Bootstrap failing gencodes miscompiled by stage1 compiler
Jeffrey A Law
law@cygnus.com
Sat Oct 31 20:56:00 GMT 1998
In message < 361BD2A4.552D@rcp.co.uk >you write:
> Hi
>
> I've done some investigations into this failure and
> have identified it as a problem with sched-insns-after_reload
> optimizations.
>
> The problem is a instruction is being reschuled before a
> instruction on which it has a dependency.
>
> You can see the incorrect/correct insn sequences if you compile
> with and without the -fno-schedule-insn2 respectively.
This should be fixed now.
jeff
More information about the Gcc-bugs
mailing list