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 'f3411a2203'
${ noResults }
infer_clone/infer/tests/codetoanalyze/objc/shared/block/block_release.m.dot

80 lines
8.7 KiB
Raw Normal View History Unescape

Add @generated comment to icfg.dot files Reviewed By: jvillard Differential Revision: D3358243 fbshipit-source-id: a47cc01
9 years ago
/* @generated */
[cfg] rename iCFG to cfg in dotty files Summary: Not sure what an "iCFG" is but the dotty is only about CFGs anyway. Diff obtained by mass-`sed`. Reviewed By: sblackshear Differential Revision: D6324280 fbshipit-source-id: b7603bb
7 years ago
digraph cfg {
Get rid of Declare_locals Reviewed By: jeremydubreil Differential Revision: D9234331 fbshipit-source-id: 70443eabe
6 years ago
"objc_blockMy_manager_blockReleaseNoLeak_1(struct CGImage).d9015a9be6e156b53b1e0a2e237f4578_1" [label="1: Start objc_blockMy_manager_blockReleaseNoLeak_1\nFormals: newImage:CGImage* a:int\nLocals: \nCaptured: newImage:CGImage* \n " color=yellow style=filled]
[debug] print procedures in alphabetical order in cfgs Summary: When looking at large CFGs, at least in `xdot`, it's often difficult to find the procedure you're looking for. Sorting the proc names puts them in alphabetical order, which makes searching one procedure easier. Reviewed By: mbouaziz Differential Revision: D7758521 fbshipit-source-id: 8e9997f
7 years ago
[clang] Add the types of captured variables to the procname of blocks Reviewed By: mbouaziz Differential Revision: D8735374 fbshipit-source-id: 5c7c0b2
6 years ago
"objc_blockMy_manager_blockReleaseNoLeak_1(struct CGImage).d9015a9be6e156b53b1e0a2e237f4578_1" -> "objc_blockMy_manager_blockReleaseNoLeak_1(struct CGImage).d9015a9be6e156b53b1e0a2e237f4578_5" ;
"objc_blockMy_manager_blockReleaseNoLeak_1(struct CGImage).d9015a9be6e156b53b1e0a2e237f4578_1" -> "objc_blockMy_manager_blockReleaseNoLeak_1(struct CGImage).d9015a9be6e156b53b1e0a2e237f4578_6" ;
"objc_blockMy_manager_blockReleaseNoLeak_1(struct CGImage).d9015a9be6e156b53b1e0a2e237f4578_2" [label="2: Exit objc_blockMy_manager_blockReleaseNoLeak_1 \n " color=yellow style=filled]
[debug] print procedures in alphabetical order in cfgs Summary: When looking at large CFGs, at least in `xdot`, it's often difficult to find the procedure you're looking for. Sorting the proc names puts them in alphabetical order, which makes searching one procedure easier. Reviewed By: mbouaziz Differential Revision: D7758521 fbshipit-source-id: 8e9997f
7 years ago
[clang] Add the types of captured variables to the procname of blocks Reviewed By: mbouaziz Differential Revision: D8735374 fbshipit-source-id: 5c7c0b2
6 years ago
"objc_blockMy_manager_blockReleaseNoLeak_1(struct CGImage).d9015a9be6e156b53b1e0a2e237f4578_3" [label="3: + \n " ]
[debug] print procedures in alphabetical order in cfgs Summary: When looking at large CFGs, at least in `xdot`, it's often difficult to find the procedure you're looking for. Sorting the proc names puts them in alphabetical order, which makes searching one procedure easier. Reviewed By: mbouaziz Differential Revision: D7758521 fbshipit-source-id: 8e9997f
7 years ago
[clang] Add the types of captured variables to the procname of blocks Reviewed By: mbouaziz Differential Revision: D8735374 fbshipit-source-id: 5c7c0b2
6 years ago
"objc_blockMy_manager_blockReleaseNoLeak_1(struct CGImage).d9015a9be6e156b53b1e0a2e237f4578_3" -> "objc_blockMy_manager_blockReleaseNoLeak_1(struct CGImage).d9015a9be6e156b53b1e0a2e237f4578_4" ;
[cfg] print dotty *after* pre-analysis Summary: Useful to understand the changes in the pre-analysis, or to inspect the CFG that checkers actually get. This means that the pre-analysis always runs when we output the dotty, but I don't really see a reason why not. In fact, we could probably *always* store the CFGs as pre-analysed. Reviewed By: mbouaziz Differential Revision: D13102952 fbshipit-source-id: 89f3102ec
6 years ago
"objc_blockMy_manager_blockReleaseNoLeak_1(struct CGImage).d9015a9be6e156b53b1e0a2e237f4578_4" [label="4: between_join_and_exit \n APPLY_ABSTRACTION; [line 24, column 5]\n " shape="box"]
[debug] print procedures in alphabetical order in cfgs Summary: When looking at large CFGs, at least in `xdot`, it's often difficult to find the procedure you're looking for. Sorting the proc names puts them in alphabetical order, which makes searching one procedure easier. Reviewed By: mbouaziz Differential Revision: D7758521 fbshipit-source-id: 8e9997f
7 years ago
[clang] Add the types of captured variables to the procname of blocks Reviewed By: mbouaziz Differential Revision: D8735374 fbshipit-source-id: 5c7c0b2
6 years ago
"objc_blockMy_manager_blockReleaseNoLeak_1(struct CGImage).d9015a9be6e156b53b1e0a2e237f4578_4" -> "objc_blockMy_manager_blockReleaseNoLeak_1(struct CGImage).d9015a9be6e156b53b1e0a2e237f4578_2" ;
[preanalysis] do not delay killing variables taken by reference Summary: Before, the liveness pre-analysis would place extra instructions in the CFG for either: 1. marking an `Ident.t` as dead, or 2. marking a `Pvar.t` as `= 0` But we have no way of marking pvars dead without setting them to 0. This is bad because setting pvars to 0 is not possible everywhere they are dead. Indeed, we only do it when we haven't seen their address being taken anyway. This prevents the following situation, recorded in our tests: ``` int address_taken() { int** x; int* y; int i = 7; y = &i; x = &y; // if we don't reason about taken addresses while adding nullify instructions, // we'll add // `nullify(y)` here and report a false NPE on the next line return **x; } ``` So we want to mark pvars as dead without nullifying them. This diff extends the `Remove_temps` SIL instruction to accept pvars as well, and so renames it to `ExitScope`. Reviewed By: da319 Differential Revision: D13102953 fbshipit-source-id: aa7f03a52
6 years ago
"objc_blockMy_manager_blockReleaseNoLeak_1(struct CGImage).d9015a9be6e156b53b1e0a2e237f4578_5" [label="5: Prune (true branch, if) \n n$10=*&newImage:CGImage* [line 24, column 9]\n PRUNE(n$10, true); [line 24, column 9]\n EXIT_SCOPE(n$10); [line 24, column 9]\n " shape="invhouse"]
[debug] print procedures in alphabetical order in cfgs Summary: When looking at large CFGs, at least in `xdot`, it's often difficult to find the procedure you're looking for. Sorting the proc names puts them in alphabetical order, which makes searching one procedure easier. Reviewed By: mbouaziz Differential Revision: D7758521 fbshipit-source-id: 8e9997f
7 years ago
[clang] Add the types of captured variables to the procname of blocks Reviewed By: mbouaziz Differential Revision: D8735374 fbshipit-source-id: 5c7c0b2
6 years ago
"objc_blockMy_manager_blockReleaseNoLeak_1(struct CGImage).d9015a9be6e156b53b1e0a2e237f4578_5" -> "objc_blockMy_manager_blockReleaseNoLeak_1(struct CGImage).d9015a9be6e156b53b1e0a2e237f4578_7" ;
[preanalysis] do not delay killing variables taken by reference Summary: Before, the liveness pre-analysis would place extra instructions in the CFG for either: 1. marking an `Ident.t` as dead, or 2. marking a `Pvar.t` as `= 0` But we have no way of marking pvars dead without setting them to 0. This is bad because setting pvars to 0 is not possible everywhere they are dead. Indeed, we only do it when we haven't seen their address being taken anyway. This prevents the following situation, recorded in our tests: ``` int address_taken() { int** x; int* y; int i = 7; y = &i; x = &y; // if we don't reason about taken addresses while adding nullify instructions, // we'll add // `nullify(y)` here and report a false NPE on the next line return **x; } ``` So we want to mark pvars as dead without nullifying them. This diff extends the `Remove_temps` SIL instruction to accept pvars as well, and so renames it to `ExitScope`. Reviewed By: da319 Differential Revision: D13102953 fbshipit-source-id: aa7f03a52
6 years ago
"objc_blockMy_manager_blockReleaseNoLeak_1(struct CGImage).d9015a9be6e156b53b1e0a2e237f4578_6" [label="6: Prune (false branch, if) \n n$10=*&newImage:CGImage* [line 24, column 9]\n PRUNE(!n$10, false); [line 24, column 9]\n NULLIFY(&newImage); [line 24, column 9]\n EXIT_SCOPE(n$10,newImage); [line 24, column 9]\n APPLY_ABSTRACTION; [line 24, column 9]\n " shape="invhouse"]
[debug] print procedures in alphabetical order in cfgs Summary: When looking at large CFGs, at least in `xdot`, it's often difficult to find the procedure you're looking for. Sorting the proc names puts them in alphabetical order, which makes searching one procedure easier. Reviewed By: mbouaziz Differential Revision: D7758521 fbshipit-source-id: 8e9997f
7 years ago
[clang] Add the types of captured variables to the procname of blocks Reviewed By: mbouaziz Differential Revision: D8735374 fbshipit-source-id: 5c7c0b2
6 years ago
"objc_blockMy_manager_blockReleaseNoLeak_1(struct CGImage).d9015a9be6e156b53b1e0a2e237f4578_6" -> "objc_blockMy_manager_blockReleaseNoLeak_1(struct CGImage).d9015a9be6e156b53b1e0a2e237f4578_3" ;
[preanalysis] do not delay killing variables taken by reference Summary: Before, the liveness pre-analysis would place extra instructions in the CFG for either: 1. marking an `Ident.t` as dead, or 2. marking a `Pvar.t` as `= 0` But we have no way of marking pvars dead without setting them to 0. This is bad because setting pvars to 0 is not possible everywhere they are dead. Indeed, we only do it when we haven't seen their address being taken anyway. This prevents the following situation, recorded in our tests: ``` int address_taken() { int** x; int* y; int i = 7; y = &i; x = &y; // if we don't reason about taken addresses while adding nullify instructions, // we'll add // `nullify(y)` here and report a false NPE on the next line return **x; } ``` So we want to mark pvars as dead without nullifying them. This diff extends the `Remove_temps` SIL instruction to accept pvars as well, and so renames it to `ExitScope`. Reviewed By: da319 Differential Revision: D13102953 fbshipit-source-id: aa7f03a52
6 years ago
"objc_blockMy_manager_blockReleaseNoLeak_1(struct CGImage).d9015a9be6e156b53b1e0a2e237f4578_7" [label="7: Call _fun_CGImageRelease \n n$11=*&newImage:CGImage* [line 25, column 22]\n n$12=_fun_CGImageRelease(n$11:CGImage*) [line 25, column 7]\n NULLIFY(&newImage); [line 25, column 7]\n EXIT_SCOPE(n$11,n$12,newImage); [line 25, column 7]\n APPLY_ABSTRACTION; [line 25, column 7]\n " shape="box"]
[debug] print procedures in alphabetical order in cfgs Summary: When looking at large CFGs, at least in `xdot`, it's often difficult to find the procedure you're looking for. Sorting the proc names puts them in alphabetical order, which makes searching one procedure easier. Reviewed By: mbouaziz Differential Revision: D7758521 fbshipit-source-id: 8e9997f
7 years ago
[clang] Add the types of captured variables to the procname of blocks Reviewed By: mbouaziz Differential Revision: D8735374 fbshipit-source-id: 5c7c0b2
6 years ago
"objc_blockMy_manager_blockReleaseNoLeak_1(struct CGImage).d9015a9be6e156b53b1e0a2e237f4578_7" -> "objc_blockMy_manager_blockReleaseNoLeak_1(struct CGImage).d9015a9be6e156b53b1e0a2e237f4578_3" ;
Get rid of Declare_locals Reviewed By: jeremydubreil Differential Revision: D9234331 fbshipit-source-id: 70443eabe
6 years ago
"blockReleaseNoLeak#My_manager#instance.0c48f80f024250b18a529440f1313af6_1" [label="1: Start My_manager_blockReleaseNoLeak\nFormals: self:My_manager*\nLocals: newImage:CGImage* context:CGContext* z:int b:_fn_(*) \n " color=yellow style=filled]
Fix control-flow graph shape when conditional is last instruction. Summary:public When a conditional is the last instruction, there will be a join node leading directly to the exit node. Some instructions, such as nullification of dead variables, and abstraction, are added to the control flow graph automatically. But, join nodes cannot contain instructions. So when a procedure ends with a conditional, there might be no place to store these instructions. This diff adds one extra node between the join and the exit node in that situation. Reviewed By: jvillard Differential Revision: D3179056 fb-gh-sync-id: 2b9cd7e fbshipit-source-id: 2b9cd7e
9 years ago
[retain cycles] Fix renaming of closures, finds more retain cycles Reviewed By: ddino Differential Revision: D7012360 fbshipit-source-id: 4a972d3
7 years ago
"blockReleaseNoLeak#My_manager#instance.0c48f80f024250b18a529440f1313af6_1" -> "blockReleaseNoLeak#My_manager#instance.0c48f80f024250b18a529440f1313af6_12" ;
[cfg] print dotty *after* pre-analysis Summary: Useful to understand the changes in the pre-analysis, or to inspect the CFG that checkers actually get. This means that the pre-analysis always runs when we output the dotty, but I don't really see a reason why not. In fact, we could probably *always* store the CFGs as pre-analysed. Reviewed By: mbouaziz Differential Revision: D13102952 fbshipit-source-id: 89f3102ec
6 years ago
"blockReleaseNoLeak#My_manager#instance.0c48f80f024250b18a529440f1313af6_2" [label="2: Exit My_manager_blockReleaseNoLeak \n NULLIFY(&newImage); [line 31, column 1]\n NULLIFY(&context); [line 31, column 1]\n " color=yellow style=filled]
initial synchronization
10 years ago
[preanalysis] do not delay killing variables taken by reference Summary: Before, the liveness pre-analysis would place extra instructions in the CFG for either: 1. marking an `Ident.t` as dead, or 2. marking a `Pvar.t` as `= 0` But we have no way of marking pvars dead without setting them to 0. This is bad because setting pvars to 0 is not possible everywhere they are dead. Indeed, we only do it when we haven't seen their address being taken anyway. This prevents the following situation, recorded in our tests: ``` int address_taken() { int** x; int* y; int i = 7; y = &i; x = &y; // if we don't reason about taken addresses while adding nullify instructions, // we'll add // `nullify(y)` here and report a false NPE on the next line return **x; } ``` So we want to mark pvars as dead without nullifying them. This diff extends the `Remove_temps` SIL instruction to accept pvars as well, and so renames it to `ExitScope`. Reviewed By: da319 Differential Revision: D13102953 fbshipit-source-id: aa7f03a52
6 years ago
"blockReleaseNoLeak#My_manager#instance.0c48f80f024250b18a529440f1313af6_3" [label="3: Return Stmt \n n$0=*&z:int [line 30, column 10]\n *&return:int=n$0 [line 30, column 3]\n NULLIFY(&z); [line 30, column 3]\n EXIT_SCOPE(n$0,z); [line 30, column 3]\n APPLY_ABSTRACTION; [line 30, column 3]\n " shape="box"]
initial synchronization
10 years ago
[retain cycles] Fix renaming of closures, finds more retain cycles Reviewed By: ddino Differential Revision: D7012360 fbshipit-source-id: 4a972d3
7 years ago
"blockReleaseNoLeak#My_manager#instance.0c48f80f024250b18a529440f1313af6_3" -> "blockReleaseNoLeak#My_manager#instance.0c48f80f024250b18a529440f1313af6_2" ;
"blockReleaseNoLeak#My_manager#instance.0c48f80f024250b18a529440f1313af6_4" [label="4: + \n " ]
initial synchronization
10 years ago
[retain cycles] Fix renaming of closures, finds more retain cycles Reviewed By: ddino Differential Revision: D7012360 fbshipit-source-id: 4a972d3
7 years ago
"blockReleaseNoLeak#My_manager#instance.0c48f80f024250b18a529440f1313af6_4" -> "blockReleaseNoLeak#My_manager#instance.0c48f80f024250b18a529440f1313af6_3" ;
[preanalysis] do not delay killing variables taken by reference Summary: Before, the liveness pre-analysis would place extra instructions in the CFG for either: 1. marking an `Ident.t` as dead, or 2. marking a `Pvar.t` as `= 0` But we have no way of marking pvars dead without setting them to 0. This is bad because setting pvars to 0 is not possible everywhere they are dead. Indeed, we only do it when we haven't seen their address being taken anyway. This prevents the following situation, recorded in our tests: ``` int address_taken() { int** x; int* y; int i = 7; y = &i; x = &y; // if we don't reason about taken addresses while adding nullify instructions, // we'll add // `nullify(y)` here and report a false NPE on the next line return **x; } ``` So we want to mark pvars as dead without nullifying them. This diff extends the `Remove_temps` SIL instruction to accept pvars as well, and so renames it to `ExitScope`. Reviewed By: da319 Differential Revision: D13102953 fbshipit-source-id: aa7f03a52
6 years ago
"blockReleaseNoLeak#My_manager#instance.0c48f80f024250b18a529440f1313af6_5" [label="5: Prune (true branch, if) \n n$1=*&context:CGContext* [line 28, column 7]\n PRUNE(n$1, true); [line 28, column 7]\n EXIT_SCOPE(n$1); [line 28, column 7]\n " shape="invhouse"]
initial synchronization
10 years ago
[retain cycles] Fix renaming of closures, finds more retain cycles Reviewed By: ddino Differential Revision: D7012360 fbshipit-source-id: 4a972d3
7 years ago
"blockReleaseNoLeak#My_manager#instance.0c48f80f024250b18a529440f1313af6_5" -> "blockReleaseNoLeak#My_manager#instance.0c48f80f024250b18a529440f1313af6_7" ;
[preanalysis] do not delay killing variables taken by reference Summary: Before, the liveness pre-analysis would place extra instructions in the CFG for either: 1. marking an `Ident.t` as dead, or 2. marking a `Pvar.t` as `= 0` But we have no way of marking pvars dead without setting them to 0. This is bad because setting pvars to 0 is not possible everywhere they are dead. Indeed, we only do it when we haven't seen their address being taken anyway. This prevents the following situation, recorded in our tests: ``` int address_taken() { int** x; int* y; int i = 7; y = &i; x = &y; // if we don't reason about taken addresses while adding nullify instructions, // we'll add // `nullify(y)` here and report a false NPE on the next line return **x; } ``` So we want to mark pvars as dead without nullifying them. This diff extends the `Remove_temps` SIL instruction to accept pvars as well, and so renames it to `ExitScope`. Reviewed By: da319 Differential Revision: D13102953 fbshipit-source-id: aa7f03a52
6 years ago
"blockReleaseNoLeak#My_manager#instance.0c48f80f024250b18a529440f1313af6_6" [label="6: Prune (false branch, if) \n n$1=*&context:CGContext* [line 28, column 7]\n PRUNE(!n$1, false); [line 28, column 7]\n EXIT_SCOPE(n$1,context); [line 28, column 7]\n APPLY_ABSTRACTION; [line 28, column 7]\n " shape="invhouse"]
initial synchronization
10 years ago
[retain cycles] Fix renaming of closures, finds more retain cycles Reviewed By: ddino Differential Revision: D7012360 fbshipit-source-id: 4a972d3
7 years ago
"blockReleaseNoLeak#My_manager#instance.0c48f80f024250b18a529440f1313af6_6" -> "blockReleaseNoLeak#My_manager#instance.0c48f80f024250b18a529440f1313af6_4" ;
[preanalysis] do not delay killing variables taken by reference Summary: Before, the liveness pre-analysis would place extra instructions in the CFG for either: 1. marking an `Ident.t` as dead, or 2. marking a `Pvar.t` as `= 0` But we have no way of marking pvars dead without setting them to 0. This is bad because setting pvars to 0 is not possible everywhere they are dead. Indeed, we only do it when we haven't seen their address being taken anyway. This prevents the following situation, recorded in our tests: ``` int address_taken() { int** x; int* y; int i = 7; y = &i; x = &y; // if we don't reason about taken addresses while adding nullify instructions, // we'll add // `nullify(y)` here and report a false NPE on the next line return **x; } ``` So we want to mark pvars as dead without nullifying them. This diff extends the `Remove_temps` SIL instruction to accept pvars as well, and so renames it to `ExitScope`. Reviewed By: da319 Differential Revision: D13102953 fbshipit-source-id: aa7f03a52
6 years ago
"blockReleaseNoLeak#My_manager#instance.0c48f80f024250b18a529440f1313af6_7" [label="7: Call _fun_CGContextRelease \n n$2=*&context:CGContext* [line 29, column 22]\n n$3=_fun_CGContextRelease(n$2:CGContext*) [line 29, column 5]\n EXIT_SCOPE(n$2,n$3,context); [line 29, column 5]\n APPLY_ABSTRACTION; [line 29, column 5]\n " shape="box"]
initial synchronization
10 years ago
[retain cycles] Fix renaming of closures, finds more retain cycles Reviewed By: ddino Differential Revision: D7012360 fbshipit-source-id: 4a972d3
7 years ago
"blockReleaseNoLeak#My_manager#instance.0c48f80f024250b18a529440f1313af6_7" -> "blockReleaseNoLeak#My_manager#instance.0c48f80f024250b18a529440f1313af6_4" ;
[preanalysis] do not delay killing variables taken by reference Summary: Before, the liveness pre-analysis would place extra instructions in the CFG for either: 1. marking an `Ident.t` as dead, or 2. marking a `Pvar.t` as `= 0` But we have no way of marking pvars dead without setting them to 0. This is bad because setting pvars to 0 is not possible everywhere they are dead. Indeed, we only do it when we haven't seen their address being taken anyway. This prevents the following situation, recorded in our tests: ``` int address_taken() { int** x; int* y; int i = 7; y = &i; x = &y; // if we don't reason about taken addresses while adding nullify instructions, // we'll add // `nullify(y)` here and report a false NPE on the next line return **x; } ``` So we want to mark pvars as dead without nullifying them. This diff extends the `Remove_temps` SIL instruction to accept pvars as well, and so renames it to `ExitScope`. Reviewed By: da319 Differential Revision: D13102953 fbshipit-source-id: aa7f03a52
6 years ago
"blockReleaseNoLeak#My_manager#instance.0c48f80f024250b18a529440f1313af6_8" [label="8: Call n$6 \n n$6=*&b:_fn_(*) [line 27, column 3]\n n$7=*&z:int [line 27, column 5]\n n$8=n$6(n$7:int) [line 27, column 3]\n NULLIFY(&b); [line 27, column 3]\n EXIT_SCOPE(n$6,n$7,n$8,b); [line 27, column 3]\n " shape="box"]
initial synchronization
10 years ago
[retain cycles] Fix renaming of closures, finds more retain cycles Reviewed By: ddino Differential Revision: D7012360 fbshipit-source-id: 4a972d3
7 years ago
"blockReleaseNoLeak#My_manager#instance.0c48f80f024250b18a529440f1313af6_8" -> "blockReleaseNoLeak#My_manager#instance.0c48f80f024250b18a529440f1313af6_5" ;
"blockReleaseNoLeak#My_manager#instance.0c48f80f024250b18a529440f1313af6_8" -> "blockReleaseNoLeak#My_manager#instance.0c48f80f024250b18a529440f1313af6_6" ;
[preanalysis] do not delay killing variables taken by reference Summary: Before, the liveness pre-analysis would place extra instructions in the CFG for either: 1. marking an `Ident.t` as dead, or 2. marking a `Pvar.t` as `= 0` But we have no way of marking pvars dead without setting them to 0. This is bad because setting pvars to 0 is not possible everywhere they are dead. Indeed, we only do it when we haven't seen their address being taken anyway. This prevents the following situation, recorded in our tests: ``` int address_taken() { int** x; int* y; int i = 7; y = &i; x = &y; // if we don't reason about taken addresses while adding nullify instructions, // we'll add // `nullify(y)` here and report a false NPE on the next line return **x; } ``` So we want to mark pvars as dead without nullifying them. This diff extends the `Remove_temps` SIL instruction to accept pvars as well, and so renames it to `ExitScope`. Reviewed By: da319 Differential Revision: D13102953 fbshipit-source-id: aa7f03a52
6 years ago
"blockReleaseNoLeak#My_manager#instance.0c48f80f024250b18a529440f1313af6_9" [label="9: BinaryOperatorStmt: Assign \n n$9=*&newImage:CGImage* [line 23, column 7]\n *&b:_fn_(*)=(_fun_objc_blockMy_manager_blockReleaseNoLeak_1,(n$9 &newImage:CGImage*)) [line 23, column 3]\n EXIT_SCOPE(n$9,newImage); [line 23, column 3]\n " shape="box"]
initial synchronization
10 years ago
[retain cycles] Fix renaming of closures, finds more retain cycles Reviewed By: ddino Differential Revision: D7012360 fbshipit-source-id: 4a972d3
7 years ago
"blockReleaseNoLeak#My_manager#instance.0c48f80f024250b18a529440f1313af6_9" -> "blockReleaseNoLeak#My_manager#instance.0c48f80f024250b18a529440f1313af6_8" ;
[preanalysis] do not delay killing variables taken by reference Summary: Before, the liveness pre-analysis would place extra instructions in the CFG for either: 1. marking an `Ident.t` as dead, or 2. marking a `Pvar.t` as `= 0` But we have no way of marking pvars dead without setting them to 0. This is bad because setting pvars to 0 is not possible everywhere they are dead. Indeed, we only do it when we haven't seen their address being taken anyway. This prevents the following situation, recorded in our tests: ``` int address_taken() { int** x; int* y; int i = 7; y = &i; x = &y; // if we don't reason about taken addresses while adding nullify instructions, // we'll add // `nullify(y)` here and report a false NPE on the next line return **x; } ``` So we want to mark pvars as dead without nullifying them. This diff extends the `Remove_temps` SIL instruction to accept pvars as well, and so renames it to `ExitScope`. Reviewed By: da319 Differential Revision: D13102953 fbshipit-source-id: aa7f03a52
6 years ago
"blockReleaseNoLeak#My_manager#instance.0c48f80f024250b18a529440f1313af6_10" [label="10: DeclStmt \n n$17=_fun___variable_initialization(&newImage:CGImage*) [line 22, column 3]\n n$15=*&context:CGContext* [line 22, column 52]\n n$16=_fun_CGBitmapContextCreateImage(n$15:CGContext*) [line 22, column 25]\n *&newImage:CGImage*=n$16 [line 22, column 3]\n EXIT_SCOPE(n$15,n$16,n$17); [line 22, column 3]\n " shape="box"]
initial synchronization
10 years ago
[retain cycles] Fix renaming of closures, finds more retain cycles Reviewed By: ddino Differential Revision: D7012360 fbshipit-source-id: 4a972d3
7 years ago
"blockReleaseNoLeak#My_manager#instance.0c48f80f024250b18a529440f1313af6_10" -> "blockReleaseNoLeak#My_manager#instance.0c48f80f024250b18a529440f1313af6_9" ;
[preanalysis] do not delay killing variables taken by reference Summary: Before, the liveness pre-analysis would place extra instructions in the CFG for either: 1. marking an `Ident.t` as dead, or 2. marking a `Pvar.t` as `= 0` But we have no way of marking pvars dead without setting them to 0. This is bad because setting pvars to 0 is not possible everywhere they are dead. Indeed, we only do it when we haven't seen their address being taken anyway. This prevents the following situation, recorded in our tests: ``` int address_taken() { int** x; int* y; int i = 7; y = &i; x = &y; // if we don't reason about taken addresses while adding nullify instructions, // we'll add // `nullify(y)` here and report a false NPE on the next line return **x; } ``` So we want to mark pvars as dead without nullifying them. This diff extends the `Remove_temps` SIL instruction to accept pvars as well, and so renames it to `ExitScope`. Reviewed By: da319 Differential Revision: D13102953 fbshipit-source-id: aa7f03a52
6 years ago
"blockReleaseNoLeak#My_manager#instance.0c48f80f024250b18a529440f1313af6_11" [label="11: DeclStmt \n n$19=_fun___variable_initialization(&context:CGContext*) [line 21, column 3]\n n$18=_fun_CGBitmapContextCreate(null:void*,(unsigned long)0:unsigned long,(unsigned long)0:unsigned long,(unsigned long)8:unsigned long,(unsigned long)0:unsigned long,null:CGColorSpace*,(unsigned int)0:unsigned int) [line 21, column 26]\n *&context:CGContext*=n$18 [line 21, column 3]\n EXIT_SCOPE(n$18,n$19); [line 21, column 3]\n " shape="box"]
initial synchronization
10 years ago
[retain cycles] Fix renaming of closures, finds more retain cycles Reviewed By: ddino Differential Revision: D7012360 fbshipit-source-id: 4a972d3
7 years ago
"blockReleaseNoLeak#My_manager#instance.0c48f80f024250b18a529440f1313af6_11" -> "blockReleaseNoLeak#My_manager#instance.0c48f80f024250b18a529440f1313af6_10" ;
[preanalysis] do not delay killing variables taken by reference Summary: Before, the liveness pre-analysis would place extra instructions in the CFG for either: 1. marking an `Ident.t` as dead, or 2. marking a `Pvar.t` as `= 0` But we have no way of marking pvars dead without setting them to 0. This is bad because setting pvars to 0 is not possible everywhere they are dead. Indeed, we only do it when we haven't seen their address being taken anyway. This prevents the following situation, recorded in our tests: ``` int address_taken() { int** x; int* y; int i = 7; y = &i; x = &y; // if we don't reason about taken addresses while adding nullify instructions, // we'll add // `nullify(y)` here and report a false NPE on the next line return **x; } ``` So we want to mark pvars as dead without nullifying them. This diff extends the `Remove_temps` SIL instruction to accept pvars as well, and so renames it to `ExitScope`. Reviewed By: da319 Differential Revision: D13102953 fbshipit-source-id: aa7f03a52
6 years ago
"blockReleaseNoLeak#My_manager#instance.0c48f80f024250b18a529440f1313af6_12" [label="12: DeclStmt \n n$20=_fun___variable_initialization(&z:int) [line 20, column 3]\n *&z:int=3 [line 20, column 3]\n EXIT_SCOPE(n$20); [line 20, column 3]\n " shape="box"]
initial synchronization
10 years ago
[retain cycles] Fix renaming of closures, finds more retain cycles Reviewed By: ddino Differential Revision: D7012360 fbshipit-source-id: 4a972d3
7 years ago
"blockReleaseNoLeak#My_manager#instance.0c48f80f024250b18a529440f1313af6_12" -> "blockReleaseNoLeak#My_manager#instance.0c48f80f024250b18a529440f1313af6_11" ;
initial synchronization
10 years ago
}