-
-
Notifications
You must be signed in to change notification settings - Fork 2k
New issue
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? Sign in to your account
macOS Meta Issue #5361
Comments
The newest version uses AA_EnableHighDpiScaling. Please test if this is the correct setting or if it needs the variable QT_AUTO_SCREEN_SCALE_FACTOR set manually. |
Well, I do not have a HDPI screen nearby, but rpcs3looks fine on a normal screen on macOS, and I believe the change is sane. |
@hcorion I just tried building RPCS3 afresh on macOS and running some simple things, like Scogger HD (installed from .pkg) and getting PPU unhandled exceptions right away on start. Is this expected or related to some of the things listed here? If not, I'll investigate further and file issues separately. |
Hi, any updates on this ? |
hcorion is currently busy irl with the education and isn't available, generally speaking. |
Hello guys, Have you got any update on the binary status for macOS? Vulkan seems to be "working" now (well at least, cmake can find it) My script 😛: Tried with clang as well (got some different errors 😝 ) I am not really an expert in Cpp and cmake stuff but with some guidance, I may be able to make it work. |
(this one is from a 3rd party lib) |
There has been no progress, in fact the Apple code paths may eventually be purged from the code base as Apple moves away from x86 |
macOS's support could be provided by using MoltenVK. Additionally, the Zink tool can be added for macOS and Linux. This shouldn't be very hard and would add macOS support for this emulator and would improve the Linux support. |
Zink is an OpenGL implementation over Vulkan, but since we have a decently maintained native OpenGL renderer, I'm fully certain that a Zink backend would not give any benefits, especially on Mac, where Zink targets MoltenVK. Edit: I forgot that Mac doesn't even support OpenGL 4.3, which is our minimum requirement. So in a way, I suppose Zink would make sense, although no idea why it would be necessarily desirable. As for MoltenVK, I remember some technical limitations being blockers, but #4996 seemed to have gone peachy, so I'm unsure where it stands now. Maybe someone with an x86 Mac can tell us if the emulator even compiles these days? Please also consider that Apple is transitioning away to ARM, full force. Mac support would need not just compliance with MoltenVK, but also adjustments made to the recompilers, as despite them technically using LLVM as a backend, as far as I'm aware, lots of x86isms are present in it. These, along with the highly limited access to Mac hardware, the minimal developer interest, and general hostility Apple seems to be showing against emulators within some spaces of their ecosystem makes them quite a hard sell at the moment for anyone involved. |
Tried compiling
Related changes seem to be from 2021-05-22, which isn't too long ago (#4996 is three years ago now) |
Im tried compiling
|
Could it be that you are using the wrong compiler? |
no, Im using |
Try building the rmacs3 branch of my fork with export Qt5_DIR="/usr/local/Cellar/qt@5/5.15.2_1/lib/cmake/Qt5"
export CXX=g++-11
export CC=gcc-11
cmake -DUSE_DISCORD_RPC=OFF -DUSE_VULKAN=ON -DUSE_ALSA=OFF -DUSE_PULSE=OFF ../rpcs3/ && make -j8 (Vulkan SDK, molten-vk, and vulkan-headers are required for Vulkan support) LLVM still doesn't work, though (see #11222). |
I would love to see a HomeBrew Formula :) |
Thanks, so Im go to next error =)
something like that?
|
In 3rdparty/libusb/libusb/libusb/os/darwin_usb.c:2151, remove Also, if you get an error about dispatch_block_t being undefined, open 3rdparty/hidapi/hidapi/mac/hid.c and add |
So, go next =)
|
That should be comment out in my fork. Are you sure you checked out the rmacs3 branch? |
Yeah, 🤦 Im on master branch. |
So, when I'm about to finish, I get a QT error:
This like seems as it already fixed, but in qt 6.1. Any ideas? |
You need to replace those Qt headers (see #11222) |
Thank you very much! Im dont know what I must doing to resolve next (maybe last) problem:
|
Oh, I forgot that you need to rebuild ffmpeg, replacing the existing .a files (no modifications to the source code are needed). |
Hmm...now I have arch error by hidapi
|
@f2re Did you manage to build it? |
@gabjauf no, Im trying to fix hid.c problem. |
See the edit in the comment I linked |
I think some of the issues mentioned in the main issue is fixed. Maybe we can close this issue and @nastys can create a new meta issue to track the work that is still pending. |
Here is a list of the issues that need to get fixed before we can start providing binaries for MacOS:
Some of the issues that can be solved afterwards:
The text was updated successfully, but these errors were encountered: