prov/cxi: Configury fixes

[raffenet]

These patches were used to successfully build the cxi provider on the Aurora system at Argonne with non-standard curl and json-c library installs.

[thomasgillis]

this PR solves my issue on LUMI #9789, thanks for the help @raffenet

[hppritcha]

@thomasgillis @raffenet
I applied these patches to CXI provider on our LANL system but when it comes to linking I''m seeing this:

In file included from prov/cxi/src/cxip_if.c:19:
./prov/cxi/include/cxip.h: In function 'cxip_cmdq_empty':
./prov/cxi/include/cxip.h:2799:16: warning: implicit declaration of function 'cxi_cq_empty'; did you mean 'cxi_eq_empty'? [-Wimplicit-function-declaration]
 2799 |         return cxi_cq_empty(cmdq->dev_cmdq);
      |                ^~~~~~~~~~~~
      |                cxi_eq_empty
  CC       prov/cxi/src/src_libfabric_la-cxip_fabric.lo
In file included from prov/cxi/src/cxip_fabric.c:20:
./prov/cxi/include/cxip.h: In function 'cxip_cmdq_empty':
./prov/cxi/include/cxip.h:2799:16: warning: implicit declaration of function 'cxi_cq_empty'; did you mean 'cxi_eq_empty'? [-Wimplicit-function-declaration]
 2799 |         return cxi_cq_empty(cmdq->dev_cmdq);
      |                ^~~~~~~~~~~~
      |                cxi_eq_empty
  CC       prov/cxi/src/src_libfabric_la-cxip_repsum.lo
In file included from prov/cxi/src/cxip_repsum.c:18:
./prov/cxi/include/cxip.h: In function 'cxip_cmdq_empty':
./prov/cxi/include/cxip.h:2799:16: warning: implicit declaration of function 'cxi_cq_empty'; did you mean 'cxi_eq_empty'? [-Wimplicit-fu

and then when it comes to linking that symbol cxi_cq_empty is undefined.

we are currently running sshot2.0.2.

Should I use git@github.com:thomasgillis/libfabric.git. instead?

[thomasgillis]

@hppritcha Yes, it seems that you have the same issue as I had on LUMI :-)
There is a branch on my fork that should work, see details here: #9835

[hppritcha]

hmmm... thanks @thomasgillis rather disturbingly if you configure with --enable-debug additional hacking is required on the cxi provider code.

[raffenet]

FWIW, I've reached out to folks at ALCF to see if there's anything that can be done to support, at minimum, build testing of cxi on the Polaris machine here at Argonne. Ideally, once cxi is able to build on a production system, CI could prevent further breaking changes from going in. @jswaro is that something that would be of interest?