[Bug]: Lowering to llvm mlir fails due to hl.add operation on pointers

[kumarak]

VAST version

v0.0.66

LLVM version

19.1.0

Operating system

macos

Description

Lowering to llvm mlir fails with llvm.add operation expecting integer as its operand type

test_ptr.c:4:15: error: unexpected error: 'llvm.add' op operand #0 must be signless integer or LLVM dialect-compatible vector of signless integer, but got '!llvm.ptr'
    return p1 + 1;
              ^
test_ptr.c:4:15: error: unexpected note: see current operation: %4 = "llvm.add"(%2, %3) <{overflowFlags = #llvm.overflow<none>}> : (!llvm.ptr, i32) -> !llvm.ptr
    return p1 + 1;
              ^
// -----// IR Dump After IRsToLLVM Failed (vast-irs-to-llvm) //----- //
"builtin.module"() ({
  "builtin.module"() <{sym_name = "/Users/akshayk/Documents/workspace/demo/vast/build/test_ptr.c"}> ({
    "llvm.func"() <{CConv = #llvm.cconv<ccc>, function_type = !llvm.func<ptr (ptr)>, linkage = #llvm.linkage<external>, sym_name = "ptr_increment", visibility_ = 0 : i64}> ({
    ^bb0(%arg0: !llvm.ptr):
      %0 = "llvm.mlir.constant"() <{value = 1 : i64}> : () -> i64
      %1 = "llvm.alloca"(%0) <{elem_type = !llvm.ptr}> : (i64) -> !llvm.ptr
      "llvm.store"(%arg0, %1) <{ordering = 0 : i64}> : (!llvm.ptr, !llvm.ptr) -> ()
      %2 = "llvm.load"(%1) <{ordering = 0 : i64}> : (!llvm.ptr) -> !llvm.ptr
      %3 = "llvm.mlir.constant"() <{value = 1 : i32}> : () -> i32
      %4 = "llvm.add"(%2, %3) <{overflowFlags = #llvm.overflow<none>}> : (!llvm.ptr, i32) -> !llvm.ptr
      "llvm.return"(%4) : (!llvm.ptr) -> ()
    }) : () -> ()
  }) {dlti.dl_spec = #dlti.dl_spec<#dlti.dl_entry<!hl.ptr<si32>, {vast.abi_align.key = 64 : i32, vast.dl.bw = 64 : i32}>, #dlti.dl_entry<!hl.ptr<ui32>, {vast.abi_align.key = 64 : i32, vast.dl.bw = 64 : i32}>>, vast.core.lang = #core<lang c>, vast.core.target_triple = "arm64-apple-macosx15.0.0"} : () -> ()
}) : () -> ()


[VAST fatal] MLIR pass manager failed when running vast passes

Steps to Reproduce

cat << EOF > test.c
unsigned int *ptr_increment(int *p1) {
    return p1 + 1;
}
EOF

vast-front -vast-emit-mlir=llvm test.c -o test.mlir