Commit Graph

50 Commits

Author SHA1 Message Date
Irmen de Jong
9c7a645e18 remove non-functional verafx.mult(). note: muls() is still there and just fine!
added documentation/source code comments to the cpu word*word multiplication routine not producing the correct upper 16 bits.
2024-09-28 00:43:05 +02:00
Irmen de Jong
32becdbced add monogfx lib to virtual target 2023-10-24 00:16:25 +02:00
Irmen de Jong
390263a34e added cx16 verafx library module 2023-09-24 23:00:40 +02:00
Irmen de Jong
e97303c226 fix word multiplication to not clobber r0 and r1 anymore
This was causing corruption in certain programs such as the cx16/amiga example.
The problem was introduced in 9.4 with the new multiply_words routine
2023-09-02 20:52:16 +02:00
Irmen de Jong
3b786c819d avoid using temp var even more 2023-09-01 23:47:01 +02:00
Irmen de Jong
be06d871b6 fix code for bitwise shifts by zero 2023-08-14 21:49:13 +02:00
Irmen de Jong
bc8126eb16 2x faster word multiplication routine 2023-08-14 18:11:30 +02:00
Irmen de Jong
4c8beefdcb slightly faster integer bytes multiplication routine 2023-08-14 17:00:16 +02:00
Irmen de Jong
bbb6c53457 slightly faster sqrt() routine for integers 2023-08-14 17:00:02 +02:00
Irmen de Jong
2c9e50873c use math.square for optimized X*X calculation (words only).
Added IR SQUARE instruction.
2023-08-14 01:05:17 +02:00
Irmen de Jong
47485e4b49 added more missing codegen for bit shifts 2023-07-16 17:42:10 +02:00
Irmen de Jong
3841cef497 implemented missing bitshift codegen (non-stack) 2023-07-15 22:26:56 +02:00
Irmen de Jong
41af63b333 remove even more X register save/restore code 2023-07-15 22:26:54 +02:00
Irmen de Jong
9e8c8973d8 remove eval stack references from asm code 2023-07-15 22:24:22 +02:00
Irmen de Jong
1f7180d9a8 math.multiply_words returns lower 16 bits of the result also in AY (to avoid repeating some load instructions) 2023-07-06 22:54:13 +02:00
Irmen de Jong
adcbe55307 replaced integer RNG with smaller and faster routine. 2022-10-22 22:01:57 +02:00
Irmen de Jong
ec5adffdc2 rnd()/rndf() routines can now be seeded with new rndseed()/rndseedf() routines. fixes #80 2022-10-22 13:34:22 +02:00
Irmen de Jong
a735939d1e removed confusing GPL software license reference and copyright header from library files. (because of exclusion in output files)
Reworded software license and exclusion clause somewhat again in attempt to make it even clearer.
2022-05-30 20:12:20 +02:00
Irmen de Jong
0a5b30e21c added fast code for x*640 2021-04-30 22:30:21 +02:00
Irmen de Jong
8b84f87217 removed fastrnd8() because it was hilariously bad, just use rnd() 2021-04-28 01:53:12 +02:00
Irmen de Jong
60e169bd87 added optimized integer square (x*x) routine 2021-03-08 23:08:47 +01:00
Irmen de Jong
dffd0a2706 added fastrnd8() with the old rnd() generator code in it, new code for rnd() uses the much better rndw() generator now. 2021-03-05 22:49:14 +01:00
Irmen de Jong
76f459ee95 amiga 2021-02-02 23:09:03 +01:00
Irmen de Jong
ee7f9d457d text editor configs 2021-01-07 01:56:31 +01:00
Irmen de Jong
d65cfbf093 fixed math.mul_word_40 that was actually doing *80... 2020-12-23 00:54:11 +01:00
Irmen de Jong
4977d1fbd5 bit shift expressions are "expanded" to the target value's datatype, now also for subroutine arguments.
implemented word bit shifts by variable number of bits.
2020-12-14 20:44:48 +01:00
Irmen de Jong
452c29574d added optimized mul 320 routine 2020-12-07 22:55:16 +01:00
Irmen de Jong
aaa20093ef cleaning up and correcting cc for builtin functions 2020-11-06 00:56:26 +01:00
Irmen de Jong
9d17421c66 implemented the arithmetic functions with new cc. fixed sgn(). 2020-11-04 02:27:29 +01:00
Irmen de Jong
3e28ed4fe4 mader versions of abs() and sgn() returning value in register 2020-10-28 22:56:13 +01:00
Irmen de Jong
aec3b82476 fixed bitshifting by more than the number of bits in the value 2020-10-19 02:05:01 +02:00
Irmen de Jong
c426f4626c added some more missing aug assign operator code 2020-10-07 22:53:18 +02:00
Irmen de Jong
af6731c9c8 preparing version 4.3 2020-09-22 21:50:56 +02:00
Irmen de Jong
25cf0d2b94 don't suggest a mult replacement routine to be used, faster ones are likely to require large tables 2020-09-22 21:19:01 +02:00
Irmen de Jong
2201765366 mult fixes 2020-09-20 00:17:33 +02:00
Irmen de Jong
48858019b7 added the last of the optimized mul_word asm routines 2020-09-14 23:54:01 +02:00
Irmen de Jong
aff6b1fca5 added some more optimized mul_word asm routines 2020-09-14 23:03:18 +02:00
Irmen de Jong
d260182ef3 added some more optimized mul_byte asm routines 2020-09-14 22:06:40 +02:00
Irmen de Jong
ea8b7ab193 reg_x removal: math.asm and some others 2020-09-13 20:38:50 +02:00
Irmen de Jong
e21aa2c8f0 better naming of the optimized math mult routines 2020-09-09 22:16:37 +02:00
Irmen de Jong
e112dfd910 implemented signed byte and word division 2020-08-29 00:00:53 +02:00
Irmen de Jong
0b55372b3b cleanup cx16 things and added call signatures. c64graphics moved into built-in libraries. 2020-08-28 21:42:53 +02:00
Irmen de Jong
d97da3bb7b implemented almost all math operations 2020-08-27 20:47:22 +02:00
Irmen de Jong
78bdbde3ae refer to ZP scratch constants from asm code via the global P8ZP constants as well 2020-08-25 19:44:08 +02:00
Irmen de Jong
ef7744dbda asm fix 2020-08-21 04:02:10 +02:00
Irmen de Jong
50a85ee6b0 attempt to optimize asm for bitshifts more. 2020-03-23 22:59:29 +01:00
Irmen de Jong
27f987f0ae fixed bit shifts, added sgn() function 2019-08-17 16:44:28 +02:00
Irmen de Jong
905d8a0c06 actually, get rid of integer pow() because a naive multiplication loop approach is way too slow 2019-03-31 18:04:19 +02:00
Irmen de Jong
d05d8de447 optimize math from Soci's suggestions 2019-02-12 23:16:13 +01:00
Irmen de Jong
5471c5211c make %asminclude work, restructure prog8lib into separate asm file 2019-01-21 00:49:48 +01:00