Update the docstring with more details on possible values of p keyword in dpnp.linalg.cond

Author: antonwolfy

dpnp/linalg/dpnp_iface_linalg.py

@@ -180,8 +180,18 @@ def cond(x, p=None):
     x : {dpnp.ndarray, usm_ndarray}
         The matrix whose condition number is sought.
     p : {None, 1, -1, 2, -2, inf, -inf, "fro"}, optional
-        Order of the norm used in the condition number computation.
-        ``inf`` means the `dpnp.inf` object, and the Frobenius norm is
+        Order of the norm used in the condition number computation:
+
+            - None: 2-norm.
+            - "fro": Frobenius norm.
+            - inf: max(sum(abs(x), axis=1)).
+            - -inf: min(sum(abs(x), axis=1)).
+            - 1: max(sum(abs(x), axis=0)).
+            - -1: min(sum(abs(x), axis=0)).
+            - 2: 2-norm (largest singular value).
+            - -2: smallest singular value.
+
+        ``inf`` means the :obj:`dpnp.inf` object, and the Frobenius norm is
         the root-of-sum-of-squares norm.
 
         Default: ``None``.