https://gcc.gnu.org/bugzilla/show_bug.cgi?id=84342
Jeffrey A. Law <law at redhat dot com> changed:
What |Removed |Added
----------------------------------------------------------------------------
Status|ASSIGNED |RESOLVED
CC| |law at redhat dot com
Resolution|--- |FIXED
--- Comment #5 from Jeffrey A. Law <law at redhat dot com> ---
I've added a marker for this BZ to Alex's change that fixed things.
Tamar -- the thing to look for are cases where a pattern emits no code, but
claims a nonzero size.
The ARM is inherently going to have problems with this because of how it uses
the ccfsm state to eliminate branches. I don't really see a good way to fix
this because at the time when we set the length information we don't have the
ccfsm state (which is only available during assembly output).
