Re: [FYI] Building the whole Debian archive with GCC 4.1: a summary

[Martin Michlmayr <tbm at cyrius dot com>]

* Mark Mitchell <mark@codesourcery.com> [2006-03-29 14:54]:
> > GCC 4.1 itself appears to be very stable, both on MIPS and AMD64.
> 
> Thank you for doing this, and for reporting the results, and for filing
> the bugs!

I've done the same with a snapshot of GCC 4.2 now on AMD64 (x86_64),
and I'm currently building on powerpc, sparc, i386 and mipsel.  So far
I have found the following regressions (some of the following PRs
turned out to be duplicates):

 - ICE in tree check: expected tree_list
   http://gcc.gnu.org/PR27094

 - Segfault with inlining producing negative shift count
   http://gcc.gnu.org/PR27088

 - ICE in add_virtual_operand, at tree-ssa-operands.c
   http://gcc.gnu.org/PR27085

 - ICE in tree check: expected ssa_name, have symbol_memory_tag in
   is_old_name, at tree-into-ssa.c:466
   http://gcc.gnu.org/PR27103

 - Incorrect integer division (wrong sign)
   http://gcc.gnu.org/PR27116

 - Segfault with -O2 on x86_64 when compiling hdf5_1.6.5-2
   http://gcc.gnu.org/PR27144

 - ICE in check_reg_live, at haifa-sched.c:4665
   http://gcc.gnu.org/PR26725

 - ICE: verify_ssa failed
   http://gcc.gnu.org/PR27093

 - ICE: verify_cgraph_node failed
   http://bugs.debian.org/361602

fixed in the meantime:

 - x86_x64 PIC problems and hidden vs anonymous namespaces
   http://bugs.debian.org/361445

 - gcc segfaults compiling pari_2.1.7-2
   http://bugs.debian.org/361715

 - ICE in merge_alias_info, at tree-ssa-copy.c
   http://gcc.gnu.org/PR27087

-- 
Martin Michlmayr
http://www.cyrius.com/