Use Eigen blas for custom sdpa (#6197)

Summary:
Pull Request resolved: #6197

OpenBlas's implementation is not thread safe. Thus when used within
parallel_for, it produces incorrect output. This has been documented in a few
places like here OpenMathLib/OpenBLAS#1441 and
OpenMathLib/OpenBLAS#2543.

I tried few options to disable openblas's multithreading but none of them
seemed to work.

It is possible that upstream openblas has fixed this. Will validate this by
pulling in latest, but in the meanwhile using eigen_blas to unblock.
ghstack-source-id: 247950073
exported-using-ghexport

Reviewed By: guangy10, nausicaasnow

Differential Revision: D64334733

fbshipit-source-id: a7117233584860c2910f45bfaa8dd8fa41f9f28d

Author: kimishpatel

kernels/optimized/lib_defs.bzl

@@ -144,7 +144,7 @@ def define_libs():
                 (
                     "^android-arm64.*$",
                     [
-                        "fbsource//third-party/openblas:openblas",
+                        "fbsource//arvr/third-party/eigen:eigen3_blas",
                     ],
                 ),
             ],