There are still some other issues not addressed here, but it's a start.
Workarounds for false-positive reports:
- `RasterizerAccelerated`: Put a gigantic array behind a `unique_ptr`,
because UBSan has a [hardcoded limit](https://stackoverflow.com/questions/64531383/c-runtime-error-using-fsanitize-undefined-object-has-a-possibly-invalid-vp)
of how big it thinks objects can be, specifically when dealing with
offset-to-top values used with multiple inheritance. Hopefully this
doesn't have a performance impact.
- `QueryCacheBase::QueryCacheBase`: Avoid an operation that UBSan thinks
is UB even though it at least arguably isn't. See the link in the
comment for more information.
Fixes for correct reports:
- `PageTable`, `Memory`: Use `uintptr_t` values instead of pointers to
avoid UB from pointer overflow (when pointer arithmetic wraps around
the address space).
- `KScheduler::Reload`: `thread->GetOwnerProcess()` can be `nullptr`;
avoid calling methods on it in this case. (The existing code returns
a garbage reference to a field, which is then passed into
`LoadWatchpointArray`, and apparently it's never used, so it's
harmless in practice but still triggers UBSan.)
- `KAutoObject::Close`: This function calls `this->Destroy()`, which
overwrites the beginning of the object with junk (specifically a free
list pointer). Then it calls `this->UnregisterWithKernel()`. UBSan
complains about a type mismatch because the vtable has been
overwritten, and I believe this is indeed UB. `UnregisterWithKernel`
also loads `m_kernel` from the 'freed' object, which seems to be
technically safe (the overwriting doesn't extend as far as that
field), but seems dubious. Switch to a `static` method and load
`m_kernel` in advance.
Allow for displaying options in the home options that are disabled with messages that explain why they are disabled.
This includes reasoning for the GPU driver installation button.
MinGW's strftime implementation does not work and cannot be used to
determine the time zone. Besides that, the string operations are
actually unnecessary since we can get the offset from
std::localtime.
Compare localtime to gmtime to find the zone offset on all platforms.
While this can be convenient in some scenarios, this will be a big problem for users trying to sideload different APK versions. If they forget the last one they had installed, they could have problems installing a new copy.
This allows to share the mappings between Nintendo and non-Nintendo controllers.
Breaks the controller configuration for existing users who are using a Nintendo controller.
(Documentation of the hint 92b3c53c92/include/SDL_hints.h (L512-L532))
Even though it compiles and runs fine on the latest Windows versions,
older LTSC builds will crash due to lacking support somewhere in the OS.
For now just disable it for MSVC until either Microsoft fixes this or we
no longer support 1809 LTSC.
Now within the Input Overlay file, there is a version that will determine when the overlay will be reset. This is intended for breaking changes like the ones we had with the additions of percentage based layouts or the addition of foldable/portrait layouts. This also includes versions for each individual layout so we don't have to reset every layout if only one is broken.
Additionally, this includes new L3/R3 buttons.
lat9nq