[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

1 files changed, 6 insertions, 6 deletions

diff --git a/llvm/test/Transforms/InstCombine/objsize.ll b/llvm/test/Transforms/InstCombine/objsize.ll
index 1285b1c3aa6..38dd33d3e83 100644
--- a/llvm/test/Transforms/InstCombine/objsize.ll
+++ b/llvm/test/Transforms/InstCombine/objsize.ll
@@ -86,7 +86,7 @@ entry:
   br i1 undef, label %bb11, label %bb12
 
 bb11:
-  %0 = getelementptr inbounds float* getelementptr inbounds ([480 x float]* @array, i32 0, i32 128), i32 -127 ; <float*> [#uses=1]
+  %0 = getelementptr inbounds float, float* getelementptr inbounds ([480 x float]* @array, i32 0, i32 128), i32 -127 ; <float*> [#uses=1]
   %1 = bitcast float* %0 to i8*                   ; <i8*> [#uses=1]
   %2 = call i32 @llvm.objectsize.i32.p0i8(i8* %1, i1 false) ; <i32> [#uses=1]
   %3 = call i8* @__memcpy_chk(i8* undef, i8* undef, i32 512, i32 %2) nounwind ; <i8*> [#uses=0]
@@ -94,7 +94,7 @@ bb11:
   unreachable
 
 bb12:
-  %4 = getelementptr inbounds float* getelementptr inbounds ([480 x float]* @array, i32 0, i32 128), i32 -127 ; <float*> [#uses=1]
+  %4 = getelementptr inbounds float, float* getelementptr inbounds ([480 x float]* @array, i32 0, i32 128), i32 -127 ; <float*> [#uses=1]
   %5 = bitcast float* %4 to i8*                   ; <i8*> [#uses=1]
   %6 = call i8* @__inline_memcpy_chk(i8* %5, i8* undef, i32 512) nounwind inlinehint ; <i8*> [#uses=0]
 ; CHECK: @__inline_memcpy_chk
@@ -153,7 +153,7 @@ define i32 @test7(i8** %esc) {
 ; CHECK-LABEL: @test7(
   %alloc = call noalias i8* @malloc(i32 48) nounwind
   store i8* %alloc, i8** %esc
-  %gep = getelementptr inbounds i8* %alloc, i32 16
+  %gep = getelementptr inbounds i8, i8* %alloc, i32 16
   %objsize = call i32 @llvm.objectsize.i32.p0i8(i8* %gep, i1 false) nounwind readonly
 ; CHECK: ret i32 32
   ret i32 %objsize
@@ -165,7 +165,7 @@ define i32 @test8(i8** %esc) {
 ; CHECK-LABEL: @test8(
   %alloc = call noalias i8* @calloc(i32 5, i32 7) nounwind
   store i8* %alloc, i8** %esc
-  %gep = getelementptr inbounds i8* %alloc, i32 5
+  %gep = getelementptr inbounds i8, i8* %alloc, i32 5
   %objsize = call i32 @llvm.objectsize.i32.p0i8(i8* %gep, i1 false) nounwind readonly
 ; CHECK: ret i32 30
   ret i32 %objsize
@@ -246,8 +246,8 @@ entry:
 
 ; technically reachable, but this malformed IR may appear as a result of constant propagation
 xpto:
-  %gep2 = getelementptr i8* %gep, i32 1
-  %gep = getelementptr i8* %gep2, i32 1
+  %gep2 = getelementptr i8, i8* %gep, i32 1
+  %gep = getelementptr i8, i8* %gep2, i32 1
   %o = call i32 @llvm.objectsize.i32.p0i8(i8* %gep, i1 true)
 ; CHECK: ret i32 undef
   ret i32 %o