Platform detection issues on Apple MacOS #1839
Comments
+1 |
Very similar issue here.
gives a
It would be nice to know the Hashrate of the baseline Mac Pro 2013 GPUs |
Sorry @minodesign we dropped apple support quite some releases ago. It does not make sense for us to maintain a platform with such low usage. |
eh imagined. |
|
@minodesign iMac Pro, for example. Or MacBook pro with AMD GPU. @MariusVanDerWijden That's a shame then. It's not about casual usage, it's about debugging and development of private chains, pool software, etc. |
I was able to replicate this problem. Any solutions out there? |
Done: add Apple support for AMD / OpenCL cards @see ethereum-mining#1839
Same issue here. macOS support is important. |
or delete the binary. |
So is this project only for Windoze and Linux fanbois? |
No. This project is effectively abandoned. |
Please update your README documentation to reflect this. |
Indeed why is the project releasing Darwin builds if support was dropped? |
macOS support is important. |
RIP eth mining on mac os then. Could you please update documentation to reflect? https://github.com/ethereum-mining/ethminer#install |
I understand if the project is taking the stance that Mac OS is no longer supported, but I think @morekid makes a good point that the documentation on the readme.md page should probably be updated then. Also, mind me asking why, if someone has already released a patch, and a pull request, someone can't simply just merge it into the code? Your build system, or whatever is compiling the releases is still building the Darwin binary: |
https://www.youtube.com/watch?v=B18ss3dbRAI Neville Longbuttom: Uh, professor. Why can't we just use magic to get rid of the virus? Severus Snape: Because, Mr. Longbottom, the logic of the wizarding world is held together by duct tape and begins to unravel at the slightest bit of scrutiny. I mean, we loan out incredibly powerful time-travel devices to 13 year olds so they can take extra classes. Nothing makes any sense, and you'd be better off if you didn't think about it. |
Here to say this, it's frustrating to find a workaround when it could be easily implemented. Especially with the GPUs that are now shipping with Macs, or the growing Hackintosh community that offer the same GPUs as their Windows counterparts. Hopefully this gets reconsidered, rather than being met with absolute stubbornness. Appreciate the work, regardless. |
Somebody got ethminer running on a new M1 MacBook with a few minor changes to the codebase: https://blog.yifangu.com/2021/02/26/mining-ethereum-on-a-m1-mac-gpu/ Looks like if help ethminer find your Mac's GPU(s) (M1 or Intel-AMD, I expect), then it spools up and works as intended. |
macbook support is important, at least should not take it down; we can use eGPU; there are many macbook users (especially software engineers and media content workers). |
Thank you for wasting my time with build instructions and binaries for it to not be supported. There's 3 hours I won't be getting back. |
I spend 3days(( |
@jamiecruickshank @prinze77 sorry guys to hear your disappointment but :
|
@AndreaLanfranchi Sounds like we need some better maintainers for this projects. A nice chunky codebase that’ll run just as well on macOS as any *nix platform, held up by some basic GPU detection logic. And a maintainer who seems to have no interest in listening to the community’s appeals— just wants to shout “no” like a child. And this, ladies and gentlemen, is why open source often sucks. |
Platform detection fails with the following message:
Unrecognized platform Apple
To Reproduce
Steps to reproduce the behavior:
Build with ETHASHCL set to on and ETHASHCUDA to off
cmake .. -DETHASHCUDA=OFF -DETHASHCL=ON
cmake --build . --config Release
Try to get a list of OpenCL devices:
./ethminer --list-devices
See error message
Expected behavior
You should see a list of available OpenCL devices.
Environment:
Additional context
Due to incorrect hardware detection, it's not possible to use miner on its purpose. Though I was able to workaround this issue by applying ugly patch:
clinfo output:
clinfo.txt
The text was updated successfully, but these errors were encountered: