RPCS3/llvm-mirror
1
0
Fork 0
mirror of https://github.com/RPCS3/llvm-mirror.git synced 2024-10-24 13:33:37 +02:00
Code Issues Projects Releases Wiki Activity
e3844a404a
llvm-mirror/test/Transforms/ObjCARC/cfg-hazards.ll

436 lines
11 KiB
LLVM
Raw Normal View History

The ARC language-specific optimizer. Credit to Dan Gohman. llvm-svn: 133108
2011-06-16 01:37:01 +02:00
; RUN: opt -S -objc-arc < %s | FileCheck %s
; rdar://9503416
; Detect loop boundaries and don't move retains and releases
; across them.
declare void @use_pointer(i8*)
declare i8* @objc_retain(i8*)
declare void @objc_release(i8*)
Don't let arbitrary calls disrupt nested retain+release pairs if the retains and releases all use the same SSA pointer value. Also, don't let CFG hazards disrupt nested retain+release pair optimizations. llvm-svn: 137399
2011-08-12 02:26:31 +02:00
declare void @callee()
[objc-arc] Do not mismatch up retains inside a for loop with releases outside said for loop in the presense of differing provenance caused by escaping blocks. This occurs due to an alloca representing a separate ownership from the original pointer. Thus consider the following pseudo-IR: objc_retain(%a) for (...) { objc_retain(%a) %block <- %a F(%block) objc_release(%block) } objc_release(%a) From the perspective of the optimizer, the %block is a separate provenance from the original %a. Thus the optimizer pairs up the inner retain for %a and the outer release from %a, resulting in segfaults. This is fixed by noting that the signature of a mismatch of retain/releases inside the for loop is a Use/CanRelease top down with an None bottom up (since bottom up the Retain-CanRelease-Use-Release sequence is completed by the inner objc_retain, but top down due to the differing provenance from the objc_release said sequence is not completed). In said case in CheckForCFGHazards, we now clear the state of %a implying that no pairing will occur. Additionally a test case is included. rdar://12969722 llvm-svn: 179747
2013-04-18 07:39:45 +02:00
declare void @block_callee(void ()*)
The ARC language-specific optimizer. Credit to Dan Gohman. llvm-svn: 133108
2011-06-16 01:37:01 +02:00
Catch more CHECK that can be converted to CHECK-LABEL in Transforms for easier debugging. No functionality change. This conversion was done with the following bash script: find test/Transforms -name "*.ll" | \ while read NAME; do echo "$NAME" if ! grep -q "^; *RUN: *llc" $NAME; then TEMP=`mktemp -t temp` cp $NAME $TEMP sed -n "s/^define [^@]*@\([A-Za-z0-9_]*\)(.*$/\1/p" < $NAME | \ while read FUNC; do sed -i '' "s/;\(.*\)\([A-Za-z0-9_]*\):\( *\)define\([^@]*\)@$FUNC\([( ]*\)\$/;\1\2-LABEL:\3define\4@$FUNC(/g" $TEMP done mv $TEMP $NAME fi done llvm-svn: 186269
2013-07-14 03:50:49 +02:00
; CHECK-LABEL: define void @test0(
The ARC language-specific optimizer. Credit to Dan Gohman. llvm-svn: 133108
2011-06-16 01:37:01 +02:00
; CHECK: call i8* @objc_retain(
; CHECK: for.body:
; CHECK-NOT: @objc
; CHECK: for.end:
; CHECK: call void @objc_release(
; CHECK: }
define void @test0(i8* %digits) {
entry:
%tmp1 = call i8* @objc_retain(i8* %digits) nounwind
Tidy up these testcases to look more like real code does. llvm-svn: 137085
2011-08-09 02:33:11 +02:00
call void @use_pointer(i8* %digits)
The ARC language-specific optimizer. Credit to Dan Gohman. llvm-svn: 133108
2011-06-16 01:37:01 +02:00
br label %for.body
for.body: ; preds = %for.body, %entry
%upcDigitIndex.01 = phi i64 [ 2, %entry ], [ %inc, %for.body ]
Tidy up these testcases to look more like real code does. llvm-svn: 137085
2011-08-09 02:33:11 +02:00
call void @use_pointer(i8* %digits)
The ARC language-specific optimizer. Credit to Dan Gohman. llvm-svn: 133108
2011-06-16 01:37:01 +02:00
%inc = add i64 %upcDigitIndex.01, 1
%cmp = icmp ult i64 %inc, 12
br i1 %cmp, label %for.body, label %for.end
for.end: ; preds = %for.body
Tidy up these testcases to look more like real code does. llvm-svn: 137085
2011-08-09 02:33:11 +02:00
call void @objc_release(i8* %digits) nounwind, !clang.imprecise_release !0
The ARC language-specific optimizer. Credit to Dan Gohman. llvm-svn: 133108
2011-06-16 01:37:01 +02:00
ret void
}
Catch more CHECK that can be converted to CHECK-LABEL in Transforms for easier debugging. No functionality change. This conversion was done with the following bash script: find test/Transforms -name "*.ll" | \ while read NAME; do echo "$NAME" if ! grep -q "^; *RUN: *llc" $NAME; then TEMP=`mktemp -t temp` cp $NAME $TEMP sed -n "s/^define [^@]*@\([A-Za-z0-9_]*\)(.*$/\1/p" < $NAME | \ while read FUNC; do sed -i '' "s/;\(.*\)\([A-Za-z0-9_]*\):\( *\)define\([^@]*\)@$FUNC\([( ]*\)\$/;\1\2-LABEL:\3define\4@$FUNC(/g" $TEMP done mv $TEMP $NAME fi done llvm-svn: 186269
2013-07-14 03:50:49 +02:00
; CHECK-LABEL: define void @test1(
The ARC language-specific optimizer. Credit to Dan Gohman. llvm-svn: 133108
2011-06-16 01:37:01 +02:00
; CHECK: call i8* @objc_retain(
; CHECK: for.body:
; CHECK-NOT: @objc
; CHECK: for.end:
; CHECK: void @objc_release(
; CHECK: }
define void @test1(i8* %digits) {
entry:
%tmp1 = call i8* @objc_retain(i8* %digits) nounwind
br label %for.body
for.body: ; preds = %for.body, %entry
%upcDigitIndex.01 = phi i64 [ 2, %entry ], [ %inc, %for.body ]
Tidy up these testcases to look more like real code does. llvm-svn: 137085
2011-08-09 02:33:11 +02:00
call void @use_pointer(i8* %digits)
call void @use_pointer(i8* %digits)
The ARC language-specific optimizer. Credit to Dan Gohman. llvm-svn: 133108
2011-06-16 01:37:01 +02:00
%inc = add i64 %upcDigitIndex.01, 1
%cmp = icmp ult i64 %inc, 12
br i1 %cmp, label %for.body, label %for.end
for.end: ; preds = %for.body
Tidy up these testcases to look more like real code does. llvm-svn: 137085
2011-08-09 02:33:11 +02:00
call void @objc_release(i8* %digits) nounwind, !clang.imprecise_release !0
The ARC language-specific optimizer. Credit to Dan Gohman. llvm-svn: 133108
2011-06-16 01:37:01 +02:00
ret void
}
Catch more CHECK that can be converted to CHECK-LABEL in Transforms for easier debugging. No functionality change. This conversion was done with the following bash script: find test/Transforms -name "*.ll" | \ while read NAME; do echo "$NAME" if ! grep -q "^; *RUN: *llc" $NAME; then TEMP=`mktemp -t temp` cp $NAME $TEMP sed -n "s/^define [^@]*@\([A-Za-z0-9_]*\)(.*$/\1/p" < $NAME | \ while read FUNC; do sed -i '' "s/;\(.*\)\([A-Za-z0-9_]*\):\( *\)define\([^@]*\)@$FUNC\([( ]*\)\$/;\1\2-LABEL:\3define\4@$FUNC(/g" $TEMP done mv $TEMP $NAME fi done llvm-svn: 186269
2013-07-14 03:50:49 +02:00
; CHECK-LABEL: define void @test2(
The ARC language-specific optimizer. Credit to Dan Gohman. llvm-svn: 133108
2011-06-16 01:37:01 +02:00
; CHECK: call i8* @objc_retain(
; CHECK: for.body:
; CHECK-NOT: @objc
; CHECK: for.end:
; CHECK: void @objc_release(
; CHECK: }
define void @test2(i8* %digits) {
entry:
%tmp1 = call i8* @objc_retain(i8* %digits) nounwind
br label %for.body
for.body: ; preds = %for.body, %entry
%upcDigitIndex.01 = phi i64 [ 2, %entry ], [ %inc, %for.body ]
Tidy up these testcases to look more like real code does. llvm-svn: 137085
2011-08-09 02:33:11 +02:00
call void @use_pointer(i8* %digits)
The ARC language-specific optimizer. Credit to Dan Gohman. llvm-svn: 133108
2011-06-16 01:37:01 +02:00
%inc = add i64 %upcDigitIndex.01, 1
%cmp = icmp ult i64 %inc, 12
br i1 %cmp, label %for.body, label %for.end
for.end: ; preds = %for.body
Tidy up these testcases to look more like real code does. llvm-svn: 137085
2011-08-09 02:33:11 +02:00
call void @use_pointer(i8* %digits)
call void @objc_release(i8* %digits) nounwind, !clang.imprecise_release !0
The ARC language-specific optimizer. Credit to Dan Gohman. llvm-svn: 133108
2011-06-16 01:37:01 +02:00
ret void
}
Don't let arbitrary calls disrupt nested retain+release pairs if the retains and releases all use the same SSA pointer value. Also, don't let CFG hazards disrupt nested retain+release pair optimizations. llvm-svn: 137399
2011-08-12 02:26:31 +02:00
; Delete nested retain+release pairs around loops.
Modify the LLVM assembly output so that it uses references to represent function attributes. This makes the LLVM assembly look better. E.g.: define void @foo() #0 { ret void } attributes #0 = { nounwind noinline ssp } llvm-svn: 175605
2013-02-20 08:21:42 +01:00
; CHECK: define void @test3(i8* %a) #0 {
Don't let arbitrary calls disrupt nested retain+release pairs if the retains and releases all use the same SSA pointer value. Also, don't let CFG hazards disrupt nested retain+release pair optimizations. llvm-svn: 137399
2011-08-12 02:26:31 +02:00
; CHECK-NEXT: entry:
Use references to attribute groups on the call/invoke instructions. Listing all of the attributes for the callee of a call/invoke instruction is way too much and makes the IR unreadable. Use references to attributes instead. llvm-svn: 175877
2013-02-22 10:09:42 +01:00
; CHECK-NEXT: tail call i8* @objc_retain(i8* %a) [[NUW:#[0-9]+]]
Don't let arbitrary calls disrupt nested retain+release pairs if the retains and releases all use the same SSA pointer value. Also, don't let CFG hazards disrupt nested retain+release pair optimizations. llvm-svn: 137399
2011-08-12 02:26:31 +02:00
; CHECK-NEXT: br label %loop
; CHECK-NOT: @objc_
; CHECK: exit:
; CHECK-NEXT: call void @objc_release(i8* %a)
; CHECK-NEXT: ret void
; CHECK-NEXT: }
define void @test3(i8* %a) nounwind {
entry:
%outer = call i8* @objc_retain(i8* %a) nounwind
%inner = call i8* @objc_retain(i8* %a) nounwind
br label %loop
loop:
call void @callee()
store i8 0, i8* %a
br i1 undef, label %loop, label %exit
exit:
call void @objc_release(i8* %a) nounwind
call void @objc_release(i8* %a) nounwind, !clang.imprecise_release !0
ret void
}
Modify the LLVM assembly output so that it uses references to represent function attributes. This makes the LLVM assembly look better. E.g.: define void @foo() #0 { ret void } attributes #0 = { nounwind noinline ssp } llvm-svn: 175605
2013-02-20 08:21:42 +01:00
; CHECK: define void @test4(i8* %a) #0 {
Don't let arbitrary calls disrupt nested retain+release pairs if the retains and releases all use the same SSA pointer value. Also, don't let CFG hazards disrupt nested retain+release pair optimizations. llvm-svn: 137399
2011-08-12 02:26:31 +02:00
; CHECK-NEXT: entry:
Use references to attribute groups on the call/invoke instructions. Listing all of the attributes for the callee of a call/invoke instruction is way too much and makes the IR unreadable. Use references to attributes instead. llvm-svn: 175877
2013-02-22 10:09:42 +01:00
; CHECK-NEXT: tail call i8* @objc_retain(i8* %a) [[NUW]]
Don't let arbitrary calls disrupt nested retain+release pairs if the retains and releases all use the same SSA pointer value. Also, don't let CFG hazards disrupt nested retain+release pair optimizations. llvm-svn: 137399
2011-08-12 02:26:31 +02:00
; CHECK-NEXT: br label %loop
; CHECK-NOT: @objc_
; CHECK: exit:
; CHECK-NEXT: call void @objc_release(i8* %a)
; CHECK-NEXT: ret void
; CHECK-NEXT: }
define void @test4(i8* %a) nounwind {
entry:
%outer = call i8* @objc_retain(i8* %a) nounwind
%inner = call i8* @objc_retain(i8* %a) nounwind
br label %loop
loop:
br label %more
more:
call void @callee()
call void @callee()
store i8 0, i8* %a
br i1 undef, label %loop, label %exit
exit:
call void @objc_release(i8* %a) nounwind
call void @objc_release(i8* %a) nounwind, !clang.imprecise_release !0
ret void
}
Modify the LLVM assembly output so that it uses references to represent function attributes. This makes the LLVM assembly look better. E.g.: define void @foo() #0 { ret void } attributes #0 = { nounwind noinline ssp } llvm-svn: 175605
2013-02-20 08:21:42 +01:00
; CHECK: define void @test5(i8* %a) #0 {
Don't let arbitrary calls disrupt nested retain+release pairs if the retains and releases all use the same SSA pointer value. Also, don't let CFG hazards disrupt nested retain+release pair optimizations. llvm-svn: 137399
2011-08-12 02:26:31 +02:00
; CHECK-NEXT: entry:
Use references to attribute groups on the call/invoke instructions. Listing all of the attributes for the callee of a call/invoke instruction is way too much and makes the IR unreadable. Use references to attributes instead. llvm-svn: 175877
2013-02-22 10:09:42 +01:00
; CHECK-NEXT: tail call i8* @objc_retain(i8* %a) [[NUW]]
Don't let arbitrary calls disrupt nested retain+release pairs if the retains and releases all use the same SSA pointer value. Also, don't let CFG hazards disrupt nested retain+release pair optimizations. llvm-svn: 137399
2011-08-12 02:26:31 +02:00
; CHECK-NEXT: call void @callee()
; CHECK-NEXT: br label %loop
; CHECK-NOT: @objc_
; CHECK: exit:
; CHECK-NEXT: call void @use_pointer(i8* %a)
; CHECK-NEXT: call void @objc_release(i8* %a)
; CHECK-NEXT: ret void
; CHECK-NEXT: }
define void @test5(i8* %a) nounwind {
entry:
%outer = tail call i8* @objc_retain(i8* %a) nounwind
%inner = tail call i8* @objc_retain(i8* %a) nounwind
call void @callee()
br label %loop
loop:
br i1 undef, label %true, label %more
true:
br label %more
more:
br i1 undef, label %exit, label %loop
exit:
call void @use_pointer(i8* %a)
call void @objc_release(i8* %a) nounwind
call void @objc_release(i8* %a) nounwind, !clang.imprecise_release !0
ret void
}
Modify the LLVM assembly output so that it uses references to represent function attributes. This makes the LLVM assembly look better. E.g.: define void @foo() #0 { ret void } attributes #0 = { nounwind noinline ssp } llvm-svn: 175605
2013-02-20 08:21:42 +01:00
; CHECK: define void @test6(i8* %a) #0 {
Don't let arbitrary calls disrupt nested retain+release pairs if the retains and releases all use the same SSA pointer value. Also, don't let CFG hazards disrupt nested retain+release pair optimizations. llvm-svn: 137399
2011-08-12 02:26:31 +02:00
; CHECK-NEXT: entry:
Use references to attribute groups on the call/invoke instructions. Listing all of the attributes for the callee of a call/invoke instruction is way too much and makes the IR unreadable. Use references to attributes instead. llvm-svn: 175877
2013-02-22 10:09:42 +01:00
; CHECK-NEXT: tail call i8* @objc_retain(i8* %a) [[NUW]]
Don't let arbitrary calls disrupt nested retain+release pairs if the retains and releases all use the same SSA pointer value. Also, don't let CFG hazards disrupt nested retain+release pair optimizations. llvm-svn: 137399
2011-08-12 02:26:31 +02:00
; CHECK-NEXT: br label %loop
; CHECK-NOT: @objc_
; CHECK: exit:
; CHECK-NEXT: call void @use_pointer(i8* %a)
; CHECK-NEXT: call void @objc_release(i8* %a)
; CHECK-NEXT: ret void
; CHECK-NEXT: }
define void @test6(i8* %a) nounwind {
entry:
%outer = tail call i8* @objc_retain(i8* %a) nounwind
%inner = tail call i8* @objc_retain(i8* %a) nounwind
br label %loop
loop:
br i1 undef, label %true, label %more
true:
call void @callee()
br label %more
more:
br i1 undef, label %exit, label %loop
exit:
call void @use_pointer(i8* %a)
call void @objc_release(i8* %a) nounwind
call void @objc_release(i8* %a) nounwind, !clang.imprecise_release !0
ret void
}
Modify the LLVM assembly output so that it uses references to represent function attributes. This makes the LLVM assembly look better. E.g.: define void @foo() #0 { ret void } attributes #0 = { nounwind noinline ssp } llvm-svn: 175605
2013-02-20 08:21:42 +01:00
; CHECK: define void @test7(i8* %a) #0 {
Don't let arbitrary calls disrupt nested retain+release pairs if the retains and releases all use the same SSA pointer value. Also, don't let CFG hazards disrupt nested retain+release pair optimizations. llvm-svn: 137399
2011-08-12 02:26:31 +02:00
; CHECK-NEXT: entry:
Use references to attribute groups on the call/invoke instructions. Listing all of the attributes for the callee of a call/invoke instruction is way too much and makes the IR unreadable. Use references to attributes instead. llvm-svn: 175877
2013-02-22 10:09:42 +01:00
; CHECK-NEXT: tail call i8* @objc_retain(i8* %a) [[NUW]]
Don't let arbitrary calls disrupt nested retain+release pairs if the retains and releases all use the same SSA pointer value. Also, don't let CFG hazards disrupt nested retain+release pair optimizations. llvm-svn: 137399
2011-08-12 02:26:31 +02:00
; CHECK-NEXT: call void @callee()
; CHECK-NEXT: br label %loop
; CHECK-NOT: @objc_
; CHECK: exit:
; CHECK-NEXT: call void @objc_release(i8* %a)
; CHECK-NEXT: ret void
; CHECK-NEXT: }
define void @test7(i8* %a) nounwind {
entry:
%outer = tail call i8* @objc_retain(i8* %a) nounwind
%inner = tail call i8* @objc_retain(i8* %a) nounwind
call void @callee()
br label %loop
loop:
br i1 undef, label %true, label %more
true:
call void @use_pointer(i8* %a)
br label %more
more:
br i1 undef, label %exit, label %loop
exit:
call void @objc_release(i8* %a) nounwind
call void @objc_release(i8* %a) nounwind, !clang.imprecise_release !0
ret void
}
Modify the LLVM assembly output so that it uses references to represent function attributes. This makes the LLVM assembly look better. E.g.: define void @foo() #0 { ret void } attributes #0 = { nounwind noinline ssp } llvm-svn: 175605
2013-02-20 08:21:42 +01:00
; CHECK: define void @test8(i8* %a) #0 {
Don't let arbitrary calls disrupt nested retain+release pairs if the retains and releases all use the same SSA pointer value. Also, don't let CFG hazards disrupt nested retain+release pair optimizations. llvm-svn: 137399
2011-08-12 02:26:31 +02:00
; CHECK-NEXT: entry:
Use references to attribute groups on the call/invoke instructions. Listing all of the attributes for the callee of a call/invoke instruction is way too much and makes the IR unreadable. Use references to attributes instead. llvm-svn: 175877
2013-02-22 10:09:42 +01:00
; CHECK-NEXT: tail call i8* @objc_retain(i8* %a) [[NUW]]
Don't let arbitrary calls disrupt nested retain+release pairs if the retains and releases all use the same SSA pointer value. Also, don't let CFG hazards disrupt nested retain+release pair optimizations. llvm-svn: 137399
2011-08-12 02:26:31 +02:00
; CHECK-NEXT: br label %loop
; CHECK-NOT: @objc_
; CHECK: exit:
; CHECK-NEXT: call void @objc_release(i8* %a)
; CHECK-NEXT: ret void
; CHECK-NEXT: }
define void @test8(i8* %a) nounwind {
entry:
%outer = tail call i8* @objc_retain(i8* %a) nounwind
%inner = tail call i8* @objc_retain(i8* %a) nounwind
br label %loop
loop:
br i1 undef, label %true, label %more
true:
call void @callee()
call void @use_pointer(i8* %a)
br label %more
more:
br i1 undef, label %exit, label %loop
exit:
call void @objc_release(i8* %a) nounwind
call void @objc_release(i8* %a) nounwind, !clang.imprecise_release !0
ret void
}
Modify the LLVM assembly output so that it uses references to represent function attributes. This makes the LLVM assembly look better. E.g.: define void @foo() #0 { ret void } attributes #0 = { nounwind noinline ssp } llvm-svn: 175605
2013-02-20 08:21:42 +01:00
; CHECK: define void @test9(i8* %a) #0 {
Don't let arbitrary calls disrupt nested retain+release pairs if the retains and releases all use the same SSA pointer value. Also, don't let CFG hazards disrupt nested retain+release pair optimizations. llvm-svn: 137399
2011-08-12 02:26:31 +02:00
; CHECK-NEXT: entry:
; CHECK-NEXT: br label %loop
; CHECK-NOT: @objc_
; CHECK: exit:
; CHECK-NEXT: ret void
; CHECK-NEXT: }
define void @test9(i8* %a) nounwind {
entry:
%outer = tail call i8* @objc_retain(i8* %a) nounwind
%inner = tail call i8* @objc_retain(i8* %a) nounwind
br label %loop
loop:
br i1 undef, label %true, label %more
true:
call void @use_pointer(i8* %a)
br label %more
more:
br i1 undef, label %exit, label %loop
exit:
call void @objc_release(i8* %a) nounwind
call void @objc_release(i8* %a) nounwind, !clang.imprecise_release !0
ret void
}
Modify the LLVM assembly output so that it uses references to represent function attributes. This makes the LLVM assembly look better. E.g.: define void @foo() #0 { ret void } attributes #0 = { nounwind noinline ssp } llvm-svn: 175605
2013-02-20 08:21:42 +01:00
; CHECK: define void @test10(i8* %a) #0 {
Don't let arbitrary calls disrupt nested retain+release pairs if the retains and releases all use the same SSA pointer value. Also, don't let CFG hazards disrupt nested retain+release pair optimizations. llvm-svn: 137399
2011-08-12 02:26:31 +02:00
; CHECK-NEXT: entry:
; CHECK-NEXT: br label %loop
; CHECK-NOT: @objc_
; CHECK: exit:
; CHECK-NEXT: ret void
; CHECK-NEXT: }
define void @test10(i8* %a) nounwind {
entry:
%outer = tail call i8* @objc_retain(i8* %a) nounwind
%inner = tail call i8* @objc_retain(i8* %a) nounwind
br label %loop
loop:
br i1 undef, label %true, label %more
true:
call void @callee()
br label %more
more:
br i1 undef, label %exit, label %loop
exit:
call void @objc_release(i8* %a) nounwind
call void @objc_release(i8* %a) nounwind, !clang.imprecise_release !0
ret void
}
Modify the LLVM assembly output so that it uses references to represent function attributes. This makes the LLVM assembly look better. E.g.: define void @foo() #0 { ret void } attributes #0 = { nounwind noinline ssp } llvm-svn: 175605
2013-02-20 08:21:42 +01:00
; CHECK: define void @test11(i8* %a) #0 {
Don't let arbitrary calls disrupt nested retain+release pairs if the retains and releases all use the same SSA pointer value. Also, don't let CFG hazards disrupt nested retain+release pair optimizations. llvm-svn: 137399
2011-08-12 02:26:31 +02:00
; CHECK-NEXT: entry:
; CHECK-NEXT: br label %loop
; CHECK-NOT: @objc_
; CHECK: exit:
; CHECK-NEXT: ret void
; CHECK-NEXT: }
define void @test11(i8* %a) nounwind {
entry:
%outer = tail call i8* @objc_retain(i8* %a) nounwind
%inner = tail call i8* @objc_retain(i8* %a) nounwind
br label %loop
loop:
br i1 undef, label %true, label %more
true:
br label %more
more:
br i1 undef, label %exit, label %loop
exit:
call void @objc_release(i8* %a) nounwind
call void @objc_release(i8* %a) nounwind, !clang.imprecise_release !0
ret void
}
; Don't delete anything if they're not balanced.
Modify the LLVM assembly output so that it uses references to represent function attributes. This makes the LLVM assembly look better. E.g.: define void @foo() #0 { ret void } attributes #0 = { nounwind noinline ssp } llvm-svn: 175605
2013-02-20 08:21:42 +01:00
; CHECK: define void @test12(i8* %a) #0 {
Don't let arbitrary calls disrupt nested retain+release pairs if the retains and releases all use the same SSA pointer value. Also, don't let CFG hazards disrupt nested retain+release pair optimizations. llvm-svn: 137399
2011-08-12 02:26:31 +02:00
; CHECK-NEXT: entry:
Use references to attribute groups on the call/invoke instructions. Listing all of the attributes for the callee of a call/invoke instruction is way too much and makes the IR unreadable. Use references to attributes instead. llvm-svn: 175877
2013-02-22 10:09:42 +01:00
; CHECK-NEXT: %outer = tail call i8* @objc_retain(i8* %a) [[NUW]]
; CHECK-NEXT: %inner = tail call i8* @objc_retain(i8* %a) [[NUW]]
Don't let arbitrary calls disrupt nested retain+release pairs if the retains and releases all use the same SSA pointer value. Also, don't let CFG hazards disrupt nested retain+release pair optimizations. llvm-svn: 137399
2011-08-12 02:26:31 +02:00
; CHECK-NEXT: br label %loop
; CHECK-NOT: @objc_
; CHECK: exit:
Use references to attribute groups on the call/invoke instructions. Listing all of the attributes for the callee of a call/invoke instruction is way too much and makes the IR unreadable. Use references to attributes instead. llvm-svn: 175877
2013-02-22 10:09:42 +01:00
; CHECK-NEXT: call void @objc_release(i8* %a) [[NUW]]
; CHECK-NEXT: call void @objc_release(i8* %a) [[NUW]], !clang.imprecise_release !0
Don't let arbitrary calls disrupt nested retain+release pairs if the retains and releases all use the same SSA pointer value. Also, don't let CFG hazards disrupt nested retain+release pair optimizations. llvm-svn: 137399
2011-08-12 02:26:31 +02:00
; CHECK-NEXT: ret void
; CHECK-NEXT: }
define void @test12(i8* %a) nounwind {
entry:
%outer = tail call i8* @objc_retain(i8* %a) nounwind
%inner = tail call i8* @objc_retain(i8* %a) nounwind
br label %loop
loop:
br i1 undef, label %true, label %more
true:
ret void
more:
br i1 undef, label %exit, label %loop
exit:
call void @objc_release(i8* %a) nounwind
call void @objc_release(i8* %a) nounwind, !clang.imprecise_release !0
ret void
}
[objc-arc] Do not mismatch up retains inside a for loop with releases outside said for loop in the presense of differing provenance caused by escaping blocks. This occurs due to an alloca representing a separate ownership from the original pointer. Thus consider the following pseudo-IR: objc_retain(%a) for (...) { objc_retain(%a) %block <- %a F(%block) objc_release(%block) } objc_release(%a) From the perspective of the optimizer, the %block is a separate provenance from the original %a. Thus the optimizer pairs up the inner retain for %a and the outer release from %a, resulting in segfaults. This is fixed by noting that the signature of a mismatch of retain/releases inside the for loop is a Use/CanRelease top down with an None bottom up (since bottom up the Retain-CanRelease-Use-Release sequence is completed by the inner objc_retain, but top down due to the differing provenance from the objc_release said sequence is not completed). In said case in CheckForCFGHazards, we now clear the state of %a implying that no pairing will occur. Additionally a test case is included. rdar://12969722 llvm-svn: 179747
2013-04-18 07:39:45 +02:00
; Do not improperly pair retains in a for loop with releases outside of a for
; loop when the proper pairing is disguised by a separate provenance represented
; by an alloca.
; rdar://12969722
; CHECK: define void @test13(i8* %a) [[NUW]] {
; CHECK: entry:
; CHECK: tail call i8* @objc_retain(i8* %a) [[NUW]]
; CHECK: loop:
; CHECK: tail call i8* @objc_retain(i8* %a) [[NUW]]
; CHECK: call void @block_callee
; CHECK: call void @objc_release(i8* %reloaded_a) [[NUW]]
; CHECK: exit:
; CHECK: call void @objc_release(i8* %a) [[NUW]]
; CHECK: }
define void @test13(i8* %a) nounwind {
entry:
%block = alloca i8*
%a1 = tail call i8* @objc_retain(i8* %a) nounwind
br label %loop
loop:
%a2 = tail call i8* @objc_retain(i8* %a) nounwind
store i8* %a, i8** %block, align 8
%casted_block = bitcast i8** %block to void ()*
call void @block_callee(void ()* %casted_block)
[opaque pointer type] Add textual IR support for explicit type parameter to load instruction Essentially the same as the GEP change in r230786. A similar migration script can be used to update test cases, though a few more test case improvements/changes were required this time around: (r229269-r229278) import fileinput import sys import re pat = re.compile(r"((?:=|:|^)\s*load (?:atomic )?(?:volatile )?(.*?))(| addrspace\(\d+\) *)\*($| *(?:%|@|null|undef|blockaddress|getelementptr|addrspacecast|bitcast|inttoptr|\[\[[a-zA-Z]|\{\{).*$)") for line in sys.stdin: sys.stdout.write(re.sub(pat, r"\1, \2\3*\4", line)) Reviewers: rafael, dexonsmith, grosser Differential Revision: http://reviews.llvm.org/D7649 llvm-svn: 230794
2015-02-27 22:17:42 +01:00
%reloaded_a = load i8*, i8** %block, align 8
[objc-arc] Do not mismatch up retains inside a for loop with releases outside said for loop in the presense of differing provenance caused by escaping blocks. This occurs due to an alloca representing a separate ownership from the original pointer. Thus consider the following pseudo-IR: objc_retain(%a) for (...) { objc_retain(%a) %block <- %a F(%block) objc_release(%block) } objc_release(%a) From the perspective of the optimizer, the %block is a separate provenance from the original %a. Thus the optimizer pairs up the inner retain for %a and the outer release from %a, resulting in segfaults. This is fixed by noting that the signature of a mismatch of retain/releases inside the for loop is a Use/CanRelease top down with an None bottom up (since bottom up the Retain-CanRelease-Use-Release sequence is completed by the inner objc_retain, but top down due to the differing provenance from the objc_release said sequence is not completed). In said case in CheckForCFGHazards, we now clear the state of %a implying that no pairing will occur. Additionally a test case is included. rdar://12969722 llvm-svn: 179747
2013-04-18 07:39:45 +02:00
call void @objc_release(i8* %reloaded_a) nounwind, !clang.imprecise_release !0
br i1 undef, label %loop, label %exit
exit:
call void @objc_release(i8* %a) nounwind, !clang.imprecise_release !0
ret void
}
Use references to attribute groups on the call/invoke instructions. Listing all of the attributes for the callee of a call/invoke instruction is way too much and makes the IR unreadable. Use references to attributes instead. llvm-svn: 175877
2013-02-22 10:09:42 +01:00
; CHECK: attributes [[NUW]] = { nounwind }
Modify the LLVM assembly output so that it uses references to represent function attributes. This makes the LLVM assembly look better. E.g.: define void @foo() #0 { ret void } attributes #0 = { nounwind noinline ssp } llvm-svn: 175605
2013-02-20 08:21:42 +01:00
IR: Make metadata typeless in assembly Now that `Metadata` is typeless, reflect that in the assembly. These are the matching assembly changes for the metadata/value split in r223802. - Only use the `metadata` type when referencing metadata from a call intrinsic -- i.e., only when it's used as a `Value`. - Stop pretending that `ValueAsMetadata` is wrapped in an `MDNode` when referencing it from call intrinsics. So, assembly like this: define @foo(i32 %v) { call void @llvm.foo(metadata !{i32 %v}, metadata !0) call void @llvm.foo(metadata !{i32 7}, metadata !0) call void @llvm.foo(metadata !1, metadata !0) call void @llvm.foo(metadata !3, metadata !0) call void @llvm.foo(metadata !{metadata !3}, metadata !0) ret void, !bar !2 } !0 = metadata !{metadata !2} !1 = metadata !{i32* @global} !2 = metadata !{metadata !3} !3 = metadata !{} turns into this: define @foo(i32 %v) { call void @llvm.foo(metadata i32 %v, metadata !0) call void @llvm.foo(metadata i32 7, metadata !0) call void @llvm.foo(metadata i32* @global, metadata !0) call void @llvm.foo(metadata !3, metadata !0) call void @llvm.foo(metadata !{!3}, metadata !0) ret void, !bar !2 } !0 = !{!2} !1 = !{i32* @global} !2 = !{!3} !3 = !{} I wrote an upgrade script that handled almost all of the tests in llvm and many of the tests in cfe (even handling many `CHECK` lines). I've attached it (or will attach it in a moment if you're speedy) to PR21532 to help everyone update their out-of-tree testcases. This is part of PR21532. llvm-svn: 224257
2014-12-15 20:07:53 +01:00
!0 = !{}
Reference in New Issue Copy Permalink