X-Authentication-Warning: delorie.com: mail set sender to djgpp-bounces using -f
From: David Winfrey <dlw AT neveranyspam DOT patriot DOT net>
Newsgroups: comp.os.msdos.djgpp
Subject: Re: BIOS 13h disk i/o buffers
Date: Mon, 05 Jan 2004 20:40:03 -0000
Organization: Posted via Supernews, http://www.supernews.com
Message-ID: <vvjit3maq0rob2@corp.supernews.com>
Sender: David Winfrey <dlw AT adams DOT patriot DOT net>
References: <vvjh4soe8ib093 AT corp DOT supernews DOT com> <200401052022 DOT i05KM4A6005335 AT envy DOT delorie DOT com>
User-Agent: tin/1.5.8-20010221 ("Blue Water") (UNIX) (Linux/2.4.23 (i686))
X-Complaints-To: abuse AT supernews DOT com
Lines: 14
To: djgpp AT delorie DOT com
DJ-Gateway: from newsgroup comp.os.msdos.djgpp
Reply-To: djgpp AT delorie DOT com

DJ Delorie <dj AT delorie DOT com> wrote:
: 
:> Is the transfer buffer that biosdisk.c uses for Int 13h i/o functions
:> guaranteed to not cross a 64K physical memory boundary?
: 
: If you're talking about DMA, no.  If you're just talking about segment
: offsets, yes.

Thanks.  My understanding of the 13h/02h and 13h/03h BIOS interrupts is
that the user-supplied buffer must not cross a 64K DMA boundary. How
does the DJGPP code handle this, or am I misreading the BIOS 
documentation?  I see that the buffer is obtained from
dpmi_allocate_dos_memory, but I'm having trouble tracing through that
to find the details.