running 4.4 libstdc++ testsuite using 4.5's libstdc++ shared library

[Mark Mitchell]

Jonathan Wakely wrote:

>> Ps: the condition_variable PR likely is also a bit pointless, we are not
>> guaranteeing full runtime binary compatibility for all the c++0x exp
>> features, sorry.

> If something must be done to the 4.4 testsuite then the test should be
> removed, because it was pretty useless to test an empty
> implementation. I'd rather do nothing though.

In general, it's fine to backport testsuite fixes.  But, it's not
*necessary* to do so.  So, Matthias, you (or we, with our Linaro hats
on), could certainly backport some testsuite fixes to the 4.4 branch if
that's helpful.

But, I think the high-order bit is Jonathan's comment above: libstdc++
is not maintaining backwards-compatibility for its complete interface.
Jonathan, Paolo, I feel that if that is the case then the SONAME for
libstdc++ should have changed beetween 4.4 and 4.5.  And, if there will
be more incompatibilities in 4.5, then we should change the SONAME
again.  Compatibility is binary: either you have it, or you don't.

Matthias is focused on the Ubuntu distribution model where they will
sometimes upgrade the toolchain without rebuilding all the packages.
But, I think it's at least as important to think about third-party
software packages.  If the package fails to run because its missing the
right libstdc++, the problem is obvious to the user; if it fails to work
at run-time because of an incompatibility, then it's just "this package
is broken".

Meanwhile, Matthias, I think you should consider the impact of the fact
that the libstdc++ maintainers are not yet in a position to guarantee
long-term binary compatibility on the dual-toolchain strategy used in
Ubuntu.

Thanks,

-- 
Mark Mitchell
CodeSourcery
mark@codesourcery.com
(650) 331-3385 x713