This is the mail archive of the
gcc-patches@gcc.gnu.org
mailing list for the GCC project.
Re: SH: PR target/22553
- From: Kaz Kojima <kkojima at rr dot iij4u dot or dot jp>
- To: joern dot rennecke at st dot com
- Cc: gcc-patches at gcc dot gnu dot org, aoliva at redhat dot com
- Date: Fri, 14 Oct 2005 23:32:45 +0900 (JST)
- Subject: Re: SH: PR target/22553
- References: <43392D60.4090604@st.com><20051014.085321.77058049.kkojima@rr.iij4u.or.jp><434FA716.10709@st.com>
Joern RENNECKE <joern.rennecke@st.com> wrote:
> I think that the fundamental bug is the way that exception handling
> places basic block
> boundaries after call_insns, without changing all the code that relies
> on basic blocks
> only being ended by jumps, labels, or the end of a function.
[snip]
Thanks for your explanation. Although I feel that I can understand
the issue more clearly, I'm afraid that the right solution for this
problem is a fairly big hammer and would be not appropriate in deep
stage3.
One temporary workround for 4.1.0 would be to disable the first schedule
insn optimization when flag_exceptions is set, with the appropriate
comment. How do you think about it?
Regards,
kaz