public inbox for gcc-patches@gcc.gnu.org
 help / color / mirror / Atom feed
From: Kai Tietz <ktietz@redhat.com>
To: gcc-patches@gcc.gnu.org
Subject: [patch tree-ssa-math-opts]: Improve bswap detection
Date: Mon, 27 Jun 2011 08:50:00 -0000	[thread overview]
Message-ID: <400461904.779176.1309161973367.JavaMail.root@zmail06.collab.prod.int.phx2.redhat.com> (raw)
In-Reply-To: <1518076949.779173.1309161963786.JavaMail.root@zmail06.collab.prod.int.phx2.redhat.com>

[-- Attachment #1: Type: text/plain, Size: 578 bytes --]

Hello,

this patch improves the bswap (32/64) detection and fixes an issue about shifted bit values out of type-size precision.

ChangeLog

2011-06-27  Kai Tietz  <ktietz@redhat.com>

        * tree-ssa-math-opts.c (do_shift_rotate): Zero bits
        out of type precision after operation.
        (find_bswap): Take for limit value the integer auto-
        promotion into account.

ChangeLog

2011-06-27  Kai Tietz  <ktietz@redhat.com>

        * gcc.dg/optimize-bswapdi-2.c: New test.

Bootstrapped and regression tested for x86_64-pc-linux-gnu. Ok for apply?

Regards,
Kai

[-- Attachment #2: bswap_fix.txt --]
[-- Type: text/plain, Size: 2344 bytes --]

Index: gcc-head/gcc/testsuite/gcc.dg/optimize-bswapdi-2.c
===================================================================
--- /dev/null
+++ gcc-head/gcc/testsuite/gcc.dg/optimize-bswapdi-2.c
@@ -0,0 +1,26 @@
+/* { dg-do compile { target arm*-*-* alpha*-*-* ia64*-*-* x86_64-*-* s390x-*-* powerpc*-*-* rs6000-*-* } } */
+/* { dg-require-effective-target stdint_types } */
+/* { dg-require-effective-target lp64 } */
+/* { dg-options "-O2 -fdump-tree-bswap" } */
+
+#include <stdint.h>
+
+/* A variant via unsigned short.  */
+
+uint64_t
+swap64_c (uint64_t x)
+{
+  uint16_t a0 = x >> 48;
+  uint16_t a1 = x >> 32;
+  uint16_t a2 = x >> 16;
+  uint16_t a3 = x;
+
+  return ((uint64_t) (((a0 >> 8) & 0xff) | ((a0 << 8) & 0xff00)))
+	| ((uint64_t) (((a1 >> 8) & 0xff) | ((a1 << 8) & 0xff00)) << 16)
+	| ((uint64_t) (((a2 >> 8) & 0xff) | ((a2 << 8) & 0xff00)) << 32)
+	| ((uint64_t) (((a3 >> 8) & 0xff) | ((a3 << 8) & 0xff00)) << 48);
+}
+
+
+/* { dg-final { scan-tree-dump-times "64 bit bswap implementation found at" 1 "bswap" } } */
+/* { dg-final { cleanup-tree-dump "bswap" } } */
Index: gcc-head/gcc/tree-ssa-math-opts.c
===================================================================
--- gcc-head.orig/gcc/tree-ssa-math-opts.c
+++ gcc-head/gcc/tree-ssa-math-opts.c
@@ -1543,6 +1543,9 @@ do_shift_rotate (enum tree_code code,
     default:
       return false;
     }
+  /* Zero unused bits for size.  */
+  if (n->size < (int)sizeof (HOST_WIDEST_INT))
+    n->n &= ((unsigned HOST_WIDEST_INT)1 << (n->size * BITS_PER_UNIT)) - 1;
   return true;
 }
 
@@ -1743,12 +1746,12 @@ find_bswap (gimple stmt)
 
   /* The last parameter determines the depth search limit.  It usually
      correlates directly to the number of bytes to be touched.  We
-     increase that number by one here in order to also cover signed ->
-     unsigned conversions of the src operand as can be seen in
-     libgcc.  */
+     increase that number by three  here in order to also
+     cover signed -> unsigned converions of the src operand as can be seen
+     in libgcc, and for initial shift/and operation of the src operand.  */
   source_expr =  find_bswap_1 (stmt, &n,
 			       TREE_INT_CST_LOW (
-				 TYPE_SIZE_UNIT (gimple_expr_type (stmt))) + 1);
+				 TYPE_SIZE_UNIT (gimple_expr_type (stmt))) + 3);
 
   if (!source_expr)
     return NULL_TREE;

       reply	other threads:[~2011-06-27  8:06 UTC|newest]

Thread overview: 4+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
     [not found] <1518076949.779173.1309161963786.JavaMail.root@zmail06.collab.prod.int.phx2.redhat.com>
2011-06-27  8:50 ` Kai Tietz [this message]
2011-06-27 10:08   ` Richard Guenther
2011-06-27 10:13     ` Kai Tietz
2011-06-27 10:16       ` Richard Guenther

Reply instructions:

You may reply publicly to this message via plain-text email
using any one of the following methods:

* Save the following mbox file, import it into your mail client,
  and reply-to-all from there: mbox

  Avoid top-posting and favor interleaved quoting:
  https://en.wikipedia.org/wiki/Posting_style#Interleaved_style

* Reply using the --to, --cc, and --in-reply-to
  switches of git-send-email(1):

  git send-email \
    --in-reply-to=400461904.779176.1309161973367.JavaMail.root@zmail06.collab.prod.int.phx2.redhat.com \
    --to=ktietz@redhat.com \
    --cc=gcc-patches@gcc.gnu.org \
    /path/to/YOUR_REPLY

  https://kernel.org/pub/software/scm/git/docs/git-send-email.html

* If your mail client supports setting the In-Reply-To header
  via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line before the message body.
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).