[BugFix][Ops] chunk_scaled_dot_kkt_fwd_kernel recompile issues (vllm-project#11577)

### What this PR does / why we need it?

This PR fixes recompilation issues in the
`chunk_scaled_dot_kkt_fwd_kernel` Triton kernel. The parameters
`bh_step`, `task_num`, and `num_core` were previously defined as
`tl.constexpr`, which caused Triton to recompile the kernel whenever
these values changed. By changing them to regular arguments, we avoid
recompilation when these values change.

### Does this PR introduce _any_ user-facing change?

No.

### How was this patch tested?


- vLLM version: v0.23.0
- vLLM main:
vllm-project/vllm@1f486d9

---------

Signed-off-by: Ascendyh <hw7osiris@outlook.com>

Author: OsirisDuan

vllm_ascend/ops/triton/fla/chunk_scaled_dot_kkt.py

@@ -23,7 +23,7 @@
         "USE_G": lambda args: args["g_cumsum"] is not None,
     }
 )
-@triton.jit(do_not_specialize=["T", "B"])
+@triton.jit(do_not_specialize=["T", "B", "bh_step", "task_num", "num_core"])
 def chunk_scaled_dot_kkt_fwd_kernel(
     k,
     beta,  # [H, B, T]
@@ -33,16 +33,16 @@ def chunk_scaled_dot_kkt_fwd_kernel(
     chunk_indices,
     T,
     B,
+    bh_step,
+    task_num,
+    num_core,
     H: tl.constexpr,
     Hg: tl.constexpr,
     K: tl.constexpr,
     BT: tl.constexpr,
     BK: tl.constexpr,
     IS_VARLEN: tl.constexpr,
     USE_G: tl.constexpr,
-    bh_step: tl.constexpr,
-    task_num: tl.constexpr,
-    num_core: tl.constexpr,
 ):
     bt_stride = B * T
     core_id = tl.program_id(0)