From: Peter Bergner <bergner@linux.ibm.com>
Date: Tue, 16 Nov 2021 18:14:22 +0000 (-0600)
Subject: rs6000: MMA test case emits wrong code when building a vector pair [PR102976]
X-Git-Tag: releases/gcc-10.4.0~472
X-Git-Url: http://git.ipfire.org/cgi-bin/gitweb.cgi?a=commitdiff_plain;h=7729d2c96d5eac9537c78d368bbc037bea13f988;p=thirdparty%2Fgcc.git

rs6000: MMA test case emits wrong code when building a vector pair [PR102976]

PR102976 shows a test case where we generate wrong code when building
a vector pair from 2 vector registers.  The bug here is that with unlucky
register assignments, we can clobber one of the input operands before
we write both registers of the output operand.  The solution is to use
early-clobbers in the assemble pair and accumulator patterns.

2021-11-16  Peter Bergner  <bergner@linux.ibm.com>

gcc/
	PR target/102976
	* config/rs6000/mma.md (*vsx_assemble_pair): Add early-clobber for
	output operand.
	(*mma_assemble_acc): Likewise.

gcc/testsuite/
	PR target/102976
	* gcc.target/powerpc/pr102976.c: New test.

(cherry picked from commit 4cdf7db9a39d18bd536d816a5751d4d3cf23808b)
---

diff --git a/gcc/config/rs6000/mma.md b/gcc/config/rs6000/mma.md
index c267a4c82e26..27f64d2cb7c3 100644
--- a/gcc/config/rs6000/mma.md
+++ b/gcc/config/rs6000/mma.md
@@ -362,8 +362,11 @@
   DONE;
 })
 
+;; We cannot update the two output registers atomically, so mark the output
+;; as an early clobber so we don't accidentally clobber the input operands.  */
+
 (define_insn_and_split "*vsx_assemble_pair"
-  [(set (match_operand:POI 0 "vsx_register_operand" "=wa")
+  [(set (match_operand:POI 0 "vsx_register_operand" "=&wa")
         (unspec:POI [(match_operand:V16QI 1 "mma_assemble_input_operand" "mwa")
 		     (match_operand:V16QI 2 "mma_assemble_input_operand" "mwa")]
                      UNSPEC_VSX_ASSEMBLE))]
@@ -395,8 +398,11 @@
   DONE;
 })
 
+;; We cannot update the four output registers atomically, so mark the output
+;; as an early clobber so we don't accidentally clobber the input operands.  */
+
 (define_insn_and_split "*mma_assemble_acc"
-  [(set (match_operand:PXI 0 "fpr_reg_operand" "=d")
+  [(set (match_operand:PXI 0 "fpr_reg_operand" "=&d")
 	(unspec:PXI [(match_operand:V16QI 1 "mma_assemble_input_operand" "mwa")
 		     (match_operand:V16QI 2 "mma_assemble_input_operand" "mwa")
 		     (match_operand:V16QI 3 "mma_assemble_input_operand" "mwa")
diff --git a/gcc/testsuite/gcc.target/powerpc/pr102976.c b/gcc/testsuite/gcc.target/powerpc/pr102976.c
new file mode 100644
index 000000000000..5a4320f8e0a3
--- /dev/null
+++ b/gcc/testsuite/gcc.target/powerpc/pr102976.c
@@ -0,0 +1,14 @@
+/* { dg-require-effective-target power10_ok } */
+/* { dg-options "-O2 -mdejagnu-cpu=power10" } */
+
+#include <altivec.h>
+void
+bug (__vector_pair *dst)
+{
+  register vector unsigned char vec0 asm ("vs44");
+  register vector unsigned char vec1 asm ("vs32");
+  __builtin_vsx_build_pair (dst, vec0, vec1);
+}
+
+/* { dg-final { scan-assembler-times {(?p)\mxxlor \d+,44,44\M} 1 } } */
+/* { dg-final { scan-assembler-times {(?p)\mxxlor \d+,32,32\M} 1 } } */