Click to expand!

Issue Type

Build/Install

Source

source

Tensorflow Version

tf 2.10

Custom Code

No

OS Platform and Distribution

MacOs ventura 13.0

Mobile device

• No response*

Python version

• No response*

Bazel version

5.3.2

GCC/Compiler version

clang 14.0 (Xcode shipped)

CUDA/cuDNN version

• No response*

GPU model and memory

• No response*

Current Behaviour?

Hello !

I am not able to build libtensorflow.so for ios.
I just `./configure` and accepted all defaults.

I can build for my mac:
`bazel build -c opt //tensorflow:libtensorflow.so`
but cross compiling for ios
`bazel build -c opt --config=ios //tensorflow:libtensorflow.so`

gives me this error

`
ERROR: /Users/SX/Documents/soundx-ai/vendor/tensorflow/tensorflow/cc/saved_model/BUILD:319:11: Compiling tensorflow/cc/saved_model/metrics.cc failed: (Aborted): wrapped_clang_pp failed: error executing command external/local_config_cc/wrapped_clang_pp '-D_FORTIFY_SOURCE=1' -fstack-protector -fcolor-diagnostics -Wall -Wthread-safety -Wself-assign -fno-omit-frame-pointer -g0 -O2 -DNDEBUG ... (remaining 32 arguments skipped)
In file included from tensorflow/cc/saved_model/metrics.cc:16:
In file included from ./tensorflow/cc/saved_model/metrics.h:25:
In file included from ./tensorflow/core/lib/monitoring/counter.h:19:
In file included from ./tensorflow/tsl/lib/monitoring/counter.h:85:
In file included from ./tensorflow/tsl/lib/monitoring/collection_registry.h:109:
In file included from ./tensorflow/tsl/lib/monitoring/collected_metrics.h:28:
In file included from ./tensorflow/tsl/lib/monitoring/metric_def.h:24:
In file included from ./tensorflow/tsl/lib/monitoring/types.h:22:
In file included from ./tensorflow/tsl/platform/types.h:21:
In file included from ./tensorflow/tsl/platform/bfloat16.h:20:
./third_party/eigen3/Eigen/Core:1:10: fatal error: 'Eigen/Core' file not found
#include "Eigen/Core"

I have a eigen3 lib installed with brew, in my /opt/homebrew/Cellar/eigen/3.4.0_1/include/eigen3/Eigen/Core , but I did not get how I could give this information to bazel. I'm not able to play with the path directly.
Any clue to solve this problem ?
Thank you!

### Standalone code to reproduce the issue

```shell
bazel build -c opt --config=ios //tensorflow:libtensorflow.so

Relevant log output

• No response*