Re: Fill more delay slots in conditional returns

[Steven Bosscher <stevenb dot gcc at gmail dot com>]

On Tue, Apr 2, 2013 at 6:56 PM, Steven Bosscher wrote:
> The SPARC back
> end also calls dbr_schedule() in its machine_reorg pass to work around
> errata in Atmel AT697F chips (LEON2-like, i.e. fairly new, see
> r179921).

Thinking about this some more: This could be fixed by inserting a
machine-specific pass just after delayed-branch scheduling, like in
the attached patch. I think the same is possible with the dbr_schedule
call in the MIPS backend.

Eric, what do you think of this approach?

With those two dbr_schedule calls out of the way, it will be a lot
easier to change things such that pass_free_cfg can run after
pass_machine_reorg (and after pass_cleanup_barriers that can be
simplified if there's still a CFG around). It will also help make the
DELAY_SLOTS hack in cfgrtl.c:rest_of_pass_free_cfg redundant.

I'm unsure how to test this patch. It's going through a bootstrap&test
cycle on sparc64-unknown-linux-gnu at the moment (on gcc62) but there
are no test cases for the errata that are being worked around...

Ciao!
Steven

Attachment: sparc_errata_as_machine_pass.diff.txt
Description: Text document