public inbox for gcc-bugs@sourceware.org
help / color / mirror / Atom feed
From: "gcc-bugzilla at gcc dot gnu dot org" <gcc-bugzilla@gcc.gnu.org>
To: gcc-bugs@gcc.gnu.org
Subject: [Bug c/16721] New: Accesses to volatile objects optimized away with -fno-strict-aliasing
Date: Mon, 26 Jul 2004 18:12:00 -0000	[thread overview]
Message-ID: <20040726181156.16721.macro@linux-mips.org> (raw)


 Certain read accesses to objects marked as "volatile" are optimized away.  
This happens when both "-O2" and "-fno-strict-aliasing" are specified.  
This has been observed with the Linux kernel which is unusable in certain
configurations due to this problem.  This used to work with 3.4.0.

Environment:
System: Linux macro 2.4.26 #3 Thu May 6 20:57:18 CEST 2004 i686 unknown unknown GNU/Linux
Architecture: i686

host: i686-pc-linux-gnu
build: i686-pc-linux-gnu
target: mips64el-unknown-linux-gnu
configured with: ../configure --prefix=/usr --enable-languages=c --disable-libmudflap --disable-shared --target=mips64el-linux

How-To-Repeat:
 The following program can be used:

$ cat volatile.c
struct data {
	volatile unsigned long *addr;
} *p;

int test()
{
	*p->addr;
	return 0;
}
$ mips64el-linux-gcc -mabi=32 -O2 -fno-strict-aliasing -mno-abicalls -S
volatile.c -o volatile.s
$ cat volatile.s
	.file	1 "volatile.c"
	.section .mdebug.abi32
	.previous
	.text
	.align	2
	.align	2
	.globl	test
	.ent	test
	.type	test, @function
test:
	.frame	$sp,0,$31		# vars= 0, regs= 0/0, args= 0, gp= 0
	.mask	0x00000000,0
	.fmask	0x00000000,0
	.set	noreorder
	.set	nomacro
	
	j	$31
	move	$2,$0

	.set	macro
	.set	reorder
	.end	test

	.comm	p,4,4
	.ident	"GCC: (GNU) 3.5.0 20040726 (experimental)"

After removing the "-fno-strict-aliasing" option, the output is correct:

$ mips64el-linux-gcc -mabi=32 -O2 -mno-abicalls -S volatile.c -o 
volatile.s
$ cat volatile.s
	.file	1 "volatile.c"
	.section .mdebug.abi32
	.previous
	.text
	.align	2
	.align	2
	.globl	test
	.ent	test
	.type	test, @function
test:
	.frame	$sp,0,$31		# vars= 0, regs= 0/0, args= 0, gp= 0
	.mask	0x00000000,0
	.fmask	0x00000000,0
	.set	noreorder
	.set	nomacro
	
	lui	$2,%hi(p)
	lw	$4,%lo(p)($2)
	move	$2,$0
	lw	$3,0($4)
	nop
	lw	$5,0($3)
	j	$31
	nop

	.set	macro
	.set	reorder
	.end	test

	.comm	p,4,4
	.ident	"GCC: (GNU) 3.5.0 20040726 (experimental)"
------- Additional Comments From macro at linux-mips dot org  2004-07-26 18:11 -------
Fix:
 None known.  Using "-fstrict-aliasing" may be a workaround, but it is not 
an option for Linux.

-- 
           Summary: Accesses to volatile objects optimized away with -fno-
                    strict-aliasing
           Product: gcc
           Version: 3.5.0
            Status: UNCONFIRMED
          Severity: normal
          Priority: P3
         Component: c
        AssignedTo: unassigned at gcc dot gnu dot org
        ReportedBy: macro at linux-mips dot org
                CC: gcc-bugs at gcc dot gnu dot org
 GCC build triplet: i686-pc-linux-gnu
  GCC host triplet: i686-pc-linux-gnu
GCC target triplet: mips64el-unknown-linux-gnu


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=16721


             reply	other threads:[~2004-07-26 18:12 UTC|newest]

Thread overview: 20+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2004-07-26 18:12 gcc-bugzilla at gcc dot gnu dot org [this message]
2004-07-26 19:04 ` [Bug c/16721] [3.5 Regression] Accesses to volatile objects optimized away falk at debian dot org
2004-07-26 19:24 ` pinskia at gcc dot gnu dot org
2004-07-26 19:30 ` falk at debian dot org
2004-07-27 12:25 ` macro at linux-mips dot org
2004-07-27 12:50 ` macro at linux-mips dot org
2004-07-27 14:48 ` falk at debian dot org
2004-07-29  2:33 ` [Bug tree-optimization/16721] " pinskia at gcc dot gnu dot org
2004-07-29 11:27 ` macro at linux-mips dot org
2004-08-03  5:58 ` pinskia at gcc dot gnu dot org
2004-09-19 13:57 ` [Bug tree-optimization/16721] [4.0 " steven at gcc dot gnu dot org
2004-09-19 15:42 ` bangerth at dealii dot org
2004-09-22 14:41 ` dnovillo at gcc dot gnu dot org
2004-09-22 16:14 ` dnovillo at gcc dot gnu dot org
2004-09-22 16:26 ` stevenb at suse dot de
2004-09-22 16:41 ` dnovillo at redhat dot com
2004-09-22 17:23 ` dnovillo at gcc dot gnu dot org
2004-09-22 17:23 ` dnovillo at gcc dot gnu dot org
2004-09-22 23:33 ` cvs-commit at gcc dot gnu dot org
2004-09-22 23:36 ` dnovillo at gcc dot gnu dot org

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=20040726181156.16721.macro@linux-mips.org \
    --to=gcc-bugzilla@gcc.gnu.org \
    --cc=gcc-bugs@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).