fix(sgl-kernel): support CUDA 13 runtime preloading for DGX Spark (#18747)

Co-authored-by: Claude Opus 4.6 <noreply@anthropic.com>

sgl-kernel/python/sgl_kernel/load_utils.py

@@ -166,6 +166,14 @@ def _load_architecture_specific_ops():
     )
 
     # All attempts failed
+    cuda_version = torch.version.cuda
+    if cuda_version and cuda_version.startswith("13"):
+        install_hint = (
+            "pip install sgl-kernel --index-url https://docs.sglang.ai/whl/cu130/"
+        )
+    else:
+        install_hint = "pip install --upgrade sgl_kernel"
+
     error_msg = f"""
 [sgl_kernel] CRITICAL: Could not load any common_ops library!
 
@@ -177,9 +185,10 @@ Attempted locations:
 GPU Info:
 - Compute capability: {compute_capability}
 - Expected variant: {variant_name}
+- CUDA version: {cuda_version}
 
 Please ensure sgl_kernel is properly installed with:
-pip install --upgrade sgl_kernel
+{install_hint}
 
 Error details from previous import attempts:
 {attempt_error_msg}
@@ -205,7 +214,7 @@ def _find_cuda_home():
 
 
 def _preload_cuda_library():
-    """Preload the CUDA runtime library to help avoid 'libcudart.so.12 not found' issues."""
+    """Preload the CUDA runtime library to help avoid 'libcudart.so not found' issues."""
     cuda_home = Path(_find_cuda_home())
 
     candidate_dirs = [
@@ -217,16 +226,22 @@ def _preload_cuda_library():
         Path("/usr/lib"),
     ]
 
+    # Determine CUDA major version to try the matching library first.
+    # On CUDA 13 systems (e.g., DGX Spark), only libcudart.so.13 exists.
+    cuda_major = torch.version.cuda.split(".")[0] if torch.version.cuda else "12"
+    lib_versions = list(dict.fromkeys([cuda_major, "13", "12"]))
+
     for base in candidate_dirs:
-        candidate = base / "libcudart.so.12"
-        if candidate.exists():
-            try:
-                cuda_runtime_lib = candidate.resolve()
-                ctypes.CDLL(str(cuda_runtime_lib), mode=ctypes.RTLD_GLOBAL)
-                logger.debug(f"Preloaded CUDA runtime under {cuda_runtime_lib}")
-                return
-            except Exception as e:
-                logger.debug(f"Failed to load {candidate}: {e}")
-                continue
+        for lib_version in lib_versions:
+            candidate = base / f"libcudart.so.{lib_version}"
+            if candidate.exists():
+                try:
+                    cuda_runtime_lib = candidate.resolve()
+                    ctypes.CDLL(str(cuda_runtime_lib), mode=ctypes.RTLD_GLOBAL)
+                    logger.debug(f"Preloaded CUDA runtime under {cuda_runtime_lib}")
+                    return
+                except Exception as e:
+                    logger.debug(f"Failed to load {candidate}: {e}")
+                    continue
 
     logger.debug("[sgl_kernel] Could not preload CUDA runtime library")