X-Authentication-Warning: delorie.com: mail set sender to djgpp-workers-bounces using -f
Date: Sat, 13 Nov 2004 23:52:05 +0200
From: "Eli Zaretskii" <eliz AT gnu DOT org>
Sender: halo1 AT zahav DOT net DOT il
To: djgpp-workers AT delorie DOT com
Message-ID: <01c4c9cb$Blat.v2.2.2$29550bc0@zahav.net.il>
Content-Transfer-Encoding: 7BIT
Content-Type: text/plain; charset=ISO-8859-1
X-Mailer: emacs 21.3.50 (via feedmail 8 I) and Blat ver 2.2.2
In-reply-to: <10411131756.AA83688@clio.rice.edu> (sandmann@clio.rice.edu)
Subject: Re: Alignment problem on Windows XP
References:  <10411131756 DOT AA83688 AT clio DOT rice DOT edu>
Reply-To: djgpp-workers AT delorie DOT com
Errors-To: nobody AT delorie DOT com
X-Mailing-List: djgpp-workers AT delorie DOT com
X-Unsubscribes-To: listserv AT delorie DOT com
Precedence: bulk

> From: sandmann AT clio DOT rice DOT edu (Charles Sandmann)
> Date: Sat, 13 Nov 2004 11:56:17 -0600 (CST)
> 
> I would first want to check the binary image used by the user with the
> problem - it could be some installation problem that generates bad
> executables?

I asked the user for the binary, being unable to reproduce the problem
on 2 XP boxes I have access to.

> The example problem has the data in BSS, so I don't see how it could
> have any address offsets.

Sorry, I don't follow: what address offsets, and why BSS data cannot
have them?  Could you please elaborate?

> If the XP DPMI provider sometimes returns blocks with alignment less than
> we assume, I think the selector base would fix this (it would run slowly,
> but the offsets would all be correct).

I'm not sure I follow this, either, but FWIW, the temacs crash (on
that user's machine; it works okay for me) during GC, which brought
the problem to our attention in the first place, clearly shows that
the selector base is aligned:

 Exiting due to signal SIGABRT
 Raised at eip=00168e5e
 eax=002f7e5c ebx=00000120 ecx=00000000 edx=00000000 esi=00000054 edi=002dfa78
 ebp=002f7f08 esp=002f7e58 program=D:\src\emacsCVS\src\temacs.exe
 cs: sel=024f  base=02f30000  limit=003affff
 ds: sel=0257  base=02f30000  limit=003affff
 es: sel=0257  base=02f30000  limit=003affff
 fs: sel=0227  base=00018030  limit=0000ffff
 gs: sel=0277  base=00000000  limit=0010ffff
 ss: sel=0257  base=02f30000  limit=003affff
 App stack: [002f8a78..002dfa78]  Exceptn stack: [002df94c..002dda0c]

 Call frame traceback EIPs:
   0x00168d84 ___djgpp_traceback_exit+48
   0x00168e5e _raise+90
   0x00152f3c _abort+55, line 5269 of msdos.c
   0x000de42c _mark_object+2599, line 5122 of alloc.c

> Since we allocate multiple memory blocks unless unixy sbrk is set, if 
> XP chose to return a block with an offset of modulo 4 bytes alignment 
> to the first block, all of the stuff in stack and/or sbrk() calls would
> be at danger - we have code that tries to do alignment but I don't know
> if it always works (probably not tested much).

The test program is a tiny one--would it at all be at danger of
hitting the multiple-block issue?

In any case, thanks for the feedback.