bslathi19/cc65
1
0
Fork 0
Code Issues Pull Requests Actions 1 Packages Projects Releases Wiki Activity
11,892 Commits 1 Branch 0 Tags
14a6c66997236dbdbfe95d243248cd5938a85900
Commit Graph

192 Commits

Author SHA1 Message Date
Kugel Fuhr
c7096ab6d7 Fix an error that was introduced by me in a4a24280f2: 
Using a pipe causes a subshell to be generated so the "exit" statements will
just leave the subshell. As a consequence, the sortedxxx.sh scripts produced
output but no error exit code. Which in turn caused an error in #2778 to slip
through undetected.
 2025-07-13 17:56:31 +02:00
Colin Leroy-Mira
20a9db757d Optimize multiple incax* and incax*/ldaxi 
- Group multiple calls to incax* and decax* into a single one
- Replace incaxN/jsr ldaxi with ldy #N+1/jsr ldaxidx

Fixes #2055
 2025-07-06 20:37:01 +02:00
Colin Leroy-Mira
fe86149203 Further optimize inc/decsp 
The optimizer can now merge incsp+decsp together, and merge
jsr incsp/decsp + jmp incsp/decsp together. Drop the code
altogether if both instructions cancel each other out.

Fixes #2748
 2025-07-06 13:30:33 +02:00
Bob Andrews
801508aecf Merge pull request #2735 from kugelfuhr/kugelfuhr/fix-2025 
Add an optimization that removes compare instructions preceeding RTS or function calls
 2025-06-26 00:29:29 +02:00
Kugel Fuhr
55d9b2dbd0 Rewrite the optimization step from c0a1b1b887 to 
remove compares not only before RTS but also befoire function calls that don't
use the flags but destroy all of them. The latter is currently the case for
all functions called. This way the optimization covers a lot more cases than
just checking for RTS.
 2025-06-25 20:07:36 +02:00
Kugel Fuhr
7a6c60ade4 Do also sort variables to satisfy sorted_codeopt.sh. 2025-06-24 20:21:43 +02:00
Kugel Fuhr
c0a1b1b887 Add an optimization step that removes compare instructions preceeding an RTS. 
Since nothing is passed back in the flags, these instructions have no effect.
Fixes #2025.
 2025-06-24 18:02:24 +02:00
Gorilla Sapiens
a7af49a763 Merge branch 'master' into c_sp 2025-06-22 19:02:24 +00:00
Kugel Fuhr
2ceaa3fabb Added a few simple optimizations that resolve most of #2527. 2025-06-18 10:34:13 +02:00
GorillaSapiens
ae54171b8a Merge branch 'master' into c_sp 2025-06-13 17:43:06 -07:00
Bob Andrews
af2ab29a91 Merge branch 'master' into c_sp 2025-06-12 23:38:48 +02:00
mrdudz
aaa1058d32 use explicit markers (comments) for the bsearch table checking, simplifies the scripts and makes them more robust too :) 2025-06-09 21:48:20 +02:00
mrdudz
ba80de5efc fix bsearch tables that must be sorted, add comment to all tables that must be sorted 2025-06-09 17:58:58 +02:00
Gorilla Sapiens
b6f42f9ab2 changed "spc" to "c_sp" 2025-06-04 06:37:59 +00:00
Gorilla Sapiens
3d118dc6e5 rename "sp" to "spc", avoid conflict with 4510 opcodes 2025-06-04 02:06:40 +00:00
Kugel Fuhr
6e18e0880a Added/improved the optimizations: 
* Added a new pass that optimizes PHA/PLA sequences
* Added a new pass that optimizes AND/EOR/ORA when an operand is known
* Added a run of an existing pass at later stages to remove code that
  otherwise goes unchanged.
* Handle binary operations in OptUnusedLoads in addition to real loads.
 2024-09-13 19:30:38 +02:00
Colin Leroy-Mira
d8a3938f2b Optimize a bit static long assignation 2023-12-01 14:20:00 +01:00
Colin Leroy-Mira
076137f41b Optimize lda/sta/lda and friends 2023-11-13 10:32:09 +01:00
acqn
70549e868e New optimizer steps to restore some possibly lost optimization with boolean due to the previous fix. 2023-10-21 23:56:13 +08:00
acqn
79c52e742f Added new opts OptBoolUnary1/OptBoolUnary2 to remove unnecessary cmp + bcastax/bnegax, 
as well as OptBoolUnary3 to "strength-reduce" certain bcastax/bnegax to boolne/booleq.
 2023-10-21 23:56:04 +08:00
acqn
e5bbdfa995 Separated boolean optimizers from bitwise unary operator optimizers. 
Renamed OptCmp6 to OptBoolCmp.
 2023-10-21 23:56:01 +08:00
mrdudz
ddab16007a run branch fixer again after replacing BRA by JMP. should fix #1936 2022-12-11 12:08:30 +01:00
compyx
2fdb1e6b0d Fix set but unused variable 2022-05-17 10:05:25 +02:00
mrdudz
43af562ce1 At the end of the optimizer run, remove jump cascades again. fixes the 
regression reported in #1652
 2022-05-08 13:22:10 +02:00
acqn
2aad72af90 Removed trailing whitespaces. 2021-03-03 10:07:24 +01:00
acqn
bb7b69f513 Added OptShiftBack for shifting the C flag into register A and back. 2020-10-20 21:55:39 +02:00
acqn
0354322413 Added OptSignExtened for testing signness right after sign extention. 2020-10-20 21:55:39 +02:00
acqn
4f3a96a535 Added new opt OptPushPop2. 
Also renamed OptPushPop to OptPushPop1.
 2020-10-20 21:55:39 +02:00
acqn
64ef562fa7 Moved all optimization steps from codeopt.c and some optimization steps from coptind.c into new separate files. 2020-09-08 14:18:22 +02:00
acqn
676b14429d Better opt chance for certain optimization steps e.g. OptPtrStore1 etc. 2020-09-04 12:00:10 +02:00
acqn
29d1999947 Rearranged the OptFunc's in alphabetic order. 2020-09-04 12:00:10 +02:00
acqn
761d00b7dc Ignore running removed old optimization steps. Neither list them in help info. 2020-09-04 12:00:10 +02:00
Steven Hugg
dca99d59e5 rearranged order of new optimizations to better handle -Oi flag 2019-03-31 22:33:22 -04:00
Steven Hugg
5d274e4bc5 added optimization for indexed 16-bit array load of form (array[i & 0x7f]) 2019-03-31 22:33:22 -04:00
Steven Hugg
ffc118cd94 added optimization for indexed pointer load of a constant, e.g.: y = ((unsigned char*)0x100)[i] 2019-03-31 22:33:22 -04:00
Greg King
a6b04f6e97 Changed most "backticks" (grave accents) into apostrophes. 
Quotations that are embraced by tick marks now look better, in most fonts.
 2019-01-05 14:57:12 -05:00
laubzega
5fa79be997 Post-review tweaks. 2018-10-16 09:55:02 +02:00
laubzega
7d9485f6bc Typo 2018-10-16 09:55:02 +02:00
laubzega
cf7f3abadd Even more optimization, cleanup, bugfix, comments. 2018-10-16 09:55:02 +02:00
laubzega
d3665b263e Stack adjustment code optimizations. 2018-10-16 09:55:02 +02:00
Chris Cacciatore
66b30f0c7a Added 'any' to --list-opt-steps. 2016-12-15 09:06:15 -08:00
Greg King
0390c34e88 Changed multi-line C comments into another style. 
The left side doesn't look unbalanced.
 2014-06-30 16:51:07 -04:00
Oliver Schmidt
4185caf855 Normalized code. 2014-03-04 01:11:19 +01:00
Oliver Schmidt
85885001b1 Removed (pretty inconsistently used) tab chars from source code base. 2013-05-09 13:57:12 +02:00
uz
e8964cde74 Quick fix for a problem in the OptLoad3 function. The fix isn't perfect and 
may fail for inline assembly code, but at least it's less buggy than before.


git-svn-id: svn://svn.cc65.org/cc65/trunk@5982 b7a2c559-68d2-44c3-8de9-860c34a00d81
 2013-02-12 19:51:11 +00:00
uz
d6290e0a0f Do an additional run of duplicate loads at the end of OptGroup4. 
git-svn-id: svn://svn.cc65.org/cc65/trunk@5959 b7a2c559-68d2-44c3-8de9-860c34a00d81
 2013-01-07 20:17:50 +00:00
uz
49e5d19950 Added an option to output the assembly after each transformation step of the 
optimizer.


git-svn-id: svn://svn.cc65.org/cc65/trunk@5781 b7a2c559-68d2-44c3-8de9-860c34a00d81
 2012-07-12 20:30:01 +00:00
uz
40a3084553 Add another form of duplicate load removal. 
git-svn-id: svn://svn.cc65.org/cc65/trunk@5778 b7a2c559-68d2-44c3-8de9-860c34a00d81
 2012-07-10 20:46:21 +00:00
uz
58530c142a Rearrange order of opt steps for better efficiency. 
git-svn-id: svn://svn.cc65.org/cc65/trunk@5774 b7a2c559-68d2-44c3-8de9-860c34a00d81
 2012-07-08 21:56:01 +00:00
uz
3365220827 Added optimization for complax. 
git-svn-id: svn://svn.cc65.org/cc65/trunk@5771 b7a2c559-68d2-44c3-8de9-860c34a00d81
 2012-07-07 19:54:24 +00:00