Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Build Errors with CL/DOCA_UROM #1002

Open
wfaderhold21 opened this issue Jul 18, 2024 · 2 comments
Open

Build Errors with CL/DOCA_UROM #1002

wfaderhold21 opened this issue Jul 18, 2024 · 2 comments

Comments

@wfaderhold21
Copy link
Collaborator

There are two issues I'm currently having with respect to CL DOCA UROM, these are simple enough to manually fix but just documenting the errors here for now:

  1. The m4 configuration file (config/m4/doca_urom.m4) when configuring with --with-doca_urom=/opt/mellanox/doca looks for the doca library files in /opt/mellanox/doca/lib. However, DOCA library files are installed in /opt/mellanox/doca/lib/x86_64-linux-gnu/ for an x86_64 system or /opt/mellanox/doca/lib/aarch64-linux-gnu/ for an ARM system. These install directories are likely specific to my machine and could be different with various users. May be beneficial to provide a --with-doca_urom-lib option on configure to allow for a user to provide a specific library path if needed.
  2. If using a strict compiler, the following error will occur:
make[3]: Entering directory '/workspace/ucc/src/components/cl/doca_urom'
  CC       libucc_cl_doca_urom_la-cl_doca_urom_common.lo
  CC       libucc_cl_doca_urom_la-cl_doca_urom.lo
  CC       libucc_cl_doca_urom_la-cl_doca_urom_lib.lo
  CC       libucc_cl_doca_urom_la-cl_doca_urom_context.lo
  CC       libucc_cl_doca_urom_la-cl_doca_urom_worker_ucc.lo
  CC       libucc_cl_doca_urom_la-cl_doca_urom_team.lo
cl_doca_urom_context.c: In function 'ucc_status_to_doca_error':
cl_doca_urom_context.c:15:5: error: enumeration value 'UCC_ERR_MEM_MAP_FAILURE' not handled in switch [-Werror=switch]
   15 |     switch (status) {
      |     ^~~~~~
cc1: all warnings being treated as errors
make[3]: *** [Makefile:633: libucc_cl_doca_urom_la-cl_doca_urom_context.lo] Error 1
@wfaderhold21
Copy link
Collaborator Author

wfaderhold21 commented Aug 6, 2024

In addition, while executing and calling ucc_finalize(), the following error occurs:

/workspace/osu/mpi/collective/osu_allreduce: symbol lookup error: /opt/hpcx/ucc/lib/ucc/libucc_cl_doca_urom.so: undefined symbol: ucc_cl_doca_urom_task_lib_destroy

Possibly forgot to include the implementation of this function?

@nsarka did you see the same error when you were testing?

@nsarka
Copy link
Collaborator

nsarka commented Aug 7, 2024

Let me check on this

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Labels
None yet
Projects
None yet
Development

No branches or pull requests

2 participants