Update on "Remove some simple config.dynamic_shapes checks - towards …

…dynamic on by default"

cc penguinwu anijain2305 EikanWang jgong5 Guobing-Chen XiaobingSuper zhuhaozhe blzheng Xia-Weiwen wenzhe-nrv jiayisunx desertfire

[ghstack-poisoned]

update vision commit hash

Update on "Better cudagraph enablement for dynamic_shapes=True with s…

…tatic shapes"

cc penguinwu EikanWang jgong5 Guobing-Chen XiaobingSuper zhuhaozhe blzheng Xia-Weiwen wenzhe-nrv jiayisunx peterbell10 desertfire

[ghstack-poisoned]

Check devices on meta functions that return inputs

[ghstack-poisoned]

Enable test_cpp_wrapper tests on ROCm

Update on "[MPS] Dispatch outer bin edges selection function"


Dispatch the selection function to prevent using `is_mps()` in `Histogram.cpp`.

cc jgong5 mingfeima XiaobingSuper sanchitintel ashokei jingxu10

[ghstack-poisoned]

skip inductor cpu tests instead of cuda

Update on "Enable quantized_max_pool3d"


**Summary**
Enable `quantized_max_pool3d` kernel to fix the issue pytorch#101386.

**Test Plan**
```
clear && python -u -m pytest -s -v test_quantized_op.py -k test_max_pool3d
clear && python -u -m pytest -s -v test_quantized_op.py -k test_max_pool3d_nhwc
```


cc jgong5 mingfeima XiaobingSuper sanchitintel ashokei jingxu10

[ghstack-poisoned]

skip dynamic shape test on cpu

Update on "[PyTorch] Add tests for empty tensors w/storage null data_…

…ptr"

Further investigation seems to show that changing this behavior (making empty tensors sometimes have non-null data_ptr) was the real problem with pytorch#98090 . Adding tests to lock down this behavior so we don't change it by accident again.

Differential Revision: [D45873002](https://our.internmc.facebook.com/intern/diff/D45873002/)

[ghstack-poisoned]