Add hacky M2 Max support

[initialed85]

Caveat: I have absolutely no idea what I'm doing

Goal: Trying to get this working on my Macbook Pro 64GB (M2 Max)

All I've done in here:

• Drop the GPU family guard from 9 to 8
• Ask gpt-6-luna to speculation on a solution to the subsequent startup error (the change to kRuntimeOverheadReserveBytes

Validation:

make -j 12

sudo sysctl iogpu.wired_limit_mb=60416

./slipstream serve --model nitinpanj/Swift-Qwen3.8-Flash-Next-Q4_0-Q8out-v3-GGUF --max-memory 59G --max-context 131072 --no-webui --host 0.0.0.0 --port 8088

Then just used Pi to fire some prompts at it (had it poking around my homelab Kube cluster for about 30 minutes; seemed to perform well).

If your take is "buddy you're an idiot this isn't sustainable for x, y and z" reasons I will absolutely not be offended because I am well out of my depth in this field.