soooo.. in OpenCL C builtins like get_global_offset have to return defined values on out of bound accesses, but apparently Intel's driver also gets this wrong. Now I'm considering if I should keep it broken as well, or actually fix it as defined in the spec...
@bashbaug any thoughts on this? Maybe we should add OpenCL CTS tests for this and annoy everybody else with this?
@karolherbst@buzzle I wonder if it would be possible to do something like a “tuned kernel with fallback”, something that is tailored to the specific system where it's booting, but still manages to fall back gracefully to through initramfs, that only gets loaded if the system where it's booting is different from the one it's tailored for.
@karolherbst highly recommended. I'm using it less these days for a number of reasons none of which are related to actual issues with the client, but I highly recommend it for anyone that needs CLI or TUI access to Mastodon. IIRC it also has a decent degree of compatibility with other server software although I haven't had much of an opportunity to test that @ihabunek