public inbox for gcc-cvs@sourceware.org
help / color / mirror / Atom feed
* [gcc r13-4423] rs6000: Generates permute index directly for little endian targets (PR100866)
@ 2022-12-01  2:07 HaoChen Gui
  0 siblings, 0 replies; only message in thread
From: HaoChen Gui @ 2022-12-01  2:07 UTC (permalink / raw)
  To: gcc-cvs

https://gcc.gnu.org/g:9d68cba5eb20442f8075b8f92d1b20a00022852f

commit r13-4423-g9d68cba5eb20442f8075b8f92d1b20a00022852f
Author: Haochen Gui <guihaoc@gcc.gnu.org>
Date:   Wed Nov 30 15:05:59 2022 +0800

    rs6000: Generates permute index directly for little endian targets (PR100866)
    
    2022-10-11  Haochen Gui <guihaoc@linux.ibm.com>
    
    gcc/
            PR target/100866
            * config/rs6000/rs6000-call.cc (swap_endian_selector_for_mode):
            Generate permute index directly for little endian targets.
            * config/rs6000/vsx.md (revb_<mode>): Call vprem directly with
            corresponding permute indexes.
    
    gcc/testsuite/
            PR target/100866
            * gcc.target/powerpc/pr100866-1.c: New.

Diff:
---
 gcc/config/rs6000/rs6000-call.cc              |  8 +++++++-
 gcc/config/rs6000/vsx.md                      |  4 ++--
 gcc/testsuite/gcc.target/powerpc/pr100866-1.c | 11 +++++++++++
 3 files changed, 20 insertions(+), 3 deletions(-)

diff --git a/gcc/config/rs6000/rs6000-call.cc b/gcc/config/rs6000/rs6000-call.cc
index 6da4de67137..c2a4e4f4e27 100644
--- a/gcc/config/rs6000/rs6000-call.cc
+++ b/gcc/config/rs6000/rs6000-call.cc
@@ -2802,6 +2802,8 @@ rs6000_gimplify_va_arg (tree valist, tree type, gimple_seq *pre_p,
   return build_va_arg_indirect_ref (addr);
 }
 
+/* The selector (perm) is expected to be used with vperm direct as the
+   function generates reversed perm for little endian with this patch.  */
 rtx
 swap_endian_selector_for_mode (machine_mode mode)
 {
@@ -2834,7 +2836,11 @@ swap_endian_selector_for_mode (machine_mode mode)
     }
 
   for (i = 0; i < 16; ++i)
-    perm[i] = GEN_INT (swaparray[i]);
+    if (BYTES_BIG_ENDIAN)
+      perm[i] = GEN_INT (swaparray[i]);
+    else
+      /* Generates the reversed perm for little endian.  */
+      perm[i] = GEN_INT (~swaparray[i] & 0x0000001f);
 
   return force_reg (V16QImode, gen_rtx_CONST_VECTOR (V16QImode,
 						     gen_rtvec_v (16, perm)));
diff --git a/gcc/config/rs6000/vsx.md b/gcc/config/rs6000/vsx.md
index fb5cf04147e..992fbc983be 100644
--- a/gcc/config/rs6000/vsx.md
+++ b/gcc/config/rs6000/vsx.md
@@ -6099,8 +6099,8 @@
 	     to the endian mode in use, i.e. in LE mode, put elements
 	     in BE order.  */
 	  rtx sel = swap_endian_selector_for_mode (<MODE>mode);
-	  emit_insn (gen_altivec_vperm_<mode> (operands[0], operands[1],
-					       operands[1], sel));
+	  emit_insn (gen_altivec_vperm_<mode>_direct (operands[0], operands[1],
+						      operands[1], sel));
 	}
     }
 
diff --git a/gcc/testsuite/gcc.target/powerpc/pr100866-1.c b/gcc/testsuite/gcc.target/powerpc/pr100866-1.c
new file mode 100644
index 00000000000..63872f21bf8
--- /dev/null
+++ b/gcc/testsuite/gcc.target/powerpc/pr100866-1.c
@@ -0,0 +1,11 @@
+/* { dg-do compile } */
+/* { dg-require-effective-target powerpc_p8vector_ok } */
+/* { dg-options "-O2 -mdejagnu-cpu=power8" } */
+/* { dg-final { scan-assembler-not {\mxxlnor\M} } } */
+
+#include <altivec.h>
+
+vector unsigned int revb (vector unsigned int a)
+{
+   return vec_revb(a);
+}

^ permalink raw reply	[flat|nested] only message in thread

only message in thread, other threads:[~2022-12-01  2:07 UTC | newest]

Thread overview: (only message) (download: mbox.gz / follow: Atom feed)
-- links below jump to the message on this page --
2022-12-01  2:07 [gcc r13-4423] rs6000: Generates permute index directly for little endian targets (PR100866) HaoChen Gui

This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for read-only IMAP folder(s) and NNTP newsgroup(s).