Add gfx950 (MI355X/CDNA4) to is_cdna() for correct Triton num_warps

MI355X (gfx950) has the same 1024-thread workgroup limit as MI300X (gfx942),
but was missing from is_cdna(), causing all Triton kernels to use num_warps=32
(2048 threads) instead of 16 (1024 threads), resulting in OutOfResources crash.

Also includes ROCm GPT-OSS BF16 routing and dequant buffer dtype fix from PR unslothai#4021
by @danielhanchen, cherry-picked for MI355X validation.

Tested on: 8x AMD Instinct MI355X (gfx950), ROCm 7.1
- Vision RL GRPO (Qwen2.5-VL-7B): 5/5 steps
- Code RL GRPO (gpt-oss-20b BF16): 20/20 steps
- gpt-oss-120b GRPO: 5/5 steps (B200 OOM'd on this)
- MoE expert LoRA + save_pretrained_merged: success

Author: GoldenGrapeGentleman

unsloth/kernels/utils.py

@@ -82,6 +82,7 @@ def is_cdna():
         "gfx940",
         "gfx941",
         "gfx942",
+        "gfx950",  # CDNA4 (MI350/MI355X)
     )