This is the mail archive of the
gcc-bugs@gcc.gnu.org
mailing list for the GCC project.
[Bug target/32325] [4.3 Regression] cc1plus ICE configuring libstdc++ on Tru64 UNIX V5.1B: SEGV in rtl_verify_flow_info
- From: "ebotcazou at gcc dot gnu dot org" <gcc-bugzilla at gcc dot gnu dot org>
- To: gcc-bugs at gcc dot gnu dot org
- Date: 20 Sep 2007 20:45:57 -0000
- Subject: [Bug target/32325] [4.3 Regression] cc1plus ICE configuring libstdc++ on Tru64 UNIX V5.1B: SEGV in rtl_verify_flow_info
- References: <bug-32325-81@http.gcc.gnu.org/bugzilla/>
- Reply-to: gcc-bugzilla at gcc dot gnu dot org
------- Comment #2 from ebotcazou at gcc dot gnu dot org 2007-09-20 20:45 -------
> A reghunt revealed that this was caused by the following patch:
>
> 2007-03-02 Eric Botcazou <ebotcazou@adacore.com>
>
> * config/alpha/alpha.c (alpha_gp_save_rtx): Insert the insns at the
> entry by means of emit_insn_at_entry.
Right. Interestingly enough, this is purely a checking failure, i.e. the IL
is correct at the end of the pass. What happens is that in
if (USING_SJLJ_EXCEPTIONS)
sjlj_build_landing_pads ();
else
dw2_build_landing_pads ();
cfun->eh->built_landing_pads = 1;
/* We've totally changed the CFG. Start over. */
find_exception_handler_labels ();
break_superblocks ();
if (USING_SJLJ_EXCEPTIONS)
commit_edge_insertions ();
dw2_build_landing_pads will trigger a CFG verification, which of course is
not the right thing to do given the comment in the code.
The solution is the original patch I posted:
http://gcc.gnu.org/ml/gcc-patches/2006-11/msg01448.html
which was OKed by RTH:
http://gcc.gnu.org/ml/gcc-patches/2007-02/msg01162.html
so I'm going to apply it.
--
http://gcc.gnu.org/bugzilla/show_bug.cgi?id=32325