[NVPTX] Enable i128 support in the NVPTX backend (#98658)

Summary: The target information needs to configure that the platform has a maximum integer size of 64 in order for it to enable i128 support. The motivation behind this patch is that the i128 libcalls seem to be the only ones used by the NVPTX backend and it would be ideal to disable those completely. That would allow LTO to optimize libcalls properly after llvm/llvm-project#98512.
intel · Jul 12, 2024 · a288d8d · a288d8d
1 parent 280eadd
commit a288d8d
Show file tree

Hide file tree

Showing 5 changed files with 644 additions and 65 deletions.
diff --git a/llvm/lib/Target/NVPTX/NVPTXISelLowering.cpp b/llvm/lib/Target/NVPTX/NVPTXISelLowering.cpp
@@ -872,6 +872,7 @@ NVPTXTargetLowering::NVPTXTargetLowering(const NVPTXTargetMachine &TM,
 
   setMinCmpXchgSizeInBits(32);
   setMaxAtomicSizeInBitsSupported(64);
+  setMaxDivRemBitWidthSupported(64);
 }
 
 const char *NVPTXTargetLowering::getTargetNodeName(unsigned Opcode) const {