2008-03-10 15:21:50 +08:00
|
|
|
; RUN: llvm-as < %s > %t.bc
|
|
|
|
|
; RUN: llvm-as < %p/testlink2.ll > %t2.bc
|
Land the long talked about "type system rewrite" patch. This
patch brings numerous advantages to LLVM. One way to look at it
is through diffstat:
109 files changed, 3005 insertions(+), 5906 deletions(-)
Removing almost 3K lines of code is a good thing. Other advantages
include:
1. Value::getType() is a simple load that can be CSE'd, not a mutating
union-find operation.
2. Types a uniqued and never move once created, defining away PATypeHolder.
3. Structs can be "named" now, and their name is part of the identity that
uniques them. This means that the compiler doesn't merge them structurally
which makes the IR much less confusing.
4. Now that there is no way to get a cycle in a type graph without a named
struct type, "upreferences" go away.
5. Type refinement is completely gone, which should make LTO much MUCH faster
in some common cases with C++ code.
6. Types are now generally immutable, so we can use "Type *" instead
"const Type *" everywhere.
Downsides of this patch are that it removes some functions from the C API,
so people using those will have to upgrade to (not yet added) new API.
"LLVM 3.0" is the right time to do this.
There are still some cleanups pending after this, this patch is large enough
as-is.
llvm-svn: 134829
2011-07-10 01:41:24 +08:00
|
|
|
; RUN: llvm-link %t.bc %t2.bc -S | FileCheck %s
|
|
|
|
|
|
|
|
|
|
; CHECK: %Ty2 = type { %Ty1* }
|
|
|
|
|
; CHECK: %Ty1 = type { %Ty2* }
|
|
|
|
|
%Ty1 = type opaque
|
|
|
|
|
%Ty2 = type { %Ty1* }
|
|
|
|
|
|
|
|
|
|
; CHECK: %intlist = type { %intlist*, i32 }
|
|
|
|
|
%intlist = type { %intlist*, i32 }
|
|
|
|
|
|
|
|
|
|
; The uses of intlist in the other file should be remapped.
|
|
|
|
|
; CHECK-NOT: {{%intlist.[0-9]}}
|
|
|
|
|
|
2013-01-10 18:49:36 +08:00
|
|
|
; CHECK: %VecSize = type { <5 x i32> }
|
|
|
|
|
; CHECK: %VecSize.{{[0-9]}} = type { <10 x i32> }
|
|
|
|
|
%VecSize = type { <5 x i32> }
|
|
|
|
|
|
Land the long talked about "type system rewrite" patch. This
patch brings numerous advantages to LLVM. One way to look at it
is through diffstat:
109 files changed, 3005 insertions(+), 5906 deletions(-)
Removing almost 3K lines of code is a good thing. Other advantages
include:
1. Value::getType() is a simple load that can be CSE'd, not a mutating
union-find operation.
2. Types a uniqued and never move once created, defining away PATypeHolder.
3. Structs can be "named" now, and their name is part of the identity that
uniques them. This means that the compiler doesn't merge them structurally
which makes the IR much less confusing.
4. Now that there is no way to get a cycle in a type graph without a named
struct type, "upreferences" go away.
5. Type refinement is completely gone, which should make LTO much MUCH faster
in some common cases with C++ code.
6. Types are now generally immutable, so we can use "Type *" instead
"const Type *" everywhere.
Downsides of this patch are that it removes some functions from the C API,
so people using those will have to upgrade to (not yet added) new API.
"LLVM 3.0" is the right time to do this.
There are still some cleanups pending after this, this patch is large enough
as-is.
llvm-svn: 134829
2011-07-10 01:41:24 +08:00
|
|
|
%Struct1 = type opaque
|
|
|
|
|
@S1GV = external global %Struct1*
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
@GVTy1 = external global %Ty1*
|
|
|
|
|
@GVTy2 = global %Ty2* null
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
; This should stay the same
|
|
|
|
|
; CHECK: @MyIntList = global %intlist { %intlist* null, i32 17 }
|
|
|
|
|
@MyIntList = global %intlist { %intlist* null, i32 17 }
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
; Nothing to link here.
|
|
|
|
|
|
|
|
|
|
; CHECK: @0 = external global i32
|
|
|
|
|
@0 = external global i32
|
|
|
|
|
; CHECK: @Inte = global i32 1
|
|
|
|
|
@Inte = global i32 1
|
|
|
|
|
|
|
|
|
|
; Intern1 is intern in both files, rename testlink2's.
|
|
|
|
|
; CHECK: @Intern1 = internal constant i32 42
|
|
|
|
|
@Intern1 = internal constant i32 42
|
|
|
|
|
|
|
|
|
|
; This should get renamed since there is a definition that is non-internal in
|
|
|
|
|
; the other module.
|
|
|
|
|
; CHECK: @Intern2{{[0-9]+}} = internal constant i32 792
|
|
|
|
|
@Intern2 = internal constant i32 792
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
; CHECK: @MyVarPtr = linkonce global { i32* } { i32* @MyVar }
|
|
|
|
|
@MyVarPtr = linkonce global { i32* } { i32* @MyVar }
|
|
|
|
|
|
|
|
|
|
; CHECK: @MyVar = global i32 4
|
|
|
|
|
@MyVar = external global i32
|
|
|
|
|
|
|
|
|
|
; Take value from other module.
|
|
|
|
|
; CHECK: AConst = constant i32 1234
|
|
|
|
|
@AConst = linkonce constant i32 123
|
|
|
|
|
|
|
|
|
|
; Renamed version of Intern1.
|
|
|
|
|
; CHECK: @Intern1{{[0-9]+}} = internal constant i32 52
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
; Globals linked from testlink2.
|
|
|
|
|
; CHECK: @Intern2 = constant i32 12345
|
|
|
|
|
|
|
|
|
|
; CHECK: @MyIntListPtr = constant
|
|
|
|
|
; CHECK: @1 = constant i32 412
|
2001-10-13 15:15:38 +08:00
|
|
|
|
2008-03-10 15:21:50 +08:00
|
|
|
|
|
|
|
|
declare i32 @foo(i32)
|
|
|
|
|
|
|
|
|
|
declare void @print(i32)
|
|
|
|
|
|
|
|
|
|
define void @main() {
|
Land the long talked about "type system rewrite" patch. This
patch brings numerous advantages to LLVM. One way to look at it
is through diffstat:
109 files changed, 3005 insertions(+), 5906 deletions(-)
Removing almost 3K lines of code is a good thing. Other advantages
include:
1. Value::getType() is a simple load that can be CSE'd, not a mutating
union-find operation.
2. Types a uniqued and never move once created, defining away PATypeHolder.
3. Structs can be "named" now, and their name is part of the identity that
uniques them. This means that the compiler doesn't merge them structurally
which makes the IR much less confusing.
4. Now that there is no way to get a cycle in a type graph without a named
struct type, "upreferences" go away.
5. Type refinement is completely gone, which should make LTO much MUCH faster
in some common cases with C++ code.
6. Types are now generally immutable, so we can use "Type *" instead
"const Type *" everywhere.
Downsides of this patch are that it removes some functions from the C API,
so people using those will have to upgrade to (not yet added) new API.
"LLVM 3.0" is the right time to do this.
There are still some cleanups pending after this, this patch is large enough
as-is.
llvm-svn: 134829
2011-07-10 01:41:24 +08:00
|
|
|
%v1 = load i32* @MyVar
|
|
|
|
|
call void @print(i32 %v1)
|
|
|
|
|
%idx = getelementptr %intlist* @MyIntList, i64 0, i32 1
|
|
|
|
|
%v2 = load i32* %idx
|
|
|
|
|
call void @print(i32 %v2)
|
|
|
|
|
%1 = call i32 @foo(i32 5)
|
|
|
|
|
%v3 = load i32* @MyVar
|
|
|
|
|
call void @print(i32 %v3)
|
|
|
|
|
%v4 = load i32* %idx
|
|
|
|
|
call void @print(i32 %v4)
|
|
|
|
|
ret void
|
2008-03-10 15:21:50 +08:00
|
|
|
}
|
2001-10-13 15:15:38 +08:00
|
|
|
|
2008-03-10 15:21:50 +08:00
|
|
|
define internal void @testintern() {
|
Land the long talked about "type system rewrite" patch. This
patch brings numerous advantages to LLVM. One way to look at it
is through diffstat:
109 files changed, 3005 insertions(+), 5906 deletions(-)
Removing almost 3K lines of code is a good thing. Other advantages
include:
1. Value::getType() is a simple load that can be CSE'd, not a mutating
union-find operation.
2. Types a uniqued and never move once created, defining away PATypeHolder.
3. Structs can be "named" now, and their name is part of the identity that
uniques them. This means that the compiler doesn't merge them structurally
which makes the IR much less confusing.
4. Now that there is no way to get a cycle in a type graph without a named
struct type, "upreferences" go away.
5. Type refinement is completely gone, which should make LTO much MUCH faster
in some common cases with C++ code.
6. Types are now generally immutable, so we can use "Type *" instead
"const Type *" everywhere.
Downsides of this patch are that it removes some functions from the C API,
so people using those will have to upgrade to (not yet added) new API.
"LLVM 3.0" is the right time to do this.
There are still some cleanups pending after this, this patch is large enough
as-is.
llvm-svn: 134829
2011-07-10 01:41:24 +08:00
|
|
|
ret void
|
2008-03-10 15:21:50 +08:00
|
|
|
}
|
2001-10-13 15:15:38 +08:00
|
|
|
|
2008-03-10 15:21:50 +08:00
|
|
|
define internal void @Testintern() {
|
Land the long talked about "type system rewrite" patch. This
patch brings numerous advantages to LLVM. One way to look at it
is through diffstat:
109 files changed, 3005 insertions(+), 5906 deletions(-)
Removing almost 3K lines of code is a good thing. Other advantages
include:
1. Value::getType() is a simple load that can be CSE'd, not a mutating
union-find operation.
2. Types a uniqued and never move once created, defining away PATypeHolder.
3. Structs can be "named" now, and their name is part of the identity that
uniques them. This means that the compiler doesn't merge them structurally
which makes the IR much less confusing.
4. Now that there is no way to get a cycle in a type graph without a named
struct type, "upreferences" go away.
5. Type refinement is completely gone, which should make LTO much MUCH faster
in some common cases with C++ code.
6. Types are now generally immutable, so we can use "Type *" instead
"const Type *" everywhere.
Downsides of this patch are that it removes some functions from the C API,
so people using those will have to upgrade to (not yet added) new API.
"LLVM 3.0" is the right time to do this.
There are still some cleanups pending after this, this patch is large enough
as-is.
llvm-svn: 134829
2011-07-10 01:41:24 +08:00
|
|
|
ret void
|
2008-03-10 15:21:50 +08:00
|
|
|
}
|
2001-11-27 03:17:06 +08:00
|
|
|
|
2008-03-10 15:21:50 +08:00
|
|
|
define void @testIntern() {
|
Land the long talked about "type system rewrite" patch. This
patch brings numerous advantages to LLVM. One way to look at it
is through diffstat:
109 files changed, 3005 insertions(+), 5906 deletions(-)
Removing almost 3K lines of code is a good thing. Other advantages
include:
1. Value::getType() is a simple load that can be CSE'd, not a mutating
union-find operation.
2. Types a uniqued and never move once created, defining away PATypeHolder.
3. Structs can be "named" now, and their name is part of the identity that
uniques them. This means that the compiler doesn't merge them structurally
which makes the IR much less confusing.
4. Now that there is no way to get a cycle in a type graph without a named
struct type, "upreferences" go away.
5. Type refinement is completely gone, which should make LTO much MUCH faster
in some common cases with C++ code.
6. Types are now generally immutable, so we can use "Type *" instead
"const Type *" everywhere.
Downsides of this patch are that it removes some functions from the C API,
so people using those will have to upgrade to (not yet added) new API.
"LLVM 3.0" is the right time to do this.
There are still some cleanups pending after this, this patch is large enough
as-is.
llvm-svn: 134829
2011-07-10 01:41:24 +08:00
|
|
|
ret void
|
2008-03-10 15:21:50 +08:00
|
|
|
}
|
2013-01-10 18:49:36 +08:00
|
|
|
|
|
|
|
|
declare void @VecSizeCrash(%VecSize)
|