Lunderberg commented on a change in pull request #7867:
URL: https://github.com/apache/tvm/pull/7867#discussion_r615947927
##########
File path: python/tvm/_ffi/runtime_ctypes.py
##########
@@ -262,6 +264,35 @@ def max_thread_dimensions(self):
"""
return json.loads(self._GetDeviceAttr(self.device_type,
self.device_id, 8))
+ @property
+ def api_version(self):
+ """Returns version number of the SDK used to compile TVM.
+
+ For example, CUDA_VERSION for cuda or VK_HEADER_VERSION for
+ Vulkan.
+
+ Returns
+ -------
+ version : int
+ The version of the SDK
+
+ """
+ return self._GetDeviceAttr(self.device_type, self.device_id, 12)
+
+ @property
+ def driver_version(self):
+ """Returns version number of the driver
+
+ Returns driver vendor's internal version number.
+ (e.g. "450.408.256" for nvidia-driver-450)
+
+ Returns
+ -------
+ version : str
Review comment:
Certainly, can do. It looks like several of the other parameters also
return None for some implementations, sometimes for cases where a parameter
doesn't make sense (e.g. warp size for cpu targets), or API limitations (e.g.
clock speed on vulkan). I'll see if I can flesh out some of the
implementations, and add documentation for each parameter on when they return
None.
--
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.
For queries about this service, please contact Infrastructure at:
[email protected]
