Recursively sync fbgemm submodules before build (pytorch#159477)

ROCm inductor benchmark builds failing fbgemm build stage https://ossci-raw-job-status.s3.amazonaws.com/log/46800456622
```
2025-07-27T08:00:32.3443858Z /var/lib/jenkins/pytorch/fbgemm/src/RowWiseSparseAdagradFused.cc:389:18: error: no matching function for call to ‘asmjit::v1_17::x86::Vec::Vec(uint32_t)’
2025-07-27T08:00:32.3444080Z   389 |         x86::Xmm partial_sum_xmm(partial_sum_vreg.id());
```

It looks like asmjit fails to build, this seems to be due to submodules of fbgemm not being updated after checking out to new commit.
Pull Request resolved: pytorch#159477
Approved by: https://github.com/pruthvistony, https://github.com/eqy

Author: jataylo

.ci/pytorch/common_utils.sh

@@ -245,7 +245,7 @@ function install_torchrec_and_fbgemm() {
     if [ "${found_whl}" == "0" ]; then
       git clone --recursive https://github.com/pytorch/fbgemm
       pushd fbgemm/fbgemm_gpu
-      git checkout "${fbgemm_commit}"
+      git checkout "${fbgemm_commit}" --recurse-submodules
       python setup.py bdist_wheel \
         --build-variant=rocm \
         -DHIP_ROOT_DIR="${ROCM_PATH}" \