Re: Versioned Symbols
Previously Stephen Frost wrote:
> It's been done already in some places from my understanding
> (libdb, libpng). The fix being hard to implement doesn't somehow make
> the problem go away.
libdb seems to have some evil hackery which puts a version string
in symbol names, it doesn't use ELF versioned symbols. I don't
see any evidence of version tricks in libpng3 sources.
> I agree, which is why I feel the right thing to do is get versioned
> symbols in our libraries.
Versioned symbols is a lot of work and creates a maintenance headache
for whoever maintains it. If you want to push those patches upstream
it means that upstream authors suddenly will have a lot of extra work to
fix a problem they are not responsible for at all.
> It's been done. I don't doubt that it's difficult.
Please do it yourself to see what it really involved.
> I don't understand how this is supposted to work. The ssh-krb5
> maintainer builds against libssl0.9.6 and uploads. The openldap
> maintainer builds ldap against 0.9.7 and uploads.
Right, that is something you can only test after the uploads since
it will be fine on each developers machines. It could be an extra
check before something hits testing perhaps.
> Hopefully this problem will be something we watch out for during a
> release at least so we can avoid it there, and maybe even develop
> scripts and whatnot to keep the problem from happening in testing.
It is pretty easy to script a test for this one could run on an
archive before a release.
Wichert.
--
Wichert Akkerman <wichert@wiggy.net> http://www.wiggy.net/
A random hacker
Reply to: