Skip site navigation (1)Skip section navigation (2)
Date:      Fri, 12 May 2000 13:31:02 -0700
From:      Peter Wemm <peter@netplex.com.au>
To:        John Baldwin <jhb@FreeBSD.org>
Cc:        "Daniel C. Sobral" <dcs@newsguy.com>, freebsd-alpha@FreeBSD.org, Boris Popov <bp@butya.kz>
Subject:   Re: Loader fixed 
Message-ID:  <20000512203102.3EC601CE1@overcee.netplex.com.au>
In-Reply-To: Message from John Baldwin <jhb@FreeBSD.org>  of "Fri, 12 May 2000 09:34:43 EDT." <200005121334.JAA10067@server.baldwin.cx> 

next in thread | previous in thread | raw e-mail | index | archive | help
John Baldwin wrote:
> 
> On 12-May-00 Daniel C. Sobral wrote:
> > Boris Popov wrote:
> >> 
> >>         There was a bug introduced by recent commit to the loader. It was
> >> caused by long != int problem. Now it fixed. However loader still has
> >> problem with sizes more 128k? and Peter Wemm has disabled FICL support on
> >> Alphas.
> > 
> > If there is a problem with sizes greater than 128Kb, that would
> > certainly explain the problems with FICL. Alas, why do you say that?
> > Have you actually seen any suspect code?
> > 
> > As for Peter disabling FICL on Alpha... it was never enabled in first
> > place!
> 
> It was enabled.  However, the problem is not 128k.  The loader from 4.0
> worked fine and it was this size:

No, it was not enabled, because it would not work on the Alpha. It crashed
mysteriously and randomly in impossible places.  Inserting debugging
printf()'s changed the location it died.  It linked into the binary to make
sure the code didn't suffer from new bitrot - but the actual call to
initialize ficl was not called.  I changed the alpha build to not even
link libficl.a and interp_forth.c into the loader.

Inserting debugging printf's changed the location that FICL failed for
inexplicable reasons.  Boris's code changed the size of the core
loader, and the "mystery" corruption moved elsewhere and affected something
outside FICL this time and caused the zfree problem.

I fully suspect that there is nothing wrong with FICL at all, just that the
code is too big and we cannot fit it all into the same executable space.

Cheers,
-Peter
--
Peter Wemm - peter@FreeBSD.org; peter@yahoo-inc.com; peter@netplex.com.au
"All of this is for nothing if we don't go to the stars" - JMS/B5



To Unsubscribe: send mail to majordomo@FreeBSD.org
with "unsubscribe freebsd-alpha" in the body of the message




Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?20000512203102.3EC601CE1>