Add version 11.1 in finding CUDA libdevice (apache#7033)

* Add CUDA 11.1 libdevice

Maybe we should have a >= check instead.
I also added a fallback to detect the version if version.txt is
missing. Calling nvcc for this has been inspired by what PyTorch
does when compiling extension modules.

python/tvm/contrib/nvcc.py

@@ -148,8 +148,20 @@ def get_cuda_version(cuda_path):
         with open(version_file_path) as f:
             version_str = f.readline().replace("\n", "").replace("\r", "")
             return float(version_str.split(" ")[2][:2])
-    except:
-        raise RuntimeError("Cannot read cuda version file")
+    except FileNotFoundError:
+        pass
+
+    cmd = [os.path.join(cuda_path, "bin", "nvcc"), "--version"]
+    proc = subprocess.Popen(cmd, stdout=subprocess.PIPE, stderr=subprocess.STDOUT)
+    (out, _) = proc.communicate()
+    out = py_str(out)
+    if proc.returncode == 0:
+        release_line = [l for l in out.split("\n") if "release" in l][0]
+        release_fields = [s.strip() for s in release_line.split(",")]
+        release_version = [f[1:] for f in release_fields if f.startswith("V")][0]
+        major_minor = ".".join(release_version.split(".")[:2])
+        return float(major_minor)
+    raise RuntimeError("Cannot read cuda version file")
 
 
 @tvm._ffi.register_func("tvm_callback_libdevice_path")
@@ -174,7 +186,7 @@ def find_libdevice_path(arch):
     selected_ver = 0
     selected_path = None
     cuda_ver = get_cuda_version(cuda_path)
-    if cuda_ver in (9.0, 9.1, 10.0, 10.1, 10.2, 11.0):
+    if cuda_ver in (9.0, 9.1, 10.0, 10.1, 10.2, 11.0, 11.1):
         path = os.path.join(lib_path, "libdevice.10.bc")
     else:
         for fn in os.listdir(lib_path):
@@ -219,6 +231,7 @@ def parse_compute_version(compute_version):
         minor = int(split_ver[1])
         return major, minor
     except (IndexError, ValueError) as err:
+        # pylint: disable=raise-missing-from
         raise RuntimeError("Compute version parsing error: " + str(err))