]
codetoanalyze/java/performance/Continue.java, codetoanalyze.java.performance.Continue.continue_outer_loop():int, 0, EXPENSIVE_EXECUTION_TIME, no_bucket, ERROR, [with estimated cost 7963049, O(1), degree = 0]
codetoanalyze/java/performance/Cost_test.java, codetoanalyze.java.performance.Cost_test.FN_loop2(int):int, 0, EXPENSIVE_EXECUTION_TIME, no_bucket, ERROR, [with estimated cost 2 + 13 ⋅ k, O(k), degree = 1,{k},Loop at line 92]
codetoanalyze/java/performance/Cost_test.java, codetoanalyze.java.performance.Cost_test.band_constant(int):void, 0, EXPENSIVE_EXECUTION_TIME, no_bucket, ERROR, [with estimated cost 1277, O(1), degree = 0]
codetoanalyze/java/performance/Cost_test.java, codetoanalyze.java.performance.Cost_test.call_inputstream_read_linear(java.io.InputStream):void, 3, EXPENSIVE_EXECUTION_TIME, no_bucket, ERROR, [with estimated cost 27304, O(1), degree = 0]
[cost] Report the biggest cost in function Summary: In the previous report, it reported the first cost of node that exceeds a threshold. However, this may hide a bigger cost of node that appears later. This diff changes this to report the biggest cost of node among the costs exceeding the threshold. Reviewed By: ezgicicek Differential Revision: D20116162 fbshipit-source-id: 06199fb46
5 years ago
codetoanalyze/java/performance/Cost_test.java, codetoanalyze.java.performance.Cost_test.call_mult_symbold_quadratic(int):void, 1, EXPENSIVE_EXECUTION_TIME, no_bucket, ERROR, [with estimated cost 14 + 5 ⋅ n + 6 ⋅ n², O(n²), degree = 2,{n},call to void Cost_test.mult_symbols_quadratic(int,int),Loop at line 211,{n},call to void Cost_test.mult_symbols_quadratic(int,int),Loop at line 211,{n},Loop at line 215]
codetoanalyze/java/performance/Cost_test.java, codetoanalyze.java.performance.Cost_test.call_mult_symbold_quadratic(int):void, 1, INTEGER_OVERFLOW_L5, no_bucket, ERROR, [Assignment,Call,<LHS trace>,Assignment,Binary operation: ([0, +oo] + 1):signed32 by call to `void Cost_test.mult_symbols_quadratic(int,int)` ]
[RFC][java] kill regexp-based line number fixer Summary: Old versions of sawja/javalib got the line numbers slightly wrong. The workaround was to do a regexp search in the source file for the right line. My understanding is that this is no longer necessary. This diff removes it. Reviewed By: jvillard Differential Revision: D19033415 fbshipit-source-id: 2da19d66d
5 years ago
codetoanalyze/java/performance/Cost_test.java, codetoanalyze.java.performance.Cost_test.ignore_boolean_symbols_linear(boolean,int):void, 0, EXPENSIVE_EXECUTION_TIME, no_bucket, ERROR, [with estimated cost 2 + 6 ⋅ n + 2 ⋅ (1+max(0, n)), O(n), degree = 1,{1+max(0, n)},Loop at line 135,{n},Loop at line 135]
codetoanalyze/java/performance/Cost_test.java, codetoanalyze.java.performance.Cost_test.loop0_bad():int, 0, EXPENSIVE_EXECUTION_TIME, no_bucket, ERROR, [with estimated cost 1202, O(1), degree = 0]
codetoanalyze/java/performance/Cost_test.java, codetoanalyze.java.performance.Cost_test.loop1_bad():int, 1, EXPENSIVE_EXECUTION_TIME, no_bucket, ERROR, [with estimated cost 1303, O(1), degree = 0]
codetoanalyze/java/performance/Cost_test.java, codetoanalyze.java.performance.Cost_test.loop3(int):int, 0, EXPENSIVE_EXECUTION_TIME, no_bucket, ERROR, [with estimated cost 237, O(1), degree = 0]
codetoanalyze/java/performance/Cost_test.java, codetoanalyze.java.performance.Cost_test.loop_on_unknown_global_linear():void, 0, EXPENSIVE_EXECUTION_TIME, no_bucket, ERROR, [with estimated cost 2 + 5 ⋅ codetoanalyze.java.performance.Cost_test.global + 5 ⋅ (1+max(0, codetoanalyze.java.performance.Cost_test.global)), O(codetoanalyze.java.performance.Cost_test.global), degree = 1,{1+max(0, codetoanalyze.java.performance.Cost_test.global)},Loop at line 203,{codetoanalyze.java.performance.Cost_test.global},Loop at line 203]
codetoanalyze/java/performance/Cost_test.java, codetoanalyze.java.performance.Cost_test.main_bad():int, 3, EXPENSIVE_EXECUTION_TIME, no_bucket, ERROR, [with estimated cost 205, O(1), degree = 0]
codetoanalyze/java/performance/Cost_test.java, codetoanalyze.java.performance.Cost_test.mult_symbols_quadratic(int,int):void, 0, EXPENSIVE_EXECUTION_TIME, no_bucket, ERROR, [with estimated cost 2 + 6 ⋅ x × y, O(x × y), degree = 2,{y},Loop at line 211,{x},Loop at line 211]
codetoanalyze/java/performance/Cost_test_deps.java, codetoanalyze.java.performance.Cost_test_deps.if_bad(int):void, 5, EXPENSIVE_EXECUTION_TIME, no_bucket, ERROR, [with estimated cost 613, O(1), degree = 0]
codetoanalyze/java/performance/Cost_test_deps.java, codetoanalyze.java.performance.Cost_test_deps.if_bad_loop():int, 1, EXPENSIVE_EXECUTION_TIME, no_bucket, ERROR, [with estimated cost 201, O(1), degree = 0]
codetoanalyze/java/performance/Cost_test_deps.java, codetoanalyze.java.performance.Cost_test_deps.if_bad_loop():int, 3, INTEGER_OVERFLOW_L5, no_bucket, ERROR, [<LHS trace>,Assignment,Binary operation: ([0, +oo] + 1):signed32]
codetoanalyze/java/performance/Cost_test_deps.java, codetoanalyze.java.performance.Cost_test_deps.loop_despite_inferbo(int):int, 1, EXPENSIVE_EXECUTION_TIME, no_bucket, ERROR, [with estimated cost 1303, O(1), degree = 0]
codetoanalyze/java/performance/Cost_test_deps.java, codetoanalyze.java.performance.Cost_test_deps.loop_no_dep1(int):int, 2, EXPENSIVE_EXECUTION_TIME, no_bucket, ERROR, [with estimated cost 605, O(1), degree = 0]
codetoanalyze/java/performance/Cost_test_deps.java, codetoanalyze.java.performance.Cost_test_deps.loop_no_dep1(int):int, 3, INTEGER_OVERFLOW_L5, no_bucket, ERROR, [<LHS trace>,Assignment,Binary operation: ([0, +oo] + 1):signed32]
codetoanalyze/java/performance/Cost_test_deps.java, codetoanalyze.java.performance.Cost_test_deps.loop_no_dep2(int):int, 2, EXPENSIVE_EXECUTION_TIME, no_bucket, ERROR, [with estimated cost 610, O(1), degree = 0]
codetoanalyze/java/performance/Cost_test_deps.java, codetoanalyze.java.performance.Cost_test_deps.loop_no_dep2(int):int, 3, INTEGER_OVERFLOW_L5, no_bucket, ERROR, [<LHS trace>,Assignment,Binary operation: ([0, +oo] + 1):signed32]
codetoanalyze/java/performance/Cost_test_deps.java, codetoanalyze.java.performance.Cost_test_deps.nested_loop():int, 1, EXPENSIVE_EXECUTION_TIME, no_bucket, ERROR, [with estimated cost 2543, O(1), degree = 0]
codetoanalyze/java/performance/Cost_test_deps.java, codetoanalyze.java.performance.Cost_test_deps.real_while():int, 2, EXPENSIVE_EXECUTION_TIME, no_bucket, ERROR, [with estimated cost 214, O(1), degree = 0]
codetoanalyze/java/performance/Cost_test_deps.java, codetoanalyze.java.performance.Cost_test_deps.real_while():int, 3, INTEGER_OVERFLOW_L5, no_bucket, ERROR, [<LHS trace>,Assignment,Assignment,<RHS trace>,Assignment,Binary operation: ([0, +oo] + [0, 29]):signed32]
codetoanalyze/java/performance/Cost_test_deps.java, codetoanalyze.java.performance.Cost_test_deps.two_loops():int, 4, INTEGER_OVERFLOW_L5, no_bucket, ERROR, [<LHS trace>,Assignment,Binary operation: ([3, +oo] + 1):signed32]
codetoanalyze/java/performance/Cost_test_deps.java, codetoanalyze.java.performance.Cost_test_deps.two_loops():int, 6, EXPENSIVE_EXECUTION_TIME, no_bucket, ERROR, [with estimated cost 545, O(1), degree = 0]
[cost] Refactor cost issue types and enable detecting allocation complexity increase on cold start Summary: - Add allocation costs to `costs-report.json` and enable diffing over allocation costs. - Also, let's be more consistent and modular in naming our cost issues. - introduce a generic issue type `X_TIME_COMPLEXITY_INCREASE` where `X` can be one of the cost kinds. If the function is on the cold start, issue can have the `COLD_START` suffix. Similarly for infinite/zero/expensive calls. - Change `PERFORMANCE_VARIATION` -> `EXECUTION_TIME_COMPLEXITY_INCREASE` - Add new issue type for `ALLOCATION_COMPLEXITY_INCREASE_COLD_START` which will be enabled by default - Refactor cost issues to be more modular and succinct. This also makes addition of a new cost kind very easy by adding the kind into the `enabled_cost_kinds` list in `CostKind.ml` Reviewed By: mbouaziz Differential Revision: D15822681 fbshipit-source-id: cf89ece59
6 years ago
codetoanalyze/java/performance/EvilCfg.java, EvilCfg.foo(int,int,boolean):void, 0, INFINITE_EXECUTION_TIME, no_bucket, ERROR, [Unbounded loop,Loop at line 15]
[RFC][java] kill regexp-based line number fixer Summary: Old versions of sawja/javalib got the line numbers slightly wrong. The workaround was to do a regexp search in the source file for the right line. My understanding is that this is no longer necessary. This diff removes it. Reviewed By: jvillard Differential Revision: D19033415 fbshipit-source-id: 2da19d66d
5 years ago
codetoanalyze/java/performance/FieldAccess.java, codetoanalyze.java.performance.FieldAccess.iterate_upto_field_size(codetoanalyze.java.performance.FieldAccess$Test):void, 0, EXPENSIVE_EXECUTION_TIME, no_bucket, ERROR, [with estimated cost 2 + 6 ⋅ test.a, O(test.a), degree = 1,{test.a},Loop at line 16]
codetoanalyze/java/performance/IntTest.java, IntTest.control_var_band_add_constant(int,int):void, 2, INTEGER_OVERFLOW_L5, no_bucket, ERROR, [<LHS trace>,Parameter `x`,Binary operation: ([-oo, +oo] + 1):signed32]
[cost] Keep excluding unqualified variables by ItvUpdatedBy Summary: This diff avoids unqualified variables by `ItvUpdatedBy` are qualified later. For example, ``` z = x & y; z = z + 1; ``` While `z` should not be selected as a control variable, it wasn't, because it was qualified by the addition. This pattern introduces FPs in many cases. Reviewed By: ngorogiannis Differential Revision: D18505894 fbshipit-source-id: 13aec3008
5 years ago
codetoanalyze/java/performance/IntTest.java, IntTest.control_var_band_add_constant(int,int):void, 4, INTEGER_OVERFLOW_L5, no_bucket, ERROR, [<LHS trace>,Parameter `x`,Binary operation: ([-oo, +oo] + 1):signed32]
[RFC][java] kill regexp-based line number fixer Summary: Old versions of sawja/javalib got the line numbers slightly wrong. The workaround was to do a regexp search in the source file for the right line. My understanding is that this is no longer necessary. This diff removes it. Reviewed By: jvillard Differential Revision: D19033415 fbshipit-source-id: 2da19d66d
5 years ago
codetoanalyze/java/performance/IntTest.java, IntTest.intValue_linear(java.lang.Integer):void, 0, EXPENSIVE_EXECUTION_TIME, no_bucket, ERROR, [with estimated cost 2 + 5 ⋅ mKBytesToSend + 3 ⋅ (1+max(0, mKBytesToSend)), O(mKBytesToSend), degree = 1,{1+max(0, mKBytesToSend)},Loop at line 9,{mKBytesToSend},Loop at line 9]
[inferbo] Add models for Java's Integer Summary: Add models for Java's Integer class. Reviewed By: skcho Differential Revision: D16668118 fbshipit-source-id: 05b3c8736
5 years ago
codetoanalyze/java/performance/IntTest.java, IntTest.static_Integer_top():void, 0, INFINITE_EXECUTION_TIME, no_bucket, ERROR, [Unbounded loop,Loop at line 15]
[RFC][java] kill regexp-based line number fixer Summary: Old versions of sawja/javalib got the line numbers slightly wrong. The workaround was to do a regexp search in the source file for the right line. My understanding is that this is no longer necessary. This diff removes it. Reviewed By: jvillard Differential Revision: D19033415 fbshipit-source-id: 2da19d66d
5 years ago
codetoanalyze/java/performance/IntTest.java, IntTest.static_Integer_top():void, 0, INTEGER_OVERFLOW_L5, no_bucket, ERROR, [<LHS trace>,Assignment,Binary operation: ([0, +oo] + 1):signed32]
codetoanalyze/java/performance/IntTest.java, IntTest.valueOf_linear(int):void, 1, EXPENSIVE_EXECUTION_TIME, no_bucket, ERROR, [with estimated cost 5 + 5 ⋅ p + 3 ⋅ (1+max(0, p)), O(p), degree = 1,{1+max(0, p)},Loop at line 20,{p},Loop at line 20]
codetoanalyze/java/performance/Invariant.java, Invariant.do_while_invariant(int,int):void, 2, EXPENSIVE_EXECUTION_TIME, no_bucket, ERROR, [with estimated cost 2 + 3 ⋅ (k - 1) + 4 ⋅ (max(1, k)), O(k), degree = 1,{max(1, k)},Loop at line 58,{k - 1},Loop at line 58]
[cost] Refactor cost issue types and enable detecting allocation complexity increase on cold start Summary: - Add allocation costs to `costs-report.json` and enable diffing over allocation costs. - Also, let's be more consistent and modular in naming our cost issues. - introduce a generic issue type `X_TIME_COMPLEXITY_INCREASE` where `X` can be one of the cost kinds. If the function is on the cold start, issue can have the `COLD_START` suffix. Similarly for infinite/zero/expensive calls. - Change `PERFORMANCE_VARIATION` -> `EXECUTION_TIME_COMPLEXITY_INCREASE` - Add new issue type for `ALLOCATION_COMPLEXITY_INCREASE_COLD_START` which will be enabled by default - Refactor cost issues to be more modular and succinct. This also makes addition of a new cost kind very easy by adding the kind into the `enabled_cost_kinds` list in `CostKind.ml` Reviewed By: mbouaziz Differential Revision: D15822681 fbshipit-source-id: cf89ece59
6 years ago
codetoanalyze/java/performance/Invariant.java, Invariant.formal_not_invariant_FP(int,int):void, 0, INFINITE_EXECUTION_TIME, no_bucket, ERROR, [Unbounded loop,Loop at line 31]
[RFC][java] kill regexp-based line number fixer Summary: Old versions of sawja/javalib got the line numbers slightly wrong. The workaround was to do a regexp search in the source file for the right line. My understanding is that this is no longer necessary. This diff removes it. Reviewed By: jvillard Differential Revision: D19033415 fbshipit-source-id: 2da19d66d
5 years ago
codetoanalyze/java/performance/Invariant.java, Invariant.formal_not_invariant_FP(int,int):void, 1, INTEGER_OVERFLOW_L5, no_bucket, ERROR, [<LHS trace>,Assignment,<RHS trace>,Assignment,Binary operation: (size + [min(0, x), max(0, x)]):signed32]
codetoanalyze/java/performance/Invariant.java, Invariant.formal_not_invariant_FP(int,int):void, 5, INTEGER_OVERFLOW_L5, no_bucket, ERROR, [<LHS trace>,Assignment,Binary operation: ([0, +oo] + 1):signed32]
codetoanalyze/java/performance/Invariant.java, Invariant.list_size_invariant(java.util.List):void, 0, EXPENSIVE_EXECUTION_TIME, no_bucket, ERROR, [with estimated cost 2 + 5 ⋅ items.length + 3 ⋅ (items.length + 1), O(items.length), degree = 1,{items.length + 1},Loop at line 66,{items.length},Loop at line 66]
codetoanalyze/java/performance/Invariant.java, Invariant.local_not_invariant_FP(int):void, 2, EXPENSIVE_EXECUTION_TIME, no_bucket, ERROR, [with estimated cost 3 + 12 ⋅ (size + 5) + 7 ⋅ (size + 5) × (5+min(1, size)) + 4 ⋅ (5+min(0, size)), O(size²), degree = 2,{5+min(0, size)},Loop at line 46,{5+min(1, size)},Loop at line 46,{size + 5},Loop at line 46]
codetoanalyze/java/performance/Invariant.java, Invariant.x_is_invariant_ok(int):void, 6, EXPENSIVE_EXECUTION_TIME, no_bucket, ERROR, [with estimated cost 8 + 6 ⋅ (size + 20), O(size), degree = 1,{size + 20},Loop at line 19]
codetoanalyze/java/performance/IteratorTest.java, IteratorTest.appendTo(java.util.Iterator):void, 0, EXPENSIVE_EXECUTION_TIME, no_bucket, ERROR, [with estimated cost 1 + 10 ⋅ parts.length + 3 ⋅ (parts.length + 1), O(parts.length), degree = 1,{parts.length + 1},Loop at line 12,{parts.length},Loop at line 12]
codetoanalyze/java/performance/IteratorTest.java, IteratorTest.linearIterable(java.lang.Iterable):void, 0, EXPENSIVE_EXECUTION_TIME, no_bucket, ERROR, [with estimated cost 9 + 10 ⋅ elements.length + 3 ⋅ (elements.length + 1), O(elements.length), degree = 1,{elements.length + 1},call to void IteratorTest.appendTo(Iterator),Loop at line 12,{elements.length},call to void IteratorTest.appendTo(Iterator),Loop at line 12]
[cost] Non-symbolic value for unknown external function call Summary: This diff returns non-symbolic value (top) for unknown external function calls because the symbolic values sometimes make it hard to understand costs. Reviewed By: ezgicicek Differential Revision: D18685715 fbshipit-source-id: 1b39c718b
5 years ago
codetoanalyze/java/performance/JsonArray.java, libraries.marauder.analytics.utils.json.JsonArray.addStringEntry(java.lang.String):void, 0, INFINITE_EXECUTION_TIME, no_bucket, ERROR, [Call to void JsonUtils.serialize(StringBuilder,String),Call to void JsonUtils.escape(StringBuilder,String),Unbounded loop,Loop at line 13]
codetoanalyze/java/performance/JsonMap.java, libraries.marauder.analytics.utils.json.JsonMap.addEntry(java.lang.String,boolean):void, 0, INFINITE_EXECUTION_TIME, no_bucket, ERROR, [Call to void JsonMap.addKeyToMap(String),Call to void JsonUtils.serialize(StringBuilder,String),Call to void JsonUtils.escape(StringBuilder,String),Unbounded loop,Loop at line 13]
codetoanalyze/java/performance/JsonMap.java, libraries.marauder.analytics.utils.json.JsonMap.addEntry(java.lang.String,double):void, 0, INFINITE_EXECUTION_TIME, no_bucket, ERROR, [Call to void JsonMap.addKeyToMap(String),Call to void JsonUtils.serialize(StringBuilder,String),Call to void JsonUtils.escape(StringBuilder,String),Unbounded loop,Loop at line 13]
codetoanalyze/java/performance/JsonMap.java, libraries.marauder.analytics.utils.json.JsonMap.addEntry(java.lang.String,java.lang.Object):void, 0, INFINITE_EXECUTION_TIME, no_bucket, ERROR, [Call to void JsonMap.addKeyToMap(String),Call to void JsonUtils.serialize(StringBuilder,String),Call to void JsonUtils.escape(StringBuilder,String),Unbounded loop,Loop at line 13]
codetoanalyze/java/performance/JsonMap.java, libraries.marauder.analytics.utils.json.JsonMap.addEntry(java.lang.String,java.lang.String):void, 0, INFINITE_EXECUTION_TIME, no_bucket, ERROR, [Call to void JsonUtils.serialize(StringBuilder,String),Call to void JsonUtils.escape(StringBuilder,String),Unbounded loop,Loop at line 13]
codetoanalyze/java/performance/JsonMap.java, libraries.marauder.analytics.utils.json.JsonMap.addEntry(java.lang.String,libraries.marauder.analytics.utils.json.JsonType):void, 0, INFINITE_EXECUTION_TIME, no_bucket, ERROR, [Call to void JsonMap.addKeyToMap(String),Call to void JsonUtils.serialize(StringBuilder,String),Call to void JsonUtils.escape(StringBuilder,String),Unbounded loop,Loop at line 13]
codetoanalyze/java/performance/JsonMap.java, libraries.marauder.analytics.utils.json.JsonMap.addEntry(java.lang.String,long):void, 0, INFINITE_EXECUTION_TIME, no_bucket, ERROR, [Call to void JsonMap.addKeyToMap(String),Call to void JsonUtils.serialize(StringBuilder,String),Call to void JsonUtils.escape(StringBuilder,String),Unbounded loop,Loop at line 13]
codetoanalyze/java/performance/JsonMap.java, libraries.marauder.analytics.utils.json.JsonMap.addKeyToMap(java.lang.String):void, 0, INFINITE_EXECUTION_TIME, no_bucket, ERROR, [Call to void JsonUtils.serialize(StringBuilder,String),Call to void JsonUtils.escape(StringBuilder,String),Unbounded loop,Loop at line 13]
codetoanalyze/java/performance/JsonString.java, libraries.marauder.analytics.utils.json.JsonString.<init>(java.lang.String), 0, INFINITE_EXECUTION_TIME, no_bucket, ERROR, [Call to StringBuilder JsonUtils.serialize(String),Call to void JsonUtils.serialize(StringBuilder,String),Call to void JsonUtils.escape(StringBuilder,String),Unbounded loop,Loop at line 13]
codetoanalyze/java/performance/JsonUtils.java, libraries.marauder.analytics.utils.json.JsonUtils.escape(java.lang.StringBuilder,java.lang.String):void, 0, BUFFER_OVERRUN_U5, no_bucket, ERROR, [<Offset trace>,Assignment,<Length trace>,Unknown value from: char[] String.toCharArray(),Array access: Offset: [-oo, +oo] (⇐ [-oo, +oo] + [0, +oo]) Size: [0, +oo]]
codetoanalyze/java/performance/JsonUtils.java, libraries.marauder.analytics.utils.json.JsonUtils.escape(java.lang.StringBuilder,java.lang.String):void, 0, INFINITE_EXECUTION_TIME, no_bucket, ERROR, [Unbounded loop,Loop at line 13]
codetoanalyze/java/performance/JsonUtils.java, libraries.marauder.analytics.utils.json.JsonUtils.escape(java.lang.StringBuilder,java.lang.String):void, 0, INTEGER_OVERFLOW_L5, no_bucket, ERROR, [<LHS trace>,Assignment,Binary operation: ([0, +oo] + 1):signed32]
codetoanalyze/java/performance/JsonUtils.java, libraries.marauder.analytics.utils.json.JsonUtils.serialize(java.lang.String):java.lang.StringBuilder, 0, INFINITE_EXECUTION_TIME, no_bucket, ERROR, [Call to void JsonUtils.serialize(StringBuilder,String),Call to void JsonUtils.escape(StringBuilder,String),Unbounded loop,Loop at line 13]
codetoanalyze/java/performance/JsonUtils.java, libraries.marauder.analytics.utils.json.JsonUtils.serialize(java.lang.StringBuilder,java.lang.String):void, 0, INFINITE_EXECUTION_TIME, no_bucket, ERROR, [Call to void JsonUtils.escape(StringBuilder,String),Unbounded loop,Loop at line 13]
[RFC][java] kill regexp-based line number fixer Summary: Old versions of sawja/javalib got the line numbers slightly wrong. The workaround was to do a regexp search in the source file for the right line. My understanding is that this is no longer necessary. This diff removes it. Reviewed By: jvillard Differential Revision: D19033415 fbshipit-source-id: 2da19d66d
5 years ago
codetoanalyze/java/performance/ListTest.java, ListTest.asList_linear(java.lang.String[]):void, 1, EXPENSIVE_EXECUTION_TIME, no_bucket, ERROR, [with estimated cost 7 + 8 ⋅ array.length + 3 ⋅ (array.length + 1), O(array.length), degree = 1,{array.length + 1},Loop at line 42,{array.length},Loop at line 42]
codetoanalyze/java/performance/ListTest.java, ListTest.call_iterate_elements_linear(java.util.List,java.util.List):void, 0, EXPENSIVE_EXECUTION_TIME, no_bucket, ERROR, [with estimated cost 26 + 5 ⋅ (l2.length + l1.length) + 3 ⋅ (l2.length + l1.length + 1), O((l2.length + l1.length)), degree = 1,{l2.length + l1.length + 1},call to void ListTest.iterate_elements_linear(List),Loop at line 59,{l2.length + l1.length},call to void ListTest.iterate_elements_linear(List),Loop at line 59]
codetoanalyze/java/performance/ListTest.java, ListTest.indexOfImpl_linear(java.util.List,java.lang.Object):int, 1, EXPENSIVE_EXECUTION_TIME, no_bucket, ERROR, [with estimated cost 4 + 11 ⋅ list.length + 3 ⋅ (list.length + 1), O(list.length), degree = 1,{list.length + 1},Loop at line 18,{list.length},Loop at line 18]
codetoanalyze/java/performance/ListTest.java, ListTest.iter_multiple_list1_linear(java.util.List,java.util.List):void, 6, EXPENSIVE_EXECUTION_TIME, no_bucket, ERROR, [with estimated cost 17 + 12 ⋅ (l2.length + l1.length) + 3 ⋅ (l2.length + l1.length + 1), O((l2.length + l1.length)), degree = 1,{l2.length + l1.length + 1},Loop at line 76,{l2.length + l1.length},Loop at line 76]
codetoanalyze/java/performance/ListTest.java, ListTest.iter_multiple_list2_linear(java.util.List,java.util.List):void, 5, EXPENSIVE_EXECUTION_TIME, no_bucket, ERROR, [with estimated cost 14 + 8 ⋅ (l2.length + l1.length) + 3 ⋅ (l2.length + l1.length + 1), O((l2.length + l1.length)), degree = 1,{l2.length + l1.length + 1},Loop at line 88,{l2.length + l1.length},Loop at line 88]
codetoanalyze/java/performance/ListTest.java, ListTest.iter_multiple_list3_linear(java.util.List,java.util.List,java.util.List):void, 6, EXPENSIVE_EXECUTION_TIME, no_bucket, ERROR, [with estimated cost 18 + 8 ⋅ (l2.length + l1.length + a.length) + 3 ⋅ (l2.length + l1.length + a.length + 1), O((l2.length + l1.length + a.length)), degree = 1,{l2.length + l1.length + a.length + 1},Loop at line 99,{l2.length + l1.length + a.length},Loop at line 99]
[inferbo] Update relation between iterator and integer value on Call Summary: This diff updates the relation between iterator (offset) and integer value not only at assignments (`x += 1`), but also at function calls (`foo()`) that increase integer values by one in their side effects. Reviewed By: ezgicicek Differential Revision: D19163214 fbshipit-source-id: 47e52f939
5 years ago
codetoanalyze/java/performance/ListTest.java, ListTest.iter_my_own_obj(java.util.List):void, 1, EXPENSIVE_EXECUTION_TIME, no_bucket, ERROR, [with estimated cost 15 + 14 ⋅ a.length + 3 ⋅ (a.length + 1), O(a.length), degree = 1,{a.length + 1},Loop at line 120,{a.length},Loop at line 120]
[inferbo] Add relation between iterator and integer value Summary: This diff extends the domain to express the relation between iterator's offset and integer value. Reviewed By: ezgicicek Differential Revision: D19143670 fbshipit-source-id: 6223bc934
5 years ago
codetoanalyze/java/performance/ListTest.java, ListTest.iter_relation_with_var(java.util.List):void, 1, EXPENSIVE_EXECUTION_TIME, no_bucket, ERROR, [with estimated cost 5 + 10 ⋅ a.length + 3 ⋅ (a.length + 1), O(a.length), degree = 1,{a.length + 1},Loop at line 104,{a.length},Loop at line 104]
[RFC][java] kill regexp-based line number fixer Summary: Old versions of sawja/javalib got the line numbers slightly wrong. The workaround was to do a regexp search in the source file for the right line. My understanding is that this is no longer necessary. This diff removes it. Reviewed By: jvillard Differential Revision: D19033415 fbshipit-source-id: 2da19d66d
5 years ago
codetoanalyze/java/performance/ListTest.java, ListTest.iterate_elements_linear(java.util.List):void, 1, EXPENSIVE_EXECUTION_TIME, no_bucket, ERROR, [with estimated cost 4 + 5 ⋅ l.length + 3 ⋅ (l.length + 1), O(l.length), degree = 1,{l.length + 1},Loop at line 59,{l.length},Loop at line 59]
codetoanalyze/java/performance/ListTest.java, ListTest.sort_comparator_nlogn(java.util.List):void, 0, EXPENSIVE_EXECUTION_TIME, no_bucket, ERROR, [with estimated cost 8 + people.length × log(people.length), O(people.length × log(people.length)), degree = 1 + 1⋅log,{people.length},Modeled call to List.sort,{people.length},Modeled call to List.sort]
codetoanalyze/java/performance/ListTest.java, ListTest.sublist(java.util.List):void, 0, EXPENSIVE_EXECUTION_TIME, no_bucket, ERROR, [with estimated cost 11 + 8 ⋅ (filesList.length - 1) + 3 ⋅ filesList.length, O(filesList.length), degree = 1,{filesList.length},Loop at line 32,{filesList.length - 1},Loop at line 32]
codetoanalyze/java/performance/Loops.java, codetoanalyze.java.performance.Loops.charsequence_length_linear(java.lang.CharSequence):void, 0, EXPENSIVE_EXECUTION_TIME, no_bucket, ERROR, [with estimated cost 2 + 5 ⋅ seq.length + 3 ⋅ (seq.length + 1), O(seq.length), degree = 1,{seq.length + 1},Loop at line 115,{seq.length},Loop at line 115]
codetoanalyze/java/performance/Loops.java, codetoanalyze.java.performance.Loops.do_while_independent_of_p(int):int, 2, EXPENSIVE_EXECUTION_TIME, no_bucket, ERROR, [with estimated cost 250, O(1), degree = 0]
codetoanalyze/java/performance/Loops.java, codetoanalyze.java.performance.Loops.dumb0(long[],int):void, 0, EXPENSIVE_EXECUTION_TIME, no_bucket, ERROR, [with estimated cost 2 + 25 ⋅ (length - 1), O(length), degree = 1,{length - 1},Loop at line 44]
codetoanalyze/java/performance/Loops.java, codetoanalyze.java.performance.Loops.dumbSort(long[],long[],int):void, 0, EXPENSIVE_EXECUTION_TIME, no_bucket, ERROR, [with estimated cost 3 + 59 ⋅ (length - 1) × (length - 1) + 8 ⋅ length, O(length × length), degree = 2,{length},Loop at line 54,{length - 1},Loop at line 55,{length - 1},Loop at line 54]
[cost] Set the minimum basic cost of node Summary: The zero cost of node does not make sense especially when the abstract memory is non-bottom. This resulted in unreasonable zero cost results sometimes, e.g. when the checker could not find appropriate control varaibles having interval values of iteration. This diff fixes this, so sets the minimum basic cost as 1, if the abstract memory at the node is non-bottom. Reviewed By: ezgicicek Differential Revision: D18199291 fbshipit-source-id: b215d10e5
5 years ago
codetoanalyze/java/performance/Loops.java, codetoanalyze.java.performance.Loops.length_of_linked_list_linear_FP(codetoanalyze.java.performance.Loops$MyLinkedList):void, 0, INFINITE_EXECUTION_TIME, no_bucket, ERROR, [Unbounded value x,call to void Loops.linear(int),Loop at line 86]
[RFC][java] kill regexp-based line number fixer Summary: Old versions of sawja/javalib got the line numbers slightly wrong. The workaround was to do a regexp search in the source file for the right line. My understanding is that this is no longer necessary. This diff removes it. Reviewed By: jvillard Differential Revision: D19033415 fbshipit-source-id: 2da19d66d
5 years ago
codetoanalyze/java/performance/Loops.java, codetoanalyze.java.performance.Loops.length_of_linked_list_linear_FP(codetoanalyze.java.performance.Loops$MyLinkedList):void, 2, INTEGER_OVERFLOW_L5, no_bucket, ERROR, [<LHS trace>,Assignment,Binary operation: ([0, +oo] + 1):signed32]
codetoanalyze/java/performance/Loops.java, codetoanalyze.java.performance.Loops.linear(int):void, 0, EXPENSIVE_EXECUTION_TIME, no_bucket, ERROR, [with estimated cost 2 + 5 ⋅ x, O(x), degree = 1,{x},Loop at line 86]
codetoanalyze/java/performance/Loops.java, codetoanalyze.java.performance.Loops.modeled_range_linear(java.nio.channels.FileChannel,java.nio.ByteBuffer):void, 2, EXPENSIVE_EXECUTION_TIME, no_bucket, ERROR, [with estimated cost 2 + 14 ⋅ FileChannel.read(...).modeled, O(FileChannel.read(...).modeled), degree = 1,{FileChannel.read(...).modeled},Modeled call to read(...)]
codetoanalyze/java/performance/Loops.java, codetoanalyze.java.performance.Loops.modeled_range_linear(java.nio.channels.FileChannel,java.nio.ByteBuffer):void, 7, INTEGER_OVERFLOW_L5, no_bucket, ERROR, [<LHS trace>,Assignment,Binary operation: ([0, +oo] + 8):signed32]
[cost] Add modeled range Summary: `ModeledRange` represents how many times the interval value can be updated by modeled functions. This domain is to support the case where there are mismatches between value of a control variable and actual number of loop iterations. For example, ``` while((c = file_channel.read(buf)) != -1) { ... } ``` the loop will iterates as the file size, but the control variable `c` does not have that value. In these cases, it assigns a symbolic value of the file size to the modeled range of `c`, then which is used when calculating the overall cost. Reviewed By: jvillard Differential Revision: D17476621 fbshipit-source-id: 9a81376e8
5 years ago
codetoanalyze/java/performance/Loops.java, codetoanalyze.java.performance.Loops.nested_do_while_FP(int):void, 0, INFINITE_EXECUTION_TIME, no_bucket, ERROR, [Unbounded loop,Loop at line 34]
[RFC][java] kill regexp-based line number fixer Summary: Old versions of sawja/javalib got the line numbers slightly wrong. The workaround was to do a regexp search in the source file for the right line. My understanding is that this is no longer necessary. This diff removes it. Reviewed By: jvillard Differential Revision: D19033415 fbshipit-source-id: 2da19d66d
5 years ago
codetoanalyze/java/performance/Loops.java, codetoanalyze.java.performance.Loops.nested_do_while_FP(int):void, 7, INTEGER_OVERFLOW_L5, no_bucket, ERROR, [<LHS trace>,Assignment,Binary operation: ([0, +oo] + 1):signed32]
codetoanalyze/java/performance/Loops.java, codetoanalyze.java.performance.Loops.similar(codetoanalyze.java.performance.Loops$C[],codetoanalyze.java.performance.Loops$C[]):boolean, 3, EXPENSIVE_EXECUTION_TIME, no_bucket, ERROR, [with estimated cost 16 + 26 ⋅ x.length, O(x.length), degree = 1,{x.length},Loop at line 77]
codetoanalyze/java/performance/Loops.java, codetoanalyze.java.performance.Loops.string_concat_linear(java.lang.String,java.lang.String):void, 1, EXPENSIVE_EXECUTION_TIME, no_bucket, ERROR, [with estimated cost 8 + 5 ⋅ (p.length + s.length) + 3 ⋅ (p.length + s.length + 1), O((p.length + s.length)), degree = 1,{p.length + s.length + 1},Loop at line 103,{p.length + s.length},Loop at line 103]
codetoanalyze/java/performance/Loops.java, codetoanalyze.java.performance.Loops.string_length_linear(java.lang.String):void, 0, EXPENSIVE_EXECUTION_TIME, no_bucket, ERROR, [with estimated cost 2 + 5 ⋅ s.length + 3 ⋅ (s.length + 1), O(s.length), degree = 1,{s.length + 1},Loop at line 98,{s.length},Loop at line 98]
[cost] Add modeled range Summary: `ModeledRange` represents how many times the interval value can be updated by modeled functions. This domain is to support the case where there are mismatches between value of a control variable and actual number of loop iterations. For example, ``` while((c = file_channel.read(buf)) != -1) { ... } ``` the loop will iterates as the file size, but the control variable `c` does not have that value. In these cases, it assigns a symbolic value of the file size to the modeled range of `c`, then which is used when calculating the overall cost. Reviewed By: jvillard Differential Revision: D17476621 fbshipit-source-id: 9a81376e8
5 years ago
codetoanalyze/java/performance/Loops.java, codetoanalyze.java.performance.Loops.unboundedSymbol():void, 0, INFINITE_EXECUTION_TIME, no_bucket, ERROR, [Unbounded value x,call to void Loops.linear(int),Loop at line 86]
[RFC][java] kill regexp-based line number fixer Summary: Old versions of sawja/javalib got the line numbers slightly wrong. The workaround was to do a regexp search in the source file for the right line. My understanding is that this is no longer necessary. This diff removes it. Reviewed By: jvillard Differential Revision: D19033415 fbshipit-source-id: 2da19d66d
5 years ago
codetoanalyze/java/performance/Loops.java, codetoanalyze.java.performance.Loops.unboundedSymbol():void, 2, INTEGER_OVERFLOW_L5, no_bucket, ERROR, [<LHS trace>,Assignment,<RHS trace>,Assignment,Binary operation: ([-oo, +oo] × [-oo, +oo]):signed32]
codetoanalyze/java/performance/Loops.java, codetoanalyze.java.performance.Loops.unboundedSymbol():void, 4, INTEGER_OVERFLOW_L5, no_bucket, ERROR, [Assignment,Call,<LHS trace>,Assignment,Binary operation: ([0, +oo] + 1):signed32 by call to `void Loops.linear(int)` ]
codetoanalyze/java/performance/MapTest.java, MapTest.entrySet_linear(java.util.Map):void, 0, EXPENSIVE_EXECUTION_TIME, no_bucket, ERROR, [with estimated cost 7 + 8 ⋅ map.length + 3 ⋅ (map.length + 1), O(map.length), degree = 1,{map.length + 1},Loop at line 17,{map.length},Loop at line 17]
codetoanalyze/java/performance/MapTest.java, MapTest.keySet_linear(java.util.Map):void, 0, EXPENSIVE_EXECUTION_TIME, no_bucket, ERROR, [with estimated cost 7 + 8 ⋅ map.length + 3 ⋅ (map.length + 1), O(map.length), degree = 1,{map.length + 1},Loop at line 13,{map.length},Loop at line 13]
codetoanalyze/java/performance/MapTest.java, MapTest.putAll_linear(java.util.Map):void, 2, EXPENSIVE_EXECUTION_TIME, no_bucket, ERROR, [with estimated cost 13 + 8 ⋅ map.length + 3 ⋅ (map.length + 1), O(map.length), degree = 1,{map.length + 1},Loop at line 28,{map.length},Loop at line 28]
codetoanalyze/java/performance/MapTest.java, MapTest.values_linear(java.util.Map):void, 1, EXPENSIVE_EXECUTION_TIME, no_bucket, ERROR, [with estimated cost 13 + 8 ⋅ (map.length + 1) + 3 ⋅ (map.length + 2), O(map.length), degree = 1,{map.length + 2},Loop at line 22,{map.length + 1},Loop at line 22]
codetoanalyze/java/performance/MathTest.java, codetoanalyze.java.performance.MathTest.call_with_max_linear(int):void, 0, EXPENSIVE_EXECUTION_TIME, no_bucket, ERROR, [with estimated cost 11 + 5 ⋅ (max(1, x)), O(x), degree = 1,{max(1, x)},call to void MathTest.linear(int),Loop at line 28]
codetoanalyze/java/performance/MathTest.java, codetoanalyze.java.performance.MathTest.linear(int):void, 0, EXPENSIVE_EXECUTION_TIME, no_bucket, ERROR, [with estimated cost 2 + 5 ⋅ p, O(p), degree = 1,{p},Loop at line 28]
codetoanalyze/java/performance/MathTest.java, codetoanalyze.java.performance.MathTest.max2_symbolic(int,int):void, 0, EXPENSIVE_EXECUTION_TIME, no_bucket, ERROR, [with estimated cost 6 + 9 ⋅ (max(x, y)), O((max(x, y))), degree = 1,{max(x, y)},Loop at line 20]
codetoanalyze/java/performance/MathTest.java, codetoanalyze.java.performance.MathTest.max_symbolic(int[]):void, 0, EXPENSIVE_EXECUTION_TIME, no_bucket, ERROR, [with estimated cost 2 + 5 ⋅ arr.length + 4 ⋅ (arr.length + 1), O(arr.length), degree = 1,{arr.length + 1},Loop at line 16,{arr.length},Loop at line 16]
[inferbo] Add models for Preconditions.checkNotNull and Preconditions.checkState Reviewed By: ngorogiannis Differential Revision: D19346368 fbshipit-source-id: 643327802
5 years ago
codetoanalyze/java/performance/PreconditionTest.java, PreconditionTest.checkNotNull_linear(java.util.ArrayList,java.lang.Object):void, 1, EXPENSIVE_EXECUTION_TIME, no_bucket, ERROR, [with estimated cost 11 + 8 ⋅ list.length + 3 ⋅ (list.length + 1), O(list.length), degree = 1,{list.length + 1},Loop at line 37,{list.length},Loop at line 37]
[inferbo][cost] Generalize String models and add models for StringBuilder Summary: `String` and `StringBuilder` both implement `CharSequence`. Let's generalize the model for `String` to `CharSequence` wherever possible and add missing models for - `StringBuilder.append` - `StringBuilder.toString` Reviewed By: skcho Differential Revision: D19558009 fbshipit-source-id: 0dfdb21af
5 years ago
codetoanalyze/java/performance/StringBuilderTest.java, StringBuilderTest.append_linear(java.lang.String):void, 1, EXPENSIVE_EXECUTION_TIME, no_bucket, ERROR, [with estimated cost 25 + 5 ⋅ (s.length + 2) + 3 ⋅ (s.length + 3), O(s.length), degree = 1,{s.length + 3},call to void StringBuilderTest.new_linear(String),Loop at line 13,{s.length + 2},call to void StringBuilderTest.new_linear(String),Loop at line 13]
codetoanalyze/java/performance/StringBuilderTest.java, StringBuilderTest.new_linear(java.lang.String):void, 1, EXPENSIVE_EXECUTION_TIME, no_bucket, ERROR, [with estimated cost 9 + 5 ⋅ s.length + 3 ⋅ (s.length + 1), O(s.length), degree = 1,{s.length + 1},Loop at line 13,{s.length},Loop at line 13]
[inferbo] Revise Java's String constructor models Summary: Java's String models were broken for - initializing a String object with a locally defined constant string (which is an `Object*` in SIL). - initializing a String object with a `char`/`byte` array This diff fixes them and also adds models for `new String ()`. Reviewed By: skcho Differential Revision: D19662180 fbshipit-source-id: 23968d0aa
5 years ago
codetoanalyze/java/performance/StringTest.java, StringTest.byte_array_constructor_linear(byte[]):void, 1, EXPENSIVE_EXECUTION_TIME, no_bucket, ERROR, [with estimated cost 6 + 5 ⋅ data.length + 3 ⋅ (data.length + 1), O(data.length), degree = 1,{data.length + 1},Loop at line 55,{data.length},Loop at line 55]
[RFC][java] kill regexp-based line number fixer Summary: Old versions of sawja/javalib got the line numbers slightly wrong. The workaround was to do a regexp search in the source file for the right line. My understanding is that this is no longer necessary. This diff removes it. Reviewed By: jvillard Differential Revision: D19033415 fbshipit-source-id: 2da19d66d
5 years ago
codetoanalyze/java/performance/StringTest.java, StringTest.index_substring_linear():java.lang.String, 0, EXPENSIVE_EXECUTION_TIME, no_bucket, ERROR, [with estimated cost 9 + this.mId.length, O(this.mId.length), degree = 1,{this.mId.length},call to int StringTest.indexof_linear(String),Modeled call to String.indexOf]
codetoanalyze/java/performance/StringTest.java, StringTest.indexof_from_linear(java.lang.String,int):int, 0, EXPENSIVE_EXECUTION_TIME, no_bucket, ERROR, [with estimated cost 3 + (-j + m.length), O((-j + m.length)), degree = 1,{-j + m.length},Modeled call to String.indexOf]
codetoanalyze/java/performance/StringTest.java, StringTest.indexof_linear(java.lang.String):int, 0, EXPENSIVE_EXECUTION_TIME, no_bucket, ERROR, [with estimated cost 2 + m.length, O(m.length), degree = 1,{m.length},Modeled call to String.indexOf]
codetoanalyze/java/performance/StringTest.java, StringTest.indexof_quadratic(java.lang.String,java.lang.String):int, 0, EXPENSIVE_EXECUTION_TIME, no_bucket, ERROR, [with estimated cost 3 + m.length × n.length, O(m.length × n.length), degree = 2,{n.length},Modeled call to String.indexOf,{m.length},Modeled call to String.indexOf]
[inferbo] Revise String.split's bounds Summary: `String.split(regexp)` returns an array that is split by the given regexp. If the regexp doesn't match, the original string is returned. Hence, the resulting array's length must be in `[1, max(1, n_u -1)]` where`n_u` is the upper bound of the string's length. Reviewed By: dulmarod Differential Revision: D19578318 fbshipit-source-id: 675af7376
5 years ago
codetoanalyze/java/performance/StringTest.java, StringTest.split_linear(java.lang.String):void, 1, EXPENSIVE_EXECUTION_TIME, no_bucket, ERROR, [with estimated cost 5 + 6 ⋅ (-1+max(2, s.length)), O(s.length), degree = 1,{-1+max(2, s.length)},Loop at line 40]
codetoanalyze/java/performance/StringTest.java, StringTest.split_with_limit_linear(java.lang.String,int):void, 1, EXPENSIVE_EXECUTION_TIME, no_bucket, ERROR, [with estimated cost 6 + 6 ⋅ (max(1, limit)), O(limit), degree = 1,{max(1, limit)},Loop at line 45]
[inferbo] Revise Java's String constructor models Summary: Java's String models were broken for - initializing a String object with a locally defined constant string (which is an `Object*` in SIL). - initializing a String object with a `char`/`byte` array This diff fixes them and also adds models for `new String ()`. Reviewed By: skcho Differential Revision: D19662180 fbshipit-source-id: 23968d0aa
5 years ago
codetoanalyze/java/performance/StringTest.java, StringTest.string_constructor_linear(java.lang.String):void, 1, EXPENSIVE_EXECUTION_TIME, no_bucket, ERROR, [with estimated cost 6 + 5 ⋅ s.length + 3 ⋅ (s.length + 1), O(s.length), degree = 1,{s.length + 1},Loop at line 71,{s.length},Loop at line 71]
[inferbo] Add model for String.substring Reviewed By: skcho Differential Revision: D19663254 fbshipit-source-id: 29b711e41
5 years ago
codetoanalyze/java/performance/StringTest.java, StringTest.substring_linear(java.lang.String,int,int):void, 1, EXPENSIVE_EXECUTION_TIME, no_bucket, ERROR, [with estimated cost 10 + 8 ⋅ (-x + y), O((-x + y)), degree = 1,{-x + y},Loop at line 86]
codetoanalyze/java/performance/StringTest.java, StringTest.substring_no_end_linear(java.lang.String,int):void, 1, EXPENSIVE_EXECUTION_TIME, no_bucket, ERROR, [with estimated cost 9 + 8 ⋅ (-x + s.length), O((-x + s.length)), degree = 1,{-x + s.length},Loop at line 81]
[RFC][java] kill regexp-based line number fixer Summary: Old versions of sawja/javalib got the line numbers slightly wrong. The workaround was to do a regexp search in the source file for the right line. My understanding is that this is no longer necessary. This diff removes it. Reviewed By: jvillard Differential Revision: D19033415 fbshipit-source-id: 2da19d66d
5 years ago
codetoanalyze/java/performance/Switch.java, codetoanalyze.java.performance.Switch.test_switch():int, 2, CONDITION_ALWAYS_TRUE, no_bucket, WARNING, [Here]
codetoanalyze/java/performance/Switch.java, codetoanalyze.java.performance.Switch.test_switch():int, 2, EXPENSIVE_EXECUTION_TIME, no_bucket, ERROR, [with estimated cost 798, O(1), degree = 0]
codetoanalyze/java/performance/Switch.java, codetoanalyze.java.performance.Switch.vanilla_switch(int):void, 0, CONDITION_ALWAYS_TRUE, no_bucket, WARNING, [Here]
codetoanalyze/java/performance/UIAnnotationTest.java, codetoanalyze.java.checkers.AllMethodsOnUiThread.bar_UIThread_linear():int, 0, EXPENSIVE_EXECUTION_TIME_UI_THREAD, no_bucket, ERROR, [with estimated cost 2 + 10 ⋅ (max(0, this.f)), O(this.f), degree = 1,{max(0, this.f)},Loop at line 37]
[cost] Introduce cost issue types for functions on UI Thread Summary: Let's introduce a set of new cost analysis issue types that are raised when the function is statically determined to run on the UI thread. For this, we rely on the existing `runs_on_ui_thread` check that is developed for RacerD. We also update the cost summary and `jsonbug.cost_item` to include whether a method is on the ui thread so that we don't repeatedly compute this at diff time for complexity increase issues. Note that `*_UI_THREAD` cost issues are assumed to be more strict than `*_COLD_START` reports at the moment. Next, we can also consider adding a new issue type that combines both such as `*_UI_THREAD_AND_COLD_START` (i.e. for methods that are both on cold start and run on ui thread). Reviewed By: ngorogiannis Differential Revision: D18428408 fbshipit-source-id: f18805716
5 years ago
codetoanalyze/java/performance/UIAnnotationTest.java, codetoanalyze.java.checkers.ExtendsClassOnUiThread.bar_UIThread_linear():int, 0, EXPENSIVE_EXECUTION_TIME_UI_THREAD, no_bucket, ERROR, [with estimated cost 11 + 10 ⋅ (max(0, this.f)), O(this.f), degree = 1,{max(0, this.f)},call to int AllMethodsOnUiThread.bar_UIThread_linear(),Loop at line 37]
[RFC][java] kill regexp-based line number fixer Summary: Old versions of sawja/javalib got the line numbers slightly wrong. The workaround was to do a regexp search in the source file for the right line. My understanding is that this is no longer necessary. This diff removes it. Reviewed By: jvillard Differential Revision: D19033415 fbshipit-source-id: 2da19d66d
5 years ago
codetoanalyze/java/performance/UIAnnotationTest.java, codetoanalyze.java.checkers.UIAnnotationTest$Annotations.loop_UIThread_linear(int):void, 0, EXPENSIVE_EXECUTION_TIME_UI_THREAD, no_bucket, ERROR, [with estimated cost 2 + 5 ⋅ x, O(x), degree = 1,{x},Loop at line 76]
codetoanalyze/java/performance/UIAnnotationTest.java, codetoanalyze.java.checkers.UIAnnotationTest$Annotations.loop_linear(int):void, 0, EXPENSIVE_EXECUTION_TIME, no_bucket, ERROR, [with estimated cost 2 + 5 ⋅ x, O(x), degree = 1,{x},Loop at line 84]
codetoanalyze/java/performance/UIAnnotationTest.java, codetoanalyze.java.checkers.UIAnnotationTest$Annotations.onBindMethod_linear(int):void, 0, EXPENSIVE_EXECUTION_TIME_UI_THREAD, no_bucket, ERROR, [with estimated cost 8 + 5 ⋅ x, O(x), degree = 1,{x},call to void UIAnnotationTest$Annotations.loop_linear(int),Loop at line 84]
codetoanalyze/java/performance/UIAnnotationTest.java, codetoanalyze.java.checkers.UIAnnotationTest$Annotations.onClick_linear(int):void, 0, EXPENSIVE_EXECUTION_TIME_UI_THREAD, no_bucket, ERROR, [with estimated cost 2 + 5 ⋅ x, O(x), degree = 1,{x},Loop at line 90]
codetoanalyze/java/performance/UIAnnotationTest.java, codetoanalyze.java.checkers.UIAnnotationTest$WeirdAnnotation.foo_linear():void, 0, EXPENSIVE_EXECUTION_TIME, no_bucket, ERROR, [with estimated cost 2 + 6 ⋅ this.f, O(this.f), degree = 1,{this.f},Loop at line 68]
[cost] Non-symbolic value for unknown external function call Summary: This diff returns non-symbolic value (top) for unknown external function calls because the symbolic values sometimes make it hard to understand costs. Reviewed By: ezgicicek Differential Revision: D18685715 fbshipit-source-id: 1b39c718b
5 years ago
codetoanalyze/java/performance/UnknownCallsTest.java, UnknownCallsTest.call_concrete_func_linear_FP(UnknownCallsTest$AbstractC):void, 0, INFINITE_EXECUTION_TIME, no_bucket, ERROR, [Unbounded loop,Loop at line 94]
codetoanalyze/java/performance/UnknownCallsTest.java, UnknownCallsTest.call_concrete_func_linear_FP(UnknownCallsTest$AbstractC):void, 1, BUFFER_OVERRUN_U5, no_bucket, ERROR, [<Length trace>,Unknown value from: int[] UnknownCallsTest$AbstractC.abstract_func(),Assignment,Array access: Offset: [-oo, +oo] Size: [0, +oo]]
codetoanalyze/java/performance/UnknownCallsTest.java, UnknownCallsTest.call_concrete_func_linear_FP(UnknownCallsTest$AbstractC):void, 1, INTEGER_OVERFLOW_U5, no_bucket, ERROR, [<LHS trace>,Unknown value from: int[] UnknownCallsTest$AbstractC.abstract_func(),Binary operation: ([0, +oo] + 1):signed32]
codetoanalyze/java/performance/UnknownCallsTest.java, UnknownCallsTest.call_loop_over_charArray_FP(java.lang.StringBuilder,java.lang.String):void, 0, INFINITE_EXECUTION_TIME, no_bucket, ERROR, [Call to void UnknownCallsTest.loop_over_charArray_FP(StringBuilder,String),Unbounded loop,Loop at line 51]
[RFC][java] kill regexp-based line number fixer Summary: Old versions of sawja/javalib got the line numbers slightly wrong. The workaround was to do a regexp search in the source file for the right line. My understanding is that this is no longer necessary. This diff removes it. Reviewed By: jvillard Differential Revision: D19033415 fbshipit-source-id: 2da19d66d
5 years ago
codetoanalyze/java/performance/UnknownCallsTest.java, UnknownCallsTest.call_may_throw_exception_constant():void, 0, EXPENSIVE_EXECUTION_TIME, no_bucket, ERROR, [with estimated cost 217, O(1), degree = 0]
codetoanalyze/java/performance/UnknownCallsTest.java, UnknownCallsTest.call_throw_exception_linear():void, 0, EXPENSIVE_EXECUTION_TIME, no_bucket, ERROR, [with estimated cost 2 + 5 ⋅ UnknownCallsTest.throw_exception().ub + 8 ⋅ (1+max(0, UnknownCallsTest.throw_exception().ub)), O(UnknownCallsTest.throw_exception().ub), degree = 1,{1+max(0, UnknownCallsTest.throw_exception().ub)},Loop at line 71,{UnknownCallsTest.throw_exception().ub},Loop at line 71]
codetoanalyze/java/performance/UnknownCallsTest.java, UnknownCallsTest.jsonArray_linear(org.json.JSONArray):void, 1, EXPENSIVE_EXECUTION_TIME, no_bucket, ERROR, [with estimated cost 5 + 5 ⋅ jsonArray.length, O(jsonArray.length), degree = 1,{jsonArray.length},Loop at line 18]
[cost] Non-symbolic value for unknown external function call Summary: This diff returns non-symbolic value (top) for unknown external function calls because the symbolic values sometimes make it hard to understand costs. Reviewed By: ezgicicek Differential Revision: D18685715 fbshipit-source-id: 1b39c718b
5 years ago
codetoanalyze/java/performance/UnknownCallsTest.java, UnknownCallsTest.loop_over_charArray_FP(java.lang.StringBuilder,java.lang.String):void, 0, BUFFER_OVERRUN_U5, no_bucket, ERROR, [<Offset trace>,Assignment,<Length trace>,Unknown value from: char[] String.toCharArray(),Array access: Offset: [-oo, +oo] (⇐ [-oo, +oo] + [0, +oo]) Size: [0, +oo]]
codetoanalyze/java/performance/UnknownCallsTest.java, UnknownCallsTest.loop_over_charArray_FP(java.lang.StringBuilder,java.lang.String):void, 0, INFINITE_EXECUTION_TIME, no_bucket, ERROR, [Unbounded loop,Loop at line 51]
codetoanalyze/java/performance/UnknownCallsTest.java, UnknownCallsTest.loop_over_charArray_FP(java.lang.StringBuilder,java.lang.String):void, 0, INTEGER_OVERFLOW_L5, no_bucket, ERROR, [<LHS trace>,Assignment,Binary operation: ([0, +oo] + 1):signed32]
[RFC][java] kill regexp-based line number fixer Summary: Old versions of sawja/javalib got the line numbers slightly wrong. The workaround was to do a regexp search in the source file for the right line. My understanding is that this is no longer necessary. This diff removes it. Reviewed By: jvillard Differential Revision: D19033415 fbshipit-source-id: 2da19d66d
5 years ago
codetoanalyze/java/performance/UnknownCallsTest.java, UnknownCallsTest.may_throw_exception():int, 1, UNREACHABLE_CODE, no_bucket, ERROR, [Here]
codetoanalyze/java/performance/UnknownCallsTest.java, UnknownCallsTest.read_max_cost(java.io.InputStream,byte[],int,int,java.util.ArrayList):int, 5, EXPENSIVE_EXECUTION_TIME, no_bucket, ERROR, [with estimated cost 21 + 5 ⋅ (byteCount + 1), O(byteCount), degree = 1,{byteCount + 1},Loop at line 46]
codetoanalyze/java/performance/UnknownCallsTest.java, UnknownCallsTest.read_sum_cost(java.io.InputStream,byte[],int,int,java.util.ArrayList):int, 3, EXPENSIVE_EXECUTION_TIME, no_bucket, ERROR, [with estimated cost 15 + 6 ⋅ 2⋅byteCount, O(2⋅byteCount), degree = 1,{2⋅byteCount},Loop at line 33]
codetoanalyze/java/performance/UnknownCallsTest.java, UnknownCallsTest.throw_exception():int, 0, UNREACHABLE_CODE, no_bucket, ERROR, [Here]
codetoanalyze/java/performance/UnknownCallsTest.java, UnknownCallsTest.unmodeled_impure_linear(java.util.ArrayList):void, 0, EXPENSIVE_EXECUTION_TIME, no_bucket, ERROR, [with estimated cost 2 + 13 ⋅ list.length + 3 ⋅ (list.length + 1), O(list.length), degree = 1,{list.length + 1},Loop at line 61,{list.length},Loop at line 61]
[cost] Add Java tests for ZERO_EXECUTION_TIME Reviewed By: skcho Differential Revision: D20117064 fbshipit-source-id: 15c5eae64
5 years ago
codetoanalyze/java/performance/Zero.java, Zero.infeasible_path_zero():void, 0, UNREACHABLE_CODE, no_bucket, ERROR, [Here]
[cost] Add traces for ZERO_* issues Summary: The issue type `ZERO_EXECUTION_TIME` actually corresponds to bottom state but has been mistakenly used to mean - unreachable nodes (program never reaching exit state) - having zero cost (e.g. for allocations). Note that, for execution costs, the latter doesn't make sense since we always incur a unit cost for the start node. Hence, a function with empty body will have unit cost. For allocations or IO however, we only incur costs for specific primitives, so a function with no allocations/IO could have a zero cost. However, there is no point reporting functions with zero cost as a specific issue type. Instead, what we want to track is the former, i.e. functions whose cost becomes 0 due to program never reaching exit state. This diff aims to split these cases into two by only reporting on the latter and adds traces to bottom/unreachable cost by creating a special category in polynomials. Next diff will rename `ZERO_XXX` to `XXX_UNREACHABLE_AT_EXIT`. Reviewed By: skcho Differential Revision: D20005774 fbshipit-source-id: 46b9abd5a
5 years ago
codetoanalyze/java/performance/Zero.java, Zero.infeasible_path_zero():void, 0, ZERO_EXECUTION_TIME, no_bucket, ERROR, [Unreachable node]