Andrew Richards

I don't have any ARM9 hardware to test on, unfortunately. But maybe you can do it. I've created a new branch named `test`. It has some fixes, changes, and adds...

Wow, thanks for the in-depth report! That's pretty interesting. I'm really surprised GCC would continue to inline in an unsafe condition -- it's not supposed to do that. There might...

OK, that's kind of interesting. I'm trying to guess at what may be going on. Does your build environment have stuff like stack protection or fortify enabled by default? I'm...

Great job! That must be it. I think the aligned attribute is not doing what it's supposed to be doing here for gcc: > This attribute specifies a minimum alignment...

Sorry, another question: Does the misaligned access occur if `-DLAY_FLOAT` is specified? (This causes float coordinates to be used instead of int16_t.) I'm trying to find a way to easily...

Thanks. Sorry I haven't finished this yet -- I'm traveling right now and don't have access to my normal testing environment (a bunch of VMs.)

I'm back from traveling. I have a way to work around this by reordering two fields and changing the alignment requirement, and it shouldn't add any extra padding. But the...

Do you have a list?

I think that's not too bad. Replace `.background()` with `.window()` to fix most of them.

Thanks for letting me know. I'll take a look when I get around to updating this for Qt6. It relies on the internals of the QPalette data to do a...