|
|
|
|
/* @generated */
|
|
|
|
|
digraph cfg {
|
|
|
|
|
"operator_new#13351464032276683690.f7532bbe196cdd8520d944e0c3e1f7b4_1" [label="1: Start operator_new\nFormals: size:unsigned long ptr:void* ptr2:void*\nLocals: \n DECLARE_LOCALS(&return); [line 22, column 1]\n " color=yellow style=filled]
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
"operator_new#13351464032276683690.f7532bbe196cdd8520d944e0c3e1f7b4_1" -> "operator_new#13351464032276683690.f7532bbe196cdd8520d944e0c3e1f7b4_3" ;
|
|
|
|
|
"operator_new#13351464032276683690.f7532bbe196cdd8520d944e0c3e1f7b4_2" [label="2: Exit operator_new \n " color=yellow style=filled]
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
"operator_new#13351464032276683690.f7532bbe196cdd8520d944e0c3e1f7b4_3" [label="3: Return Stmt \n n$0=*&ptr2:void* [line 23, column 10]\n *&return:void*=n$0 [line 23, column 3]\n " shape="box"]
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
"operator_new#13351464032276683690.f7532bbe196cdd8520d944e0c3e1f7b4_3" -> "operator_new#13351464032276683690.f7532bbe196cdd8520d944e0c3e1f7b4_2" ;
|
|
|
|
|
"test#18241244337164948030.afc14f193ad97442f67ac7183be789bc_1" [label="1: Start test\nFormals: \nLocals: i:int* x:int \n DECLARE_LOCALS(&return,&i,&x); [line 10, column 1]\n " color=yellow style=filled]
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
"test#18241244337164948030.afc14f193ad97442f67ac7183be789bc_1" -> "test#18241244337164948030.afc14f193ad97442f67ac7183be789bc_6" ;
|
|
|
|
|
"test#18241244337164948030.afc14f193ad97442f67ac7183be789bc_2" [label="2: Exit test \n " color=yellow style=filled]
|
|
|
|
|
|
|
|
|
|
|
[clang] enforce that `instruction` always returns one SIL expression
Summary:
Previously, the type of `trans_result` contained a list of SIL expressions.
However, most of the time we expect to get exactly one, and getting a different
number is a soft(!) error, usually returning `-1`.
This splits `trans_result` into `control`, which contains the information
needed for temporary computation (hence when we don't necessarily know the
return value yet), and a new version of `trans_result` that includes `control`,
the previous `exps` list but replaced by a single `return` expression instead,
and a couple other values that made sense to move out of `control`. This allows
some flexibility in the frontend compared to enforcing exactly one return
expression always: if they are not known yet we stick to `control` instead (see
eg `compute_controls_to_parent`).
This creates more garbage temporary identifiers, however they do not show up in
the final cfg. Instead, we see that temporary IDs are now often not
consecutive...
The most painful complication is in the treatment of `DeclRefExpr`, which was
actually returning *two* expressions: the method name and the `this` object.
Now the method name is a separate (optional) field in `trans_result`.
Reviewed By: mbouaziz
Differential Revision: D7881088
fbshipit-source-id: 41ad3b5
7 years ago
|
|
|
"test#18241244337164948030.afc14f193ad97442f67ac7183be789bc_3" [label="3: Call delete \n n$1=*&i:int* [line 14, column 10]\n n$2=_fun___delete(n$1:int*) [line 14, column 3]\n " shape="box"]
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
"test#18241244337164948030.afc14f193ad97442f67ac7183be789bc_3" -> "test#18241244337164948030.afc14f193ad97442f67ac7183be789bc_2" ;
|
[clang] enforce that `instruction` always returns one SIL expression
Summary:
Previously, the type of `trans_result` contained a list of SIL expressions.
However, most of the time we expect to get exactly one, and getting a different
number is a soft(!) error, usually returning `-1`.
This splits `trans_result` into `control`, which contains the information
needed for temporary computation (hence when we don't necessarily know the
return value yet), and a new version of `trans_result` that includes `control`,
the previous `exps` list but replaced by a single `return` expression instead,
and a couple other values that made sense to move out of `control`. This allows
some flexibility in the frontend compared to enforcing exactly one return
expression always: if they are not known yet we stick to `control` instead (see
eg `compute_controls_to_parent`).
This creates more garbage temporary identifiers, however they do not show up in
the final cfg. Instead, we see that temporary IDs are now often not
consecutive...
The most painful complication is in the treatment of `DeclRefExpr`, which was
actually returning *two* expressions: the method name and the `this` object.
Now the method name is a separate (optional) field in `trans_result`.
Reviewed By: mbouaziz
Differential Revision: D7881088
fbshipit-source-id: 41ad3b5
7 years ago
|
|
|
"test#18241244337164948030.afc14f193ad97442f67ac7183be789bc_4" [label="4: CXXNewExpr \n n$3=_fun___new(sizeof(t=int):unsigned long) [line 13, column 3]\n " shape="box"]
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
"test#18241244337164948030.afc14f193ad97442f67ac7183be789bc_4" -> "test#18241244337164948030.afc14f193ad97442f67ac7183be789bc_3" ;
|
[clang] enforce that `instruction` always returns one SIL expression
Summary:
Previously, the type of `trans_result` contained a list of SIL expressions.
However, most of the time we expect to get exactly one, and getting a different
number is a soft(!) error, usually returning `-1`.
This splits `trans_result` into `control`, which contains the information
needed for temporary computation (hence when we don't necessarily know the
return value yet), and a new version of `trans_result` that includes `control`,
the previous `exps` list but replaced by a single `return` expression instead,
and a couple other values that made sense to move out of `control`. This allows
some flexibility in the frontend compared to enforcing exactly one return
expression always: if they are not known yet we stick to `control` instead (see
eg `compute_controls_to_parent`).
This creates more garbage temporary identifiers, however they do not show up in
the final cfg. Instead, we see that temporary IDs are now often not
consecutive...
The most painful complication is in the treatment of `DeclRefExpr`, which was
actually returning *two* expressions: the method name and the `this` object.
Now the method name is a separate (optional) field in `trans_result`.
Reviewed By: mbouaziz
Differential Revision: D7881088
fbshipit-source-id: 41ad3b5
7 years ago
|
|
|
"test#18241244337164948030.afc14f193ad97442f67ac7183be789bc_5" [label="5: DeclStmt \n n$4=_fun___new(sizeof(t=int):unsigned long) [line 12, column 12]\n *&i:int*=n$4 [line 12, column 3]\n " shape="box"]
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
"test#18241244337164948030.afc14f193ad97442f67ac7183be789bc_5" -> "test#18241244337164948030.afc14f193ad97442f67ac7183be789bc_4" ;
|
|
|
|
|
"test#18241244337164948030.afc14f193ad97442f67ac7183be789bc_6" [label="6: DeclStmt \n *&x:int=2 [line 11, column 3]\n " shape="box"]
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
"test#18241244337164948030.afc14f193ad97442f67ac7183be789bc_6" -> "test#18241244337164948030.afc14f193ad97442f67ac7183be789bc_5" ;
|
|
|
|
|
"test_placement#7589029240520377616.7f92d4e10c030674dddd1682731c0ba3_1" [label="1: Start test_placement\nFormals: ptr:void* ptr2:int*\nLocals: p:A* \n DECLARE_LOCALS(&return,&p); [line 26, column 1]\n " color=yellow style=filled]
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
"test_placement#7589029240520377616.7f92d4e10c030674dddd1682731c0ba3_1" -> "test_placement#7589029240520377616.7f92d4e10c030674dddd1682731c0ba3_3" ;
|
|
|
|
|
"test_placement#7589029240520377616.7f92d4e10c030674dddd1682731c0ba3_2" [label="2: Exit test_placement \n " color=yellow style=filled]
|
|
|
|
|
|
|
|
|
|
|
[clang] enforce that `instruction` always returns one SIL expression
Summary:
Previously, the type of `trans_result` contained a list of SIL expressions.
However, most of the time we expect to get exactly one, and getting a different
number is a soft(!) error, usually returning `-1`.
This splits `trans_result` into `control`, which contains the information
needed for temporary computation (hence when we don't necessarily know the
return value yet), and a new version of `trans_result` that includes `control`,
the previous `exps` list but replaced by a single `return` expression instead,
and a couple other values that made sense to move out of `control`. This allows
some flexibility in the frontend compared to enforcing exactly one return
expression always: if they are not known yet we stick to `control` instead (see
eg `compute_controls_to_parent`).
This creates more garbage temporary identifiers, however they do not show up in
the final cfg. Instead, we see that temporary IDs are now often not
consecutive...
The most painful complication is in the treatment of `DeclRefExpr`, which was
actually returning *two* expressions: the method name and the `this` object.
Now the method name is a separate (optional) field in `trans_result`.
Reviewed By: mbouaziz
Differential Revision: D7881088
fbshipit-source-id: 41ad3b5
7 years ago
|
|
|
"test_placement#7589029240520377616.7f92d4e10c030674dddd1682731c0ba3_3" [label="3: DeclStmt \n n$3=*&ptr:void* [line 26, column 60]\n n$1=*&ptr2:int* [line 26, column 65]\n *&ptr2:int*=(n$1 + 1) [line 26, column 65]\n n$2=*&ptr2:int* [line 26, column 65]\n n$4=_fun___placement_new(sizeof(t=A):unsigned long,n$3:void*,n$2:void*) [line 26, column 55]\n n$5=_fun_A_A(n$4:A*) [line 26, column 73]\n *&p:A*=n$4 [line 26, column 45]\n " shape="box"]
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
"test_placement#7589029240520377616.7f92d4e10c030674dddd1682731c0ba3_3" -> "test_placement#7589029240520377616.7f92d4e10c030674dddd1682731c0ba3_2" ;
|
|
|
|
|
}
|
