Re: [hatari-devel] Upstream / downstream code in Hatari

[ Thread Index | Date Index | More lists.tuxfamily.org/hatari-devel Archives ]


Hi,

On 21.8.2025 16.46, Nicolas Pomarède wrote:
Le 21/08/2025 à 15:35, Eero Tamminen a écrit :
Before it's applied, I'd like to know which parts of Hatari are:
- Used as-is
- Modified
- Ignored

In Previous, so that I can document both upstream[1] and downstream[2] project considerations in Hatari "coding.txt".  Andreas?


[1] Nicolas, currently only WinAUE CPU core code is used as-is, i.e. code to that should go through WinUAE, right?

yes, that's the preferred way to do change in the CPU core, especially when it's syntax/compilation warnings, as such changes usually benefit to WinUAE too. This makes it easier later to merge changes from WinUAE into Hatari by limiting the number of possible differences.

There're also some cases where change are more Atari specific (eg bus error cases, timing of vectored vs autovectored interrupts, ...) and adding the changes to WinUAE require more work from Toni because it's hard to test/replicate in Amiga environment. In such cases I sometimes modify CPU core in Hatari directly, while waiting for a possible inclusion in WinUAE (or not if it's really too Atari specific :) )

So rule for "coding.txt" could be something like this:
----
Changes to '*.[ch]' files under src/cpu/ should go through WinUAE upstream project, unless change is inside WINUAE_FOR_HATARI ifdef.
---

	- Eero




Mail converted by MHonArc 2.6.19+ http://listengine.tuxfamily.org/