cc65-devel Mailing List for cc65
cc65 - a freeware C compiler for 6502 based systems
Brought to you by:
gpz
You can subscribe to this list here.
| 2013 |
Jan
|
Feb
|
Mar
|
Apr
(2) |
May
(75) |
Jun
(146) |
Jul
(34) |
Aug
(44) |
Sep
(44) |
Oct
(15) |
Nov
(15) |
Dec
(41) |
|---|---|---|---|---|---|---|---|---|---|---|---|---|
| 2014 |
Jan
(5) |
Feb
(14) |
Mar
(335) |
Apr
(101) |
May
(28) |
Jun
(4) |
Jul
|
Aug
(7) |
Sep
(41) |
Oct
(27) |
Nov
(17) |
Dec
(6) |
| 2015 |
Jan
(21) |
Feb
(9) |
Mar
(12) |
Apr
(14) |
May
(58) |
Jun
(48) |
Jul
(26) |
Aug
(16) |
Sep
(16) |
Oct
|
Nov
(27) |
Dec
(1) |
| 2016 |
Jan
(39) |
Feb
(9) |
Mar
(4) |
Apr
(17) |
May
(53) |
Jun
(46) |
Jul
(12) |
Aug
(9) |
Sep
(6) |
Oct
(4) |
Nov
|
Dec
(5) |
| 2017 |
Jan
(26) |
Feb
(24) |
Mar
(57) |
Apr
(19) |
May
(91) |
Jun
(110) |
Jul
(21) |
Aug
(21) |
Sep
(11) |
Oct
(13) |
Nov
(37) |
Dec
(15) |
| 2018 |
Jan
(2) |
Feb
(21) |
Mar
(14) |
Apr
(4) |
May
|
Jun
(2) |
Jul
(17) |
Aug
(35) |
Sep
(56) |
Oct
(4) |
Nov
(5) |
Dec
(3) |
| 2019 |
Jan
(9) |
Feb
|
Mar
(14) |
Apr
(12) |
May
(46) |
Jun
(42) |
Jul
(25) |
Aug
|
Sep
(2) |
Oct
(27) |
Nov
(8) |
Dec
|
| 2020 |
Jan
(5) |
Feb
(15) |
Mar
(4) |
Apr
(13) |
May
(8) |
Jun
(15) |
Jul
(82) |
Aug
(8) |
Sep
(5) |
Oct
(9) |
Nov
(22) |
Dec
(23) |
| 2021 |
Jan
(2) |
Feb
(58) |
Mar
(20) |
Apr
(30) |
May
(19) |
Jun
(2) |
Jul
|
Aug
(11) |
Sep
(2) |
Oct
(5) |
Nov
(4) |
Dec
|
| 2022 |
Jan
(15) |
Feb
|
Mar
(13) |
Apr
(22) |
May
(9) |
Jun
(7) |
Jul
(10) |
Aug
|
Sep
(2) |
Oct
|
Nov
(23) |
Dec
|
| 2023 |
Jan
(17) |
Feb
|
Mar
(4) |
Apr
(5) |
May
|
Jun
(2) |
Jul
|
Aug
|
Sep
|
Oct
(1) |
Nov
(12) |
Dec
(2) |
| 2024 |
Jan
(21) |
Feb
(14) |
Mar
(15) |
Apr
(1) |
May
|
Jun
(23) |
Jul
(17) |
Aug
(3) |
Sep
(2) |
Oct
(2) |
Nov
(1) |
Dec
(2) |
| 2025 |
Jan
(4) |
Feb
|
Mar
|
Apr
|
May
(15) |
Jun
|
Jul
(11) |
Aug
|
Sep
(4) |
Oct
|
Nov
|
Dec
|
| 2026 |
Jan
(3) |
Feb
(6) |
Mar
|
Apr
(3) |
May
(9) |
Jun
|
Jul
|
Aug
|
Sep
|
Oct
|
Nov
|
Dec
|
|
From: Mark J. R. <mar...@gm...> - 2026-05-11 14:32:02
|
PETSCII control codes are handled by the CHROUT kernal call. They'll work
if you use studio functions - regular printf() or putchar(). But I believe
the conio routines bypass the kernal for performance, so the control codes
don't do anything. -
Mark J. Reed <mar...@gm...>
On Mon, May 11, 2026 at 10:23 Leif Bloomquist <lei...@gm...>
wrote:
> Hi all,
>
> Is there a way to have cprintf() emit PETSCII control characters (i.e.
> color changes) in the middle of a string? I know there is a
> character mapping "under the hood" that can be disabled, so I tried
> this:
>
> #define PETSCII_RED 0x1C
> #define PETSCII_WHITE 0x05
>
> // Override charmaps?
> #pragma charmap (PETSCII_RED, PETSCII_RED);
> #pragma charmap (PETSCII_WHITE, PETSCII_WHITE);
>
> cprintf("%cRED %cWHITE\n\r", PETSCII_RED, PETSCII_WHITE); // Shows
> random characters instead
>
>
> I also tried including #include "cbm_petscii_charmap.h" with no
> difference.
>
> I know using textcolor() is the "proper" way (and it works), but it
> makes the code unnecessarily complicated, especially for long,
> colorful strings.
>
>
> What's the best approach here?
>
> Thanks
> -Leif
>
>
> _______________________________________________________
> cc6...@li...
> https://lists.sourceforge.net/lists/listinfo/cc65-devel
>
|
|
From: groepaz <gr...@gm...> - 2026-05-11 14:30:59
|
Am Montag, 11. Mai 2026, 16:22:33 Mitteleuropäische Sommerzeit schrieb Leif
Bloomquist:
> Hi all,
>
> Is there a way to have cprintf() emit PETSCII control characters (i.e.
> color changes) in the middle of a string? I know there is a
> character mapping "under the hood" that can be disabled, so I tried
> this:
>
> #define PETSCII_RED 0x1C
> #define PETSCII_WHITE 0x05
>
> // Override charmaps?
> #pragma charmap (PETSCII_RED, PETSCII_RED);
> #pragma charmap (PETSCII_WHITE, PETSCII_WHITE);
>
> cprintf("%cRED %cWHITE\n\r", PETSCII_RED, PETSCII_WHITE); // Shows
> random characters instead
>
>
> I also tried including #include "cbm_petscii_charmap.h" with no
> difference.
>
> I know using textcolor() is the "proper" way (and it works), but it
> makes the code unnecessarily complicated, especially for long,
> colorful strings.
>
>
> What's the best approach here?
You'll have to use regular stdio, conio does not interpret control codes (note
that you cant mix conio and stdio, that will give undefined results even if it
might work by chance)
--
https://cc65.github.io https://rr.pokefinder.org
https://vice-emu.sourceforge.net https://magicdisk.untergrund.net
Intellectual property suffers in most cases from a significant lack of the
intellectual part.
<yepp in c.l.python>
|
|
From: Leif B. <lei...@gm...> - 2026-05-11 14:22:55
|
Hi all,
Is there a way to have cprintf() emit PETSCII control characters (i.e.
color changes) in the middle of a string? I know there is a
character mapping "under the hood" that can be disabled, so I tried
this:
#define PETSCII_RED 0x1C
#define PETSCII_WHITE 0x05
// Override charmaps?
#pragma charmap (PETSCII_RED, PETSCII_RED);
#pragma charmap (PETSCII_WHITE, PETSCII_WHITE);
cprintf("%cRED %cWHITE\n\r", PETSCII_RED, PETSCII_WHITE); // Shows
random characters instead
I also tried including #include "cbm_petscii_charmap.h" with no difference.
I know using textcolor() is the "proper" way (and it works), but it
makes the code unnecessarily complicated, especially for long,
colorful strings.
What's the best approach here?
Thanks
-Leif
|
|
From: Eric L. <hom...@gm...> - 2026-05-09 19:01:28
|
Thanks all. So it seems I can simply use clock_gettime(CLOCK_REALTIME, &start); then clock_gettime(CLOCK_REALTIME, &end); to calculate the time difference. I used another clock application to test and it seems pretty accurate. Is there anything else I need to do before calling those functions? Thanks. On Wed, May 6, 2026 at 2:25 PM Oliver Schmidt via cc65-devel < cc6...@li...> wrote: > Hi, > > > [...] I am not sure if current cc65 has any >> > convenience functions for that but you can always address the registers. >> > > https://github.com/cc65/cc65/blob/master/libsrc/c64/settime.s > https://github.com/cc65/cc65/blob/master/libsrc/c64/gettime.s > https://github.com/cc65/cc65/blob/master/libsrc/cbm/getres.s > > >> > Just don't forget to initialise correctly the 50//60 Hz flag. >> > > https://github.com/cc65/cc65/blob/master/libsrc/c64/tmcommon.s > > Regards, > Oliver > _______________________________________________________ > cc6...@li... > https://lists.sourceforge.net/lists/listinfo/cc65-devel > |
|
From: Oliver S. <ol...@we...> - 2026-05-06 18:25:36
|
Hi, > [...] I am not sure if current cc65 has any > > convenience functions for that but you can always address the registers. > https://github.com/cc65/cc65/blob/master/libsrc/c64/settime.s https://github.com/cc65/cc65/blob/master/libsrc/c64/gettime.s https://github.com/cc65/cc65/blob/master/libsrc/cbm/getres.s > > Just don't forget to initialise correctly the 50//60 Hz flag. > https://github.com/cc65/cc65/blob/master/libsrc/c64/tmcommon.s Regards, Oliver |
|
From: groepaz <gr...@gm...> - 2026-05-06 14:39:06
|
Am Mittwoch, 6. Mai 2026, 15:38:31 Mitteleuropäische Sommerzeit schrieb silverdr--- via cc65-devel: > > On 6 May 2026, at 02:06, Eric Lee <hom...@gm...> wrote: > > > > Thanks Marc and Janne. So in CC65 + C64U, what's the best approach to > > calculate time? > In C64, and presumably also in C64U, the CIAs have relatively reliable TOD > (for Time Of Day) registers with granularity to tenths of a second. > Whenever you want to measure time in human readable format (rather than e. > g. CPU clock cycles) in a 64 the TODs (rather than IRQ driven software > "jiffy clock") are the way to go. I am not sure if current cc65 has any > convenience functions for that but you can always address the registers. > Just don't forget to initialise correctly the 50//60 Hz flag. That said, if you need more precision, you can use the regular CIA timers - in that case i am sure there are no readily made utility functions for this though :) -- https://cc65.github.io https://rr.pokefinder.org https://vice-emu.sourceforge.net https://magicdisk.untergrund.net |
|
From: <sil...@sr...> - 2026-05-06 13:56:51
|
> On 6 May 2026, at 02:06, Eric Lee <hom...@gm...> wrote: > > Thanks Marc and Janne. So in CC65 + C64U, what's the best approach to calculate time? In C64, and presumably also in C64U, the CIAs have relatively reliable TOD (for Time Of Day) registers with granularity to tenths of a second. Whenever you want to measure time in human readable format (rather than e. g. CPU clock cycles) in a 64 the TODs (rather than IRQ driven software "jiffy clock") are the way to go. I am not sure if current cc65 has any convenience functions for that but you can always address the registers. Just don't forget to initialise correctly the 50//60 Hz flag. -- SD! |
|
From: groepaz <gr...@gm...> - 2026-05-06 12:12:46
|
Am Mittwoch, 6. Mai 2026, 04:06:27 Mitteleuropäische Sommerzeit schrieb Eric Lee: > Thanks Marc and Janne. So in CC65 + C64U, what's the best approach to > calculate time? use the TOD clock of the CIA -- https://cc65.github.io https://rr.pokefinder.org https://vice-emu.sourceforge.net https://magicdisk.untergrund.net I wouldn't recommend sex, drugs or insanity for everyone, but they've always worked for me. <Hunter S. Thompson> |
|
From: Eric L. <hom...@gm...> - 2026-05-06 02:06:51
|
Thanks Marc and Janne. So in CC65 + C64U, what's the best approach to calculate time? Eric On Thu, Apr 30, 2026 at 6:43 AM Marc Rintsch <ma...@ri...> wrote: > On 19/04/2026 03:49, Eric Lee wrote: > > > I used cc65 to write a small 1541 disk speed benchmark. I noticed that > > while using cbm_open() + cbm_write(), the writing speed is almost ~10x > > higher than reading via cbm_open() + cbm_read(). > > `clock()` doesn't work correctly. Its value is incremented in the IRQ > handler of the system, but that is disabled in timing critical > communication with the IEC bus. It affects both reading and writing, so > you get higher bytes/second values for both operations than they > actually are transferring. Writing then seems to be affected ten times > worse. > > Ciao, > Marc 'BlackJack' Rintsch > -- > »Whenever the C++ language designers had two competing ideas as to how > they should solve some problem, they said, “OK, we'll do them both”. > So the language is too baroque for my taste.« — Donald E. Knuth > > > _______________________________________________________ > cc6...@li... > https://lists.sourceforge.net/lists/listinfo/cc65-devel > |
|
From: Marc R. <ma...@ri...> - 2026-04-30 10:43:34
|
On 19/04/2026 03:49, Eric Lee wrote: > I used cc65 to write a small 1541 disk speed benchmark. I noticed that > while using cbm_open() + cbm_write(), the writing speed is almost ~10x > higher than reading via cbm_open() + cbm_read(). `clock()` doesn't work correctly. Its value is incremented in the IRQ handler of the system, but that is disabled in timing critical communication with the IEC bus. It affects both reading and writing, so you get higher bytes/second values for both operations than they actually are transferring. Writing then seems to be affected ten times worse. Ciao, Marc 'BlackJack' Rintsch -- »Whenever the C++ language designers had two competing ideas as to how they should solve some problem, they said, “OK, we'll do them both”. So the language is too baroque for my taste.« — Donald E. Knuth |
|
From: Janne J. <ice...@gm...> - 2026-04-20 06:23:17
|
> I used cc65 to write a small 1541 disk speed benchmark. I noticed that while using cbm_open() + cbm_write(), the writing speed is almost ~10x higher than reading via cbm_open() + cbm_read(). The test method is testing sequential read / write speed. I configured a 16KB buffer w/ 0 then write to disk to test the writing speed. I tested on VICE and C64U, got the same result. Here's a code snippet I implemented: https://github.com/meow-watermelon/c64/blob/master/diskspeed/diskspeed.c#L57-L85 > Could anyone show me some hints? Thanks in advance. Without even knowing, my guess is that your writes are (at least in part) benchmarking filling up the buffer, and not writing (the last sector) to disk. For reads, that is of course not possible, so you get truer numbers from it. A possible improvement would be to printf() after the close, might help, or worst case, ask for a small read after the write and see how long it takes until that finishes, since it will have to flush the writes presumably in order to get a new read in. -- May the most significant bit of your life be positive. |
|
From: Eric L. <hom...@gm...> - 2026-04-19 01:49:53
|
Hi, I used cc65 to write a small 1541 disk speed benchmark. I noticed that while using cbm_open() + cbm_write(), the writing speed is almost ~10x higher than reading via cbm_open() + cbm_read(). The test method is testing sequential read / write speed. I configured a 16KB buffer w/ 0 then write to disk to test the writing speed. I tested on VICE and C64U, got the same result. Here's a code snippet I implemented: https://github.com/meow-watermelon/c64/blob/master/diskspeed/diskspeed.c#L57-L85 Could anyone show me some hints? Thanks in advance. Eric |
|
From: Oliver S. <ol...@we...> - 2026-02-12 00:11:58
|
Hi,
Yes, indeed. It was fixed by acqn years ago IIRC (and it only "worked" by
> chance for some specific situations before).
Thanks for pointing out :-)
I'd write something like
>
> char *display[3] = {"Wait", "Load", "Paus"};
>
> myself (because incomplete types look wrong to me in any case)
Thanks for your input!
Regards,
Oliver
|
|
From: <gr...@gm...> - 2026-02-11 20:51:30
|
Yes, indeed. It was fixed by acqn years ago IIRC (and it only "worked" by
chance for some specific situations before).
I'd write something like
char *display[3] = {"Wait", "Load", "Paus"};
myself (because incomplete types look wrong to me in any case)
Am Mittwoch, 11. Februar 2026, 20:37:18 Mitteleuropäische Normalzeit schrieb
Mark J. Reed:
> > char display[][] = {"Wait", "Load", "Paus"};
>
> AFAIK that's never been legal in C. If cc65 used to allow it, I'd say that
> was a bug that has now been fixed?
>
> On Wed, Feb 11, 2026 at 12:28 PM Oliver Schmidt via cc65-devel <
>
> cc6...@li...> wrote:
> > Hi,
> >
> > Does really nobody have anything to say about my question(s) below?
> >
> > Wondering,
> > Oliver
> >
> > On Tue, Jan 27, 2026 at 1:18 AM Oliver Schmidt <ol...@we...> wrote:
> >> Hi,
> >>
> >> the line
> >>
> >> char display[][] = {"Wait", "Load", "Paus"};
> >>
> >> used to compile. Now I get
> >>
> >> Array of incomplete element type 'char[]'
> >>
> >> I'd like to verify if this change in behavior is intended. If yes, is
> >>
> >> char display[][5] = {"Wait", "Load", "Paus"};
> >>
> >> the expected fix? Or is there another, more elegant way?
> >>
> >> Thanks in advance,
> >> Oliver
> >
> > _______________________________________________________
> > cc6...@li...
> > https://lists.sourceforge.net/lists/listinfo/cc65-devel
--
http://hitmen.eu http://rr.pokefinder.org
http://vice-emu.sourceforge.net http://magicdisk.untergrund.net
"Virtual" means never knowing where your next byte is coming from.
|
|
From: Mark J. R. <mar...@gm...> - 2026-02-11 19:37:44
|
> char display[][] = {"Wait", "Load", "Paus"};
AFAIK that's never been legal in C. If cc65 used to allow it, I'd say that
was a bug that has now been fixed?
On Wed, Feb 11, 2026 at 12:28 PM Oliver Schmidt via cc65-devel <
cc6...@li...> wrote:
> Hi,
>
> Does really nobody have anything to say about my question(s) below?
>
> Wondering,
> Oliver
>
>
> On Tue, Jan 27, 2026 at 1:18 AM Oliver Schmidt <ol...@we...> wrote:
>
>> Hi,
>>
>> the line
>>
>> char display[][] = {"Wait", "Load", "Paus"};
>>
>> used to compile. Now I get
>>
>> Array of incomplete element type 'char[]'
>>
>> I'd like to verify if this change in behavior is intended. If yes, is
>>
>> char display[][5] = {"Wait", "Load", "Paus"};
>>
>> the expected fix? Or is there another, more elegant way?
>>
>> Thanks in advance,
>> Oliver
>>
> _______________________________________________________
> cc6...@li...
> https://lists.sourceforge.net/lists/listinfo/cc65-devel
>
--
Mark J. Reed <mar...@gm...>
|
|
From: Oliver S. <ol...@we...> - 2026-02-11 17:28:15
|
Hi,
Does really nobody have anything to say about my question(s) below?
Wondering,
Oliver
On Tue, Jan 27, 2026 at 1:18 AM Oliver Schmidt <ol...@we...> wrote:
> Hi,
>
> the line
>
> char display[][] = {"Wait", "Load", "Paus"};
>
> used to compile. Now I get
>
> Array of incomplete element type 'char[]'
>
> I'd like to verify if this change in behavior is intended. If yes, is
>
> char display[][5] = {"Wait", "Load", "Paus"};
>
> the expected fix? Or is there another, more elegant way?
>
> Thanks in advance,
> Oliver
>
|
|
From: <gr...@gm...> - 2026-02-11 17:24:18
|
Am Mittwoch, 11. Februar 2026, 17:59:00 Mitteleuropäische Normalzeit schrieb Physiker via cc65-devel: > How can I use this include file if it contains duplicates? E.g OPC_NOP_abx Well, you can't - the dupes need to be commented out of course (like other dupes already are). I guess "6502X" isn't used in the tests, so noone noticed this before :) (It would be great if you could fix that file accordingly and make a PR) -- http://hitmen.eu http://rr.pokefinder.org http://vice-emu.sourceforge.net http://magicdisk.untergrund.net When animal-rights activists and right-to-life protesters are marching outside your laboratory, then you know you've definitely made progress in your artificial life research. <Donald A. Smith> |
|
From: Physiker <vor...@gm...> - 2026-02-11 16:59:12
|
How can I use this include file if it contains duplicates? E.g OPC_NOP_abx |
|
From: Oliver S. <ol...@we...> - 2026-01-27 00:19:01
|
Hi,
the line
char display[][] = {"Wait", "Load", "Paus"};
used to compile. Now I get
Array of incomplete element type 'char[]'
I'd like to verify if this change in behavior is intended. If yes, is
char display[][5] = {"Wait", "Load", "Paus"};
the expected fix? Or is there another, more elegant way?
Thanks in advance,
Oliver
|
|
From: <gr...@gm...> - 2026-01-24 14:03:46
|
Am Samstag, 24. Januar 2026, 11:22:56 Mitteleuropäische Normalzeit schrieb
Physiker via cc65-devel:
> Die Fehlermeldung verschwindet, wenn ich den Aufruf über den function ptr
> auskommentiere.
>
> Bsp.:
>
> typedef int fnc_t(int);
> void myfunc(fnc_t *func){
> ...
> (*func)(10);
> ...
> }
Please make a ticket on github
--
http://hitmen.eu http://rr.pokefinder.org
http://vice-emu.sourceforge.net http://magicdisk.untergrund.net
When Im working I code in whatever language is required, Ive even had to code
on a Mac.
<radiantx>
|
|
From: Physiker <vor...@gm...> - 2026-01-24 10:23:13
|
Die Fehlermeldung verschwindet, wenn ich den Aufruf über den function ptr auskommentiere.
Bsp.:
typedef int fnc_t(int);
void myfunc(fnc_t *func){
...
(*func)(10);
...
}
|
|
From: Vorsichtphysiker <vor...@gm...> - 2025-09-29 19:11:49
|
I just want to tell that I will open an issue, soon. I found out several things. In short: The CA65 assembler does not distinguish between parentheses in the macro parameter list and parentheses as part of expressions. The white space between the macro name and replace expression does not terminate an eventual parameter list. The assembler just takes the first pair of parentheses as the start/end of the macro parameters. This is broken at least since january 2022. Macros containing a parameter list did not show the issue, because then the assembler considers eventual following parentheses as part of the replace expression and not part of the parameter list. This enabled a simple workaroundm so that one could define macros with parentheses in the expression part, anyway. If you just put an empty pair of parenthesis directly after the macro name, the following parentheses were also considered as part of the replace expression and not part of the parameter list like in the case above. Due to a code change in july this year, the workaround is not working anymore. This is the reason I just realized the problem now and not already some years ago. I will provide a bisect log, my testcode and my conclusion. However, it needs a little time to write all that down into the ticket. So I beg for a little patience. This is a typical example of the manifestation of a long existing bug due to some apparently unrelated code change. Yours Monte Am 28.09.25 um 13:17 schrieb Physiker via cc65-devel: > I would like to tell the last version I know which worked, but the > hash before the pull scrolled out of my terminal buffer. So I realized > too late to give you that info. Hope I can add later on. > > 28.09.2025 13:04:13 groepaz--- via cc65-devel > <cc6...@li...>: > > Am Sonntag, 28. September 2025, 12:11:00 Mitteleuropäische > Sommerzeit schrieb > Stefan via cc65-devel: > > I recently pulled my cc65 checkout, recompiled, installed and > then realized > that some CA65 macros I use for years don't work anymore. > Version: cc65 > V2.19 - Git 93c1b659e > > > It seems that parentheses in arguments are not avaluated > correctly, anymore. > However, parentheses in parameter definitions seem to work but > not the same > as before. > Here is an isolated testcase: > > (1) .define foo 100 > (2) .define bar foo+200 > (3) .define baz (foo+200) > (4) .define x1() foo+200 > (5) .define x2() (foo+200) > (6) .define x3(par) foo+par > > (1), (2) are unproblematic, (6) also. > I can understand that (4) and (5) are conercases, anyway they > worked before. > (3) does not work too. > > Here is the assembler output: > > ca65 define_test.s > Assembling include file #3 > define_test.s:11: Error: Expected a parameter name but found > ‘100’ > define_test.s:9: Note: Expanded from macro here > define_test.s:11: Error: Expected ‘)’ but found ‘100’ > define_test.s:9: Note: Expanded from macro here > define_test.s:12: Error: Expected a parameter name but found ‘)’ > define_test.s:13: Error: Expected a parameter name but found ‘)’ > > > Please open a bug ticket - and provide a complete test case for > each :) > Without one its not really possible to tell what is happening, and > why. > > -- > > http://hitmen.eu http://rr.pokefinder.org > http://vice-emu.sourceforge.net http://magicdisk.untergrund.net > > Digital files cannot be made uncopyable, any more than water can > be made not > wet. > > > > > > > _______________________________________________________ > cc6...@li... > https://lists.sourceforge.net/lists/listinfo/cc65-devel > > > > _______________________________________________________ > cc6...@li... > https://lists.sourceforge.net/lists/listinfo/cc65-devel |
|
From: Physiker <vor...@gm...> - 2025-09-28 11:17:58
|
I would like to tell the last version I know which worked, but the hash before the pull scrolled out of my terminal buffer. So I realized too late to give you that info. Hope I can add later on. 28.09.2025 13:04:13 groepaz--- via cc65-devel <cc6...@li...>: > Am Sonntag, 28. September 2025, 12:11:00 Mitteleuropäische Sommerzeit schrieb > Stefan via cc65-devel: >> I recently pulled my cc65 checkout, recompiled, installed and then realized >> that some CA65 macros I use for years don't work anymore. Version: cc65 >> V2.19 - Git 93c1b659e >> >> >> It seems that parentheses in arguments are not avaluated correctly, anymore. >> However, parentheses in parameter definitions seem to work but not the same >> as before. >> Here is an isolated testcase: >> >> (1) .define foo 100 >> (2) .define bar foo+200 >> (3) .define baz (foo+200) >> (4) .define x1() foo+200 >> (5) .define x2() (foo+200) >> (6) .define x3(par) foo+par >> >> (1), (2) are unproblematic, (6) also. >> I can understand that (4) and (5) are conercases, anyway they worked before. >> (3) does not work too. >> >> Here is the assembler output: >> >> ca65 define_test.s >> Assembling include file #3 >> define_test.s:11: Error: Expected a parameter name but found ‘100’ >> define_test.s:9: Note: Expanded from macro here >> define_test.s:11: Error: Expected ‘)’ but found ‘100’ >> define_test.s:9: Note: Expanded from macro here >> define_test.s:12: Error: Expected a parameter name but found ‘)’ >> define_test.s:13: Error: Expected a parameter name but found ‘)’ > > Please open a bug ticket - and provide a complete test case for each :) > Without one its not really possible to tell what is happening, and why. > > -- > > http://hitmen.eu http://rr.pokefinder.org > http://vice-emu.sourceforge.net http://magicdisk.untergrund.net > > Digital files cannot be made uncopyable, any more than water can be made not > wet. > > > > > > > _______________________________________________________ > cc6...@li... > https://lists.sourceforge.net/lists/listinfo/cc65-devel |
|
From: <gr...@gm...> - 2025-09-28 11:03:25
|
Am Sonntag, 28. September 2025, 12:11:00 Mitteleuropäische Sommerzeit schrieb Stefan via cc65-devel: > I recently pulled my cc65 checkout, recompiled, installed and then realized > that some CA65 macros I use for years don't work anymore. Version: cc65 > V2.19 - Git 93c1b659e > > > It seems that parentheses in arguments are not avaluated correctly, anymore. > However, parentheses in parameter definitions seem to work but not the same > as before. > Here is an isolated testcase: > > (1) .define foo 100 > (2) .define bar foo+200 > (3) .define baz (foo+200) > (4) .define x1() foo+200 > (5) .define x2() (foo+200) > (6) .define x3(par) foo+par > > (1), (2) are unproblematic, (6) also. > I can understand that (4) and (5) are conercases, anyway they worked before. > (3) does not work too. > > Here is the assembler output: > > ca65 define_test.s > Assembling include file #3 > define_test.s:11: Error: Expected a parameter name but found ‘100’ > define_test.s:9: Note: Expanded from macro here > define_test.s:11: Error: Expected ‘)’ but found ‘100’ > define_test.s:9: Note: Expanded from macro here > define_test.s:12: Error: Expected a parameter name but found ‘)’ > define_test.s:13: Error: Expected a parameter name but found ‘)’ Please open a bug ticket - and provide a complete test case for each :) Without one its not really possible to tell what is happening, and why. -- http://hitmen.eu http://rr.pokefinder.org http://vice-emu.sourceforge.net http://magicdisk.untergrund.net Digital files cannot be made uncopyable, any more than water can be made not wet. |
|
From: Stefan <vor...@gm...> - 2025-09-28 10:11:14
|
<div style="font-family: verdana;font-size: 12.0px;color: rgb(0,0,0);"><span style="background-color: rgb(255,255,255);">I recently pulled my cc65 checkout, recompiled, installed and then realized that some CA65 macros I use for years don't work anymore. Version: cc65 V2.19 - Git 93c1b659e<br></span></div> <div style="font-family: verdana;font-size: 12.0px;color: rgb(0,0,0);"><span style="background-color: rgb(255,255,255);"> </span></div> <div style="font-family: verdana;font-size: 12.0px;color: rgb(0,0,0);"><span style="background-color: rgb(255,255,255);">It seems that parentheses in arguments are not avaluated correctly, anymore.</span></div> <div style="font-family: verdana;font-size: 12.0px;color: rgb(0,0,0);"><span style="background-color: rgb(255,255,255);">However, parentheses in parameter definitions seem to work but not the same as before.</span></div> <div style="font-family: verdana;font-size: 12.0px;color: rgb(0,0,0);"><span style="background-color: rgb(255,255,255);"> </span></div> <div style="font-family: verdana;font-size: 12.0px;color: rgb(0,0,0);"><span style="background-color: rgb(255,255,255);">Here is an isolated testcase:</span></div> <div style="font-family: verdana;font-size: 12.0px;color: rgb(0,0,0);"><span style="background-color: rgb(255,255,255);"> </span></div> <div style="font-family: verdana;font-size: 12.0px;color: rgb(0,0,0);"><span style="background-color: rgb(255,255,255);">(1) .define foo 100<br>(2) .define bar foo+200<br>(3) .define baz (foo+200)<br>(4) .define x1() foo+200<br>(5) .define x2() (foo+200)<br>(6) .define x3(par) foo+par</span></div> <div style="font-family: verdana;font-size: 12.0px;color: rgb(0,0,0);"><span style="background-color: rgb(255,255,255);"> </span></div> <div style="font-family: verdana;font-size: 12.0px;color: rgb(0,0,0);"><span style="background-color: rgb(255,255,255);">(1), (2) are unproblematic, (6) also.</span></div> <div style="font-family: verdana;font-size: 12.0px;color: rgb(0,0,0);"><span style="background-color: rgb(255,255,255);">I can understand that (4) and (5) are conercases, anyway they worked before.</span></div> <div style="font-family: verdana;font-size: 12.0px;color: rgb(0,0,0);"><span style="background-color: rgb(255,255,255);">(3) does not work too.</span></div> <div style="font-family: verdana;font-size: 12.0px;color: rgb(0,0,0);"><span style="background-color: rgb(255,255,255);"> </span></div> <div style="font-family: verdana;font-size: 12.0px;color: rgb(0,0,0);"><span style="background-color: rgb(255,255,255);">Here is the assembler output:</span></div> <div style="font-family: verdana;font-size: 12.0px;color: rgb(0,0,0);"><span style="background-color: rgb(255,255,255);"> </span></div> <div style="font-family: verdana;font-size: 12.0px;color: rgb(0,0,0);"><span style="background-color: rgb(255,255,255);">ca65 define_test.s <br>Assembling include file #3<br>define_test.s:11: Error: Expected a parameter name but found ‘100’<br>define_test.s:9: Note: Expanded from macro here<br>define_test.s:11: Error: Expected ‘)’ but found ‘100’<br>define_test.s:9: Note: Expanded from macro here<br>define_test.s:12: Error: Expected a parameter name but found ‘)’<br>define_test.s:13: Error: Expected a parameter name but found ‘)’<br></span></div> <div style="font-family: verdana;font-size: 12.0px;color: rgb(0,0,0);"><span style="background-color: rgb(255,255,255);"> </span></div> <div style="font-family: verdana;font-size: 12.0px;color: rgb(0,0,0);"><span style="background-color: rgb(255,255,255);">Greetings Monte</span></div> <div style="font-family: verdana;font-size: 12.0px;color: rgb(0,0,0);"><span style="background-color: rgb(255,255,255);"> </span></div> |