2013-07-25 16:57:02 +08:00
|
|
|
; Test STOCs that are presented as selects.
|
|
|
|
|
;
|
|
|
|
|
; RUN: llc < %s -mtriple=s390x-linux-gnu -mcpu=z196 | FileCheck %s
|
|
|
|
|
|
2016-11-28 21:34:08 +08:00
|
|
|
; Run the test again to make sure it still works the same even
|
|
|
|
|
; in the presence of the load-store-on-condition-2 facility.
|
|
|
|
|
; RUN: llc < %s -mtriple=s390x-linux-gnu -mcpu=z13 | FileCheck %s
|
|
|
|
|
|
2013-07-25 16:57:02 +08:00
|
|
|
declare void @foo(i32 *)
|
|
|
|
|
|
|
|
|
|
; Test the simple case, with the loaded value first.
|
|
|
|
|
define void @f1(i32 *%ptr, i32 %alt, i32 %limit) {
|
|
|
|
|
; CHECK-LABEL: f1:
|
|
|
|
|
; CHECK: clfi %r4, 42
|
[SystemZ] Be more careful about inverting CC masks
System z branches have a mask to select which of the 4 CC values should
cause the branch to be taken. We can invert a branch by inverting the mask.
However, not all instructions can produce all 4 CC values, so inverting
the branch like this can lead to some oddities. For example, integer
comparisons only produce a CC of 0 (equal), 1 (less) or 2 (greater).
If an integer EQ is reversed to NE before instruction selection,
the branch will test for 1 or 2. If instead the branch is reversed
after instruction selection (by inverting the mask), it will test for
1, 2 or 3. Both are correct, but the second isn't really canonical.
This patch therefore keeps track of which CC values are possible
and uses this when inverting a mask.
Although this is mostly cosmestic, it fixes undefined behavior
for the CIJNLH in branch-08.ll. Another fix would have been
to mask out bit 0 when generating the fused compare and branch,
but the point of this patch is that we shouldn't need to do that
in the first place.
The patch also makes it easier to reuse CC results from other instructions.
llvm-svn: 187495
2013-07-31 20:30:20 +08:00
|
|
|
; CHECK: stoche %r3, 0(%r2)
|
2013-07-25 16:57:02 +08:00
|
|
|
; CHECK: br %r14
|
|
|
|
|
%cond = icmp ult i32 %limit, 42
|
2018-07-20 20:12:10 +08:00
|
|
|
%orig = load i32, i32 *%ptr
|
2013-07-25 16:57:02 +08:00
|
|
|
%res = select i1 %cond, i32 %orig, i32 %alt
|
|
|
|
|
store i32 %res, i32 *%ptr
|
|
|
|
|
ret void
|
|
|
|
|
}
|
|
|
|
|
|
|
|
|
|
; ...and with the loaded value second
|
|
|
|
|
define void @f2(i32 *%ptr, i32 %alt, i32 %limit) {
|
|
|
|
|
; CHECK-LABEL: f2:
|
|
|
|
|
; CHECK: clfi %r4, 42
|
|
|
|
|
; CHECK: stocl %r3, 0(%r2)
|
|
|
|
|
; CHECK: br %r14
|
|
|
|
|
%cond = icmp ult i32 %limit, 42
|
2018-07-20 20:12:10 +08:00
|
|
|
%orig = load i32, i32 *%ptr
|
2013-07-25 16:57:02 +08:00
|
|
|
%res = select i1 %cond, i32 %alt, i32 %orig
|
|
|
|
|
store i32 %res, i32 *%ptr
|
|
|
|
|
ret void
|
|
|
|
|
}
|
|
|
|
|
|
|
|
|
|
; Test cases where the value is explicitly sign-extended to 64 bits, with the
|
|
|
|
|
; loaded value first.
|
|
|
|
|
define void @f3(i32 *%ptr, i64 %alt, i32 %limit) {
|
|
|
|
|
; CHECK-LABEL: f3:
|
|
|
|
|
; CHECK: clfi %r4, 42
|
[SystemZ] Be more careful about inverting CC masks
System z branches have a mask to select which of the 4 CC values should
cause the branch to be taken. We can invert a branch by inverting the mask.
However, not all instructions can produce all 4 CC values, so inverting
the branch like this can lead to some oddities. For example, integer
comparisons only produce a CC of 0 (equal), 1 (less) or 2 (greater).
If an integer EQ is reversed to NE before instruction selection,
the branch will test for 1 or 2. If instead the branch is reversed
after instruction selection (by inverting the mask), it will test for
1, 2 or 3. Both are correct, but the second isn't really canonical.
This patch therefore keeps track of which CC values are possible
and uses this when inverting a mask.
Although this is mostly cosmestic, it fixes undefined behavior
for the CIJNLH in branch-08.ll. Another fix would have been
to mask out bit 0 when generating the fused compare and branch,
but the point of this patch is that we shouldn't need to do that
in the first place.
The patch also makes it easier to reuse CC results from other instructions.
llvm-svn: 187495
2013-07-31 20:30:20 +08:00
|
|
|
; CHECK: stoche %r3, 0(%r2)
|
2013-07-25 16:57:02 +08:00
|
|
|
; CHECK: br %r14
|
|
|
|
|
%cond = icmp ult i32 %limit, 42
|
2018-07-20 20:12:10 +08:00
|
|
|
%orig = load i32, i32 *%ptr
|
2013-07-25 16:57:02 +08:00
|
|
|
%ext = sext i32 %orig to i64
|
|
|
|
|
%res = select i1 %cond, i64 %ext, i64 %alt
|
|
|
|
|
%trunc = trunc i64 %res to i32
|
|
|
|
|
store i32 %trunc, i32 *%ptr
|
|
|
|
|
ret void
|
|
|
|
|
}
|
|
|
|
|
|
|
|
|
|
; ...and with the loaded value second
|
|
|
|
|
define void @f4(i32 *%ptr, i64 %alt, i32 %limit) {
|
|
|
|
|
; CHECK-LABEL: f4:
|
|
|
|
|
; CHECK: clfi %r4, 42
|
|
|
|
|
; CHECK: stocl %r3, 0(%r2)
|
|
|
|
|
; CHECK: br %r14
|
|
|
|
|
%cond = icmp ult i32 %limit, 42
|
2018-07-20 20:12:10 +08:00
|
|
|
%orig = load i32, i32 *%ptr
|
2013-07-25 16:57:02 +08:00
|
|
|
%ext = sext i32 %orig to i64
|
|
|
|
|
%res = select i1 %cond, i64 %alt, i64 %ext
|
|
|
|
|
%trunc = trunc i64 %res to i32
|
|
|
|
|
store i32 %trunc, i32 *%ptr
|
|
|
|
|
ret void
|
|
|
|
|
}
|
|
|
|
|
|
|
|
|
|
; Test cases where the value is explicitly zero-extended to 32 bits, with the
|
|
|
|
|
; loaded value first.
|
|
|
|
|
define void @f5(i32 *%ptr, i64 %alt, i32 %limit) {
|
|
|
|
|
; CHECK-LABEL: f5:
|
|
|
|
|
; CHECK: clfi %r4, 42
|
[SystemZ] Be more careful about inverting CC masks
System z branches have a mask to select which of the 4 CC values should
cause the branch to be taken. We can invert a branch by inverting the mask.
However, not all instructions can produce all 4 CC values, so inverting
the branch like this can lead to some oddities. For example, integer
comparisons only produce a CC of 0 (equal), 1 (less) or 2 (greater).
If an integer EQ is reversed to NE before instruction selection,
the branch will test for 1 or 2. If instead the branch is reversed
after instruction selection (by inverting the mask), it will test for
1, 2 or 3. Both are correct, but the second isn't really canonical.
This patch therefore keeps track of which CC values are possible
and uses this when inverting a mask.
Although this is mostly cosmestic, it fixes undefined behavior
for the CIJNLH in branch-08.ll. Another fix would have been
to mask out bit 0 when generating the fused compare and branch,
but the point of this patch is that we shouldn't need to do that
in the first place.
The patch also makes it easier to reuse CC results from other instructions.
llvm-svn: 187495
2013-07-31 20:30:20 +08:00
|
|
|
; CHECK: stoche %r3, 0(%r2)
|
2013-07-25 16:57:02 +08:00
|
|
|
; CHECK: br %r14
|
|
|
|
|
%cond = icmp ult i32 %limit, 42
|
2018-07-20 20:12:10 +08:00
|
|
|
%orig = load i32, i32 *%ptr
|
2013-07-25 16:57:02 +08:00
|
|
|
%ext = zext i32 %orig to i64
|
|
|
|
|
%res = select i1 %cond, i64 %ext, i64 %alt
|
|
|
|
|
%trunc = trunc i64 %res to i32
|
|
|
|
|
store i32 %trunc, i32 *%ptr
|
|
|
|
|
ret void
|
|
|
|
|
}
|
|
|
|
|
|
|
|
|
|
; ...and with the loaded value second
|
|
|
|
|
define void @f6(i32 *%ptr, i64 %alt, i32 %limit) {
|
|
|
|
|
; CHECK-LABEL: f6:
|
|
|
|
|
; CHECK: clfi %r4, 42
|
|
|
|
|
; CHECK: stocl %r3, 0(%r2)
|
|
|
|
|
; CHECK: br %r14
|
|
|
|
|
%cond = icmp ult i32 %limit, 42
|
2018-07-20 20:12:10 +08:00
|
|
|
%orig = load i32, i32 *%ptr
|
2013-07-25 16:57:02 +08:00
|
|
|
%ext = zext i32 %orig to i64
|
|
|
|
|
%res = select i1 %cond, i64 %alt, i64 %ext
|
|
|
|
|
%trunc = trunc i64 %res to i32
|
|
|
|
|
store i32 %trunc, i32 *%ptr
|
|
|
|
|
ret void
|
|
|
|
|
}
|
|
|
|
|
|
|
|
|
|
; Check the high end of the aligned STOC range.
|
|
|
|
|
define void @f7(i32 *%base, i32 %alt, i32 %limit) {
|
|
|
|
|
; CHECK-LABEL: f7:
|
|
|
|
|
; CHECK: clfi %r4, 42
|
[SystemZ] Be more careful about inverting CC masks
System z branches have a mask to select which of the 4 CC values should
cause the branch to be taken. We can invert a branch by inverting the mask.
However, not all instructions can produce all 4 CC values, so inverting
the branch like this can lead to some oddities. For example, integer
comparisons only produce a CC of 0 (equal), 1 (less) or 2 (greater).
If an integer EQ is reversed to NE before instruction selection,
the branch will test for 1 or 2. If instead the branch is reversed
after instruction selection (by inverting the mask), it will test for
1, 2 or 3. Both are correct, but the second isn't really canonical.
This patch therefore keeps track of which CC values are possible
and uses this when inverting a mask.
Although this is mostly cosmestic, it fixes undefined behavior
for the CIJNLH in branch-08.ll. Another fix would have been
to mask out bit 0 when generating the fused compare and branch,
but the point of this patch is that we shouldn't need to do that
in the first place.
The patch also makes it easier to reuse CC results from other instructions.
llvm-svn: 187495
2013-07-31 20:30:20 +08:00
|
|
|
; CHECK: stoche %r3, 524284(%r2)
|
2013-07-25 16:57:02 +08:00
|
|
|
; CHECK: br %r14
|
[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
|
|
|
%ptr = getelementptr i32, i32 *%base, i64 131071
|
2013-07-25 16:57:02 +08:00
|
|
|
%cond = icmp ult i32 %limit, 42
|
2018-07-20 20:12:10 +08:00
|
|
|
%orig = load i32, i32 *%ptr
|
2013-07-25 16:57:02 +08:00
|
|
|
%res = select i1 %cond, i32 %orig, i32 %alt
|
|
|
|
|
store i32 %res, i32 *%ptr
|
|
|
|
|
ret void
|
|
|
|
|
}
|
|
|
|
|
|
|
|
|
|
; Check the next word up. Other sequences besides this one would be OK.
|
|
|
|
|
define void @f8(i32 *%base, i32 %alt, i32 %limit) {
|
|
|
|
|
; CHECK-LABEL: f8:
|
|
|
|
|
; CHECK: agfi %r2, 524288
|
|
|
|
|
; CHECK: clfi %r4, 42
|
[SystemZ] Be more careful about inverting CC masks
System z branches have a mask to select which of the 4 CC values should
cause the branch to be taken. We can invert a branch by inverting the mask.
However, not all instructions can produce all 4 CC values, so inverting
the branch like this can lead to some oddities. For example, integer
comparisons only produce a CC of 0 (equal), 1 (less) or 2 (greater).
If an integer EQ is reversed to NE before instruction selection,
the branch will test for 1 or 2. If instead the branch is reversed
after instruction selection (by inverting the mask), it will test for
1, 2 or 3. Both are correct, but the second isn't really canonical.
This patch therefore keeps track of which CC values are possible
and uses this when inverting a mask.
Although this is mostly cosmestic, it fixes undefined behavior
for the CIJNLH in branch-08.ll. Another fix would have been
to mask out bit 0 when generating the fused compare and branch,
but the point of this patch is that we shouldn't need to do that
in the first place.
The patch also makes it easier to reuse CC results from other instructions.
llvm-svn: 187495
2013-07-31 20:30:20 +08:00
|
|
|
; CHECK: stoche %r3, 0(%r2)
|
2013-07-25 16:57:02 +08:00
|
|
|
; CHECK: br %r14
|
[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
|
|
|
%ptr = getelementptr i32, i32 *%base, i64 131072
|
2013-07-25 16:57:02 +08:00
|
|
|
%cond = icmp ult i32 %limit, 42
|
2018-07-20 20:12:10 +08:00
|
|
|
%orig = load i32, i32 *%ptr
|
2013-07-25 16:57:02 +08:00
|
|
|
%res = select i1 %cond, i32 %orig, i32 %alt
|
|
|
|
|
store i32 %res, i32 *%ptr
|
|
|
|
|
ret void
|
|
|
|
|
}
|
|
|
|
|
|
|
|
|
|
; Check the low end of the STOC range.
|
|
|
|
|
define void @f9(i32 *%base, i32 %alt, i32 %limit) {
|
|
|
|
|
; CHECK-LABEL: f9:
|
|
|
|
|
; CHECK: clfi %r4, 42
|
[SystemZ] Be more careful about inverting CC masks
System z branches have a mask to select which of the 4 CC values should
cause the branch to be taken. We can invert a branch by inverting the mask.
However, not all instructions can produce all 4 CC values, so inverting
the branch like this can lead to some oddities. For example, integer
comparisons only produce a CC of 0 (equal), 1 (less) or 2 (greater).
If an integer EQ is reversed to NE before instruction selection,
the branch will test for 1 or 2. If instead the branch is reversed
after instruction selection (by inverting the mask), it will test for
1, 2 or 3. Both are correct, but the second isn't really canonical.
This patch therefore keeps track of which CC values are possible
and uses this when inverting a mask.
Although this is mostly cosmestic, it fixes undefined behavior
for the CIJNLH in branch-08.ll. Another fix would have been
to mask out bit 0 when generating the fused compare and branch,
but the point of this patch is that we shouldn't need to do that
in the first place.
The patch also makes it easier to reuse CC results from other instructions.
llvm-svn: 187495
2013-07-31 20:30:20 +08:00
|
|
|
; CHECK: stoche %r3, -524288(%r2)
|
2013-07-25 16:57:02 +08:00
|
|
|
; CHECK: br %r14
|
[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
|
|
|
%ptr = getelementptr i32, i32 *%base, i64 -131072
|
2013-07-25 16:57:02 +08:00
|
|
|
%cond = icmp ult i32 %limit, 42
|
2018-07-20 20:12:10 +08:00
|
|
|
%orig = load i32, i32 *%ptr
|
2013-07-25 16:57:02 +08:00
|
|
|
%res = select i1 %cond, i32 %orig, i32 %alt
|
|
|
|
|
store i32 %res, i32 *%ptr
|
|
|
|
|
ret void
|
|
|
|
|
}
|
|
|
|
|
|
|
|
|
|
; Check the next word down, with the same comments as f8.
|
|
|
|
|
define void @f10(i32 *%base, i32 %alt, i32 %limit) {
|
|
|
|
|
; CHECK-LABEL: f10:
|
|
|
|
|
; CHECK: agfi %r2, -524292
|
|
|
|
|
; CHECK: clfi %r4, 42
|
[SystemZ] Be more careful about inverting CC masks
System z branches have a mask to select which of the 4 CC values should
cause the branch to be taken. We can invert a branch by inverting the mask.
However, not all instructions can produce all 4 CC values, so inverting
the branch like this can lead to some oddities. For example, integer
comparisons only produce a CC of 0 (equal), 1 (less) or 2 (greater).
If an integer EQ is reversed to NE before instruction selection,
the branch will test for 1 or 2. If instead the branch is reversed
after instruction selection (by inverting the mask), it will test for
1, 2 or 3. Both are correct, but the second isn't really canonical.
This patch therefore keeps track of which CC values are possible
and uses this when inverting a mask.
Although this is mostly cosmestic, it fixes undefined behavior
for the CIJNLH in branch-08.ll. Another fix would have been
to mask out bit 0 when generating the fused compare and branch,
but the point of this patch is that we shouldn't need to do that
in the first place.
The patch also makes it easier to reuse CC results from other instructions.
llvm-svn: 187495
2013-07-31 20:30:20 +08:00
|
|
|
; CHECK: stoche %r3, 0(%r2)
|
2013-07-25 16:57:02 +08:00
|
|
|
; CHECK: br %r14
|
[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
|
|
|
%ptr = getelementptr i32, i32 *%base, i64 -131073
|
2013-07-25 16:57:02 +08:00
|
|
|
%cond = icmp ult i32 %limit, 42
|
2018-07-20 20:12:10 +08:00
|
|
|
%orig = load i32, i32 *%ptr
|
2013-07-25 16:57:02 +08:00
|
|
|
%res = select i1 %cond, i32 %orig, i32 %alt
|
|
|
|
|
store i32 %res, i32 *%ptr
|
|
|
|
|
ret void
|
|
|
|
|
}
|
|
|
|
|
|
|
|
|
|
; Try a frame index base.
|
|
|
|
|
define void @f11(i32 %alt, i32 %limit) {
|
|
|
|
|
; CHECK-LABEL: f11:
|
|
|
|
|
; CHECK: brasl %r14, foo@PLT
|
[SystemZ] Be more careful about inverting CC masks
System z branches have a mask to select which of the 4 CC values should
cause the branch to be taken. We can invert a branch by inverting the mask.
However, not all instructions can produce all 4 CC values, so inverting
the branch like this can lead to some oddities. For example, integer
comparisons only produce a CC of 0 (equal), 1 (less) or 2 (greater).
If an integer EQ is reversed to NE before instruction selection,
the branch will test for 1 or 2. If instead the branch is reversed
after instruction selection (by inverting the mask), it will test for
1, 2 or 3. Both are correct, but the second isn't really canonical.
This patch therefore keeps track of which CC values are possible
and uses this when inverting a mask.
Although this is mostly cosmestic, it fixes undefined behavior
for the CIJNLH in branch-08.ll. Another fix would have been
to mask out bit 0 when generating the fused compare and branch,
but the point of this patch is that we shouldn't need to do that
in the first place.
The patch also makes it easier to reuse CC results from other instructions.
llvm-svn: 187495
2013-07-31 20:30:20 +08:00
|
|
|
; CHECK: stoche {{%r[0-9]+}}, {{[0-9]+}}(%r15)
|
2013-07-25 16:57:02 +08:00
|
|
|
; CHECK: brasl %r14, foo@PLT
|
|
|
|
|
; CHECK: br %r14
|
|
|
|
|
%ptr = alloca i32
|
|
|
|
|
call void @foo(i32 *%ptr)
|
|
|
|
|
%cond = icmp ult i32 %limit, 42
|
2018-07-20 20:12:10 +08:00
|
|
|
%orig = load i32, i32 *%ptr
|
2013-07-25 16:57:02 +08:00
|
|
|
%res = select i1 %cond, i32 %orig, i32 %alt
|
|
|
|
|
store i32 %res, i32 *%ptr
|
|
|
|
|
call void @foo(i32 *%ptr)
|
|
|
|
|
ret void
|
|
|
|
|
}
|
|
|
|
|
|
|
|
|
|
; Test that conditionally-executed stores do not use STOC, since STOC
|
|
|
|
|
; is allowed to trap even when the condition is false.
|
|
|
|
|
define void @f12(i32 %a, i32 %b, i32 *%dest) {
|
|
|
|
|
; CHECK-LABEL: f12:
|
|
|
|
|
; CHECK-NOT: stoc
|
|
|
|
|
; CHECK: br %r14
|
|
|
|
|
entry:
|
|
|
|
|
%cmp = icmp ule i32 %a, %b
|
|
|
|
|
br i1 %cmp, label %store, label %exit
|
|
|
|
|
|
|
|
|
|
store:
|
|
|
|
|
store i32 %b, i32 *%dest
|
|
|
|
|
br label %exit
|
|
|
|
|
|
|
|
|
|
exit:
|
|
|
|
|
ret void
|
|
|
|
|
}
|