1
0
mirror of https://github.com/RPCS3/llvm-mirror.git synced 2024-10-24 21:42:54 +02:00
llvm-mirror/test/Transforms/InstCombine/objsize.ll

279 lines
8.5 KiB
LLVM
Raw Normal View History

; Test a pile of objectsize bounds checking.
; RUN: opt < %s -instcombine -S | FileCheck %s
; We need target data to get the sizes of the arrays and structures.
target datalayout = "e-p:32:32:32-i1:8:8-i8:8:8-i16:16:16-i32:32:32-i64:32:64-f32:32:32-f64:32:64-v64:64:64-v128:128:128-a0:0:64-f80:128:128"
@a = private global [60 x i8] zeroinitializer, align 1 ; <[60 x i8]*>
@.str = private constant [8 x i8] c"abcdefg\00" ; <[8 x i8]*>
define i32 @foo() nounwind {
; CHECK-LABEL: @foo(
; CHECK-NEXT: ret i32 60
%1 = call i32 @llvm.objectsize.i32.p0i8(i8* getelementptr inbounds ([60 x i8]* @a, i32 0, i32 0), i1 false)
ret i32 %1
}
define i8* @bar() nounwind {
; CHECK-LABEL: @bar(
entry:
%retval = alloca i8*
%0 = call i32 @llvm.objectsize.i32.p0i8(i8* getelementptr inbounds ([60 x i8]* @a, i32 0, i32 0), i1 false)
%cmp = icmp ne i32 %0, -1
; CHECK: br i1 true
br i1 %cmp, label %cond.true, label %cond.false
cond.true:
%1 = load i8** %retval
ret i8* %1
cond.false:
%2 = load i8** %retval
ret i8* %2
}
define i32 @f() nounwind {
; CHECK-LABEL: @f(
; CHECK-NEXT: ret i32 0
%1 = call i32 @llvm.objectsize.i32.p0i8(i8* getelementptr ([60 x i8]* @a, i32 1, i32 0), i1 false)
ret i32 %1
}
@window = external global [0 x i8]
define i1 @baz() nounwind {
; CHECK-LABEL: @baz(
; CHECK-NEXT: objectsize
%1 = tail call i32 @llvm.objectsize.i32.p0i8(i8* getelementptr inbounds ([0 x i8]* @window, i32 0, i32 0), i1 false)
%2 = icmp eq i32 %1, -1
ret i1 %2
}
define void @test1(i8* %q, i32 %x) nounwind noinline {
; CHECK-LABEL: @test1(
; CHECK: objectsize.i32.p0i8
entry:
%0 = call i32 @llvm.objectsize.i32.p0i8(i8* getelementptr inbounds ([0 x i8]* @window, i32 0, i32 10), i1 false) ; <i64> [#uses=1]
%1 = icmp eq i32 %0, -1 ; <i1> [#uses=1]
br i1 %1, label %"47", label %"46"
"46": ; preds = %entry
unreachable
"47": ; preds = %entry
unreachable
}
@.str5 = private constant [9 x i32] [i32 97, i32 98, i32 99, i32 100, i32 0, i32
101, i32 102, i32 103, i32 0], align 4
define i32 @test2() nounwind {
; CHECK-LABEL: @test2(
; CHECK-NEXT: ret i32 34
%1 = call i32 @llvm.objectsize.i32.p0i8(i8* getelementptr (i8* bitcast ([9 x i32]* @.str5 to i8*), i32 2), i1 false)
ret i32 %1
}
; rdar://7674946
@array = internal global [480 x float] zeroinitializer ; <[480 x float]*> [#uses=1]
declare i8* @__memcpy_chk(i8*, i8*, i32, i32) nounwind
declare i32 @llvm.objectsize.i32.p0i8(i8*, i1) nounwind readonly
declare i8* @__inline_memcpy_chk(i8*, i8*, i32) nounwind inlinehint
define void @test3() nounwind {
; CHECK-LABEL: @test3(
entry:
br i1 undef, label %bb11, label %bb12
bb11:
[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
2015-02-27 20:29:02 +01:00
%0 = getelementptr inbounds float, float* getelementptr inbounds ([480 x float]* @array, i32 0, i32 128), i32 -127 ; <float*> [#uses=1]
%1 = bitcast float* %0 to i8* ; <i8*> [#uses=1]
%2 = call i32 @llvm.objectsize.i32.p0i8(i8* %1, i1 false) ; <i32> [#uses=1]
%3 = call i8* @__memcpy_chk(i8* undef, i8* undef, i32 512, i32 %2) nounwind ; <i8*> [#uses=0]
; CHECK: unreachable
unreachable
bb12:
[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
2015-02-27 20:29:02 +01:00
%4 = getelementptr inbounds float, float* getelementptr inbounds ([480 x float]* @array, i32 0, i32 128), i32 -127 ; <float*> [#uses=1]
%5 = bitcast float* %4 to i8* ; <i8*> [#uses=1]
%6 = call i8* @__inline_memcpy_chk(i8* %5, i8* undef, i32 512) nounwind inlinehint ; <i8*> [#uses=0]
; CHECK: @__inline_memcpy_chk
unreachable
}
; rdar://7718857
%struct.data = type { [100 x i32], [100 x i32], [1024 x i8] }
define i32 @test4(i8** %esc) nounwind ssp {
; CHECK-LABEL: @test4(
entry:
%0 = alloca %struct.data, align 8
%1 = bitcast %struct.data* %0 to i8*
%2 = call i32 @llvm.objectsize.i32.p0i8(i8* %1, i1 false) nounwind
; CHECK-NOT: @llvm.objectsize
; CHECK: @llvm.memset.p0i8.i32(i8* %1, i8 0, i32 1824, i32 8, i1 false)
%3 = call i8* @__memset_chk(i8* %1, i32 0, i32 1824, i32 %2) nounwind
store i8* %1, i8** %esc
ret i32 0
}
; rdar://7782496
@s = external global i8*
define i8* @test5(i32 %n) nounwind ssp {
; CHECK-LABEL: @test5(
entry:
%0 = tail call noalias i8* @malloc(i32 20) nounwind
%1 = tail call i32 @llvm.objectsize.i32.p0i8(i8* %0, i1 false)
%2 = load i8** @s, align 8
; CHECK-NOT: @llvm.objectsize
; CHECK: @llvm.memcpy.p0i8.p0i8.i32(i8* %0, i8* %1, i32 10, i32 1, i1 false)
%3 = tail call i8* @__memcpy_chk(i8* %0, i8* %2, i32 10, i32 %1) nounwind
ret i8* %0
}
define void @test6(i32 %n) nounwind ssp {
; CHECK-LABEL: @test6(
entry:
%0 = tail call noalias i8* @malloc(i32 20) nounwind
%1 = tail call i32 @llvm.objectsize.i32.p0i8(i8* %0, i1 false)
%2 = load i8** @s, align 8
; CHECK-NOT: @llvm.objectsize
; CHECK: @__memcpy_chk(i8* %0, i8* %1, i32 30, i32 20)
%3 = tail call i8* @__memcpy_chk(i8* %0, i8* %2, i32 30, i32 %1) nounwind
ret void
}
declare i8* @__memset_chk(i8*, i32, i32, i32) nounwind
declare noalias i8* @malloc(i32) nounwind
define i32 @test7(i8** %esc) {
; CHECK-LABEL: @test7(
%alloc = call noalias i8* @malloc(i32 48) nounwind
store i8* %alloc, i8** %esc
[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
2015-02-27 20:29:02 +01:00
%gep = getelementptr inbounds i8, i8* %alloc, i32 16
%objsize = call i32 @llvm.objectsize.i32.p0i8(i8* %gep, i1 false) nounwind readonly
; CHECK: ret i32 32
ret i32 %objsize
}
declare noalias i8* @calloc(i32, i32) nounwind
define i32 @test8(i8** %esc) {
; CHECK-LABEL: @test8(
%alloc = call noalias i8* @calloc(i32 5, i32 7) nounwind
store i8* %alloc, i8** %esc
[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
2015-02-27 20:29:02 +01:00
%gep = getelementptr inbounds i8, i8* %alloc, i32 5
%objsize = call i32 @llvm.objectsize.i32.p0i8(i8* %gep, i1 false) nounwind readonly
; CHECK: ret i32 30
ret i32 %objsize
}
declare noalias i8* @strdup(i8* nocapture) nounwind
declare noalias i8* @strndup(i8* nocapture, i32) nounwind
; CHECK-LABEL: @test9(
define i32 @test9(i8** %esc) {
%call = tail call i8* @strdup(i8* getelementptr inbounds ([8 x i8]* @.str, i64 0, i64 0)) nounwind
store i8* %call, i8** %esc, align 8
%1 = tail call i32 @llvm.objectsize.i32.p0i8(i8* %call, i1 true)
; CHECK: ret i32 8
ret i32 %1
}
; CHECK-LABEL: @test10(
define i32 @test10(i8** %esc) {
%call = tail call i8* @strndup(i8* getelementptr inbounds ([8 x i8]* @.str, i64 0, i64 0), i32 3) nounwind
store i8* %call, i8** %esc, align 8
%1 = tail call i32 @llvm.objectsize.i32.p0i8(i8* %call, i1 true)
; CHECK: ret i32 4
ret i32 %1
}
; CHECK-LABEL: @test11(
define i32 @test11(i8** %esc) {
%call = tail call i8* @strndup(i8* getelementptr inbounds ([8 x i8]* @.str, i64 0, i64 0), i32 7) nounwind
store i8* %call, i8** %esc, align 8
%1 = tail call i32 @llvm.objectsize.i32.p0i8(i8* %call, i1 true)
; CHECK: ret i32 8
ret i32 %1
}
; CHECK-LABEL: @test12(
define i32 @test12(i8** %esc) {
%call = tail call i8* @strndup(i8* getelementptr inbounds ([8 x i8]* @.str, i64 0, i64 0), i32 8) nounwind
store i8* %call, i8** %esc, align 8
%1 = tail call i32 @llvm.objectsize.i32.p0i8(i8* %call, i1 true)
; CHECK: ret i32 8
ret i32 %1
}
; CHECK-LABEL: @test13(
define i32 @test13(i8** %esc) {
%call = tail call i8* @strndup(i8* getelementptr inbounds ([8 x i8]* @.str, i64 0, i64 0), i32 57) nounwind
store i8* %call, i8** %esc, align 8
%1 = tail call i32 @llvm.objectsize.i32.p0i8(i8* %call, i1 true)
; CHECK: ret i32 8
ret i32 %1
}
; CHECK-LABEL: @PR13390(
define i32 @PR13390(i1 %bool, i8* %a) {
entry:
%cond = or i1 %bool, true
br i1 %cond, label %return, label %xpto
xpto:
%select = select i1 %bool, i8* %select, i8* %a
%select2 = select i1 %bool, i8* %a, i8* %select2
%0 = tail call i32 @llvm.objectsize.i32.p0i8(i8* %select, i1 true)
%1 = tail call i32 @llvm.objectsize.i32.p0i8(i8* %select2, i1 true)
%2 = add i32 %0, %1
; CHECK: ret i32 undef
ret i32 %2
return:
ret i32 42
}
; CHECK-LABEL: @PR13621(
define i32 @PR13621(i1 %bool) nounwind {
entry:
%cond = or i1 %bool, true
br i1 %cond, label %return, label %xpto
; technically reachable, but this malformed IR may appear as a result of constant propagation
xpto:
[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
2015-02-27 20:29:02 +01:00
%gep2 = getelementptr i8, i8* %gep, i32 1
%gep = getelementptr i8, i8* %gep2, i32 1
%o = call i32 @llvm.objectsize.i32.p0i8(i8* %gep, i1 true)
; CHECK: ret i32 undef
ret i32 %o
return:
ret i32 7
}
2013-04-20 19:39:52 +02:00
@globalalias = internal alias [60 x i8]* @a
2013-04-20 19:39:52 +02:00
; CHECK-LABEL: @test18(
2013-04-20 19:39:52 +02:00
; CHECK-NEXT: ret i32 60
define i32 @test18() {
%bc = bitcast [60 x i8]* @globalalias to i8*
%1 = call i32 @llvm.objectsize.i32.p0i8(i8* %bc, i1 false)
2013-04-20 19:39:52 +02:00
ret i32 %1
}
@globalalias2 = weak alias [60 x i8]* @a
2013-04-20 19:39:52 +02:00
; CHECK-LABEL: @test19(
2013-04-20 19:39:52 +02:00
; CHECK: llvm.objectsize
define i32 @test19() {
%bc = bitcast [60 x i8]* @globalalias2 to i8*
%1 = call i32 @llvm.objectsize.i32.p0i8(i8* %bc, i1 false)
2013-04-20 19:39:52 +02:00
ret i32 %1
}