Re: llvmpipe - was: Re: Updated Debian Ports installation images 2025-12-06

On 21/4/26 3:12 pm, John Paul Adrian Glaubitz wrote:
> The LLVM error messages don't mean that the binaries were built with LLVM/Clang
> but that LLVM is being used for software-rendering. The underlying software is
> called llvmpipe.
>
> See:https://docs.mesa3d.org/drivers/llvmpipe.html
>
> Since LLVM does not support all features on 32-bit PowerPC, llvmpipe has some
> issues on this architecture.

Thanks for the info. See, I found the same thing when I first looked up 
this "LLVM error relocation type not implemented" message a while back. 
Most results returned related to Clang and/or linker. I'm not trying to 
compile, I'm running a binary, so that's too late. Even now, most 
results talk about linkers and not graphics even when using "llvmpipe" 
as a search term. Clearly, the error is too vague and LLVMPipe needs to 
mention itself, and how it relates to graphics. Then it would be more 
obvious. Typical computer errors. And something like "LLVM" when 
unpacked describes a hardware emulator to me, so these people should 
come up with better names. :-?

So, most G4 Mac people using newest graphic cards without full drivers 
would be affected, while G5 people should be fine when needing a 
fallback. This would explain why mine breaks. My kernel is 64 bit but I 
found out I had a 32 bit userland, so must have accidentally picked the 
wrong install image. My Debian 64 has no such issues, if just a bit slow 
using software rendering, since I don't have a HD6000 or 3d supported 
card plugged in.

Slightly off topic, but at this point, I wonder if AI could be used for 
good use to scrape through the AMD driver sources and locate endian 
issues. Then suggest how to fix. I'm thinking AI could be good for 
finding endian issues in frequently updated code that's not practical to 
fix by hand for every software update.