On Tue, Apr 3, 2012 at 6:56 PM, Joseph S. Myers joseph@codesourcery.com wrote:
(e) Existing practice for cases that do use different dynamic linkers is to use a separate library directory, not just dynamic linker name, as in lib32 and lib64 for MIPS or libx32 for x32; it's certainly a lot easier to make two sets of libraries work in parallel if you have separate library directories like that. So it would seem more appropriate to define a directory libhf for ARM (meaning you need a binutils patch as well to handle that directory, I think), and these different Debian-style names could be implemented separately in a multiarch patch if someone submits one that properly accounts for my review comments on previous patch versions (failure to produce such a fixed patch being why Debian multiarch directory support has not got into GCC so far).
The thread doesn't seem to be wrapping up, instead it appears to go in circles :-)
As a glibc maintainer, when it comes to this issue, I am guided by:
(1) This is a distribution problem and the solution needs to come from a consensus among the distributions.
(2) The gcc/glibc community should listen to the distributions.
The distributions have the most experience when it comes to whole-system issues. I certainly don't have that experience. Unfortunately *I* give the distributions a B- or C+ for communication. Please make it easy for me to give you an A. It is exceedingly difficult for me to review solutions that span multiple patches, emails, mailing lists, and communities. The best way to avoid rehashing old problems is to document the design and get sign off from the interested parties.
If I see uncoordinated and conflicting responses from the distributions... I get worried.
Is there a proposal on a wiki that begins to summarize the suggestions and solution?
Cheers, Carlos.