AArch64/GISel: fall back to SDAG if a Darwin function needs varargs regs saved.

[t.p.northover]

Summary

A function on Darwin with win64cc needs to save any GPRs that might contain varargs arguments. GlobalISel can't do that yet, so we should fall back.

[paquette]

LGTM

[t.p.northover]

Thanks, it turns out 1c10d5b175992 actually implemented support (including windows) so I was on a stale branch. Sorry about that.

llvm/lib/Target/AArch64/GISel/AArch64CallLowering.cpp

for (auto &KV : BoolArgs) {
           OrigReg, MIRBuilder.buildAssertZExt(WideTy, WideReg, 1).getReg(0));
     }
   }
 
   AArch64FunctionInfo *FuncInfo = MF.getInfo<AArch64FunctionInfo>();
   uint64_t StackOffset = Assigner.StackOffset;
   if (F.isVarArg()) {
     auto &Subtarget = MF.getSubtarget<AArch64Subtarget>();
-    if (!Subtarget.isTargetDarwin()) {
+    if (!Subtarget.isTargetDarwin() || Subtarget.isCallingConvWin64(F.getCallingConv())) {
         // FIXME: we need to reimplement saveVarArgsRegisters from
       // AArch64ISelLowering.
       return false;
     }
 
     // We currently pass all varargs at 8-byte alignment, or 4 in ILP32.
     StackOffset =
         alignTo(Assigner.StackOffset, Subtarget.isTargetILP32() ? 4 : 8);

llvm/test/CodeGen/AArch64/aarch64_win64cc_vararg.ll

 ; NOTE: Assertions have been autogenerated by utils/update_llc_test_checks.py
 ; RUN: llc < %s -mtriple=aarch64-linux-gnu | FileCheck %s
+; RUN: llc < %s -mtriple=arm64-apple-ios -global-isel=1 -global-isel-abort=0 | FileCheck %s --check-prefix=CHECK-DARWIN
 
 define win64cc void @pass_va(i32 %count, ...) nounwind {
+; CHECK-LINUX-LABEL: pass_va:
+; CHECK-LINUX:       // %bb.0: // %entry
+; CHECK-LINUX-NEXT:    sub sp, sp, #96
+; CHECK-LINUX-NEXT:    add x8, sp, #40
+; CHECK-LINUX-NEXT:    add x0, sp, #40
+; CHECK-LINUX-NEXT:    stp x30, x18, [sp, #16] // 16-byte Folded Spill
+; CHECK-LINUX-NEXT:    stp x3, x4, [sp, #56]
+; CHECK-LINUX-NEXT:    stp x1, x2, [sp, #40]
+; CHECK-LINUX-NEXT:    stp x5, x6, [sp, #72]
+; CHECK-LINUX-NEXT:    str x7, [sp, #88]
+; CHECK-LINUX-NEXT:    str x8, [sp, #8]
+; CHECK-LINUX-NEXT:    bl other_func
+; CHECK-LINUX-NEXT:    ldp x30, x18, [sp, #16] // 16-byte Folded Reload
+; CHECK-LINUX-NEXT:    add sp, sp, #96
+; CHECK-LINUX-NEXT:    ret
+;
 ; CHECK-LABEL: pass_va:
 ; CHECK:       // %bb.0: // %entry
 ; CHECK-NEXT:    sub sp, sp, #96
 ; CHECK-NEXT:    add x8, sp, #40
 ; CHECK-NEXT:    add x0, sp, #40
 ; CHECK-NEXT:    stp x30, x18, [sp, #16] // 16-byte Folded Spill
 ; CHECK-NEXT:    stp x3, x4, [sp, #56]
 ; CHECK-NEXT:    stp x1, x2, [sp, #40]
 ; CHECK-NEXT:    stp x5, x6, [sp, #72]
 ; CHECK-NEXT:    str x7, [sp, #88]
 ; CHECK-NEXT:    str x8, [sp, #8]
 ; CHECK-NEXT:    bl other_func
 ; CHECK-NEXT:    ldp x30, x18, [sp, #16] // 16-byte Folded Reload
 ; CHECK-NEXT:    add sp, sp, #96
 ; CHECK-NEXT:    ret
+;
+; CHECK-DARWIN-LABEL: pass_va:
+; CHECK-DARWIN:       ; %bb.0: ; %entry
+; CHECK-DARWIN-NEXT:    str x18, [sp, #-96]! ; 8-byte Folded Spill
+; CHECK-DARWIN-NEXT:    add x8, sp, #40
+; CHECK-DARWIN-NEXT:    add x0, sp, #40
+; CHECK-DARWIN-NEXT:    stp x29, x30, [sp, #16] ; 16-byte Folded Spill
+; CHECK-DARWIN-NEXT:    stp x3, x4, [sp, #56]
+; CHECK-DARWIN-NEXT:    stp x1, x2, [sp, #40]
+; CHECK-DARWIN-NEXT:    stp x5, x6, [sp, #72]
+; CHECK-DARWIN-NEXT:    str x7, [sp, #88]
+; CHECK-DARWIN-NEXT:    str x8, [sp, #8]
+; CHECK-DARWIN-NEXT:    bl _other_func
+; CHECK-DARWIN-NEXT:    ldp x29, x30, [sp, #16] ; 16-byte Folded Reload
+; CHECK-DARWIN-NEXT:    ldr x18, [sp], #96 ; 8-byte Folded Reload
+; CHECK-DARWIN-NEXT:    ret
 entry:
   %ap = alloca ptr, align 8
   call void @llvm.va_start(ptr %ap)
   %ap2 = load ptr, ptr %ap, align 8
   call void @other_func(ptr %ap2)
   ret void
 }
 
 declare void @other_func(ptr) local_unnamed_addr
 
 declare void @llvm.va_start(ptr) nounwind
 declare void @llvm.va_copy(ptr, ptr) nounwind
 
 define win64cc ptr @f9(i64 %a0, i64 %a1, i64 %a2, i64 %a3, i64 %a4, i64 %a5, i64 %a6, i64 %a7, i64 %a8, ...) nounwind {
+; CHECK-LINUX-LABEL: f9:
+; CHECK-LINUX:       // %bb.0: // %entry
+; CHECK-LINUX-NEXT:    str x18, [sp, #-16]! // 8-byte Folded Spill
+; CHECK-LINUX-NEXT:    add x8, sp, #24
+; CHECK-LINUX-NEXT:    add x0, sp, #24
+; CHECK-LINUX-NEXT:    str x8, [sp, #8]
+; CHECK-LINUX-NEXT:    ldr x18, [sp], #16 // 8-byte Folded Reload
+; CHECK-LINUX-NEXT:    ret
+;
 ; CHECK-LABEL: f9:
 ; CHECK:       // %bb.0: // %entry
 ; CHECK-NEXT:    str x18, [sp, #-16]! // 8-byte Folded Spill
 ; CHECK-NEXT:    add x8, sp, #24
 ; CHECK-NEXT:    add x0, sp, #24
 ; CHECK-NEXT:    str x8, [sp, #8]
 ; CHECK-NEXT:    ldr x18, [sp], #16 // 8-byte Folded Reload
 ; CHECK-NEXT:    ret
+;
+; CHECK-DARWIN-LABEL: f9:
+; CHECK-DARWIN:       ; %bb.0: ; %entry
+; CHECK-DARWIN-NEXT:    str x18, [sp, #-16]! ; 8-byte Folded Spill
+; CHECK-DARWIN-NEXT:    add x8, sp, #24
+; CHECK-DARWIN-NEXT:    add x0, sp, #24
+; CHECK-DARWIN-NEXT:    str x8, [sp, #8]
+; CHECK-DARWIN-NEXT:    ldr x18, [sp], #16 ; 8-byte Folded Reload
+; CHECK-DARWIN-NEXT:    ret
 entry:
   %ap = alloca ptr, align 8
   call void @llvm.va_start(ptr %ap)
   %ap2 = load ptr, ptr %ap, align 8
   ret ptr %ap2
 }
 
 define win64cc ptr @f8(i64 %a0, i64 %a1, i64 %a2, i64 %a3, i64 %a4, i64 %a5, i64 %a6, i64 %a7, ...) nounwind {
+; CHECK-LINUX-LABEL: f8:
+; CHECK-LINUX:       // %bb.0: // %entry
+; CHECK-LINUX-NEXT:    str x18, [sp, #-16]! // 8-byte Folded Spill
+; CHECK-LINUX-NEXT:    add x8, sp, #16
+; CHECK-LINUX-NEXT:    add x0, sp, #16
+; CHECK-LINUX-NEXT:    str x8, [sp, #8]
+; CHECK-LINUX-NEXT:    ldr x18, [sp], #16 // 8-byte Folded Reload
+; CHECK-LINUX-NEXT:    ret
+;
 ; CHECK-LABEL: f8:
 ; CHECK:       // %bb.0: // %entry
 ; CHECK-NEXT:    str x18, [sp, #-16]! // 8-byte Folded Spill
 ; CHECK-NEXT:    add x8, sp, #16
 ; CHECK-NEXT:    add x0, sp, #16
 ; CHECK-NEXT:    str x8, [sp, #8]
 ; CHECK-NEXT:    ldr x18, [sp], #16 // 8-byte Folded Reload
 ; CHECK-NEXT:    ret
+;
+; CHECK-DARWIN-LABEL: f8:
+; CHECK-DARWIN:       ; %bb.0: ; %entry
+; CHECK-DARWIN-NEXT:    str x18, [sp, #-16]! ; 8-byte Folded Spill
+; CHECK-DARWIN-NEXT:    add x8, sp, #16
+; CHECK-DARWIN-NEXT:    add x0, sp, #16
+; CHECK-DARWIN-NEXT:    str x8, [sp, #8]
+; CHECK-DARWIN-NEXT:    ldr x18, [sp], #16 ; 8-byte Folded Reload
+; CHECK-DARWIN-NEXT:    ret
 entry:
   %ap = alloca ptr, align 8
   call void @llvm.va_start(ptr %ap)
   %ap2 = load ptr, ptr %ap, align 8
   ret ptr %ap2
 }
 
 define win64cc ptr @f7(i64 %a0, i64 %a1, i64 %a2, i64 %a3, i64 %a4, i64 %a5, i64 %a6, ...) nounwind {
+; CHECK-LINUX-LABEL: f7:
+; CHECK-LINUX:       // %bb.0: // %entry
+; CHECK-LINUX-NEXT:    str x18, [sp, #-32]! // 8-byte Folded Spill
+; CHECK-LINUX-NEXT:    add x8, sp, #24
+; CHECK-LINUX-NEXT:    add x0, sp, #24
+; CHECK-LINUX-NEXT:    str x7, [sp, #24]
+; CHECK-LINUX-NEXT:    str x8, [sp, #8]
+; CHECK-LINUX-NEXT:    ldr x18, [sp], #32 // 8-byte Folded Reload
+; CHECK-LINUX-NEXT:    ret
+;
 ; CHECK-LABEL: f7:
 ; CHECK:       // %bb.0: // %entry
 ; CHECK-NEXT:    str x18, [sp, #-32]! // 8-byte Folded Spill
 ; CHECK-NEXT:    add x8, sp, #24
 ; CHECK-NEXT:    add x0, sp, #24
 ; CHECK-NEXT:    str x7, [sp, #24]
 ; CHECK-NEXT:    str x8, [sp, #8]
 ; CHECK-NEXT:    ldr x18, [sp], #32 // 8-byte Folded Reload
 ; CHECK-NEXT:    ret
+;
+; CHECK-DARWIN-LABEL: f7:
+; CHECK-DARWIN:       ; %bb.0: ; %entry
+; CHECK-DARWIN-NEXT:    str x18, [sp, #-32]! ; 8-byte Folded Spill
+; CHECK-DARWIN-NEXT:    add x8, sp, #24
+; CHECK-DARWIN-NEXT:    add x0, sp, #24
+; CHECK-DARWIN-NEXT:    str x7, [sp, #24]
+; CHECK-DARWIN-NEXT:    str x8, [sp, #8]
+; CHECK-DARWIN-NEXT:    ldr x18, [sp], #32 ; 8-byte Folded Reload
+; CHECK-DARWIN-NEXT:    ret
 entry:
   %ap = alloca ptr, align 8
   call void @llvm.va_start(ptr %ap)
   %ap2 = load ptr, ptr %ap, align 8
   ret ptr %ap2
 }