public inbox for gcc-prs@sourceware.org
help / color / mirror / Atom feed
From: torek@bsdi.com
To: gcc-gnats@gcc.gnu.org
Subject: optimization/7690: gcc 2.95.3: argument destroyed under -O2 on IA32
Date: Thu, 22 Aug 2002 15:56:00 -0000	[thread overview]
Message-ID: <20020822220202.7531.qmail@sources.redhat.com> (raw)


>Number:         7690
>Category:       optimization
>Synopsis:       gcc 2.95.3: argument destroyed under -O2 on IA32
>Confidential:   no
>Severity:       serious
>Priority:       medium
>Responsible:    unassigned
>State:          open
>Class:          sw-bug
>Submitter-Id:   net
>Arrival-Date:   Thu Aug 22 15:06:00 PDT 2002
>Closed-Date:
>Last-Modified:
>Originator:     torek@bsdi.com
>Release:        gcc 2.95.3
>Organization:
>Environment:
BSD/OS 4.3
>Description:
The stripped-down C code below generates IA32 assembly that destroys the "name" parameter before making another call, but only when using -O2.

This code is simplified from a real program (squid) and no longer does anything useful on its own, so you must look at the generated assembly.

I was hoping this is a known bug with a known fix that I can find somewhere so as to patch it.
>How-To-Repeat:
typedef unsigned long long off_t;
typedef unsigned int size_t;

off_t f1(char *buf, size_t sz, const char *name);
void f2(int);

void bug(char *buf, size_t sz, const char *name, unsigned type)
{
    off_t off;

    off = f1(buf, sz, name);
    memcpy(buf + off, &type, sizeof(type));
    f2(off <= sz);
}

compile with -O2, examine assembly, note this section:

        movl 8(%ebp),%edi
        movl 12(%ebp),%eax
        xorl %edx,%edx
        movl %eax,12(%ebp)
        movl %edx,16(%ebp)

note that 16(%ebp) used to hold the parameter "name" before this last movl, which (since %edx is now 0) clobbers it.

Using -O1 -fexpensive-optimizations appears to suffice to trigger the problem (but -O1 alone does not).
>Fix:

>Release-Note:
>Audit-Trail:
>Unformatted:


             reply	other threads:[~2002-08-22 22:06 UTC|newest]

Thread overview: 5+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2002-08-22 15:56 torek [this message]
2002-08-22 18:06 Chris Torek
2002-08-23  0:06 Chris Torek
2002-12-06 13:16 bangerth
2003-03-12 18:47 neroden

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=20020822220202.7531.qmail@sources.redhat.com \
    --to=torek@bsdi.com \
    --cc=gcc-gnats@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).