This is the mail archive of the
gcc-patches@gcc.gnu.org
mailing list for the GCC project.
Re: sh barrier_align vs. fill_eager_delay_slots
- To: cpopetz at cygnus dot com (Clinton Popetz)
- Subject: Re: sh barrier_align vs. fill_eager_delay_slots
- From: Joern Rennecke <amylaar at cygnus dot co dot uk>
- Date: Mon, 6 Mar 2000 21:03:12 GMT
- Cc: gcc-patches at gcc dot gnu dot org dot amylaar@pasanda.cygnus.co.uk
- Newsgroups: cygnus.egcs.patches
: --- 2770,2787 ----
: prev = prev_real_insn (prev))
: {
: + jump_to_next = 0;
: if (GET_CODE (PATTERN (prev)) == USE
: || GET_CODE (PATTERN (prev)) == CLOBBER)
: continue;
: if (GET_CODE (PATTERN (prev)) == SEQUENCE)
: ! {
: ! prev = XVECEXP (PATTERN (prev), 0, 1);
: ! if (INSN_UID (prev) == INSN_UID (next))
: ! {
: ! /* Delay slot was filled with insn at jump target. */
: ! jump_to_next = 1;
There should be a 'continue;' here to avoid a mismatch in the credit tally
before and after reorg. Otherwise it looks OK.
: ! }
: ! }
: !
: if (slot &&
: get_attr_in_delay_slot (prev) == IN_DELAY_SLOT_YES)