------- Comment #8 from dominiq at lps dot ens dot fr 2010-06-09 18:39 ------- > > The following http://gcc.gnu.org/ml/gcc-patches/2010-06/msg00692.html > > It definitly avoids the ICE, but it would be nice to know if libstdc++ > > testsuite passes. > > It does fix the bootstrap failure. I am currently regtesting (at gfortran > -m32) > and should have the results for libstdc++ in a couple of hours.
On x86_64-apple-darwin10.3.0 I have bootstrapped successfully revision 160431 with revision 160380 reverted and revision 160454 with revision 160380 and the patch in http://gcc.gnu.org/ml/gcc-patches/2010-06/msg00692.html. For the later the regtest failures are those reported in http://gcc.gnu.org/ml/gcc-testresults/2010-06/msg00587.html . Note that the 2*16 failures for libstdc++ are due to revision 159362 (see also the regress logs for ppc). Martin Jambor knows about them and I'll fill a PR for yesterday later tonight or tomorrow (filling PR against darwin sucks!-). Obviously (as a duplicate) the patch fixes also pr44453. -- http://gcc.gnu.org/bugzilla/show_bug.cgi?id=44460