PATCH to remove unnecessary NOTE_INSN_DELETED
Jeffrey A Law
Tue Nov 30 23:59:00 GMT 1999
In message < 9911041921.AA25366@vlsi1.ultra.nyu.edu >you write:
> > Just FYI, I started to look into this. I've hit one stumbling
> > block so far (expressions of the form uid_luid[REGNO_FIRST_UID (xxx
> > but I guess that can be worked around somehow.
> Instead of recording the UID record the actual insn perhaps.
> I think that has a problem when insns are moved since the ideas is that
> the information would record a position independent of what actual
> insn was at that position.
Well, from the insn you still get to the luid.
The point being if luids where good enough for this use beforehand, then all
we have to do it assign them to insns in the same way as they showed up in
the luid table. If we can do that, then the existing code works as-is.
We weren't updating luids in the table when we moved insns around in cse
to the best of my knowledge, so movement of an insn shouldn't be an issue.
More information about the Gcc-patches