2010-02-01 18:27:38 +00:00
|
|
|
; "PLAIN" - No optimizations. This tests the target-independent
|
|
|
|
|
; constant folder.
|
2010-01-28 18:08:26 +00:00
|
|
|
; RUN: opt -S -o - < %s | FileCheck --check-prefix=PLAIN %s
|
2010-02-01 18:27:38 +00:00
|
|
|
|
|
|
|
|
; "OPT" - Optimizations but no targetdata. This tests target-independent
|
|
|
|
|
; folding in the optimizers.
|
2010-01-28 18:08:26 +00:00
|
|
|
; RUN: opt -S -o - -instcombine -globalopt < %s | FileCheck --check-prefix=OPT %s
|
|
|
|
|
|
2010-02-01 18:27:38 +00:00
|
|
|
; "TO" - Optimizations and targetdata. This tests target-dependent
|
|
|
|
|
; folding in the optimizers.
|
|
|
|
|
; RUN: opt -S -o - -instcombine -globalopt -default-data-layout="e-p:64:64:64-i1:8:8-i8:8:8-i16:16:16-i32:32:32-i64:64:64" < %s | FileCheck --check-prefix=TO %s
|
|
|
|
|
|
|
|
|
|
; "SCEV" - ScalarEvolution but no targetdata.
|
|
|
|
|
; RUN: opt -analyze -scalar-evolution < %s | FileCheck --check-prefix=SCEV %s
|
|
|
|
|
|
|
|
|
|
|
2010-01-28 18:08:26 +00:00
|
|
|
; The automatic constant folder in opt does not have targetdata access, so
|
|
|
|
|
; it can't fold gep arithmetic, in general. However, the constant folder run
|
2010-02-01 18:27:38 +00:00
|
|
|
; from instcombine and global opt can use targetdata.
|
2010-01-28 18:08:26 +00:00
|
|
|
|
|
|
|
|
; PLAIN: @G8 = global i8* getelementptr (i8* inttoptr (i32 1 to i8*), i32 -1)
|
|
|
|
|
; PLAIN: @G1 = global i1* getelementptr (i1* inttoptr (i32 1 to i1*), i32 -1)
|
|
|
|
|
; PLAIN: @F8 = global i8* getelementptr (i8* inttoptr (i32 1 to i8*), i32 -2)
|
|
|
|
|
; PLAIN: @F1 = global i1* getelementptr (i1* inttoptr (i32 1 to i1*), i32 -2)
|
|
|
|
|
; PLAIN: @H8 = global i8* getelementptr (i8* null, i32 -1)
|
|
|
|
|
; PLAIN: @H1 = global i1* getelementptr (i1* null, i32 -1)
|
2010-02-01 18:27:38 +00:00
|
|
|
; OPT: @G8 = global i8* getelementptr (i8* inttoptr (i32 1 to i8*), i32 -1)
|
|
|
|
|
; OPT: @G1 = global i1* getelementptr (i1* inttoptr (i32 1 to i1*), i32 -1)
|
|
|
|
|
; OPT: @F8 = global i8* getelementptr (i8* inttoptr (i32 1 to i8*), i32 -2)
|
|
|
|
|
; OPT: @F1 = global i1* getelementptr (i1* inttoptr (i32 1 to i1*), i32 -2)
|
|
|
|
|
; OPT: @H8 = global i8* getelementptr (i8* null, i32 -1)
|
|
|
|
|
; OPT: @H1 = global i1* getelementptr (i1* null, i32 -1)
|
|
|
|
|
; TO: @G8 = global i8* null
|
|
|
|
|
; TO: @G1 = global i1* null
|
|
|
|
|
; TO: @F8 = global i8* inttoptr (i64 -1 to i8*)
|
|
|
|
|
; TO: @F1 = global i1* inttoptr (i64 -1 to i1*)
|
|
|
|
|
; TO: @H8 = global i8* inttoptr (i64 -1 to i8*)
|
|
|
|
|
; TO: @H1 = global i1* inttoptr (i64 -1 to i1*)
|
|
|
|
|
|
|
|
|
|
@G8 = global i8* getelementptr (i8* inttoptr (i32 1 to i8*), i32 -1)
|
|
|
|
|
@G1 = global i1* getelementptr (i1* inttoptr (i32 1 to i1*), i32 -1)
|
|
|
|
|
@F8 = global i8* getelementptr (i8* inttoptr (i32 1 to i8*), i32 -2)
|
|
|
|
|
@F1 = global i1* getelementptr (i1* inttoptr (i32 1 to i1*), i32 -2)
|
|
|
|
|
@H8 = global i8* getelementptr (i8* inttoptr (i32 0 to i8*), i32 -1)
|
|
|
|
|
@H1 = global i1* getelementptr (i1* inttoptr (i32 0 to i1*), i32 -1)
|
|
|
|
|
|
|
|
|
|
; The target-independent folder should be able to do some clever
|
|
|
|
|
; simplifications on sizeof, alignof, and offsetof expressions. The
|
|
|
|
|
; target-dependent folder should fold these down to constants.
|
|
|
|
|
|
|
|
|
|
; PLAIN: @a = constant i64 mul (i64 ptrtoint (double* getelementptr (double* null, i32 1) to i64), i64 2310)
|
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.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@134829 91177308-0d34-0410-b5e6-96231b3b80d8
2011-07-09 17:41:24 +00:00
|
|
|
; PLAIN: @b = constant i64 ptrtoint (double* getelementptr ({ i1, double }* null, i64 0, i32 1) to i64)
|
2010-02-01 18:27:38 +00:00
|
|
|
; PLAIN: @c = constant i64 mul nuw (i64 ptrtoint (double* getelementptr (double* null, i32 1) to i64), i64 2)
|
|
|
|
|
; PLAIN: @d = constant i64 mul nuw (i64 ptrtoint (double* getelementptr (double* null, i32 1) to i64), i64 11)
|
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.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@134829 91177308-0d34-0410-b5e6-96231b3b80d8
2011-07-09 17:41:24 +00:00
|
|
|
; PLAIN: @e = constant i64 ptrtoint (double* getelementptr ({ double, float, double, double }* null, i64 0, i32 2) to i64)
|
2010-02-01 18:27:38 +00:00
|
|
|
; PLAIN: @f = constant i64 1
|
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.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@134829 91177308-0d34-0410-b5e6-96231b3b80d8
2011-07-09 17:41:24 +00:00
|
|
|
; PLAIN: @g = constant i64 ptrtoint (double* getelementptr ({ i1, double }* null, i64 0, i32 1) to i64)
|
2010-02-10 06:13:07 +00:00
|
|
|
; PLAIN: @h = constant i64 ptrtoint (i1** getelementptr (i1** null, i32 1) to i64)
|
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.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@134829 91177308-0d34-0410-b5e6-96231b3b80d8
2011-07-09 17:41:24 +00:00
|
|
|
; PLAIN: @i = constant i64 ptrtoint (i1** getelementptr ({ i1, i1* }* null, i64 0, i32 1) to i64)
|
2010-02-01 18:27:38 +00:00
|
|
|
; OPT: @a = constant i64 mul (i64 ptrtoint (double* getelementptr (double* null, i32 1) to i64), i64 2310)
|
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.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@134829 91177308-0d34-0410-b5e6-96231b3b80d8
2011-07-09 17:41:24 +00:00
|
|
|
; OPT: @b = constant i64 ptrtoint (double* getelementptr ({ i1, double }* null, i64 0, i32 1) to i64)
|
2010-02-01 18:27:38 +00:00
|
|
|
; OPT: @c = constant i64 mul (i64 ptrtoint (double* getelementptr (double* null, i32 1) to i64), i64 2)
|
|
|
|
|
; OPT: @d = constant i64 mul (i64 ptrtoint (double* getelementptr (double* null, i32 1) to i64), i64 11)
|
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.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@134829 91177308-0d34-0410-b5e6-96231b3b80d8
2011-07-09 17:41:24 +00:00
|
|
|
; OPT: @e = constant i64 ptrtoint (double* getelementptr ({ double, float, double, double }* null, i64 0, i32 2) to i64)
|
2010-02-01 18:27:38 +00:00
|
|
|
; OPT: @f = constant i64 1
|
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.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@134829 91177308-0d34-0410-b5e6-96231b3b80d8
2011-07-09 17:41:24 +00:00
|
|
|
; OPT: @g = constant i64 ptrtoint (double* getelementptr ({ i1, double }* null, i64 0, i32 1) to i64)
|
2010-02-10 06:13:07 +00:00
|
|
|
; OPT: @h = constant i64 ptrtoint (i1** getelementptr (i1** null, i32 1) to i64)
|
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.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@134829 91177308-0d34-0410-b5e6-96231b3b80d8
2011-07-09 17:41:24 +00:00
|
|
|
; OPT: @i = constant i64 ptrtoint (i1** getelementptr ({ i1, i1* }* null, i64 0, i32 1) to i64)
|
2010-02-01 18:27:38 +00:00
|
|
|
; TO: @a = constant i64 18480
|
|
|
|
|
; TO: @b = constant i64 8
|
|
|
|
|
; TO: @c = constant i64 16
|
|
|
|
|
; TO: @d = constant i64 88
|
|
|
|
|
; TO: @e = constant i64 16
|
|
|
|
|
; TO: @f = constant i64 1
|
2010-02-02 01:41:39 +00:00
|
|
|
; TO: @g = constant i64 8
|
2010-02-10 06:13:07 +00:00
|
|
|
; TO: @h = constant i64 8
|
|
|
|
|
; TO: @i = constant i64 8
|
2010-02-01 18:27:38 +00:00
|
|
|
|
|
|
|
|
@a = constant i64 mul (i64 3, i64 mul (i64 ptrtoint ({[7 x double], [7 x double]}* getelementptr ({[7 x double], [7 x double]}* null, i64 11) to i64), i64 5))
|
|
|
|
|
@b = constant i64 ptrtoint ([13 x double]* getelementptr ({i1, [13 x double]}* null, i64 0, i32 1) to i64)
|
|
|
|
|
@c = constant i64 ptrtoint (double* getelementptr ({double, double, double, double}* null, i64 0, i32 2) to i64)
|
|
|
|
|
@d = constant i64 ptrtoint (double* getelementptr ([13 x double]* null, i64 0, i32 11) to i64)
|
|
|
|
|
@e = constant i64 ptrtoint (double* getelementptr ({double, float, double, double}* null, i64 0, i32 2) to i64)
|
|
|
|
|
@f = constant i64 ptrtoint (<{ i16, i128 }>* getelementptr ({i1, <{ i16, i128 }>}* null, i64 0, i32 1) to i64)
|
2010-02-02 01:41:39 +00:00
|
|
|
@g = constant i64 ptrtoint ({double, double}* getelementptr ({i1, {double, double}}* null, i64 0, i32 1) to i64)
|
2010-02-10 06:13:07 +00:00
|
|
|
@h = constant i64 ptrtoint (double** getelementptr (double** null, i64 1) to i64)
|
|
|
|
|
@i = constant i64 ptrtoint (double** getelementptr ({i1, double*}* null, i64 0, i32 1) to i64)
|
2010-02-01 18:27:38 +00:00
|
|
|
|
|
|
|
|
; The target-dependent folder should cast GEP indices to integer-sized pointers.
|
|
|
|
|
|
|
|
|
|
; PLAIN: @M = constant i64* getelementptr (i64* null, i32 1)
|
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.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@134829 91177308-0d34-0410-b5e6-96231b3b80d8
2011-07-09 17:41:24 +00:00
|
|
|
; PLAIN: @N = constant i64* getelementptr ({ i64, i64 }* null, i32 0, i32 1)
|
2010-02-01 18:27:38 +00:00
|
|
|
; PLAIN: @O = constant i64* getelementptr ([2 x i64]* null, i32 0, i32 1)
|
|
|
|
|
; OPT: @M = constant i64* getelementptr (i64* null, i32 1)
|
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.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@134829 91177308-0d34-0410-b5e6-96231b3b80d8
2011-07-09 17:41:24 +00:00
|
|
|
; OPT: @N = constant i64* getelementptr ({ i64, i64 }* null, i32 0, i32 1)
|
2010-02-01 18:27:38 +00:00
|
|
|
; OPT: @O = constant i64* getelementptr ([2 x i64]* null, i32 0, i32 1)
|
|
|
|
|
; TO: @M = constant i64* inttoptr (i64 8 to i64*)
|
|
|
|
|
; TO: @N = constant i64* inttoptr (i64 8 to i64*)
|
|
|
|
|
; TO: @O = constant i64* inttoptr (i64 8 to i64*)
|
|
|
|
|
|
2010-03-10 19:00:54 +00:00
|
|
|
@M = constant i64* getelementptr (i64* null, i32 1)
|
|
|
|
|
@N = constant i64* getelementptr ({ i64, i64 }* null, i32 0, i32 1)
|
|
|
|
|
@O = constant i64* getelementptr ([2 x i64]* null, i32 0, i32 1)
|
2010-02-01 18:27:38 +00:00
|
|
|
|
2010-03-10 19:31:51 +00:00
|
|
|
; Fold GEP of a GEP. Theoretically some of these cases could be folded
|
|
|
|
|
; without using targetdata, however that's not implemented yet.
|
|
|
|
|
|
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.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@134829 91177308-0d34-0410-b5e6-96231b3b80d8
2011-07-09 17:41:24 +00:00
|
|
|
; PLAIN: @Z = global i32* getelementptr inbounds (i32* getelementptr inbounds ([3 x { i32, i32 }]* @ext, i64 0, i64 1, i32 0), i64 1)
|
|
|
|
|
; OPT: @Z = global i32* getelementptr (i32* getelementptr inbounds ([3 x { i32, i32 }]* @ext, i64 0, i64 1, i32 0), i64 1)
|
|
|
|
|
; TO: @Z = global i32* getelementptr inbounds ([3 x { i32, i32 }]* @ext, i64 0, i64 1, i32 1)
|
2010-03-10 19:31:51 +00:00
|
|
|
|
|
|
|
|
@ext = external global [3 x { i32, i32 }]
|
|
|
|
|
@Z = global i32* getelementptr inbounds (i32* getelementptr inbounds ([3 x { i32, i32 }]* @ext, i64 0, i64 1, i32 0), i64 1)
|
|
|
|
|
|
2010-02-01 18:27:38 +00:00
|
|
|
; Duplicate all of the above as function return values rather than
|
|
|
|
|
; global initializers.
|
|
|
|
|
|
2010-01-28 18:08:26 +00:00
|
|
|
; PLAIN: define i8* @goo8() nounwind {
|
2010-02-01 18:27:38 +00:00
|
|
|
; PLAIN: %t = bitcast i8* getelementptr (i8* inttoptr (i32 1 to i8*), i32 -1) to i8*
|
|
|
|
|
; PLAIN: ret i8* %t
|
2010-01-28 18:08:26 +00:00
|
|
|
; PLAIN: }
|
|
|
|
|
; PLAIN: define i1* @goo1() nounwind {
|
2010-02-01 18:27:38 +00:00
|
|
|
; PLAIN: %t = bitcast i1* getelementptr (i1* inttoptr (i32 1 to i1*), i32 -1) to i1*
|
|
|
|
|
; PLAIN: ret i1* %t
|
2010-01-28 18:08:26 +00:00
|
|
|
; PLAIN: }
|
|
|
|
|
; PLAIN: define i8* @foo8() nounwind {
|
2010-02-01 18:27:38 +00:00
|
|
|
; PLAIN: %t = bitcast i8* getelementptr (i8* inttoptr (i32 1 to i8*), i32 -2) to i8*
|
|
|
|
|
; PLAIN: ret i8* %t
|
2010-01-28 18:08:26 +00:00
|
|
|
; PLAIN: }
|
|
|
|
|
; PLAIN: define i1* @foo1() nounwind {
|
2010-02-01 18:27:38 +00:00
|
|
|
; PLAIN: %t = bitcast i1* getelementptr (i1* inttoptr (i32 1 to i1*), i32 -2) to i1*
|
|
|
|
|
; PLAIN: ret i1* %t
|
2010-01-28 18:08:26 +00:00
|
|
|
; PLAIN: }
|
|
|
|
|
; PLAIN: define i8* @hoo8() nounwind {
|
2010-02-01 18:27:38 +00:00
|
|
|
; PLAIN: %t = bitcast i8* getelementptr (i8* null, i32 -1) to i8*
|
|
|
|
|
; PLAIN: ret i8* %t
|
2010-01-28 18:08:26 +00:00
|
|
|
; PLAIN: }
|
|
|
|
|
; PLAIN: define i1* @hoo1() nounwind {
|
2010-02-01 18:27:38 +00:00
|
|
|
; PLAIN: %t = bitcast i1* getelementptr (i1* null, i32 -1) to i1*
|
|
|
|
|
; PLAIN: ret i1* %t
|
2010-01-28 18:08:26 +00:00
|
|
|
; PLAIN: }
|
|
|
|
|
; OPT: define i8* @goo8() nounwind {
|
2010-02-01 18:27:38 +00:00
|
|
|
; OPT: ret i8* getelementptr (i8* inttoptr (i32 1 to i8*), i32 -1)
|
2010-01-28 18:08:26 +00:00
|
|
|
; OPT: }
|
|
|
|
|
; OPT: define i1* @goo1() nounwind {
|
2010-02-01 18:27:38 +00:00
|
|
|
; OPT: ret i1* getelementptr (i1* inttoptr (i32 1 to i1*), i32 -1)
|
2010-01-28 18:08:26 +00:00
|
|
|
; OPT: }
|
|
|
|
|
; OPT: define i8* @foo8() nounwind {
|
2010-02-01 18:27:38 +00:00
|
|
|
; OPT: ret i8* getelementptr (i8* inttoptr (i32 1 to i8*), i32 -2)
|
2010-01-28 18:08:26 +00:00
|
|
|
; OPT: }
|
|
|
|
|
; OPT: define i1* @foo1() nounwind {
|
2010-02-01 18:27:38 +00:00
|
|
|
; OPT: ret i1* getelementptr (i1* inttoptr (i32 1 to i1*), i32 -2)
|
2010-01-28 18:08:26 +00:00
|
|
|
; OPT: }
|
|
|
|
|
; OPT: define i8* @hoo8() nounwind {
|
2010-02-01 18:27:38 +00:00
|
|
|
; OPT: ret i8* getelementptr (i8* null, i32 -1)
|
2010-01-28 18:08:26 +00:00
|
|
|
; OPT: }
|
|
|
|
|
; OPT: define i1* @hoo1() nounwind {
|
2010-02-01 18:27:38 +00:00
|
|
|
; OPT: ret i1* getelementptr (i1* null, i32 -1)
|
2010-01-28 18:08:26 +00:00
|
|
|
; OPT: }
|
2010-02-01 18:27:38 +00:00
|
|
|
; TO: define i8* @goo8() nounwind {
|
|
|
|
|
; TO: ret i8* null
|
|
|
|
|
; TO: }
|
|
|
|
|
; TO: define i1* @goo1() nounwind {
|
|
|
|
|
; TO: ret i1* null
|
|
|
|
|
; TO: }
|
|
|
|
|
; TO: define i8* @foo8() nounwind {
|
|
|
|
|
; TO: ret i8* inttoptr (i64 -1 to i8*)
|
|
|
|
|
; TO: }
|
|
|
|
|
; TO: define i1* @foo1() nounwind {
|
|
|
|
|
; TO: ret i1* inttoptr (i64 -1 to i1*)
|
|
|
|
|
; TO: }
|
|
|
|
|
; TO: define i8* @hoo8() nounwind {
|
|
|
|
|
; TO: ret i8* inttoptr (i64 -1 to i8*)
|
|
|
|
|
; TO: }
|
|
|
|
|
; TO: define i1* @hoo1() nounwind {
|
|
|
|
|
; TO: ret i1* inttoptr (i64 -1 to i1*)
|
|
|
|
|
; TO: }
|
|
|
|
|
; SCEV: Classifying expressions for: @goo8
|
|
|
|
|
; SCEV: %t = bitcast i8* getelementptr (i8* inttoptr (i32 1 to i8*), i32 -1) to i8*
|
|
|
|
|
; SCEV: --> ((-1 * sizeof(i8)) + inttoptr (i32 1 to i8*))
|
|
|
|
|
; SCEV: Classifying expressions for: @goo1
|
|
|
|
|
; SCEV: %t = bitcast i1* getelementptr (i1* inttoptr (i32 1 to i1*), i32 -1) to i1*
|
|
|
|
|
; SCEV: --> ((-1 * sizeof(i1)) + inttoptr (i32 1 to i1*))
|
|
|
|
|
; SCEV: Classifying expressions for: @foo8
|
|
|
|
|
; SCEV: %t = bitcast i8* getelementptr (i8* inttoptr (i32 1 to i8*), i32 -2) to i8*
|
|
|
|
|
; SCEV: --> ((-2 * sizeof(i8)) + inttoptr (i32 1 to i8*))
|
|
|
|
|
; SCEV: Classifying expressions for: @foo1
|
|
|
|
|
; SCEV: %t = bitcast i1* getelementptr (i1* inttoptr (i32 1 to i1*), i32 -2) to i1*
|
|
|
|
|
; SCEV: --> ((-2 * sizeof(i1)) + inttoptr (i32 1 to i1*))
|
|
|
|
|
; SCEV: Classifying expressions for: @hoo8
|
|
|
|
|
; SCEV: --> (-1 * sizeof(i8))
|
|
|
|
|
; SCEV: Classifying expressions for: @hoo1
|
|
|
|
|
; SCEV: --> (-1 * sizeof(i1))
|
2010-01-28 18:08:26 +00:00
|
|
|
|
|
|
|
|
define i8* @goo8() nounwind {
|
2010-02-01 18:27:38 +00:00
|
|
|
%t = bitcast i8* getelementptr (i8* inttoptr (i32 1 to i8*), i32 -1) to i8*
|
|
|
|
|
ret i8* %t
|
2010-01-28 18:08:26 +00:00
|
|
|
}
|
|
|
|
|
define i1* @goo1() nounwind {
|
2010-02-01 18:27:38 +00:00
|
|
|
%t = bitcast i1* getelementptr (i1* inttoptr (i32 1 to i1*), i32 -1) to i1*
|
|
|
|
|
ret i1* %t
|
2010-01-28 18:08:26 +00:00
|
|
|
}
|
|
|
|
|
define i8* @foo8() nounwind {
|
2010-02-01 18:27:38 +00:00
|
|
|
%t = bitcast i8* getelementptr (i8* inttoptr (i32 1 to i8*), i32 -2) to i8*
|
|
|
|
|
ret i8* %t
|
2010-01-28 18:08:26 +00:00
|
|
|
}
|
|
|
|
|
define i1* @foo1() nounwind {
|
2010-02-01 18:27:38 +00:00
|
|
|
%t = bitcast i1* getelementptr (i1* inttoptr (i32 1 to i1*), i32 -2) to i1*
|
|
|
|
|
ret i1* %t
|
2010-01-28 18:08:26 +00:00
|
|
|
}
|
|
|
|
|
define i8* @hoo8() nounwind {
|
2010-02-01 18:27:38 +00:00
|
|
|
%t = bitcast i8* getelementptr (i8* inttoptr (i32 0 to i8*), i32 -1) to i8*
|
|
|
|
|
ret i8* %t
|
2010-01-28 18:08:26 +00:00
|
|
|
}
|
|
|
|
|
define i1* @hoo1() nounwind {
|
2010-02-01 18:27:38 +00:00
|
|
|
%t = bitcast i1* getelementptr (i1* inttoptr (i32 0 to i1*), i32 -1) to i1*
|
|
|
|
|
ret i1* %t
|
|
|
|
|
}
|
|
|
|
|
|
|
|
|
|
; PLAIN: define i64 @fa() nounwind {
|
|
|
|
|
; PLAIN: %t = bitcast i64 mul (i64 ptrtoint (double* getelementptr (double* null, i32 1) to i64), i64 2310) to i64
|
|
|
|
|
; PLAIN: ret i64 %t
|
|
|
|
|
; PLAIN: }
|
|
|
|
|
; PLAIN: define i64 @fb() nounwind {
|
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.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@134829 91177308-0d34-0410-b5e6-96231b3b80d8
2011-07-09 17:41:24 +00:00
|
|
|
; PLAIN: %t = bitcast i64 ptrtoint (double* getelementptr ({ i1, double }* null, i64 0, i32 1) to i64) to i64
|
2010-02-01 18:27:38 +00:00
|
|
|
; PLAIN: ret i64 %t
|
|
|
|
|
; PLAIN: }
|
|
|
|
|
; PLAIN: define i64 @fc() nounwind {
|
|
|
|
|
; PLAIN: %t = bitcast i64 mul nuw (i64 ptrtoint (double* getelementptr (double* null, i32 1) to i64), i64 2) to i64
|
|
|
|
|
; PLAIN: ret i64 %t
|
|
|
|
|
; PLAIN: }
|
|
|
|
|
; PLAIN: define i64 @fd() nounwind {
|
|
|
|
|
; PLAIN: %t = bitcast i64 mul nuw (i64 ptrtoint (double* getelementptr (double* null, i32 1) to i64), i64 11) to i64
|
|
|
|
|
; PLAIN: ret i64 %t
|
|
|
|
|
; PLAIN: }
|
|
|
|
|
; PLAIN: define i64 @fe() nounwind {
|
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.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@134829 91177308-0d34-0410-b5e6-96231b3b80d8
2011-07-09 17:41:24 +00:00
|
|
|
; PLAIN: %t = bitcast i64 ptrtoint (double* getelementptr ({ double, float, double, double }* null, i64 0, i32 2) to i64) to i64
|
2010-02-01 18:27:38 +00:00
|
|
|
; PLAIN: ret i64 %t
|
|
|
|
|
; PLAIN: }
|
|
|
|
|
; PLAIN: define i64 @ff() nounwind {
|
|
|
|
|
; PLAIN: %t = bitcast i64 1 to i64
|
|
|
|
|
; PLAIN: ret i64 %t
|
|
|
|
|
; PLAIN: }
|
2010-02-02 01:41:39 +00:00
|
|
|
; PLAIN: define i64 @fg() nounwind {
|
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.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@134829 91177308-0d34-0410-b5e6-96231b3b80d8
2011-07-09 17:41:24 +00:00
|
|
|
; PLAIN: %t = bitcast i64 ptrtoint (double* getelementptr ({ i1, double }* null, i64 0, i32 1) to i64) to i64
|
2010-02-02 01:41:39 +00:00
|
|
|
; PLAIN: ret i64 %t
|
|
|
|
|
; PLAIN: }
|
2010-02-10 06:13:07 +00:00
|
|
|
; PLAIN: define i64 @fh() nounwind {
|
2010-02-25 16:45:19 +00:00
|
|
|
; PLAIN: %t = bitcast i64 ptrtoint (i1** getelementptr (i1** null, i32 1) to i64) to i64
|
2010-02-10 06:13:07 +00:00
|
|
|
; PLAIN: ret i64 %t
|
|
|
|
|
; PLAIN: }
|
|
|
|
|
; PLAIN: define i64 @fi() nounwind {
|
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.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@134829 91177308-0d34-0410-b5e6-96231b3b80d8
2011-07-09 17:41:24 +00:00
|
|
|
; PLAIN: %t = bitcast i64 ptrtoint (i1** getelementptr ({ i1, i1* }* null, i64 0, i32 1) to i64) to i64
|
2010-02-25 16:45:19 +00:00
|
|
|
; PLAIN: ret i64 %t
|
|
|
|
|
; PLAIN: }
|
2010-02-01 18:27:38 +00:00
|
|
|
; OPT: define i64 @fa() nounwind {
|
|
|
|
|
; OPT: ret i64 mul (i64 ptrtoint (double* getelementptr (double* null, i32 1) to i64), i64 2310)
|
|
|
|
|
; OPT: }
|
|
|
|
|
; OPT: define i64 @fb() nounwind {
|
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.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@134829 91177308-0d34-0410-b5e6-96231b3b80d8
2011-07-09 17:41:24 +00:00
|
|
|
; OPT: ret i64 ptrtoint (double* getelementptr ({ i1, double }* null, i64 0, i32 1) to i64)
|
2010-02-01 18:27:38 +00:00
|
|
|
; OPT: }
|
|
|
|
|
; OPT: define i64 @fc() nounwind {
|
|
|
|
|
; OPT: ret i64 mul nuw (i64 ptrtoint (double* getelementptr (double* null, i32 1) to i64), i64 2)
|
|
|
|
|
; OPT: }
|
|
|
|
|
; OPT: define i64 @fd() nounwind {
|
|
|
|
|
; OPT: ret i64 mul nuw (i64 ptrtoint (double* getelementptr (double* null, i32 1) to i64), i64 11)
|
|
|
|
|
; OPT: }
|
|
|
|
|
; OPT: define i64 @fe() nounwind {
|
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.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@134829 91177308-0d34-0410-b5e6-96231b3b80d8
2011-07-09 17:41:24 +00:00
|
|
|
; OPT: ret i64 ptrtoint (double* getelementptr ({ double, float, double, double }* null, i64 0, i32 2) to i64)
|
2010-02-01 18:27:38 +00:00
|
|
|
; OPT: }
|
|
|
|
|
; OPT: define i64 @ff() nounwind {
|
|
|
|
|
; OPT: ret i64 1
|
|
|
|
|
; OPT: }
|
2010-02-02 01:41:39 +00:00
|
|
|
; OPT: define i64 @fg() nounwind {
|
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.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@134829 91177308-0d34-0410-b5e6-96231b3b80d8
2011-07-09 17:41:24 +00:00
|
|
|
; OPT: ret i64 ptrtoint (double* getelementptr ({ i1, double }* null, i64 0, i32 1) to i64)
|
2010-02-02 01:41:39 +00:00
|
|
|
; OPT: }
|
2010-02-10 06:13:07 +00:00
|
|
|
; OPT: define i64 @fh() nounwind {
|
|
|
|
|
; OPT: ret i64 ptrtoint (i1** getelementptr (i1** null, i32 1) to i64)
|
|
|
|
|
; OPT: }
|
|
|
|
|
; OPT: define i64 @fi() nounwind {
|
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.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@134829 91177308-0d34-0410-b5e6-96231b3b80d8
2011-07-09 17:41:24 +00:00
|
|
|
; OPT: ret i64 ptrtoint (i1** getelementptr ({ i1, i1* }* null, i64 0, i32 1) to i64)
|
2010-02-10 06:13:07 +00:00
|
|
|
; OPT: }
|
2010-02-01 18:27:38 +00:00
|
|
|
; TO: define i64 @fa() nounwind {
|
|
|
|
|
; TO: ret i64 18480
|
|
|
|
|
; TO: }
|
|
|
|
|
; TO: define i64 @fb() nounwind {
|
|
|
|
|
; TO: ret i64 8
|
|
|
|
|
; TO: }
|
|
|
|
|
; TO: define i64 @fc() nounwind {
|
|
|
|
|
; TO: ret i64 16
|
|
|
|
|
; TO: }
|
|
|
|
|
; TO: define i64 @fd() nounwind {
|
|
|
|
|
; TO: ret i64 88
|
|
|
|
|
; TO: }
|
|
|
|
|
; TO: define i64 @fe() nounwind {
|
|
|
|
|
; TO: ret i64 16
|
|
|
|
|
; TO: }
|
|
|
|
|
; TO: define i64 @ff() nounwind {
|
|
|
|
|
; TO: ret i64 1
|
|
|
|
|
; TO: }
|
2010-02-02 01:41:39 +00:00
|
|
|
; TO: define i64 @fg() nounwind {
|
|
|
|
|
; TO: ret i64 8
|
|
|
|
|
; TO: }
|
2010-02-10 06:13:07 +00:00
|
|
|
; TO: define i64 @fh() nounwind {
|
|
|
|
|
; TO: ret i64 8
|
|
|
|
|
; TO: }
|
|
|
|
|
; TO: define i64 @fi() nounwind {
|
|
|
|
|
; TO: ret i64 8
|
|
|
|
|
; TO: }
|
2010-02-01 18:27:38 +00:00
|
|
|
; SCEV: Classifying expressions for: @fa
|
|
|
|
|
; SCEV: %t = bitcast i64 mul (i64 ptrtoint (double* getelementptr (double* null, i32 1) to i64), i64 2310) to i64
|
|
|
|
|
; SCEV: --> (2310 * sizeof(double))
|
|
|
|
|
; SCEV: Classifying expressions for: @fb
|
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.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@134829 91177308-0d34-0410-b5e6-96231b3b80d8
2011-07-09 17:41:24 +00:00
|
|
|
; SCEV: %t = bitcast i64 ptrtoint (double* getelementptr ({ i1, double }* null, i64 0, i32 1) to i64) to i64
|
2010-02-01 18:27:38 +00:00
|
|
|
; SCEV: --> alignof(double)
|
|
|
|
|
; SCEV: Classifying expressions for: @fc
|
|
|
|
|
; SCEV: %t = bitcast i64 mul nuw (i64 ptrtoint (double* getelementptr (double* null, i32 1) to i64), i64 2) to i64
|
|
|
|
|
; SCEV: --> (2 * sizeof(double))
|
|
|
|
|
; SCEV: Classifying expressions for: @fd
|
|
|
|
|
; SCEV: %t = bitcast i64 mul nuw (i64 ptrtoint (double* getelementptr (double* null, i32 1) to i64), i64 11) to i64
|
|
|
|
|
; SCEV: --> (11 * sizeof(double))
|
|
|
|
|
; SCEV: Classifying expressions for: @fe
|
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.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@134829 91177308-0d34-0410-b5e6-96231b3b80d8
2011-07-09 17:41:24 +00:00
|
|
|
; SCEV: %t = bitcast i64 ptrtoint (double* getelementptr ({ double, float, double, double }* null, i64 0, i32 2) to i64) to i64
|
2010-02-01 18:27:38 +00:00
|
|
|
; SCEV: --> offsetof({ double, float, double, double }, 2)
|
|
|
|
|
; SCEV: Classifying expressions for: @ff
|
|
|
|
|
; SCEV: %t = bitcast i64 1 to i64
|
|
|
|
|
; SCEV: --> 1
|
2010-02-02 01:41:39 +00:00
|
|
|
; SCEV: Classifying expressions for: @fg
|
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.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@134829 91177308-0d34-0410-b5e6-96231b3b80d8
2011-07-09 17:41:24 +00:00
|
|
|
; SCEV: %t = bitcast i64 ptrtoint (double* getelementptr ({ i1, double }* null, i64 0, i32 1) to i64) to i64
|
2010-02-02 01:41:39 +00:00
|
|
|
; SCEV: --> alignof(double)
|
2010-02-10 06:13:07 +00:00
|
|
|
; SCEV: Classifying expressions for: @fh
|
2010-02-25 16:45:19 +00:00
|
|
|
; SCEV: %t = bitcast i64 ptrtoint (i1** getelementptr (i1** null, i32 1) to i64) to i64
|
2010-02-10 06:13:07 +00:00
|
|
|
; SCEV: --> sizeof(i1*)
|
|
|
|
|
; SCEV: Classifying expressions for: @fi
|
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.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@134829 91177308-0d34-0410-b5e6-96231b3b80d8
2011-07-09 17:41:24 +00:00
|
|
|
; SCEV: %t = bitcast i64 ptrtoint (i1** getelementptr ({ i1, i1* }* null, i64 0, i32 1) to i64) to i64
|
2010-02-10 06:13:07 +00:00
|
|
|
; SCEV: --> alignof(i1*)
|
2010-02-01 18:27:38 +00:00
|
|
|
|
|
|
|
|
define i64 @fa() nounwind {
|
|
|
|
|
%t = bitcast i64 mul (i64 3, i64 mul (i64 ptrtoint ({[7 x double], [7 x double]}* getelementptr ({[7 x double], [7 x double]}* null, i64 11) to i64), i64 5)) to i64
|
|
|
|
|
ret i64 %t
|
|
|
|
|
}
|
|
|
|
|
define i64 @fb() nounwind {
|
|
|
|
|
%t = bitcast i64 ptrtoint ([13 x double]* getelementptr ({i1, [13 x double]}* null, i64 0, i32 1) to i64) to i64
|
|
|
|
|
ret i64 %t
|
|
|
|
|
}
|
|
|
|
|
define i64 @fc() nounwind {
|
|
|
|
|
%t = bitcast i64 ptrtoint (double* getelementptr ({double, double, double, double}* null, i64 0, i32 2) to i64) to i64
|
|
|
|
|
ret i64 %t
|
|
|
|
|
}
|
|
|
|
|
define i64 @fd() nounwind {
|
|
|
|
|
%t = bitcast i64 ptrtoint (double* getelementptr ([13 x double]* null, i64 0, i32 11) to i64) to i64
|
|
|
|
|
ret i64 %t
|
|
|
|
|
}
|
|
|
|
|
define i64 @fe() nounwind {
|
|
|
|
|
%t = bitcast i64 ptrtoint (double* getelementptr ({double, float, double, double}* null, i64 0, i32 2) to i64) to i64
|
|
|
|
|
ret i64 %t
|
|
|
|
|
}
|
|
|
|
|
define i64 @ff() nounwind {
|
|
|
|
|
%t = bitcast i64 ptrtoint (<{ i16, i128 }>* getelementptr ({i1, <{ i16, i128 }>}* null, i64 0, i32 1) to i64) to i64
|
|
|
|
|
ret i64 %t
|
|
|
|
|
}
|
2010-02-02 01:41:39 +00:00
|
|
|
define i64 @fg() nounwind {
|
|
|
|
|
%t = bitcast i64 ptrtoint ({double, double}* getelementptr ({i1, {double, double}}* null, i64 0, i32 1) to i64) to i64
|
|
|
|
|
ret i64 %t
|
|
|
|
|
}
|
2010-02-10 06:13:07 +00:00
|
|
|
define i64 @fh() nounwind {
|
|
|
|
|
%t = bitcast i64 ptrtoint (double** getelementptr (double** null, i32 1) to i64) to i64
|
|
|
|
|
ret i64 %t
|
|
|
|
|
}
|
|
|
|
|
define i64 @fi() nounwind {
|
|
|
|
|
%t = bitcast i64 ptrtoint (double** getelementptr ({i1, double*}* null, i64 0, i32 1) to i64) to i64
|
|
|
|
|
ret i64 %t
|
|
|
|
|
}
|
2010-02-01 18:27:38 +00:00
|
|
|
|
|
|
|
|
; PLAIN: define i64* @fM() nounwind {
|
|
|
|
|
; PLAIN: %t = bitcast i64* getelementptr (i64* null, i32 1) to i64*
|
|
|
|
|
; PLAIN: ret i64* %t
|
|
|
|
|
; PLAIN: }
|
|
|
|
|
; PLAIN: define i64* @fN() nounwind {
|
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.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@134829 91177308-0d34-0410-b5e6-96231b3b80d8
2011-07-09 17:41:24 +00:00
|
|
|
; PLAIN: %t = bitcast i64* getelementptr ({ i64, i64 }* null, i32 0, i32 1) to i64*
|
2010-02-01 18:27:38 +00:00
|
|
|
; PLAIN: ret i64* %t
|
|
|
|
|
; PLAIN: }
|
|
|
|
|
; PLAIN: define i64* @fO() nounwind {
|
|
|
|
|
; PLAIN: %t = bitcast i64* getelementptr ([2 x i64]* null, i32 0, i32 1) to i64*
|
|
|
|
|
; PLAIN: ret i64* %t
|
|
|
|
|
; PLAIN: }
|
|
|
|
|
; OPT: define i64* @fM() nounwind {
|
|
|
|
|
; OPT: ret i64* getelementptr (i64* null, i32 1)
|
|
|
|
|
; OPT: }
|
|
|
|
|
; OPT: define i64* @fN() nounwind {
|
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.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@134829 91177308-0d34-0410-b5e6-96231b3b80d8
2011-07-09 17:41:24 +00:00
|
|
|
; OPT: ret i64* getelementptr ({ i64, i64 }* null, i32 0, i32 1)
|
2010-02-01 18:27:38 +00:00
|
|
|
; OPT: }
|
|
|
|
|
; OPT: define i64* @fO() nounwind {
|
|
|
|
|
; OPT: ret i64* getelementptr ([2 x i64]* null, i32 0, i32 1)
|
|
|
|
|
; OPT: }
|
|
|
|
|
; TO: define i64* @fM() nounwind {
|
|
|
|
|
; TO: ret i64* inttoptr (i64 8 to i64*)
|
|
|
|
|
; TO: }
|
|
|
|
|
; TO: define i64* @fN() nounwind {
|
|
|
|
|
; TO: ret i64* inttoptr (i64 8 to i64*)
|
|
|
|
|
; TO: }
|
|
|
|
|
; TO: define i64* @fO() nounwind {
|
|
|
|
|
; TO: ret i64* inttoptr (i64 8 to i64*)
|
|
|
|
|
; TO: }
|
|
|
|
|
; SCEV: Classifying expressions for: @fM
|
|
|
|
|
; SCEV: %t = bitcast i64* getelementptr (i64* null, i32 1) to i64*
|
|
|
|
|
; SCEV: --> sizeof(i64)
|
|
|
|
|
; SCEV: Classifying expressions for: @fN
|
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.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@134829 91177308-0d34-0410-b5e6-96231b3b80d8
2011-07-09 17:41:24 +00:00
|
|
|
; SCEV: %t = bitcast i64* getelementptr ({ i64, i64 }* null, i32 0, i32 1) to i64*
|
2010-02-01 18:27:38 +00:00
|
|
|
; SCEV: --> sizeof(i64)
|
|
|
|
|
; SCEV: Classifying expressions for: @fO
|
|
|
|
|
; SCEV: %t = bitcast i64* getelementptr ([2 x i64]* null, i32 0, i32 1) to i64*
|
|
|
|
|
; SCEV: --> sizeof(i64)
|
|
|
|
|
|
|
|
|
|
define i64* @fM() nounwind {
|
2010-03-10 19:00:54 +00:00
|
|
|
%t = bitcast i64* getelementptr (i64* null, i32 1) to i64*
|
2010-02-01 18:27:38 +00:00
|
|
|
ret i64* %t
|
|
|
|
|
}
|
|
|
|
|
define i64* @fN() nounwind {
|
2010-03-10 19:00:54 +00:00
|
|
|
%t = bitcast i64* getelementptr ({ i64, i64 }* null, i32 0, i32 1) to i64*
|
2010-02-01 18:27:38 +00:00
|
|
|
ret i64* %t
|
|
|
|
|
}
|
|
|
|
|
define i64* @fO() nounwind {
|
2010-03-10 19:00:54 +00:00
|
|
|
%t = bitcast i64* getelementptr ([2 x i64]* null, i32 0, i32 1) to i64*
|
2010-02-01 18:27:38 +00:00
|
|
|
ret i64* %t
|
2010-01-28 18:08:26 +00:00
|
|
|
}
|
2010-03-10 19:31:51 +00:00
|
|
|
|
|
|
|
|
; PLAIN: define i32* @fZ() nounwind {
|
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.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@134829 91177308-0d34-0410-b5e6-96231b3b80d8
2011-07-09 17:41:24 +00:00
|
|
|
; PLAIN: %t = bitcast i32* getelementptr inbounds (i32* getelementptr inbounds ([3 x { i32, i32 }]* @ext, i64 0, i64 1, i32 0), i64 1) to i32*
|
2010-03-10 19:31:51 +00:00
|
|
|
; PLAIN: ret i32* %t
|
|
|
|
|
; PLAIN: }
|
|
|
|
|
; OPT: define i32* @fZ() nounwind {
|
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.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@134829 91177308-0d34-0410-b5e6-96231b3b80d8
2011-07-09 17:41:24 +00:00
|
|
|
; OPT: ret i32* getelementptr inbounds (i32* getelementptr inbounds ([3 x { i32, i32 }]* @ext, i64 0, i64 1, i32 0), i64 1)
|
2010-03-10 19:31:51 +00:00
|
|
|
; OPT: }
|
|
|
|
|
; TO: define i32* @fZ() nounwind {
|
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.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@134829 91177308-0d34-0410-b5e6-96231b3b80d8
2011-07-09 17:41:24 +00:00
|
|
|
; TO: ret i32* getelementptr inbounds ([3 x { i32, i32 }]* @ext, i64 0, i64 1, i32 1)
|
2010-03-10 19:31:51 +00:00
|
|
|
; TO: }
|
|
|
|
|
; SCEV: Classifying expressions for: @fZ
|
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.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@134829 91177308-0d34-0410-b5e6-96231b3b80d8
2011-07-09 17:41:24 +00:00
|
|
|
; SCEV: %t = bitcast i32* getelementptr inbounds (i32* getelementptr inbounds ([3 x { i32, i32 }]* @ext, i64 0, i64 1, i32 0), i64 1) to i32*
|
2010-03-10 19:31:51 +00:00
|
|
|
; SCEV: --> ((3 * sizeof(i32)) + @ext)
|
|
|
|
|
|
|
|
|
|
define i32* @fZ() nounwind {
|
|
|
|
|
%t = bitcast i32* getelementptr inbounds (i32* getelementptr inbounds ([3 x { i32, i32 }]* @ext, i64 0, i64 1, i32 0), i64 1) to i32*
|
|
|
|
|
ret i32* %t
|
|
|
|
|
}
|
