[PATCH] Do not set BB_SUPERBLOCK in loop-unroll.c:split_edge_and_insert

Steven Bosscher stevenb.gcc@gmail.com
Mon Mar 5 07:15:00 GMT 2007


On 3/5/07, Jan Hubicka <hubicka@ucw.cz> wrote:
> There is no information given to the passes what kind of control flow is
> constructed by splitting/expanding, so it still seems to me that
> find_many_sub_basic_blocks is needed unless we wnat to reorganize all
> machine dependent expansion to in place adjust CFG.

I agree it's needed in some cases.  I mentioned expand and insn
splitting. But for "normal" passes, we shouldn't have to do this. Even
passes that call the expanders in a less uncontroled way than expand
should be able to fix the CFG in place.

I fully agree something like find_many_sub_basic_blocks is a sensible
thing to have to handle cases like cfgexpand.

Gr.
Steven



More information about the Gcc-patches mailing list