Re: Fwd: cselib_record_set breaks due to auto_inc_dec

[Kenneth Zadeck <zadeck at naturalbridge dot com>]

you need to talk to someone with a higher rtl rating than myself. sorry.
On 12/20/2012 08:41 AM, Paulo Matos wrote:
> > -----Original Message-----
> > From: gcc-owner@gcc.gnu.org [mailto:gcc-owner@gcc.gnu.org] On Behalf Of
> > Kenneth Zadeck
> > Sent: 20 December 2012 12:46
> > To: David Edelsohn; GCC Development
> > Subject: Re: Fwd: cselib_record_set breaks due to auto_inc_dec
> >
> > if i had to guess what was going on here, and since i do not have the
> > example or even know the platform, i am just guessing, i would say that
> > at some earlier pass, the sub expression in the regnote and the second
> > set in the parallel were the same expression and that someone changed
> > the second set but did not delete the note or modify it.
> >
> > and once they became different, it was just a case of "dead man walking".
>
> Hi Kenneth,
>
> Thanks for the comment. However, this happens in dse2 pass and the previous pass is pro_and_epilogue when this exact insn is added.
> We are the ones that add the expression to the expr list of the insn when we add the prologue insn for debug purposes but we don't expect cselib through dse2 to assume the insn in the expr list to be an extra set.
>
> Is it not allowed to add expressions onto an insn expr list?