Re: [PATCH] change GLIBC PPC64/ELF2 ABI default to 2.17

[Michael Matz <matz at suse dot de>]

Hi,

[Leave me CCed please]

First, let me say "Arghh!" to all this.

On Wed, 29 Jan 2014, Steven Munroe wrote:

> Each distribution decides how it will support a new platform and we have
> requests to allow for back-ports of PPC64LE ELF2 to a GLIBC-2.17. 

Meee :-/

> We are not asking to apply this patch retroactively to the GLIBC-2.17
> source. But we like to get this patch accepted upstream to serve notice
> to all involved (in PPC64LE) that GLIBC-2.17 is the oldest symbol set,
> and promise this it final word on the topic.

I want some more promises.  In particular, even though the symbol version 
is 2.17 the ABI, the symbol list, about everything is defined by the 2.19 
release.  If there are discrepancies then 2.19 is right.  It's right even 
if it was a ABI-affecting bug and the backports fixed that bug.  It will 
stay right forever.

> We also ask that all distributions apply this patch and ELF2 ABI patches
> if they planing to support the PPC64LE platform and doing a back port of
> the to GLIBC-2.17 or 2.18.

While we technically didn't release any product with ppc64le there are 
quite a number of things available for openSUSE in our buildservice since 
roughly two months.  And a simple mass rebuild is not enough, as others 
here claimed.  The existing binaries won't run anymore as soon as glibc 
got rebuild (i.e. early) and hence the rebuild itself will fail.  We'll 
first have to rebuild with a glibc that provides both, 2.18 and 2.17 
symbols (with 2.17 being the default symversion), and then remove the 2.18 
set and rebuild again.  And we'll have to do that not only for the distro 
itself, but for all (public) repositories based on the ppc64le.

That's quite a hassle being dumped on us :-/  I'm sort of willing to go 
through that _if_ the backrev is some really large advantage to someone.  
Is it?


Ciao,
Michael.