[inferbo] Visit every nodes at narrowing

Reviewed By: mbouaziz

Differential Revision: D13320983

fbshipit-source-id: 1b31c3c17
master
Sungkeun Cho 6 years ago committed by Facebook Github Bot
parent a689301c53
commit 98d05044fb

@ -344,15 +344,21 @@ module MakeUsingWTO (TransferFunctions : TransferFunctions.SIL) = struct
L.(die InternalError) "Could not compute the pre of a node"
let rec exec_wto_component ~pp_instr cfg proc_data inv_map head ~is_loop_head ~is_narrowing rest
=
let rec exec_wto_component ~pp_instr cfg proc_data inv_map head ~is_loop_head ~is_narrowing
~is_first_visit rest =
match exec_wto_node ~pp_instr cfg proc_data inv_map head ~is_loop_head ~is_narrowing with
| inv_map, ReachedFixPoint ->
inv_map
if is_narrowing && is_first_visit then
exec_wto_rest ~pp_instr cfg proc_data inv_map head ~is_narrowing rest
else inv_map
| inv_map, DidNotReachFixPoint ->
let inv_map = exec_wto_partition ~pp_instr cfg proc_data ~is_narrowing inv_map rest in
exec_wto_component ~pp_instr cfg proc_data inv_map head ~is_loop_head:true ~is_narrowing
rest
exec_wto_rest ~pp_instr cfg proc_data inv_map head ~is_narrowing rest
and exec_wto_rest ~pp_instr cfg proc_data inv_map head ~is_narrowing rest =
let inv_map = exec_wto_partition ~pp_instr cfg proc_data ~is_narrowing inv_map rest in
exec_wto_component ~pp_instr cfg proc_data inv_map head ~is_loop_head:true ~is_narrowing
~is_first_visit:false rest
and exec_wto_partition ~pp_instr cfg proc_data ~is_narrowing inv_map
@ -369,7 +375,7 @@ module MakeUsingWTO (TransferFunctions : TransferFunctions.SIL) = struct
| Component {head; rest; next} ->
let inv_map =
exec_wto_component ~pp_instr cfg proc_data inv_map head ~is_loop_head:false ~is_narrowing
rest
~is_first_visit:true rest
in
exec_wto_partition ~pp_instr cfg proc_data ~is_narrowing inv_map next

@ -167,6 +167,11 @@ codetoanalyze/c/bufferoverrun/models.c, strncpy_good5_FP, 3, BUFFER_OVERRUN_L1,
codetoanalyze/c/bufferoverrun/nested_loop.c, nested_loop, 7, BUFFER_OVERRUN_L2, no_bucket, ERROR, [<Offset trace>,Assignment,<Length trace>,Array declaration,Array access: Offset: [0, 10] Size: 10]
codetoanalyze/c/bufferoverrun/nested_loop.c, nested_loop3_Bad, 6, BUFFER_OVERRUN_L2, no_bucket, ERROR, [<Offset trace>,Assignment,<Length trace>,Array declaration,Array access: Offset: [0, 10] Size: 10]
codetoanalyze/c/bufferoverrun/nested_loop.c, nested_loop4_Bad, 6, BUFFER_OVERRUN_L2, no_bucket, ERROR, [<Offset trace>,Assignment,<Length trace>,Array declaration,Array access: Offset: [0, 10] Size: 10]
codetoanalyze/c/bufferoverrun/nested_loop.c, nested_loop_narrowing_Bad, 4, CONDITION_ALWAYS_TRUE, no_bucket, WARNING, [Here]
codetoanalyze/c/bufferoverrun/nested_loop.c, nested_loop_narrowing_Bad, 5, CONDITION_ALWAYS_TRUE, no_bucket, WARNING, [Here]
codetoanalyze/c/bufferoverrun/nested_loop.c, nested_loop_narrowing_Bad, 6, BUFFER_OVERRUN_L2, no_bucket, ERROR, [<Offset trace>,Assignment,<Length trace>,Array declaration,Array access: Offset: [0, 10] Size: 10]
codetoanalyze/c/bufferoverrun/nested_loop.c, nested_loop_narrowing_Good, 4, CONDITION_ALWAYS_TRUE, no_bucket, WARNING, [Here]
codetoanalyze/c/bufferoverrun/nested_loop.c, nested_loop_narrowing_Good, 5, CONDITION_ALWAYS_TRUE, no_bucket, WARNING, [Here]
codetoanalyze/c/bufferoverrun/nested_loop_with_label.c, nested_loop_with_label, 4, CONDITION_ALWAYS_TRUE, no_bucket, WARNING, [Here]
codetoanalyze/c/bufferoverrun/nested_loop_with_label.c, nested_loop_with_label, 6, BUFFER_OVERRUN_L2, no_bucket, ERROR, [<Offset trace>,Assignment,<Length trace>,Array declaration,Array access: Offset: [0, 10] Size: 10]
codetoanalyze/c/bufferoverrun/nested_loop_with_label.c, nested_loop_with_label, 7, CONDITION_ALWAYS_TRUE, no_bucket, WARNING, [Here]

@ -52,3 +52,29 @@ void nested_loop4_Bad() {
}
}
}
void nested_loop_narrowing_Good() {
int i = 0;
int a[10];
while (1) {
while (1) {
a[i] = 0;
for (i = 0; i < 5; i++) {
}
}
}
}
void nested_loop_narrowing_Bad() {
int i = 0;
int a[10];
while (1) {
while (1) {
a[i] = 0;
for (i = 0; i < 10; i++) {
}
}
}
}

Loading…
Cancel
Save