From abca87c57847cd04f5e05935ff9ae5121c1ecb1d Mon Sep 17 00:00:00 2001
From: Bill Schmidt <wschmidt@linux.ibm.com>
Date: Thu, 27 Aug 2020 11:17:06 -0500
Subject: [PATCH] rs6000: Support ELFv2 sibcall for indirect calls [PR96787]

Prior to P10, ELFv2 hasn't implemented nonlocal sibcalls.  Now that we do,
we need to be sure that r12 is set up prior to such a call.

2020-08-27  Bill Schmidt  <wschmidt@linux.ibm.com>

gcc/
	PR target/96787
	* config/rs6000/rs6000.c (rs6000_sibcall_aix): Support
	indirect call for ELFv2.

gcc/testsuite/

	PR target/96787
	* gcc.target/powerpc/pr96787-1.c: New.
	* gcc.target/powerpc/pr96787-2.c: New.
---
 gcc/config/rs6000/rs6000.c                   | 19 +++++++++++++-
 gcc/testsuite/gcc.target/powerpc/pr96787-1.c | 38 ++++++++++++++++++++++++++++
 gcc/testsuite/gcc.target/powerpc/pr96787-2.c | 35 +++++++++++++++++++++++++
 3 files changed, 91 insertions(+), 1 deletion(-)
 create mode 100644 gcc/testsuite/gcc.target/powerpc/pr96787-1.c
 create mode 100644 gcc/testsuite/gcc.target/powerpc/pr96787-2.c

diff --git a/gcc/config/rs6000/rs6000.c b/gcc/config/rs6000/rs6000.c
index 1c1caa9..0954527 100644
--- a/gcc/config/rs6000/rs6000.c
+++ b/gcc/config/rs6000/rs6000.c
@@ -24833,14 +24833,27 @@ rs6000_sibcall_aix (rtx value, rtx func_desc, rtx tlsarg, rtx cookie)
 {
   rtx call[2];
   rtx insn;
+  rtx r12 = NULL_RTX;
+  rtx func_addr = func_desc;
 
   gcc_assert (INTVAL (cookie) == 0);
 
   if (global_tlsarg)
     tlsarg = global_tlsarg;
 
+  /* For ELFv2, r12 and CTR need to hold the function address
+     for an indirect call.  */
+  if (GET_CODE (func_desc) != SYMBOL_REF && DEFAULT_ABI == ABI_ELFv2)
+    {
+      r12 = gen_rtx_REG (Pmode, 12);
+      if (!rtx_equal_p (r12, func_desc))
+	emit_move_insn (r12, func_desc);
+      func_addr = gen_rtx_REG (Pmode, CTR_REGNO);
+      emit_move_insn (func_addr, r12);
+    }
+
   /* Create the call.  */
-  call[0] = gen_rtx_CALL (VOIDmode, gen_rtx_MEM (SImode, func_desc), tlsarg);
+  call[0] = gen_rtx_CALL (VOIDmode, gen_rtx_MEM (SImode, func_addr), tlsarg);
   if (value != NULL_RTX)
     call[0] = gen_rtx_SET (value, call[0]);
 
@@ -24853,6 +24866,10 @@ rs6000_sibcall_aix (rtx value, rtx func_desc, rtx tlsarg, rtx cookie)
   if (!rs6000_pcrel_p (cfun))
     use_reg (&CALL_INSN_FUNCTION_USAGE (insn),
 	     gen_rtx_REG (Pmode, TOC_REGNUM));
+
+  /* Note use of r12.  */
+  if (r12)
+    use_reg (&CALL_INSN_FUNCTION_USAGE (insn), r12);
 }
 
 /* Expand code to perform a call under the SYSV4 ABI.  */
diff --git a/gcc/testsuite/gcc.target/powerpc/pr96787-1.c b/gcc/testsuite/gcc.target/powerpc/pr96787-1.c
new file mode 100644
index 0000000..3c58e63
--- /dev/null
+++ b/gcc/testsuite/gcc.target/powerpc/pr96787-1.c
@@ -0,0 +1,38 @@
+/* { dg-do compile } */
+/* { dg-require-effective-target powerpc_elfv2 } */
+/* { dg-require-effective-target power10_ok } */
+/* { dg-options "-O2 -mdejagnu-cpu=power10" } */
+
+/* Verify that we generate an indirect sibcall for ELFv2 on P10 and
+   later, with r12 and CTR containing the function address.  PR96787.  */
+
+extern int f (int);
+
+int main ()
+{
+  if (f (3) != 6)
+    return 1;
+  return 0;
+}
+
+
+int g (int a)
+{
+  return a * 2;
+}
+
+
+int h (int a)
+{
+  return a + 2;
+}
+
+int __attribute__((__noinline__)) f (int a)
+{
+  int (*x) (int) = a % 2 ? &g : &h;
+  (*x) (a);
+}
+
+/* { dg-final { scan-assembler {\mmtctr 12\M} } } */
+/* { dg-final { scan-assembler {\mbctr\M} } } */
+/* { dg-final { scan-assembler-not {\mbctrl\M} } } */
diff --git a/gcc/testsuite/gcc.target/powerpc/pr96787-2.c b/gcc/testsuite/gcc.target/powerpc/pr96787-2.c
new file mode 100644
index 0000000..b10ab7a
--- /dev/null
+++ b/gcc/testsuite/gcc.target/powerpc/pr96787-2.c
@@ -0,0 +1,35 @@
+/* { dg-do run } */
+/* { dg-require-effective-target power10_hw } */
+/* { dg-options "-O2 -mdejagnu-cpu=power10" } */
+
+/* Verify that we generate an indirect sibcall for ELFv2 on P10 and
+   later, with r12 and CTR containing the function address.  PR96787.  */
+
+extern void abort (void);
+extern int f (int);
+
+int main ()
+{
+  if (f (3) != 6)
+    abort ();
+  return 0;
+}
+
+
+int g (int a)
+{
+  return a * 2;
+}
+
+
+int h (int a)
+{
+  return a + 2;
+}
+
+int __attribute__((__noinline__)) f (int a)
+{
+  int (*x) (int) = a % 2 ? &g : &h;
+  (*x) (a);
+}
+
-- 
2.7.4