llvm-project/llvm/test/Transforms/InstCombine/phi.ll

882 lines
22 KiB
LLVM
Raw Normal View History

; This test makes sure that these instructions are properly eliminated.
;
2009-11-02 03:22:20 +08:00
; RUN: opt < %s -instcombine -S | FileCheck %s
target datalayout = "e-p:64:64:64-i1:8:8-i8:8:8-i16:16:16-i32:32:32-i64:64:64-f32:32:32-f64:64:64-v64:64:64-v128:128:128-a0:0:64-s0:64:64-f80:128:128:n8:16:32:64"
define i32 @test1(i32 %A, i1 %b) {
BB0:
br i1 %b, label %BB1, label %BB2
2009-11-02 03:22:20 +08:00
BB1:
; Combine away one argument PHI nodes
2009-11-02 03:22:20 +08:00
%B = phi i32 [ %A, %BB0 ]
ret i32 %B
2009-11-02 03:22:20 +08:00
BB2:
ret i32 %A
; CHECK-LABEL: @test1(
2009-11-02 03:22:20 +08:00
; CHECK: BB1:
; CHECK-NEXT: ret i32 %A
}
define i32 @test2(i32 %A, i1 %b) {
BB0:
br i1 %b, label %BB1, label %BB2
2009-11-02 03:22:20 +08:00
BB1:
br label %BB2
2009-11-02 03:22:20 +08:00
BB2:
; Combine away PHI nodes with same values
2009-11-02 03:22:20 +08:00
%B = phi i32 [ %A, %BB0 ], [ %A, %BB1 ]
ret i32 %B
; CHECK-LABEL: @test2(
2009-11-02 03:22:20 +08:00
; CHECK: BB2:
; CHECK-NEXT: ret i32 %A
}
define i32 @test3(i32 %A, i1 %b) {
BB0:
br label %Loop
2009-11-02 03:22:20 +08:00
Loop:
; PHI has same value always.
2009-11-02 03:22:20 +08:00
%B = phi i32 [ %A, %BB0 ], [ %B, %Loop ]
br i1 %b, label %Loop, label %Exit
2009-11-02 03:22:20 +08:00
Exit:
ret i32 %B
; CHECK-LABEL: @test3(
2009-11-02 03:22:20 +08:00
; CHECK: Exit:
; CHECK-NEXT: ret i32 %A
}
define i32 @test4(i1 %b) {
BB0:
; Loop is unreachable
ret i32 7
Loop: ; preds = %L2, %Loop
; PHI has same value always.
2009-11-02 03:22:20 +08:00
%B = phi i32 [ %B, %L2 ], [ %B, %Loop ]
br i1 %b, label %L2, label %Loop
L2: ; preds = %Loop
br label %Loop
; CHECK-LABEL: @test4(
2009-11-02 03:22:20 +08:00
; CHECK: Loop:
; CHECK-NEXT: br i1 %b
}
define i32 @test5(i32 %A, i1 %b) {
BB0:
br label %Loop
Loop: ; preds = %Loop, %BB0
; PHI has same value always.
2009-11-02 03:22:20 +08:00
%B = phi i32 [ %A, %BB0 ], [ undef, %Loop ]
br i1 %b, label %Loop, label %Exit
Exit: ; preds = %Loop
ret i32 %B
; CHECK-LABEL: @test5(
2009-11-02 03:22:20 +08:00
; CHECK: Loop:
; CHECK-NEXT: br i1 %b
; CHECK: Exit:
; CHECK-NEXT: ret i32 %A
}
2009-11-02 03:22:20 +08:00
define i32 @test6(i16 %A, i1 %b) {
BB0:
2009-11-02 03:22:20 +08:00
%X = zext i16 %A to i32
br i1 %b, label %BB1, label %BB2
2009-11-02 03:22:20 +08:00
BB1:
%Y = zext i16 %A to i32
br label %BB2
2009-11-02 03:22:20 +08:00
BB2:
;; Suck casts into phi
2009-11-02 03:22:20 +08:00
%B = phi i32 [ %X, %BB0 ], [ %Y, %BB1 ]
ret i32 %B
; CHECK-LABEL: @test6(
2009-11-02 03:22:20 +08:00
; CHECK: BB2:
; CHECK: zext i16 %A to i32
; CHECK-NEXT: ret i32
}
define i32 @test7(i32 %A, i1 %b) {
BB0:
br label %Loop
Loop: ; preds = %Loop, %BB0
; PHI is dead.
2009-11-02 03:22:20 +08:00
%B = phi i32 [ %A, %BB0 ], [ %C, %Loop ]
%C = add i32 %B, 123
br i1 %b, label %Loop, label %Exit
2007-01-15 15:29:29 +08:00
Exit: ; preds = %Loop
ret i32 0
; CHECK-LABEL: @test7(
2009-11-02 03:22:20 +08:00
; CHECK: Loop:
; CHECK-NEXT: br i1 %b
2007-01-15 15:29:29 +08:00
}
define i32* @test8({ i32, i32 } *%A, i1 %b) {
BB0:
[opaque pointer type] Add textual IR support for explicit type parameter to getelementptr instruction One of several parallel first steps to remove the target type of pointers, replacing them with a single opaque pointer type. This adds an explicit type parameter to the gep instruction so that when the first parameter becomes an opaque pointer type, the type to gep through is still available to the instructions. * This doesn't modify gep operators, only instructions (operators will be handled separately) * Textual IR changes only. Bitcode (including upgrade) and changing the in-memory representation will be in separate changes. * geps of vectors are transformed as: getelementptr <4 x float*> %x, ... ->getelementptr float, <4 x float*> %x, ... Then, once the opaque pointer type is introduced, this will ultimately look like: getelementptr float, <4 x ptr> %x with the unambiguous interpretation that it is a vector of pointers to float. * address spaces remain on the pointer, not the type: getelementptr float addrspace(1)* %x ->getelementptr float, float addrspace(1)* %x Then, eventually: getelementptr float, ptr addrspace(1) %x Importantly, the massive amount of test case churn has been automated by same crappy python code. I had to manually update a few test cases that wouldn't fit the script's model (r228970,r229196,r229197,r229198). The python script just massages stdin and writes the result to stdout, I then wrapped that in a shell script to handle replacing files, then using the usual find+xargs to migrate all the files. update.py: import fileinput import sys import re ibrep = re.compile(r"(^.*?[^%\w]getelementptr inbounds )(((?:<\d* x )?)(.*?)(| addrspace\(\d\)) *\*(|>)(?:$| *(?:%|@|null|undef|blockaddress|getelementptr|addrspacecast|bitcast|inttoptr|\[\[[a-zA-Z]|\{\{).*$))") normrep = re.compile( r"(^.*?[^%\w]getelementptr )(((?:<\d* x )?)(.*?)(| addrspace\(\d\)) *\*(|>)(?:$| *(?:%|@|null|undef|blockaddress|getelementptr|addrspacecast|bitcast|inttoptr|\[\[[a-zA-Z]|\{\{).*$))") def conv(match, line): if not match: return line line = match.groups()[0] if len(match.groups()[5]) == 0: line += match.groups()[2] line += match.groups()[3] line += ", " line += match.groups()[1] line += "\n" return line for line in sys.stdin: if line.find("getelementptr ") == line.find("getelementptr inbounds"): if line.find("getelementptr inbounds") != line.find("getelementptr inbounds ("): line = conv(re.match(ibrep, line), line) elif line.find("getelementptr ") != line.find("getelementptr ("): line = conv(re.match(normrep, line), line) sys.stdout.write(line) apply.sh: for name in "$@" do python3 `dirname "$0"`/update.py < "$name" > "$name.tmp" && mv "$name.tmp" "$name" rm -f "$name.tmp" done The actual commands: From llvm/src: find test/ -name *.ll | xargs ./apply.sh From llvm/src/tools/clang: find test/ -name *.mm -o -name *.m -o -name *.cpp -o -name *.c | xargs -I '{}' ../../apply.sh "{}" From llvm/src/tools/polly: find test/ -name *.ll | xargs ./apply.sh After that, check-all (with llvm, clang, clang-tools-extra, lld, compiler-rt, and polly all checked out). The extra 'rm' in the apply.sh script is due to a few files in clang's test suite using interesting unicode stuff that my python script was throwing exceptions on. None of those files needed to be migrated, so it seemed sufficient to ignore those cases. Reviewers: rafael, dexonsmith, grosser Differential Revision: http://reviews.llvm.org/D7636 llvm-svn: 230786
2015-02-28 03:29:02 +08:00
%X = getelementptr inbounds { i32, i32 }, { i32, i32 } *%A, i32 0, i32 1
br i1 %b, label %BB1, label %BB2
BB1:
[opaque pointer type] Add textual IR support for explicit type parameter to getelementptr instruction One of several parallel first steps to remove the target type of pointers, replacing them with a single opaque pointer type. This adds an explicit type parameter to the gep instruction so that when the first parameter becomes an opaque pointer type, the type to gep through is still available to the instructions. * This doesn't modify gep operators, only instructions (operators will be handled separately) * Textual IR changes only. Bitcode (including upgrade) and changing the in-memory representation will be in separate changes. * geps of vectors are transformed as: getelementptr <4 x float*> %x, ... ->getelementptr float, <4 x float*> %x, ... Then, once the opaque pointer type is introduced, this will ultimately look like: getelementptr float, <4 x ptr> %x with the unambiguous interpretation that it is a vector of pointers to float. * address spaces remain on the pointer, not the type: getelementptr float addrspace(1)* %x ->getelementptr float, float addrspace(1)* %x Then, eventually: getelementptr float, ptr addrspace(1) %x Importantly, the massive amount of test case churn has been automated by same crappy python code. I had to manually update a few test cases that wouldn't fit the script's model (r228970,r229196,r229197,r229198). The python script just massages stdin and writes the result to stdout, I then wrapped that in a shell script to handle replacing files, then using the usual find+xargs to migrate all the files. update.py: import fileinput import sys import re ibrep = re.compile(r"(^.*?[^%\w]getelementptr inbounds )(((?:<\d* x )?)(.*?)(| addrspace\(\d\)) *\*(|>)(?:$| *(?:%|@|null|undef|blockaddress|getelementptr|addrspacecast|bitcast|inttoptr|\[\[[a-zA-Z]|\{\{).*$))") normrep = re.compile( r"(^.*?[^%\w]getelementptr )(((?:<\d* x )?)(.*?)(| addrspace\(\d\)) *\*(|>)(?:$| *(?:%|@|null|undef|blockaddress|getelementptr|addrspacecast|bitcast|inttoptr|\[\[[a-zA-Z]|\{\{).*$))") def conv(match, line): if not match: return line line = match.groups()[0] if len(match.groups()[5]) == 0: line += match.groups()[2] line += match.groups()[3] line += ", " line += match.groups()[1] line += "\n" return line for line in sys.stdin: if line.find("getelementptr ") == line.find("getelementptr inbounds"): if line.find("getelementptr inbounds") != line.find("getelementptr inbounds ("): line = conv(re.match(ibrep, line), line) elif line.find("getelementptr ") != line.find("getelementptr ("): line = conv(re.match(normrep, line), line) sys.stdout.write(line) apply.sh: for name in "$@" do python3 `dirname "$0"`/update.py < "$name" > "$name.tmp" && mv "$name.tmp" "$name" rm -f "$name.tmp" done The actual commands: From llvm/src: find test/ -name *.ll | xargs ./apply.sh From llvm/src/tools/clang: find test/ -name *.mm -o -name *.m -o -name *.cpp -o -name *.c | xargs -I '{}' ../../apply.sh "{}" From llvm/src/tools/polly: find test/ -name *.ll | xargs ./apply.sh After that, check-all (with llvm, clang, clang-tools-extra, lld, compiler-rt, and polly all checked out). The extra 'rm' in the apply.sh script is due to a few files in clang's test suite using interesting unicode stuff that my python script was throwing exceptions on. None of those files needed to be migrated, so it seemed sufficient to ignore those cases. Reviewers: rafael, dexonsmith, grosser Differential Revision: http://reviews.llvm.org/D7636 llvm-svn: 230786
2015-02-28 03:29:02 +08:00
%Y = getelementptr { i32, i32 }, { i32, i32 } *%A, i32 0, i32 1
br label %BB2
BB2:
;; Suck GEPs into phi
%B = phi i32* [ %X, %BB0 ], [ %Y, %BB1 ]
ret i32* %B
; CHECK-LABEL: @test8(
2009-11-02 03:22:20 +08:00
; CHECK-NOT: phi
; CHECK: BB2:
[opaque pointer type] Add textual IR support for explicit type parameter to getelementptr instruction One of several parallel first steps to remove the target type of pointers, replacing them with a single opaque pointer type. This adds an explicit type parameter to the gep instruction so that when the first parameter becomes an opaque pointer type, the type to gep through is still available to the instructions. * This doesn't modify gep operators, only instructions (operators will be handled separately) * Textual IR changes only. Bitcode (including upgrade) and changing the in-memory representation will be in separate changes. * geps of vectors are transformed as: getelementptr <4 x float*> %x, ... ->getelementptr float, <4 x float*> %x, ... Then, once the opaque pointer type is introduced, this will ultimately look like: getelementptr float, <4 x ptr> %x with the unambiguous interpretation that it is a vector of pointers to float. * address spaces remain on the pointer, not the type: getelementptr float addrspace(1)* %x ->getelementptr float, float addrspace(1)* %x Then, eventually: getelementptr float, ptr addrspace(1) %x Importantly, the massive amount of test case churn has been automated by same crappy python code. I had to manually update a few test cases that wouldn't fit the script's model (r228970,r229196,r229197,r229198). The python script just massages stdin and writes the result to stdout, I then wrapped that in a shell script to handle replacing files, then using the usual find+xargs to migrate all the files. update.py: import fileinput import sys import re ibrep = re.compile(r"(^.*?[^%\w]getelementptr inbounds )(((?:<\d* x )?)(.*?)(| addrspace\(\d\)) *\*(|>)(?:$| *(?:%|@|null|undef|blockaddress|getelementptr|addrspacecast|bitcast|inttoptr|\[\[[a-zA-Z]|\{\{).*$))") normrep = re.compile( r"(^.*?[^%\w]getelementptr )(((?:<\d* x )?)(.*?)(| addrspace\(\d\)) *\*(|>)(?:$| *(?:%|@|null|undef|blockaddress|getelementptr|addrspacecast|bitcast|inttoptr|\[\[[a-zA-Z]|\{\{).*$))") def conv(match, line): if not match: return line line = match.groups()[0] if len(match.groups()[5]) == 0: line += match.groups()[2] line += match.groups()[3] line += ", " line += match.groups()[1] line += "\n" return line for line in sys.stdin: if line.find("getelementptr ") == line.find("getelementptr inbounds"): if line.find("getelementptr inbounds") != line.find("getelementptr inbounds ("): line = conv(re.match(ibrep, line), line) elif line.find("getelementptr ") != line.find("getelementptr ("): line = conv(re.match(normrep, line), line) sys.stdout.write(line) apply.sh: for name in "$@" do python3 `dirname "$0"`/update.py < "$name" > "$name.tmp" && mv "$name.tmp" "$name" rm -f "$name.tmp" done The actual commands: From llvm/src: find test/ -name *.ll | xargs ./apply.sh From llvm/src/tools/clang: find test/ -name *.mm -o -name *.m -o -name *.cpp -o -name *.c | xargs -I '{}' ../../apply.sh "{}" From llvm/src/tools/polly: find test/ -name *.ll | xargs ./apply.sh After that, check-all (with llvm, clang, clang-tools-extra, lld, compiler-rt, and polly all checked out). The extra 'rm' in the apply.sh script is due to a few files in clang's test suite using interesting unicode stuff that my python script was throwing exceptions on. None of those files needed to be migrated, so it seemed sufficient to ignore those cases. Reviewers: rafael, dexonsmith, grosser Differential Revision: http://reviews.llvm.org/D7636 llvm-svn: 230786
2015-02-28 03:29:02 +08:00
; CHECK-NEXT: %B = getelementptr { i32, i32 }, { i32, i32 }* %A
2009-11-02 03:22:20 +08:00
; CHECK-NEXT: ret i32* %B
}
define i32 @test9(i32* %A, i32* %B) {
entry:
%c = icmp eq i32* %A, null
br i1 %c, label %bb1, label %bb
bb:
%C = load i32, i32* %B, align 1
br label %bb2
bb1:
%D = load i32, i32* %A, align 1
br label %bb2
bb2:
%E = phi i32 [ %C, %bb ], [ %D, %bb1 ]
ret i32 %E
; CHECK-LABEL: @test9(
; CHECK: bb2:
; CHECK-NEXT: phi i32* [ %B, %bb ], [ %A, %bb1 ]
; CHECK-NEXT: %E = load i32, i32* %{{[^,]*}}, align 1
; CHECK-NEXT: ret i32 %E
}
define i32 @test10(i32* %A, i32* %B) {
entry:
%c = icmp eq i32* %A, null
br i1 %c, label %bb1, label %bb
bb:
%C = load i32, i32* %B, align 16
br label %bb2
bb1:
%D = load i32, i32* %A, align 32
br label %bb2
bb2:
%E = phi i32 [ %C, %bb ], [ %D, %bb1 ]
ret i32 %E
; CHECK-LABEL: @test10(
; CHECK: bb2:
; CHECK-NEXT: phi i32* [ %B, %bb ], [ %A, %bb1 ]
; CHECK-NEXT: %E = load i32, i32* %{{[^,]*}}, align 16
; CHECK-NEXT: ret i32 %E
}
; PR1777
declare i1 @test11a()
define i1 @test11() {
entry:
%a = alloca i32
%i = ptrtoint i32* %a to i64
%b = call i1 @test11a()
br i1 %b, label %one, label %two
one:
%x = phi i64 [%i, %entry], [%y, %two]
%c = call i1 @test11a()
br i1 %c, label %two, label %end
two:
%y = phi i64 [%i, %entry], [%x, %one]
%d = call i1 @test11a()
br i1 %d, label %one, label %end
end:
%f = phi i64 [ %x, %one], [%y, %two]
; Change the %f to %i, and the optimizer suddenly becomes a lot smarter
; even though %f must equal %i at this point
%g = inttoptr i64 %f to i32*
store i32 10, i32* %g
%z = call i1 @test11a()
ret i1 %z
; CHECK-LABEL: @test11(
; CHECK-NOT: phi i32
; CHECK: ret i1 %z
}
define i64 @test12(i1 %cond, i8* %Ptr, i64 %Val) {
entry:
%tmp41 = ptrtoint i8* %Ptr to i64
%tmp42 = zext i64 %tmp41 to i128
br i1 %cond, label %end, label %two
two:
%tmp36 = zext i64 %Val to i128 ; <i128> [#uses=1]
%tmp37 = shl i128 %tmp36, 64 ; <i128> [#uses=1]
%ins39 = or i128 %tmp42, %tmp37 ; <i128> [#uses=1]
br label %end
end:
%tmp869.0 = phi i128 [ %tmp42, %entry ], [ %ins39, %two ]
%tmp32 = trunc i128 %tmp869.0 to i64 ; <i64> [#uses=1]
%tmp29 = lshr i128 %tmp869.0, 64 ; <i128> [#uses=1]
%tmp30 = trunc i128 %tmp29 to i64 ; <i64> [#uses=1]
%tmp2 = add i64 %tmp32, %tmp30
ret i64 %tmp2
; CHECK-LABEL: @test12(
; CHECK-NOT: zext
; CHECK: end:
; CHECK-NEXT: phi i64 [ 0, %entry ], [ %Val, %two ]
; CHECK-NOT: phi
; CHECK: ret i64
}
declare void @test13f(double, i32)
define void @test13(i1 %cond, i32 %V1, double %Vald) {
entry:
%tmp42 = zext i32 %V1 to i128
br i1 %cond, label %end, label %two
two:
%Val = bitcast double %Vald to i64
%tmp36 = zext i64 %Val to i128 ; <i128> [#uses=1]
%tmp37 = shl i128 %tmp36, 64 ; <i128> [#uses=1]
%ins39 = or i128 %tmp42, %tmp37 ; <i128> [#uses=1]
br label %end
end:
%tmp869.0 = phi i128 [ %tmp42, %entry ], [ %ins39, %two ]
%tmp32 = trunc i128 %tmp869.0 to i32
%tmp29 = lshr i128 %tmp869.0, 64 ; <i128> [#uses=1]
%tmp30 = trunc i128 %tmp29 to i64 ; <i64> [#uses=1]
%tmp31 = bitcast i64 %tmp30 to double
call void @test13f(double %tmp31, i32 %tmp32)
ret void
; CHECK-LABEL: @test13(
; CHECK-NOT: zext
; CHECK: end:
; CHECK-NEXT: phi double [ 0.000000e+00, %entry ], [ %Vald, %two ]
; CHECK-NEXT: call void @test13f(double {{[^,]*}}, i32 %V1)
; CHECK: ret void
}
define i640 @test14a(i320 %A, i320 %B, i1 %b1) {
BB0:
%a = zext i320 %A to i640
%b = zext i320 %B to i640
br label %Loop
Loop:
%C = phi i640 [ %a, %BB0 ], [ %b, %Loop ]
br i1 %b1, label %Loop, label %Exit
Exit: ; preds = %Loop
ret i640 %C
; CHECK-LABEL: @test14a(
; CHECK: Loop:
; CHECK-NEXT: phi i320
}
define i160 @test14b(i320 %A, i320 %B, i1 %b1) {
BB0:
%a = trunc i320 %A to i160
%b = trunc i320 %B to i160
br label %Loop
Loop:
%C = phi i160 [ %a, %BB0 ], [ %b, %Loop ]
br i1 %b1, label %Loop, label %Exit
Exit: ; preds = %Loop
ret i160 %C
; CHECK-LABEL: @test14b(
; CHECK: Loop:
; CHECK-NEXT: phi i160
}
declare i64 @test15a(i64)
define i64 @test15b(i64 %A, i1 %b) {
; CHECK-LABEL: @test15b(
entry:
%i0 = zext i64 %A to i128
%i1 = shl i128 %i0, 64
%i = or i128 %i1, %i0
br i1 %b, label %one, label %two
; CHECK: entry:
; CHECK-NEXT: br i1 %b
one:
%x = phi i128 [%i, %entry], [%y, %two]
%x1 = lshr i128 %x, 64
%x2 = trunc i128 %x1 to i64
%c = call i64 @test15a(i64 %x2)
%c1 = zext i64 %c to i128
br label %two
; CHECK: one:
; CHECK-NEXT: phi i64
; CHECK-NEXT: %c = call i64 @test15a
two:
%y = phi i128 [%i, %entry], [%c1, %one]
%y1 = lshr i128 %y, 64
%y2 = trunc i128 %y1 to i64
%d = call i64 @test15a(i64 %y2)
%d1 = trunc i64 %d to i1
br i1 %d1, label %one, label %end
; CHECK: two:
; CHECK-NEXT: phi i64
; CHECK-NEXT: phi i64
; CHECK-NEXT: %d = call i64 @test15a
end:
%g = trunc i128 %y to i64
ret i64 %g
; CHECK: end:
; CHECK-NEXT: ret i64
}
; PR6512 - Shouldn't merge loads from different addr spaces.
define i32 @test16(i32 addrspace(1)* %pointer1, i32 %flag, i32* %pointer2)
nounwind {
entry:
%retval = alloca i32, align 4 ; <i32*> [#uses=2]
%pointer1.addr = alloca i32 addrspace(1)*, align 4 ; <i32 addrspace(1)**>
%flag.addr = alloca i32, align 4 ; <i32*> [#uses=2]
%pointer2.addr = alloca i32*, align 4 ; <i32**> [#uses=2]
%res = alloca i32, align 4 ; <i32*> [#uses=4]
store i32 addrspace(1)* %pointer1, i32 addrspace(1)** %pointer1.addr
store i32 %flag, i32* %flag.addr
store i32* %pointer2, i32** %pointer2.addr
store i32 10, i32* %res
%tmp = load i32, i32* %flag.addr ; <i32> [#uses=1]
%tobool = icmp ne i32 %tmp, 0 ; <i1> [#uses=1]
br i1 %tobool, label %if.then, label %if.else
return: ; preds = %if.end
%tmp7 = load i32, i32* %retval ; <i32> [#uses=1]
ret i32 %tmp7
if.end: ; preds = %if.else, %if.then
%tmp6 = load i32, i32* %res ; <i32> [#uses=1]
store i32 %tmp6, i32* %retval
br label %return
if.then: ; preds = %entry
%tmp1 = load i32 addrspace(1)*, i32 addrspace(1)** %pointer1.addr ; <i32 addrspace(1)*>
[opaque pointer type] Add textual IR support for explicit type parameter to getelementptr instruction One of several parallel first steps to remove the target type of pointers, replacing them with a single opaque pointer type. This adds an explicit type parameter to the gep instruction so that when the first parameter becomes an opaque pointer type, the type to gep through is still available to the instructions. * This doesn't modify gep operators, only instructions (operators will be handled separately) * Textual IR changes only. Bitcode (including upgrade) and changing the in-memory representation will be in separate changes. * geps of vectors are transformed as: getelementptr <4 x float*> %x, ... ->getelementptr float, <4 x float*> %x, ... Then, once the opaque pointer type is introduced, this will ultimately look like: getelementptr float, <4 x ptr> %x with the unambiguous interpretation that it is a vector of pointers to float. * address spaces remain on the pointer, not the type: getelementptr float addrspace(1)* %x ->getelementptr float, float addrspace(1)* %x Then, eventually: getelementptr float, ptr addrspace(1) %x Importantly, the massive amount of test case churn has been automated by same crappy python code. I had to manually update a few test cases that wouldn't fit the script's model (r228970,r229196,r229197,r229198). The python script just massages stdin and writes the result to stdout, I then wrapped that in a shell script to handle replacing files, then using the usual find+xargs to migrate all the files. update.py: import fileinput import sys import re ibrep = re.compile(r"(^.*?[^%\w]getelementptr inbounds )(((?:<\d* x )?)(.*?)(| addrspace\(\d\)) *\*(|>)(?:$| *(?:%|@|null|undef|blockaddress|getelementptr|addrspacecast|bitcast|inttoptr|\[\[[a-zA-Z]|\{\{).*$))") normrep = re.compile( r"(^.*?[^%\w]getelementptr )(((?:<\d* x )?)(.*?)(| addrspace\(\d\)) *\*(|>)(?:$| *(?:%|@|null|undef|blockaddress|getelementptr|addrspacecast|bitcast|inttoptr|\[\[[a-zA-Z]|\{\{).*$))") def conv(match, line): if not match: return line line = match.groups()[0] if len(match.groups()[5]) == 0: line += match.groups()[2] line += match.groups()[3] line += ", " line += match.groups()[1] line += "\n" return line for line in sys.stdin: if line.find("getelementptr ") == line.find("getelementptr inbounds"): if line.find("getelementptr inbounds") != line.find("getelementptr inbounds ("): line = conv(re.match(ibrep, line), line) elif line.find("getelementptr ") != line.find("getelementptr ("): line = conv(re.match(normrep, line), line) sys.stdout.write(line) apply.sh: for name in "$@" do python3 `dirname "$0"`/update.py < "$name" > "$name.tmp" && mv "$name.tmp" "$name" rm -f "$name.tmp" done The actual commands: From llvm/src: find test/ -name *.ll | xargs ./apply.sh From llvm/src/tools/clang: find test/ -name *.mm -o -name *.m -o -name *.cpp -o -name *.c | xargs -I '{}' ../../apply.sh "{}" From llvm/src/tools/polly: find test/ -name *.ll | xargs ./apply.sh After that, check-all (with llvm, clang, clang-tools-extra, lld, compiler-rt, and polly all checked out). The extra 'rm' in the apply.sh script is due to a few files in clang's test suite using interesting unicode stuff that my python script was throwing exceptions on. None of those files needed to be migrated, so it seemed sufficient to ignore those cases. Reviewers: rafael, dexonsmith, grosser Differential Revision: http://reviews.llvm.org/D7636 llvm-svn: 230786
2015-02-28 03:29:02 +08:00
%arrayidx = getelementptr i32, i32 addrspace(1)* %tmp1, i32 0 ; <i32 addrspace(1)*> [#uses=1]
%tmp2 = load i32, i32 addrspace(1)* %arrayidx ; <i32> [#uses=1]
store i32 %tmp2, i32* %res
br label %if.end
if.else: ; preds = %entry
%tmp3 = load i32*, i32** %pointer2.addr ; <i32*> [#uses=1]
[opaque pointer type] Add textual IR support for explicit type parameter to getelementptr instruction One of several parallel first steps to remove the target type of pointers, replacing them with a single opaque pointer type. This adds an explicit type parameter to the gep instruction so that when the first parameter becomes an opaque pointer type, the type to gep through is still available to the instructions. * This doesn't modify gep operators, only instructions (operators will be handled separately) * Textual IR changes only. Bitcode (including upgrade) and changing the in-memory representation will be in separate changes. * geps of vectors are transformed as: getelementptr <4 x float*> %x, ... ->getelementptr float, <4 x float*> %x, ... Then, once the opaque pointer type is introduced, this will ultimately look like: getelementptr float, <4 x ptr> %x with the unambiguous interpretation that it is a vector of pointers to float. * address spaces remain on the pointer, not the type: getelementptr float addrspace(1)* %x ->getelementptr float, float addrspace(1)* %x Then, eventually: getelementptr float, ptr addrspace(1) %x Importantly, the massive amount of test case churn has been automated by same crappy python code. I had to manually update a few test cases that wouldn't fit the script's model (r228970,r229196,r229197,r229198). The python script just massages stdin and writes the result to stdout, I then wrapped that in a shell script to handle replacing files, then using the usual find+xargs to migrate all the files. update.py: import fileinput import sys import re ibrep = re.compile(r"(^.*?[^%\w]getelementptr inbounds )(((?:<\d* x )?)(.*?)(| addrspace\(\d\)) *\*(|>)(?:$| *(?:%|@|null|undef|blockaddress|getelementptr|addrspacecast|bitcast|inttoptr|\[\[[a-zA-Z]|\{\{).*$))") normrep = re.compile( r"(^.*?[^%\w]getelementptr )(((?:<\d* x )?)(.*?)(| addrspace\(\d\)) *\*(|>)(?:$| *(?:%|@|null|undef|blockaddress|getelementptr|addrspacecast|bitcast|inttoptr|\[\[[a-zA-Z]|\{\{).*$))") def conv(match, line): if not match: return line line = match.groups()[0] if len(match.groups()[5]) == 0: line += match.groups()[2] line += match.groups()[3] line += ", " line += match.groups()[1] line += "\n" return line for line in sys.stdin: if line.find("getelementptr ") == line.find("getelementptr inbounds"): if line.find("getelementptr inbounds") != line.find("getelementptr inbounds ("): line = conv(re.match(ibrep, line), line) elif line.find("getelementptr ") != line.find("getelementptr ("): line = conv(re.match(normrep, line), line) sys.stdout.write(line) apply.sh: for name in "$@" do python3 `dirname "$0"`/update.py < "$name" > "$name.tmp" && mv "$name.tmp" "$name" rm -f "$name.tmp" done The actual commands: From llvm/src: find test/ -name *.ll | xargs ./apply.sh From llvm/src/tools/clang: find test/ -name *.mm -o -name *.m -o -name *.cpp -o -name *.c | xargs -I '{}' ../../apply.sh "{}" From llvm/src/tools/polly: find test/ -name *.ll | xargs ./apply.sh After that, check-all (with llvm, clang, clang-tools-extra, lld, compiler-rt, and polly all checked out). The extra 'rm' in the apply.sh script is due to a few files in clang's test suite using interesting unicode stuff that my python script was throwing exceptions on. None of those files needed to be migrated, so it seemed sufficient to ignore those cases. Reviewers: rafael, dexonsmith, grosser Differential Revision: http://reviews.llvm.org/D7636 llvm-svn: 230786
2015-02-28 03:29:02 +08:00
%arrayidx4 = getelementptr i32, i32* %tmp3, i32 0 ; <i32*> [#uses=1]
%tmp5 = load i32, i32* %arrayidx4 ; <i32> [#uses=1]
store i32 %tmp5, i32* %res
br label %if.end
}
; PR4413
declare i32 @ext()
; CHECK-LABEL: @test17(
define i32 @test17(i1 %a) {
entry:
br i1 %a, label %bb1, label %bb2
bb1: ; preds = %entry
%0 = tail call i32 @ext() ; <i32> [#uses=1]
br label %bb2
bb2: ; preds = %bb1, %entry
%cond = phi i1 [ true, %bb1 ], [ false, %entry ] ; <i1> [#uses=1]
; CHECK-NOT: %val = phi i32 [ %0, %bb1 ], [ 0, %entry ]
%val = phi i32 [ %0, %bb1 ], [ 0, %entry ] ; <i32> [#uses=1]
%res = select i1 %cond, i32 %val, i32 0 ; <i32> [#uses=1]
; CHECK: ret i32 %cond
ret i32 %res
}
define i1 @test18(i1 %cond) {
%zero = alloca i32
%one = alloca i32
br i1 %cond, label %true, label %false
true:
br label %ret
false:
br label %ret
ret:
%ptr = phi i32* [ %zero, %true ] , [ %one, %false ]
%isnull = icmp eq i32* %ptr, null
ret i1 %isnull
; CHECK-LABEL: @test18(
; CHECK: ret i1 false
}
define i1 @test19(i1 %cond, double %x) {
br i1 %cond, label %true, label %false
true:
br label %ret
false:
br label %ret
ret:
%p = phi double [ %x, %true ], [ 0x7FF0000000000000, %false ]; RHS = +infty
%cmp = fcmp ule double %x, %p
ret i1 %cmp
; CHECK-LABEL: @test19(
; CHECK: ret i1 true
}
define i1 @test20(i1 %cond) {
%a = alloca i32
%b = alloca i32
%c = alloca i32
br i1 %cond, label %true, label %false
true:
br label %ret
false:
br label %ret
ret:
%p = phi i32* [ %a, %true ], [ %b, %false ]
%r = icmp eq i32* %p, %c
ret i1 %r
; CHECK-LABEL: @test20(
; CHECK: ret i1 false
}
define i1 @test21(i1 %c1, i1 %c2) {
%a = alloca i32
%b = alloca i32
%c = alloca i32
br i1 %c1, label %true, label %false
true:
br label %loop
false:
br label %loop
loop:
%p = phi i32* [ %a, %true ], [ %b, %false ], [ %p, %loop ]
%r = icmp eq i32* %p, %c
br i1 %c2, label %ret, label %loop
ret:
ret i1 %r
; CHECK-LABEL: @test21(
; CHECK: ret i1 false
}
define void @test22() {
; CHECK-LABEL: @test22(
entry:
br label %loop
loop:
%phi = phi i32 [ 0, %entry ], [ %y, %loop ]
%y = add i32 %phi, 1
%o = or i32 %y, %phi
%e = icmp eq i32 %o, %y
br i1 %e, label %loop, label %ret
; CHECK: br i1 %e
ret:
ret void
}
define i32 @test23(i32 %A, i1 %b, i32 * %P) {
BB0:
br label %Loop
Loop: ; preds = %Loop, %BB0
; PHI has same value always.
%B = phi i32 [ %A, %BB0 ], [ 42, %Loop ]
%D = add i32 %B, 19
store i32 %D, i32* %P
br i1 %b, label %Loop, label %Exit
Exit: ; preds = %Loop
%E = add i32 %B, 19
ret i32 %E
; CHECK-LABEL: @test23(
; CHECK: %phitmp = add i32 %A, 19
; CHECK: Loop:
; CHECK-NEXT: %B = phi i32 [ %phitmp, %BB0 ], [ 61, %Loop ]
; CHECK: Exit:
; CHECK-NEXT: ret i32 %B
}
define i32 @test24(i32 %A, i1 %cond) {
BB0:
%X = add nuw nsw i32 %A, 1
br i1 %cond, label %BB1, label %BB2
BB1:
%Y = add nuw i32 %A, 1
br label %BB2
BB2:
%C = phi i32 [ %X, %BB0 ], [ %Y, %BB1 ]
ret i32 %C
; CHECK-LABEL: @test24(
; CHECK-NOT: phi
; CHECK: BB2:
; CHECK-NEXT: %C = add nuw i32 %A, 1
; CHECK-NEXT: ret i32 %C
}
; Same as test11, but used to be missed due to a bug.
declare i1 @test25a()
define i1 @test25() {
entry:
%a = alloca i32
%i = ptrtoint i32* %a to i64
%b = call i1 @test25a()
br i1 %b, label %one, label %two
one:
%x = phi i64 [%y, %two], [%i, %entry]
%c = call i1 @test25a()
br i1 %c, label %two, label %end
two:
%y = phi i64 [%x, %one], [%i, %entry]
%d = call i1 @test25a()
br i1 %d, label %one, label %end
end:
%f = phi i64 [ %x, %one], [%y, %two]
; Change the %f to %i, and the optimizer suddenly becomes a lot smarter
; even though %f must equal %i at this point
%g = inttoptr i64 %f to i32*
store i32 10, i32* %g
%z = call i1 @test25a()
ret i1 %z
; CHECK-LABEL: @test25(
; CHECK-NOT: phi i32
; CHECK: ret i1 %z
}
declare i1 @test26a()
define i1 @test26(i32 %n) {
entry:
%a = alloca i32
%i = ptrtoint i32* %a to i64
%b = call i1 @test26a()
br label %one
one:
%x = phi i64 [%y, %two], [%w, %three], [%i, %entry]
%c = call i1 @test26a()
switch i32 %n, label %end [
i32 2, label %two
i32 3, label %three
]
two:
%y = phi i64 [%x, %one], [%w, %three]
%d = call i1 @test26a()
switch i32 %n, label %end [
i32 10, label %one
i32 30, label %three
]
three:
%w = phi i64 [%y, %two], [%x, %one]
%e = call i1 @test26a()
br i1 %e, label %one, label %two
end:
%f = phi i64 [ %x, %one], [%y, %two]
; Change the %f to %i, and the optimizer suddenly becomes a lot smarter
; even though %f must equal %i at this point
%g = inttoptr i64 %f to i32*
store i32 10, i32* %g
%z = call i1 @test26a()
ret i1 %z
; CHECK-LABEL: @test26(
; CHECK-NOT: phi i32
; CHECK: ret i1 %z
}
; CHECK-LABEL: @test27(
; CHECK: ret i32 undef
define i32 @test27(i1 %b) {
entry:
br label %done
done:
%y = phi i32 [ undef, %entry ]
ret i32 %y
}
; We should be able to fold the zexts to the other side of the phi
; even though there's a constant value input to the phi. This is
; because we can shrink that constant to the smaller phi type.
define i1 @PR24766(i8 %x1, i8 %x2, i8 %condition) {
entry:
%conv = sext i8 %condition to i32
switch i32 %conv, label %epilog [
i32 0, label %sw1
i32 1, label %sw2
]
sw1:
%cmp1 = icmp eq i8 %x1, %x2
%frombool1 = zext i1 %cmp1 to i8
br label %epilog
sw2:
%cmp2 = icmp sle i8 %x1, %x2
%frombool2 = zext i1 %cmp2 to i8
br label %epilog
epilog:
%conditionMet = phi i8 [ 0, %entry ], [ %frombool2, %sw2 ], [ %frombool1, %sw1 ]
%tobool = icmp ne i8 %conditionMet, 0
ret i1 %tobool
; CHECK-LABEL: @PR24766(
; CHECK: %[[RES:.*]] = phi i1 [ false, %entry ], [ %cmp2, %sw2 ], [ %cmp1, %sw1 ]
; CHECK-NEXT: ret i1 %[[RES]]
}
; Same as above (a phi with more than 2 operands), but no constants
define i1 @PR24766_no_constants(i8 %x1, i8 %x2, i8 %condition, i1 %another_condition) {
entry:
%frombool0 = zext i1 %another_condition to i8
%conv = sext i8 %condition to i32
switch i32 %conv, label %epilog [
i32 0, label %sw1
i32 1, label %sw2
]
sw1:
%cmp1 = icmp eq i8 %x1, %x2
%frombool1 = zext i1 %cmp1 to i8
br label %epilog
sw2:
%cmp2 = icmp sle i8 %x1, %x2
%frombool2 = zext i1 %cmp2 to i8
br label %epilog
epilog:
%conditionMet = phi i8 [ %frombool0, %entry ], [ %frombool2, %sw2 ], [ %frombool1, %sw1 ]
%tobool = icmp ne i8 %conditionMet, 0
ret i1 %tobool
; CHECK-LABEL: @PR24766_no_constants(
; CHECK: %[[RES:.*]] = phi i1 [ %another_condition, %entry ], [ %cmp2, %sw2 ], [ %cmp1, %sw1 ]
; CHECK-NEXT: ret i1 %[[RES]]
}
; Same as above (a phi with more than 2 operands), but two constants
define i1 @PR24766_two_constants(i8 %x1, i8 %x2, i8 %condition) {
entry:
%conv = sext i8 %condition to i32
switch i32 %conv, label %epilog [
i32 0, label %sw1
i32 1, label %sw2
]
sw1:
%cmp1 = icmp eq i8 %x1, %x2
%frombool1 = zext i1 %cmp1 to i8
br label %epilog
sw2:
%cmp2 = icmp sle i8 %x1, %x2
%frombool2 = zext i1 %cmp2 to i8
br label %epilog
epilog:
%conditionMet = phi i8 [ 0, %entry ], [ 1, %sw2 ], [ %frombool1, %sw1 ]
%tobool = icmp ne i8 %conditionMet, 0
ret i1 %tobool
; CHECK-LABEL: @PR24766_two_constants(
; CHECK: %[[RES:.*]] = phi i1 [ false, %entry ], [ true, %sw2 ], [ %cmp1, %sw1 ]
; CHECK-NEXT: ret i1 %[[RES]]
}
; Same as above (a phi with more than 2 operands), but two constants and two variables
define i1 @PR24766_two_constants_two_var(i8 %x1, i8 %x2, i8 %condition) {
entry:
%conv = sext i8 %condition to i32
switch i32 %conv, label %epilog [
i32 0, label %sw1
i32 1, label %sw2
i32 2, label %sw3
]
sw1:
%cmp1 = icmp eq i8 %x1, %x2
%frombool1 = zext i1 %cmp1 to i8
br label %epilog
sw2:
%cmp2 = icmp sle i8 %x1, %x2
%frombool2 = zext i1 %cmp2 to i8
br label %epilog
sw3:
%cmp3 = icmp sge i8 %x1, %x2
%frombool3 = zext i1 %cmp3 to i8
br label %epilog
epilog:
%conditionMet = phi i8 [ 0, %entry ], [ %frombool2, %sw2 ], [ %frombool1, %sw1 ], [ 1, %sw3 ]
%tobool = icmp ne i8 %conditionMet, 0
ret i1 %tobool
; CHECK-LABEL: @PR24766_two_constants_two_var(
; CHECK: %[[RES:.*]] = phi i1 [ false, %entry ], [ %cmp2, %sw2 ], [ %cmp1, %sw1 ], [ true, %sw3 ]
; CHECK-NEXT: ret i1 %[[RES]]
}
; CHECK-LABEL: phi_allnonzeroconstant
; CHECK-NOT: phi i32
; CHECK: ret i1 false
define i1 @phi_allnonzeroconstant(i1 %c, i32 %a, i32 %b) {
entry:
br i1 %c, label %if.then, label %if.else
if.then: ; preds = %entry
br label %if.end
if.else: ; preds = %entry
call void @dummy()
br label %if.end
if.end: ; preds = %if.else, %if.then
%x.0 = phi i32 [ 1, %if.then ], [ 2, %if.else ]
%or = or i32 %x.0, %a
%cmp1 = icmp eq i32 %or, 0
ret i1 %cmp1
}
declare void @dummy()
; CHECK-LABEL: @phi_knownnonzero_eq
; CHECK-LABEL: if.then:
; CHECK-NOT: select
; CHECK-LABEL: if.end:
; CHECK: phi i32 [ 1, %if.then ]
define i1 @phi_knownnonzero_eq(i32 %n, i32 %s, i32* nocapture readonly %P) {
entry:
%tobool = icmp slt i32 %n, %s
br i1 %tobool, label %if.end, label %if.then
if.then: ; preds = %entry
%0 = load i32, i32* %P
%cmp = icmp eq i32 %n, %0
%1 = select i1 %cmp, i32 1, i32 2
br label %if.end
if.end: ; preds = %entry, %if.then
%a.0 = phi i32 [ %1, %if.then ], [ %n, %entry ]
%cmp1 = icmp eq i32 %a.0, 0
ret i1 %cmp1
}
; CHECK-LABEL: @phi_knownnonzero_ne
; CHECK-LABEL: if.then:
; CHECK-NOT: select
; CHECK-LABEL: if.end:
; CHECK: phi i32 [ 1, %if.then ]
define i1 @phi_knownnonzero_ne(i32 %n, i32 %s, i32* nocapture readonly %P) {
entry:
%tobool = icmp slt i32 %n, %s
br i1 %tobool, label %if.end, label %if.then
if.then: ; preds = %entry
%0 = load i32, i32* %P
%cmp = icmp eq i32 %n, %0
%1 = select i1 %cmp, i32 1, i32 2
br label %if.end
if.end: ; preds = %entry, %if.then
%a.0 = phi i32 [ %1, %if.then ], [ %n, %entry ]
%cmp1 = icmp ne i32 %a.0, 0
ret i1 %cmp1
}
; CHECK-LABEL: @phi_knownnonzero_eq_2
; CHECK-LABEL: if.then:
; CHECK-NOT: select
; CHECK-LABEL: if.end:
; CHECK: phi i32 [ 2, %if.else ]
define i1 @phi_knownnonzero_eq_2(i32 %n, i32 %s, i32* nocapture readonly %P) {
entry:
%tobool = icmp slt i32 %n, %s
br i1 %tobool, label %if.then, label %if.end
if.then:
%tobool2 = icmp slt i32 %n, %s
br i1 %tobool2, label %if.else, label %if.end
if.else: ; preds = %entry
%0 = load i32, i32* %P
%cmp = icmp eq i32 %n, %0
%1 = select i1 %cmp, i32 1, i32 2
br label %if.end
if.end: ; preds = %entry, %if.then
%a.0 = phi i32 [ %1, %if.else], [ %n, %entry ], [2, %if.then]
%cmp1 = icmp eq i32 %a.0, 0
ret i1 %cmp1
}
; CHECK-LABEL: @phi_knownnonzero_ne_2
; CHECK-LABEL: if.then:
; CHECK-NOT: select
; CHECK-LABEL: if.end:
; CHECK: phi i32 [ 2, %if.else ]
define i1 @phi_knownnonzero_ne_2(i32 %n, i32 %s, i32* nocapture readonly %P) {
entry:
%tobool = icmp slt i32 %n, %s
br i1 %tobool, label %if.then, label %if.end
if.then:
%tobool2 = icmp slt i32 %n, %s
br i1 %tobool2, label %if.else, label %if.end
if.else: ; preds = %entry
%0 = load i32, i32* %P
%cmp = icmp eq i32 %n, %0
%1 = select i1 %cmp, i32 1, i32 2
br label %if.end
if.end: ; preds = %entry, %if.then
%a.0 = phi i32 [ %1, %if.else], [ %n, %entry ], [2, %if.then]
%cmp1 = icmp ne i32 %a.0, 0
ret i1 %cmp1
}