On Feb 24, 2015, at 5:44 PM, Paul Hargrove <phhargr...@lbl.gov> wrote:
>
> FIRST:
> I believe that *something* should have occurred when no dl component could be
> built.
> Either the build should have been aborted or it could/should have switched to
> building everything static.
> However, the failure at runtime should not have been the eventual outcome.
Yes, it looks like I missed this case.
This is a good question: what should we do here?
1. Abort the configure (e.g., insist that the user install libltdl or
--disable-dlopen)
2. Fall back to a --disable-dlopen build
--> I looked into #2; at first blush, it looks kinda hard to do. :-\ I.e., by
the time we figure out that neither dl component will build, all the "whether
dl functionality will be available or not" decisions have been made (and are
difficult to un-make). It would require some re-structuring -- e.g., deferring
the "whether dl functionality will be available or not" decisions.
> SECOND:
> On {Free,Net,Open}BSD dlopen() appears in libc, not in libdl.
> So, I suspect one *should* be able to compile dl:dlopen on all these systems
> with the proper configure tests.
Cool; I'll fix this. ...done.
--
Jeff Squyres
jsquy...@cisco.com
For corporate legal information go to:
http://www.cisco.com/web/about/doing_business/legal/cri/
