[TF2XLA] Do not promote autoclustering on compilation failure

When the explicit compilation scope fails, the right solution is to try to compile less (or to file a bug).
Suggesting autoclustering in these cases is rarely helpful.

PiperOrigin-RevId: 307089449
Change-Id: I082e9303cd3cff438ab6d56beaf39179e8632cc6

Author: cheshire

tensorflow/compiler/jit/kernels/xla_ops.cc

@@ -358,13 +358,6 @@ void XlaLocalLaunchBase::Compute(OpKernelContext* ctx) {
         ctx, function_, /*has_ref_vars=*/has_ref_vars_, platform_info_,
         resources_, constants_, /*lazy=*/false, &client, &variables, &kernel,
         &executable);
-    if (!s.ok() && (platform_info_.device_type().type_string() == DEVICE_CPU ||
-                    platform_info_.device_type().type_string() == DEVICE_GPU)) {
-      // Suggest auto jit if the failure was with GPU or CPU.
-      errors::AppendToMessage(&s,
-                              xla::status_macros::kPossibleAutoJitAlternative);
-    }
-
     OP_REQUIRES_OK(ctx, s);
   }