PPC: Add CTR-register clobber to builtin setjmp

Because the builtin longjmp implementation uses a CTR-based indirect jump, when
the control flow arrives at the builtin setjmp call, the CTR register has
necessarily been clobbered. Correspondingly, this adds CTR to the list of
implicit definitions of the builtin setjmp pseudo instruction.

We don't need to add CTR to the implicit definitions of builtin longjmp
because, even though it does clobber the CTR register, the control flow cannot
return to inside the loop unless there is also a builtin setjmp call.

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@186488 91177308-0d34-0410-b5e6-96231b3b80d8

lib/Target/PowerPC/PPCCTRLoops.cpp

@@ -233,6 +233,13 @@ bool PPCCTRLoops::mightUseCTR(const Triple &TT, BasicBlock *BB) {
 #endif
 
           case Intrinsic::longjmp:
+
+          // Exclude eh_sjlj_setjmp; we don't need to exclude eh_sjlj_longjmp
+          // because, although it does clobber the counter register, the
+          // control can't then return to inside the loop unless there is also
+          // an eh_sjlj_setjmp.
+          case Intrinsic::eh_sjlj_setjmp:
+
           case Intrinsic::memcpy:
           case Intrinsic::memmove:
           case Intrinsic::memset:

lib/Target/PowerPC/PPCInstr64Bit.td

@@ -280,6 +280,7 @@ def MFCR8 : XFXForm_3<31, 19, (outs g8rc:$rT), (ins),
 } // neverHasSideEffects = 1
 
 let hasSideEffects = 1, isBarrier = 1, usesCustomInserter = 1 in {
+  let Defs = [CTR8] in
   def EH_SjLj_SetJmp64  : Pseudo<(outs gprc:$dst), (ins memr:$buf),
                             "#EH_SJLJ_SETJMP64",
                             [(set i32:$dst, (PPCeh_sjlj_setjmp addr:$buf))]>,

lib/Target/PowerPC/PPCInstrInfo.td

@@ -1093,6 +1093,7 @@ def TAILBA   : IForm<18, 0, 0, (outs), (ins abscalltarget:$dst),
 }
 
 let hasSideEffects = 1, isBarrier = 1, usesCustomInserter = 1 in {
+  let Defs = [CTR] in
   def EH_SjLj_SetJmp32  : Pseudo<(outs gprc:$dst), (ins memr:$buf),
                             "#EH_SJLJ_SETJMP32",
                             [(set i32:$dst, (PPCeh_sjlj_setjmp addr:$buf))]>,

test/CodeGen/PowerPC/sj-ctr-loop.ll

@@ -0,0 +1,50 @@
+; RUN: llc -mtriple=powerpc64-unknown-linux-gnu < %s | FileCheck %s
+target datalayout = "E-p:64:64:64-i1:8:8-i8:8:8-i16:16:16-i32:32:32-i64:64:64-f32:32:32-f64:64:64-f128:128:128-v128:128:128-n32:64"
+target triple = "powerpc64-unknown-linux-gnu"
+
+%struct.__jmp_buf_tag.1.15.17.21.25.49.53.55 = type { [64 x i64], i32, %struct.__sigset_t.0.14.16.20.24.48.52.54, [8 x i8] }
+%struct.__sigset_t.0.14.16.20.24.48.52.54 = type { [16 x i64] }
+
+@env_sigill = external global [1 x %struct.__jmp_buf_tag.1.15.17.21.25.49.53.55], align 16
+
+; CHECK-LABEL: @main
+; CHECK-NOT: mtctr
+
+; Function Attrs: nounwind
+define void @main() #0 {
+entry:
+  br i1 undef, label %return, label %if.end
+
+if.end:                                           ; preds = %entry
+  br i1 undef, label %for.body.lr.ph, label %for.end.thread
+
+for.end.thread:                                   ; preds = %if.end
+  br label %return
+
+for.body.lr.ph:                                   ; preds = %if.end
+  br label %for.body
+
+for.cond:                                         ; preds = %for.body
+  %cmp2 = icmp slt i32 %inc, undef
+  br i1 %cmp2, label %for.body, label %for.end
+
+for.body:                                         ; preds = %for.cond, %for.body.lr.ph
+  %i.032 = phi i32 [ 0, %for.body.lr.ph ], [ %inc, %for.cond ]
+  %0 = call i32 @llvm.eh.sjlj.setjmp(i8* bitcast ([1 x %struct.__jmp_buf_tag.1.15.17.21.25.49.53.55]* @env_sigill to i8*))
+  %inc = add nsw i32 %i.032, 1
+  br i1 false, label %if.else, label %for.cond
+
+if.else:                                          ; preds = %for.body
+  unreachable
+
+for.end:                                          ; preds = %for.cond
+  unreachable
+
+return:                                           ; preds = %for.end.thread, %entry
+  ret void
+}
+
+; Function Attrs: nounwind
+declare i32 @llvm.eh.sjlj.setjmp(i8*) #0
+
+attributes #0 = { nounwind }