X-Authentication-Warning: delorie.com: mail set sender to djgpp-bounces using -f
X-Recipient: djgpp AT delorie DOT com
DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed;
        d=gmail.com; s=gamma;
        h=mime-version:in-reply-to:references:date:message-id:subject:from:to
         :content-type:content-transfer-encoding;
        bh=B45D4JTx6lAV1ovNbTTthkDqcYNstIuefgJe71p+Gqk=;
        b=bJii+QJvmGH7HB8tJUsS0H4mCGfJn0WOoxbpl2un4YHVS9rgX8sa0w3u9wpM+jge9r
         vEgVbVQGs25plkzMPwozYFkehPqZVYSYQxIF15pJMNCCaq8AL8iC2GhK6aGGX2vlnmKv
         EeJDTO0wgSqAHc5l5pGCsT5oe61S2arRi0eXU=
MIME-Version: 1.0
In-Reply-To: <6d89c59a-0e99-4fe4-be98-0d8f2ced65d3@g30g2000yqc.googlegroups.com>
References: <201109171356 DOT 53360 DOT juan DOT guerrero AT gmx DOT de>
	<CAA2C=vD-Lzrt2z=gkkQcrgmssdh6F7vgU_kwVgu6o9iPGykMJQ AT mail DOT gmail DOT com>
	<2dc464f3-6f8c-4139-8dcb-0719e25116bf AT i39g2000yqn DOT googlegroups DOT com>
	<CAA2C=vAphY=pug=oSg-5Az05kZybY0NUYg6SS8d9xKL6f3AwRA AT mail DOT gmail DOT com>
	<ab129657-7f07-425e-861e-b4abeb0319e4 AT g31g2000yqh DOT googlegroups DOT com>
	<CAA2C=vAZ0r0g55PpGpJWL3CMhAX7bqGFivdi6fY0NgjPjqwrVw AT mail DOT gmail DOT com>
	<a9a46165-40d4-4cb4-8fa2-8930bbd5b3d2 AT o9g2000vbo DOT googlegroups DOT com>
	<ea39a0dc-55e2-4590-9b1a-fc10677490e4 AT 1g2000yqm DOT googlegroups DOT com>
	<CAA2C=vBhm0Ak33ddOehVvPWPqNP2OxWBH7A_6ZqP5fS1AH67dA AT mail DOT gmail DOT com>
	<6d89c59a-0e99-4fe4-be98-0d8f2ced65d3 AT g30g2000yqc DOT googlegroups DOT com>
Date: Wed, 21 Sep 2011 09:38:43 +0300
Message-ID: <CAA2C=vBWuDCtb3XKeVWB3Agi=PHQWEc0uapAYoUQXt-nWrOSZQ@mail.gmail.com>
Subject: Re: Hammer of Thyrion (was Re: Isues concerning the INT 21 ...)
From: Ozkan Sezer <sezeroz AT gmail DOT com>
To: djgpp AT delorie DOT com
Content-Type: text/plain; charset=ISO-8859-1
Content-Transfer-Encoding: 8bit
X-MIME-Autoconverted: from quoted-printable to 8bit by delorie.com id p8L6cm7E004172
Reply-To: djgpp AT delorie DOT com
Errors-To: nobody AT delorie DOT com
X-Mailing-List: djgpp AT delorie DOT com
X-Unsubscribes-To: listserv AT delorie DOT com
Precedence: bulk

On Wed, Sep 21, 2011 at 3:43 AM, Rugxulo <rugxulo AT gmail DOT com> wrote:
> Hi again,
>
> On Sep 20, 12:47 am, Ozkan Sezer <seze DOT  DOT  DOT  AT gmail DOT com> wrote:
>> On Tue, Sep 20, 2011 at 5:38 AM, Rugxulo <rugx DOT  DOT  DOT  AT gmail DOT com> wrote:
>>
>> >>>  (Though you're using old CWSDPMI versions, dunno why, latest is r7.)
>>
>> >> I don't know how tested that one is, so I decided to
>> >> stay safe and use r5+latest updates to it
>>
>> > It's tested fairly well, and CWS considers it stable. DJ's Zip Picker
>> > suggests it by default now
>>
>> OK then. Will test it for next versions.
>
> I'm on a "big" RAM machine here. So when I test in native FreeDOS,
> there's gobs of memory. Apparently it doesn't like that, tries to
> malloc a negative amount by default.   ;-)   I assume this is an old
> Quake (2 GB??) limitation. That's using r7. When using r5, it can't

Hrmph..

sys_dos.c :: void *dos_getmaxlockedmem(int *size)
{
[...]
 __dpmi_free_mem_info meminfo;
 __dpmi_meminfo info;
 int working_size;
[...]
 working_size = meminfo.maximum_locked_page_allocation_in_pages * 4096;

I guess the problem lies in the above lines. You see, we're
using a signed int for working_size, whereas all members of
the __dpmi_free_mem_info structure are unsigned long.
So yes, a 2GB limit due to bad data type usage. Will try
fixing it _and_ limiting it to 2GB properly.

> load all the page tables, so similar memory problem. (CWS would
> probably find this funny. And yes it's using /beta/ 2.04, probably
> from CVS since stub says Sep. 2011.)

Yes, the current 1.5.0 build is made using djgpp v2.04 / 20110910
cvs snapshot

>
> The solution in both cases is to manually use "-mem 64". I honestly
> wonder if maybe you should clamp max. memory usage to a fixed
> reasonable amount to avoid this. As mentioned in the docs, 32 MB is
> suggested, so it's not like it needs all RAM anyways. Well, or you

Yes it has more than it needs with -mem 64

> could ignore it, most people probably? won't run natively due to
> stupid SB-incompatible sound cards anyways. (I halfway wish PC speaker
> w/ digital samples was supported, but I'd have to dig up some sources
> to figure that out, probably not easy for a dummy like myself. Yeah
> yeah, just use DOSEMU or DOSBox.)
>
>> > Yeah, 1.5.0 is officially out! That was fast! A quick play-through
>> > again of the demo (DOS version under DOSEMU) showed no problems.
>>
>> Thanks for testing!
>
> BTW, I just now tried again under DOSEMU. The console command "music
> blah" works great (tested one .mp3, two .ogg) though of course stops
> the song when entering a new hub or restarting the game.

Well, thats very normal: if you rip your cdrom, make mp3 or ogg
with them and put the files under the game accordingly, it will play
the game's music upon every level change.

>
> (BTW, no mouse wheel / button 3 in DOSEMU unlike when using CTMouse in
> native FreeDOS, but I never found a good use for that yet anyways.
> Crouch?? Anyways, I don't prefer the mouse, only keyboard.)
>

Changing weapon, scrolling inventory, ...

>> > (Stupid mill key, heh.)
>>
>> :)
>
> Seriously, it's a quite fun game. I guess I should go ahead and grab
> the full version off eBay or wherever (preferably not Steam). But man,
> the demo is ultra confusing if you don't know where to go! (I've
> probably completed the demo 50 times over the past year.) Heh, not
> exactly obvious to jump here, shoot there, open this, grind that,
> etc.   ;-)
>
>

On Wed, Sep 21, 2011 at 7:10 AM, Charles Sandmann <cwsdpmi AT earthlink DOT net> wrote:
> The memory code in Quake has some horrible hacks to try to
> make it run on like an 8MB machine under Windows 95.  In

Well the issue he's describing doesn't need running under w95.

> those days the thought of a 2GB machine would be unbelievable.
> (That code tries to make sure it's all paged in, locked in memory,
> and everything else is paged out).
>
> I suggest using sbrk() when you need to use more than 2GB of memory :-)
>

> No so funny, but it shows to not make assumptions.
>

Exactly.

--
O.S.