Darren J Moffat writes:
> Dale Ghent wrote:
> > Recently, over in OpenAFS land, a issue arose with the OpenAFS code  
> > failing to build on snv due to a side effect of the August-ish  
> > introduction of NFS RDMA as a part of PSARC 2007/347.
> > 
> > In OpenAFS code, we have a struct we use called "conn". OpenAFS sucks  
> > in RPC and NFS-related header files as part of its AFS->NFS translator  
> > functionality. With the introduction of NFS RDMA, its attendant header  
> > file, rpc_rdma.h, introduces a struct of the same name[1]. Hence the  
> > build failure with the expected redefinition errors.
> > 
> > I'm not sure what the general policy is regarding namespace of such  
> > things, and it could be that the NFS RDMA code is compliant in such  
> > regard... but I can't help but to think that giving that a name of  
> > "conn" is a bit too generic? Would "rdma_conn" be more appropriate?
> 
> Ignoring any standards issues I'd personally say that both the OpenAFS 
> and NFS RDMA codebases are partly to blame.  "conn" is just too generic 
> a name for a type/variable.  Both codebases IMO should have used an 
> appropriate prefix on this eg:  afs_conn, rdma_conn.  I'm a little 
> surprised that the RDMA code didn't have a prefix on that given that 
> other types declared in that same header do.

Indeed.  But as the boundary for the namespace in question is the ON
consolidation (those aren't public interfaces that AFS is using), I
think it's possible to draw a slightly sharper distinction.

Filing an RFE asking for NFS RDMA to stop camping on that too-generic
symbol sounds like a fine thing to me, but it's hard to say that it's
completely wrong.

-- 
James Carlson, Solaris Networking              <[email protected]>
Sun Microsystems / 35 Network Drive        71.232W   Vox +1 781 442 2084
MS UBUR02-212 / Burlington MA 01803-2757   42.496N   Fax +1 781 442 1677
_______________________________________________
opensolaris-code mailing list
[email protected]
http://mail.opensolaris.org/mailman/listinfo/opensolaris-code

Reply via email to