This is the mail archive of the
gcc@gcc.gnu.org
mailing list for the GCC project.
Re: How early are COND_EXEC patterns created?
- To: meissner at cygnus dot com
- Subject: Re: How early are COND_EXEC patterns created?
- From: John Wehle <john at feith dot com>
- Date: Tue, 2 May 2000 00:55:56 -0400 (EDT)
- Cc: gcc at gcc dot gnu dot org
> Looking at toplev.c, there is a call to if_convert after the first jump pass
> (in my original code, the cond_exec code was called after each of the two flow
> passes).
However process_if_block checks reload_completed before calling
cond_exec_process_if_block. What makes things unclear (at least
to me) is that dead_or_predicable (used by find_if_case_1 and
find_if_case_2) calls cond_exec_process_insns. It looks like
find_if_case_[12] are only called before reload_completed if
! HAVE_conditional_execution. Does COND_EXEC patterns only
appear after reload_completed or is there some code path I'm
missing that allows them to be created earlier? What's the
intent of the design?
-- John
-------------------------------------------------------------------------
| Feith Systems | Voice: 1-215-646-8000 | Email: john@feith.com |
| John Wehle | Fax: 1-215-540-5495 | |
-------------------------------------------------------------------------