================ @@ -1056,6 +1056,23 @@ def CC_Intel_OCL_BI : CallingConv<[ CCDelegateTo<CC_X86_32_C> ]>; +def CC_X86_64_Preserve_None : CallingConv<[ + // We don't preserve general registers, so all of them can be used to pass + // arguments except + // - RBP frame pointer + // - R10 'nest' parameter + // - RBX base pointer + // - R16 - R31 these are not available everywhere + CCIfType<[i32], CCAssignToReg<[EDI, ESI, EDX, ECX, R8D, R9D, + R11D, R12D, R13D, R14D, R15D, EAX]>>, + + CCIfType<[i64], CCAssignToReg<[RDI, RSI, RDX, RCX, R8, R9, + R11, R12, R13, R14, R15, RAX]>>, + + // Otherwise it's the same as the regular C calling convention. + CCDelegateTo<CC_X86_64_C> ---------------- weiguozhi wrote:
@jyknight could you help to review the code that checks the combination of swift attributes and preserve_none in https://github.com/llvm/llvm-project/pull/76868/commits/dc1bc55634045adb5b0ab465b16dcf73f1e77144 https://github.com/llvm/llvm-project/pull/76868 _______________________________________________ cfe-commits mailing list cfe-commits@lists.llvm.org https://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits