[Pkg-xmpp-devel] gsasl_1.4.0-2_i386.changes REJECTED

cascardo at minaslivre.org cascardo at minaslivre.org
Thu Feb 4 16:20:35 UTC 2010


On Thu, Feb 04, 2010 at 02:04:29PM -0200, cascardo at minaslivre.org wrote:
> I completely agree with you in this. I would have waited for any other
> ABI break to use the opportunity to start using symbol versioning. I
> don't find any other advantage to use symbol versioning besides avoiding
> ABI breaks and soversion bumps. I'm going to read Ulrich Drepper's
> articles again to try to find out any other advantage of using symbol
> versioning and any reason to break ABI when starting using symbol
> versioning.

Quoting from [0]:
"Depending on whether the DSO an object is linked against had symbols or
not, the reference to the DSO requires symbols or not."

But this would be interpreted as: our binary has not been linked to a
DSO with symbols. So its reference do not require symbols. However, it
does not say that it must NOT have symbols. And I think they mean
"symbol versions" by symbols.

The following paragraph may shed some light into how we can make this
work without breaking the ABI. I'm just not sure that we are still in
time to do this for gsasl. But for the other libraries, that's sure we
could try to start using symbol versioning without breaking anything. I
still haven't read it with attention and did not reach any conclusions.
I'm going go make some experiments with real code before concluding
anything anyway.

"In case only the object file with the reference does not use
versioning but the object with the definition does, then the reference
only matches the base definition.  The base definition is the one with
index numbers 1 and 2 (1 is the unspecified name, 2 is the name given
later to the baseline of symbols once the library started using symbol
versioning).  The static linker is guaranteed to use this indeces for
the base definition.  If there is no symbol definition with such an
version index and there is exactly one version for which this symbol
is defined, then this version is accepted (this was mostly implemented
for dlopen() calls as it will normally not happen when the static
linker is used).  Otherwise, if more then one version of the symbol is
available, none of the definitions is accepted and the search
continues with the next object."

[0] http://people.redhat.com/drepper/symbol-versioning
-------------- next part --------------
A non-text attachment was scrubbed...
Name: not available
Type: application/pgp-signature
Size: 198 bytes
Desc: Digital signature
URL: <http://lists.alioth.debian.org/pipermail/pkg-xmpp-devel/attachments/20100204/802c5db6/attachment.pgp>


More information about the Pkg-xmpp-devel mailing list