ROMIO fails to build if level_zero is installed

[acgoldma]

Thank you for taking the time to submit an issue!

Background information

What version of Open MPI are you using? (e.g., v3.0.5, v4.0.2, git branch name and hash, etc.)

https://github.com/open-mpi/ompi/pull/10224

Describe how Open MPI was installed (e.g., from a source/distribution tarball, from a git clone, from an operating system distribution package, etc.)

I was attempting to build OMPI rpm

If you are building/installing from a git clone, please copy-n-paste the output from git submodule status.

https://github.com/open-mpi/ompi/pull/10224 openpmix @ 6692c28 prrte @ 7ae2c08

Please describe the system on which you are running

• Operating system/version: RHEL 8.4
• Computer hardware: Intel E5-2699 v3
• Network type: N/A

---

Details of the problem

Looks like romio341 detects Level-Zero on my system and causes issues.

  CC       src/gavl/mpl_gavl.lo
../../../../../3rd-party/romio341/mpl/src/gpu/mpl_gpu_ze.c: In function 'MPL_gpu_ipc_handle_create':
../../../../../3rd-party/romio341/mpl/src/gpu/mpl_gpu_ze.c:123:11: warning: implicit declaration of function 'zeDriverGetMemIpcHandle'; did you mean 'zeMemGetIpcHandle'
? [-Wimplicit-function-declaration]
     ret = zeDriverGetMemIpcHandle(global_ze_driver_handle, ptr, ipc_handle);
           ^~~~~~~~~~~~~~~~~~~~~~~
           zeMemGetIpcHandle
../../../../../3rd-party/romio341/mpl/src/gpu/mpl_gpu_ze.c: In function 'MPL_gpu_ipc_handle_map':
../../../../../3rd-party/romio341/mpl/src/gpu/mpl_gpu_ze.c:139:9: warning: implicit declaration of function 'zeDriverOpenMemIpcHandle'; did you mean 'zeMemOpenIpcHandle
'? [-Wimplicit-function-declaration]
         zeDriverOpenMemIpcHandle(global_ze_driver_handle,
         ^~~~~~~~~~~~~~~~~~~~~~~~
         zeMemOpenIpcHandle
../../../../../3rd-party/romio341/mpl/src/gpu/mpl_gpu_ze.c:140:69: error: 'MPL_gpu_ipc_mem_handle_t' {aka 'struct _ze_ipc_mem_handle_t'} has no member named 'global_dev
_id'
                                  global_ze_devices_handle[ipc_handle.global_dev_id],
                                                                     ^
../../../../../3rd-party/romio341/mpl/src/gpu/mpl_gpu_ze.c:141:44: error: 'MPL_gpu_ipc_mem_handle_t' {aka 'struct _ze_ipc_mem_handle_t'} has no member named 'handle'
                                  ipc_handle.handle, ZE_IPC_MEMORY_FLAG_NONE, ptr);
                                            ^
../../../../../3rd-party/romio341/mpl/src/gpu/mpl_gpu_ze.c:141:53: error: 'ZE_IPC_MEMORY_FLAG_NONE' undeclared (first use in this function); did you mean 'ZE_IPC_MEMORY
_FLAG_TBD'?
                                  ipc_handle.handle, ZE_IPC_MEMORY_FLAG_NONE, ptr);
                                                     ^~~~~~~~~~~~~~~~~~~~~~~
                                                     ZE_IPC_MEMORY_FLAG_TBD

Looks like it is not using the correct API?

[wenduwan]

@acgoldma Sorry for the 2-year wait. Does this problem still happen with latest romio and ompi 5.0.3?

[acgoldma]

did not see an issue recently.

[sjb017]

Hello, I work with Adam, I think we need to reopen this issue.

I see the same build error when trying to build from the v5.0.3 source RPM.

[acgoldma]

https://github.com/open-mpi/ompi/blob/42c744e00eba2da1f904d2b94f33d2769e744867/3rd-party/romio341/mpl/configure.ac#L999

There appears to be no easy way to disable this code from configure.

Also, the ROMIO version is using a pre-release spec (v0.95) for level-zero, so that will need to be updated. Many things changes in the release spec (v1.0+).

[juselius]

A quick fix is to run sed -i 's/have_ze=yes/have_ze=no/' /build/ompi/3rd-party/romio341/mpl/configure.ac to disable level-zero in romio completely.