[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

1 files changed, 19 insertions, 19 deletions

diff --git a/test/CodeGen/PowerPC/ia-mem-r0.ll b/test/CodeGen/PowerPC/ia-mem-r0.ll
index 4ab17edc5b1..0ce6cc89f1b 100644
--- a/test/CodeGen/PowerPC/ia-mem-r0.ll
+++ b/test/CodeGen/PowerPC/ia-mem-r0.ll
@@ -16,76 +16,76 @@ define void @test1({ i8*, void (i8*, i8*)* } %fn_arg) {
   %1 = bitcast [18 x i64]* %regs to i64*
   call void asm sideeffect "std  14, $0", "=*m"(i64* %1)
   %2 = bitcast [18 x i64]* %regs to i8*
-  %3 = getelementptr i8* %2, i32 8
+  %3 = getelementptr i8, i8* %2, i32 8
   %4 = bitcast i8* %3 to i64*
   call void asm sideeffect "std  15, $0", "=*m"(i64* %4)
   %5 = bitcast [18 x i64]* %regs to i8*
-  %6 = getelementptr i8* %5, i32 16
+  %6 = getelementptr i8, i8* %5, i32 16
   %7 = bitcast i8* %6 to i64*
   call void asm sideeffect "std  16, $0", "=*m"(i64* %7)
   %8 = bitcast [18 x i64]* %regs to i8*
-  %9 = getelementptr i8* %8, i32 24
+  %9 = getelementptr i8, i8* %8, i32 24
   %10 = bitcast i8* %9 to i64*
   call void asm sideeffect "std  17, $0", "=*m"(i64* %10)
   %11 = bitcast [18 x i64]* %regs to i8*
-  %12 = getelementptr i8* %11, i32 32
+  %12 = getelementptr i8, i8* %11, i32 32
   %13 = bitcast i8* %12 to i64*
   call void asm sideeffect "std  18, $0", "=*m"(i64* %13)
   %14 = bitcast [18 x i64]* %regs to i8*
-  %15 = getelementptr i8* %14, i32 40
+  %15 = getelementptr i8, i8* %14, i32 40
   %16 = bitcast i8* %15 to i64*
   call void asm sideeffect "std  19, $0", "=*m"(i64* %16)
   %17 = bitcast [18 x i64]* %regs to i8*
-  %18 = getelementptr i8* %17, i32 48
+  %18 = getelementptr i8, i8* %17, i32 48
   %19 = bitcast i8* %18 to i64*
   call void asm sideeffect "std  20, $0", "=*m"(i64* %19)
   %20 = bitcast [18 x i64]* %regs to i8*
-  %21 = getelementptr i8* %20, i32 56
+  %21 = getelementptr i8, i8* %20, i32 56
   %22 = bitcast i8* %21 to i64*
   call void asm sideeffect "std  21, $0", "=*m"(i64* %22)
   %23 = bitcast [18 x i64]* %regs to i8*
-  %24 = getelementptr i8* %23, i32 64
+  %24 = getelementptr i8, i8* %23, i32 64
   %25 = bitcast i8* %24 to i64*
   call void asm sideeffect "std  22, $0", "=*m"(i64* %25)
   %26 = bitcast [18 x i64]* %regs to i8*
-  %27 = getelementptr i8* %26, i32 72
+  %27 = getelementptr i8, i8* %26, i32 72
   %28 = bitcast i8* %27 to i64*
   call void asm sideeffect "std  23, $0", "=*m"(i64* %28)
   %29 = bitcast [18 x i64]* %regs to i8*
-  %30 = getelementptr i8* %29, i32 80
+  %30 = getelementptr i8, i8* %29, i32 80
   %31 = bitcast i8* %30 to i64*
   call void asm sideeffect "std  24, $0", "=*m"(i64* %31)
   %32 = bitcast [18 x i64]* %regs to i8*
-  %33 = getelementptr i8* %32, i32 88
+  %33 = getelementptr i8, i8* %32, i32 88
   %34 = bitcast i8* %33 to i64*
   call void asm sideeffect "std  25, $0", "=*m"(i64* %34)
   %35 = bitcast [18 x i64]* %regs to i8*
-  %36 = getelementptr i8* %35, i32 96
+  %36 = getelementptr i8, i8* %35, i32 96
   %37 = bitcast i8* %36 to i64*
   call void asm sideeffect "std  26, $0", "=*m"(i64* %37)
   %38 = bitcast [18 x i64]* %regs to i8*
-  %39 = getelementptr i8* %38, i32 104
+  %39 = getelementptr i8, i8* %38, i32 104
   %40 = bitcast i8* %39 to i64*
   call void asm sideeffect "std  27, $0", "=*m"(i64* %40)
   %41 = bitcast [18 x i64]* %regs to i8*
-  %42 = getelementptr i8* %41, i32 112
+  %42 = getelementptr i8, i8* %41, i32 112
   %43 = bitcast i8* %42 to i64*
   call void asm sideeffect "std  28, $0", "=*m"(i64* %43)
   %44 = bitcast [18 x i64]* %regs to i8*
-  %45 = getelementptr i8* %44, i32 120
+  %45 = getelementptr i8, i8* %44, i32 120
   %46 = bitcast i8* %45 to i64*
   call void asm sideeffect "std  29, $0", "=*m"(i64* %46)
   %47 = bitcast [18 x i64]* %regs to i8*
-  %48 = getelementptr i8* %47, i32 128
+  %48 = getelementptr i8, i8* %47, i32 128
   %49 = bitcast i8* %48 to i64*
   call void asm sideeffect "std  30, $0", "=*m"(i64* %49)
   %50 = bitcast [18 x i64]* %regs to i8*
-  %51 = getelementptr i8* %50, i32 136
+  %51 = getelementptr i8, i8* %50, i32 136
   %52 = bitcast i8* %51 to i64*
   call void asm sideeffect "std  31, $0", "=*m"(i64* %52)
-  %53 = getelementptr { i8*, void (i8*, i8*)* }* %fn, i32 0, i32 1
+  %53 = getelementptr { i8*, void (i8*, i8*)* }, { i8*, void (i8*, i8*)* }* %fn, i32 0, i32 1
   %.funcptr = load void (i8*, i8*)** %53
-  %54 = getelementptr { i8*, void (i8*, i8*)* }* %fn, i32 0, i32 0
+  %54 = getelementptr { i8*, void (i8*, i8*)* }, { i8*, void (i8*, i8*)* }* %fn, i32 0, i32 0
   %.ptr = load i8** %54
   %55 = load i8** %sp
   call void %.funcptr(i8* %.ptr, i8* %55)