[TF][2.17] Build with cuda

[smuzaffar]

No description provided.

[cmsbuild]

A new Pull Request was created by @smuzaffar for branch IB/CMSSW_15_1_X/tf.

@cmsbuild, @iarspider, @smuzaffar can you please review it and eventually sign? Thanks.
@antoniovilela, @mandrenguyen, @rappoccio, @sextonkennedy you are the release manager for this.
cms-bot commands are listed here

[cmsbuild]

cms-bot internal usage

[smuzaffar]

test parameters:

• full_cmssw = true
• enable = gpu

[smuzaffar]

please test

[cmsbuild]

-1

Summary: https://cmssdt.cern.ch/SDT/jenkins-artifacts/pull-request-integration/PR-10994e/45176/summary.html
COMMIT: ca52a23
CMSSW: CMSSW_15_1_TF_X_2025-03-18-1100/el8_amd64_gcc12
Additional Tests: ROCM,CUDA
User test area: For local testing, you can use /cvmfs/cms-ci.cern.ch/week1/cms-sw/cmsdist/9758/45176/install.sh to create a dev area with all the needed externals and cmssw changes.

External Build

I found compilation error when building:

Requested to quit.
Requested to quit.
Requested to quit.
* The action "build-external+tensorflow-sources+2.17.0-2c3ffa93f7d19a98bde49aabef31572d" was not completed successfully because Failed to build tensorflow-sources. Log file in /data/cmsbld/jenkins/workspace/ib-run-pr-tests/testBuildDir/BUILD/el8_amd64_gcc12/external/tensorflow-sources/2.17.0-2c3ffa93f7d19a98bde49aabef31572d/log. Final lines of the log file:

external/local_xla/xla/service/gpu/gpu_prim.h(48): error: no instance of overloaded function "cub::ThreadLoadVolatilePointer" matches the specified type
__attribute__((device)) __inline__ __attribute__((always_inline)) Eigen::half ThreadLoadVolatilePointer(
^

external/local_xla/xla/service/gpu/gpu_prim.h(63): error: no instance of overloaded function "cub::ThreadLoadVolatilePointer" matches the specified type
ThreadLoadVolatilePointer(tsl::bfloat16 *ptr,