Mail Archives: djgpp/1992/02/24/18:30:57
hi,
Below are some minor problems (and some fixes) I found in DJGPP.
Since I just tried to subscribe to this list please CC me directly on
any reply. Several problems were already reported to DJ directly (who
kindly point me to this list).
* emu387 has a bug when comparing two negative numbers. Basically there are
compared as two positive and obviously return the opposite result. The
fix is trivial in compare.cc to add the following two lines just before
the diff sign tests, for the return statements:
if (a.sign) // The two numbers are negative.
diff = -diff;
* MouseWarp does not update last_event as it should and therefore does
not function. below is the modified MouseWarp from mouse.c
void MouseWarp(int x, int y)
{
REGISTERS r;
r.ax = 4;
r.cx = last_event.x = x;
r.dx = last_event.y = y;
int33(&r);
}
* Since no driver for ATI version 1 existed, I modified the version 2
provided. The modified version was mailed to DJ.
* CPP is defining 'unix'. Although, in general, this should not impose any
problem, any attempt to port a unix program that is using this constant
may cause problems. DJGPP is not running on unix system. Also how about
defining a constant for DJGPP (__DJGPP__!?).
* MouseGetEvent will always "eat" characters even if no M_KEYPRESS event
was specified. This is not the intuitive behaviour. If no M_KEYPRESS
event was requested then keyboard buffer should not be tempered with.
The fix is trivial in mouse.c:
instead of "if (kbhit())", do "if (flags & M_KEYPRESS && kbhit())"
* The line drawing in the graphics package can be optimized significatly
by unfolding the dot plotting into the line drawing and updating the
frame buffer pointer directly (no multiplication with 0xd0000000 is
required any more). Clipping should be done in advance instead
of at the dot plotting level.
I tried to unfold the dot plotting into the line drawing without
clipping and found a speedup factor of 2 to 3.
Before I dive further, are you aware of anyone else trying to do the
same thing?
* The small program below generates a bogus warning about ':' WITHIN a
comments. I have no idea why. The warning is:
bug.c:23:Rest of line ignored. 1st junk character valued 58 (:).
A script of the environment set and compilation follows.
These are minor things! I am VERY happy with the quality of this compiler.
In fact for a reason I do not understand, its generated code is much smaller
than BC++!!!
Thanks for such a fine compiler.
Gershon
-----------------------------------------------------------------------------
/* GNUPLOT - standard.c */
/*
* Copyright (C) 1986, 1987, 1990, 1991 Thomas Williams, Colin Kelley
*
* Permission to use, copy, and distribute this software and its
* documentation for any purpose with or without fee is hereby granted,
* provided that the above copyright notice appear in all copies and
* that both that copyright notice and this permission notice appear
* in supporting documentation.
*
* Permission to modify the software is granted, but not the right to
* distribute the modified code. Modifications are to be distributed
* as patches to released version.
*
* This software is provided "as is" without express or implied warranty.
*
*
* AUTHORS
*
* Original Software:
* Thomas Williams, Colin Kelley.
*
* Gnuplot 2.0 additions:
* Russell Lang, Dave Kotz, John Campbell.
*
* Gnuplot 3.0 additions:
* Gershon Elber and many others.
*
* Send your comments or suggestions to
* pixar!info-gnuplot AT sun DOT com.
* This is a mailing list; to join it send a note to
* pixar!info-gnuplot-request AT sun DOT com.
* Send bug reports to
* pixar!bug-gnuplot AT sun DOT com.
*/
/* jzero for x in [0,8]
* Index 5849, 19.22 digits precision
*/
double pjzero[] = {
0.4933787251794133561816813446e+21,
-0.11791576291076105360384408e+21,
0.6382059341072356562289432465e+19,
-0.1367620353088171386865416609e+18,
0.1434354939140346111664316553e+16,
-0.8085222034853793871199468171e+13,
0.2507158285536881945555156435e+11,
-0.4050412371833132706360663322e+8,
0.2685786856980014981415848441e+5
};
double qjzero[] = {
0.4933787251794133562113278438e+21,
0.5428918384092285160200195092e+19,
0.3024635616709462698627330784e+17,
0.1127756739679798507056031594e+15,
0.3123043114941213172572469442e+12,
0.669998767298223967181402866e+9,
0.1114636098462985378182402543e+7,
0.1363063652328970604442810507e+4,
0.1e+1
};
-----------------------------------------------------------------------------
Script V1.1 session started Sat Feb 22 09:02:08 1992
Microsoft(R) MS-DOS(R) Version 5.00
(C)Copyright Microsoft Corp 1981-1991.
C:\TEMP > djgcc
C:\TEMP >
C:\TEMP > path d:\;c:\bat;c:\tools\djgpp\bin;c:\bc\bin;c:\util;c:\unix;c:\windows;c:\cards;c:\graphics\bin;c:\msdos5;c:\gif
C:\TEMP >
C:\TEMP > set GCCBIN=c:\tools\djgpp\bin
C:\TEMP > set GCCINC=c:/tools/djgpp/include
C:\TEMP > set GCCLIB=c:/tools/djgpp/lib
C:\TEMP > set GCCTMP=c:/temp
C:\TEMP > set GO32TMP=c:/temp
C:\TEMP > set GO32=emu c:/tools/djgpp/emu387/emu387 ansi driver c:/tools/djgpp/drivers/ati.grd gw 640 gh 480
C:\TEMP > set GRMODE=
C:\TEMP >
C:\TEMP >
C:\TEMP > c:\msdos5\mem/c
Conventional Memory :
Name Size in Decimal Size in Hex
------------- --------------------- -------------
MSDOS 16464 ( 16.1K) 4050
HIMEM 1184 ( 1.2K) 4A0
TDH386 8656 ( 8.5K) 21D0
MOUSE 15424 ( 15.1K) 3C40
NANSI 2992 ( 2.9K) BB0
RAMDRIVE 1184 ( 1.2K) 4A0
COMMAND 3968 ( 3.9K) F80
CONV 3664 ( 3.6K) E50
FASTBF26 1440 ( 1.4K) 5A0
CMDEDIT 9712 ( 9.5K) 25F0
FDREAD 192 ( 0.2K) C0
SCRIPT 18432 ( 18.0K) 4800
PUSHDIR 768 ( 0.8K) 300
COMMAND 4608 ( 4.5K) 1200
FREE 64 ( 0.1K) 40
FREE 32 ( 0.0K) 20
FREE 32 ( 0.0K) 20
FREE 566096 (552.8K) 8A350
Total FREE : 566224 (553.0K)
Total bytes available to programs : 566224 (553.0K)
Largest executable program size : 565616 (552.4K)
1441792 bytes total contiguous extended memory
0 bytes available contiguous extended memory
1171456 bytes available XMS memory
MS-DOS resident in High Memory Area
C:\TEMP >
C:\C\GNUPLOT > gcc -c bug.c
bug.c:23:Rest of line ignored. 1st junk character valued 58 (:).
C:\C\GNUPLOT > exit
Script completed Sat Feb 22 09:03:04 1992
- Raw text -