From mboxrd@z Thu Jan  1 00:00:00 1970
Return-Path: <gcc-bugs-return-480334-listarch-gcc-bugs=gcc.gnu.org@gcc.gnu.org>
Received: (qmail 68908 invoked by alias); 14 Mar 2015 14:31:34 -0000
Mailing-List: contact gcc-bugs-help@gcc.gnu.org; run by ezmlm
Precedence: bulk
List-Id: <gcc-bugs.gcc.gnu.org>
List-Archive: <http://gcc.gnu.org/ml/gcc-bugs/>
List-Post: <mailto:gcc-bugs@gcc.gnu.org>
List-Help: <mailto:gcc-bugs-help@gcc.gnu.org>
Sender: gcc-bugs-owner@gcc.gnu.org
Received: (qmail 68874 invoked by uid 48); 14 Mar 2015 14:31:31 -0000
From: "fuz at fuz dot su" <gcc-bugzilla@gcc.gnu.org>
To: gcc-bugs@gcc.gnu.org
Subject: [Bug tree-optimization/65424] New: gcc does not recognize byte swaps implemented as loop.
Date: Sat, 14 Mar 2015 14:31:00 -0000
X-Bugzilla-Reason: CC
X-Bugzilla-Type: new
X-Bugzilla-Watch-Reason: None
X-Bugzilla-Product: gcc
X-Bugzilla-Component: tree-optimization
X-Bugzilla-Version: 5.0
X-Bugzilla-Keywords:
X-Bugzilla-Severity: normal
X-Bugzilla-Who: fuz at fuz dot su
X-Bugzilla-Status: UNCONFIRMED
X-Bugzilla-Priority: P3
X-Bugzilla-Assigned-To: unassigned at gcc dot gnu.org
X-Bugzilla-Target-Milestone: ---
X-Bugzilla-Flags:
X-Bugzilla-Changed-Fields: bug_id short_desc product version bug_status bug_severity priority component assigned_to reporter attachments.created
Message-ID: <bug-65424-4@http.gcc.gnu.org/bugzilla/>
Content-Type: text/plain; charset="UTF-8"
Content-Transfer-Encoding: 7bit
X-Bugzilla-URL: http://gcc.gnu.org/bugzilla/
Auto-Submitted: auto-generated
MIME-Version: 1.0
X-SW-Source: 2015-03/txt/msg01478.txt.bz2

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=65424

            Bug ID: 65424
           Summary: gcc does not recognize byte swaps implemented as loop.
           Product: gcc
           Version: 5.0
            Status: UNCONFIRMED
          Severity: normal
          Priority: P3
         Component: tree-optimization
          Assignee: unassigned at gcc dot gnu.org
          Reporter: fuz at fuz dot su

Created attachment 35034
  --> https://gcc.gnu.org/bugzilla/attachment.cgi?id=35034&action=edit
endianess-aware reading / writing macros

gcc (r221432) is able to recognize the following code and implements it with
either a bswap or a straight load on x86.

    #include <stdint.h>

    extern uint32_t
    rbe32(const uint8_t a[4])
    {
        uint32_t x = 0;

        x |= a[0] << 24;
        x |= a[1] << 16;
        x |= a[2] <<  8;
        x |= a[3] <<  0;

        return x;
    }

    extern uint32_t
    rle32(const uint8_t a[4])
    {
        uint32_t x = 0;

        x |= a[0] <<  0;
        x |= a[1] <<  8;
        x |= a[2] << 16;
        x |= a[3] << 24;

        return x;
    }

it isn't able to recognize the following two functions which yield identical
results:

    extern uint32_t
    rle32(const uint8_t a[4])
    {
        int i;
        uint32_t x = 0;

        for (i = 0; i < 4; i++)
                x |= a[i] << 8 * i;

        return x;
    }

    extern uint32_t
    rbe32(const uint8_t a[4])
    {
        int i;
        uint32_t x = 0;

        for (i = 0; i < 4; i++)
                x |= a[i] << 8 * (3 - i);

        return x;
    }

It would be great if gcc was able to recognize these implementations of
endianess-aware reads; this kind of code is produced by macros I use to
generate a bunch of read / write functions for differently sized types.
Attached is a file containing the macros I want to use.