Skip to content

Fix CUDA configuration

Karl Wette requested to merge ANU-CGA/lalsuite:cuda-rpath-fix into master

Description

As reported in #560 (closed) the CUDA configuration code is being called before the C compiler and libtool has been configured, so the -R option meant for libtool is being passed to the C compiler directly.

This MR splits the CUDA configuration macro in gnuscripts/lalsuite_build.m4 into 2 macros:

  • LALSUITE_WITH_CUDA() is called early to define ./configure command line arguments
  • LALSUITE_USE_CUDA() is called after C compiler/libtool has been configured

A few other changes in .gitlab-ci.yml:

  • In the cuda job, check whether ./configure --with-cuda=... works with a non-system path (i.e. one which doesn't begin with /usr), as it's relatively common for the CUDA toolkit to be installed in a non-system location
  • For top-level make dist-style jobs, upload config.log and Makefile from the package-level directories (as well as the top level) on failure, to aid debugging

API Changes and Justification

Backwards Compatible Changes

  • This change does not modify any class/function/struct/type definitions in a public C header file or any Python class/function definitions
  • This change adds new classes/functions/structs/types to a public C header file or Python module

Backwards Incompatible Changes

  • This change modifies an existing class/function/struct/type definition in a public C header file or Python module
  • This change removes an existing class/function/struct/type from a public C header file or Python module

Review Status

@rodrigo.tenorio Can you please check out this MR and confirm that this fixes #560 (closed) ?

Edited by Karl Wette

Merge request reports