maxjhandsome
/
llvm-project
forked from OSchip/llvm-project
1
0
Fork 0
Code Issues Pull Requests Packages Releases Wiki Activity
llvm-project/llvm/test/CodeGen/Hexagon/hwloop-lt.ll

410 lines
10 KiB
LLVM
Raw Normal View History

[Hexagon] Adding a number of other tests for min/max instructions and loading i1s. llvm-svn: 239935
2015-06-18 04:29:33 +08:00
; RUN: llc -march=hexagon -O3 < %s | FileCheck %s
Extend Hexagon hardware loop generation to handle various additional cases: - variety of compare instructions, - loops with no preheader, - arbitrary lower and upper bounds. llvm-svn: 174904
2013-02-12 05:37:55 +08:00
[Hexagon] Generate more hardware loops Refactored parts of the hardware loop pass to generate more. Also, added more tests. Differential Revision: http://reviews.llvm.org/D9568 llvm-svn: 236896
2015-05-09 04:18:21 +08:00
; CHECK-LABEL: @test_pos1_ir_slt
Extend Hexagon hardware loop generation to handle various additional cases: - variety of compare instructions, - loops with no preheader, - arbitrary lower and upper bounds. llvm-svn: 174904
2013-02-12 05:37:55 +08:00
; CHECK: loop0
; a < b
define void @test_pos1_ir_slt(i8* nocapture %p, i32 %a, i32 %b) nounwind {
entry:
%cmp3 = icmp slt i32 8531, %b
br i1 %cmp3, label %for.body.lr.ph, label %for.end
[Hexagon] Generate more hardware loops Refactored parts of the hardware loop pass to generate more. Also, added more tests. Differential Revision: http://reviews.llvm.org/D9568 llvm-svn: 236896
2015-05-09 04:18:21 +08:00
for.body.lr.ph:
Extend Hexagon hardware loop generation to handle various additional cases: - variety of compare instructions, - loops with no preheader, - arbitrary lower and upper bounds. llvm-svn: 174904
2013-02-12 05:37:55 +08:00
br label %for.body
[Hexagon] Generate more hardware loops Refactored parts of the hardware loop pass to generate more. Also, added more tests. Differential Revision: http://reviews.llvm.org/D9568 llvm-svn: 236896
2015-05-09 04:18:21 +08:00
for.body:
Extend Hexagon hardware loop generation to handle various additional cases: - variety of compare instructions, - loops with no preheader, - arbitrary lower and upper bounds. llvm-svn: 174904
2013-02-12 05:37:55 +08:00
%i.04 = phi i32 [ 8531, %for.body.lr.ph ], [ %inc, %for.body ]
[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 inbounds i8, i8* %p, i32 %i.04
[opaque pointer type] Add textual IR support for explicit type parameter to load instruction Essentially the same as the GEP change in r230786. A similar migration script can be used to update test cases, though a few more test case improvements/changes were required this time around: (r229269-r229278) import fileinput import sys import re pat = re.compile(r"((?:=|:|^)\s*load (?:atomic )?(?:volatile )?(.*?))(| addrspace\(\d+\) *)\*($| *(?:%|@|null|undef|blockaddress|getelementptr|addrspacecast|bitcast|inttoptr|\[\[[a-zA-Z]|\{\{).*$)") for line in sys.stdin: sys.stdout.write(re.sub(pat, r"\1, \2\3*\4", line)) Reviewers: rafael, dexonsmith, grosser Differential Revision: http://reviews.llvm.org/D7649 llvm-svn: 230794
2015-02-28 05:17:42 +08:00
%0 = load i8, i8* %arrayidx, align 1
Extend Hexagon hardware loop generation to handle various additional cases: - variety of compare instructions, - loops with no preheader, - arbitrary lower and upper bounds. llvm-svn: 174904
2013-02-12 05:37:55 +08:00
%conv = zext i8 %0 to i32
%add = add nsw i32 %conv, 1
%conv1 = trunc i32 %add to i8
store i8 %conv1, i8* %arrayidx, align 1
%inc = add nsw i32 %i.04, 1
%cmp = icmp slt i32 %inc, %b
br i1 %cmp, label %for.body, label %for.end
[Hexagon] Generate more hardware loops Refactored parts of the hardware loop pass to generate more. Also, added more tests. Differential Revision: http://reviews.llvm.org/D9568 llvm-svn: 236896
2015-05-09 04:18:21 +08:00
for.end:
Extend Hexagon hardware loop generation to handle various additional cases: - variety of compare instructions, - loops with no preheader, - arbitrary lower and upper bounds. llvm-svn: 174904
2013-02-12 05:37:55 +08:00
ret void
}
[Hexagon] Generate more hardware loops Refactored parts of the hardware loop pass to generate more. Also, added more tests. Differential Revision: http://reviews.llvm.org/D9568 llvm-svn: 236896
2015-05-09 04:18:21 +08:00
; CHECK-LABEL: @test_pos2_ir_slt
Extend Hexagon hardware loop generation to handle various additional cases: - variety of compare instructions, - loops with no preheader, - arbitrary lower and upper bounds. llvm-svn: 174904
2013-02-12 05:37:55 +08:00
; CHECK: loop0
; a < b
define void @test_pos2_ir_slt(i8* nocapture %p, i32 %a, i32 %b) nounwind {
entry:
%cmp3 = icmp slt i32 9152, %b
br i1 %cmp3, label %for.body.lr.ph, label %for.end
[Hexagon] Generate more hardware loops Refactored parts of the hardware loop pass to generate more. Also, added more tests. Differential Revision: http://reviews.llvm.org/D9568 llvm-svn: 236896
2015-05-09 04:18:21 +08:00
for.body.lr.ph:
Extend Hexagon hardware loop generation to handle various additional cases: - variety of compare instructions, - loops with no preheader, - arbitrary lower and upper bounds. llvm-svn: 174904
2013-02-12 05:37:55 +08:00
br label %for.body
[Hexagon] Generate more hardware loops Refactored parts of the hardware loop pass to generate more. Also, added more tests. Differential Revision: http://reviews.llvm.org/D9568 llvm-svn: 236896
2015-05-09 04:18:21 +08:00
for.body:
Extend Hexagon hardware loop generation to handle various additional cases: - variety of compare instructions, - loops with no preheader, - arbitrary lower and upper bounds. llvm-svn: 174904
2013-02-12 05:37:55 +08:00
%i.04 = phi i32 [ 9152, %for.body.lr.ph ], [ %inc, %for.body ]
[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 inbounds i8, i8* %p, i32 %i.04
[opaque pointer type] Add textual IR support for explicit type parameter to load instruction Essentially the same as the GEP change in r230786. A similar migration script can be used to update test cases, though a few more test case improvements/changes were required this time around: (r229269-r229278) import fileinput import sys import re pat = re.compile(r"((?:=|:|^)\s*load (?:atomic )?(?:volatile )?(.*?))(| addrspace\(\d+\) *)\*($| *(?:%|@|null|undef|blockaddress|getelementptr|addrspacecast|bitcast|inttoptr|\[\[[a-zA-Z]|\{\{).*$)") for line in sys.stdin: sys.stdout.write(re.sub(pat, r"\1, \2\3*\4", line)) Reviewers: rafael, dexonsmith, grosser Differential Revision: http://reviews.llvm.org/D7649 llvm-svn: 230794
2015-02-28 05:17:42 +08:00
%0 = load i8, i8* %arrayidx, align 1
Extend Hexagon hardware loop generation to handle various additional cases: - variety of compare instructions, - loops with no preheader, - arbitrary lower and upper bounds. llvm-svn: 174904
2013-02-12 05:37:55 +08:00
%conv = zext i8 %0 to i32
%add = add nsw i32 %conv, 1
%conv1 = trunc i32 %add to i8
store i8 %conv1, i8* %arrayidx, align 1
%inc = add nsw i32 %i.04, 2
%cmp = icmp slt i32 %inc, %b
br i1 %cmp, label %for.body, label %for.end
[Hexagon] Generate more hardware loops Refactored parts of the hardware loop pass to generate more. Also, added more tests. Differential Revision: http://reviews.llvm.org/D9568 llvm-svn: 236896
2015-05-09 04:18:21 +08:00
for.end:
Extend Hexagon hardware loop generation to handle various additional cases: - variety of compare instructions, - loops with no preheader, - arbitrary lower and upper bounds. llvm-svn: 174904
2013-02-12 05:37:55 +08:00
ret void
}
[Hexagon] Generate more hardware loops Refactored parts of the hardware loop pass to generate more. Also, added more tests. Differential Revision: http://reviews.llvm.org/D9568 llvm-svn: 236896
2015-05-09 04:18:21 +08:00
; CHECK-LABEL: @test_pos4_ir_slt
Extend Hexagon hardware loop generation to handle various additional cases: - variety of compare instructions, - loops with no preheader, - arbitrary lower and upper bounds. llvm-svn: 174904
2013-02-12 05:37:55 +08:00
; CHECK: loop0
; a < b
define void @test_pos4_ir_slt(i8* nocapture %p, i32 %a, i32 %b) nounwind {
entry:
%cmp3 = icmp slt i32 18851, %b
br i1 %cmp3, label %for.body.lr.ph, label %for.end
[Hexagon] Generate more hardware loops Refactored parts of the hardware loop pass to generate more. Also, added more tests. Differential Revision: http://reviews.llvm.org/D9568 llvm-svn: 236896
2015-05-09 04:18:21 +08:00
for.body.lr.ph:
Extend Hexagon hardware loop generation to handle various additional cases: - variety of compare instructions, - loops with no preheader, - arbitrary lower and upper bounds. llvm-svn: 174904
2013-02-12 05:37:55 +08:00
br label %for.body
[Hexagon] Generate more hardware loops Refactored parts of the hardware loop pass to generate more. Also, added more tests. Differential Revision: http://reviews.llvm.org/D9568 llvm-svn: 236896
2015-05-09 04:18:21 +08:00
for.body:
Extend Hexagon hardware loop generation to handle various additional cases: - variety of compare instructions, - loops with no preheader, - arbitrary lower and upper bounds. llvm-svn: 174904
2013-02-12 05:37:55 +08:00
%i.04 = phi i32 [ 18851, %for.body.lr.ph ], [ %inc, %for.body ]
[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 inbounds i8, i8* %p, i32 %i.04
[opaque pointer type] Add textual IR support for explicit type parameter to load instruction Essentially the same as the GEP change in r230786. A similar migration script can be used to update test cases, though a few more test case improvements/changes were required this time around: (r229269-r229278) import fileinput import sys import re pat = re.compile(r"((?:=|:|^)\s*load (?:atomic )?(?:volatile )?(.*?))(| addrspace\(\d+\) *)\*($| *(?:%|@|null|undef|blockaddress|getelementptr|addrspacecast|bitcast|inttoptr|\[\[[a-zA-Z]|\{\{).*$)") for line in sys.stdin: sys.stdout.write(re.sub(pat, r"\1, \2\3*\4", line)) Reviewers: rafael, dexonsmith, grosser Differential Revision: http://reviews.llvm.org/D7649 llvm-svn: 230794
2015-02-28 05:17:42 +08:00
%0 = load i8, i8* %arrayidx, align 1
Extend Hexagon hardware loop generation to handle various additional cases: - variety of compare instructions, - loops with no preheader, - arbitrary lower and upper bounds. llvm-svn: 174904
2013-02-12 05:37:55 +08:00
%conv = zext i8 %0 to i32
%add = add nsw i32 %conv, 1
%conv1 = trunc i32 %add to i8
store i8 %conv1, i8* %arrayidx, align 1
%inc = add nsw i32 %i.04, 4
%cmp = icmp slt i32 %inc, %b
br i1 %cmp, label %for.body, label %for.end
[Hexagon] Generate more hardware loops Refactored parts of the hardware loop pass to generate more. Also, added more tests. Differential Revision: http://reviews.llvm.org/D9568 llvm-svn: 236896
2015-05-09 04:18:21 +08:00
for.end:
Extend Hexagon hardware loop generation to handle various additional cases: - variety of compare instructions, - loops with no preheader, - arbitrary lower and upper bounds. llvm-svn: 174904
2013-02-12 05:37:55 +08:00
ret void
}
[Hexagon] Generate more hardware loops Refactored parts of the hardware loop pass to generate more. Also, added more tests. Differential Revision: http://reviews.llvm.org/D9568 llvm-svn: 236896
2015-05-09 04:18:21 +08:00
; CHECK-LABEL: @test_pos8_ir_slt
Extend Hexagon hardware loop generation to handle various additional cases: - variety of compare instructions, - loops with no preheader, - arbitrary lower and upper bounds. llvm-svn: 174904
2013-02-12 05:37:55 +08:00
; CHECK: loop0
; a < b
define void @test_pos8_ir_slt(i8* nocapture %p, i32 %a, i32 %b) nounwind {
entry:
%cmp3 = icmp slt i32 25466, %b
br i1 %cmp3, label %for.body.lr.ph, label %for.end
[Hexagon] Generate more hardware loops Refactored parts of the hardware loop pass to generate more. Also, added more tests. Differential Revision: http://reviews.llvm.org/D9568 llvm-svn: 236896
2015-05-09 04:18:21 +08:00
for.body.lr.ph:
Extend Hexagon hardware loop generation to handle various additional cases: - variety of compare instructions, - loops with no preheader, - arbitrary lower and upper bounds. llvm-svn: 174904
2013-02-12 05:37:55 +08:00
br label %for.body
[Hexagon] Generate more hardware loops Refactored parts of the hardware loop pass to generate more. Also, added more tests. Differential Revision: http://reviews.llvm.org/D9568 llvm-svn: 236896
2015-05-09 04:18:21 +08:00
for.body:
Extend Hexagon hardware loop generation to handle various additional cases: - variety of compare instructions, - loops with no preheader, - arbitrary lower and upper bounds. llvm-svn: 174904
2013-02-12 05:37:55 +08:00
%i.04 = phi i32 [ 25466, %for.body.lr.ph ], [ %inc, %for.body ]
[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 inbounds i8, i8* %p, i32 %i.04
[opaque pointer type] Add textual IR support for explicit type parameter to load instruction Essentially the same as the GEP change in r230786. A similar migration script can be used to update test cases, though a few more test case improvements/changes were required this time around: (r229269-r229278) import fileinput import sys import re pat = re.compile(r"((?:=|:|^)\s*load (?:atomic )?(?:volatile )?(.*?))(| addrspace\(\d+\) *)\*($| *(?:%|@|null|undef|blockaddress|getelementptr|addrspacecast|bitcast|inttoptr|\[\[[a-zA-Z]|\{\{).*$)") for line in sys.stdin: sys.stdout.write(re.sub(pat, r"\1, \2\3*\4", line)) Reviewers: rafael, dexonsmith, grosser Differential Revision: http://reviews.llvm.org/D7649 llvm-svn: 230794
2015-02-28 05:17:42 +08:00
%0 = load i8, i8* %arrayidx, align 1
Extend Hexagon hardware loop generation to handle various additional cases: - variety of compare instructions, - loops with no preheader, - arbitrary lower and upper bounds. llvm-svn: 174904
2013-02-12 05:37:55 +08:00
%conv = zext i8 %0 to i32
%add = add nsw i32 %conv, 1
%conv1 = trunc i32 %add to i8
store i8 %conv1, i8* %arrayidx, align 1
%inc = add nsw i32 %i.04, 8
%cmp = icmp slt i32 %inc, %b
br i1 %cmp, label %for.body, label %for.end
[Hexagon] Generate more hardware loops Refactored parts of the hardware loop pass to generate more. Also, added more tests. Differential Revision: http://reviews.llvm.org/D9568 llvm-svn: 236896
2015-05-09 04:18:21 +08:00
for.end:
Extend Hexagon hardware loop generation to handle various additional cases: - variety of compare instructions, - loops with no preheader, - arbitrary lower and upper bounds. llvm-svn: 174904
2013-02-12 05:37:55 +08:00
ret void
}
[Hexagon] Generate more hardware loops Refactored parts of the hardware loop pass to generate more. Also, added more tests. Differential Revision: http://reviews.llvm.org/D9568 llvm-svn: 236896
2015-05-09 04:18:21 +08:00
; CHECK-LABEL: @test_pos16_ir_slt
Extend Hexagon hardware loop generation to handle various additional cases: - variety of compare instructions, - loops with no preheader, - arbitrary lower and upper bounds. llvm-svn: 174904
2013-02-12 05:37:55 +08:00
; CHECK: loop0
; a < b
define void @test_pos16_ir_slt(i8* nocapture %p, i32 %a, i32 %b) nounwind {
entry:
%cmp3 = icmp slt i32 9295, %b
br i1 %cmp3, label %for.body.lr.ph, label %for.end
[Hexagon] Generate more hardware loops Refactored parts of the hardware loop pass to generate more. Also, added more tests. Differential Revision: http://reviews.llvm.org/D9568 llvm-svn: 236896
2015-05-09 04:18:21 +08:00
for.body.lr.ph:
Extend Hexagon hardware loop generation to handle various additional cases: - variety of compare instructions, - loops with no preheader, - arbitrary lower and upper bounds. llvm-svn: 174904
2013-02-12 05:37:55 +08:00
br label %for.body
[Hexagon] Generate more hardware loops Refactored parts of the hardware loop pass to generate more. Also, added more tests. Differential Revision: http://reviews.llvm.org/D9568 llvm-svn: 236896
2015-05-09 04:18:21 +08:00
for.body:
Extend Hexagon hardware loop generation to handle various additional cases: - variety of compare instructions, - loops with no preheader, - arbitrary lower and upper bounds. llvm-svn: 174904
2013-02-12 05:37:55 +08:00
%i.04 = phi i32 [ 9295, %for.body.lr.ph ], [ %inc, %for.body ]
[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 inbounds i8, i8* %p, i32 %i.04
[opaque pointer type] Add textual IR support for explicit type parameter to load instruction Essentially the same as the GEP change in r230786. A similar migration script can be used to update test cases, though a few more test case improvements/changes were required this time around: (r229269-r229278) import fileinput import sys import re pat = re.compile(r"((?:=|:|^)\s*load (?:atomic )?(?:volatile )?(.*?))(| addrspace\(\d+\) *)\*($| *(?:%|@|null|undef|blockaddress|getelementptr|addrspacecast|bitcast|inttoptr|\[\[[a-zA-Z]|\{\{).*$)") for line in sys.stdin: sys.stdout.write(re.sub(pat, r"\1, \2\3*\4", line)) Reviewers: rafael, dexonsmith, grosser Differential Revision: http://reviews.llvm.org/D7649 llvm-svn: 230794
2015-02-28 05:17:42 +08:00
%0 = load i8, i8* %arrayidx, align 1
Extend Hexagon hardware loop generation to handle various additional cases: - variety of compare instructions, - loops with no preheader, - arbitrary lower and upper bounds. llvm-svn: 174904
2013-02-12 05:37:55 +08:00
%conv = zext i8 %0 to i32
%add = add nsw i32 %conv, 1
%conv1 = trunc i32 %add to i8
store i8 %conv1, i8* %arrayidx, align 1
%inc = add nsw i32 %i.04, 16
%cmp = icmp slt i32 %inc, %b
br i1 %cmp, label %for.body, label %for.end
[Hexagon] Generate more hardware loops Refactored parts of the hardware loop pass to generate more. Also, added more tests. Differential Revision: http://reviews.llvm.org/D9568 llvm-svn: 236896
2015-05-09 04:18:21 +08:00
for.end:
Extend Hexagon hardware loop generation to handle various additional cases: - variety of compare instructions, - loops with no preheader, - arbitrary lower and upper bounds. llvm-svn: 174904
2013-02-12 05:37:55 +08:00
ret void
}
[Hexagon] Generate more hardware loops Refactored parts of the hardware loop pass to generate more. Also, added more tests. Differential Revision: http://reviews.llvm.org/D9568 llvm-svn: 236896
2015-05-09 04:18:21 +08:00
; CHECK-LABEL: @test_pos1_ri_slt
Extend Hexagon hardware loop generation to handle various additional cases: - variety of compare instructions, - loops with no preheader, - arbitrary lower and upper bounds. llvm-svn: 174904
2013-02-12 05:37:55 +08:00
; CHECK: loop0
; a < b
define void @test_pos1_ri_slt(i8* nocapture %p, i32 %a, i32 %b) nounwind {
entry:
%cmp3 = icmp slt i32 %a, 31236
br i1 %cmp3, label %for.body.lr.ph, label %for.end
[Hexagon] Generate more hardware loops Refactored parts of the hardware loop pass to generate more. Also, added more tests. Differential Revision: http://reviews.llvm.org/D9568 llvm-svn: 236896
2015-05-09 04:18:21 +08:00
for.body.lr.ph:
Extend Hexagon hardware loop generation to handle various additional cases: - variety of compare instructions, - loops with no preheader, - arbitrary lower and upper bounds. llvm-svn: 174904
2013-02-12 05:37:55 +08:00
br label %for.body
[Hexagon] Generate more hardware loops Refactored parts of the hardware loop pass to generate more. Also, added more tests. Differential Revision: http://reviews.llvm.org/D9568 llvm-svn: 236896
2015-05-09 04:18:21 +08:00
for.body:
Extend Hexagon hardware loop generation to handle various additional cases: - variety of compare instructions, - loops with no preheader, - arbitrary lower and upper bounds. llvm-svn: 174904
2013-02-12 05:37:55 +08:00
%i.04 = phi i32 [ %a, %for.body.lr.ph ], [ %inc, %for.body ]
[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 inbounds i8, i8* %p, i32 %i.04
[opaque pointer type] Add textual IR support for explicit type parameter to load instruction Essentially the same as the GEP change in r230786. A similar migration script can be used to update test cases, though a few more test case improvements/changes were required this time around: (r229269-r229278) import fileinput import sys import re pat = re.compile(r"((?:=|:|^)\s*load (?:atomic )?(?:volatile )?(.*?))(| addrspace\(\d+\) *)\*($| *(?:%|@|null|undef|blockaddress|getelementptr|addrspacecast|bitcast|inttoptr|\[\[[a-zA-Z]|\{\{).*$)") for line in sys.stdin: sys.stdout.write(re.sub(pat, r"\1, \2\3*\4", line)) Reviewers: rafael, dexonsmith, grosser Differential Revision: http://reviews.llvm.org/D7649 llvm-svn: 230794
2015-02-28 05:17:42 +08:00
%0 = load i8, i8* %arrayidx, align 1
Extend Hexagon hardware loop generation to handle various additional cases: - variety of compare instructions, - loops with no preheader, - arbitrary lower and upper bounds. llvm-svn: 174904
2013-02-12 05:37:55 +08:00
%conv = zext i8 %0 to i32
%add = add nsw i32 %conv, 1
%conv1 = trunc i32 %add to i8
store i8 %conv1, i8* %arrayidx, align 1
%inc = add nsw i32 %i.04, 1
%cmp = icmp slt i32 %inc, 31236
br i1 %cmp, label %for.body, label %for.end
[Hexagon] Generate more hardware loops Refactored parts of the hardware loop pass to generate more. Also, added more tests. Differential Revision: http://reviews.llvm.org/D9568 llvm-svn: 236896
2015-05-09 04:18:21 +08:00
for.end:
Extend Hexagon hardware loop generation to handle various additional cases: - variety of compare instructions, - loops with no preheader, - arbitrary lower and upper bounds. llvm-svn: 174904
2013-02-12 05:37:55 +08:00
ret void
}
[Hexagon] Generate more hardware loops Refactored parts of the hardware loop pass to generate more. Also, added more tests. Differential Revision: http://reviews.llvm.org/D9568 llvm-svn: 236896
2015-05-09 04:18:21 +08:00
; CHECK-LABEL: @test_pos2_ri_slt
Extend Hexagon hardware loop generation to handle various additional cases: - variety of compare instructions, - loops with no preheader, - arbitrary lower and upper bounds. llvm-svn: 174904
2013-02-12 05:37:55 +08:00
; CHECK: loop0
; a < b
define void @test_pos2_ri_slt(i8* nocapture %p, i32 %a, i32 %b) nounwind {
entry:
%cmp3 = icmp slt i32 %a, 22653
br i1 %cmp3, label %for.body.lr.ph, label %for.end
[Hexagon] Generate more hardware loops Refactored parts of the hardware loop pass to generate more. Also, added more tests. Differential Revision: http://reviews.llvm.org/D9568 llvm-svn: 236896
2015-05-09 04:18:21 +08:00
for.body.lr.ph:
Extend Hexagon hardware loop generation to handle various additional cases: - variety of compare instructions, - loops with no preheader, - arbitrary lower and upper bounds. llvm-svn: 174904
2013-02-12 05:37:55 +08:00
br label %for.body
[Hexagon] Generate more hardware loops Refactored parts of the hardware loop pass to generate more. Also, added more tests. Differential Revision: http://reviews.llvm.org/D9568 llvm-svn: 236896
2015-05-09 04:18:21 +08:00
for.body:
Extend Hexagon hardware loop generation to handle various additional cases: - variety of compare instructions, - loops with no preheader, - arbitrary lower and upper bounds. llvm-svn: 174904
2013-02-12 05:37:55 +08:00
%i.04 = phi i32 [ %a, %for.body.lr.ph ], [ %inc, %for.body ]
[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 inbounds i8, i8* %p, i32 %i.04
[opaque pointer type] Add textual IR support for explicit type parameter to load instruction Essentially the same as the GEP change in r230786. A similar migration script can be used to update test cases, though a few more test case improvements/changes were required this time around: (r229269-r229278) import fileinput import sys import re pat = re.compile(r"((?:=|:|^)\s*load (?:atomic )?(?:volatile )?(.*?))(| addrspace\(\d+\) *)\*($| *(?:%|@|null|undef|blockaddress|getelementptr|addrspacecast|bitcast|inttoptr|\[\[[a-zA-Z]|\{\{).*$)") for line in sys.stdin: sys.stdout.write(re.sub(pat, r"\1, \2\3*\4", line)) Reviewers: rafael, dexonsmith, grosser Differential Revision: http://reviews.llvm.org/D7649 llvm-svn: 230794
2015-02-28 05:17:42 +08:00
%0 = load i8, i8* %arrayidx, align 1
Extend Hexagon hardware loop generation to handle various additional cases: - variety of compare instructions, - loops with no preheader, - arbitrary lower and upper bounds. llvm-svn: 174904
2013-02-12 05:37:55 +08:00
%conv = zext i8 %0 to i32
%add = add nsw i32 %conv, 1
%conv1 = trunc i32 %add to i8
store i8 %conv1, i8* %arrayidx, align 1
%inc = add nsw i32 %i.04, 2
%cmp = icmp slt i32 %inc, 22653
br i1 %cmp, label %for.body, label %for.end
[Hexagon] Generate more hardware loops Refactored parts of the hardware loop pass to generate more. Also, added more tests. Differential Revision: http://reviews.llvm.org/D9568 llvm-svn: 236896
2015-05-09 04:18:21 +08:00
for.end:
Extend Hexagon hardware loop generation to handle various additional cases: - variety of compare instructions, - loops with no preheader, - arbitrary lower and upper bounds. llvm-svn: 174904
2013-02-12 05:37:55 +08:00
ret void
}
[Hexagon] Generate more hardware loops Refactored parts of the hardware loop pass to generate more. Also, added more tests. Differential Revision: http://reviews.llvm.org/D9568 llvm-svn: 236896
2015-05-09 04:18:21 +08:00
; CHECK-LABEL: @test_pos4_ri_slt
Extend Hexagon hardware loop generation to handle various additional cases: - variety of compare instructions, - loops with no preheader, - arbitrary lower and upper bounds. llvm-svn: 174904
2013-02-12 05:37:55 +08:00
; CHECK: loop0
; a < b
define void @test_pos4_ri_slt(i8* nocapture %p, i32 %a, i32 %b) nounwind {
entry:
%cmp3 = icmp slt i32 %a, 1431
br i1 %cmp3, label %for.body.lr.ph, label %for.end
[Hexagon] Generate more hardware loops Refactored parts of the hardware loop pass to generate more. Also, added more tests. Differential Revision: http://reviews.llvm.org/D9568 llvm-svn: 236896
2015-05-09 04:18:21 +08:00
for.body.lr.ph:
Extend Hexagon hardware loop generation to handle various additional cases: - variety of compare instructions, - loops with no preheader, - arbitrary lower and upper bounds. llvm-svn: 174904
2013-02-12 05:37:55 +08:00
br label %for.body
[Hexagon] Generate more hardware loops Refactored parts of the hardware loop pass to generate more. Also, added more tests. Differential Revision: http://reviews.llvm.org/D9568 llvm-svn: 236896
2015-05-09 04:18:21 +08:00
for.body:
Extend Hexagon hardware loop generation to handle various additional cases: - variety of compare instructions, - loops with no preheader, - arbitrary lower and upper bounds. llvm-svn: 174904
2013-02-12 05:37:55 +08:00
%i.04 = phi i32 [ %a, %for.body.lr.ph ], [ %inc, %for.body ]
[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 inbounds i8, i8* %p, i32 %i.04
[opaque pointer type] Add textual IR support for explicit type parameter to load instruction Essentially the same as the GEP change in r230786. A similar migration script can be used to update test cases, though a few more test case improvements/changes were required this time around: (r229269-r229278) import fileinput import sys import re pat = re.compile(r"((?:=|:|^)\s*load (?:atomic )?(?:volatile )?(.*?))(| addrspace\(\d+\) *)\*($| *(?:%|@|null|undef|blockaddress|getelementptr|addrspacecast|bitcast|inttoptr|\[\[[a-zA-Z]|\{\{).*$)") for line in sys.stdin: sys.stdout.write(re.sub(pat, r"\1, \2\3*\4", line)) Reviewers: rafael, dexonsmith, grosser Differential Revision: http://reviews.llvm.org/D7649 llvm-svn: 230794
2015-02-28 05:17:42 +08:00
%0 = load i8, i8* %arrayidx, align 1
Extend Hexagon hardware loop generation to handle various additional cases: - variety of compare instructions, - loops with no preheader, - arbitrary lower and upper bounds. llvm-svn: 174904
2013-02-12 05:37:55 +08:00
%conv = zext i8 %0 to i32
%add = add nsw i32 %conv, 1
%conv1 = trunc i32 %add to i8
store i8 %conv1, i8* %arrayidx, align 1
%inc = add nsw i32 %i.04, 4
%cmp = icmp slt i32 %inc, 1431
br i1 %cmp, label %for.body, label %for.end
[Hexagon] Generate more hardware loops Refactored parts of the hardware loop pass to generate more. Also, added more tests. Differential Revision: http://reviews.llvm.org/D9568 llvm-svn: 236896
2015-05-09 04:18:21 +08:00
for.end:
Extend Hexagon hardware loop generation to handle various additional cases: - variety of compare instructions, - loops with no preheader, - arbitrary lower and upper bounds. llvm-svn: 174904
2013-02-12 05:37:55 +08:00
ret void
}
[Hexagon] Generate more hardware loops Refactored parts of the hardware loop pass to generate more. Also, added more tests. Differential Revision: http://reviews.llvm.org/D9568 llvm-svn: 236896
2015-05-09 04:18:21 +08:00
; CHECK-LABEL: @test_pos8_ri_slt
Extend Hexagon hardware loop generation to handle various additional cases: - variety of compare instructions, - loops with no preheader, - arbitrary lower and upper bounds. llvm-svn: 174904
2013-02-12 05:37:55 +08:00
; CHECK: loop0
; a < b
define void @test_pos8_ri_slt(i8* nocapture %p, i32 %a, i32 %b) nounwind {
entry:
%cmp3 = icmp slt i32 %a, 22403
br i1 %cmp3, label %for.body.lr.ph, label %for.end
[Hexagon] Generate more hardware loops Refactored parts of the hardware loop pass to generate more. Also, added more tests. Differential Revision: http://reviews.llvm.org/D9568 llvm-svn: 236896
2015-05-09 04:18:21 +08:00
for.body.lr.ph:
Extend Hexagon hardware loop generation to handle various additional cases: - variety of compare instructions, - loops with no preheader, - arbitrary lower and upper bounds. llvm-svn: 174904
2013-02-12 05:37:55 +08:00
br label %for.body
[Hexagon] Generate more hardware loops Refactored parts of the hardware loop pass to generate more. Also, added more tests. Differential Revision: http://reviews.llvm.org/D9568 llvm-svn: 236896
2015-05-09 04:18:21 +08:00
for.body:
Extend Hexagon hardware loop generation to handle various additional cases: - variety of compare instructions, - loops with no preheader, - arbitrary lower and upper bounds. llvm-svn: 174904
2013-02-12 05:37:55 +08:00
%i.04 = phi i32 [ %a, %for.body.lr.ph ], [ %inc, %for.body ]
[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 inbounds i8, i8* %p, i32 %i.04
[opaque pointer type] Add textual IR support for explicit type parameter to load instruction Essentially the same as the GEP change in r230786. A similar migration script can be used to update test cases, though a few more test case improvements/changes were required this time around: (r229269-r229278) import fileinput import sys import re pat = re.compile(r"((?:=|:|^)\s*load (?:atomic )?(?:volatile )?(.*?))(| addrspace\(\d+\) *)\*($| *(?:%|@|null|undef|blockaddress|getelementptr|addrspacecast|bitcast|inttoptr|\[\[[a-zA-Z]|\{\{).*$)") for line in sys.stdin: sys.stdout.write(re.sub(pat, r"\1, \2\3*\4", line)) Reviewers: rafael, dexonsmith, grosser Differential Revision: http://reviews.llvm.org/D7649 llvm-svn: 230794
2015-02-28 05:17:42 +08:00
%0 = load i8, i8* %arrayidx, align 1
Extend Hexagon hardware loop generation to handle various additional cases: - variety of compare instructions, - loops with no preheader, - arbitrary lower and upper bounds. llvm-svn: 174904
2013-02-12 05:37:55 +08:00
%conv = zext i8 %0 to i32
%add = add nsw i32 %conv, 1
%conv1 = trunc i32 %add to i8
store i8 %conv1, i8* %arrayidx, align 1
%inc = add nsw i32 %i.04, 8
%cmp = icmp slt i32 %inc, 22403
br i1 %cmp, label %for.body, label %for.end
[Hexagon] Generate more hardware loops Refactored parts of the hardware loop pass to generate more. Also, added more tests. Differential Revision: http://reviews.llvm.org/D9568 llvm-svn: 236896
2015-05-09 04:18:21 +08:00
for.end:
Extend Hexagon hardware loop generation to handle various additional cases: - variety of compare instructions, - loops with no preheader, - arbitrary lower and upper bounds. llvm-svn: 174904
2013-02-12 05:37:55 +08:00
ret void
}
[Hexagon] Generate more hardware loops Refactored parts of the hardware loop pass to generate more. Also, added more tests. Differential Revision: http://reviews.llvm.org/D9568 llvm-svn: 236896
2015-05-09 04:18:21 +08:00
; CHECK-LABEL: @test_pos16_ri_slt
Extend Hexagon hardware loop generation to handle various additional cases: - variety of compare instructions, - loops with no preheader, - arbitrary lower and upper bounds. llvm-svn: 174904
2013-02-12 05:37:55 +08:00
; CHECK: loop0
; a < b
define void @test_pos16_ri_slt(i8* nocapture %p, i32 %a, i32 %b) nounwind {
entry:
%cmp3 = icmp slt i32 %a, 21715
br i1 %cmp3, label %for.body.lr.ph, label %for.end
[Hexagon] Generate more hardware loops Refactored parts of the hardware loop pass to generate more. Also, added more tests. Differential Revision: http://reviews.llvm.org/D9568 llvm-svn: 236896
2015-05-09 04:18:21 +08:00
for.body.lr.ph:
Extend Hexagon hardware loop generation to handle various additional cases: - variety of compare instructions, - loops with no preheader, - arbitrary lower and upper bounds. llvm-svn: 174904
2013-02-12 05:37:55 +08:00
br label %for.body
[Hexagon] Generate more hardware loops Refactored parts of the hardware loop pass to generate more. Also, added more tests. Differential Revision: http://reviews.llvm.org/D9568 llvm-svn: 236896
2015-05-09 04:18:21 +08:00
for.body:
Extend Hexagon hardware loop generation to handle various additional cases: - variety of compare instructions, - loops with no preheader, - arbitrary lower and upper bounds. llvm-svn: 174904
2013-02-12 05:37:55 +08:00
%i.04 = phi i32 [ %a, %for.body.lr.ph ], [ %inc, %for.body ]
[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 inbounds i8, i8* %p, i32 %i.04
[opaque pointer type] Add textual IR support for explicit type parameter to load instruction Essentially the same as the GEP change in r230786. A similar migration script can be used to update test cases, though a few more test case improvements/changes were required this time around: (r229269-r229278) import fileinput import sys import re pat = re.compile(r"((?:=|:|^)\s*load (?:atomic )?(?:volatile )?(.*?))(| addrspace\(\d+\) *)\*($| *(?:%|@|null|undef|blockaddress|getelementptr|addrspacecast|bitcast|inttoptr|\[\[[a-zA-Z]|\{\{).*$)") for line in sys.stdin: sys.stdout.write(re.sub(pat, r"\1, \2\3*\4", line)) Reviewers: rafael, dexonsmith, grosser Differential Revision: http://reviews.llvm.org/D7649 llvm-svn: 230794
2015-02-28 05:17:42 +08:00
%0 = load i8, i8* %arrayidx, align 1
Extend Hexagon hardware loop generation to handle various additional cases: - variety of compare instructions, - loops with no preheader, - arbitrary lower and upper bounds. llvm-svn: 174904
2013-02-12 05:37:55 +08:00
%conv = zext i8 %0 to i32
%add = add nsw i32 %conv, 1
%conv1 = trunc i32 %add to i8
store i8 %conv1, i8* %arrayidx, align 1
%inc = add nsw i32 %i.04, 16
%cmp = icmp slt i32 %inc, 21715
br i1 %cmp, label %for.body, label %for.end
[Hexagon] Generate more hardware loops Refactored parts of the hardware loop pass to generate more. Also, added more tests. Differential Revision: http://reviews.llvm.org/D9568 llvm-svn: 236896
2015-05-09 04:18:21 +08:00
for.end:
Extend Hexagon hardware loop generation to handle various additional cases: - variety of compare instructions, - loops with no preheader, - arbitrary lower and upper bounds. llvm-svn: 174904
2013-02-12 05:37:55 +08:00
ret void
}
[Hexagon] Generate more hardware loops Refactored parts of the hardware loop pass to generate more. Also, added more tests. Differential Revision: http://reviews.llvm.org/D9568 llvm-svn: 236896
2015-05-09 04:18:21 +08:00
; CHECK-LABEL: @test_pos1_rr_slt
Extend Hexagon hardware loop generation to handle various additional cases: - variety of compare instructions, - loops with no preheader, - arbitrary lower and upper bounds. llvm-svn: 174904
2013-02-12 05:37:55 +08:00
; CHECK: loop0
; a < b
define void @test_pos1_rr_slt(i8* nocapture %p, i32 %a, i32 %b) nounwind {
entry:
%cmp3 = icmp slt i32 %a, %b
br i1 %cmp3, label %for.body.lr.ph, label %for.end
[Hexagon] Generate more hardware loops Refactored parts of the hardware loop pass to generate more. Also, added more tests. Differential Revision: http://reviews.llvm.org/D9568 llvm-svn: 236896
2015-05-09 04:18:21 +08:00
for.body.lr.ph:
Extend Hexagon hardware loop generation to handle various additional cases: - variety of compare instructions, - loops with no preheader, - arbitrary lower and upper bounds. llvm-svn: 174904
2013-02-12 05:37:55 +08:00
br label %for.body
[Hexagon] Generate more hardware loops Refactored parts of the hardware loop pass to generate more. Also, added more tests. Differential Revision: http://reviews.llvm.org/D9568 llvm-svn: 236896
2015-05-09 04:18:21 +08:00
for.body:
Extend Hexagon hardware loop generation to handle various additional cases: - variety of compare instructions, - loops with no preheader, - arbitrary lower and upper bounds. llvm-svn: 174904
2013-02-12 05:37:55 +08:00
%i.04 = phi i32 [ %a, %for.body.lr.ph ], [ %inc, %for.body ]
[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 inbounds i8, i8* %p, i32 %i.04
[opaque pointer type] Add textual IR support for explicit type parameter to load instruction Essentially the same as the GEP change in r230786. A similar migration script can be used to update test cases, though a few more test case improvements/changes were required this time around: (r229269-r229278) import fileinput import sys import re pat = re.compile(r"((?:=|:|^)\s*load (?:atomic )?(?:volatile )?(.*?))(| addrspace\(\d+\) *)\*($| *(?:%|@|null|undef|blockaddress|getelementptr|addrspacecast|bitcast|inttoptr|\[\[[a-zA-Z]|\{\{).*$)") for line in sys.stdin: sys.stdout.write(re.sub(pat, r"\1, \2\3*\4", line)) Reviewers: rafael, dexonsmith, grosser Differential Revision: http://reviews.llvm.org/D7649 llvm-svn: 230794
2015-02-28 05:17:42 +08:00
%0 = load i8, i8* %arrayidx, align 1
Extend Hexagon hardware loop generation to handle various additional cases: - variety of compare instructions, - loops with no preheader, - arbitrary lower and upper bounds. llvm-svn: 174904
2013-02-12 05:37:55 +08:00
%conv = zext i8 %0 to i32
%add = add nsw i32 %conv, 1
%conv1 = trunc i32 %add to i8
store i8 %conv1, i8* %arrayidx, align 1
%inc = add nsw i32 %i.04, 1
%cmp = icmp slt i32 %inc, %b
br i1 %cmp, label %for.body, label %for.end
[Hexagon] Generate more hardware loops Refactored parts of the hardware loop pass to generate more. Also, added more tests. Differential Revision: http://reviews.llvm.org/D9568 llvm-svn: 236896
2015-05-09 04:18:21 +08:00
for.end:
Extend Hexagon hardware loop generation to handle various additional cases: - variety of compare instructions, - loops with no preheader, - arbitrary lower and upper bounds. llvm-svn: 174904
2013-02-12 05:37:55 +08:00
ret void
}
[Hexagon] Generate more hardware loops Refactored parts of the hardware loop pass to generate more. Also, added more tests. Differential Revision: http://reviews.llvm.org/D9568 llvm-svn: 236896
2015-05-09 04:18:21 +08:00
; CHECK-LABEL: @test_pos2_rr_slt
Extend Hexagon hardware loop generation to handle various additional cases: - variety of compare instructions, - loops with no preheader, - arbitrary lower and upper bounds. llvm-svn: 174904
2013-02-12 05:37:55 +08:00
; CHECK: loop0
; a < b
define void @test_pos2_rr_slt(i8* nocapture %p, i32 %a, i32 %b) nounwind {
entry:
%cmp3 = icmp slt i32 %a, %b
br i1 %cmp3, label %for.body.lr.ph, label %for.end
[Hexagon] Generate more hardware loops Refactored parts of the hardware loop pass to generate more. Also, added more tests. Differential Revision: http://reviews.llvm.org/D9568 llvm-svn: 236896
2015-05-09 04:18:21 +08:00
for.body.lr.ph:
Extend Hexagon hardware loop generation to handle various additional cases: - variety of compare instructions, - loops with no preheader, - arbitrary lower and upper bounds. llvm-svn: 174904
2013-02-12 05:37:55 +08:00
br label %for.body
[Hexagon] Generate more hardware loops Refactored parts of the hardware loop pass to generate more. Also, added more tests. Differential Revision: http://reviews.llvm.org/D9568 llvm-svn: 236896
2015-05-09 04:18:21 +08:00
for.body:
Extend Hexagon hardware loop generation to handle various additional cases: - variety of compare instructions, - loops with no preheader, - arbitrary lower and upper bounds. llvm-svn: 174904
2013-02-12 05:37:55 +08:00
%i.04 = phi i32 [ %a, %for.body.lr.ph ], [ %inc, %for.body ]
[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 inbounds i8, i8* %p, i32 %i.04
[opaque pointer type] Add textual IR support for explicit type parameter to load instruction Essentially the same as the GEP change in r230786. A similar migration script can be used to update test cases, though a few more test case improvements/changes were required this time around: (r229269-r229278) import fileinput import sys import re pat = re.compile(r"((?:=|:|^)\s*load (?:atomic )?(?:volatile )?(.*?))(| addrspace\(\d+\) *)\*($| *(?:%|@|null|undef|blockaddress|getelementptr|addrspacecast|bitcast|inttoptr|\[\[[a-zA-Z]|\{\{).*$)") for line in sys.stdin: sys.stdout.write(re.sub(pat, r"\1, \2\3*\4", line)) Reviewers: rafael, dexonsmith, grosser Differential Revision: http://reviews.llvm.org/D7649 llvm-svn: 230794
2015-02-28 05:17:42 +08:00
%0 = load i8, i8* %arrayidx, align 1
Extend Hexagon hardware loop generation to handle various additional cases: - variety of compare instructions, - loops with no preheader, - arbitrary lower and upper bounds. llvm-svn: 174904
2013-02-12 05:37:55 +08:00
%conv = zext i8 %0 to i32
%add = add nsw i32 %conv, 1
%conv1 = trunc i32 %add to i8
store i8 %conv1, i8* %arrayidx, align 1
%inc = add nsw i32 %i.04, 2
%cmp = icmp slt i32 %inc, %b
br i1 %cmp, label %for.body, label %for.end
[Hexagon] Generate more hardware loops Refactored parts of the hardware loop pass to generate more. Also, added more tests. Differential Revision: http://reviews.llvm.org/D9568 llvm-svn: 236896
2015-05-09 04:18:21 +08:00
for.end:
Extend Hexagon hardware loop generation to handle various additional cases: - variety of compare instructions, - loops with no preheader, - arbitrary lower and upper bounds. llvm-svn: 174904
2013-02-12 05:37:55 +08:00
ret void
}
[Hexagon] Generate more hardware loops Refactored parts of the hardware loop pass to generate more. Also, added more tests. Differential Revision: http://reviews.llvm.org/D9568 llvm-svn: 236896
2015-05-09 04:18:21 +08:00
; CHECK-LABEL: @test_pos4_rr_slt
Extend Hexagon hardware loop generation to handle various additional cases: - variety of compare instructions, - loops with no preheader, - arbitrary lower and upper bounds. llvm-svn: 174904
2013-02-12 05:37:55 +08:00
; CHECK: loop0
; a < b
define void @test_pos4_rr_slt(i8* nocapture %p, i32 %a, i32 %b) nounwind {
entry:
%cmp3 = icmp slt i32 %a, %b
br i1 %cmp3, label %for.body.lr.ph, label %for.end
[Hexagon] Generate more hardware loops Refactored parts of the hardware loop pass to generate more. Also, added more tests. Differential Revision: http://reviews.llvm.org/D9568 llvm-svn: 236896
2015-05-09 04:18:21 +08:00
for.body.lr.ph:
Extend Hexagon hardware loop generation to handle various additional cases: - variety of compare instructions, - loops with no preheader, - arbitrary lower and upper bounds. llvm-svn: 174904
2013-02-12 05:37:55 +08:00
br label %for.body
[Hexagon] Generate more hardware loops Refactored parts of the hardware loop pass to generate more. Also, added more tests. Differential Revision: http://reviews.llvm.org/D9568 llvm-svn: 236896
2015-05-09 04:18:21 +08:00
for.body:
Extend Hexagon hardware loop generation to handle various additional cases: - variety of compare instructions, - loops with no preheader, - arbitrary lower and upper bounds. llvm-svn: 174904
2013-02-12 05:37:55 +08:00
%i.04 = phi i32 [ %a, %for.body.lr.ph ], [ %inc, %for.body ]
[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 inbounds i8, i8* %p, i32 %i.04
[opaque pointer type] Add textual IR support for explicit type parameter to load instruction Essentially the same as the GEP change in r230786. A similar migration script can be used to update test cases, though a few more test case improvements/changes were required this time around: (r229269-r229278) import fileinput import sys import re pat = re.compile(r"((?:=|:|^)\s*load (?:atomic )?(?:volatile )?(.*?))(| addrspace\(\d+\) *)\*($| *(?:%|@|null|undef|blockaddress|getelementptr|addrspacecast|bitcast|inttoptr|\[\[[a-zA-Z]|\{\{).*$)") for line in sys.stdin: sys.stdout.write(re.sub(pat, r"\1, \2\3*\4", line)) Reviewers: rafael, dexonsmith, grosser Differential Revision: http://reviews.llvm.org/D7649 llvm-svn: 230794
2015-02-28 05:17:42 +08:00
%0 = load i8, i8* %arrayidx, align 1
Extend Hexagon hardware loop generation to handle various additional cases: - variety of compare instructions, - loops with no preheader, - arbitrary lower and upper bounds. llvm-svn: 174904
2013-02-12 05:37:55 +08:00
%conv = zext i8 %0 to i32
%add = add nsw i32 %conv, 1
%conv1 = trunc i32 %add to i8
store i8 %conv1, i8* %arrayidx, align 1
%inc = add nsw i32 %i.04, 4
%cmp = icmp slt i32 %inc, %b
br i1 %cmp, label %for.body, label %for.end
[Hexagon] Generate more hardware loops Refactored parts of the hardware loop pass to generate more. Also, added more tests. Differential Revision: http://reviews.llvm.org/D9568 llvm-svn: 236896
2015-05-09 04:18:21 +08:00
for.end:
Extend Hexagon hardware loop generation to handle various additional cases: - variety of compare instructions, - loops with no preheader, - arbitrary lower and upper bounds. llvm-svn: 174904
2013-02-12 05:37:55 +08:00
ret void
}
[Hexagon] Generate more hardware loops Refactored parts of the hardware loop pass to generate more. Also, added more tests. Differential Revision: http://reviews.llvm.org/D9568 llvm-svn: 236896
2015-05-09 04:18:21 +08:00
; CHECK-LABEL: @test_pos8_rr_slt
Extend Hexagon hardware loop generation to handle various additional cases: - variety of compare instructions, - loops with no preheader, - arbitrary lower and upper bounds. llvm-svn: 174904
2013-02-12 05:37:55 +08:00
; CHECK: loop0
; a < b
define void @test_pos8_rr_slt(i8* nocapture %p, i32 %a, i32 %b) nounwind {
entry:
%cmp3 = icmp slt i32 %a, %b
br i1 %cmp3, label %for.body.lr.ph, label %for.end
[Hexagon] Generate more hardware loops Refactored parts of the hardware loop pass to generate more. Also, added more tests. Differential Revision: http://reviews.llvm.org/D9568 llvm-svn: 236896
2015-05-09 04:18:21 +08:00
for.body.lr.ph:
Extend Hexagon hardware loop generation to handle various additional cases: - variety of compare instructions, - loops with no preheader, - arbitrary lower and upper bounds. llvm-svn: 174904
2013-02-12 05:37:55 +08:00
br label %for.body
[Hexagon] Generate more hardware loops Refactored parts of the hardware loop pass to generate more. Also, added more tests. Differential Revision: http://reviews.llvm.org/D9568 llvm-svn: 236896
2015-05-09 04:18:21 +08:00
for.body:
Extend Hexagon hardware loop generation to handle various additional cases: - variety of compare instructions, - loops with no preheader, - arbitrary lower and upper bounds. llvm-svn: 174904
2013-02-12 05:37:55 +08:00
%i.04 = phi i32 [ %a, %for.body.lr.ph ], [ %inc, %for.body ]
[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 inbounds i8, i8* %p, i32 %i.04
[opaque pointer type] Add textual IR support for explicit type parameter to load instruction Essentially the same as the GEP change in r230786. A similar migration script can be used to update test cases, though a few more test case improvements/changes were required this time around: (r229269-r229278) import fileinput import sys import re pat = re.compile(r"((?:=|:|^)\s*load (?:atomic )?(?:volatile )?(.*?))(| addrspace\(\d+\) *)\*($| *(?:%|@|null|undef|blockaddress|getelementptr|addrspacecast|bitcast|inttoptr|\[\[[a-zA-Z]|\{\{).*$)") for line in sys.stdin: sys.stdout.write(re.sub(pat, r"\1, \2\3*\4", line)) Reviewers: rafael, dexonsmith, grosser Differential Revision: http://reviews.llvm.org/D7649 llvm-svn: 230794
2015-02-28 05:17:42 +08:00
%0 = load i8, i8* %arrayidx, align 1
Extend Hexagon hardware loop generation to handle various additional cases: - variety of compare instructions, - loops with no preheader, - arbitrary lower and upper bounds. llvm-svn: 174904
2013-02-12 05:37:55 +08:00
%conv = zext i8 %0 to i32
%add = add nsw i32 %conv, 1
%conv1 = trunc i32 %add to i8
store i8 %conv1, i8* %arrayidx, align 1
%inc = add nsw i32 %i.04, 8
%cmp = icmp slt i32 %inc, %b
br i1 %cmp, label %for.body, label %for.end
[Hexagon] Generate more hardware loops Refactored parts of the hardware loop pass to generate more. Also, added more tests. Differential Revision: http://reviews.llvm.org/D9568 llvm-svn: 236896
2015-05-09 04:18:21 +08:00
for.end:
Extend Hexagon hardware loop generation to handle various additional cases: - variety of compare instructions, - loops with no preheader, - arbitrary lower and upper bounds. llvm-svn: 174904
2013-02-12 05:37:55 +08:00
ret void
}
[Hexagon] Generate more hardware loops Refactored parts of the hardware loop pass to generate more. Also, added more tests. Differential Revision: http://reviews.llvm.org/D9568 llvm-svn: 236896
2015-05-09 04:18:21 +08:00
; CHECK-LABEL: @test_pos16_rr_slt
Extend Hexagon hardware loop generation to handle various additional cases: - variety of compare instructions, - loops with no preheader, - arbitrary lower and upper bounds. llvm-svn: 174904
2013-02-12 05:37:55 +08:00
; CHECK: loop0
; a < b
define void @test_pos16_rr_slt(i8* nocapture %p, i32 %a, i32 %b) nounwind {
entry:
%cmp3 = icmp slt i32 %a, %b
br i1 %cmp3, label %for.body.lr.ph, label %for.end
[Hexagon] Generate more hardware loops Refactored parts of the hardware loop pass to generate more. Also, added more tests. Differential Revision: http://reviews.llvm.org/D9568 llvm-svn: 236896
2015-05-09 04:18:21 +08:00
for.body.lr.ph:
Extend Hexagon hardware loop generation to handle various additional cases: - variety of compare instructions, - loops with no preheader, - arbitrary lower and upper bounds. llvm-svn: 174904
2013-02-12 05:37:55 +08:00
br label %for.body
[Hexagon] Generate more hardware loops Refactored parts of the hardware loop pass to generate more. Also, added more tests. Differential Revision: http://reviews.llvm.org/D9568 llvm-svn: 236896
2015-05-09 04:18:21 +08:00
for.body:
Extend Hexagon hardware loop generation to handle various additional cases: - variety of compare instructions, - loops with no preheader, - arbitrary lower and upper bounds. llvm-svn: 174904
2013-02-12 05:37:55 +08:00
%i.04 = phi i32 [ %a, %for.body.lr.ph ], [ %inc, %for.body ]
[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 inbounds i8, i8* %p, i32 %i.04
[opaque pointer type] Add textual IR support for explicit type parameter to load instruction Essentially the same as the GEP change in r230786. A similar migration script can be used to update test cases, though a few more test case improvements/changes were required this time around: (r229269-r229278) import fileinput import sys import re pat = re.compile(r"((?:=|:|^)\s*load (?:atomic )?(?:volatile )?(.*?))(| addrspace\(\d+\) *)\*($| *(?:%|@|null|undef|blockaddress|getelementptr|addrspacecast|bitcast|inttoptr|\[\[[a-zA-Z]|\{\{).*$)") for line in sys.stdin: sys.stdout.write(re.sub(pat, r"\1, \2\3*\4", line)) Reviewers: rafael, dexonsmith, grosser Differential Revision: http://reviews.llvm.org/D7649 llvm-svn: 230794
2015-02-28 05:17:42 +08:00
%0 = load i8, i8* %arrayidx, align 1
Extend Hexagon hardware loop generation to handle various additional cases: - variety of compare instructions, - loops with no preheader, - arbitrary lower and upper bounds. llvm-svn: 174904
2013-02-12 05:37:55 +08:00
%conv = zext i8 %0 to i32
%add = add nsw i32 %conv, 1
%conv1 = trunc i32 %add to i8
store i8 %conv1, i8* %arrayidx, align 1
%inc = add nsw i32 %i.04, 16
%cmp = icmp slt i32 %inc, %b
br i1 %cmp, label %for.body, label %for.end
[Hexagon] Generate more hardware loops Refactored parts of the hardware loop pass to generate more. Also, added more tests. Differential Revision: http://reviews.llvm.org/D9568 llvm-svn: 236896
2015-05-09 04:18:21 +08:00
for.end:
Extend Hexagon hardware loop generation to handle various additional cases: - variety of compare instructions, - loops with no preheader, - arbitrary lower and upper bounds. llvm-svn: 174904
2013-02-12 05:37:55 +08:00
ret void
}