llvm-6502/test/CodeGen/SystemZ/cmpxchg-03.ll
David Blaikie 198d8baafb [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

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@230786 91177308-0d34-0410-b5e6-96231b3b80d8
2015-02-27 19:29:02 +00:00

144 lines
4.3 KiB
LLVM

; Test 32-bit compare and swap.
;
; RUN: llc < %s -mtriple=s390x-linux-gnu | FileCheck %s
; Check the low end of the CS range.
define i32 @f1(i32 %cmp, i32 %swap, i32 *%src) {
; CHECK-LABEL: f1:
; CHECK: cs %r2, %r3, 0(%r4)
; CHECK: br %r14
%pair = cmpxchg i32 *%src, i32 %cmp, i32 %swap seq_cst seq_cst
%val = extractvalue { i32, i1 } %pair, 0
ret i32 %val
}
; Check the high end of the aligned CS range.
define i32 @f2(i32 %cmp, i32 %swap, i32 *%src) {
; CHECK-LABEL: f2:
; CHECK: cs %r2, %r3, 4092(%r4)
; CHECK: br %r14
%ptr = getelementptr i32, i32 *%src, i64 1023
%pair = cmpxchg i32 *%ptr, i32 %cmp, i32 %swap seq_cst seq_cst
%val = extractvalue { i32, i1 } %pair, 0
ret i32 %val
}
; Check the next word up, which should use CSY instead of CS.
define i32 @f3(i32 %cmp, i32 %swap, i32 *%src) {
; CHECK-LABEL: f3:
; CHECK: csy %r2, %r3, 4096(%r4)
; CHECK: br %r14
%ptr = getelementptr i32, i32 *%src, i64 1024
%pair = cmpxchg i32 *%ptr, i32 %cmp, i32 %swap seq_cst seq_cst
%val = extractvalue { i32, i1 } %pair, 0
ret i32 %val
}
; Check the high end of the aligned CSY range.
define i32 @f4(i32 %cmp, i32 %swap, i32 *%src) {
; CHECK-LABEL: f4:
; CHECK: csy %r2, %r3, 524284(%r4)
; CHECK: br %r14
%ptr = getelementptr i32, i32 *%src, i64 131071
%pair = cmpxchg i32 *%ptr, i32 %cmp, i32 %swap seq_cst seq_cst
%val = extractvalue { i32, i1 } %pair, 0
ret i32 %val
}
; Check the next word up, which needs separate address logic.
; Other sequences besides this one would be OK.
define i32 @f5(i32 %cmp, i32 %swap, i32 *%src) {
; CHECK-LABEL: f5:
; CHECK: agfi %r4, 524288
; CHECK: cs %r2, %r3, 0(%r4)
; CHECK: br %r14
%ptr = getelementptr i32, i32 *%src, i64 131072
%pair = cmpxchg i32 *%ptr, i32 %cmp, i32 %swap seq_cst seq_cst
%val = extractvalue { i32, i1 } %pair, 0
ret i32 %val
}
; Check the high end of the negative aligned CSY range.
define i32 @f6(i32 %cmp, i32 %swap, i32 *%src) {
; CHECK-LABEL: f6:
; CHECK: csy %r2, %r3, -4(%r4)
; CHECK: br %r14
%ptr = getelementptr i32, i32 *%src, i64 -1
%pair = cmpxchg i32 *%ptr, i32 %cmp, i32 %swap seq_cst seq_cst
%val = extractvalue { i32, i1 } %pair, 0
ret i32 %val
}
; Check the low end of the CSY range.
define i32 @f7(i32 %cmp, i32 %swap, i32 *%src) {
; CHECK-LABEL: f7:
; CHECK: csy %r2, %r3, -524288(%r4)
; CHECK: br %r14
%ptr = getelementptr i32, i32 *%src, i64 -131072
%pair = cmpxchg i32 *%ptr, i32 %cmp, i32 %swap seq_cst seq_cst
%val = extractvalue { i32, i1 } %pair, 0
ret i32 %val
}
; Check the next word down, which needs separate address logic.
; Other sequences besides this one would be OK.
define i32 @f8(i32 %cmp, i32 %swap, i32 *%src) {
; CHECK-LABEL: f8:
; CHECK: agfi %r4, -524292
; CHECK: cs %r2, %r3, 0(%r4)
; CHECK: br %r14
%ptr = getelementptr i32, i32 *%src, i64 -131073
%pair = cmpxchg i32 *%ptr, i32 %cmp, i32 %swap seq_cst seq_cst
%val = extractvalue { i32, i1 } %pair, 0
ret i32 %val
}
; Check that CS does not allow an index.
define i32 @f9(i32 %cmp, i32 %swap, i64 %src, i64 %index) {
; CHECK-LABEL: f9:
; CHECK: agr %r4, %r5
; CHECK: cs %r2, %r3, 0(%r4)
; CHECK: br %r14
%add1 = add i64 %src, %index
%ptr = inttoptr i64 %add1 to i32 *
%pair = cmpxchg i32 *%ptr, i32 %cmp, i32 %swap seq_cst seq_cst
%val = extractvalue { i32, i1 } %pair, 0
ret i32 %val
}
; Check that CSY does not allow an index.
define i32 @f10(i32 %cmp, i32 %swap, i64 %src, i64 %index) {
; CHECK-LABEL: f10:
; CHECK: agr %r4, %r5
; CHECK: csy %r2, %r3, 4096(%r4)
; CHECK: br %r14
%add1 = add i64 %src, %index
%add2 = add i64 %add1, 4096
%ptr = inttoptr i64 %add2 to i32 *
%pair = cmpxchg i32 *%ptr, i32 %cmp, i32 %swap seq_cst seq_cst
%val = extractvalue { i32, i1 } %pair, 0
ret i32 %val
}
; Check that a constant %cmp value is loaded into a register first.
define i32 @f11(i32 %dummy, i32 %swap, i32 *%ptr) {
; CHECK-LABEL: f11:
; CHECK: lhi %r2, 1001
; CHECK: cs %r2, %r3, 0(%r4)
; CHECK: br %r14
%pair = cmpxchg i32 *%ptr, i32 1001, i32 %swap seq_cst seq_cst
%val = extractvalue { i32, i1 } %pair, 0
ret i32 %val
}
; Check that a constant %swap value is loaded into a register first.
define i32 @f12(i32 %cmp, i32 *%ptr) {
; CHECK-LABEL: f12:
; CHECK: lhi [[SWAP:%r[0-9]+]], 1002
; CHECK: cs %r2, [[SWAP]], 0(%r3)
; CHECK: br %r14
%pair = cmpxchg i32 *%ptr, i32 %cmp, i32 1002 seq_cst seq_cst
%val = extractvalue { i32, i1 } %pair, 0
ret i32 %val
}